From rws@inko.no Fri Aug 1 03:34:00 1997 From: rws@inko.no (Rolf Welde Skeie) Date: Fri, 01 Aug 1997 03:34:00 -0000 Subject: ICMP Message-ID: <33E1B7FA.32695A3B@scali.no> Hi! Is icmp development supported? (can't find ICMP_* nor struct icmp/icmphdr) -- Rolf Welde Skeie Senior Design Engineer =+= __ mailto:rws@scali.no_____http://www.scali.com___====+====__ Tel +47 6384 6700 Scali AS ====+==== Fax +47 6384 4005 Hvamstubben 17 ====+==== Home +44 181 946 0650 2013 Skjetten =+= Mob +47 9095 4406 NORWAY o==V==o - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From rws@inko.no Fri Aug 1 03:40:00 1997 From: rws@inko.no (Rolf Welde Skeie) Date: Fri, 01 Aug 1997 03:40:00 -0000 Subject: Where are Reg* implemented? References: <3.0.1.32.19970731172659.00692430@elwood.pionet.net> <33E1B654.90B5720@scali.no> Message-ID: <33E1BE77.BB7E7E01@scali.no> Rolf Welde Skeie wrote: > (BTW: shared libraries can be placed before the target(-o xxx)...) Wrong again. The SunOS5 linker can have libraries placed before the target, that's why I tried it per default under GnuWin32... Topic closed ;-) -- Rolf Welde Skeie Senior Design Engineer =+= __ mailto:rws@scali.no_____http://www.scali.com___====+====__ Tel +47 6384 6700 Scali AS ====+==== Fax +47 6384 4005 Hvamstubben 17 ====+==== Home +44 181 946 0650 2013 Skjetten =+= Mob +47 9095 4406 NORWAY o==V==o - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From giac@dalim.de Fri Aug 1 04:29:00 1997 From: giac@dalim.de (Philippe GIACINTI) Date: Fri, 01 Aug 1997 04:29:00 -0000 Subject: helpful hints to build dlls in Win32 using cygwin (gnu) References: <33E0F327.DF8B81B3@Eng.Sun.COM> <199708010456.OAA10715@mundook.cs.mu.OZ.AU> Message-ID: <9708011319.ZM28568@pion.kehl.dalim.de> Fergus Henderson writes: >Padmakar Vishnubhatt writes: > >>I did exactly as u recommended in your hints - and yet I get the >>following error when I run the linker (ld). Am I missing something in >>the cygwin distribution? >> >>/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libccrt0.o)(.text+0x >>8a):libccrt0.cc: undefined reference to `GetModuleHandleA@4' > >Your problem is due to `_impure_ptr'. (By the way, to debug these >sorts of problems, use the `-Map' and `--cref' options to `ld'.) >libccrt0.o should not be linked in to DLLs at all. >The reason that it does get linked in is because it defines `_impure_ptr', >which is referenced by your code (implicitly, e.g. because stdout >is #defined as `_impure_ptr._stdout'), and which is >not defined anywhere else. > >You need to provide a definition of `_impure_ptr'. >You also need to initialize it (otherwise references to stdout >etc. in your DLL will be bogus). > The problem of definition and initialization of _impure_ptr is solved in my previous patch. If you use this patch you must link your dll with libccrt0.o and then you have to add -lkernel32 to your link line ... regards. -- ========================================================================== | Philippe GIACINTI | | | | DALiM GmbH R&D email: giac@dalim.de | | Daimler Strasse 2, tel: +49.7851.9196-28 | | D-77694 Kehl-Sundheim Germany fax: +49.7851.73576 | | | ========================================================================== - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Fri Aug 1 04:29:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Fri, 01 Aug 1997 04:29:00 -0000 Subject: New patch available Message-ID: <01BC9E7C.798C1C50@gater.krystalbank.msk.ru> Hi! - console code no longer alway translates CR to NL when tty support is enabled. The behavior depends now of termios settings. - signals should terminate blocking socket calls. - Chris Faylor's (cfg@bbc.com) patch for wait and SIGCHLD raising logic - cygwin processes now waits non-cygwin processes correctly, no more time races problems with SIGCHLD raising and process's exit. - utmp/wtmp files are filled now for every cygwin's session (both local session and telnet session). "Who" command works now. Create /var/run and /var/log directories to enable utmp logging. - syslog() call handles now %m macro. - signal raising keys (ctrl-C, ctrl-\ etc) restarts suspended tty output. New internet servers/clients added to remote.tar.gz -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pete@horus.cix.co.uk Fri Aug 1 07:14:00 1997 From: pete@horus.cix.co.uk (Pete Jordan) Date: Fri, 01 Aug 1997 07:14:00 -0000 Subject: Perl5.004_01 patch for b18 References: Message-ID: cgf@bbc.com (Chris Faylor) wrote: > Actually since, to the best of my knowledge, the patch that is pending > for perl5.004_02 is actually the patch on my web site, then I wouldn't > be too certain that things will work much better at all. Well, I didn't have serious problems building your patched code - the biggest hassle was with patch itself and that won't affect the 5.004_02 distribution. If the niggles I mentioned in my mails to you are fixed ($perlpath problem with the inconsistency in which directory config.sh is called from, "read answ" kludge around for Win95, path delimiter inconsistency in built GNUmakefile and MakeMaker generating filenames with "::" in for man files), all should be well AFAICS. Pete Jordan = = = = = = = = = = = = = = = = = = = = = = = Horus Communications http://www.horus.cix.co.uk/ = = = = = = = = = = = = = = = = = = = = = = = "'Not twisted,' Salzy once said of her own passion, 'it is helical. That sounds better.'" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jazz@softway.com Fri Aug 1 07:47:00 1997 From: jazz@softway.com (Jason Zions) Date: Fri, 01 Aug 1997 07:47:00 -0000 Subject: Where are Reg* implemented? References: <199708010304.XAA31283@solid.ssd.ornl.gov> Message-ID: <33E1F632.1B2D@softway.com> > My impression ( an what I always do ) is to put the options before the > files to be compiled or linked as following: > > g++ -o outfilename -O -L libpath -l lib file.c file.o ... Usually won't work. *Most* options need to go up front; -O, -L, etc. Library options (-l) *must* be specified in the *correct order* so that resolution can be controlled. By listing a library, -lfoo, before any .o files have been loaded, you're trying to resolve entrypoints from the library before the loader has seen any unresolved names. First list the .o files, then the libraries you want the loader to use to resolve entrypoints, in the order you want it to try. If two libraries both have a particular entrypoint, the first one the loader sees (once it sees the unresolved name from another file) is the one it loads, and it silently ignores the latter one (because the name has already been resolved). This is basic stuff, folks. Read a Unix system's "ld" man page. Jason - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From blroberts@xactware.com Fri Aug 1 08:12:00 1997 From: blroberts@xactware.com (Barry Roberts) Date: Fri, 01 Aug 1997 08:12:00 -0000 Subject: Creating new windows Message-ID: Under NT: cmd /c start bash Replace cmd with command under 95. >>> Ian Collins 07/31 11:59 AM >>> Does anyone know if it is possible to create a new window in which to run a bash shell from within a bash shell? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From huotte@rpi.edu Fri Aug 1 08:12:00 1997 From: huotte@rpi.edu (Ed Huott) Date: Fri, 01 Aug 1997 08:12:00 -0000 Subject: Creating new windows References: <01BC9E40.140579D0@modem2.kiwiplan2.co.nz> Message-ID: <199708011512.LAA71142@alumni.rpi.edu> ian@kiwiplan.co.nz wrote: > > Does anyone know if it is possible to create a new window in which to run a bash shell from within a bash shell? You can use the Windows 'start' command to fire up a copy of bash in a new window. At the bash (or DOS) prompt, just invoke: $ start bash If you have used Explorer (File Manager) on Win95 to create a .PIF file for starting bash with a particular set of options from an icon, you can invoke that from the command line as well: $ start bash.pif Enjoy. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From huotte@rpi.edu Fri Aug 1 08:12:00 1997 From: huotte@rpi.edu (Ed Huott) Date: Fri, 01 Aug 1997 08:12:00 -0000 Subject: Creating new windows References: <01BC9E40.140579D0@modem2.kiwiplan2.co.nz> Message-ID: <199708011512.LAA74460@alumni.rpi.edu> ian@kiwiplan.co.nz wrote: > > Does anyone know if it is possible to create a new window in which to run a bash shell from within a bash shell? You can use the Windows 'start' command to fire up a copy of bash in a new window. At the bash (or DOS) prompt, just invoke: $ start bash If you have used Explorer (File Manager) on Win95 to create a .PIF file for starting bash with a particular set of options from an icon, you can invoke that from the command line as well: $ start bash.pif Enjoy. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From wdoyle@bode.cdsp.neu.edu Fri Aug 1 08:15:00 1997 From: wdoyle@bode.cdsp.neu.edu (Patrick Doyle) Date: Fri, 01 Aug 1997 08:15:00 -0000 Subject: Problem with diff3 References: Message-ID: <199708011525.LAA00336@cdsp.neu.edu> I have (successfully) build rcs-5.7 using gnuwin32 b18. Configure failed due to a problem with diff3 To reproduce the error echo 0 > conftest0 echo 1 > conftest1 diff3 -E -m -L 0 -L 1 -L 2 conftest0 conftest1 /dev/null On ultrix 4.3 and Irix 6.2 I get <<<<<<< 0 0 ======= >>>>>>> 2 With win95 and gnuwin32 b18 I get diff3: /dev/null: No such file or directory The problem is that "diff3" attempts to perform a 'stat()' system call on each of the file names and 'stat()' fails (returns ENOENT) for "/dev/null". You can fix this at least four different ways: 1) Change offending line to read (something like): cat /dev/null | diff3 -E -m -L 0 -L 1 -L 2 conftest0 conftest1 - 2) Change the offending line to read (something like): diff3 -E -m -L 0 -L 1 -L 2 conftest0 conftest1 NUL 3) Change "diff3" to compare the filename to be opened with "/dev/null" and not call 'stat()' if they are equal. (It's a hack, but it solves the problem). 4) Change the implementation of 'stat()' in CYGWIN32.DLL to produce the correct (i.e. unix-like) result for files called "/dev/null". I believe that 'open()' already does this. -- --patrick - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From weiku@solid.ssd.ornl.gov Fri Aug 1 08:28:00 1997 From: weiku@solid.ssd.ornl.gov (Ku Wei) Date: Fri, 01 Aug 1997 08:28:00 -0000 Subject: Where are Reg* implemented? References: <33E1F632.1B2D@softway.com> Message-ID: Good point Jason about the ld man page. However, we are discuessing gcc or g++, not ld. It is clear that we need to put the main() entry ( in one of the .c or .cxx files) as the first of the source file list ( including .o and .c). However, the libraries are not 'source files' to g++. Instead, they are assigned as 'options' by '-l'. Therefore, it is really the compiler's responsibility to add these options to correct position in parameter list passed to 'ld' later. For example, for IBM AIX C compiler, you are only allowed to put parameters "before" the source file list: xlC [options] files On the other hand, g++ allows two different syntex: g++ [option|filenam] ... Anyway, why use same name in your own code as the name in libraries to confuse the linker ? Sincerely, Wei Ku *************************************** Department of Physics and Astronomy The University of Tennessee 1408 Circle Drive Knoxville, Tennessee 37996-1200 weiku@utkux.utcc.utk.edu --------------------------------------- Solid State Division Oak Ridge National Laboratory P.O.Box 2008 Oak Ridge, TN 37831-6032 Phone: (423) 574-5795 Fax: (423) 574-4143 weiku@solid.ssd.ornl.gov *************************************** On Fri, 1 Aug 1997, Jason Zions wrote: > > My impression ( an what I always do ) is to put the options before the > > files to be compiled or linked as following: > > > > g++ -o outfilename -O -L libpath -l lib file.c file.o ... > > Usually won't work. *Most* options need to go up front; -O, -L, etc. > > Library options (-l) *must* be specified in the *correct order* so that > resolution can be controlled. By listing a library, -lfoo, before any .o > files have been loaded, you're trying to resolve entrypoints from the > library before the loader has seen any unresolved names. > > First list the .o files, then the libraries you want the loader to use > to resolve entrypoints, in the order you want it to try. If two > libraries both have a particular entrypoint, the first one the loader > sees (once it sees the unresolved name from another file) is the one it > loads, and it silently ignores the latter one (because the name has > already been resolved). > > This is basic stuff, folks. Read a Unix system's "ld" man page. > > Jason > > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From topcoder@mindspring.com Fri Aug 1 10:20:00 1997 From: topcoder@mindspring.com (Derek Greene) Date: Fri, 01 Aug 1997 10:20:00 -0000 Subject: directx References: <3.0.3.32.19970731094435.00697934@mail.mindspring.com> <33e65e37.5155151@smtp.netzone.com> Message-ID: <3.0.3.32.19970731225124.00698270@mail.mindspring.com> umm can i get the mirror site for ming32? At 04:14 AM 8/1/97 GMT, you wrote: >There is a pointer on the cygnus home page to >the home page of a guy that did some work on it. > >also you may want to check out mingw32, and the patch for it >SDK_4_ming_cygwin32.tar.gz on > ftp://wwwdim.uqac.uquebec.ca/pub/gnu-win32 >at least I think that's the right address, I guess they're busy >tonight, can't connect. > >anyway the patch lets you use the platform SDK headers/libs from microsuck >with mingw32/cygwin32 I don't think I ported any of >the sample apps that use directX, but it shouldn't >be particularly difficult. and at least all of the calls will be available, > > >On Thu, 31 Jul 1997 09:44:35 -0400, you wrote: > >>has ANYONE managed to port directx to gnu-win32? puh-leez say yes, puh-leez >>say yes....:) >>- >>For help on using this list (especially unsubscribing), send a message to >>"gnu-win32-request@cygnus.com" with one line of text: "help". >> > >(*jeffdb@netzone.nospam.com) >Remove *&.nospam from the above to reply > Mikey > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Fri Aug 1 10:20:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Fri, 01 Aug 1997 10:20:00 -0000 Subject: ICMP Message-ID: <01BC9EC0.DB9E8600@sos> Rolf Welde Skeie wrote: > Hi! > > Is icmp development supported? > (can't find ICMP_* nor struct icmp/icmphdr) Traditional Unix ICMP is not supported by Windows sockets :-( There are some undocumented extensions (icmp.dll) to provide ICMP functionality. Search for ICMP_API.ZIP on the Net. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From cgf@bbc.com Fri Aug 1 10:20:00 1997 From: cgf@bbc.com (Chris Faylor) Date: Fri, 01 Aug 1997 10:20:00 -0000 Subject: .tar.gz References: <01BC9DC5.447A11A0.bwomack@ptc.com> Message-ID: I don't see how anyone could ever argue that gzip is "non-portable". In fact that isn't what this thread is about. Of course it is "portable"; it's got something called a "configure" script which goes to great pains to ensure that it will run on as many platforms as possible. The point that was being made is that 'gzip' is not distributed, by default, with many computer systems. Most UNIX distributions do have a 'compress' program, however. That doesn't mean that compress is more "portable" but it does mean that it is more widely available. The original poster was not saying that the compress code was more "portable". He was pointing out that if you expect to find 'gzip' on every UNIX platform you'll be disappointed. In article <01BC9DC5.447A11A0.bwomack@ptc.com>, Brian David Womack wrote: >Very good point. I've felt that this debate about gzip being >non-portable was bogus as well. Every UNIX and PC I've been >using has had a cross-platform compile of gzip. Never had >a problem. > >On Wednesday, July 30, 1997 4:33 PM, Jason Zions [SMTP:jazz@softway.com] wrote: >> > But g*zip is not portable, either, only *compress is "standard" on unix. >> > You can just as well install gtar, under that name, too (or link on Linux). >> >> gzip/gunzip are a helluva lot more portable than compress/uncompress; >> the code is universally available, and unencumbered by the Unisys patent >> on LZW compression. The requirement for a patent license to ship >> compress has driven us away from providing it; we use gzip/gunzip. -- http://www.bbc.com/ cgf@bbc.com "Strange how unreal VMS=>UNIX Solutions Boston Business Computing the real can be." - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From b.backhaus@bbbo.dortmund.netsurf.de Fri Aug 1 11:45:00 1997 From: b.backhaus@bbbo.dortmund.netsurf.de (Bernd Backhaus) Date: Fri, 01 Aug 1997 11:45:00 -0000 Subject: Porting from DJGPP code References: <33e127d4.8193813@smtp.netzone.com> Message-ID: <6b384$c-JEB@bbbo.dortmund.netsurf.de> *jeffdb@netzone.nospam.com wrote 01.08.97 00:14 in <33e127d4.8193813@smtp.netzone.com>: > (*jeffdb@netzone.nospam.com) > Remove *&.nospam from the above to reply While this is off-topic somehow, might I still remind you and others that nospam.com is a valid domain with a very unhappy SMTP server which has to bounce thousands of mails every day from people who did not bother to edit the faked address... Cheers, Bernd. -- --------------------------------------------------------------------------- Bernd Backhaus email: b.backhaus@dortmund.netsurf.de Brueckstr. 20 Compuserve: 100111,3061 44787 Bochum Fidonet: 2:2448/615.14 Germany - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ebritten@uci.edu Fri Aug 1 12:31:00 1997 From: ebritten@uci.edu (Eric Britten) Date: Fri, 01 Aug 1997 12:31:00 -0000 Subject: directx References: <3.0.3.32.19970731094435.00697934@mail.mindspring.com> Message-ID: <33E23580.F0D012D6@uci.edu> Derek Greene wrote: Check out.... http://dragon.acadiau.ca/~025668b/develop/cygnus/ I have also used gnuwin32 to use OLE & COM objects. I think directx uses a similar interface. It's not easy, you must use C not C++ for now. C++ cleans up your code alot, but it doesn't compile the Microsoft headers at all. For some reason the g++ is more peticular about the location of * in a function prototype that returns a pointer. The directx sample doesn't have any problems since it uses C. For instance, int * __stdcall foo(); gives an error. but int __stdcall * foo(); does not. Another thing is that if you declare the calling convention for a member function of a class, the calling convention is ignored. This may pose a problem since OLE expects stdcall for member functions of COM objects. Hope this helps. Eric Britten > has ANYONE managed to port directx to gnu-win32? puh-leez say yes, > puh-leez > say yes....:) > - > For help on using this list (especially unsubscribing), send a message > to > "gnu-win32-request@cygnus.com" with one line of text: "help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gbischof@verity.com Fri Aug 1 12:31:00 1997 From: gbischof@verity.com (Gusel Bischoff) Date: Fri, 01 Aug 1997 12:31:00 -0000 Subject: Creating new windows Message-ID: <2.2.32.19970801193858.00698340@devmail> If you have Mortice Kern Systems (MKS) installed on your PC, you should use start.exe. 1. call bash 2. inside bash call start bash and it will open a new window with bash prompt. Gusel Bischoff At 05:59 AM 8/1/97 +1200, you wrote: >Does anyone know if it is possible to create a new window in which to run a bash shell from within a bash shell? >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mc90085@central.ntua.gr Fri Aug 1 13:52:00 1997 From: mc90085@central.ntua.gr (Dionyshs Papadopoulos) Date: Fri, 01 Aug 1997 13:52:00 -0000 Subject: gcc -pipe works? Message-ID: <33E24CB8.D30771BD@central.ntua.gr> Thanks for all the patches Sergey I wonder if the gcc -pipe bug in b18 still remains or have been fixed with your patches - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ahc@lanl.gov Fri Aug 1 14:22:00 1997 From: ahc@lanl.gov (Allen H. Cogbill) Date: Fri, 01 Aug 1997 14:22:00 -0000 Subject: How does Cygnus compare to RSX Message-ID: <33E25377.5B32@lanl.gov> I've never used the Cygnus-Win32 development environment, but I have used RSXNT, which is an extension of the EMX project (for OS/2) to Win32. How does Cygnus compare? Does it provide the same capabilities? RSX seems especially easy to use (to me, at least), and appears to have few bugs. What isn't available is a g77 port, whereas I note that is is available under Cygnus. Any information greatly appreciated. Allen Cogbill Los Alamos National Lab ahc@lanl.gov - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bert_driehuis@nl.compuware.com Fri Aug 1 14:22:00 1997 From: bert_driehuis@nl.compuware.com (Bert Driehuis) Date: Fri, 01 Aug 1997 14:22:00 -0000 Subject: Unix domain sockets Message-ID: <33E24865.AB504930@nl.compuware.com> Hi Folks, I noticed that UNIX domain sockets (AF_UNIX) don't work with gnu-win32 beta 18 yet (or rather, that they don't work for me :-) I must admit to not having looked in the source code. Anyone know of patches or workarounds that address this? Cheers, -- Bert -- Bert Driehuis -- Compuware Amsterdam MIS dept -- phone: +31-20-3116119 "There is no such word as `impossible' in my dictionary. In fact, I think all the words between `herring' and `marmelade' are missing." -- Dirk Gently - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From cgf@bbc.com Fri Aug 1 14:55:00 1997 From: cgf@bbc.com (Chris Faylor) Date: Fri, 01 Aug 1997 14:55:00 -0000 Subject: New patch available -- ZSH anyone? References: <01BC9E7C.798C1C50@gater.krystalbank.msk.ru> Message-ID: With Sergey's new cygwin (plus my changes) it is now possible to produce a working version of ZSH. I'll probably be putting a binary on my web site soon. I haven't tried in a while, but I think ZSH may just configure "out of the box", too. This is just a heads up for anyone who wants to try building ZSH now. In article < 01BC9E7C.798C1C50@gater.krystalbank.msk.ru >, Sergey Okhapkin wrote: >- console code no longer alway translates CR to NL when tty support is >enabled. The behavior depends now of termios settings. >- signals should terminate blocking socket calls. >- Chris Faylor's (cgf@bbc.com) patch for wait and SIGCHLD raising logic - >cygwin processes now waits non-cygwin processes correctly, no more time >races problems with SIGCHLD raising and process's exit. >- utmp/wtmp files are filled now for every cygwin's session (both local >session and telnet session). "Who" command works now. Create /var/run and >/var/log directories to enable utmp logging. >- syslog() call handles now %m macro. >- signal raising keys (ctrl-C, ctrl-\ etc) restarts suspended tty output. > >New internet servers/clients added to remote.tar.gz -- http://www.bbc.com/ cgf@bbc.com "Strange how unreal VMS=>UNIX Solutions Boston Business Computing the real can be." - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From topcoder@mindspring.com Fri Aug 1 14:55:00 1997 From: topcoder@mindspring.com (Derek Greene) Date: Fri, 01 Aug 1997 14:55:00 -0000 Subject: directx References: <3.0.3.32.19970731094435.00697934@mail.mindspring.com> <33E23580.F0D012D6@uci.edu> Message-ID: <3.0.3.32.19970801045705.0069f524@mail.mindspring.com> thanks for the replies guys, but, i already knew about that URL, but he hasnt done enuf for me to worry with his stuff At 12:14 PM 8/1/97 -0700, you wrote: >Derek Greene wrote: > >Check out.... > http://dragon.acadiau.ca/~025668b/develop/cygnus/ > >I have also used gnuwin32 to use OLE & COM objects. I think directx >uses a similar interface. It's not easy, you must use C not C++ for >now. C++ cleans up your code alot, but it doesn't compile the Microsoft >headers at all. For some reason the g++ is more peticular about the >location of * in a function prototype that returns a pointer. > >The directx sample doesn't have any problems since it uses C. > >For instance, >int * __stdcall foo(); > >gives an error. but > >int __stdcall * foo(); > >does not. > >Another thing is that if you declare the calling convention for a member >function of a class, the calling convention is ignored. This may pose a >problem since OLE expects stdcall for member functions of COM objects. > >Hope this helps. > >Eric Britten > >> has ANYONE managed to port directx to gnu-win32? puh-leez say yes, >> puh-leez >> say yes....:) >> - >> For help on using this list (especially unsubscribing), send a message >> to >> "gnu-win32-request@cygnus.com" with one line of text: "help". > > > > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From rfrink@infinet.com Fri Aug 1 17:51:00 1997 From: rfrink@infinet.com (Rick Frink) Date: Fri, 01 Aug 1997 17:51:00 -0000 Subject: NT: 'nlist' function missing, equivalent in gnu-win32 ? Message-ID: <3.0.1.32.19970801204236.00917790@mail.infinet.com> >X-From_: rfrink@infinet.com Fri Aug 1 20:18:41 1997 >Return-Path: rfrink@infinet.com >X-Sender: rfrink@mail.infinet.com >Date: Fri, 01 Aug 1997 20:10:03 -0400 >To: ucd-snmp@ece.ucdavis.edu, gnu-win@cygnus.com >From: Rick Frink >Subject: NT: 'nlist' function missing, equivalent in gnu-win32 ? >Cc: rfrink@en-sam.etroy.com > > >Hello, > > I am trying to compile the UCD-CMU SNMP pkg under NT 4.0 with the >cygnus gnu-win32 pkg and am stuck. snmplib/system.c includes , >and then calls the 'nlist' function, but, my setup does not have nlist.h. >Also, I believe that nlist.h includes a.out.h, which does not seem to be >present. > >I am wondering if there is something equivalent in the NT gnu-win32 >world ? Any help/advice/guidance, would be much appreciated. Thanks. > >Rick > > > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dumser@ti.com Fri Aug 1 17:51:00 1997 From: dumser@ti.com (James Dumser) Date: Fri, 01 Aug 1997 17:51:00 -0000 Subject: New patch available References: <01BC9E7C.798C1C50@gater.krystalbank.msk.ru> Message-ID: On Fri, 01 Aug 1997 13:11, Sergey Okhapkin wrote: >- console code no longer alway translates CR to NL when tty support is > enabled. The behavior depends now of termios settings. >- signals should terminate blocking socket calls. >- Chris Faylor's (cfg@bbc.com) patch for wait and SIGCHLD raising > logic - cygwin processes now waits non-cygwin processes correctly, no > more time races problems with SIGCHLD raising and process's exit. >- utmp/wtmp files are filled now for every cygwin's session (both local > session and telnet session). "Who" command works now. Create /var/run > and /var/log directories to enable utmp logging. >- syslog() call handles now %m macro. >- signal raising keys (ctrl-C, ctrl-\ etc) restarts suspended tty > output. > >New internet servers/clients added to remote.tar.gz Installed this on my NT 4.0SP3 workstation and ran into a couple of gotchas: - Telneting into your own machine appears to work correctly, but telneting from some other machines (Suns running SunOS 4.1.3 in my experiments) did not echo terminal input -- commands only appeared after a return (you had to type blind). I compared stty -all outputs and did not find anything different. I also tried telneting from my PC to a Unix box then back to my PC; this also didn't work correctly. [Telneting from Solaris (5.5.1) does work correctly.] - stty rows x doesn't work. "Locally" (non-telnet), the command is accepted but has no effect (as indicated by a subsequent stty -all). When logged in through telnet, stty rows x gives "stty: standard input: No such file or directory." -- James Dumser 972.462.5335 dumser@ti.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jazz@softway.com Fri Aug 1 17:51:00 1997 From: jazz@softway.com (Jason Zions) Date: Fri, 01 Aug 1997 17:51:00 -0000 Subject: Where are Reg* implemented? References: Message-ID: <33E27A57.1D12@softway.com> > Good point Jason about the ld man page. However, we are discuessing gcc > or g++, not ld. A difference that makes no difference; see below. > It is clear that we need to put the main() entry ( in one of the .c or > .cxx files) as the first of the source file list ( including .o and .c). > However, the libraries are not 'source files' to g++. Instead, they are > assigned as 'options' by '-l'. Therefore, it is really the compiler's > responsibility to add these options to correct position in parameter list > passed to 'ld' later. And how on earth is gcc supposed to guess at this? There are times when what *must* be passed to the loader is this: file1.o file2.o -lmylib file3.o file4.0 -lotherlib If gcc can't preserve this ordering, it's fundamentally broken. Jason P.S. I'd address you informally, as you did me, but I admit to being stumped as to whether I should call you Ku, Wei, or both (and in which order - I've seen your address come with your name in both orders... :-). - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gordoni@cygnus.com Fri Aug 1 17:51:00 1997 From: gordoni@cygnus.com (Gordon Irlam) Date: Fri, 01 Aug 1997 17:51:00 -0000 Subject: Revised licensing terms Message-ID: <199708020025.RAA11413@snuffle.cygnus.com> Several weeks ago I posted our planned new licensing terms for Cygwin32. We have received a lot of useful feedback from people in response. Cygnus listens. In response we have decided to further revise the licensing terms. We want to ensure Cygwin32 continues to be made available to the net community under terms that they find reasonable. We want to continue to work with, cooperate with, and to be considered a part of the net community. We draw our roots from the net, and we want to contribute back to the net. And so it is with Cygnwin32. We do not want a portion of the net community to be unable to use Cygwin32 because of the licensing terms. (The new licensing terms were drafted in such a way as to ensure Cygwin32 continued to be freely available to the net community. The only restriction we imposed was on use of Cygwin32 by Cygnus' embedded competitors. However, we perhaps a little naively failed to realize that some people on the net community would have problems with the licensing terms, not because the license prevented them from using the software, but because they were unhappy with a license that contained any restrictions, irrespective of whether the restrictions directly impacted their use of the software. This had not been our intent in coming up with the license). Our new licensing terms are still being prepared so I can't present them to you yet, however I can explain their intended flavor. Our new licensing terms will permit Cygwin32 to be used under the terms of the GPL, and also under a separate commercial use license that frees people from the constraints of the GPL. The precise terms of this separate commercial use license have yet to be determined. Once we have determined the precise terms we will announce the new license here. One of the things we are presently considering is charging a fee for use of Cygwin32 under the commercial use license. Use of Cygwin32 under the terms of the GPL would remain free of charge. (If we decide to charge fees for commercial use, they would probably be in line with the fees we charge for our commercial GNUPro compilers: in the range $2-3k for a single developer, $7-15k for up to five developers, and $25-50k for up to twenty five developers). If you have any thoughts on our charging fees for the commercial use of Cygwin32, please feel free to email me. To ensure no one might have been mislead by our previously announced new licensing terms, you are free to make use of the current beta 18 release of Cygwin32 under the previously announced Cygwin32 Version 1.0 license. regards, Gordon Irlam Cygnus Solutions - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From *jeffdb@netzone.nospam.com Fri Aug 1 18:47:00 1997 From: *jeffdb@netzone.nospam.com (Mikey) Date: Fri, 01 Aug 1997 18:47:00 -0000 Subject: directx References: <3.0.3.32.19970731094435.00697934@mail.mindspring.com> <33E23580.F0D012D6@uci.edu> Message-ID: <33e38f56.707231@smtp.netzone.com> The headers in SDK_4_ming_cygwin32 will compile using g++, that isn't the major problem, the reason you have to use C for the INTERFACE_() declarations is that the system dll's were built with VC++, which has a different name mangling convention than g++. Unless you want to try to figure out msvc++ name mangling and implement that for gnu-win32 g++, you will never be able to use any of the C++ routines in the system libraries, you will always have to "thunk" to C. On Fri, 01 Aug 1997 12:14:08 -0700, you wrote: >Derek Greene wrote: > >Check out.... > http://dragon.acadiau.ca/~025668b/develop/cygnus/ > >I have also used gnuwin32 to use OLE & COM objects. I think directx >uses a similar interface. It's not easy, you must use C not C++ for >now. C++ cleans up your code alot, but it doesn't compile the Microsoft >headers at all. For some reason the g++ is more peticular about the >location of * in a function prototype that returns a pointer. > >The directx sample doesn't have any problems since it uses C. > >For instance, >int * __stdcall foo(); > >gives an error. but > >int __stdcall * foo(); > >does not. > >Another thing is that if you declare the calling convention for a member >function of a class, the calling convention is ignored. This may pose a >problem since OLE expects stdcall for member functions of COM objects. > >Hope this helps. > >Eric Britten > >> has ANYONE managed to port directx to gnu-win32? puh-leez say yes, >> puh-leez >> say yes....:) >> - >> For help on using this list (especially unsubscribing), send a message >> to >> "gnu-win32-request@cygnus.com" with one line of text: "help". > > > >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > (*jeffdb@netzone.nospam.com) Remove *&.nospam from the above to reply Mikey - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From wgreathouse@smva.com Fri Aug 1 18:47:00 1997 From: wgreathouse@smva.com (William Greathouse) Date: Fri, 01 Aug 1997 18:47:00 -0000 Subject: Where are Reg* implemented? References: Message-ID: <33E2589B.D48D8ED8@smva.com> Ku Wei wrote: > > Good point Jason about the ld man page. However, we are discuessing > gcc > or g++, not ld. > > It is clear that we need to put the main() entry ( in one of the .c or > .cxx files) as the first of the source file list ( including .o and > .c). > However, the libraries are not 'source files' to g++. Instead, they > are > assigned as 'options' by '-l'. Therefore, it is really the compiler's > responsibility to add these options to correct position in parameter > list > passed to 'ld' later. You said the operative word -- "options". Libraries (-l) are _not_ options in the normal sense. They are "special" object modules to include in the final output. > > For example, for IBM AIX C compiler, you are only allowed to put > parameters "before" the source file list: xlC [options] files > On the other hand, g++ allows two different syntex: g++ > [option|filenam] ... AIX will allow you to put -l at the end (as is typical of most compiler/linkers) as it is not an option but a shorthand representation of a special object file to search during linking. It is also necessary to note that AIX does not implement ld in a "normal" manner -- it will search all object modules in a circular fashion, left to right and back to the front, until all references are resolved (or a complete pass does not resolve a symbol). Typical linkers do one pass left to right to resolve references -- each object is evaluated once for unresolved external references, which must be resolved by an object to its right on the link line. In this case it would not make sense to include the library before the main object, as no portion of the library would be incorporated in to the output, because no external references exist when it is scanned. > Anyway, why use same name in your own code as the name in libraries to > confuse the linker ? > > Sincerely, > Wei Ku > Enjoy the debate, Bill begin: vcard fn: William Greathouse n: Greathouse;William org: SMV America Engineering adr: 8380 Darrow Road;;;Twinsburg;OH;44087; email;internet: wgreathouse@smva.com title: Senior Engineer tel;work: (216) 425-1340 x4006 tel;fax: (216) 405-7684 note: wgg@netcom.com -- Home Account x-mozilla-cpt: ;0 x-mozilla-html: TRUE end: vcard From mbenz@sabre.com Fri Aug 1 20:29:00 1997 From: mbenz@sabre.com (Micheal A. Benzinger) Date: Fri, 01 Aug 1997 20:29:00 -0000 Subject: VIM 5.0k References: <9707290755.AA26000@hagar.aspentec.com> Message-ID: <3.0.3.32.19970801223019.00916100@pop.flash.net> Jerome, I got vim-5.0l and finally figured out what the problem was. It turns out that when using termcap, you need to modify vim.h as follows: 513 #if defined(AMIGA) || defined(__linux__) || defined(__CYGWIN32__) || defined(__QNX__) 514 # define TBUFSZ 2048 /* buffer size for termcap entry */ 515 #else 516 # define TBUFSZ 1024 /* buffer size for termcap entry */ 517 #endif The defined(__CYGWIN32__) needs to be added to up the buffer size for the terminal type 'linux'. If this is not done, the linux entry in the termcap file overflows the character buffer and overwrites the stack. Now that I have gotten as far as you, I will work on trying to figure out why it is so darned slooooooow. It seems that this is related to the I/O. When was debugging to determine the problem, an open for the termcap file took over thirty seconds. Mike Benzinger At 09:54 AM 7/29/97 +0200, Jerome Gay wrote: > > >I got a lot of problem too trying to build vim5.0k for gnu-win32, and it's >still not working properly. >Do you have binary mounts or text mounts ? (just type mount and check if you >have text=binary or text!=binary for each mount) >You'll need binary mounts to build it. > >I'm able to build vim (using ncurses), but when it runs, the response is >desperately slow (this has been identified in the vim faq but for win95, I >got the same symtoms on NT4SP3 too). > >Any idea folks ? > >Jerome - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ebritten@uci.edu Fri Aug 1 21:34:00 1997 From: ebritten@uci.edu (Eric Britten) Date: Fri, 01 Aug 1997 21:34:00 -0000 Subject: directx References: <3.0.3.32.19970731094435.00697934@mail.mindspring.com> <33E23580.F0D012D6@uci.edu> <33e38f56.707231@smtp.netzone.com> Message-ID: <33E2B8E3.1B3E4DD8@uci.edu> I am glad that the headers in SDK_4_ming_cygwin32 will compile using g++. I just downloaded the files to patch my mssdk platform headers. However the name mangling conventions of VC++ are irrelavent. If you write code that uses C++ and OLE it doesn't matter what compiler you because you only work with interface pointers. All functions that your program calls in other DLLs are declared and exported in C with no name mangling(this doesn't count interface functions but you call them through pointers). One calls the C functions to get interface pointers to objects. You can work with C++ or C with these functions. It's easier in C++ because you type less. In C you must pass a pointer to the object as the first parameter of any interface function. In C++ this is the equivalent of the this pointer which is handled implicitly. For example: pUnknown->AddRef(pUnknown); //C pUnknown->AddRef(); //C++ And it gets worse if you declare your own objects in C as opposed to C++. I have built a shell extension with modifications of the SDK headers and a shell extension example. Although it crashed my system I know that there were no link errors. Hopefully the headers in SDK_4_ming_cygwin32 will allow me to compile, link and debug a working shell extension without crashing my Win95 machine. Thus far I have been unable to so. Eric Britten Mikey wrote: > The headers in SDK_4_ming_cygwin32 > will compile using g++, that isn't > the major problem, the reason you have to use C for > the INTERFACE_() declarations is that the > system dll's were built with VC++, > which has a different name mangling convention > than g++. > > Unless you want to try to figure out msvc++ name mangling > and implement that for gnu-win32 g++, you will never > be able to use any of the C++ routines in the system > libraries, you will always have to "thunk" > to C. > > On Fri, 01 Aug 1997 12:14:08 -0700, you wrote: > > >Derek Greene wrote: > > > >Check out.... > > http://dragon.acadiau.ca/~025668b/develop/cygnus/ > > > >I have also used gnuwin32 to use OLE & COM objects. I think directx > >uses a similar interface. It's not easy, you must use C not C++ for > >now. C++ cleans up your code alot, but it doesn't compile the > Microsoft > >headers at all. For some reason the g++ is more peticular about the > >location of * in a function prototype that returns a pointer. > > > >The directx sample doesn't have any problems since it uses C. > > > >For instance, > >int * __stdcall foo(); > > > >gives an error. but > > > >int __stdcall * foo(); > > > >does not. > > > >Another thing is that if you declare the calling convention for a > member > >function of a class, the calling convention is ignored. This may > pose a > >problem since OLE expects stdcall for member functions of COM > objects. > > > >Hope this helps. > > > >Eric Britten > > > >> has ANYONE managed to port directx to gnu-win32? puh-leez say yes, > >> puh-leez > >> say yes....:) > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sat Aug 2 01:05:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sat, 02 Aug 1997 01:05:00 -0000 Subject: gcc -pipe works? Message-ID: <01BC9F3C.995AA540@sos> Dionyshs Papadopoulos wrote: > Thanks for all the patches Sergey > I wonder if the gcc -pipe bug in b18 still > remains or have been fixed with your patches It's the bug of gcc driver itself, but not cygwin.dll. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From grhgkim@tiger.korea.ac.kr Sat Aug 2 04:41:00 1997 From: grhgkim@tiger.korea.ac.kr (HgKim) Date: Sat, 02 Aug 1997 04:41:00 -0000 Subject: installation problem? Message-ID: <01BC9F84.495F19C0@Hgkim.korea.ac.kr> When compiling, following message out. "GCC.EXE: installation problem, cannot exec `cpp': No such file or directory" How can I do fix this problem? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From marcusd@cathcart.sysc.pdx.edu Sat Aug 2 10:23:00 1997 From: marcusd@cathcart.sysc.pdx.edu (Marcus G. Daniels) Date: Sat, 02 Aug 1997 10:23:00 -0000 Subject: open vs stat and install behavior Message-ID: I think `install' should be able to deal with executable filenames without an .exe suffix. The b18 behavior is that `stat' will succeed finding an executable named "t", but `open' will fail. For example, the following program correctly prints the file size of itself, but fails on the open. If there is an important reason for the different open and stat behavior, I propose the patch that follows to install.c. #include #include #include #include main() { struct stat statbuf; const char *file = "t"; if (stat (file, &statbuf) == 1) { fprintf (stderr, "stat failed to find `%s'\n",file); exit (1); } printf ("%d\n", statbuf.st_size); if (open (file, O_BINARY | O_RDONLY) == -1) { fprintf (stderr, "open failed to find `%s'\n", file); exit (1); } } /* Local Variables: compile-command: "gcc t.c -o t" End: */ *** install.c Tue Apr 15 00:57:17 1997 --- install.c.new Sat Aug 2 10:00:52 1997 *************** *** 309,321 **** Return 0 if successful, 1 if an error occurs. */ static int ! install_file_in_file (char *from, char *to) { int to_created; int no_need_to_chown; ! if (copy_file (from, to, &to_created)) ! return 1; if (strip_files) strip (to); no_need_to_chown = (to_created --- 309,346 ---- Return 0 if successful, 1 if an error occurs. */ static int ! install_file_in_file (char *from_, char *to_) { int to_created; int no_need_to_chown; ! char *from, *to; ! int fromfd = open (from, O_RDONLY | O_BINARY, 0); ! if (fromfd == -1) ! { ! from = alloca (strlen (from_) + 5); ! to = alloca (strlen (to_) + 5); ! strcpy (from, from_); ! strcat (from, ".exe"); ! strcpy (to, to_); ! strcat (to, ".exe"); ! fromfd = open (from, O_RDONLY | O_BINARY, 0); ! if (fromfd == -1) ! { ! from = from_; ! to = to_; ! } ! else ! close (fromfd); ! } ! else ! { ! close (fromfd); ! from = from_; ! to = to_; ! } ! if (copy_file (from, to, &to_created)) ! return 1; if (strip_files) strip (to); no_need_to_chown = (to_created - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Francois.Chenais@insat.com Sat Aug 2 10:23:00 1997 From: Francois.Chenais@insat.com (Francois Chenais) Date: Sat, 02 Aug 1997 10:23:00 -0000 Subject: bash for dos/win Message-ID: <199708021727.TAA22719@golgoth.inlandsys.com> Hello, Is bash exists for DOS/windows ? If yes, where can I get it ? Thanks for all informations. [Francois] --- Francois Chenais Yumi Yumi .oooO Man Oooo. chenais@insat.com ( ) Blong ( ) http://www.insat.com/Club/chenais \_) VANUATU (_/ ----------------------------------------------------------- Linux --- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kirsh@geocities.com Sat Aug 2 10:42:00 1997 From: kirsh@geocities.com (naftali kirsh) Date: Sat, 02 Aug 1997 10:42:00 -0000 Subject: Having problem with "make" References: <33DE52F1.37DA@infocom.net> Message-ID: <31FE0DF4.F94E7AD6@geocities.com> Rhea Elliott wrote: > I am a newbie to windows programming. Just installed cygwin32. > Compiles working perfectly but, when I run make, I get a lot of frame > messages followed by another message "Probably corrupted stack. I am > running "Minimalist" and using crtdll.dll instead of cygwin.dll. > Would > this have something do do with my problem? If so, how can I get > around > it? > > -Thank you for any help you can give me. > > Dana Elliott > - > For help on using this list (especially unsubscribing), send a message > to > "gnu-win32-request@cygnus.com" with one line of text: "help". I had this problem too, I think the solution for this is to set the GCC_EXEC_PREFIX to the right directory. Naftali Kirsh (kirsh@geocities.com) - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From rdorgan@worldnet.att.net Sat Aug 2 11:01:00 1997 From: rdorgan@worldnet.att.net (Tom) Date: Sat, 02 Aug 1997 11:01:00 -0000 Subject: bash window banner Message-ID: <33E375DA.1895@worldnet.att.net> Greetings: I notice that bash changes the window banner based on the last command executed. I always configure my x windows to show hostname:pwd. It allows me to leave my prompt very small as well as avoiding the nervous pwd'ing I usually do with only a '$ ' for PS1. Is there any way to configure this in the current bash executable or would it require a programming change? Oh, and I am amazed at the progress the gnu-win32 package is making. tom dorgan. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mc4420@mclink.it Sat Aug 2 14:02:00 1997 From: mc4420@mclink.it (Giandomenico De Sanctis) Date: Sat, 02 Aug 1997 14:02:00 -0000 Subject: Visual Basic and GCC Message-ID: <199708022053.WAA16887@ammi.mclink.it> For who is interested using GNU C-Win32 with Visual Basic, I follow this steps: Note: I installed the Minimal GCC-Win32, then the note is referred at this package; also, the installation is on Win 95, I not have any experience with Win NT 1) compiling a C DLL, callable by VB 4: 1.1- prepare the .DEF file with the functions exported by the library, for example EXPORTS func1 func2 1.2- execute the dlltool utility ('myl' is the name of library in the example): dlltool --def myl.def --output-exp myl.exp --output-lib myl.a --dllname myl.dll with this you obtained the .EXP file for use in successive step, and the .A file for using the DLL in others C/C++ programs 1.3- compile/link the DLL with: gcc -o myl.dll myl.c myl.exp -dll -mrtd -windows -mpentium use '-windows' only if the DLL call any Win functions; '-mpentium' optimize the code for Pentium, use '-m386' or '-m486' for other processors; '-mrtd' is essential, because force the use of the STDCALL calling convention, that is also used by VB 4 With this step you obtain the MYL.DLL. 1.4- if you want debug the DLL, only by a C calling test program, not by VB, add the '-ggdb3' parameter; after, to debug execute gdb testp.exe assuming 'testp.exe' is the test C program 1.5- remember that VB 4 integer type correspond to short C type (2 bytes) in Win32, because the integer C type is 4-bytes long 2) obtaining a link definition file to use an existing DLL with a GCC program/DLL: 2.1- if you have the MS Visual C .LIB file for the library, prepare the list of exported functions with: nm exil.lib >exil.def (of course, 'exil' is the existing library) manually edit exil.def, leaving only the names of exported functions, without leading underscore, but leaving the trailing @xx if any 2.2- if you not have the .LIB, in explorer right click on the DLL file, and choose preview; then, manually (the clipboard is not active !) annotate the names of functions, and prepare the .DEF file 2.3- execute dlltool: dlltool -k --def exil.def --output-lib exil.a --dllname exil.dll the '-k' option eliminate the trailing @xx from function names in the .A files, and is necessary when using the STDCALL convention (then, always when compiling DLL for VB). - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From grhgkim@tiger.korea.ac.kr Sat Aug 2 19:59:00 1997 From: grhgkim@tiger.korea.ac.kr (HgKim) Date: Sat, 02 Aug 1997 19:59:00 -0000 Subject: gcc compiler? Message-ID: <01BCA004.8D706D00@Hgkim.korea.ac.kr> When compiling, following message : ld : cannot open crt0.o : No such file or directory. gcc : Internal compiler error : program ld got fatal signal 1. Please suggest what is the best way to deal with this problem. Thanks for your help - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From cgf@bbc.com Sat Aug 2 21:57:00 1997 From: cgf@bbc.com (Chris Faylor) Date: Sat, 02 Aug 1997 21:57:00 -0000 Subject: bash window banner References: <33E375DA.1895@worldnet.att.net> Message-ID: In article < 33E375DA.1895@worldnet.att.net >, Tom wrote: >I notice that bash changes the window banner based on the last command >executed. I always configure my x windows to show hostname:pwd. It >allows me to leave my prompt very small as well as avoiding the nervous >pwd'ing I usually do with only a '$ ' for PS1. > >Is there any way to configure this in the current bash executable or >would it require a programming change? Sergey's newer cygwin.dll's have an option (suggested by me *blush*) where setting the environment variable CYGWIN_NOTITLE causes the title to remain unchanged. You might want to check that out. Sergey's homepage is: http://www.lexa.ru/sos -- http://www.bbc.com/ cgf@bbc.com "Strange how unreal VMS=>UNIX Solutions Boston Business Computing the real can be." - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pact@parallel.cs.pku.edu.cn Sat Aug 2 22:33:00 1997 From: pact@parallel.cs.pku.edu.cn (PACT Group) Date: Sat, 02 Aug 1997 22:33:00 -0000 Subject: Question about X11R6.3 Message-ID: I have installed gnu-win32 and X11R6.3 b18 on my NT 4.0 (chinese version ). But when I run the X app , I always find this error reported by NT : [0x0259bac6] instruction refered to memory address [0x100476c0] which can not be written , and the program is terminated . I don't know if this problem has some relations with my chinese vesion NT ? Zheng Gengbin - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jeffdbREMOVETHIS@netzone.com Sun Aug 3 00:16:00 1997 From: jeffdbREMOVETHIS@netzone.com (Mikey) Date: Sun, 03 Aug 1997 00:16:00 -0000 Subject: directx References: <3.0.3.32.19970731094435.00697934@mail.mindspring.com> <33E23580.F0D012D6@uci.edu> <33e38f56.707231@smtp.netzone.com> <33E2B8E3.1B3E4DD8@uci.edu> <33e2ec32.17151848@smtp.netzone.com> <33E3A33C.BB10DB2B@uci.edu> Message-ID: <33e52dae.19014284@smtp.netzone.com> On Sat, 02 Aug 1997 14:14:37 -0700, you wrote: >Mikey wrote: > >> On Fri, 01 Aug 1997 21:34:43 -0700, you wrote: >> >> >I am glad that the headers in SDK_4_ming_cygwin32 >> >will compile using g++. I just downloaded the files to patch my >> mssdk >> >platform headers. >> > >> >However the name mangling conventions of VC++ are irrelavent. If you >> >> >write code that uses C++ and OLE it doesn't matter what compiler you >> >> It matters a lot, if you're trying to use pre written vc++ code, >> because the declarations >> of the INTERFACE() functions will be in c++, so you have to rewrite >> them >> to C. and I hate typing ;^). > >You have a point there. However, I built the shellext sample in the >platform sdk, it was a while ago, so don't ask me how I did it. I had >to change all sorts of header declarations. Anyway it was written in >C++, and it did link correctly. However it also crashed when loaded by >Explorer. I think one of its problems was the fact that the compiler >complained about member functions declared with stdcall, it said >attribute ignored. If it defaulted to C calling convention the stack >would get corrupted pretty quickly. In any event C works better for >now. My current project uses OLE for drag and drop. It uses the C >interface successfully to create my own objects(IDropTarget) and >manipulate others such as IDataObject. > >> >because you only work with interface pointers. All functions that >> your >> >program calls in other DLLs are declared and exported in C with no >> name >> >mangling(this doesn't count interface functions but you call them >> >> My point was that if you try to use the headers using g++ declaring >> your classes and expecting to be able to link with the standard >> libraries, you won't be able to. >> > >I don't see why this is so. If you are linking with a VC++ dll that >exports classes, then you will have problems, but I know of no such >instance where this happens. Maybe I will try it out. If you have any >error messages that support this send it to me. > > >> >through pointers). One calls the C functions to get interface >> pointers >> >to objects. You can work with C++ or C with these functions. It's >> >easier in C++ because you type less. In C you must pass a pointer to >> >> >the object as the first parameter of any interface function. In C++ >> >this is the equivalent of the this pointer which is handled >> implicitly. >> >For example: >> > >> >pUnknown->AddRef(pUnknown); //C >> >pUnknown->AddRef(); //C++ >> > >> >And it gets worse if you declare your own objects in C as opposed to >> >C++. >> > >> >I have built a shell extension with modifications of the SDK headers >> and >> >> Was it dropext.dll? >> Unfortunately although I was able to build dropext.dll, from the >> win32api.exe >> samples, it did the same to me, crashed the system after I added the >> .reg file, >> so I backed out my changes to the shellapi.h shlobj.h and shlguid.h >> files, untill >> I could get a better handle on what was going on. >> > >Yes it was dropext.dll. I had to link(ld) with the msvc++ library >uuid.lib for the IID constants, I don't know how or why this works. I JUST FIGURED IT ALL OUT!!! This works because uuid.lib contains the actual .obj files what you ended up with was a statically linked binary, run ar -t uuid.lib and you will see what I mean. I am glad that you pointed this out to me, you have saved me a lot of time and frustration THANKS ;^))))))) The reason the shellext example crashed was probably because the obj files in shell32.lib that contain the CLSID IID UID etc. structures weren't there in libshell32.a, here are the changes I had to make to crtdll/Makefile to get the correct libraries for shell32 and uuid. liboldnames.a: oldnames.def dlltool --dllname crtdll.dll --def $< --output-lib $@ -k %.def: $(SDK)/lib/%.lib set -xv;\ [ -d obj/i386/ ] || mkdir -p obj/i386/;\ cp $< . ;\ Objs=`ar -t $(notdir $<)|grep -v \\.dll`;\ [ -z "$$Objs" ] && exit 0;\ ar -x $(notdir $<) $$Objs;\ ar -d $(notdir $<) $$Objs;\ for file in $$Objs;do \ objcopy $$file `basename $$file .obj`.o;\ rm $$file;\ done ../bin/lib2def $(notdir $<) > $@ rm $(notdir $<) lib%.a: %.def dlltool --dllname $*.dll --def $< --output-lib $@ -k set -xv;\ Objs=`echo *.o`;\ [ "$$Objs" = "*.o" ] && exit 0;\ ar urvs $@ $$Objs;\ rm *.o # to keep a copy of the .def files uncomment the next 6 lines #%.def: $(SDK)/lib/%.lib # -mkdir -p defs # ../bin/lib2def $< > defs/$@ also you might want to add echo fakefunc$RANDOM to lib2def.sh right after echo LIBRARY etc. so the empty uuid.def file dosen't give you a syntax error. I just built dropext.dll installed it and it works perfect!!! >Since I wasn't using the lib to import function calls I thought it would >be ok. > >> I'm pretty sure now that a major part of the problem was that I didn't >> >> take the time to "fix" the IID CLSID etc... pointers, to the >> __imp_IID* >> references in the import libraries, which is what I'm working on right >> now. >> >> So I'm afraid the SDK patches aren't going to help you very much. >> > >I found that some IID constants are in uuid.lib, however I am sure they >can befound elsewhere as well. > >> >a shell extension example. Although it crashed my system I know that >> >> >there were no link errors. >> > >> >Hopefully the headers in SDK_4_ming_cygwin32 will allow me to >> compile, >> >link and debug a working shell extension without crashing my Win95 >> >machine. Thus far I have been unable to so. >> > > >Eric Britten > > (jeffdbREMOVETHIS@netzone.com) delete REMOVETHIS from the above to reply Mikey - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From david@coent.demon.co.uk Sun Aug 3 04:57:00 1997 From: david@coent.demon.co.uk (David Coe) Date: Sun, 03 Aug 1997 04:57:00 -0000 Subject: VIM 5.0k References: <3.0.3.32.19970801223019.00916100@pop.flash.net> Message-ID: <33E42F80.82C@coent.demon.co.uk> Applying Chris Faylor's and Michael Benzinger's patches to vim-5.0l together with one to define TCIOFLUSH in pty.c seems to get things going for console, athena and motif (lesstif-0.79) versions. Thanks good people! I don't find any of the sluggishness complained of on NT4SP3 + Sergey's 1 Aug coolview; however, I have only termcap-1.3 installed with the entries for linux from mc and xterm from rxvt. Maybe this is a problem confined to terminfo or ncurses. I do find a whole slew of X11 errors ("could not find menu widget ...", "cannot find callback list ...") if gvim is forked and detached from the calling shell. Using gvim -f for either of the gui-enabled versions fixes this. Is this something for Sergey or Geoffrey's attention? -- Dr David Coe \=\ 58 Fairlawn Drive, East Grinstead \=\ Tel +44 1342 326860 West Sussex, RH19 1NT, United Kingdom \=\ Fax +44 1342 316019 From David.Billinghurst@restech.cra.com.au Sun Aug 3 16:17:00 1997 From: David.Billinghurst@restech.cra.com.au (Billinghurst, David) Date: Sun, 03 Aug 1997 16:17:00 -0000 Subject: Successful port of rcs-5.7 to gnu-win32 b18 Message-ID: I have ported rcs-5.7 to gnu-win32 b18. Build procedure 1. Apply patches below to rcs-5.7 source tree 2. In rcs-5.7 directory ./configure make (cd man; make install) (cd src; make x=.exe install) 3. To check installation get a working "who am i" command make installcheck NOTE: Two tests are dissabled. I think the problems are in the test routine rather than with rcs. What I did 1. configure had problems with "diff3 file1 file2 /dev/null" This is a gnu-win32 problem 2. The patch to src/rcslex.c is an rcs bug 3. I think the problem with rcstest is a gnu-win32 bug +++++++++++++++++++++++++++++++++++++++++ (Mr) David Billinghurst Comalco Research Centre PO Box 316, Thomastown, Vic, Australia, 3074 Phone: +61 3 9469 0642 FAX: +61 3 9462 2700 Email: David.Billinghurst@restech.cra.com.au It's not my speeling that's the trouble. It's just taht I can't tipe. *** configure.orig Fri Jun 16 16:57:49 1995 --- configure Sun Aug 03 21:25:03 1997 *************** *** 405,410 **** --- 405,411 ---- echo 0 >conftest0 echo 0 >conftest0c echo 1 >conftest1 + echo -n > conftestn cat >conftestok <<'EOF' d1 1 a1 1 *************** *** 637,643 **** rcs_cv_prog_diff3_bin=$DIFF3 case $rcs_cv_prog_diff3_bin in '') ! PATH=$diff3PATH sh -c "exec diff3 -E -m -L 0 -L 1 -L 2 conftest0 conftest1 /dev/null" >conftestout 2>/dev/null case $? in 1) if cmp -s - conftestout <<'EOF' --- 638,644 ---- rcs_cv_prog_diff3_bin=$DIFF3 case $rcs_cv_prog_diff3_bin in '') ! PATH=$diff3PATH sh -c "exec diff3 -E -m -L 0 -L 1 -L 2 conftest0 conftest1 conftestn" >conftestout 2>/dev/null case $? in 1) if cmp -s - conftestout <<'EOF' *** src/rcslex.c.orig Sun Aug 03 21:34:51 1997 --- src/rcslex.c Sun Aug 03 21:36:59 1997 *************** *** 1113,1119 **** /* Open NAME for reading, yield its descriptor, and set *STATUS. */ { int fd = fdSafer(open(name, O_RDONLY ! # if OPEN_O_BINARY | (strchr(type,'b') ? OPEN_O_BINARY : 0) # endif )); --- 1113,1119 ---- /* Open NAME for reading, yield its descriptor, and set *STATUS. */ { int fd = fdSafer(open(name, O_RDONLY ! # if OPEN_O_BINARY && !(large_memory && maps_memory) | (strchr(type,'b') ? OPEN_O_BINARY : 0) # endif )); *** src/rcstest.orig Fri Jun 16 16:19:23 1995 --- src/rcstest Sun Aug 03 22:15:53 1997 *************** *** 239,247 **** rcs -l1.1.1 $q a.c && ci -r1.1.1.2 -mm $q a.c && rcs -b1.1.1 $q a.c && ! test " `co -p $q a.c`" = ' 1.1.1.2' || { echo "#rcs -b1.1.1 failed"; exit 1; } rcs -b $q a.c && ! test " `co -p $q a.c`" = ' 2.2' || { echo "#rcs -b failed"; exit 1; } echo 2.3 >a.c || exit 2 rcs -U $q a.c || { echo "#rcs -U failed"; exit 1; } --- 239,247 ---- rcs -l1.1.1 $q a.c && ci -r1.1.1.2 -mm $q a.c && rcs -b1.1.1 $q a.c && ! test " `co -p $q a.c`" = ' 1.1.1.2' || { echo "#rcs -b1.1.1 failed"; } rcs -b $q a.c && ! test " `co -p $q a.c`" = ' 2.2' || { echo "#rcs -b failed"; } echo 2.3 >a.c || exit 2 rcs -U $q a.c || { echo "#rcs -U failed"; exit 1; } - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From racke@gundel.han.de Sun Aug 3 19:19:00 1997 From: racke@gundel.han.de (Stefan Hornburg) Date: Sun, 03 Aug 1997 19:19:00 -0000 Subject: gdbm References: <33DE5EDF.C5D53B59@Eng.Sun.COM> Message-ID: Padmakar Vishnubhatt writes: > > Is gdbm a part of win32? If not, I'd like to port this to win32. Could > you point me to the sources for gdbm? gdbm is short for GNU Dbm. You get it at the GNU mirrors around the world - or grab it from http://www.delorie.com/gnu/ . -- This site is powered by Linux 2.0.29 and his friends GNU CC 2.7.2, C Library 5.4.23, C++ Library 2.7.2.1, Binutils 2.8.1.0.1. Surf to Racke's Webnode (URL: http://www.han.de/~racke )! - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From chin.cheekai@singapore.sterling.com Sun Aug 3 19:59:00 1997 From: chin.cheekai@singapore.sterling.com (Chin Chee-Kai) Date: Sun, 03 Aug 1997 19:59:00 -0000 Subject: request for a simple search for mail archive References: <199706101938.MAA06102@singapore.sterling.com> Message-ID: Hi Geoffrey, Understand that you guys at Cygnus are busy and only doing all of gnu-win32 parttime. Just a suggestion here; I wonder if it is possible to provide a simple search mechanism at the mailing list page http://www.cynus.com/ml/gnu-win32/ that just does a simple fgrep search and returns the mails that contain the given word. It'd be really useful, as I'm looking for queries/responses regarding ODBC calls using gnu-win32. Thanks. Chin Chee-Kai - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From giac@dalim.de Sun Aug 3 23:11:00 1997 From: giac@dalim.de (Philippe GIACINTI) Date: Sun, 03 Aug 1997 23:11:00 -0000 Subject: patch: correct behaviour of dlls ( init & fork ) Message-ID: <9708040809.ZM11642@pion.kehl.dalim.de> Hi, These changes are based on original cdk b18 distribution + the patch "gnu-win32-b18-patches.diff" created by Mumit Khan with almost all Sergey Okhapkin's patches (which is included in my tar file or that you can get at http://www.xraylith.wisc.edu/~khan/software/gnu-win32 ). Regards. P.S: 1) you may have some troubles with relocations that can be solved in adding a third link pass. This is only a workaround and I think that the problem is deeper. 2) I you have problems while using the macro, try to remove the fixup ("asm(.idata$3 .....);") from the macro and add it at the end of your code. -- ========================================================================== | Philippe GIACINTI | | | | DALiM GmbH R&D email: giac@dalim.de | | Daimler Strasse 2, tel: +49.7851.9196-28 | | D-77694 Kehl-Sundheim Germany fax: +49.7851.73576 | | | ========================================================================== - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gay@aspentec.com Mon Aug 4 02:56:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Mon, 04 Aug 1997 02:56:00 -0000 Subject: latest Sergey's update (08/01) crashes on Win95 Message-ID: <9708040956.AA28128@hagar.aspentec.com> Hi all, I just installed Sergey's latest coolview, dated Aug,1st... when I start bash, WIndows95 crahes completely (hard-reset necessary)... any ideas ? Jerome - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From michael@ime.rwth-aachen.de Mon Aug 4 05:25:00 1997 From: michael@ime.rwth-aachen.de (Michael Holzapfel) Date: Mon, 04 Aug 1997 05:25:00 -0000 Subject: How does Cygnus compare to RSX Message-ID: <33E5C55F.2311@ime.rwth-aachen.de> Allen H. Cogbill wrote: > > I've never used the Cygnus-Win32 development environment, but I > have used RSXNT, which is an extension of the EMX project (for > OS/2) to Win32. How does Cygnus compare? Does it provide the > same capabilities? RSX seems especially easy to use (to me, at > least), and appears to have few bugs. What isn't available is > a g77 port, whereas I note that is is available under Cygnus. You can get the OS/2(emx) binaries of the g77-Port. They run under RSX as well. Long filenames are not supported on NT, because the binaries run in DOS-DPMI mode. You can recompile the sources (g77src.zip + gccsrc1..3.zip) with adding the -Zwin32 switch in order to get Win32-console Apps. Maybe i'll provide these binaires in the next release of g77. Because of my experiences with emx (under OS/2), i would prefer the RSX-compiler. Nevertheless, cygnus seems to be the mainstream gcc implementation on Win32. > > Any information greatly appreciated. > > Allen Cogbill > Los Alamos National Lab > ahc@lanl.gov > - > For help on using this list (especially unsubscribing), send a message to > "gnu-win32-request@cygnus.com" with one line of text: "help". -- Michael Holzapfel -------------------------------------------------------- Institut fuer Maschinenelemente und Maschinengestaltung RWTH-Aachen Tel. 0049-(0)-241-805661 Fax 0049-(0)-241-8888256 Email michael@ime.rwth-aachen.de WWW http://www.ime.rwth-aachen.de/~michael/index.html -------------------------------------------------------- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From grhgkim@tiger.korea.ac.kr Mon Aug 4 05:25:00 1997 From: grhgkim@tiger.korea.ac.kr (HgKim) Date: Mon, 04 Aug 1997 05:25:00 -0000 Subject: lorder & tsort Message-ID: <01BCA11C.CDE00C60@Hgkim.korea.ac.kr> I can't find command 'lorder' and 'tsort'. How can I get these command for windows95? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jont@harlequin.co.uk Mon Aug 4 08:12:00 1997 From: jont@harlequin.co.uk (Jon Thackray) Date: Mon, 04 Aug 1997 08:12:00 -0000 Subject: dllimport for gcc Message-ID: <199708041507.QAA28158@zaphod.long.harlequin.co.uk> Is this going to be implemented for the win32 in beta-19? Is there some other way of achieving the same effect? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From marcus@bighorn.dr.lucent.com Mon Aug 4 08:25:00 1997 From: marcus@bighorn.dr.lucent.com (marcus@bighorn.dr.lucent.com) Date: Mon, 04 Aug 1997 08:25:00 -0000 Subject: .tar.gz Message-ID: <199708041525.JAA00273@chorus.dr.lucent.com> Brian David Womack wrote: > Very good point. I've felt that this debate about gzip being > non-portable was bogus as well. Every UNIX and PC I've been > using has had a cross-platform compile of gzip. Never had > a problem. I had thought that the issue wasn't the use of gzip/gunzip itself, but the -z flag to tar, which causes gzip/gunzip to be invoked if using gnu tar, but may do other things on other versions of tar. I don't think that anybody has questioned gzipping archives as being non-portable, just how to handle them. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jman@lx.net Mon Aug 4 09:43:00 1997 From: jman@lx.net (jman) Date: Mon, 04 Aug 1997 09:43:00 -0000 Subject: latest Sergey's update (08/01) crashes on Win95 References: <9708040956.AA28128@hagar.aspentec.com> Message-ID: <3.0.2.32.19970804105544.00842a90@lx.net> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Let me guess your on win95... I have the same problem just mailed the list bout it. I can't figure it out. Something to do with ls dircolors an bash I think? If you run ls in dos instead of bash mine dies too. I had to downgrade to a previous coolview I had laying around... At 11:55 AM 8/4/97 +0200, Jerome Gay wrote: > >Hi all, > >I just installed Sergey's latest coolview, dated Aug,1st... when I start >bash, WIndows95 crahes completely (hard-reset necessary)... any ideas ? > >Jerome > > > >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > -----BEGIN PGP SIGNATURE----- Version: PGP for Personal Privacy 5.0 Charset: noconv iQA/AwUBM+X7fr6leZrIz2AbEQIKwACgokq0KOX2i12noEjLY5yFm0LlLkgAoKxe ioEra710R0GPEmqySLHuPnaC =1gjo -----END PGP SIGNATURE----- ------------------------------------------------------- Jason L. Esman aka _Jman Owner Den Internet Services System Admin. Network Consulting http://www.deninc.com | irc.deninc.com, irc.lx.net Email jman@lx.net or root@lx.net Finger jman@lx.net for PGP Public Keys... ------------------------------------------------------- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gthall@icg.tu-graz.ac.at Mon Aug 4 09:43:00 1997 From: gthall@icg.tu-graz.ac.at (Gerhard G. Thallinger) Date: Mon, 04 Aug 1997 09:43:00 -0000 Subject: High resolution clock References: Message-ID: On Fri, 25 Jul 1997, Joerg Lepler wrote: > > Hi, > > Does anyone know about such clock functions for the win32 environment? > There is the (undocumented) function NtQueryPerformanceCounter which returns a 64 bit counter value and the frequency of the counter update. I use this function on a Everex Dual Pentium Pro with 200 MHz. On this computer the call to this function takes about 14 microseconds and the counter has a resolution of 5 ns (= 1 / 200 MHz). On single processor machines calls may take significantly longer and the resolution is much less, so your mileage may vary. See the recent discussion on comp.os.ms-windows.programmer.nt.kernel-mode for details. I attach an example how to use this function. BTW: Which functions did you use on SUN and SGI ? HTH Gerhard G. Thallinger Vexcel Imaging GmbH ---------------------------Start of program profile.c ------------------------- /* * This code is supplied 'as is', no warranty whatsoever implied. * Use at your own risk. */ #include #include #include typedef LONG NTSTATUS; typedef NTSTATUS WINBASEAPI (WINAPI *NtQPC)(LARGE_INTEGER* pCounter, LARGE_INTEGER* pFrequency); #undef __FUNC__ #define __FUNC__ "GetFuncAddress()" NtQPC GetFuncAddress() { NtQPC pTemp = NULL; NTSTATUS result = 0; int loadLib = 0; HMODULE hNtDll = NULL; hNtDll = GetModuleHandle("ntdll"); if (hNtDll == NULL) { hNtDll = LoadLibrary("ntdll"); loadLib = 1; } if (hNtDll != NULL) { pTemp = (NtQPC) GetProcAddress(hNtDll, "NtQueryPerformanceCounter"); if (loadLib != 0) { FreeLibrary(hNtDll); } } return(pTemp); } #undef __FUNC__ #define __FUNC__ "main()" main(int argc, char* argv[]) { NTSTATUS result = 0; int iLoop; LARGE_INTEGER llCount; LARGE_INTEGER llFrequency; NtQPC NtQueryPerformanceCounter = NULL; NtQueryPerformanceCounter = GetFuncAddress(); if (NtQueryPerformanceCounter != NULL) { for (iLoop = 0; iLoop < 1000000;iLoop++) { result = NtQueryPerformanceCounter(&llCount,&llFrequency); } } else { fprintf(stderr,"Could not determine function address"); } return(0); } ---------------------------End of program profile.c ------------------------- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mdash@novell.com Mon Aug 4 10:11:00 1997 From: mdash@novell.com (Michael Scheer) Date: Mon, 04 Aug 1997 10:11:00 -0000 Subject: cdk.exe install problem Message-ID: I downloaded the beta18 cdk.exe. The installation fails early on with the following error from installshield: "The decompression of %s failed. There may no be enough free disk space in the TEMP directory." But I've got over 700 meg free in the volume holding $TEMP. Ideas? Thanks. --Mike Scheer - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ian@cygnus.com Mon Aug 4 10:31:00 1997 From: ian@cygnus.com (ian@cygnus.com) Date: Mon, 04 Aug 1997 10:31:00 -0000 Subject: Where are Reg* implemented? References: <33E2589B.D48D8ED8@smva.com> Message-ID: <199708041730.NAA26400@tweedledumb.cygnus.com> >It is also necessary to note that AIX does not implement ld in a >"normal" manner -- it will search all object modules in a circular >fashion, left to right and back to the front, until all references are >resolved (or a complete pass does not resolve a symbol). Typical >linkers do one pass left to right to resolve references -- each object >is evaluated once for unresolved external references, which must be >resolved by an object to its right on the link line. In this case it >would not make sense to include the library before the main object, as >no portion of the library would be incorporated in to the output, >because no external references exist when it is scanned. Technically, what AIX does is include the entire archive in the link. It then uses garbage collection--a feature of the AIX linker--to discard contents of the archive which are not needed. The real point, as I see it, is that you can't really argue about the behaviour of the GNU linker based on AIX. The GNU linker is a standard Unix linker. The AIX linker is not. Ian - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From lfm@pgroup.com Mon Aug 4 12:01:00 1997 From: lfm@pgroup.com (Larry Meadows) Date: Mon, 04 Aug 1997 12:01:00 -0000 Subject: stack size issues, multi-threading Message-ID: <199708041900.MAA23239@pacific.pgroup.com> I apologize if this is an FAQ, I looked... 1. How do I increase the stack size for an application? Is this done at link time, or can it be done at run time. 2. How are stacks allocated for multi-threaded applications. tia lfm - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From david@coent.demon.co.uk Mon Aug 4 14:17:00 1997 From: david@coent.demon.co.uk (David Coe) Date: Mon, 04 Aug 1997 14:17:00 -0000 Subject: Successful port of rcs-5.7 to gnu-win32 b18 References: Message-ID: <33E60146.5054@coent.demon.co.uk> Billinghurst, David wrote: > > I have ported rcs-5.7 to gnu-win32 b18. > > Build procedure > > 1. Apply patches below to rcs-5.7 source tree I needed the following additional patch from Lassi Tuura to get past the mmap stage in make: *** src/conf.sh.orig Fri Jun 16 07:19:23 1995 --- src/conf.sh Mon Aug 04 16:59:17 1997 *************** *** 1025,1030 **** --- 1025,1034 ---- # Early editions of SunOS 5.4 are reported to have problems with mmap # that generate NUL bytes in RCS files with a Solaris 2.2 NFS server. # This has been reported to be fixed as of patch 101945-10. + 'CYGWIN32'*) + # GNU win32 does not permit mmapped file to be opened (returns with + # "not owner"). So which is signal rcs gets is immaterial. + has_map_fd=0 has_mmap=1 has_madvise=0 mmap_signal=SIGBUS ;; *) $PREPARE_CC || exit if ($CL -DTRY_MAP_FD=1 a.c $L && $aout &2 -- Dr David Coe \=\ 58 Fairlawn Drive, East Grinstead \=\ Tel +44 1342 326860 West Sussex, RH19 1NT, United Kingdom \=\ Fax +44 1342 316019 - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mdash@novell.com Mon Aug 4 14:17:00 1997 From: mdash@novell.com (Michael Scheer) Date: Mon, 04 Aug 1997 14:17:00 -0000 Subject: cdk.exe install problem Message-ID: >I had the same problem, and got a response from somebody that helped me. >There is a bug in the installation. It seems that it fails when there are >too many files in the TEMP directory, not just when it's out of space. Try >removing some files from your TEMP directory. It worked for me. > > Hope this help, > Robert Seeger Thanks. That seems to have done it, and I did not have that much sitting around in $TEMP. Strange. --Mike Scheer - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Robert_Seeger@baynetworks.com Mon Aug 4 14:17:00 1997 From: Robert_Seeger@baynetworks.com (Robert Seeger) Date: Mon, 04 Aug 1997 14:17:00 -0000 Subject: cdk.exe install problem Message-ID: <3.0.32.19970804151739.009265c0@bl-mail2.corpeast.baynetworks.com> I had the same problem, and got a response from somebody that helped me. There is a bug in the installation. It seems that it fails when there are too many files in the TEMP directory, not just when it's out of space. Try removing some files from your TEMP directory. It worked for me. Hope this help, Robert Seeger At 11:10 AM 8/4/97 -0600, Michael Scheer wrote: >I downloaded the beta18 cdk.exe. The installation fails early on with >the following error from installshield: > >"The decompression of %s failed. There may no be enough free disk >space in the TEMP directory." > >But I've got over 700 meg free in the volume holding $TEMP. Ideas? >Thanks. > >--Mike Scheer > > >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > ======================================== Robert Seeger Network Engineer Bay Networks Telephone: (518)237-2087 Pager: (800)SKY-8888 Pin#1264792 Fax: (518)237-4190 Email: rseeger@baynetworks.com Address: 224 5th Ave, Apt#2 Lansingburgh, NY 12182 ======================================== - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dahms@ifk20.mach.uni-karlsruhe.de Mon Aug 4 17:28:00 1997 From: dahms@ifk20.mach.uni-karlsruhe.de (dahms@ifk20.mach.uni-karlsruhe.de) Date: Mon, 04 Aug 1997 17:28:00 -0000 Subject: Revised licensing terms Message-ID: <009B84D2.0C35C240.20684@ifk20.mach.uni-karlsruhe.de> Hi Gordon, you wrote: : under the terms of the GPL would remain free of charge. (If we decide : to charge fees for commercial use, they would probably be in line with : the fees we charge for our commercial GNUPro compilers: in the range : $2-3k for a single developer, $7-15k for up to five developers, and : $25-50k for up to twenty five developers). If you have any thoughts : on our charging fees for the commercial use of Cygwin32, please feel : free to email me. AFAIK OpenNT is cheaper. I hope that commercial use for non-competitors remains free. The previous license fitted well for us, in contrast to GPL only. Bye, Heribert (dahms@ifk20.mach.uni-karlsruhe.de) - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From p.west@mailbox.uq.edu.au Mon Aug 4 19:26:00 1997 From: p.west@mailbox.uq.edu.au (Peter B. West) Date: Mon, 04 Aug 1997 19:26:00 -0000 Subject: Ineradicable file %_ in W95 Message-ID: <3.0.2.32.19970805122636.00699e4c@dingo.cc.uq.edu.au> Dear subscribers, After using the GNU tools, bash and vim (W95 version), I have a file called "%_" (no quotes) in one of my directories. I can't get rid of it. I don't know whether it was created by a GNU process or by Turbo Pascal for DOS (yes, I'm afraid I was using it.) If anyone has seen anything similar while using gnuw32, or if anyone can suggest a way of cleaning it up, I would be most grateful. Your faithfully, Peter __ /__ Peter B. West / p.west@mailbox.uq.edu.au / "Master, to whom shall we go?" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From p.west@mailbox.uq.edu.au Mon Aug 4 19:36:00 1997 From: p.west@mailbox.uq.edu.au (Peter B. West) Date: Mon, 04 Aug 1997 19:36:00 -0000 Subject: Starting bash from Explorer Message-ID: <3.0.2.32.19970805123718.00699e4c@dingo.cc.uq.edu.au> One and all, My apologies if this has been mentioned before, but I haven't seen it in a few months. I have modified the file Doshere.inf, part of the power toys package from MS. Doshere installs in the right-button menu of Explorer, and offers to open a Dos box in the specified directory. bashhere.inf makes a similar offer of a bash shell. I find it useful. Your faithfully, Peter ; "bash prompt here" by Peter West ; Based on:- ; "Command Prompt Here" PowerToy ; ; Parts Copyright 1996 Microsoft Corporation ; ; NOTES: ; This setup is currently working under Windows 95. My configuration includes ; Service Pack 1 and a number of recent fixes downloaded from MS. Mail me if ; you need more info on this. ; ; This file is based directly on the PowerToys doshere.inf, which installs ; entries in the registry to allow startup, via right-clicking on a directory ; in Explorer, of a console window rooted in the selected directory. After ; installation of doshere.inf, the menu entry - ; Command _Prompt Here ; appears in the right-click menu. ; ; In this case (bashhere.inf), the menu item is ; bash shel_L here ; ; I experienced some frustration in trying to get the "cd" to work ; properly, until I decided to experiment from a bare-bones bash console. I ; set the command entry to what you see below: ; %BashWinBin%\bash.exe ; and it worked! By default, the bash window was cd'd to the selected directory. ; Another bonus is that long file names are preserved. ; ; *** REMEMBER to define %BashWinBin% appropriately for your system *** ; ; **WARNING* I have not attempted to do this in NT (I don't have NT), so I have ; NO idea whether it will work. ; ; SETUP: ; From wherever you download the file, right-click on the file in Explorer, ; and select Install. The new entry will appear in the Explorer right-click ; menu list when you select a drive or directory. ; ; This .inf file is copied to the \windows\inf directory, and entries, as detailed ; below, are made in the registry. ; ; UNINSTALLING: ; Use Control Panel|Add/Remove Programs ; [version] signature="$CHICAGO$" [BashHereInstall] CopyFiles = BashHere.Files.Inf AddReg = BashHere.Reg [DefaultInstall] CopyFiles = BashHere.Files.Inf AddReg = BashHere.Reg ; WARNING - untested [DefaultInstall.ntx86] CopyFiles = BashHere.Files.Inf AddReg = BashHere.Reg.NT [DefaultUnInstall] DelFiles = BashHere.Files.Inf DelReg = BashHere.Reg ; WARNING - untested [DefaultUnInstall.ntx86] DelFiles = BashHere.Files.Inf DelReg = BashHere.Reg.NT [SourceDisksNames] 55="BASH Prompt Here","",1 [SourceDisksFiles] BASHHERE.INF=55 [DestinationDirs] BashHere.Files.Inf = 17 [BashHere.Files.Inf] BASHHERE.INF [BashHere.Reg] HKLM,%UDHERE% HKLM,%UDHERE%,DisplayName,,"%BashHereName%" HKLM,%UDHERE%,UninstallString,,"%10%\rundll.exe setupx.dll,InstallHinfSection DefaultUninstall 132 %17%\BashHere.inf" HKCR,Directory\Shell\BashHere,,,"%BashHereAccel%" HKCR,Directory\Shell\BashHere\command,,,"%BashWinBin%\bash.exe" HKCR,Drive\Shell\BashHere,,,"%BashHereAccel%" HKCR,Drive\Shell\BashHere\command,,,"%BashWinBin%\bash.exe" ; WARNING -untested [BashHere.Reg.NT] HKLM,%UDHERE% HKLM,%UDHERE%,DisplayName,,"%BashHereName%" HKLM,%UDHERE%,UninstallString,,"rundll32.exe syssetup.dll,SetupInfObjectInstallAction DefaultUninstall 132 %17%\BashHere.inf" HKCR,Directory\Shell\BashHere,,,"%BashHereAccel%" HKCR,Directory\Shell\BashHere\command,,,"%BashWinBin%\bash.exe" HKCR,Drive\Shell\BashHere,,,"%BashHereAccel%" HKCR,Drive\Shell\BashHere\command,,,"%BashWinBin%\bash.exe" [Strings] BashHereName="bash prompt here Power Toy" BashHereAccel="bash shel&L here" UDHERE="Software\Microsoft\Windows\CurrentVersion\Uninstall\BashHere" ; ***** N.B. ***** Remember to redefine this for your system BashWinBin="E:\gnu-w32\H-i386-cygwin32\bin" From sos@prospect.com.ru Mon Aug 4 21:02:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Mon, 04 Aug 1997 21:02:00 -0000 Subject: latest Sergey's update (08/01) crashes on Win95 Message-ID: <01BCA176.1E192AF0@sos> Jerome Gay wrote: > I just installed Sergey's latest coolview, dated Aug,1st... when I start > bash, WIndows95 crahes completely (hard-reset necessary)... any ideas ? Do you have a valid /etc/passwd and /etc/group? -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From o@globeset.com Mon Aug 4 21:02:00 1997 From: o@globeset.com (Steve o) Date: Mon, 04 Aug 1997 21:02:00 -0000 Subject: an x environment Message-ID: <33E6A36C.36D4@io.com> Hi, Having recently been transplanted from an X environment to winnt, I am homesick and wish to recreate my old desktop. My plan was enabled when Sergey ported rxvt, well, not just ported but added a much needed /dev/tty interface. wow. fvwm2 ported without a hitch. xemacs was quite a different story. I managed to get temacs.exe to compile, but only after I hacked out what seemed like a very useless, and very system dependent code, called unexec.c. Turns out, to my surprise, temacs reads itself and dumps out an executable copy, all via unexec. hmm. I poked around, looked at what emacs-19.34 did, tried a couple things, but really, I was out of my leaque. Are .exe files in COFF format, or elf? What happens in crt0.o that has to be accounted for? Can I mix the Win32'isms in unexnt.c with the un-adulterated xemacs code? Maybe I could lift some .exe generating code from ld? So for now I've given up, but maybe by posting this, some guru will become intrigued and, while waiting for a compile, will patch unexec.c and casually mention it in an email to this list. As a side note, I used "configure i386-cygwin-linux" to get configure to work. This got me thinking that since the cygwin API is apporaching that of Linux, it may be possible to write a loader that could execute native linux binaries under cygwin, a la the WINE project only reversed. Imagine, the whole Red-hat distribution ported to cygwin, and no .exe extensions... -steve - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mka@redes.int.com.mx Mon Aug 4 22:09:00 1997 From: mka@redes.int.com.mx (Michael Anderson) Date: Mon, 04 Aug 1997 22:09:00 -0000 Subject: Revised licensing terms References: <009B84D2.0C35C240.20684@ifk20.mach.uni-karlsruhe.de> Message-ID: <33E6B5BF.18FB78B@redes.int.com.mx> dahms@ifk20.mach.uni-karlsruhe.de wrote: > I hope that commercial use for non-competitors remains free. > The previous license fitted well for us, in contrast to GPL only. I agree, the Version 1.0 license worked well for me as an independent consultant. The GPL is useless and fees as suggested would force me to other alternatives. -- Mike Anderson mka@redes.int.com.mx Guanajuato, GTO, Mexico - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From loki@dragoncat.net Mon Aug 4 22:10:00 1997 From: loki@dragoncat.net (Jeremy Blackman) Date: Mon, 04 Aug 1997 22:10:00 -0000 Subject: cdk.exe install problem References: Message-ID: On Mon, 4 Aug 1997, Michael Scheer wrote: > I downloaded the beta18 cdk.exe. The installation fails early on with > the following error from installshield: > > "The decompression of %s failed. There may no be enough free disk > space in the TEMP directory." > > But I've got over 700 meg free in the volume holding $TEMP. Ideas? > Thanks. I oughtta make a template for this sort of reply. ;) [deep breath, into rehearsed speech] Having had to build one too many InstallShield setups at my workplace, I can tell you that this is a bug in InstallShield, not a bug in Cygnus' installer. If you have more than a certain number of files in your TEMP directory - or if you have files named the same thing InstallShield wishes to name its files (probably from past runs of other InstallShield install scripts) which it cannot, for some reason, write over - InstallShield will fail. And not gracefully, either - this failure is indicated by that dialog box you mentioned. :) Clearing out your TEMP directory entirely should do it - that will get rid of both any files with conflicting names, and the potential 'too many files' problem. :) [out of rehearsed reply] Hope that helps! - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jeffdbREMOVETHIS@netzone.com Tue Aug 5 00:03:00 1997 From: jeffdbREMOVETHIS@netzone.com (Mikey) Date: Tue, 05 Aug 1997 00:03:00 -0000 Subject: stack size issues, multi-threading References: <199708041900.MAA23239@pacific.pgroup.com> Message-ID: <33e8cff5.6761684@smtp.netzone.com> gcc -Wl,--heap,1024,--stack,4096 -o blah blah.c On Mon, 4 Aug 1997 12:00:55 -0700 (PDT), you wrote: >I apologize if this is an FAQ, I looked... > >1. How do I increase the stack size for an application? Is this done >at link time, or can it be done at run time. > >2. How are stacks allocated for multi-threaded applications. > >tia > >lfm >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > (jeffdbREMOVETHIS@netzone.com) delete REMOVETHIS from the above to reply Mikey - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From yadda@msn.com Tue Aug 5 00:03:00 1997 From: yadda@msn.com (Christine Behrens) Date: Tue, 05 Aug 1997 00:03:00 -0000 Subject: About your compiler Message-ID: Hi, my name is Tony Philippsen. I have recently decided to delve into the world of computer programming. I have need of a C++ compiler for an online course in this language, and I was given your name as a source for a compiler. Unfortunately, my knowledge in this area is nothing less than pathetic, and I am completely confused by your download page (not your fault). The course requires a compiler that can "write to standard output", "it must support templates", and "must support exceptions". I was wondering if yours met all of these criteria. I read your "Read Me" file, and could not discern if your compiler met these requirements because I have no idea what they mean. I would appreciate some help in answering these questions if you have the time. Also, if this compiler ( I don't even know what a compiler really is!) meets these requirements, I sure could use any help you would be willing to give me regarding downloading the proper files to my computer. As I said, I was able to access your download area, but have no idea which files I need to insure a working compiler is installed at my end. Please Help, Tony Philippsen yadda@msn.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kuku@gilberto.physik.RWTH-Aachen.DE Tue Aug 5 01:34:00 1997 From: kuku@gilberto.physik.RWTH-Aachen.DE (Christoph Kukulies) Date: Tue, 05 Aug 1997 01:34:00 -0000 Subject: configure switches for cross building Message-ID: <199708050837.KAA10233@gil.physik.rwth-aachen.de> Could someone explain to me what the benefits of ommitting/disabling ./configure ..... --disable-libgloss --disable-gdb --disable-tcl --disable-tk are? What is the tk/tcl stuff for, anyway? -- Chris Christoph P. U. Kukulies kuku@gil.physik.rwth-aachen.de - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gay@aspentec.com Tue Aug 5 01:49:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Tue, 05 Aug 1997 01:49:00 -0000 Subject: Ineradicable file %_ in W95 Message-ID: <9708050849.AA38606@hagar.aspentec.com> Looks like the good ol' Unix 'how_to_remove_that_f***in_file'... try 'rm -i *' , for each file in the current directory, it will prompt you before removing it, just confirm only this %_ file... hope this helps... Jerome "ZoSo" -----Original Message----- From: Peter B. West To: gnu-win32@cygnus.com Date: Tuesday, August 05, 1997 10:44 AM Subject: Ineradicable file %_ in W95 >Dear subscribers, > >After using the GNU tools, bash and vim (W95 version), I have a file called >"%_" (no quotes) in one of my directories. I can't get rid of it. I don't >know whether it was created by a GNU process or by Turbo Pascal for DOS >(yes, I'm afraid I was using it.) > >If anyone has seen anything similar while using gnuw32, or if anyone can >suggest a way of cleaning it up, I would be most grateful. > >Your faithfully, >Peter > __ /__ Peter B. West > / p.west@mailbox.uq.edu.au > / "Master, to whom shall we go?" >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gay@aspentec.com Tue Aug 5 04:19:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Tue, 05 Aug 1997 04:19:00 -0000 Subject: Where to get GNU tools for Alpha/NT4 ? Message-ID: <9708051118.AA27161@hagar.aspentec.com> Is someone working on the port of GNU tools (bash...) for NT4 running on Alpha ? I'll support it actively... Jerome "ZoSo" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mdash@novell.com Tue Aug 5 05:57:00 1997 From: mdash@novell.com (Michael Scheer) Date: Tue, 05 Aug 1997 05:57:00 -0000 Subject: cdk.exe installed, new problem Message-ID: Thanks to all who got me past the installshield problem; cleaning out $TEMP did the trick. Now an attempt to compile a small C program results in gcc: Internal compiler error: program cpp got fata signal 5. Corrupted binary? I would think (hope?) that installshield did the right thing unpacking things. Other info: 1. I've installed on a drive other than C:, due to space problems. 2. I've created C:/tmp and C:/bin/sh. 3. I have not mounted anything anywhere. The use of mount is unclear from the faq, and I have found no additional doc. Thanks in advance for your help. --Mike Scheer - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kuku@gilberto.physik.RWTH-Aachen.DE Tue Aug 5 07:03:00 1997 From: kuku@gilberto.physik.RWTH-Aachen.DE (Christoph Kukulies) Date: Tue, 05 Aug 1997 07:03:00 -0000 Subject: need help on building a cross envmnt Message-ID: <199708051407.QAA11199@gil.physik.rwth-aachen.de> It's too weird: I'm still having a couple of stumbling blocks getting a gnu-win32 cross compiler built. (OS is FreeBSD 3.0-current), this is what I typed: ./configure -v --target=i386-cygwin32 --host=i386-unknown-freebsd3.0 I had to hand edit some glitches in texinfo/info/terminal.c: #if defined (HAVE_TERMIOS_H) || defined (HAVE_TERMIO_H) ttybuff.c_iflag &= (~ISTRIP & ~INLCR & ~IGNCR & ~ICRNL & ~IXON); #ifdef __FreeBSD__ <<<<< ttybuff.c_oflag &= (~ONLCR ); <<<<< #else <<<<< ttybuff.c_oflag &= (~ONLCR & ~OCRNL); #endif I don't like putting things like this inmidst the code and it should be better filtered out by appropriate config define's. The other was an inconsistent definition of sys_errlist in some source file. Now the biggest stumbling block seems to me this one (if not others will occur): making sub-all in emacs rm -f SYSCALLS.c tmp-SYSCALLS.s cat ./sys-types.h ./sys-protos.h > SYSCALLS.c /home/kuku/gnuwin32/cdk/gcc/xgcc -B/home/kuku/gnuwin32/cdk/gcc/ -DCROSS_COMPILE -DIN_GCC -g -I./include -I. -I. -I./config -aux-info SYSCALLS.c.X -S -o tmp-SYSCALLS.s SYSCALLS.c SYSCALLS.c:86: stddef.h: No such file or directory *** Error code 1 Stop. *** Error code 1 Stop. I see stddef in gcc/ginclude and in /usr/include (the latter not being relevant here, I think) but I wonder what's screwed in the cdk distribution such that it doesn't compile smoothly. -- Chris Christoph P. U. Kukulies kuku@gil.physik.rwth-aachen.de - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jelling@iitb.fhg.de Tue Aug 5 07:26:00 1997 From: jelling@iitb.fhg.de (Andreas Jellinghaus (bei rf)) Date: Tue, 05 Aug 1997 07:26:00 -0000 Subject: exception trapped in cygnus exception handler Message-ID: <19970805162551.29191@bsipc10.iitb.fhg.de> i'm using gnu win32 beta 18, and i failed to build a dll : ld : In cygwin_except_handler (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) Exception trapped! anyone a idea, what i can try now ? regards, andreas --- logfie : echo 'int main() { return 0; } ' > fixup.c echo 'asm (".section .idata\n" ".long 0,0,0,0,0,0,0,0"); ' >> fixup.c gcc -c -o fixup.o fixup.c echo EXPORTS > iitb.def nm dll.o \ |grep '^........ [T] _'\ |sed 's/[^_]*_//' >> iitb.def echo "step 1 - ld" step 1 - ld ld \ --dll \ --base-file iitb.base \ --output iitb.dll \ --entry _dll_entry@12 \ -L/usr/H-i386-cygwin32/i386-cygwin32/lib/ \ dll.o echo "step 2 - dlltool" step 2 - dlltool dlltool \ --as=as \ --base-file iitb.base \ --def iitb.def \ --dllname iitb.dll \ --output-exp iitb.exp echo "step 3 - ld (mit exp)" step 3 - ld (mit exp) ld \ --dll \ --entry _dll_entry@12 \ --output iitb.dll \ -L/usr/H-i386-cygwin32/i386-cygwin32/lib/ \ iitb.exp dll.o fixup.o -lcygwin -lkernel32 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) In cygwin_except_handler (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) Exception trapped! (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) exception C0000005 at 42D792 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) exception: ax 0 bx 2EBC068 cx 2EBCC9C dx 28 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) exception: si 4000 di FFFFFE00 bp 244EE50 sp 244EE40 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) exception is: STATUS_ACCESS_VIOLATION (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) Stack trace: (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 0: sp = 0x244EC74, pc = 0x1000CEC2 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 1: sp = 0x244EC90, pc = 0x77F94072 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 2: sp = 0x244ECB4, pc = 0x77F88A53 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 3: sp = 0x244ED40, pc = 0x77F75E82 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 4: sp = 0x244EE50, pc = 0x42D895 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 5: sp = 0x244EEA4, pc = 0x43094E (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 6: sp = 0x244F2BC, pc = 0x41FA39 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 7: sp = 0x244F314, pc = 0x416DCF (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 8: sp = 0x244F344, pc = 0x1000C102 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 9: sp = 0x244FF94, pc = 0x1000C113 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 10: sp = 0x244FFA0, pc = 0x4408A9 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 11: sp = 0x244FFB0, pc = 0x40103B (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 12: sp = 0x244FFC0, pc = 0x77F1AFC1 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) frame 13: sp = 0x244FFF0, pc = 0x0 (D:\GNUWIN32\H-i386-cygwin32\bin\ld.exe 1083) End of stack trace make: *** [dll] Error 2 - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From topcoder@mindspring.com Tue Aug 5 08:51:00 1997 From: topcoder@mindspring.com (Pam Greene) Date: Tue, 05 Aug 1997 08:51:00 -0000 Subject: Revised licensing terms References: <009B84D2.0C35C240.20684@ifk20.mach.uni-karlsruhe.de> <33E6B5BF.18FB78B@redes.int.com.mx> Message-ID: <3.0.3.32.19970804234559.00698608@mail.mindspring.com> considering my interest is in game programming and game related applications, i dont even see where i would BE a competitor. Aside from that, i am only 13, i could afford that price even if i WAS a competitor!!! At 12:10 AM 8/5/97 -0500, you wrote: >dahms@ifk20.mach.uni-karlsruhe.de wrote: > >> I hope that commercial use for non-competitors remains free. >> The previous license fitted well for us, in contrast to GPL only. > >I agree, the Version 1.0 license worked well for me as an independent >consultant. The GPL is useless and fees as suggested would force me to >other alternatives. > >-- > >Mike Anderson >mka@redes.int.com.mx >Guanajuato, GTO, Mexico >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jelling@iitb.fhg.de Tue Aug 5 09:27:00 1997 From: jelling@iitb.fhg.de (Andreas Jellinghaus (bei rf)) Date: Tue, 05 Aug 1997 09:27:00 -0000 Subject: gnu-win32 beta18 did not compile on a linux system Message-ID: <19970805182645.28480@bsipc10.iitb.fhg.de> hi. i was running configure with --prefix=$HOME/gnu --target=i386-unknown-cygwin32 i had to do this, to get compile and install done : cp install-sh i386-unknown-cygwin32 ln -s "../config.sub" i386-unknown-cygwin32/ ln -s "../config.sub" i386-unknown-cygwin32/winsup cp gcc/getopt.h winsup cp gcc/getopt.h i386-unknown-cygwin32/winsup i hope this is enough to locate the bugs. if not, i have logfiles of all configure, and make calls here. note : i downloaded the cdk source from ftp.sunet.se. there are md5sums of xc[a-h], but the source is xc[a-j], and has different md5sums. i assumed, that the md5sum file is not up-to-date... regards, andreas - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bub@io.com Tue Aug 5 11:23:00 1997 From: bub@io.com (Steve o) Date: Tue, 05 Aug 1997 11:23:00 -0000 Subject: an x environment References: <33E6A36C.36D4@io.com> <33E754EE.46094188@redes.int.com.mx> Message-ID: <33E76B9D.60B1@io.com> Michael Anderson wrote: > > Steve o wrote: > > > > fvwm2 ported without a hitch. > > Hi Steve: > > I've been using bash windows and vi for my work and wondering just how I > might get X and fvwm working. Can you tell me the steps needed to do > this? Which X server are you using and what other software needs to be > installed? Config files? If you posted the process to the newsgroup, it > might help out others wondering how to do this. I am using Hummingbird's Exceed X Server, http://www.hummingbird.com/products/exceed/ I set the environment variable DISPLAY=:0 I installed the "X client libs for NT" from http://www.cygnus.com/misc/gnu-win32/ Fvwm needs xpm so I got that from: ftp://ftp.x.org/contrib/libraries/xpm-3.4j.tar.gz Fvwm can be obtained from: http://www.hpc.uh.edu/fvwm/ The steps to install are: tar -xvzf package.tar.gz cd package xmkmf -a make make install Sometimes the install target in the Makefile will have: cp foo /usr/local/bin/foo These need to be changed to: cp foo.exe /usr/local/bin/foo.exe There are some sample .fvwm2rc files included in the distribution, copy one of these to your home directory. If you run fvwm, you'll want to set the Exceed Xconfig->Window Mode->window manager to X. You'll need Sergey's coolview and remote.tar.gz. -steve - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From leisner@sdsp.mc.xerox.com Tue Aug 5 11:23:00 1997 From: leisner@sdsp.mc.xerox.com (Marty Leisner) Date: Tue, 05 Aug 1997 11:23:00 -0000 Subject: Porting from DJGPP code References: <33E0CF88.7BE9@inter.net.il> Message-ID: <9708051803.AA16122@gnu.sdsp.mc.xerox.com> > Dear Sirs > > Most of my source codes are in DJGPP, and I was wondering > whether it is possible to port these to GNU-Win32, and how > would you rate the effort needed. > > Thanks > > Yoav > - > For help on using this list (especially unsubscribing), send a message to > "gnu-win32-request@cygnus.com" with one line of text: "help". If you try to stay with ansi/posix style code, you should not have much of a problem...but if you start using dos-isms... It probably is as hard to port to win32 as some other unix platform. It many cases if code runs on unix platforms, its easy to port to win32 and djgpp. -- marty leisner@sdsp.mc.xerox.com The Feynman problem solving Algorithm 1) Write down the problem 2) Think real hard 3) Write down the answer Murray Gel-mann in the NY Times - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jeffdbREMOVETHIS@netzone.com Tue Aug 5 13:15:00 1997 From: jeffdbREMOVETHIS@netzone.com (Mikey) Date: Tue, 05 Aug 1997 13:15:00 -0000 Subject: stack size issues, multi-threading References: <199708051550.IAA01312@pacific.pgroup.com> Message-ID: <33e788e4.2210376@smtp.netzone.com> .../cdk/ld/emultempl/pe.em The obvious place ;^). Pointed out by Sergey about 2 months ago. On Tue, 5 Aug 1997 08:50:00 -0700 (PDT), you wrote: >> gcc -Wl,--heap,1024,--stack,4096 -o blah blah.c > >Thanks. Boy, I don't even see that in the source code, much less >the documentation. > >lfm > (jeffdbREMOVETHIS@netzone.com) delete REMOVETHIS from the above to reply Mikey - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From uyar@athene.informatik.uni-bonn.de Tue Aug 5 14:46:00 1997 From: uyar@athene.informatik.uni-bonn.de (sed) Date: Tue, 05 Aug 1997 14:46:00 -0000 Subject: basic questions Message-ID: <33E79F36.7B99@athene.informatik.uni-bonn.de> Hi, I've some Question : Is it possible to use b18 on nt3.5 and if its possible which functionality I would miss ? Why isn't it possible to use a dll, which was imported with dlltool, without explicit linking of libcygwin.a ? Without linking libcygwin.a I get no error at compile-time, but an exception at run-time. Can someone precisly explain how to build a correct library ,,libfoo.a" for a DLL ,,foo.dll", which only exists in binary form ? (please all the steps) Which options should be provided later to gcc for the prog which will use functions inside of this DLL ? Thanks in advance !! - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From smithj@pobox.com Tue Aug 5 14:58:00 1997 From: smithj@pobox.com (Jeffery G. Smith) Date: Tue, 05 Aug 1997 14:58:00 -0000 Subject: lesstif 0.79 Message-ID: <33E7A15B.C7@pobox.com> I swear that someone had posted that lesstif had been ported under b18. I was unable to find a reference in the list archives or any info on the ported software link at Cygnus. If anyone knows about this, please send me whatever info you have. Everything went reasonably well once I got the X11R6.3 stuff and until I hit the yacc code. byacc accepts the .y file fine but then gcc barfs with gcc -I. -I../../../include -I./../../../include/Motif-1.2 -I./../../../lib/Mrm -I/usr/local/gnuwin32/X11R6.3/include -g -DLESSTIF_VERBOSE -Wall -ansi -pedantic -c -o uilparse.o uilparse.c y.tab.c: In function `yymalloc': y.tab.c:534: warning: implicit declaration of function `malloc' y.tab.c: In function `yyrealloc': y.tab.c:543: warning: implicit declaration of function `realloc' y.tab.c: At top level: y.tab.c:553: syntax error before `yygrow' y.tab.c:559: syntax error before `if' yacc.y:35: warning: `False' defined but not used yacc.y:36: warning: `True' defined but not used y.tab.c:525: warning: `yyss' defined but not used y.tab.c:526: warning: `yyvs' defined but not used y.tab.c:527: warning: `yystacksize' defined but not used y.tab.c:533: warning: `yymalloc' defined but not used y.tab.c:542: warning: `yyrealloc' defined but not used gcc: Internal compiler error: program cc1 got fatal signal 33 make: *** [uilparse.o] Error 1 - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gm-goodwin@worldnet.att.net Tue Aug 5 18:36:00 1997 From: gm-goodwin@worldnet.att.net (Gregory Goodwin) Date: Tue, 05 Aug 1997 18:36:00 -0000 Subject: No readme included in b18 Message-ID: <3.0.3.32.19970805203429.007aec60@postoffice.worldnet.att.net> I just downloaded/installed the full binary installation of build 18 for Win95. The install program mentioned that I would need to view the readme to set up the environment. However, the readme was not installed. Nor was there a copy of the License Agreement. Thought you should know. Regards, Greg - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mphalpin@iag.net Tue Aug 5 19:43:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Tue, 05 Aug 1997 19:43:00 -0000 Subject: gnu-win32 beta18 did not compile on a linux system Message-ID: <199708060233.WAA29587@iag.net> Why are you compiled Gnuwin32 on a Linux system? Gnuwin32 is made for Windows systems. -SJ ---------- > From: Andreas Jellinghaus (bei rf) > To: gnu-win32@cygnus.com > Subject: gnu-win32 beta18 did not compile on a linux system > Date: Tuesday, August 05, 1997 12:26 PM > > hi. > > i was running configure with --prefix=$HOME/gnu --target=i386-unknown-cygwin32 > i had to do this, to get compile and install done : > cp install-sh i386-unknown-cygwin32 > ln -s "../config.sub" i386-unknown-cygwin32/ > ln -s "../config.sub" i386-unknown-cygwin32/winsup > cp gcc/getopt.h winsup > cp gcc/getopt.h i386-unknown-cygwin32/winsup > > i hope this is enough to locate the bugs. if not, i have logfiles of > all configure, and make calls here. > > note : i downloaded the cdk source from ftp.sunet.se. > there are md5sums of xc[a-h], but the source is xc[a-j], and has > different md5sums. i assumed, that the md5sum file is not up-to-date... > > regards, andreas > > - > For help on using this list (especially unsubscribing), send a message to > "gnu-win32-request@cygnus.com" with one line of text: "help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From rhino@tfs.net Tue Aug 5 20:13:00 1997 From: rhino@tfs.net (Stephen P. Green) Date: Tue, 05 Aug 1997 20:13:00 -0000 Subject: lesstif 0.79 References: <33E7A15B.C7@pobox.com> Message-ID: <33E7EC20.80FE60BB@tfs.net> Jeffery G. Smith wrote: > I swear that someone had posted that lesstif had been ported under > b18. > I was unable to find a reference in the list archives or any info on > the > ported software link at Cygnus. > > If anyone knows about this, please send me whatever info you have. > Everything went reasonably well once I got the X11R6.3 stuff and until > I > hit the yacc code. > > byacc accepts the .y file fine but then gcc barfs with > > gcc -I. -I../../../include -I./../../../include/Motif-1.2 > -I./../../../lib/Mrm > -I/usr/local/gnuwin32/X11R6.3/include > -g -DLESSTIF_VERBOSE -Wall -ansi -pedantic -c -o uilparse.o > uilparse.c > y.tab.c: In function `yymalloc': > y.tab.c:534: warning: implicit declaration of function `malloc' > y.tab.c: In function `yyrealloc': > y.tab.c:543: warning: implicit declaration of function `realloc' > y.tab.c: At top level: > y.tab.c:553: syntax error before `yygrow' > y.tab.c:559: syntax error before `if' > yacc.y:35: warning: `False' defined but not used > yacc.y:36: warning: `True' defined but not used > y.tab.c:525: warning: `yyss' defined but not used > y.tab.c:526: warning: `yyvs' defined but not used > y.tab.c:527: warning: `yystacksize' defined but not used > y.tab.c:533: warning: `yymalloc' defined but not used > y.tab.c:542: warning: `yyrealloc' defined but not used > gcc: Internal compiler error: program cc1 got fatal signal 33 > make: *** [uilparse.o] Error 1 > - > For help on using this list (especially unsubscribing), send a message > to > "gnu-win32-request@cygnus.com" with one line of text: "help". I don't ever remember any ports of LessTif yet, but if anyone is willing to try it I will gladly and actively support it. I've been need this done anyway for quite some time. If anyone has already done this or is working on it, please let me know. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From iverson@cisco.com Tue Aug 5 20:27:00 1997 From: iverson@cisco.com (Tim Iverson) Date: Tue, 05 Aug 1997 20:27:00 -0000 Subject: Revised licensing terms References: <33E6B5BF.18FB78B@redes.int.com.mx> Message-ID: <199708060314.UAA09111@stars.cisco.com> IMHO, Cygnus certainly deserves to be paid by anyone wanting to sell a commercial product containing their work (ie. the DLL). If you can't make a profit on your product with their DLL licensing fees, then obviously you are not adding enough value. However, the level of support provided by this beta group goes far beyond what anyone would expect from any group not directly involved in development. That deserves more than a little consideration, too. I suspect this support would vanish should Cygnus decide to release only commercially (ie. you pay, or you can't use it). Releasing also under the GPL should satisfy just about everyone else. I was surprised when Cygnus called embedded developers competition in the previously proposed license. I'd say embedded folks are more likely to be customers (Cisco certainly is), not competition. Compiler and tool developers are competition. Personally, I am pleased that the GPL will be offered. This would allow my group within Cisco to use (and fix) the GCC/W95 tool suite in-house until it has proven itself worthy of purchasing a support contract from Cygnus. It also ensures propagation of bug-fixes, something that improves quality across our entire industry. - Tim Iverson iverson@cisco.com +---------------- | Date: Tue, 05 Aug 1997 00:10:23 -0500 | From: Michael Anderson | To: gnu-win32@cygnus.com | Subject: Re: Revised licensing terms | | dahms@ifk20.mach.uni-karlsruhe.de wrote: | > I hope that commercial use for non-competitors remains free. | > The previous license fitted well for us, in contrast to GPL only. | | I agree, the Version 1.0 license worked well for me as an independent | consultant. The GPL is useless and fees as suggested would force me to | other alternatives. | | -- | | Mike Anderson | mka@redes.int.com.mx | Guanajuato, GTO, Mexico - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From rhino@tfs.net Tue Aug 5 20:27:00 1997 From: rhino@tfs.net (Stephen P. Green) Date: Tue, 05 Aug 1997 20:27:00 -0000 Subject: Assembler Message-ID: <33E7EF73.121AB62E@tfs.net> I am interested in learning assembly language. I know this is almost a completely dead language, but I find it intringing. I have two questions: First is their any assembler included in the b18 release, or any unix gnu assembler that could be ported(please include a link if it has already been done or any other programs that might serve the same purpose) and Secondly, any literature that you could point me to as a good starting point. I know that this request is a little bit of a deviation from our normal topic, but I beg you to show me a little patience and help me. Thanks to you all... - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Tue Aug 5 21:17:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Tue, 05 Aug 1997 21:17:00 -0000 Subject: lesstif 0.79 Message-ID: <01BCA241.62126D40@sos> Jeffery G. Smith wrote: > byacc accepts the .y file fine but then gcc barfs with > > gcc -I. -I../../../include -I./../../../include/Motif-1.2 > -I./../../../lib/Mrm > -I/usr/local/gnuwin32/X11R6.3/include > -g -DLESSTIF_VERBOSE -Wall -ansi -pedantic -c -o uilparse.o uilparse.c > y.tab.c: In function `yymalloc': > y.tab.c:534: warning: implicit declaration of function `malloc' > y.tab.c: In function `yyrealloc': > y.tab.c:543: warning: implicit declaration of function `realloc' > y.tab.c: At top level: > y.tab.c:553: syntax error before `yygrow' > y.tab.c:559: syntax error before `if' Gcc complains about "inline" keyword. Remove it, or remove -ansi and -pedantic command line options. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From colin@bird.fu.is.saga-u.ac.jp Tue Aug 5 21:17:00 1997 From: colin@bird.fu.is.saga-u.ac.jp (Colin Peters) Date: Tue, 05 Aug 1997 21:17:00 -0000 Subject: About your compiler Message-ID: <01BCA269.7CD179A0@gbird0> Note to mailing list: I only cc'd this to the list because I am not sure about some of my comments, and because I would like to hear any feedback anyone wishes to make about my "newbie guide" appendix included at the end of this mail. I apologize for the bandwidth usage. Christine Behrens[SMTP:yadda@msn.com] wrote: > Hi, my name is Tony Philippsen. I have recently decided to delve into the >world of computer programming. I have need of a C++ compiler for an online >course in this language, and I was given your name as a source for a compiler. > Unfortunately, my knowledge in this area is nothing less than pathetic, and I >am completely confused by your download page (not your fault). The course >requires a compiler that can "write to standard output", "it must support >templates", and "must support exceptions". I was wondering if yours met all >of these criteria. I read your "Read Me" file, and could not discern if your >compiler met these requirements because I have no idea what they mean. I >would appreciate some help in answering these questions if you have the time. >Also, if this compiler ( I don't even know what a compiler really is!) meets >these requirements, I sure could use any help you would be willing to give me >regarding downloading the proper files to my computer. As I said, I was able >to access your download area, but have no idea which files I need to insure a >working compiler is installed at my end. Hi there, First of all, note that the mailing list you sent to is read by both the Cygnus developers and a large number of active users. I'm a user, so I am not the final authority on anything at all. However, I will try to give you some hints based on my own knowledge. Tip number one is that you have your work cut out for you. The GNU compiler is not a model of user-friendliness by any stretch of the imagination. However, I believe it is worth learning how to use, so just try to be patient and persistent. About your needs (remember, I am NOT an authority. I have the strong feeling that someone will flame me for my lack of knowledge on points two and three.): - Must be able to write to standard output. I know of no C/C++ compiler on Earth that cannot do that. No problem. - Support templates. GCC supports templates. In fact it was one of the first compilers to do so. There are, however, still some odd bits in the implementation which may make it tricky to use. Still, it's probably no worse than any other C++ compiler, and you can always ask the mailing list or GCC newsgroups for help. Reading the documentation might be helpful as well. - Supports exceptions. GCC was also one of the first compilers to support exceptions. However, I am not sure myself exactly how to use them under GCC, and I don't know if they are entirely stable. You can always hope. Now, downloading the compiler itself: By following links or using an FTP program you should be able to get to the ...gnu-win32/gnu-win32-b18/ directory on Cygnus' FTP server or one of the mirrors. There is a list of mirrors on the Cygwin32 Project page and on my homepage ( http://www.geocities.com/Tokyo/Towers/6162/gcc.html ). In that directory you will find a file README.txt, which you should download, since it contains instructions for setting up the compiler, among other things. Then you can either download the extremely large "cdk.exe" file from that directory, or download all the files in the directory "cdk-split". Remember to download these files in BINARY mode. They should be the exact same size on your hard drive as shown in the directory listing in your browser. If you downloaded the files from cdk-split do the following at a DOS prompt (i.e. console window on Win NT or 95): copy /b xaa + xab + xac + ... + xag + xah cdk.exe (replace the ... with the other filenames in alphabetical order). Then, if that worked, you can delete xa*.*. Now run cdk.exe, that will start an installation program which should uncompress the compiler and copy it to the directory you want. After that you still have to edit your autoexec.bat (or system environment variables on NT) to set environment variables and the PATH for your system so that the compiler will work. See the README.txt file for details. Test out the compiler by compiling a simple program like "hello world". Use your favorite text editor (not a word processor, unless you make sure you save in ASCII text; notepad works fine, but vim or ntemacs is better) to create a file with these contents: #include int main () { printf("Hello, world.\n"); } And save it as "hello.c", then, in the directory where you saved it (again using a DOS-console) run the following command: gcc -o hello.exe hello.c This should produce a file hello.exe and when you type hello you should get the line: Hello, world. as your output. If it didn't work there is some problem with your configuration. Check whether you set up all the environment variables correctly (you must reboot for changes to take effect on Windows 95) and if you can't figure it out note the exact error messages and mail me or the mailing list. The remainder of this document is an appendix to the README for my own extensions to GNU-Win32, the Minimalist GNU-Win32, and it describes some of the options and other programs you will need to use to make your own programs. One more thing though: if you are compiling C++ you might want to use "g++" instead of "gcc" (I think the command line options are basically the same) because this automatically includes some extra libraries which your online course probably assumes are being included (specifically "iostreams"). Good luck, Colin. Appendix A: A Newbie's Guide to Using GNU-Win32 This is just a little section I threw together because I keep getting messages like "I just started learning C/C++ and I can't figure out how to use the compiler." Hopefully this will help cut down on those a bit. There is complete, if somewhat outdated, documentation for the GNU compiler and tools (especially of interest are the "binary utilities") both online at the Cygnus Technical Library and included with the cdk.exe download (unfortunately in man file format). This section is just an attempt to get you started. A.1 Definitions First of all here are the definitions of a few terms you need to understand in order to get anything out of the rest of the discussion. Source code (or "source") - Text written, generally, by a human being containing one or more statements, definitions, expressions or functions written in a programming language, such as C or C++. Source code also refers to a text file containing such text. You can usually tell what language a source file is by the file name. If it ends in (i.e. it's "extension" is) ".c" then it is C language code, if the extension is ".cpp" then it is C++. Object code - Series of instructions and/or data in a form which could be executed by a computer, but not necessarily a complete program. Object code is stored in files that end with ".o" (or sometimes ".obj", but that is not the GNU convention). Usually one creates a single object file to match each source file by compiling the source files (see below). Object files are not complete programs because they can contain calls to functions or data not defined in the object file and they may have no single "starting point" where the program could begin. Compiler - A program (or set of programs) that converts source code into object code, generally creating one object file for each source file input into it. Gcc is usually called the GNU "compiler" but actually it is a program that figures out which of several compilers to call to actually compile your source (there are separate C and C++ compilers) and also calls the linker (see below) to link the executable. Executable - Another word for a program or a file that contains a program. In the Win32 world an executable file ends with the extension ".exe". Linker - A program that takes a set of object files and builds a complete program from them by resolving all the undefined references in one object file to where the data or function is actually defined in other object files which are part of the program. The GNU linker is "ld", but generally you run it indirectly using gcc. Archive or Library - A collection of object files put together as a single file for easy access. Usually a library has a file name like "libxxx.a" where xxx is the 'real' name of the library. For example, the basic Mingw32 library is libmingw32.a, which is automatically linked to all Mingw32 programs. A.2 Compiling The simplest way to make an executable is to invoke gcc like this: gcc -o foo.exe foo.c This command builds the executable foo.exe from the source file foo.c. The "-o" option tells gcc that the next item on the command line is the name of the new file to create (or [o]utput). If you don't use this option then your executable will end up with a name like "a.out". Gcc decides which compiler to use based on the name of the files supplied to it. Foo.c is compiled using the C compiler, if the file was named foo.cpp or foo.cc then it would be compiled with the C++ compiler. (Actually some people use a capitol 'C' for C++ and lower case 'c' for C, but that is not a good idea on Win32 systems for obvious reasons, and I personally don't think it's a good idea under UNIX either.) That's fine, but generally you will want to compile more complex programs with several source files. You could just add all the source file names to gcc, but that would mean every time you changed one file you would have to re-compile all of them. So usually you will compile object files and link them later. To compile an object file use a command line like this: gcc -c -o foo.o foo.c The -c option tells gcc to [c]ompile only. This will create an object file foo.o from the C language source file foo.c. A.3 Linking Now you have a bunch of object files and you want to make an executable. You can do this using a command line like this: gcc -o foo.exe foo.o bar.o foobar.o Gcc knows that .o files are object files, and sees that there are no source files in the input, so it simply calls the linker to build the executable foo.exe from the object files foo.o bar.o and foobar.o. If all goes well you should be able to execute foo.exe simply by typing "foo" at the command prompt, or by double clicking on the foo program's icon in Explorer. A.4 Libraries For large projects it is convenient to put together sets of related utility functions and data in libraries. You also will want to use libraries supplied by other companies or programmers. Given a library libgoo.a that contains some utility functions you want to use in your foo program you can link foo.exe like this: gcc -o foo.exe foo.o bar.o foobar.o -lgoo This assumes that libgoo.a is located in one of the directories on your LIBRARY_PATH. On the other hand, if you need to add a directory to the library path you can use the -L option. For example if goo was in the directory C:\goolib (or /goolib in UNIX-speak) then you could do this: gcc -o foo.exe foo.o bar.o foobar.o -L/goolib -lgoo Or if libgoo.a was in the current directory you could use this: gcc -o foo.exe foo.o bar.o foobar.o -L. -lgoo You can also include libraries directly using their full file names, like this: gcc -o foo.exe foo.o bar.o foobar.o libgoo.a Linking with a library is just like including all the object files inside the library on your link command line, except for one difference. If none of the object files (or other libraries) on the command line before a library contain references to an object file inside the library then that object file will not be added to the program at all. This means that this will not work: gcc -o foo.exe -lgoo foo.o bar.o foobar.o Because none of the objects in libgoo.a will be added to your program and all the references from foo.o bar.o and foobar.o will fail to be resolved. You need to put a library on the command line AFTER all other objects or libraries that might reference it have been included. To create your own libraries use the ar program. The following command adds foo.o and bar.o to the foobar library, replacing old copies if they exist, and creating the library from scratch if it didn't exist in the first place: ar -ru libfoobar.a foo.o bar.o That's it for now. Happy programming. Colin Peters (colin@bird.fu.is.saga-u.ac.jp) -- Colin Peters - Saga Univ. Dept. of Information Science -- colin@bird.fu.is.saga-u.ac.jp - finger for PGP public key -- http://www.fu.is.saga-u.ac.jp/~colin/index.html -- http://www.geocities.com/Tokyo/Towers/6162/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From colin@bird.fu.is.saga-u.ac.jp Tue Aug 5 22:24:00 1997 From: colin@bird.fu.is.saga-u.ac.jp (Colin Peters) Date: Tue, 05 Aug 1997 22:24:00 -0000 Subject: stack size issues, multi-threading Message-ID: <01BCA273.B0B32340@gbird0> Larry Meadows[SMTP:lfm@pgroup.com] wrote: >1. How do I increase the stack size for an application? Is this done >at link time, or can it be done at run time. > >2. How are stacks allocated for multi-threaded applications. As Mikey said: gcc -Wl,--heap,1024,--stack,4096 -o blah blah.c for setting the heap and or stack of an executable. Now for question 2: You can set the stack size of threads as an argument to CreateThread under Win32: HANDLE CreateThread( LPSECURITY_ATTRIBUTES lpsa, DWORD cbStack, LPTHREAD_START_ROUTINE lpStartAddr, LPVOID lpvThreadParm, DWORD fdwCreate, LPDWORD lpIDThread); Set cbStack to the number of bytes you wish to commit as an initial stack size. You can also pass zero, in which case the same value as was used for the primary thread (this is 0x1000, or 4 KB, and is not affected by the --stack option) will be used. It is my understanding that this size is not an upper limit on the stack size of the thread, but only an initial size (and size to grow by). This is based on the following passage from "Advanced Windows": ---quote--- The cbStack parameter specifies how much address space the thread is allowed to use for its own stack. Every thread owns its very own stack. When CreateProcess starts an application, it calls CreateThread to initialize the process's primary thread. For the cbStack parameter, CreateProcess uses the value stored inside the executable file. You can control this value using the linker's /STACK switch: /STACK:[reserve] [,commit] The reserve argument sets the amount of memory the system should reserve in the address space for the thread's stack. The default is 1 MB. The commit argument specifies the amount of reserved address space that should initially be committed to the stack. The default is 1 page. (See Chapter 6 for a discussion of reserving and committing memory.) As the code in your thread executes, it is quite possible that you'll require more than 1 page of memory. When your thread overflows its stack, an exception is generated. (See Chapter 14 for detailed information about handling exceptions.) The system catches the exception and commits another page (or whatever you specified for the commit argument) to the reserved space, which allows your thread's stacks to grow dynamically as needed. When calling CreateThread you can pass 0 to the cbStack parameter. In this case, CreateThread creates a stack for the new thread using the commit argument embedded in the EXE file by the linker. The amount of reserved space is always 1 MB. The system sets a limit of 1 MB to stop functions that recurse endlessly. ---endquote--- The --stack option appears to change the "reserve" value (or upper limit) on stack size. However, there seems to be no way to change the upper limit on a thread's stack space (still, "1 MB of stack!? What are you thinking!?"). It does seem that, other than possible efficiency issues, there's not much point in setting the stack size, since it will grow dynamically anyway (that is, if someone hasn't installed a buggy exception handler and messed things up). Hope this helps, Colin. -- Colin Peters - Saga Univ. Dept. of Information Science -- colin@bird.fu.is.saga-u.ac.jp - finger for PGP public key -- http://www.fu.is.saga-u.ac.jp/~colin/index.html -- http://www.geocities.com/Tokyo/Towers/6162/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Ian@kiwiplan.co.nz Tue Aug 5 23:59:00 1997 From: Ian@kiwiplan.co.nz (Ian Collins) Date: Tue, 05 Aug 1997 23:59:00 -0000 Subject: Gnu-Win32 Spooler Message-ID: <07551A47EAD6D01186430060B025C6F41C13@ntserver.kiwiplan2.co.nz> I'm a newby to gnu-win32, but not to Unix. Are there any printing facilities with this package? The command print seems to do something, but nothing comes out on any of my NT printers. How is this configured? Many Regards, Ian Collins - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From David.Billinghurst@restech.cra.com.au Wed Aug 6 00:35:00 1997 From: David.Billinghurst@restech.cra.com.au (Billinghurst, David) Date: Wed, 06 Aug 1997 00:35:00 -0000 Subject: f77 of fc script for f2c Message-ID: Does anyone know of a working fc or f77 script for the f2c package. +++++++++++++++++++++++++++++++++++++++++ (Mr) David Billinghurst Comalco Research Centre PO Box 316, Thomastown, Vic, Australia, 3074 Phone: +61 3 9469 0642 FAX: +61 3 9462 2700 Email: David.Billinghurst@restech.cra.com.au It's not my speeling that's the trouble. It's just taht I can't tipe. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From iverson@lionheart.com Wed Aug 6 01:19:00 1997 From: iverson@lionheart.com (Tim Iverson) Date: Wed, 06 Aug 1997 01:19:00 -0000 Subject: Assembler References: <33E7EF73.121AB62E@tfs.net> Message-ID: The quick reply is, "Yes, gas is the GNU assembler and it works just fine." Now for the longer version ... I find several of your comments more than a little humorous. Personally, I'd never hire any programmer for systems (driver or kernel) or embedded programming that couldn't skillfully write an assembly routine at need. It is a vital skill in many areas of business. Also, for which processor do you want to write in assembly? I would assume that you intend to target the i386 family since you're asking the question of the gnu-win32 mailing list. I would strongly urge you to learn assembly for any general purpose, non-RISC, CPU (eg. PDP-11/70, Motorola 68K family, etc.) instead for your first try. The intel x86 family is more of a special purpose stack processor and as such you will miss out on learning many of the more common constructions used by more general CPUs. Of course, there is a large market for x86 assembly programmers, so learning it is far from a waste of time. Just remember that Intel is one of the weird ones, so what you learn may not carry well to the rest of the world. Lastly, if you're going to learn Intel assembly, DO NOT USE GAS! Use just about any other commercial or PD assembler instead. Gas was never intended for humans, though it can be made to serve if you preprocess with cpp and/or m4. It's syntax is also different than every other Intel assembler in existence. Your knowledge won't carry anywhere. Oh, btw, you will want to pick up a copy of the CPU reference manual for your target CPU. And for examples, it's far better to look at assembly written by a human (even bad assembly ;-) than to look at compiler output. - Tim Iverson iverson@cisco.com +---------------- | Date: Tue, 05 Aug 1997 22:28:51 -0500 | From: "Stephen P. Green" | To: "gnu-win32@cygnus.com" | Subject: Assembler | | I am interested in learning assembly language. I know this is almost a | completely dead language, but I find it intringing. I have two | questions: First is their any assembler included in the b18 release, or | any unix gnu assembler that could be ported(please include a link if it | has already been done or any other programs that might serve the same | purpose) and Secondly, any literature that you could point me to as a | good starting point. I know that this request is a little bit of a | deviation from our normal topic, but I beg you to show me a little | patience and help me. Thanks to you all... - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kuku@gilberto.physik.rwth-aachen.de Wed Aug 6 01:19:00 1997 From: kuku@gilberto.physik.rwth-aachen.de (Christoph Kukulies) Date: Wed, 06 Aug 1997 01:19:00 -0000 Subject: gnu-win32 beta18 did not compile on a linux system References: <199708060233.WAA29587@iag.net> Message-ID: <19970806102246.53836@gil.physik.rwth-aachen.de> On Tue, Aug 05, 1997 at 10:34:39PM -0400, Sonic Junior wrote: > > Why are you compiled Gnuwin32 on a Linux system? Gnuwin32 is made for > Windows systems. gnuwin32 is a compiler environment that targets Win32 (Windows 95/Windows NT) Where the system is hosted is another matter :-) In other words, he (the original author) wanted to build a cross compiler for Linux. > > -SJ > > ---------- > > From: Andreas Jellinghaus (bei rf) > > To: gnu-win32@cygnus.com > > Subject: gnu-win32 beta18 did not compile on a linux system > > Date: Tuesday, August 05, 1997 12:26 PM > > > > hi. > > > > i was running configure with --prefix=$HOME/gnu > --target=i386-unknown-cygwin32 > > i had to do this, to get compile and install done : > > cp install-sh i386-unknown-cygwin32 > > ln -s "../config.sub" i386-unknown-cygwin32/ > > ln -s "../config.sub" i386-unknown-cygwin32/winsup > > cp gcc/getopt.h winsup > > cp gcc/getopt.h i386-unknown-cygwin32/winsup > > > > i hope this is enough to locate the bugs. if not, i have logfiles of > > all configure, and make calls here. > > > > note : i downloaded the cdk source from ftp.sunet.se. > > there are md5sums of xc[a-h], but the source is xc[a-j], and has > > different md5sums. i assumed, that the md5sum file is not up-to-date... > > > > regards, andreas > > > > - > > For help on using this list (especially unsubscribing), send a message to > > "gnu-win32-request@cygnus.com" with one line of text: "help". > - > For help on using this list (especially unsubscribing), send a message to > "gnu-win32-request@cygnus.com" with one line of text: "help". -- --Chris Christoph P. U. Kukulies kuku@gil.physik.rwth-aachen.de - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kuku@gilberto.physik.RWTH-Aachen.DE Wed Aug 6 02:16:00 1997 From: kuku@gilberto.physik.RWTH-Aachen.DE (Christoph Kukulies) Date: Wed, 06 Aug 1997 02:16:00 -0000 Subject: rm -rf -> Circular directory structure Message-ID: <199708060919.LAA15850@gil.physik.rwth-aachen.de> I'm having a weird problem under NT 4.0 (I remember to have seen this with b17 and NT3.51 also): D:\>rm -rf somedir rm: WARNING: Circular directory structure. This almost certainly means that you have a corrupted file system. NOTIFY YOUR SYSTEM MANAGER. Cycle detected: somedir/FOO/FOO is the same file as somedir CHKDSK doesn't flag any errors on the filesystem and I tend to believe this isn't the cause anyway. The interesting thing is: When I remove the file somedir/FOO/FOO the rm command is able to run through without errors. -- Chris Christoph P. U. Kukulies kuku@gil.physik.rwth-aachen.de - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From STEPHANE.S.P.PERIO@tts.thomson.fr Wed Aug 6 02:29:00 1997 From: STEPHANE.S.P.PERIO@tts.thomson.fr (STEPHANE.S.P.PERIO@tts.thomson.fr) Date: Wed, 06 Aug 1997 02:29:00 -0000 Subject: Building GDB problem Message-ID: <2200011106081997/A04771/INDRE/11B832C03900*> <@MHS> Hi, I apologize if this is an FAQ, but i see nothing about that in the mailing lists. I'm using gnu win32 b18 on NT 4.0 SP3 and i failed to build GDB. Under bash, in the directory my-gdb, i run ../gdb/configure and it recognized correctly the i386-pc-cygwin32 environment. When i run make, the first problem i have is that the include files malloc.h and bfd.h are missing but i found a copy of them in /cdk/newlib/libc/stdlib/malloc.h and /gnuwin32/b18/include/bfd.h. Are these two files the correct ones ? If i use them, the make stops when compiling the i386-tdep.c file with the error : ../gdb/i386-tdep.c: In function `get_longjmp_target': ../gdb/i386-tdep.c:592: `SP_ARG0' undeclared (first use this function) ../gdb/i386-tdep.c:592: (each undeclared identifier is reported only once ../gdb/i386-tdep.c:592: for each function it appears in.) ../gdb/i386-tdep.c: In function `set_assembly_language_command': ../gdb/i386-tdep.c:698: `bfd_mach_i386_i386' (first use this function) ../gdb/i386-tdep.c:698: `bfd_mach_i386_i8086' (first use this function) ../gdb/i386-tdep.c: In function _initialize_i386_tdep': ../gdb/i386-tdep.c:709: `bfd_mach_i386_i386' (first use this function) Can someone help me ? Thanks in advance. S. PERIO - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gay@aspentec.com Wed Aug 6 02:29:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Wed, 06 Aug 1997 02:29:00 -0000 Subject: login and telnetd Message-ID: <9708060928.AA38659@hagar.aspentec.com> I'm using b18 with latest coolview (01-Aug) under NT4SP3, and got a problem with login.exe After playing around with user rights, I got login.exe almost working... (BTW, Sergey do you know exactly how to configure a user with User Manager to have login.exe working fine ?) I enter login name and password, which are ok, and I get : login: no shell: /bin/sh: Not owner in my /etc/passwd, I have : 'sysroot::500:0:::/bin/sh' among other things (sysroot is the name of the builtin NT administrator, renamed) telnetd : when running 'inetd -d /etc/inetd.conf' in one bash session, and running telnet (Sergey's telnet.exe) in another bash session, trying to connect on localhost, I got : telnet client side : Trying 127.0.0.1... setuid called, and it's not done yet. Connected to pathfinder.[...] Escape character is '^]'. Connection closed by foreign host. in bash session where inetd was started : ... someone wants telnet accept, ctrl 3 + Closing from 17 1022 execl /usr/local/libexec/in.telnetd 1022 reaped, status 0x100 Is anyone using telnetd successfully ? Any idea ? BTW, when I run rxvt, though it's apparently working fine, it does not source my $HOME/.bashrc file, and I also get several 'setuid called, and it's not done yet' (in the terminal where I start rxvt and also in the rxvt window itself)..... - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ronen@netmanage.co.il Wed Aug 6 03:08:00 1997 From: ronen@netmanage.co.il (Ronen Pinchook) Date: Wed, 06 Aug 1997 03:08:00 -0000 Subject: Gnu-Win32 Spooler References: <07551A47EAD6D01186430060B025C6F41C13@ntserver.kiwiplan2.co.nz> Message-ID: how do I get off this list --- On Wed, 6 Aug 1997 19:00:22 +1200 Ian Collins wrote: > I'm a newby to gnu-win32, but not to Unix. > Are there any printing facilities with this package? > The command print seems to do something, but nothing comes out on any of > my NT printers. > How is this configured? > > Many Regards, > Ian Collins > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mark@solcom.co.uk Wed Aug 6 03:33:00 1997 From: mark@solcom.co.uk (Mark Hannah) Date: Wed, 06 Aug 1997 03:33:00 -0000 Subject: Beginner's Problem Message-ID: When attempting to install GCC only using, make install-gcc I got the following message :- rm -rf /install/native/H-i386-cygwin32/lib/gcc-lib/i386-pc-cygwin32/cygnus-2.7.2-970404/include mkdir /install/native/H-i386-cygwin32/lib/gcc-lib/i386-pc-cygwin32/cygnus-2.7.2-970404/include chmod a+rx /install/native/H-i386-cygwin32/lib/gcc-lib/i386-pc-cygwin32/cygnus-2.7.2-970404/include (cd include; \ tar -cf - .; exit 0) | (cd /install/native/H-i386-cygwin32/lib/gcc-lib/i386-pc-cygwin32/cygnus-2.7.2-970404/include; tar xpBf - ) tar: Could not change access and modification times of objc: Permission denied tar: Could not change access and modification times of .: Permission denied make: *** [install-headers-tar] Error 2 make: *** [install-gcc] Error 2 Can anyone help ? I would be interested to hear from anyone who has managed to re-build all the development kit under windows NT 4.0. Thanks in advance. Mark. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From rich@rdp.introl.com Wed Aug 6 03:33:00 1997 From: rich@rdp.introl.com (Richard Pennington) Date: Wed, 06 Aug 1997 03:33:00 -0000 Subject: gnu-win32 beta18 did not compile on a linux system References: <199708060233.WAA29587@iag.net> Message-ID: <199708061032.FAA13097@rdp.introl.com> > > Why are you compiled Gnuwin32 on a Linux system? Gnuwin32 is made for > Windows systems. > > -SJ > For the same reason I cross compiled it under NetBSD. To allow me to develop software for Windows from a *real* operating system. :) Of course, I just cross compiled the compiler, assembler, linker, etc. and use Mingw32. Works great. -Rich -- Richard Pennington Introl Corporation, Milwaukee, WI USA Email: rich@introl.com Phone: +1 414-273-6100 Fax: +1 414-273-6106 US and Canada: 1 800-327-7171 Cross development tools for the 68HC05, 68HC08, 6809, 68HC11, 68HC12, 68HC16, and 68XXX: http://www.introl.com ftp://ftp.introl.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gay@aspentec.com Wed Aug 6 07:11:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Wed, 06 Aug 1997 07:11:00 -0000 Subject: termcap issues... Message-ID: <9708061342.AA38522@hagar.aspentec.com> I'm using b18 with latest coolview (01-Aug) under NT4SP3... I got some strange problems with termcap... I've exported TERM=linux If I export also TERMCAP=/etc/termcap (containing the linux entry), less (3.32) behave strangely, it starts paging from the top of the sreen, overwriting previous information... Maybe that's because I use bash inside an NT Command Prompt with scroll-bar (I set a buffer size of 1000 lines in the shortcut properties). BUT, if I set TERMCAP=/etc/termcap/linux, less behaves _normally_, but with a 'Warning: terminal not fully functional' I'm using termcap 1.3 (compiled with -D_STRICT_ANSI to avoid error regarding bcopy and stuff...) Has anyone similar experience ? Any ideas ? Jerome "ZoSo" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From hchang@ee.usyd.edu.au Wed Aug 6 07:11:00 1997 From: hchang@ee.usyd.edu.au (Hung Hsin Chang) Date: Wed, 06 Aug 1997 07:11:00 -0000 Subject: rxvt and telnetd Message-ID: Need procedures for porting rxvt and telnetd. Please help. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gay@aspentec.com Wed Aug 6 08:19:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Wed, 06 Aug 1997 08:19:00 -0000 Subject: login and telnetd Message-ID: <9708061519.AA33062@hagar.aspentec.com> Sergey- Thanks a lot... After a few tries, login is working fine, I had to give my administrator account the "Replace a process token" right, "Increases quotas" is set by default for administrators, "Logon as a service" does not seem to be required to run login locally. I still cannot have telnetd running... Though rexecd is ok ,I can run rxvt via rexec, but then : - bash starts in directory '/HOME/GAY' (though $HOME=/home/gay), but if I type 'cd' then 'pwd' I got '/home/gay' - my .bashrc is not executed (maybe linked to this uppercase/lowercase problem for my home dir), - I got lots of messages like 'free_tty: count < 0' in rxvt, - does not work in rxvt, I have to type + !... I've run into a lot of newbie problems to setup gnuwin32 ok, I'm trying to write-down what I may consider as 'must_be_done' steps. If anyone is interested or wants to add some input... you're welcome *:o)8 Jerome "ZoSo" -----Original Message----- From: Sergey Okhapkin To: 'Jerome Gay' Cc: cygnus Date: Wednesday, August 06, 1997 3:18 PM Subject: RE: login and telnetd >Jerome Gay wrote: >> (BTW, Sergey do you know exactly how to configure a user with User >Manager >> to have login.exe working fine ?) >> I enter login name and password, which are ok, and I get : >> login: no shell: /bin/sh: Not owner > >The rights required are "Act as part of operation system" (SE_TCB_NAME) for >LogonUser call, and "Assign primary token" (SE_ASSIGNPRIMARYTOKEN_NAME) > and "Increase quotas" (SE_INCREASE_QUOTA_NAME) for CreateProcessAsUser >call. > > > >-- >Sergey Okhapkin, http://www.lexa.ru/sos >Moscow, Russia >Looking for a job. > > > > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dmeans@bellsouth.net Wed Aug 6 08:19:00 1997 From: dmeans@bellsouth.net (David Means) Date: Wed, 06 Aug 1997 08:19:00 -0000 Subject: latest Sergey's update (08/01) crashes on Win95 References: <9708040956.AA28128@hagar.aspentec.com> Message-ID: <33E87C1B.9A2DF836@bellsouth.net> Jerome Gay wrote: > > Hi all, > > I just installed Sergey's latest coolview, dated Aug,1st... when I > start > bash, WIndows95 crahes completely (hard-reset necessary)... any ideas > ? > > Jerome > > - > For help on using this list (especially unsubscribing), send a message > to > "gnu-win32-request@cygnus.com" with one line of text: "help". -- Ditto me too. I've got a valid /etc/passwd and /etc/group. Here's what it does: BASH.EXE: /etc/profile: error 0 BASH.EXE: /home/users/dmeans/.bash_login: cannot execute binary file BASH.EXE-2.01$ .bash_login is a sym link to .profile. I know that /etc/profile isn't getting sourced because pwd reports /dev/c/ and not $HOME, and $PATH is not set. I don't start bash in /dev/c nor do I cd to it, so I don't know how it's getting there. If I execute a command at the prompt, the command will run, i.e., mount or /bin/ls, but the system will hang afterwards. The mouse still works, but nothing will get the system back save a three-finger-salute ( ctrl-alt-del ). Sometimes it will take a moment or two before it hangs, at least it did after the mount command. Thanks, David Means mailto:dmeans@bellsouth.net - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From lhall@rfk.com Wed Aug 6 08:19:00 1997 From: lhall@rfk.com (Larry Hall) Date: Wed, 06 Aug 1997 08:19:00 -0000 Subject: rm -rf -> Circular directory structure Message-ID: <2.2.32.19970806133414.0097c55c@ma.ultranet.com> At 11:19 AM 8/6/97 +0200, Christoph Kukulies wrote: >I'm having a weird problem under NT 4.0 (I remember to >have seen this with b17 and NT3.51 also): > >D:\>rm -rf somedir >rm: WARNING: Circular directory structure. >This almost certainly means that you have a corrupted file system. >NOTIFY YOUR SYSTEM MANAGER. >Cycle detected: >somedir/FOO/FOO >is the same file as >somedir > >CHKDSK doesn't flag any errors on the filesystem and I tend to believe >this isn't the cause anyway. > >The interesting thing is: When I remove the file somedir/FOO/FOO >the rm command is able to run through without errors. > > >-- >Chris Christoph P. U. Kukulies kuku@gil.physik.rwth-aachen.de Chris, This should be in the FAQ (I expect its not at the moment). The problem is caused by a faulty inode mapping algorithm in all the betas. There is no problem with your file system. It merely means that the faulty inode mapping for files and directories thinks there's more than 1 file and/or directory for the generated inode number, which is certainly a bad thing for traditional UNIX file systems. The error hinders rm from working correctly but does not indicate file system problems. You'll just need to delete the intended files/directories another way (most likely via Windows or DOS tools). I believe Sergey's latest DLL has some fixes to mitigate this problem. I don't think there's a complete fix yet (correct me if I'm wrong folks). Larry Hall lhall@rfk.com RFK Partners, Inc. (617) 239-1053 8 Grove Street (617) 239-1655 - FAX Wellesley, MA 02181 - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Wed Aug 6 08:46:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Wed, 06 Aug 1997 08:46:00 -0000 Subject: login and telnetd Message-ID: <01BCA28C.F4981520@gater.krystalbank.msk.ru> Jerome Gay wrote: > (BTW, Sergey do you know exactly how to configure a user with User Manager > to have login.exe working fine ?) > I enter login name and password, which are ok, and I get : > login: no shell: /bin/sh: Not owner The rights required are "Act as part of operation system" (SE_TCB_NAME) for LogonUser call, and "Assign primary token" (SE_ASSIGNPRIMARYTOKEN_NAME) and "Increase quotas" (SE_INCREASE_QUOTA_NAME) for CreateProcessAsUser call. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ian@cygnus.com Wed Aug 6 08:46:00 1997 From: ian@cygnus.com (ian@cygnus.com) Date: Wed, 06 Aug 1997 08:46:00 -0000 Subject: Assembler References: Message-ID: <199708061546.LAA02093@tweedledumb.cygnus.com> >Lastly, if you're going to learn Intel assembly, DO NOT USE GAS! Use >just about any other commercial or PD assembler instead. Gas was never >intended for humans, though it can be made to serve if you preprocess >with cpp and/or m4. It's syntax is also different than every other Intel >assembler in existence. Your knowledge won't carry anywhere. Although gas was not originally intended for humans, I feel I should mention that it now supports macros, conditionals, and other helpful constructs, so it can now be reasonably used for hand assembly code. The gas ix86 syntax is not like MASM or Intel assemblers, but it is (or should be) identical to that accepted by AT&T style ix86 assemblers, including those used on all ix86 Unix systems I know of. However, I agree that you'll be better off using some other ix86 assembler. Ian - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From p.dalgaard@kubism.ku.dk Wed Aug 6 09:13:00 1997 From: p.dalgaard@kubism.ku.dk (Peter Dalgaard BSA) Date: Wed, 06 Aug 1997 09:13:00 -0000 Subject: Gnu-Win32 Spooler References: <07551A47EAD6D01186430060B025C6F41C13@ntserver.kiwiplan2.co.nz> Message-ID: Ronen Pinchook writes: > > how do I get off this list > .. > - > For help on using this list (especially unsubscribing), send a message to > "gnu-win32-request@cygnus.com" with one line of text: "help". > Have you considered sending a message to "gnu-win32-request@cygnus.com" with one line of text: "help"? -- O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk) FAX: (+45) 35327907 - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jelling@iitb.fhg.de Wed Aug 6 09:13:00 1997 From: jelling@iitb.fhg.de (Andreas Jellinghaus (bei rf)) Date: Wed, 06 Aug 1997 09:13:00 -0000 Subject: dll building : segmentation faults Message-ID: <19970806130447.28713@bsilx1.iitb.fhg.de> gcc -c dllentry.c -o dllentry.o ld -s --base-file my_dll.base --entry _dll_entry@12 \ --dll --output my_dll.jnk dllentry.o dlltool --base-file my_dll.base --def my_dll.def --dllname my_dll.dll \ --output-exp my_dll.exp --output-lib my_dll.a ld -s --dll --entry _dll_entry@12 --output my_dll.dll my_dll.exp dllentry.o working dllentry.c : #include asm(".section .idata$3\n" ".long 0,0,0,0,0,0,0,0"); WINAPI dll_entry(int a,int b,int c) { return 1; } char* dummy() { return ""; } not working dllentry.c : #include asm(".section .idata$3\n" ".long 0,0,0,0,0,0,0,0"); WINAPI dll_entry(int a,int b,int c) { return 1; } void dummy() { return ; } (also not working int dummy, int* dummy, ...). any hints ? i'm using gnu win 32, beta 18 (i tried this on winnt and linux with the same result). andreas - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From paul.toseland@virgin.net Wed Aug 6 09:50:00 1997 From: paul.toseland@virgin.net (Matthew Toseland) Date: Wed, 06 Aug 1997 09:50:00 -0000 Subject: B18 bash, make, ls not compatible with IE4Preview2? Message-ID: <33E8581F.1BC1@virgin.net> System: Win95, Service Pack 1, Microsoft Plus!, Internet Explorer 4.0 Preview 2, 32Mb RAM, several big Drivespace 3's P150+ H:\Matthew\CPP\CACHIMPT>make -f import.mak gcc -ggdb -c twowaydw.cpp (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) In cygwin_except_handler (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) Exception trapped! (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) exception C0000005 at 7E2EB4 4F (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) exception: ax 42087C bx 1 cx 254E3F0 dx 815AB034 (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) exception: si 7E2EF044 di 10 0479B4 bp 254E594 sp 254E3CC (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) exception is: STATUS_ACCESS_ VIOLATION (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) Stack trace: (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 0: sp = 0x254E1E4, pc = 0x1000CEC2 (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 1: sp = 0x254E200, pc = 0xBFF7663C (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 2: sp = 0x254E224, pc = 0xBFF859D9 (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 3: sp = 0x254E2BC, pc = 0xFFECBAD7 (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 4: sp = 0x254E594, pc = 0x10011145 (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 5: sp = 0x254E630, pc = 0x100111F4 (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 6: sp = 0x254E718, pc = 0x1001136F (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 7: sp = 0x254E784, pc = 0x10011360 (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 8: sp = 0x254E78C, pc = 0x402542 (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 9: sp = 0x254E7C4, pc = 0x4026B6 (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 10: sp = 0x254E7D8, pc = 0x402944 (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 11: sp = 0x254E810, pc = 0x40176E (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 12: sp = 0x254E828, pc = 0x40CA7E (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 13: sp = 0x254E838, pc = 0x40C35E (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 14: sp = 0x254E870, pc = 0x40B787 (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) frame 15: sp = 0x254E88C, pc = 0x40C59B (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\MAKE.EXE 1001) End of stack trace (more sta ck frames may be present) MAKE.EXE: *** [twowaydw.o] Error 2 But: H:\Matthew\CPP\CACHIMPT>gcc -ggdb -c twowaydw.cpp H:\Matthew\CPP\CACHIMPT> When I go into bash and try to do an ls: H:\Matthew\CPP\CACHIMPT>bash bash$ ls (H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\BASH.EXE 1001) In cygwin_except_handler [it hangs at this point, and I have to kill both H:\GNUWIN32\B18\H-I386-CYGWIN32\BIN\BASH.EXE and Bash from Ctrl+Alt+Del] Should I just revert to Platform Preview 1 [which works], or is this fixable? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mark@solcom.co.uk Wed Aug 6 09:50:00 1997 From: mark@solcom.co.uk (Mark Hannah) Date: Wed, 06 Aug 1997 09:50:00 -0000 Subject: beginners Problem :2 Message-ID: When I try to re-build the cdk using configure (+options for native compiler) make all make install I get the following error message :- src/install-sh -c patch /install/native/H-i386-cygwin32/bin/patch install: patch does not exist, trying with .exe appended /src/install-sh -c -m 644 /src/patch/patch.man /install/native/H-i386-cygwin32/man/man1/patch.1 /src/install-sh -c -m 644 libreadline.a /install/native/H-i386-cygwin32/lib/libreadline.a ranlib /install/native/H-i386-cygwin32/lib/libreadline.a /src/install-sh -c -m 644 /src/readline/readline.h /install/native/H-i386-cygwin32/include/readline/readline.h /src/install-sh -c -m 644 /src/readline/keymaps.h /install/native/H-i386-cygwin32/include/readline/keymaps.h /src/install-sh -c -m 644 /src/readline/chardefs.h /install/native/H-i386-cygwin32/include/readline/chardefs.h Installing libtcl7.6.a ranlib: libtcl7.6.a: No such file or directory make: *** [install-binaries] Error 9 make: *** [install] Error 2 make: *** [install-tcl] Error 2 Can anyone help ? Thanks in advance, Mark. _______________________________________________ Mark G Hannah Email:mark@solcom.co.uk SolCom Systems SolCom House Tel: +44 1506 461707 Meikle Road Fax: +44 1506 461717 Kirkton Campus Livingston http://www.solcom.co.uk EH54 7DE 06/08/97 11:26:58 _______________________________________________ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jont@harlequin.co.uk Wed Aug 6 09:50:00 1997 From: jont@harlequin.co.uk (Jon Thackray) Date: Wed, 06 Aug 1997 09:50:00 -0000 Subject: Assembler References: <33E7EF73.121AB62E@tfs.net> Message-ID: <199708061118.MAA18965@zaphod.long.harlequin.co.uk> Stephen P. Green writes: > I am interested in learning assembly language. I know this is almost a > completely dead language, but I find it intringing. I have two > questions: First is their any assembler included in the b18 release, or > any unix gnu assembler that could be ported(please include a link if it > has already been done or any other programs that might serve the same > purpose) and Secondly, any literature that you could point me to as a > good starting point. I know that this request is a little bit of a > deviation from our normal topic, but I beg you to show me a little > patience and help me. Thanks to you all... b18 contains as, the gnu assembler. This differs from most commercial x86 assemblers both in the instruction format it accepts and the argument order it uses. There should be info pages on it, but I doubt this will be sufficient to anble you to learn assembler language, assuming you never learnt assembler for any other processor. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kuku@gilberto.physik.rwth-aachen.de Wed Aug 6 09:50:00 1997 From: kuku@gilberto.physik.rwth-aachen.de (Christoph Kukulies) Date: Wed, 06 Aug 1997 09:50:00 -0000 Subject: bfd, ld, and dlltool patches References: <199707181824.MAA06085@chorus.dr.lucent.com> Message-ID: <19970806165229.35697@gil.physik.rwth-aachen.de> On Fri, Jul 18, 1997 at 12:24:20PM -0600, marcus@bighorn.dr.lucent.com wrote: > OK, well, I just saw the test message from this morning, but I haven't seen > any sign of the two previous mailings of this message from Wednesday and > Thursday. If they did actually make it to the mailing list, I'm sorry for > the repeat, and would somebody please tell me to stop! :-) > > > I have been doing some work in my spare time (actually it was a while ago > on b17.1) to build a cross-compiler environment to generate NT code on a > Sun and interwork with MS DLLs. To this end, I have about 350 lines of > patches to bfd and ld files, and 1200 lines of patches to dlltool. These > have been forwarded to the b18 versions, although I haven't had time to > do any further development. > > At this point, I can link with many of the microsoft .lib libraries. The > remaining problem seems to be in handling some of the segment types where > ld complains that it is ignoring multiple instances of a segment, aparently > because of a mis-interpretation of communil data header information. I > can produce a .lib that is almost acceptable to MS LINK, but it seems that > LINK wants the file names inside the archive to be identical, unlike the > dt0.o, dt1.o, etc. files produced by dlltool. There isn't an obvious way > to get bfd to produce an archive with internal file names to be the same > even though it can handle this case in an existing archive just fine. Perhaps > playing with storing the files in memory instead of in disk files would > be appropriate here, since the problem seems to be in that aspect. > > Most of the changes to dlltool were to (nearly) eliminate the use of the > assembler to produce the .o files and to simply write the .o files directly > with bfd. I have converted all the code to deal with generating the import > tables, but have not yet attacked the export tables. I also generate an > import table terminator that is compatible with the MS .libs, so the > fixup.c kludges should no longer be necessary. > > My question here is, does the list want to see the patches posted here or not? > If not, what mechanism would be appropriate? Since b19 is forthcoming and > some of these changes may be useful for that, I'd like to get some of these > changes submitted for consideration by cygnus, and perhaps save them some > work if they haven't already done equivalent work. I would like to see the patches posted, either as a gzipped/uuencoded mail appendix or put up for ftp somewhere. > > The context diff is 1419 lines long... > > marcus hall > Lucent Technologies > - > For help on using this list (especially unsubscribing), send a message to > "gnu-win32-request@cygnus.com" with one line of text: "help". -- Chris Christoph P. U. Kukulies kuku@gil.physik.rwth-aachen.de - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From collver@dnc.net Wed Aug 6 09:50:00 1997 From: collver@dnc.net (Reuben Collver) Date: Wed, 06 Aug 1997 09:50:00 -0000 Subject: Assembler Message-ID: <19970806161604850.AAA227@powell.dnc.net> > I am interested in learning assembly language. I know this is almost a > completely dead language, but I find it intringing. I have two > questions: First is their any assembler included in the b18 release, or > any unix gnu assembler that could be ported(please include a link if it > has already been done or any other programs that might serve the same > purpose) and Secondly, any literature that you could point me to as a > good starting point. I know that this request is a little bit of a > deviation from our normal topic, but I beg you to show me a little > patience and help me. Thanks to you all... The gnu assembler 'AS' is included with Cygnus gnu-win32, as well as the DJGPP package. You can download a GREAT and complete book on assembly from the net. The book is oriented toward intel syntax and DOS, however it is still a great reference. The llink seems to be down right now, but I used it a coulpe days ago. You can download it in adobe acrobat or postscript format to print a neat looking hardcopy, or you can browse the html formatted one. Here is the URL: http://webster.ucr.edu/Page_asm/ArtOfAsm.html The GNU assembler uses AT&T assembler syntax. NASM, a free assembler you can get from the net, uses INTEL syntax. NASM also produces coff, obj, elf, a.out, and pure binary output. COFF is useful if you wish to link with code for djgpp or cygnus gcc. The URL for NASM is http://www.cryogen.com/Nasm If you want some pointers for using AT&T syntax, look at the Brennan's reference for inline assembly with DJGPP: http://www.rt66.com/~brennan/djgpp/djgpp_asm.html hope this was helpful, Ben. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gay@aspentec.com Wed Aug 6 09:50:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Wed, 06 Aug 1997 09:50:00 -0000 Subject: exc C0000005 when running from NT Command Prompt Message-ID: <9708061220.AA34509@hagar.aspentec.com> Folks, I'using b18 with latest coolview (01-Aug) under NT4SP3. I compiled a simple C program which does some string manipulation. When run from bash, my program runs fine, when run from NT Command Prompt, I got an exception at the very end of the execution (at the very end of the main()), traceback is : In cygwin_except_handler exc C0000005 at ... Exception trapped! exception C0000005 at ... ... Any ideas ? Jerome "ZoSo" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jazz@softway.com Wed Aug 6 09:50:00 1997 From: jazz@softway.com (Jason Zions) Date: Wed, 06 Aug 1997 09:50:00 -0000 Subject: Assembler References: Message-ID: <33E8A9A1.587F@softway.com> If you're still looking at x86 processors, the CPU reference manuals for Intel's current chips are available in downloadable PDF form on Intel's web site. You'll still kill a tree printing it out, but there's no middleman driving up costs. (The Pentium MMX manual was about 800 pages; be a hero and print it double-sided. :-) Jason - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Wed Aug 6 10:37:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Wed, 06 Aug 1997 10:37:00 -0000 Subject: login and telnetd Message-ID: <01BCA2B1.1D721C60@sos> Jerome Gay wrote: > >> type 'cd' then 'pwd' I got '/home/gay' > > > >What is the real case stored on filesystem? > > On file system (NTFS) it's stored in lowercase (/home/gay) pwd always reports the real case. > >Run "rxvt -e bash -login" > > Do you mean I've to start rxvt with these options each time ? Yes. Or "rxvt -ls". Read the rxvt (and bash) manuals:-) > >Depends of X server software you're using. BS works fine with StarNet's > xwin32 and have the same nasty behavior with DEC eXcursion. > > I'm using ReflectionX 6.0 from WRQ Try to play with X server's keyboard mappings (if any). -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gebhart@informatics.net Wed Aug 6 10:37:00 1997 From: gebhart@informatics.net (Toby Gebhart) Date: Wed, 06 Aug 1997 10:37:00 -0000 Subject: No readme included in b18 Message-ID: <01BCA262.EE017960@informatics.net> I noticed the same, but I found a readme.txt file with the needed variables in it, I'm just not sure how to incorporate them. I'm using Win95, so I thought that maybe I would just add the strings in the autoexec.bat in the following format: GCC_EXEC_PREFIX=C:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\ Obviously, this didn't work. Where exactly do I put the environment variables? I wish it was more clear, or maybe I'm missing the boat entirely...... I'm new to the list, please don't flame me away with "We've already answered this question 5 million times!". Toby Gebhart mailto:gebhart@informatics.net -----Original Message----- From: Gregory Goodwin [SMTP:gm-goodwin@worldnet.att.net] Sent: Tuesday, August 05, 1997 8:34 PM To: gnu-win32@cygnus.com Subject: No readme included in b18 I just downloaded/installed the full binary installation of build 18 for Win95. The install program mentioned that I would need to view the readme to set up the environment. However, the readme was not installed. Nor was there a copy of the License Agreement. Thought you should know. Regards, Greg - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gay@aspentec.com Wed Aug 6 10:37:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Wed, 06 Aug 1997 10:37:00 -0000 Subject: Problem with Sergey's telnetd, rxvt,vim Message-ID: <9708061626.AA40482@hagar.aspentec.com> Jerzy- I'm having the same kind of problems with telnetd and rxvt... Did you solve this problems ? How ? Thanks Jerome "ZoSo" -----Original Message----- From: Jerzy Schubert To: Gilligan <_kunglao@prairienet.org> Cc: GNU Win 32 Date: Friday, July 25, 1997 1:56 AM Subject: Problem with Sergey's telnetd, rxvt,vim >I had exactly (just the pids were different) error messages >Gillian had here. > >I remembered then the advice from *jeffdb@netzone.nospam.com (Mikey) > > "after you change over to binary mode, you will need > to reinstall from cdk.exe and user.exe .... " > >I've done that, then reinstall Sergey cygwin.dll and libcygwin.a >and I'v got the login prompt: > >crusher:/ # telnet bewerly >Trying 10.11.4.66... >Connected to bewerly.ppe.bb-data.de. >Escape character is '^]'. > >CYGWIN32/NT 4.0 (BEWERLY) (tty1) > >login: jsc >Password: >Login incorrect >login: > >So one problem gone away. >However i still cannot login >login keeps saying "Login incorrect" > >Here my passwd and group file, i see that the groupid in the passwd file >does not matched the groupid in the group files, those were made with >the utility mkpasswd -l >Administrator::500:513:::/bin/sh >Gast::501:513:::/bin/sh >IUSR_YUTA::1000:513:Internet-Gastkonto::/bin/sh >jsc::1001:513:Jerzy Schubert:/home/jsc:/bin/sh >root::1002:513:admin account:/home/root:/bin/sh > >Administratoren::544:0::: >Benutzer::545:0::: >Gdste::546:0::: >Hauptbenutzer::547:0::: >Replikations-Operator::552:0::: >Sicherungs-Operatoren::551:0::: > >I could start inetd from com box or from bash observing the behavior >above. > >When i tried to test service using autoexnt > net start autoexnt >with autoexnt.bat as: >g:\gnuwin32\b18\H-i386-cygwin32\local\sbin\inetd -d >Here i cannot say instead /usr/local/sbin/inetd -d can i > >autoexnt says in the same dos box: > >AutoExNT wird gestartet (starting) >AutoExNT wurde erfolgreich gestartet (started successfully) > >In the TaskManager list i can indeed see inetd running >but when i say telnet i get >crusher:/ # telnet bewerly >Trying 10.11.4.66... >telnet: Unable to connect to remote host: Connection refused > >I stopped simple internet services, as suggested before >but as i could started inetd by hand it probably has nothing to do with >it. > > >Other problem still here is with rxvt >rxvt: Failed to open input method >seteuid called, and it's not done yet. >ptydev="/dev/tty2" >ttydev="/dev/tty2" >seteuid called and it's not done yet (4 times) >i get rxvt window but nothing in it. (no shell) >I use X11R6.3 binary provided by Arlindo da Silva build with b18. > > >And i also have problem with vim in the shell box, which many >people mentioned before, i did not understood the replay from Sergey > > >Jerzy > > > >On Tue, 22 Jul 1997, Gilligan wrote: > >> Ok, I've setup my inetd.conf, copied login.w95 to login, ran inetd -d. >> I can telnet in, but it crashes, perhaps looking for the login script. >> Here's inetd's output: >> >> registered /usr/local/libexec/in.telnetd on 14 >> someone wants telnet >> accept, ctrl 3 >> + Closing from 14 >> 1060 execl /usr/local/libexec/in.telnetd >> 1060 reaped, status 0x100 >> >> On the telnet client side, I get: >> >> Trying 192.168.1.2... >> Connected to bozo. >> Escape character is '^]'. >> >> CYGWIN32/95 4.0 (BOZO) (tty10) >> >> Connection closed by foreign host. >> . >> . >> . > >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gay@aspentec.com Wed Aug 6 10:37:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Wed, 06 Aug 1997 10:37:00 -0000 Subject: New patch available Message-ID: <9708061640.AA38100@hagar.aspentec.com> James- Sorry to bother you, but I cannot get telnetd to work on NT4SP3... Did you tweak something to get it work ? Did you set some special user rights in NT ? How are you running inetd (as a service, or from command line) ? I'll appreciate your help. Jerom "ZoSo" -----Original Message----- From: James Dumser To: Sergey Okhapkin Cc: gnu-win32@cygnus.com Date: Saturday, August 02, 1997 7:25 AM Subject: re: New patch available >On Fri, 01 Aug 1997 13:11, Sergey Okhapkin wrote: >>- console code no longer alway translates CR to NL when tty support is >> enabled. The behavior depends now of termios settings. >>- signals should terminate blocking socket calls. >>- Chris Faylor's (cfg@bbc.com) patch for wait and SIGCHLD raising >> logic - cygwin processes now waits non-cygwin processes correctly, no >> more time races problems with SIGCHLD raising and process's exit. >>- utmp/wtmp files are filled now for every cygwin's session (both local > >session and telnet session). "Who" command works now. Create /var/run >> and /var/log directories to enable utmp logging. >>- syslog() call handles now %m macro. >>- signal raising keys (ctrl-C, ctrl-\ etc) restarts suspended tty >> output. >> >>New internet servers/clients added to remote.tar.gz > >Installed this on my NT 4.0SP3 workstation and ran into a couple of gotchas: > - Telneting into your own machine appears to work correctly, but >telneting from some other machines (Suns running SunOS 4.1.3 in my > experiments) did not echo terminal input -- commands only appeared > after a return (you had to type blind). I compared stty -all outputs > and did not find anything different. I also tried telneting from my > PC to a Unix box then back to my PC; this also didn't work > correctly. [Telneting from Solaris (5.5.1) does work correctly.] > - stty rows x doesn't work. "Locally" (non-telnet), the command is > accepted but has no effect (as indicated by a subsequent stty -all). > When logged in through telnet, stty rows x gives "stty: standard > input: No such file or directory." > >-- >James Dumser 972.462.5335 dumser@ti.com >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Wed Aug 6 10:37:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Wed, 06 Aug 1997 10:37:00 -0000 Subject: login and telnetd Message-ID: <01BCA2AE.AC9079D0@sos> Jerome Gay wrote: > I still cannot have telnetd running... > Though rexecd is ok ,I can run rxvt via rexec, but then : It's a news for me :-) I didn't try rexecd :-) > - bash starts in directory '/HOME/GAY' (though $HOME=/home/gay), but if I > type 'cd' then 'pwd' I got '/home/gay' What is the real case stored on filesystem? > - my .bashrc is not executed (maybe linked to this uppercase/lowercase > problem for my home dir), Run "rxvt -e bash -login" > - I got lots of messages like 'free_tty: count < 0' in rxvt, How to reproduce the bug? > - does not work in rxvt, I have to type + > !... Depends of X server software you're using. BS works fine with StarNet's xwin32 and have the same nasty behavior with DEC eXcursion. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From lgd@atml.co.uk Wed Aug 6 10:37:00 1997 From: lgd@atml.co.uk (Luke Diamand) Date: Wed, 06 Aug 1997 10:37:00 -0000 Subject: Building GDB problem References: <2200011106081997/A04771/INDRE/11B832C03900*> <@MHS> Message-ID: <199708061714.SAA31728@inkcap.atml.co.uk> Hi! STEPHANE.S.P.PERIO@tts.thomson.fr writes: > Hi, > I apologize if this is an FAQ, but i see nothing about that in the mailing > lists. > I'm using gnu win32 b18 on NT 4.0 SP3 and i failed to build GDB. > Under bash, in the directory my-gdb, i run ../gdb/configure and it recognized > correctly the i386-pc-cygwin32 environment. I think your problem is that in order to build gdb, you also need to have already built bfd, mmalloc, readline, and various other bits and bobs. The simplest thing to do is to build the whole lot in one go - it takes a long time and a lot of disk space but is easier all round. You want to say: % mkdir /gnuwin32/b18/build % cd /gnuwin32/b18/build % ../cdk/configure % make Cheers! Luke Diamand -- Advanced Telecommunications Modules Ltd Cambridge - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gay@aspentec.com Wed Aug 6 10:37:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Wed, 06 Aug 1997 10:37:00 -0000 Subject: login and telnetd Message-ID: <9708061729.AA17633@hagar.aspentec.com> >Jerome Gay wrote: >> I still cannot have telnetd running... telnetd gives 'readped, status 0x100' each time I try to connect.... >> Though rexecd is ok ,I can run rxvt via rexec, but then : > >It's a news for me :-) I didn't try rexecd :-) Great ! >> - bash starts in directory '/HOME/GAY' (though $HOME=/home/gay), but if I >> type 'cd' then 'pwd' I got '/home/gay' > >What is the real case stored on filesystem? On file system (NTFS) it's stored in lowercase (/home/gay) >> - my .bashrc is not executed (maybe linked to this uppercase/lowercase >> problem for my home dir), > >Run "rxvt -e bash -login" Do you mean I've to start rxvt with these options each time ? >> - I got lots of messages like 'free_tty: count < 0' in rxvt, > >How to reproduce the bug? I'll try to find it out... >> - does not work in rxvt, I have to type + >> !... > >Depends of X server software you're using. BS works fine with StarNet's xwin32 and have the same nasty behavior with DEC eXcursion. I'm using ReflectionX 6.0 from WRQ Cheers, Jerome "ZoSo" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From shultsb@kenyon.edu Wed Aug 6 13:07:00 1997 From: shultsb@kenyon.edu (Benjamin Shults) Date: Wed, 06 Aug 1997 13:07:00 -0000 Subject: same install problem but solution does not work Message-ID: <33E8D917.F095CB6A@kenyon.edu> When I try to run cdk, I get the same "not enough room in TEMP" error that others have mentioned. I emptied TEMP and tried again...same error. I created a new TEMP directory and changed the value of the TEMP variable...same error. I have over 300MB on the disk. Please help. Benji - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fjh@cs.mu.OZ.AU Wed Aug 6 13:20:00 1997 From: fjh@cs.mu.OZ.AU (Fergus Henderson) Date: Wed, 06 Aug 1997 13:20:00 -0000 Subject: rm -rf -> Circular directory structure References: <199708060919.LAA15850@gil.physik.rwth-aachen.de> Message-ID: <199708061334.XAA06600@mundook.cs.mu.OZ.AU> Christoph Kukulies writes: >I'm having a weird problem under NT 4.0 (I remember to >have seen this with b17 and NT3.51 also): > >D:\>rm -rf somedir >rm: WARNING: Circular directory structure. >This almost certainly means that you have a corrupted file system. >NOTIFY YOUR SYSTEM MANAGER. >Cycle detected: >somedir/FOO/FOO >is the same file as >somedir This is because the algorithm that cygwin uses to compute inode numbers just hashes the file name (or something like that) rather than actually figuring out a properly unique number. When you get a hash collision, `rm' gets confused. -- Fergus Henderson | "I have always known that the pursuit WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From KMorris692@aol.com Wed Aug 6 14:33:00 1997 From: KMorris692@aol.com (KMorris692@aol.com) Date: Wed, 06 Aug 1997 14:33:00 -0000 Subject: trouble unpacking cdk-src Message-ID: <970806173212_596400449@emout14.mail.aol.com> I downloaded csd-src file xca thru xcj. I'm having trouble cat-ing them together. The result I get is too small. I suspect the files are being treated as text and cat is finding an end-of-file character, but that's only a guess on my part. Any advice on other methods? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fjh@cs.mu.OZ.AU Wed Aug 6 14:33:00 1997 From: fjh@cs.mu.OZ.AU (Fergus Henderson) Date: Wed, 06 Aug 1997 14:33:00 -0000 Subject: dllimport for gcc References: <199708041507.QAA28158@zaphod.long.harlequin.co.uk> Message-ID: <199708061330.XAA13972@mundook.cs.mu.OZ.AU> Jon Thackray writes: >Subject: dllimport for gcc >Is this going to be implemented for the win32 in beta-19? Is there >some other way of achieving the same effect? If `dllimport' means what I think it means, then yes, there is: #define foo (*__imp_foo) See < http://www.cs.mu.oz.au/~fjh/gnu-win32/gnu-win32/how-to-build-dlls.html >. -- Fergus Henderson | "I have always known that the pursuit WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From cgf@bbc.com Wed Aug 6 14:47:00 1997 From: cgf@bbc.com (Chris Faylor) Date: Wed, 06 Aug 1997 14:47:00 -0000 Subject: latest Sergey's update (08/01) crashes on Win95 References: <9708040956.AA28128@hagar.aspentec.com> <33E87C1B.9A2DF836@bellsouth.net> Message-ID: In article < 33E87C1B.9A2DF836@bellsouth.net >, David Means wrote: >Jerome Gay wrote: >>I just installed Sergey's latest coolview, dated Aug,1st... when I >>start bash, WIndows95 crahes completely (hard-reset necessary)... any >>ideas? > >Ditto me too. I've got a valid /etc/passwd and /etc/group. I have contributed some code to the latest coolview release which seems to be misbehaving on Windows 95. I am actively tracking it down now, but it looks like it is a really strange problem; possibly a bug in Windows 95. So, don't blame Sergey. It's probably my fault. -- http://www.bbc.com/ cgf@bbc.com "Strange how unreal VMS=>UNIX Solutions Boston Business Computing the real can be." - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From cgf@bbc.com Wed Aug 6 14:50:00 1997 From: cgf@bbc.com (Chris Faylor) Date: Wed, 06 Aug 1997 14:50:00 -0000 Subject: termcap issues... References: <9708061342.AA38522@hagar.aspentec.com> Message-ID: In article < 9708061342.AA38522@hagar.aspentec.com >, Jerome Gay wrote: >I'm using b18 with latest coolview (01-Aug) under NT4SP3... >I got some strange problems with termcap... >I've exported TERM=linux >If I export also TERMCAP=/etc/termcap (containing the linux entry), less >(3.32) behave strangely, it starts paging from the top of the sreen, >overwriting previous information... >Maybe that's because I use bash inside an NT Command Prompt with scroll-bar >(I set a buffer size of 1000 lines in the shortcut properties). I notice the same thing. Actually, I think it is a bug in less's handling of scrolling regions. It is definitely triggered by the scrollbar. >BUT, if I set TERMCAP=/etc/termcap/linux, less behaves _normally_, but with >a 'Warning: terminal not fully functional' You're pointing the TERMCAP environment variable to a non-existent file. That probably forces less to redraw the screen rather than attempt to scroll. That will work but it will be slow. -- http://www.bbc.com/ cgf@bbc.com "Strange how unreal VMS=>UNIX Solutions Boston Business Computing the real can be." - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pete@horus.cix.co.uk Wed Aug 6 17:04:00 1997 From: pete@horus.cix.co.uk (Pete Jordan) Date: Wed, 06 Aug 1997 17:04:00 -0000 Subject: latest Sergey's update (08/01) crashes on Win95 References: <33E87C1B.9A2DF836@bellsouth.net> Message-ID: dmeans@bellsouth.net (David Means) wrote: > Ditto me too. I've got a valid /etc/passwd and /etc/group. Here too (Win95 OSR2). > If I execute a command at the prompt, the command will run, i.e., mount > or /bin/ls, but the system will hang afterwards. The mouse still works, > but nothing will get the system back save a three-finger-salute ( > ctrl-alt-del ). Sometimes it will take a moment or two before it hangs, > at least it did after the mount command. Exactly the same symptoms. Pete Jordan = = = = = = = = = = = = = = = = = = = = = = = Horus Communications http://www.horus.cix.co.uk/ = = = = = = = = = = = = = = = = = = = = = = = "'Not twisted,' Salzy once said of her own passion, 'it is helical. That sounds better.'" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ashfielb@nbnet.nb.ca Wed Aug 6 21:17:00 1997 From: ashfielb@nbnet.nb.ca (Bruce Ashfield) Date: Wed, 06 Aug 1997 21:17:00 -0000 Subject: Imake Question ? Message-ID: <33E938C9.626A78D0@nbnet.nb.ca> Hi, I'll first give the standard disclaimer: I've searched through the mailing list archive, but didn't see (or missed) any reported problems similar to mine. Also, I've been using gnu-win32 since b17 and have been nothing but pleased. Now for the problem: Recently, I downloaded the X11R6.3 gnu-win32 port and installed it on my system. (NT4.0, running b18). Now that I've finally gotten rid of the application faults, etc. .. I want to compile/port some of my favorite work and Linux applications. My problem is with imake, more specifically it doesn't generate a valid Makefile. I'll admit that I'm not familiar with the guts of Imake, but I'm not a novice user, so I've played around with the templates, to the point where Imake runs. But .. generated Makefiles never correctly compile the application, since things like linking in the X libs are not present, or dependencies are messed up. It appears that others have X applications running and compiling, so I was hoping for some advice/flames/pointers to where I might have gone wrong. If any more info is needed I'd be more than happy to provide details. Thanks in advance, Bruce Ashfield - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From weiku@UTKUX1.UTK.EDU Wed Aug 6 21:17:00 1997 From: weiku@UTKUX1.UTK.EDU (Wei Ku) Date: Wed, 06 Aug 1997 21:17:00 -0000 Subject: No readme included in b18 Message-ID: <199708070417.AAA22996@solid.ssd.ornl.gov> The attached batch file setcygnu.bat is an example to set the necessary variable and call bash.exe. The other file .bash_profile is needed to be put in you home directory ( //D/MYDOCU~1/WEI/HOME in my case). You need to adjust the property of the batch file such that it allocate enough initial environment memory ( 768 in my case ) so that the set command won't cause 'out of environment variable space' error. Hope this help. Sincerely, Wei Ku *************************************** Department of Physics and Astronomy The University of Tennessee 1408 Circle Drive Knoxville, Tennessee 37996-1200 weiku@utkux.utcc.utk.edu --------------------------------------- Solid State Division Oak Ridge National Laboratory P.O.Box 2008 Oak Ridge, TN 37831-6032 Phone: (423) 574-5795 Fax: (423) 574-4143 weiku@solid.ssd.ornl.gov *************************************** -----Original Message----- From: Toby Gebhart To: gnu-win32@cygnus.com Date: Wednesday, August 06, 1997 11:55 PM Subject: RE: No readme included in b18 > >I noticed the same, but I found a readme.txt file with the needed variables in it, I'm just not sure how to incorporate them. I'm using Win95, so I thought that maybe I would just add the strings in the autoexec.bat in the following format: >GCC_EXEC_PREFIX=C:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\ >Obviously, this didn't work. Where exactly do I put the environment variables? I wish it was more clear, or maybe I'm missing the boat entirely...... I'm new to the list, please don't flame me away with "We've already answered this question 5 million times!". > >Toby Gebhart > mailto:gebhart@informatics.net >-----Original Message----- >From: Gregory Goodwin [SMTP:gm-goodwin@worldnet.att.net] >Sent: Tuesday, August 05, 1997 8:34 PM >To: gnu-win32@cygnus.com >Subject: No readme included in b18 > >I just downloaded/installed the full binary installation of build 18 for >Win95. The install program mentioned that I would need to view the readme >to set up the environment. However, the readme was not installed. Nor was >there a copy of the License Agreement. > >Thought you should know. > >Regards, >Greg >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > -------------- next part -------------- set GCC_EXEC_PREFIX=E:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\ set TCL_LIBRARY=C:/gnuwin32/b18/tcl/lib/tcl7.6 set GDBTK_LIBRARY=C:/gnuwin32/b18/share/gdbtcl set PATH=e:\gnuwin32\b18\H-i386-cygwin32\bin;E:\gnuwin32\b18\tcl\bin;E:\gnuwin32\b18\usr\local\bin;E:\gnuwin32\b18\usr\X11R6.3\bin;e:\mydocu~1\wei\home\bin;e:\mydocu~1\wei\home\lapw;%PATH% set LIBRARY_PATH=E:\Cygnus\b18\H-i386-cygwin32\lib set MAN_PATH=E:\Cygnus\man set C_INCLUDE_PATH=E:\Cygnus\include set CPLUS_INCLUDE_PATH=E:\Cygnus\include\g++ set HOME=//D/MYDOCU~1/WEI/HOME set DISPLAY=localhost:0.0 E:\gnuwin32\b18\H-i386-cygwin32\bin\bash.201.Sergey.exe -login rem E:\gnuwin32\b18\H-i386-cygwin32\bin\bash.201.exe -login rem E:\gnuwin32\b18\H-i386-cygwin32\bin\bash.200.exe -login rem E:\gnuwin32\b18\H-i386-cygwin32\bin\bash.exe -login From weiqigao@a.crl.com Wed Aug 6 21:17:00 1997 From: weiqigao@a.crl.com (Weiqi Gao) Date: Wed, 06 Aug 1997 21:17:00 -0000 Subject: Assembler References: <33E7EF73.121AB62E@tfs.net> <199708061118.MAA18965@zaphod.long.harlequin.co.uk> Message-ID: <33E929B2.70C77F2B@a.crl.com> I mailed this link to a follower-upper, should have mailed it to the list. Sorry. The Netwide Assembler (for Intel x86, DOS, Windows, Linux) is available from the following page: http://www.dcs.warwick.ac.uk/~jules/nasm1.html Fascinating! -- Weiqi Gao weiqigao@a.crl.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Wed Aug 6 22:42:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Wed, 06 Aug 1997 22:42:00 -0000 Subject: exc C0000005 when running from NT Command Prompt Message-ID: <01BCA316.4A0A0570@gater.krystalbank.msk.ru> Jerome Gay wrote: > When run from bash, my program runs fine, when run from NT Command Prompt, I > got an exception at the very end of the execution (at the very end of the > main()), traceback is : > > In cygwin_except_handler exc C0000005 at ... > Exception trapped! > exception C0000005 at ... > .. What is the trap address? It's very important - probably it will help to find why this version of cygwin.dll do not works on w95... -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From david@coent.demon.co.uk Wed Aug 6 23:39:00 1997 From: david@coent.demon.co.uk (David Coe) Date: Wed, 06 Aug 1997 23:39:00 -0000 Subject: lesstif 0.79 References: <33E7A15B.C7@pobox.com> Message-ID: <33E96D01.20FD@coent.demon.co.uk> Jeffery G. Smith wrote: > > I swear that someone had posted that lesstif had been ported under b18. > I was unable to find a reference in the list archives or any info on the > ported software link at Cygnus. As they often say on this list (:-) .. it compiles out of the box. My notes for lesstif-0.79 just say take out -g (otherwise the test set is **huge**) and -pipe (or gas fails). Seems to work fine on things like vim. Best wishes ] -- Dr David Coe \=\ 58 Fairlawn Drive, East Grinstead \=\ Tel +44 1342 326860 West Sussex, RH19 1NT, United Kingdom \=\ Fax +44 1342 316019 - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From david@coent.demon.co.uk Thu Aug 7 00:33:00 1997 From: david@coent.demon.co.uk (David Coe) Date: Thu, 07 Aug 1997 00:33:00 -0000 Subject: f77 of fc script for f2c References: Message-ID: <33E97987.4629@coent.demon.co.uk> Billinghurst, David wrote: > > Does anyone know of a working fc or f77 script for the f2c package. > There's one on the Scilab site: http://www-rocq.inria.fr/scilab/ and a full g77 ported by Mumit Khan on: http://www.xraylith.wisc.edu/~khan/software/gnu-win32/ -- Dr David Coe \=\ 58 Fairlawn Drive, East Grinstead \=\ Tel +44 1342 326860 West Sussex, RH19 1NT, United Kingdom \=\ Fax +44 1342 316019 - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gay@aspentec.com Thu Aug 7 00:33:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Thu, 07 Aug 1997 00:33:00 -0000 Subject: termcap issues... Message-ID: <9708070717.AA36967@hagar.aspentec.com> From: Larry Hall (RFK Partners Inc) >At 03:41 PM 8/6/97 +0200, you wrote: >> >>I'm using b18 with latest coolview (01-Aug) under NT4SP3... >>I got some strange problems with termcap... >>I've exported TERM=linux >>If I export also TERMCAP=/etc/termcap (containing the linux entry), less >>(3.32) behave strangely, it starts paging from the top of the sreen, >>overwriting previous information... >>Maybe that's because I use bash inside an NT Command Prompt with scroll-bar >>(I set a buffer size of 1000 lines in the shortcut properties). >>BUT, if I set TERMCAP=/etc/termcap/linux, less behaves _normally_, but with >>a 'Warning: terminal not fully functional' >> >>I'm using termcap 1.3 (compiled with -D_STRICT_ANSI to avoid error regarding >>bcopy and stuff...) >> >>Has anyone similar experience ? >>Any ideas ? >> >>Jerome "ZoSo" >> > >Are you using a gnuwin32 version of less or a Windows one? >I have some troubles with either.... I've built less-332 with gnuwin32-b18.... These problems only occur in windows w/ toolbars..... Jerome "ZoSo" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Janne.Leppdnen@metla.fi Thu Aug 7 01:29:00 1997 From: Janne.Leppdnen@metla.fi (Janne Leppdnen) Date: Thu, 07 Aug 1997 01:29:00 -0000 Subject: DirectX Message-ID: <33E9874E.5D40@metla.fi> Is it possible to use GNU-Win32 to make DirectX application?? And if yes where can I get import libraries or is there a tool which makes them from ".dll". And if no.. why it is not possible? And will it soon be? (answers via email, please) - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mark@solcom.co.uk Thu Aug 7 02:06:00 1997 From: mark@solcom.co.uk (Mark Hannah) Date: Thu, 07 Aug 1997 02:06:00 -0000 Subject: Beginners Problem 3 Message-ID: When trying to run "make install" I get the following messgae when building 'flex' :- cd /install/native/H-i386-cygwin32/bin && ln flex flex++ (c:\bin\sh.exe 1099) In cygwin_except_handler ork_helper: WaitForMultipleObjects timed out c:\bin\sh.exe 1099) In cygwin_except_handler ork_helper: child died before initialization with win32 error 0 bin/sh: fork: No more processes make: *** [install] Error 128 make: *** [install-flex] Error 2 Can anyone hold ? What am I doing so wrong ? Thanks in advance. Mark. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jsc@ppe.bb-data.de Thu Aug 7 03:06:00 1997 From: jsc@ppe.bb-data.de (Jerzy Schubert) Date: Thu, 07 Aug 1997 03:06:00 -0000 Subject: local paths and cd // i.e cd //g not working References: <9708061626.AA40482@hagar.aspentec.com> Message-ID: local paths seems not to work correctly, see below ls /usr vs. ls usr form within / also staying in f.e. /usr/local/sbin and doing cd //g then ls gives still the contents of /usr/local/sbin ( //g hapened to be / ) after that cd ..; ls gives the contens of the root directory *** 1 Problem *** mount => Device Directory Type Flags g:\gnuwin32\b18\H-i386-cygwin32 /usr native no-mixed,text=binary g:\home /home native no-mixed,text=binary \\.\tape1: /dev/st1 native no-mixed,text=binary \\.\tape0: /dev/st0 native no-mixed,text=binary \\.\b: /dev/fd1 native no-mixed,text=binary \\.\a: /dev/fd0 native no-mixed,text=binary g: / native no-mixed,text=binary ------------------------------------------------------------------------- in the root directory: ls /usr => X11R6.3 bin etc i386-cygwin32 lib libexec local man share ucb var adm local ------------------------------------------------------------------------- ls usr => ls: usr: No such file or directory ------------------------------------------------------------------------- *** 2 Problem *** sh-2.01$ cd /usr/local/sbin sh-2.01$ ls inetd.exe sh-2.01$ ls inetd.exe sh-2.01$ cd //g sh-2.01$ ls inetd.exe on the other hand switching drives from g to e works i.e cd //e; ls is ok, but then, cd //e/winnt/system32;ls gives the contents of system32 cd //e;ls STIL gives the contents of system 32 Any ideas? Jerzy - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jeffdbREMOVETHIS@netzone.com Thu Aug 7 07:58:00 1997 From: jeffdbREMOVETHIS@netzone.com (Mikey) Date: Thu, 07 Aug 1997 07:58:00 -0000 Subject: Problem with Sergey's telnetd, rxvt,vim References: <9708061626.AA40482@hagar.aspentec.com> Message-ID: <33e9e1bb.43885425@smtp.netzone.com> I don't know if this is the problem, but try pressing the first character of your password twice, on my win95 box the telnet prompt eats the first character. On Wed, 6 Aug 1997 18:25:21 +0200, you wrote: > Jerzy- > >I'm having the same kind of problems with telnetd and rxvt... >Did you solve this problems ? How ? >Thanks > >Jerome "ZoSo" > > >-----Original Message----- >From: Jerzy Schubert >To: Gilligan <_kunglao@prairienet.org> >Cc: GNU Win 32 >Date: Friday, July 25, 1997 1:56 AM >Subject: Problem with Sergey's telnetd, rxvt,vim > > > >>I had exactly (just the pids were different) error messages >>Gillian had here. >> >>I remembered then the advice from *jeffdb@netzone.nospam.com (Mikey) >> >> "after you change over to binary mode, you will need >> to reinstall from cdk.exe and user.exe .... " >> >>I've done that, then reinstall Sergey cygwin.dll and libcygwin.a >>and I'v got the login prompt: >> >>crusher:/ # telnet bewerly >>Trying 10.11.4.66... >>Connected to bewerly.ppe.bb-data.de. >>Escape character is '^]'. >> >>CYGWIN32/NT 4.0 (BEWERLY) (tty1) >> >>login: jsc >>Password: >>Login incorrect >>login: >> >>So one problem gone away. >>However i still cannot login >>login keeps saying "Login incorrect" >> >>Here my passwd and group file, i see that the groupid in the passwd file >>does not matched the groupid in the group files, those were made with >>the utility mkpasswd -l >>Administrator::500:513:::/bin/sh >>Gast::501:513:::/bin/sh >>IUSR_YUTA::1000:513:Internet-Gastkonto::/bin/sh >>jsc::1001:513:Jerzy Schubert:/home/jsc:/bin/sh >>root::1002:513:admin account:/home/root:/bin/sh >> >>Administratoren::544:0::: >>Benutzer::545:0::: >>Gdste::546:0::: >>Hauptbenutzer::547:0::: >>Replikations-Operator::552:0::: >>Sicherungs-Operatoren::551:0::: >> >>I could start inetd from com box or from bash observing the behavior >>above. >> >>When i tried to test service using autoexnt >> net start autoexnt >>with autoexnt.bat as: >>g:\gnuwin32\b18\H-i386-cygwin32\local\sbin\inetd -d >>Here i cannot say instead /usr/local/sbin/inetd -d can i >> >>autoexnt says in the same dos box: >> >>AutoExNT wird gestartet (starting) >>AutoExNT wurde erfolgreich gestartet (started successfully) >> >>In the TaskManager list i can indeed see inetd running >>but when i say telnet i get >>crusher:/ # telnet bewerly >>Trying 10.11.4.66... >>telnet: Unable to connect to remote host: Connection refused >> >>I stopped simple internet services, as suggested before >>but as i could started inetd by hand it probably has nothing to do with >>it. >> >> >>Other problem still here is with rxvt >>rxvt: Failed to open input method >>seteuid called, and it's not done yet. >>ptydev="/dev/tty2" >>ttydev="/dev/tty2" >>seteuid called and it's not done yet (4 times) >>i get rxvt window but nothing in it. (no shell) >>I use X11R6.3 binary provided by Arlindo da Silva build with b18. >> >> >>And i also have problem with vim in the shell box, which many >>people mentioned before, i did not understood the replay from Sergey >> >> >>Jerzy >> >> >> >>On Tue, 22 Jul 1997, Gilligan wrote: >> >>> Ok, I've setup my inetd.conf, copied login.w95 to login, ran inetd -d. >>> I can telnet in, but it crashes, perhaps looking for the login script. >>> Here's inetd's output: >>> >>> registered /usr/local/libexec/in.telnetd on 14 >>> someone wants telnet >>> accept, ctrl 3 >>> + Closing from 14 >>> 1060 execl /usr/local/libexec/in.telnetd >>> 1060 reaped, status 0x100 >>> >>> On the telnet client side, I get: >>> >>> Trying 192.168.1.2... >>> Connected to bozo. >>> Escape character is '^]'. >>> >>> CYGWIN32/95 4.0 (BOZO) (tty10) >>> >>> Connection closed by foreign host. >>> . >>> . >>> . >> >>- >>For help on using this list (especially unsubscribing), send a message to >>"gnu-win32-request@cygnus.com" with one line of text: "help". >> > >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > (jeffdbREMOVETHIS@netzone.com) delete REMOVETHIS from the above to reply Mikey - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From adg@dibe.unige.it Thu Aug 7 08:07:00 1997 From: adg@dibe.unige.it (alessandro de gloria) Date: Thu, 07 Aug 1997 08:07:00 -0000 Subject: gcc and gnu-win32 Message-ID: <33E9EAEC.A1B28481@dibe.unige.it> Dear Sir, I got the gcc source from a gnu site and I tried to compile it with gnu-winnt32 under nt-4.0 The configure doesn't recognize the host. If I force the host as a i386-winnt the compilation abort. What can I do ? Best Regards, Alessandro De Gloria Prof. Alessandro De Gloria University of Genoa -DIBE Via Opera Pia 11A 16145 Genova - Italy Tel & Fax +39-10-3532785 e-mail: adg@dibe.unige.it - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From rsc@vingmed.no Thu Aug 7 08:07:00 1997 From: rsc@vingmed.no (Robert Schmidt) Date: Thu, 07 Aug 1997 08:07:00 -0000 Subject: ls is too slow Message-ID: <33E9BE4F.6AACE5BC@vingmed.no> ls is too slow, especially when doing a recursive (-R) list, because each and every file is opened to see if it is a symbolic link. Is it possible to disable this behaviour? I don't use symlinks yet, anyway. -- Robert Schmidt Software Developer / Vingmed Sound tel +47 67124237 fax +47 67124355 Private tel +47 22352293 WWW http://www.nvg.unit.no/~rsc jeg lukker et oye og ser halvt jeg lukker begge og ser alt -- seigmen - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From wdoyle@bode.cdsp.neu.edu Thu Aug 7 08:07:00 1997 From: wdoyle@bode.cdsp.neu.edu (Patrick Doyle) Date: Thu, 07 Aug 1997 08:07:00 -0000 Subject: Microsoft Networking Message-ID: <199708071314.JAA02240@cdsp.neu.edu> Does anybody out there use CYGWIN32 on a set of computers networked using Microsoft networking? We are experiencing a problem in which files accessed over the network have totally bogus sizes (3236843376 for example). Is this a known problem with Microsoft networking? Is it a known problem with B18? Does Sergey's updated CYGWIN.DLL fix this? Any help and/or replies would be greatly appreciated. -- --patrick - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jont@harlequin.co.uk Thu Aug 7 08:34:00 1997 From: jont@harlequin.co.uk (Jon Thackray) Date: Thu, 07 Aug 1997 08:34:00 -0000 Subject: dllimport for gcc References: <199708041507.QAA28158@zaphod.long.harlequin.co.uk> <199708061330.XAA13972@mundook.cs.mu.OZ.AU> Message-ID: <199708071053.LAA19915@zaphod.long.harlequin.co.uk> Fergus Henderson writes: > Jon Thackray writes: > > >Subject: dllimport for gcc > >Is this going to be implemented for the win32 in beta-19? Is there > >some other way of achieving the same effect? > > If `dllimport' means what I think it means, then yes, there is: > > #define foo (*__imp_foo) Perhaps you mean #define foo (*_imp_foo) as presumably the C compiler will automatically prepend the remaining _. > See < http://www.cs.mu.oz.au/~fjh/gnu-win32/gnu-win32/how-to-build-dlls.html >. I think you meant < http://www.cs.mu.oz.au/~fjh/gnu-win32/how-to-build-dlls.html > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dmeans@bellsouth.net Thu Aug 7 08:34:00 1997 From: dmeans@bellsouth.net (David Means) Date: Thu, 07 Aug 1997 08:34:00 -0000 Subject: latest Sergey's update (08/01) crashes on Win95 References: <9708040956.AA28128@hagar.aspentec.com> <33E87C1B.9A2DF836@bellsouth.net> Message-ID: <33E9EB07.7DCA5022@bellsouth.net> Chris Faylor wrote: { snip ... } > I have contributed some code to the latest coolview release which > seems > to be misbehaving on Windows 95. { snip } > > So, don't blame Sergey. It's probably my fault. No worries! ;-) David - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jont@harlequin.co.uk Thu Aug 7 08:34:00 1997 From: jont@harlequin.co.uk (Jon Thackray) Date: Thu, 07 Aug 1997 08:34:00 -0000 Subject: Assembler References: <33E8A9A1.587F@softway.com> Message-ID: <199708071046.LAA19678@zaphod.long.harlequin.co.uk> Jason Zions writes: > If you're still looking at x86 processors, the CPU reference manuals for > Intel's current chips are available in downloadable PDF form on Intel's > web site. You'll still kill a tree printing it out, but there's no > middleman driving up costs. (The Pentium MMX manual was about 800 pages; > be a hero and print it double-sided. :-) Care to give an exact URL for this? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From rsc@vingmed.no Thu Aug 7 08:34:00 1997 From: rsc@vingmed.no (Robert Schmidt) Date: Thu, 07 Aug 1997 08:34:00 -0000 Subject: exc C0000005 when running from NT Command Prompt References: <01BCA316.4A0A0570@gater.krystalbank.msk.ru> Message-ID: <33E9BDB9.FF520146@vingmed.no> Sergey Okhapkin wrote: > > Jerome Gay wrote: > > When run from bash, my program runs fine, when run from NT Command > Prompt, I > > got an exception at the very end of the execution (at the very end > of the > > main()), traceback is : > > > > In cygwin_except_handler exc C0000005 at ... > > Exception trapped! > > exception C0000005 at ... > > .. > > What is the trap address? It's very important - probably it will help > to find why this version of cygwin.dll do not works on w95... I also often get this exception when using gnu-win32 tools in NT 4.0, but they only occur when I use them from within 4NT batch programs. I don't think it has ever happened when executing from the command line. The exceptions occuring with the b18 tools have bothered me so much that I've had to replace the tools I use most (grep, diff, tc and ls) with other NT ports. By the way - how does uwin32 (or whatever its name is) compare to gnu-win32? -- Robert Schmidt Software Developer / Vingmed Sound tel +47 67124237 fax +47 67124355 Private tel +47 22352293 WWW http://www.nvg.unit.no/~rsc jeg lukker et oye og ser halvt jeg lukker begge og ser alt -- seigmen - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From giac@dalim.de Thu Aug 7 09:45:00 1997 From: giac@dalim.de (Philippe GIACINTI) Date: Thu, 07 Aug 1997 09:45:00 -0000 Subject: exc C0000005 when running from NT Command Prompt References: <9708061220.AA34509@hagar.aspentec.com> Message-ID: <9708071523.ZM988@pion.kehl.dalim.de> Hi, Are you sure you have replaced all "cygwin.dll" in your environment ? When you start your application within bash, you are using the cygwin.dll which is in same directory as bash ...\H-i386-cygwin32\bin, BUT when you start it from NT Command prompt may be another cygwin.dll ( in SYSTEM32 or SYSTEM ) is loaded. Best regards. -- ========================================================================== | Philippe GIACINTI | | | | DALiM GmbH R&D email: giac@dalim.de | | Daimler Strasse 2, tel: +49.7851.9196-28 | | D-77694 Kehl-Sundheim Germany fax: +49.7851.73576 | | | ========================================================================== - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dumser@ti.com Thu Aug 7 09:45:00 1997 From: dumser@ti.com (James Dumser) Date: Thu, 07 Aug 1997 09:45:00 -0000 Subject: rexecd strangeness Message-ID: rexecd does work, but there appear to be problems with pty handling in this case. To test, I'm using eXceed's Xstart to start rxvt (/usr/local/X11R6.3/bin/rxvt -ls). With the "show host response" option set, I get rxvt: Failed to open input method seteuid called, and it's not done yet. ptydev="/dev/tty4" ttydev="/dev/tty4" as rxvt is coming up. Then for every command I issue from rxvt, I get three "free tty: count < 0" messages before execution and two after: sh$ ls foo free tty: count < 0 free tty: count < 0 free tty: count < 0 ls: foo: No such file or directory free tty: count < 0 free tty: count < 0 (it does not make any difference what command is executed). ps verifies that I am on pty4: TTY PID PPID PGID WIN32-PID UID COMMAND 00 1000 1000 0 166 1007 d:\local\sbin\inetd.exe 02 1282 1282 0 89 1007 d:\local\sbin\inetd.exe 03 1288 1288 0 274 1007 d:\local\sbin\inetd.exe 04I 1297 1288 0 230 1007 d:\local\X11R6.3\bin\rxvt.exe 04 1298 1297 1298 250 1007 c:\bin\sh.exe 04 1306 1298 1306 159 1007 ...\H-i386-cygwin32\bin\ps.exe but who does not show anything on tty4 (as it does if I telneted in): A0207776 tty0 Aug 01 15:03 (local) A0207776 tty1 Aug 07 09:00 (local) A0207776 tty2 Aug 07 09:02 (local) A0207776 tty3 Aug 07 09:14 (local) One other thing I've noticed in using rxvt: Normally, I have a nice block cursor, but starting vim makes it disappear -- it's kind of hard to run vim without a cursor! :) Is there some xrdb entry I can set to prevent this, or is there another way around it? (Again, using eXceed for NT 5.1.) -- James Dumser 972.462.5335 dumser@ti.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Thu Aug 7 09:45:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Thu, 07 Aug 1997 09:45:00 -0000 Subject: latest Sergey's update (08/01) crashes on Win95 Message-ID: <01BCA354.4ACF9630@gater.krystalbank.msk.ru> Chris Faylor wrote: > >Ditto me too. I've got a valid /etc/passwd and /etc/group. > > I have contributed some code to the latest coolview release which seems > to be misbehaving on Windows 95. I am actively tracking it down now, but > it looks like it is a really strange problem; possibly a bug in Windows 95. > While searching for a win95-specific problems in the last cygwin.dll, I have found the following bugs in Windows95: 1. The application hangs when tries to resume non-suspended thread. Workaround: suspend the thread before resuming :-) 2. MsgWaitForMultipleObjects() returns WAIT_FAILED with errno ERROR_INVALID_HANDLE instead of WAIT_OBJECT_0+N when the waited process is killed with three-finger salute. Workaround: just wait again :-) 3. InterlockedIncrement()/InterlockedDecrement() returns the temperature on the Shri-Lanka coast instead of operation result :-) This results in incorrect tty allocation and the messages like "free_tty: count < 0". I don't know a good workaround, and didn't fix it yet. I think, the tty support should be completely disabled on #$%@'95. > So, don't blame Sergey. It's probably my fault. Try to write to askbill@microsoft.com :-) -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From giac@dalim.de Thu Aug 7 09:45:00 1997 From: giac@dalim.de (Philippe GIACINTI) Date: Thu, 07 Aug 1997 09:45:00 -0000 Subject: dll building : segmentation faults References: <19970806130447.28713@bsilx1.iitb.fhg.de> Message-ID: <9708071517.ZM982@pion.kehl.dalim.de> Hi, 1) The fixup asm(".section .idata$3 ...") is here to terminate the Import Directory Table in which each element is 5x32 bits long and which should be terminated with a NULL element. So I think the correct fixup should be asm(".section .idata$3\n" ".long 0,0,0,0,0"); only 5 zeros ... 2) May be can you try to put the fixup at end of your file, I had problems which were solved that way ... ( don't ask me why ) #include /* asm(".section .idata$3\n" ".long 0,0,0,0,0,0,0,0"); not here */ WINAPI dll_entry(int a,int b,int c) { return 1; } void dummy() { return ; } /* put it here */ asm(".section .idata$3\n" ".long 0,0,0,0,0"); 3) I had also some troubles with strange relocations that I have solved in adding a third link pass ... Hope this helps, Best regards. -- ========================================================================== | Philippe GIACINTI | | | | DALiM GmbH R&D email: giac@dalim.de | | Daimler Strasse 2, tel: +49.7851.9196-28 | | D-77694 Kehl-Sundheim Germany fax: +49.7851.73576 | | | ========================================================================== - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From lhall@rfk.com Thu Aug 7 09:45:00 1997 From: lhall@rfk.com (Larry Hall (RFK Partners Inc)) Date: Thu, 07 Aug 1997 09:45:00 -0000 Subject: local paths and cd // i.e cd //g not working Message-ID: <2.2.32.19970807164311.009432b8@ma.ultranet.com> At 12:04 PM 8/7/97 +0200, Jerzy Schubert wrote: >local paths seems not to work correctly, see below ls /usr vs. ls usr >form within / >also staying in f.e. /usr/local/sbin >and doing cd //g >then ls gives still the contents of /usr/local/sbin >( //g hapened to be / ) >after that cd ..; ls gives the contens of the root directory > Jerzy, Try making a /usr directory, for example. I think you'll find "ls usr" works OK then (I can't tell for sure because I'm at work with only b17 and it only kinda works here). Keep in mind that the basis for all of these tools is a UNIX environment. In UNIX, you cannot mount a directory to a location without creating that location first. The Cygnus version of mount is a little peculiar in that sense since it doesn't require the mounting directory to exist first. Regardless, the tools will require that the files and directories you give them exist. If what you type matches the mount point exactly, Cygnus will resolve it anyway. Otherwise, you need to provide the structure to make it work. Symbolic links may be another way to solve this problem... Larry Hall lhall@rfk.com RFK Partners, Inc. (617) 239-1053 8 Grove Street (617) 239-1655 - FAX Wellesley, MA, 02181 - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mchase@ix.netcom.com Thu Aug 7 09:45:00 1997 From: mchase@ix.netcom.com (Michael A. Chase) Date: Thu, 07 Aug 1997 09:45:00 -0000 Subject: rm -rf -> Circular directory structure Message-ID: <199708071351.IAA23425@dfw-ix1.ix.netcom.com> I'm using Sergey's cygwin.dll dated 31 July and bash.exe dated 4 July. When I try to delete a large directory tree, I still see this problem. Mac :}) -----Original Message----- From: Larry Hall To: Christoph Kukulies ; gnu-win32@cygnus.com Date: Wednesday, August 06, 1997 10:54 Subject: Re: rm -rf -> Circular directory structure >At 11:19 AM 8/6/97 +0200, Christoph Kukulies wrote: >>I'm having a weird problem under NT 4.0 (I remember to >>have seen this with b17 and NT3.51 also): >> >>D:\>rm -rf somedir >>rm: WARNING: Circular directory structure. >>This almost certainly means that you have a corrupted file system. >>NOTIFY YOUR SYSTEM MANAGER. >>Cycle detected: >>somedir/FOO/FOO >>is the same file as >>somedir >> >>CHKDSK doesn't flag any errors on the filesystem and I tend to believe >>this isn't the cause anyway. >> >>The interesting thing is: When I remove the file somedir/FOO/FOO >>the rm command is able to run through without errors. >> >> >>-- >>Chris Christoph P. U. Kukulies kuku@gil.physik.rwth-aachen.de > >Chris, > >This should be in the FAQ (I expect its not at the moment). The problem >is caused by a faulty inode mapping algorithm in all the betas. There is >no problem with your file system. It merely means that the faulty inode >mapping for files and directories thinks there's more than 1 file and/or >directory for the generated inode number, which is certainly a bad thing >for traditional UNIX file systems. The error hinders rm from working >correctly but does not indicate file system problems. You'll just need >to delete the intended files/directories another way (most likely via >Windows or DOS tools). I believe Sergey's latest DLL has some fixes to >mitigate this problem. I don't think there's a complete fix yet (correct >me if I'm wrong folks). > >Larry Hall lhall@rfk.com >RFK Partners, Inc. (617) 239-1053 >8 Grove Street (617) 239-1655 - FAX >Wellesley, MA 02181 > >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From DKATSULIS@worldnet.att.net Thu Aug 7 12:40:00 1997 From: DKATSULIS@worldnet.att.net (DIANE KATSULIS) Date: Thu, 07 Aug 1997 12:40:00 -0000 Subject: documentation Message-ID: <33EA24F9.3A54@WORLDNET.ATT.NET> - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From wbach@iicm.tu-graz.acdot Thu Aug 7 12:41:00 1997 From: wbach@iicm.tu-graz.acdot (wbach@iicm.tu-graz.acdot) Date: Thu, 07 Aug 1997 12:41:00 -0000 Subject: patch - strange behaviour Message-ID: Hi gnuwinners, I've tried to apply Mumit Khan's patches to b18 and experienced very strage behaviour of patch: patch aborts a after some patches, typically a few files, with the message "Assertion hunk failed ..." (don't remember exact wording). If I split the patch file into seven parts, I can apply the separate patch files without problems. I can live with that, but it's inconvenient. Has anyone got an idea what's going on? TIA. Cheers, Wolfgang - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jsc@ppe.bb-data.de Thu Aug 7 12:41:00 1997 From: jsc@ppe.bb-data.de (Jerzy Schubert) Date: Thu, 07 Aug 1997 12:41:00 -0000 Subject: local paths and cd // i.e cd //g not working References: <2.2.32.19970807164311.009432b8@ma.ultranet.com> Message-ID: Unfortunately this is not the case from the dos box: g: (g is the root directory) umount /usr mkdir usr mount -b g:\gnuwin32\b18\H-i386-cygwin32 /usr ls usr gives nothing ls /usr works as before same story after starting shell sh cd / ls usr gives nothing i tried to change in regedit mounting for root from g: / to g:\ / but mount output looks the same then i logged out and in, still the same the i rebooted the machine, logged in under same account and all the mount points were effektively lost mount gives the default mount points from HK Users-Default-Software-cygnus and so on up to mount points with defaults values f.e. for / c: text setting the root in Current User has no effect. However THIS default setting for root i could change with umount / mount -b g:\gnuwin32\b18\H-i386-cygwin32 / normally those two actions had not much effect. umount / was silent mount -b / said always "busy" any ideas? On Thu, 7 Aug 1997, Larry Hall (RFK Partners Inc) wrote: > At 12:04 PM 8/7/97 +0200, Jerzy Schubert wrote: > >local paths seems not to work correctly, see below ls /usr vs. ls usr > >form within / > >also staying in f.e. /usr/local/sbin > >and doing cd //g > >then ls gives still the contents of /usr/local/sbin > >( //g hapened to be / ) > >after that cd ..; ls gives the contens of the root directory > > > > > Jerzy, > > Try making a /usr directory, for example. I think you'll find "ls usr" > works OK then (I can't tell for sure because I'm at work with only b17 > and it only kinda works here). Keep in mind that the basis for all of > these tools is a UNIX environment. In UNIX, you cannot mount a > directory to a location without creating that location first. The > Cygnus version of mount is a little peculiar in that sense since it > doesn't require the mounting directory to exist first. Regardless, > the tools will require that the files and directories you give them > exist. If what you type matches the mount point exactly, Cygnus will > resolve it anyway. Otherwise, you need to provide the structure to > make it work. Symbolic links may be another way to solve this > problem... > > Larry Hall lhall@rfk.com > RFK Partners, Inc. (617) 239-1053 > 8 Grove Street (617) 239-1655 - FAX > Wellesley, MA, 02181 > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pete@horus.cix.co.uk Thu Aug 7 12:41:00 1997 From: pete@horus.cix.co.uk (Pete Jordan) Date: Thu, 07 Aug 1997 12:41:00 -0000 Subject: latest Sergey's update (08/01) crashes on Win95 References: Message-ID: cgf@bbc.com (Chris Faylor) wrote: > possibly a bug in Windows 95. How could you possibly even /suggest/ such a thing?... >;) Pete Jordan = = = = = = = = = = = = = = = = = = = = = = = Horus Communications http://www.horus.cix.co.uk/ = = = = = = = = = = = = = = = = = = = = = = = "'Not twisted,' Salzy once said of her own passion, 'it is helical. That sounds better.'" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From DKATSULIS@worldnet.att.net Thu Aug 7 12:41:00 1997 From: DKATSULIS@worldnet.att.net (DIANE KATSULIS) Date: Thu, 07 Aug 1997 12:41:00 -0000 Subject: compiling windows apps Message-ID: <33EA2537.4E33@WORLDNET.ATT.NET> I'm looking for more information on compiling windows applications with the gnu-win32 cdk. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jdennis@ultranet.com Thu Aug 7 12:41:00 1997 From: jdennis@ultranet.com (John R. Dennis) Date: Thu, 07 Aug 1997 12:41:00 -0000 Subject: Assembler References: <199708071046.LAA19678@zaphod.long.harlequin.co.uk> Message-ID: <199708071814.OAA21740@elektra.ultra.net> >>>>> "Jon" == Jon Thackray writes: Jon> Jason Zions writes: >> If you're still looking at x86 processors, the CPU reference >> manuals for Intel's current chips are available in downloadable >> PDF form on Intel's web site. You'll still kill a tree printing >> it out, but there's no middleman driving up costs. (The Pentium >> MMX manual was about 800 pages; be a hero and print it >> double-sided. :-) Jon> Care to give an exact URL for this? I'm not the original poster, but its: http://developer.intel.com/design/perftool/cbts/ John - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From topcoder@mindspring.com Thu Aug 7 17:55:00 1997 From: topcoder@mindspring.com (Pam Greene) Date: Thu, 07 Aug 1997 17:55:00 -0000 Subject: Assembler References: <33E8A9A1.587F@softway.com> <199708071046.LAA19678@zaphod.long.harlequin.co.uk> Message-ID: <3.0.3.32.19970807085845.006a916c@mail.mindspring.com> www.x86.org has all the manual's At 11:46 AM 8/7/97 +0100, you wrote: >Jason Zions writes: > > If you're still looking at x86 processors, the CPU reference manuals for > > Intel's current chips are available in downloadable PDF form on Intel's > > web site. You'll still kill a tree printing it out, but there's no > > middleman driving up costs. (The Pentium MMX manual was about 800 pages; > > be a hero and print it double-sided. :-) > >Care to give an exact URL for this? >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mbenz@sabre.com Thu Aug 7 18:21:00 1997 From: mbenz@sabre.com (Micheal A. Benzinger) Date: Thu, 07 Aug 1997 18:21:00 -0000 Subject: New patch available References: <9708061640.AA38100@hagar.aspentec.com> Message-ID: <3.0.3.32.19970807202120.00912630@pop.flash.net> Jerome, I, too, had problems getting telnetd using NT4SP3 as you. To rectify these problems I took two steps: 1) I had to add user rights as specified at the bottom of Sergey's home page to the user that was used to start up the inetd service. 2) I had to ensure that I had a valid /etc/passwd and /etc/group and, in my case, they had to be UNIX type files with only linefeeds and not DOS type files with lines feeds and carriage returns. After I made these two changes I was able to log in from any of our UNIX hosts as well as other NT work stations. I hope this helps, Mike Benzinger P.S. I was able to get VIM 5.0l to compile and run after applying the patches made by Chris Faylor in version 5.0h and the modification to vim.h that I found. After this VIM seems to be rolling along fairly nicely except for the fact that at work I am experiencing some problems with "dd" and "j" not updating the screen when deleting a line or scrolling at the bottom of the screen. At 06:39 PM 8/6/97 +0200, Jerome Gay wrote: > James- > >Sorry to bother you, but I cannot get telnetd to work on NT4SP3... >Did you tweak something to get it work ? >Did you set some special user rights in NT ? >How are you running inetd (as a service, or from command line) ? >I'll appreciate your help. > >Jerom "ZoSo" > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From venkman1@ix.netcom.com Thu Aug 7 21:10:00 1997 From: venkman1@ix.netcom.com (Narayan Natarajan) Date: Thu, 07 Aug 1997 21:10:00 -0000 Subject: Beginner's Question: history command Message-ID: <3.0.1.32.19970808000956.00918840@popd.ix.netcom.com> Hi, My installation of bash does not seem to read the ~/.bash_history although it writes to it each time I quit bash. I tried therefore to initiate the history by: history -r ~/.bash_history to see if it loads the existing history and this failed. I did read the faq's and man pages etc. nothing seems to have any more info on this. Any help appreciated. Also, I realize that my question may not be of a sufficiently "techie" nature to be answered or even read, and so I would like to know if there is another mailing group which may be less techie where such unfortunates as me may recieve some help. Narayan | Narayan _/~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~/ |Natarajan _/ venkman1@ix.netcom.com http://www.netcom.com/~venkman1 _/ | _/ (810)-475-0667 . .____ _ _/ | _/ "The heavens call to you, and * . x -(____/ | _ / | _/ circle about you, displaying to ________/____/ \ ( )----- | _/ you their eternal splendors, and ,/ _ ( ------- | _/ your eye gazes only to earth." <________\ // (_)----- | _/ -- Dante --<== * |_____/ _ |/ -(_____) (_)----- |~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From digit@postoffice.ptd.net Thu Aug 7 21:44:00 1997 From: digit@postoffice.ptd.net (John Northrop) Date: Thu, 07 Aug 1997 21:44:00 -0000 Subject: libxpm.dll Message-ID: <199708080444.VAA15991@cygnus.com> I have b18 installed and working rather well. A problem I encountered while trying to build XPM was a reference to "libxpm.dll". If someone could let me know how to correct this I would appreciate it. Thanks John - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jazz@softway.com Thu Aug 7 22:49:00 1997 From: jazz@softway.com (Jason Zions) Date: Thu, 07 Aug 1997 22:49:00 -0000 Subject: Assembler References: <33E8A9A1.587F@softway.com> <199708071046.LAA19678@zaphod.long.harlequin.co.uk> Message-ID: <33EAA35C.217B@softway.com> >> If you're still looking at x86 processors, the CPU reference manuals for >> Intel's current chips are available in downloadable PDF form on Intel's >> web site. > > Care to give an exact URL for this? http://developer.intel.com/design/pro/MANUALS/242691.HTM Jason - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From greg.maynard@cimtegration.com Thu Aug 7 22:49:00 1997 From: greg.maynard@cimtegration.com (Greg Maynard) Date: Thu, 07 Aug 1997 22:49:00 -0000 Subject: Error : Ld : cannot load '-lgcc' no such file or directory Message-ID: <32098040.5E78780D@cimtegration.com> I have been working with GCC for the past couple days, basically tweaking things out because I am very slow at getting things to WORK properly ;).. Anyhow, when I run gcc it gives me the error 'Ld : cannot load '-lgcc' no such file or directory' and I don't know where the heck that is coming from or how I can fix it.. Someone must know ;).. Any help is appreciated.. Thanks in advance, Greg. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From iverson@cisco.com Thu Aug 7 22:49:00 1997 From: iverson@cisco.com (Tim Iverson) Date: Thu, 07 Aug 1997 22:49:00 -0000 Subject: Default alignment? Message-ID: <199708080505.WAA00263@stars.cisco.com> I need to change the default alignment for all structures and unions from 4 bytes to 2 bytes. Is there a handy -falign=2 option? I didn't see one in the info files, but that doesn't mean it isn't there. I also tried #pragma pack(2) -- it would be fairly easy to just put that into the top level headers. However it doesn't seem to work here, even though MSC supports it and gcc does too on Unixen that have it. It is simply too large a task to go and mark every structure and union with __attribute__ ((aligned (2), packed)). Naturally, I'd prefer not to rebuild GCC just to add a new -f option. Anyone got some good news for me on this? ;-) Thanks, - Tim Iverson iverson@cisco.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Thu Aug 7 23:58:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Thu, 07 Aug 1997 23:58:00 -0000 Subject: rm -rf -> Circular directory structure Message-ID: <01BCA3EA.21533FF0@gater.krystalbank.msk.ru> Michael A. Chase wrote: > I'm using Sergey's cygwin.dll dated 31 July and bash.exe dated 4 July. > When I try to delete a large directory tree, I still see this problem. > The hash function used to convert filename to inode number is very simple. Do you know any fast and good hash? -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From denis.trcek@ijs.si Fri Aug 8 00:10:00 1997 From: denis.trcek@ijs.si (DT) Date: Fri, 08 Aug 1997 00:10:00 -0000 Subject: Sockets/POSIX etc. Message-ID: <33EAC6EE.65F2@ijs.si> Dear all, has anybody succesfully ported SSLeay0.6.6 to GniWin32? If so, please le me know, how SO_ERROR is to be handeled (and FILE *SSL_LOG=stderr; FILE *SSL_ERR=stderr; in ssl_lib.c, as well). I' ve succeded to compile it with some modification, but crypto routines crash at run-time... On the other hand, the program can be compiled on Solaris and Linux without problems. Thanks for response. D. Trcek IJS/E6, Ljubljana, SI - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Don.Sharp@dddandr.octacon.co.uk Fri Aug 8 01:00:00 1997 From: Don.Sharp@dddandr.octacon.co.uk (Don Sharp) Date: Fri, 08 Aug 1997 01:00:00 -0000 Subject: Beginner's Question: history command References: <3.0.1.32.19970808000956.00918840@popd.ix.netcom.com> Message-ID: <33EAD1B7.2F01@dddandr.octacon.co.uk> As I recall this is because bash attempts to read the file and compares the result with the stat size. The CR/LFs distort this picture and so bash decides that something went wrong and it better ignore it. You probably need to make your mounts binary. Can't we have default binary mounts please. Cheers Don Sharp Narayan Natarajan wrote: > > Hi, > > My installation of bash does not seem to read the ~/.bash_history although > it writes to it each time I quit bash. I tried therefore to initiate the > history by: > history -r ~/.bash_history > to see if it loads the existing history and this failed. > > I did read the faq's and man pages etc. nothing seems to have any more info > on this. > > Any help appreciated. > > Also, I realize that my question may not be of a sufficiently "techie" > nature to be answered or even read, and so I would like to know if there is > another mailing group which may be less techie where such unfortunates as > me may recieve some help. > > Narayan > | Narayan _/~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~/ > - > For help on using this list (especially unsubscribing), send a message to > "gnu-win32-request@cygnus.com" with one line of text: "help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From iverson@cisco.com Fri Aug 8 01:38:00 1997 From: iverson@cisco.com (Tim Iverson) Date: Fri, 08 Aug 1997 01:38:00 -0000 Subject: Debugging info for commercial linkers? Message-ID: <199708080834.BAA01228@stars.cisco.com> I just noticed that my linker (PharLap) isn't picking up debugging information from the GCC compiled objects. I'm using -gcoff (obviously stabs won't work). According to objdump, the output file looks like it has all the right debug stuff in it, but the linker is ignoring it. Is coff debug format just too different from pe-i386? I was under the impression that the formats were almost identical. Since the linker isn't dying on the coff debug records, I suspect it isn't too different. Anyone wanna point me to a pe-i386 format reference so I can fix it? Please don't tell me the -gcoff format is correct for pe-i386 debugging; that would mean I'll have to add OMF support to BFD -- I'm stuck with the linker. ;-( Thanks, - Tim Iverson iverson@cisco.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jont@harlequin.co.uk Fri Aug 8 03:24:00 1997 From: jont@harlequin.co.uk (Jon Thackray) Date: Fri, 08 Aug 1997 03:24:00 -0000 Subject: Error : Ld : cannot load '-lgcc' no such file or directory References: <32098040.5E78780D@cimtegration.com> Message-ID: <199708081019.LAA11268@zaphod.long.harlequin.co.uk> Greg Maynard writes: > I have been working with GCC for the past couple days, basically > tweaking things out because I am very slow at getting things to WORK > properly ;).. Anyhow, when I run gcc it gives me the error 'Ld : cannot > load '-lgcc' no such file or directory' and I don't know where the heck > that is coming from or how I can fix it.. Someone must know ;).. Any > help is appreciated.. -lgcc means it is looking for libgcc.a. That can be found in gcc-lib/i386-cygwin32/cygnus-2.7.2-961023 (at least that's where it is for beta 17.1). The real question though, is why are you requiring libgcc.a at all? You might try nm libgcc.a to see if any of the symbols exported (those marked T or ?) are in your code. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pete@horus.cix.co.uk Fri Aug 8 04:05:00 1997 From: pete@horus.cix.co.uk (Pete Jordan) Date: Fri, 08 Aug 1997 04:05:00 -0000 Subject: patch - strange behaviour References: Message-ID: wbach@iicm.tu-graz.ac.at (Wolfgang Bachmann) wrote: > patch aborts a after some patches, typically a few files, with the > message "Assertion hunk failed ..." (don't remember exact wording). > > If I split the patch file into seven parts, I can apply the separate > patch files without problems. I can live with that, but it's > inconvenient. > > Has anyone got an idea what's going on? Yep. The patch (2.1) that comes with GNU-Win32 b18 is broken - I got exactly the same problem with Chris faylor's Perl patches. Grab and build patch 2.4 from your friendly local GNU archive. Apropos nothing, from the patch ChangeLog: > Thu Jun 10 21:13:47 1993 Paul Eggert (eggert@twinsun.com) > > * patchlevel.h: PATCH_VERSION 2.1. > (The name `patch-2.0.12g12' is too long for traditional Unix.) Well, it's as good a reason to bump up from 2.0.x to 2.1 as any, I suppose... Pete Jordan = = = = = = = = = = = = = = = = = = = = = = = Horus Communications http://www.horus.cix.co.uk/ = = = = = = = = = = = = = = = = = = = = = = = "'Not twisted,' Salzy once said of her own passion, 'it is helical. That sounds better.'" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From wdoyle@bode.cdsp.neu.edu Fri Aug 8 06:09:00 1997 From: wdoyle@bode.cdsp.neu.edu (Patrick Doyle) Date: Fri, 08 Aug 1997 06:09:00 -0000 Subject: patch - strange behaviour References: Message-ID: <199708081319.JAA02801@cdsp.neu.edu> I've tried to apply Mumit Khan's patches to b18 and experienced very strage behaviour of patch: patch aborts a after some patches, typically a few files, with the message "Assertion hunk failed ..." (don't remember exact wording). If I split the patch file into seven parts, I can apply the separate patch files without problems. I can live with that, but it's inconvenient. Has anyone got an idea what's going on? Yes, "patch" reads its patch file in TEXT mode (rather than BINARY mode) and occasionally caches a location in the file by performing an 'ftell()' type of operation. It expects to be able to perform an 'fseek()' to get back to that location in the file, but this doesn't work for text files (where CR's are silently eaten). I worked around it the brute force way by replacing calls to 'fseek()' with calls to 'my_brute_force_fseek()', which looped until 'ftell()' returned the desired position, reading and discarding one character at a time along the way. This is basically what I recall doing... I can send you a patch (ha ha) if you would like. -- --patrick - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mchase@ix.netcom.com Fri Aug 8 06:34:00 1997 From: mchase@ix.netcom.com (Michael A. Chase) Date: Fri, 08 Aug 1997 06:34:00 -0000 Subject: trouble unpacking cdk-src Message-ID: <199708081334.IAA26357@dfw-ix15.ix.netcom.com> Your filesystems are mounted in text mode by default by the initial gnuwin32 installation. That causes all files written by cat to have LF changed to CRLF and stop short at the MSDOS EOF mark (Ctrl-Z). That's part of why the installation instructions tell you to use the MSDOS/Windows copy command to splice the files back together. I don't have the output file's name in front of me, but the command should be something like this: copy /b xca+xcb+xcc+xcd+xce+xcf+xcg+xch+xci+xcj cdk-src.tar.gz Mac :}) -----Original Message----- From: KMorris692@aol.com To: gnu-win32@cygnus.com Date: Wednesday, August 06, 1997 21:53 Subject: trouble unpacking cdk-src >I downloaded csd-src file xca thru xcj. I'm having trouble cat-ing them >together. The result I get is too small. I suspect the files are being >treated as text and cat is finding an end-of-file character, but that's only >a guess on my part. > >Any advice on other methods? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From cgf@bbc.com Fri Aug 8 06:45:00 1997 From: cgf@bbc.com (Chris Faylor) Date: Fri, 08 Aug 1997 06:45:00 -0000 Subject: latest Sergey's update (08/01) crashes on Win95 References: Message-ID: In article < memo.19970807190129.45233C@horus.cix.co.uk >, Pete Jordan wrote: >cgf@bbc.com (Chris Faylor) wrote: > >> possibly a bug in Windows 95. > >How could you possibly even /suggest/ such a thing?... >;) You're right. (hangs head in shame) I'm sorry, Bill. -- http://www.bbc.com/ cgf@bbc.com "Strange how unreal VMS=>UNIX Solutions Boston Business Computing the real can be." - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From cgf@bbc.com Fri Aug 8 06:47:00 1997 From: cgf@bbc.com (Chris Faylor) Date: Fri, 08 Aug 1997 06:47:00 -0000 Subject: patch - strange behaviour References: Message-ID: In article < v02130500b00fc1f21707@[193.83.160.71] >, Wolfgang Bachmann wrote: >I've tried to apply Mumit Khan's patches to b18 and experienced very strage >behaviour of patch: > >patch aborts a after some patches, typically a few files, with the message >"Assertion hunk failed ..." (don't remember exact wording). > >If I split the patch file into seven parts, I can apply the separate patch >files without problems. I can live with that, but it's inconvenient. > >Has anyone got an idea what's going on? Yeah, you need a new patch. The one supplied in cdk.exe is apparently broken. Grab the latest GNU sources from prep.ai.mit.edu:/pub/gnu and patch should configure and build. -- http://www.bbc.com/ cgf@bbc.com "Strange how unreal VMS=>UNIX Solutions Boston Business Computing the real can be." - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From thoki@csi.com Fri Aug 8 07:52:00 1997 From: thoki@csi.com (Thorsten Kitz) Date: Fri, 08 Aug 1997 07:52:00 -0000 Subject: Fw: Berkley DB 2.2.6 (alias libdb) Message-ID: Hey, I tried to compile the Berkley database system (db 2.2.6 or libdb) with beta 18 and all worked well. I didn't get any compiler errors or warnings, but when it came to testing (with TCL testscripts, distributed in the libdb package) nothing really worked. Here is an extract from the all.out file: ------------------------------------------------------- cut here -------------------------------------------------------------------------- run_method: DB_BTREE 1 1 12:20:36 (00:00:00) Test001: DB_BTREE 10000 equal key/data pairs Test001.a: put/get loop Test001.b: dump file dbtest: unable to create/retrieve page 56 dbtest: TESTDIR/test001.db: put: more blocks returned than retrieved Test001: key/data mismatch: |-1| || run_method: DB_BTREE 2 2 12:20:47 (00:00:00) Test002: DB_BTREE 10000 key pairs Test002.a: put/get loop Test002.b: dump file dbtest: unable to create/retrieve page 54 dbtest: TESTDIR/test002.db: put: more blocks returned than retrieved Test002: data mismatch for key -1: || |abcdefghijklmnopqrstuvwxyzabcdefghijklmnopqrstuvwxyz| (E:\temp\db-2.2.6\dbtest.exe 1004) In cygwin_except_handler (E:\temp\db-2.2.6\dbtest.exe 1004) Exception trapped! (E:\temp\db-2.2.6\dbtest.exe 1004) exception C0000005 at 414032 (E:\temp\db-2.2.6\dbtest.exe 1004) exception: ax FFFEFFFF bx 33B7B58 cx 33B7B54 dx FFFFFFFC (E:\temp\db-2.2.6\dbtest.exe 1004) exception: si 423 di D539 bp 243DACC sp 243DAC0 (E:\temp\db-2.2.6\dbtest.exe 1004) exception is: STATUS_ACCESS_VIOLATION (E:\temp\db-2.2.6\dbtest.exe 1004) Stack trace: (E:\temp\db-2.2.6\dbtest.exe 1004) frame 0: sp = 0x243D8F4, pc = 0x1000CEC2 (E:\temp\db-2.2.6\dbtest.exe 1004) frame 1: sp = 0x243D910, pc = 0x77FA4512 (E:\temp\db-2.2.6\dbtest.exe 1004) frame 2: sp = 0x243D934, pc = 0x77F98EEB (E:\temp\db-2.2.6\dbtest.exe 1004) frame 3: sp = 0x243D9C0, pc = 0x77F86266 (E:\temp\db-2.2.6\dbtest.exe 1004) frame 4: sp = 0x243DACC, pc = 0x41B9B5 (E:\temp\db-2.2.6\dbtest.exe 1004) frame 5: sp = 0x243DB14, pc = 0x4108F8 (E:\temp\db-2.2.6\dbtest.exe 1004) frame 6: sp = 0x243DB6C, pc = 0x42DB9E (E:\temp\db-2.2.6\dbtest.exe 1004) frame 7: sp = 0x243DB8C, pc = 0x42A082 (E:\temp\db-2.2.6\dbtest.exe 1004) frame 8: sp = 0x243DBB0, pc = 0x4294D9 (E:\temp\db-2.2.6\dbtest.exe 1004) frame 9: sp = 0x243DC08, pc = 0x40960F (E:\temp\db-2.2.6\dbtest.exe 1004) frame 10: sp = 0x243DC80, pc = 0x40893E (E:\temp\db-2.2.6\dbtest.exe 1004) frame 11: sp = 0x243DCA4, pc = 0x25627FB (E:\temp\db-2.2.6\dbtest.exe 1004) frame 12: sp = 0x243DE14, pc = 0x257D05B (E:\temp\db-2.2.6\dbtest.exe 1004) frame 13: sp = 0x243DE30, pc = 0x257D3A9 (E:\temp\db-2.2.6\dbtest.exe 1004) frame 14: sp = 0x243DE74, pc = 0x256259D (E:\temp\db-2.2.6\dbtest.exe 1004) frame 15: sp = 0x243DFF4, pc = 0x256492B (E:\temp\db-2.2.6\dbtest.exe 1004) End of stack trace (more stack frames may be present) FAIL: test002 DB_BTREE run_method: DB_BTREE 3 3 12:21:01 (00:00:00) Test003: DB_BTREE filename=key filecontents=data pairs Test003.a: put/get loop Error 0 run_method: DB_BTREE 4 4 12:21:02 (00:00:00) Test004: DB_BTREE 10000 delete small key; medium data pairs Test004.a: put/get loop Test004.b: get/delete loop dbtest: unable to create/retrieve page 54 dbtest: TESTDIR/test004.db: put: more blocks returned than retrieved FAIL:12:21:35 (00:00:33) Test004:-1: expected -1-1-1-1-1-1-1-1-1-1, got run_method: DB_BTREE 5 5 12:21:36 (00:00:00) Test005: DB_BTREE 10000 delete small key; medium data pairs(with close) Test005.a: put/get loop dbtest: TESTDIR/test004.db: file size not a multiple of the pagesize Test005.b: get/delete loop invalid command name "dbopen: Error 0 " ------------------------------------------------------- cut here -------------------------------------------------------------------------- As you see, there are several problems, starting from access problems to the database up to system exceptions. Since it compiled well and my knowledge of debugging foreign C++ programs is very limited, does anybody knows an answer for this or does anybody have successfully compiled and tested the Berkley DB? Thanks, Thorsten. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kcantrel@dg-rtp.dg.com Fri Aug 8 08:08:00 1997 From: kcantrel@dg-rtp.dg.com (Ken Cantrell) Date: Fri, 08 Aug 1997 08:08:00 -0000 Subject: (Fwd) Mail Delivery Failure. Message-ID: <9708081507.AA05763@dg-rtp.dg.com> I seem to have found a bug with "tar." I'm using b18 (without any of Sergey's modifications). The command tar -cvf t.tar p6_das2bin.c p6_das2bin.h Makefile changelog works fine, but tar -cvf t08081997-10:46.tar p6_das2bin.c p6_das2bin.h Makefile changelog on the same set of files give the following errors: (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) In cygwin_except_handler (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) Exception trapped! (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) exception C0000005 at 40A9D7 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) exception: ax 2E84072 bx 41D070cx2E84091 dx 2E84097 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) exception: si 241FA0E di 0 bp 241F27C sp 241F218 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) exception is: STATUS_ACCESS_VIOLATION (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) Stack trace: (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe1268) frame 0: sp = 0x241F04C, pc = 0x1000CEC2 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) frame 1: sp =0x241F068, pc = 0x77F94072 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) frame 2: sp= 0x241F08C, pc = 0x77F88A53 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) frame 3: sp = 0x241F118, pc = 0x77F75E82 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) frame 4: sp = 0x241F27C, pc = 0x402046 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) frame 5: sp = 0x241F2E4, pc = 0x404300 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) frame 6: sp = 0x241F304, pc = 0x40E1ED (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) frame 7: sp = 0x241F31C, pc = 0x1000C102 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) frame 8: sp = 0x241FF94, pc = 0x1000C113 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) frame 9: sp = 0x241FFA0, pc = 0x41A7C1 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) frame 10: sp = 0x241FFB0, pc =0x40103B (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) frame 11: sp = 0x241FFC0, pc = 0x77F1AFC1 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) frame 12: sp = 0x241FFF0, pc = 0x0 (c:\gnuwin32\b18\h-i386-cygwin32\bin\tar.exe 1268) End of stack trace ***************************************************** Ken Cantrell Performance Modeling Data General, RTP kcantrel@dg.com; (919) 248 - 5879 **** Don't forget to Smile...that's why God gave us Fangs **** - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jont@harlequin.co.uk Fri Aug 8 09:00:00 1997 From: jont@harlequin.co.uk (Jon Thackray) Date: Fri, 08 Aug 1997 09:00:00 -0000 Subject: Problems with as and MSVC link Message-ID: <199708081555.QAA16561@zaphod.long.harlequin.co.uk> I have recently had reason to attempt to link files generated by as using MSVC link. For some reason, link tries to seek 4 bytes beyond the end of the file in every case. I have got around this by hand editing the object files to include four bytes of nulls, after which link seems quite happy. ld is happy with both the original files and the modified ones. Does anyone know why link wants the extra bytes? Should as be modified to provide them anyway? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From p.west@mailbox.uq.edu.au Fri Aug 8 09:39:00 1997 From: p.west@mailbox.uq.edu.au (Peter B. West) Date: Fri, 08 Aug 1997 09:39:00 -0000 Subject: Ineradicable file %_ revisited Message-ID: <3.0.2.32.19970808105139.0069ef90@dingo.cc.uq.edu.au> Fellow mailing list masochists, Thanks for all of the suggestions regarding file %_, which will not go away. I fear some more information is required. Bare ls in the containing directory yields: ls: %_: No such file or directory If there are other files in the directory, the preceding message is output first, followed by the listing of the "real" files. Similarly, rm -i * yields rm: %_: No such file or directory Reverting to Explorer, and examining properties of %_ gives: Size 0 bytes (0 bytes) MS-DOS name: Created unknown Modified unknown Accessed unknown Attempting to delete the little limpet gives: Cannot delete %_: Cannot find the specified file etc.. I mention this because of the possibility that this anomaly is GNU created. This is not necessarily the case, of course. All (well-meaning) suggestions gratefully accepted. Yours faithfully, Peter __ /__ Peter B. West / p.west@mailbox.uq.edu.au / "Master, to whom shall we go?" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pete@horus.cix.co.uk Fri Aug 8 09:39:00 1997 From: pete@horus.cix.co.uk (Pete Jordan) Date: Fri, 08 Aug 1997 09:39:00 -0000 Subject: A tool for executing Windows and DOS programs from bash Message-ID: I've written a little launch program that translates pathnames from GNU-Win32 format to Windows format so I can (for example) launch Word from bash by typing: $word foo.doc or: $word /usr/baz.doc The way it works is you create a symlink ("word") to my program (debash.exe) and add a line in an "/etc/debash" file that maps "word" onto the actual location of the target program and describes something of the sort of parameters that program requires. The path portion of each filename parameter is then matched against the mount table (with the path defaulting to `pwd`) and appropriate substitutions are made before using spawnvp() to launch the target app. It's a bit of a kludge ATM, but it does what I want so... If anyone is interested, I can make it available - GPLed of course :) Pete Jordan = = = = = = = = = = = = = = = = = = = = = = = Horus Communications http://www.horus.cix.co.uk/ = = = = = = = = = = = = = = = = = = = = = = = "Is a polar bear a rectangular bear after a coordinate transform?" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mh@mike.franken.de Fri Aug 8 10:27:00 1997 From: mh@mike.franken.de (Michael Hirmke) Date: Fri, 08 Aug 1997 10:27:00 -0000 Subject: patch - strange behaviour References: Message-ID: <6bVrzY9ppfB@mike.franken.de> Hi Wolfgang, [...] > > Has anyone got an idea what's going on? Yes, the b18 patch.exe is buggy. > > TIA. > > > > Cheers, Wolfgang Bye. Michael. -- Michael Hirmke | Telefon +49 (911) 557999 Georg-Strobel-Strasse 81 | FAX +49 (911) 557664 90489 Nuernberg | E-Mail mh@mike.franken.de | WWW http://minimike.franken.de/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From root@jacob.remcomp.fr Fri Aug 8 11:52:00 1997 From: root@jacob.remcomp.fr (root) Date: Fri, 08 Aug 1997 11:52:00 -0000 Subject: Ctrl-Z References: <199708081334.IAA26357@dfw-ix15.ix.netcom.com> Message-ID: > > Your filesystems are mounted in text mode by default by the initial > gnuwin32 installation. That causes all files written by cat to have LF > changed to CRLF and stop short at the MSDOS EOF mark (Ctrl-Z). > This is CRAZY. We have discussed this a thousand times in this list, in a span of several years: Ctrl-Z IS NOT USED FOR MARKING THE END OF FILE SINCE MS-DOS 2.0!!!!!!!!!!!! Does gnu-win32 want to be compatible with DOS 1.0? The real problem behind this, and that is why it makes me mad, is that nobody at Cygnus is reading this bug reports, so the same unfixed bugs remain forever and people fall into them over and over again. NOBODY uses Ctrl-Z to end a text file since AGES, at least 10 years or so. WAKE-UP GUYS!!! -- Jacob Navia Logiciels/Informatique 41 rue Maurice Ravel Tel 01 48.23.51.44 93430 Villetaneuse Fax 01 48.23.95.39 France - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From loki@dragoncat.net Fri Aug 8 11:52:00 1997 From: loki@dragoncat.net (Jeremy Blackman) Date: Fri, 08 Aug 1997 11:52:00 -0000 Subject: (Fwd) Mail Delivery Failure. References: <9708081507.AA05763@dg-rtp.dg.com> Message-ID: On Fri, 8 Aug 1997, Ken Cantrell wrote: > The command > tar -cvf t.tar p6_das2bin.c p6_das2bin.h Makefile changelog > works fine, but > tar -cvf t08081997-10:46.tar p6_das2bin.c p6_das2bin.h Makefile changelog I think this is because DOS/Windows can't use ':' in filenames that way. Look at it this way: a:blah.tar <--- blah.tar on drive a: t08081997-10:46.tar <--- 46.tar on drive t08081997-10: Tar, of course, being a UNIX program, doesn't inherently -know- this is a bad thing. :) - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From iverson@lionheart.com Fri Aug 8 12:09:00 1997 From: iverson@lionheart.com (Tim Iverson) Date: Fri, 08 Aug 1997 12:09:00 -0000 Subject: Debugging info for commercial linkers? References: <33EB5DC0.60E3@softway.com> Message-ID: Thanks for the info -- I suspected this. Do you mean that only the local variable information is different? My linker isn't picking up *anything* but global symbols. In particular, it's not finding the line numbers. Is there perhaps a web reference to pe-i386 object format somewhere? Thanks, - Tim Iverson iverson@cisco.com +---------------- | Date: Fri, 08 Aug 1997 11:56:16 -0600 | From: Jason Zions | To: Tim Iverson | CC: gnu-win32@cygnus.com | Subject: Re: Debugging info for commercial linkers? | | While porting gdb to OpenNT, we've discovered that the MS compilers and | gcc generate local variable debug information differently; that is, even - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ebritten@uci.edu Fri Aug 8 12:09:00 1997 From: ebritten@uci.edu (Eric Britten) Date: Fri, 08 Aug 1997 12:09:00 -0000 Subject: DirectX References: <33E9874E.5D40@metla.fi> Message-ID: <33EB5E53.908E81AF@uci.edu> It is possible. There is the headers and import lib for DirectDraw atleast at http://dragon.acadiau.ca/~025668b/develop/cygnus/ Check the mail archives for July and August at http://www.cygnus.com/ml/gnu-win32 for other mail about directx. Eric Britten Janne Leppdnen wrote: > Is it possible to use GNU-Win32 to make DirectX application?? > > And if yes where can I get import libraries or is there a tool which > makes them from ".dll". > > And if no.. why it is not possible? And will it soon be? > > (answers via email, please) > - > For help on using this list (especially unsubscribing), send a message > to > "gnu-win32-request@cygnus.com" with one line of text: "help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mh@mike.franken.de Fri Aug 8 12:09:00 1997 From: mh@mike.franken.de (Michael Hirmke) Date: Fri, 08 Aug 1997 12:09:00 -0000 Subject: documentation References: <33EA24F9.3A54@WORLDNET.ATT.NET> Message-ID: <6bVrz-0KpfB@mike.franken.de> Hi DIANE, You are perfectly right :))) Bye. Michael. -- Michael Hirmke | Telefon +49 (911) 557999 Georg-Strobel-Strasse 81 | FAX +49 (911) 557664 90489 Nuernberg | E-Mail mh@mike.franken.de | WWW http://minimike.franken.de/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From krr2@newton.ruph.cornell.edu Fri Aug 8 14:41:00 1997 From: krr2@newton.ruph.cornell.edu (Kyle R. Rose) Date: Fri, 08 Aug 1997 14:41:00 -0000 Subject: Visual C++ build Message-ID: <9708082033.AA18900@dirac> This is something you guys probably don't do a lot of anymore, but is it possible to rebuild the entire system using VC++ instead of the included GCC? Is this easier or harder than I expect? (For reference, I expect a real PIA...) Thanks, -------------------------------------------+---------------------------------- Kyle R. Rose | Cornell University, Class of '98 Email: krr2@cornell.edu | Department of Computer Science http://www.people.cornell.edu/pages/krr2/ | Department of Mathematics -------------------------------------------+---------------------------------- "I don't have low self-esteem.....I have | Discuss CORNELL HOCKEY ONLINE! low esteem for everyone else." - Daria | http://www.csuglab.cornell.edu/ -------------------------------------------+ cgi-bin/krr2/hockeymain.cgi - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From lfm@pgroup.com Fri Aug 8 14:41:00 1997 From: lfm@pgroup.com (Larry Meadows) Date: Fri, 08 Aug 1997 14:41:00 -0000 Subject: linking w/ microsoft libraries that reference kernel32 functions Message-ID: <199708082140.OAA23372@pacific.pgroup.com> I'm trying to link with the Intel BLAS library, intended for use with microsoft C and/or the Intel reference compiler. It has some code in it that references several kernel32 DLL functions with names __imp__xxx . These are unresolved by ld, and if I try to create an assembly file that defines them and sets them to the appropriate gnu interface function, I get program that won't execute: bash: ./a.exe: (null) bash$ gdb -nw a.exe GNU gdb 4.16.1 Copyright 1997 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "i386-cygwin32"... (gdb) run Starting program: /tmp/a.exe Error creating process /tmp/a.exe, (error 193) Obviously this must be an FAQ, but I can't find it. Thanks. lfm - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From john.cook@kla-tencor.com Fri Aug 8 14:41:00 1997 From: john.cook@kla-tencor.com (John Cook) Date: Fri, 08 Aug 1997 14:41:00 -0000 Subject: Ctrl-Z Message-ID: Unfortunately, even if most applications have ceased to use this CTRL-Z as an EOF marker, DOS itself and even NT's CMD.EXE continue to do so in some circumstances. 'copy /a filea fileb' treats the source file as 'ASCII' and appends that CTRL-Z to the end of the file (unless a CTRL-Z is already there). I'm not arguing in favor of this by any means; I started one of the thousand discussions about this last year upon noticing that gnuwin32's cat stopped upon finding CTRL-Z in a binary file. I'd prefer that cygwin32 treat all files as binary by default and never use (or interpret) CR or CTRL-Z for text files. NT's CMD.EXE even appears quite happy with text files which use only LF instead of CR-LF as line separators (I don't know if win95 is as tolerant). --John >---------- >From: root@jacob.remcomp.fr[SMTP:root@jacob.remcomp.fr] >Sent: Friday, August 08, 1997 11:12 AM >To: mchase@ix.netcom.com >Cc: gnu-win32@cygnus.com >Subject: Re: Ctrl-Z > >> Your filesystems are mounted in text mode by default by the initial >> gnuwin32 installation. That causes all files written by cat to have LF >> changed to CRLF and stop short at the MSDOS EOF mark (Ctrl-Z). > >This is CRAZY. We have discussed this a thousand times in this list, >in a span of several years: > >Ctrl-Z IS NOT USED FOR MARKING THE END OF FILE SINCE MS-DOS 2.0!!!!!!!!!!!! > >Does gnu-win32 want to be compatible with DOS 1.0? > >The real problem behind this, and that is why it makes me mad, is that nobody >at Cygnus is reading this bug reports, so the same unfixed bugs remain >forever >and people fall into them over and over again. > >NOBODY uses Ctrl-Z to end a text file since AGES, at least 10 years or so. >WAKE-UP GUYS!!! > >-- >Jacob Navia Logiciels/Informatique >41 rue Maurice Ravel Tel 01 48.23.51.44 >93430 Villetaneuse Fax 01 48.23.95.39 >France >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jeffdbREMOVETHIS@netzone.com Fri Aug 8 15:03:00 1997 From: jeffdbREMOVETHIS@netzone.com (Mikey) Date: Fri, 08 Aug 1997 15:03:00 -0000 Subject: Ctrl-Z References: Message-ID: <33ec9735.1491438@smtp.netzone.com> Sadly Microsuck still uses ^Z in their SDK header files. In fact they use it to "hide" information from some of their own tools. On Fri, 8 Aug 1997 20:12:08 +0200 (MET DST), you wrote: >> >> Your filesystems are mounted in text mode by default by the initial >> gnuwin32 installation. That causes all files written by cat to have LF >> changed to CRLF and stop short at the MSDOS EOF mark (Ctrl-Z). >> > >This is CRAZY. We have discussed this a thousand times in this list, >in a span of several years: > >Ctrl-Z IS NOT USED FOR MARKING THE END OF FILE SINCE MS-DOS 2.0!!!!!!!!!!!! > >Does gnu-win32 want to be compatible with DOS 1.0? > >The real problem behind this, and that is why it makes me mad, is that nobody >at Cygnus is reading this bug reports, so the same unfixed bugs remain forever >and people fall into them over and over again. > >NOBODY uses Ctrl-Z to end a text file since AGES, at least 10 years or so. >WAKE-UP GUYS!!! > >-- >Jacob Navia Logiciels/Informatique >41 rue Maurice Ravel Tel 01 48.23.51.44 >93430 Villetaneuse Fax 01 48.23.95.39 >France >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > (jeffdbREMOVETHIS@netzone.com) delete REMOVETHIS from the above to reply Mikey - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jazz@softway.com Fri Aug 8 15:41:00 1997 From: jazz@softway.com (Jason Zions) Date: Fri, 08 Aug 1997 15:41:00 -0000 Subject: Debugging info for commercial linkers? References: <199708080834.BAA01228@stars.cisco.com> Message-ID: <33EB5DC0.60E3@softway.com> While porting gdb to OpenNT, we've discovered that the MS compilers and gcc generate local variable debug information differently; that is, even if you tell MSVC to generate COFF-compatible debug information, it still outputs debug information for local vars in non-COFF form. It seems likely that your PharLap linker is expecting MS-form debug data and isn't finding it in a gcc-generated .o file. Alpha-test versions of gcc and g++ for OpenNT are available from http://www.opennt.com in the tools warehouse (under Support). As soon as we're confident we did the port right, the source code diffs will be made available there as well. Of interest: our port of gcc/g++ correctly generates code for NT on DEC Alpha hardware. Jason - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jazz@softway.com Fri Aug 8 15:41:00 1997 From: jazz@softway.com (Jason Zions) Date: Fri, 08 Aug 1997 15:41:00 -0000 Subject: Debugging info for commercial linkers? References: Message-ID: <33EB9018.7A66@softway.com> Variables are completely hosed. Line numbers for functions are partly hosed; here's a quote from one of my engineers which talks about the problem. > [gcc] generates line number information in the "COFF standard" > way: One particular opcode says "this is the beginning line number > of the function" and all others are relative. (At least I assume > it's coff standard, given that that's how gcc works, and there > aren't any options in that regard.) There's another debug opcode > that says "end of function", and in COFF standard, it's relative > to the beginning of the function. > > However, MS generates the end line number as an absolute (w.r.t. the > file) line number. [ ... ] Now, however, it turns out that MS > ALSO changed whether the relative lines are 0 or 1 based, so MS > generated objects are off by 1. There isn't a *really* good reference for the NT PE format. Chunks of it are discussed in various MSDN articles; some of it can be derived from various public headers in the SDK. Some of it, though, requires heavy duty bit-grovelling detective work. Jason - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From weiku@UTKUX1.UTK.EDU Fri Aug 8 15:41:00 1997 From: weiku@UTKUX1.UTK.EDU (Wei Ku) Date: Fri, 08 Aug 1997 15:41:00 -0000 Subject: A tool for executing Windows and DOS programs from bash Message-ID: <199708082226.SAA23931@solid.ssd.ornl.gov> Sound very interesting. Would you please put it on the web or mail a copy to me ? I would suggest you not to post it to the mailing list unless it is highly demanded since there are some people are unhappy to receive a large mail which is not necessary to them. Sincerely, Wei Ku *************************************** Department of Physics and Astronomy The University of Tennessee 1408 Circle Drive Knoxville, Tennessee 37996-1200 weiku@utkux.utcc.utk.edu --------------------------------------- Solid State Division Oak Ridge National Laboratory P.O.Box 2008 Oak Ridge, TN 37831-6032 Phone: (423) 574-5795 Fax: (423) 574-4143 weiku@solid.ssd.ornl.gov *************************************** -----Original Message----- From: Pete Jordan To: gnu-win32@cygnus.com Date: Friday, August 08, 1997 4:52 PM Subject: A tool for executing Windows and DOS programs from bash >I've written a little launch program that translates pathnames from >GNU-Win32 format to Windows format so I can (for example) launch Word from >bash by typing: > > $word foo.doc >or: > $word /usr/baz.doc > >The way it works is you create a symlink ("word") to my program >(debash.exe) and add a line in an "/etc/debash" file that maps "word" onto >the actual location of the target program and describes something of the >sort of parameters that program requires. > >The path portion of each filename parameter is then matched against the >mount table (with the path defaulting to `pwd`) and appropriate >substitutions are made before using spawnvp() to launch the target app. > >It's a bit of a kludge ATM, but it does what I want so... > >If anyone is interested, I can make it available - GPLed of course :) > >Pete Jordan >= = = = = = = = = = = = = = = = = = = = = = = >Horus Communications > http://www.horus.cix.co.uk/ >= = = = = = = = = = = = = = = = = = = = = = = >"Is a polar bear a rectangular bear after > a coordinate transform?" >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From iverson@cisco.com Fri Aug 8 16:29:00 1997 From: iverson@cisco.com (Tim Iverson) Date: Fri, 08 Aug 1997 16:29:00 -0000 Subject: Debugging info for commercial linkers? References: <33EB9018.7A66@softway.com> Message-ID: <199708082324.QAA08599@stars.cisco.com> Thanks! This is incredibly helpful. While it may not be the best news for my application, it probably means that it would be quicker and surer for me to just add OMF support (which is very well-documented) to BFD. Is there some compelling reason folks are even trying to support pe-i386 given that MS seems uninterested in documenting it? I guess if MSC outputs it, that might be a tiny reason. ;-) - Tim +---------------- | Date: Fri, 08 Aug 1997 15:31:04 -0600 | From: Jason Zions | To: Tim Iverson | CC: gnu-win32@cygnus.com | Subject: Re: Debugging info for commercial linkers? | References: < m0wwuFW-000SUZC@lionheart.com > | | Variables are completely hosed. Line numbers for functions are partly ... | There isn't a *really* good reference for the NT PE format. Chunks of it | are discussed in various MSDN articles; some of it can be derived from | various public headers in the SDK. Some of it, though, requires heavy | duty bit-grovelling detective work. | | Jason - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mh@mike.franken.de Fri Aug 8 16:29:00 1997 From: mh@mike.franken.de (Michael Hirmke) Date: Fri, 08 Aug 1997 16:29:00 -0000 Subject: vim again Message-ID: <6bVr-1q4pfB@mike.franken.de> Hi *, I found a new problem with vim 5.0h; sometimes when I want to edit a new file with ":e somefile", I get the following message: Can't read file /tmp/vo001209 I then can't read the file. If I leave vim and start it again with the filename as parameter, it works as expected. Anyone any idea what the problem might be ? Thx and bye. Michael. -- Michael Hirmke | Telefon +49 (911) 557999 Georg-Strobel-Strasse 81 | FAX +49 (911) 557664 90489 Nuernberg | E-Mail mh@mike.franken.de | WWW http://minimike.franken.de/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pete@horus.cix.co.uk Fri Aug 8 17:56:00 1997 From: pete@horus.cix.co.uk (Pete Jordan) Date: Fri, 08 Aug 1997 17:56:00 -0000 Subject: A tool for executing Windows and DOS programs from bash References: <199708082226.SAA23931@solid.ssd.ornl.gov> Message-ID: weiku@UTKUX1.UTK.EDU (Wei Ku) wrote: > Sound very interesting. Would you please put it on the web or mail a > copy to me ? I'll put it on my website - watch this space (I need to put the usual GNU preamble as a comment in the source). > I would suggest you not to post it to the mailing list unless it is > highly demanded since there are some people are unhappy to receive a > large mail which is not necessary to them. Indeed - that's why I didn't include it as an attachment to my original message :) Pete Jordan = = = = = = = = = = = = = = = = = = = = = = = Horus Communications http://www.horus.cix.co.uk/ = = = = = = = = = = = = = = = = = = = = = = = "'Not twisted,' Salzy once said of her own passion, 'it is helical. That sounds better.'" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From lfm@pgroup.com Fri Aug 8 17:56:00 1997 From: lfm@pgroup.com (Larry Meadows) Date: Fri, 08 Aug 1997 17:56:00 -0000 Subject: alignment of common blocks Message-ID: <199708090054.RAA24261@pacific.pgroup.com> For performance reasons, we would like to be able to control the alignment of common blocks (.comm name,size). We'd probably like to be able to control the alignment of local common blocks (.lcomm name,size) as well. It seems that the cygwin32 stuff aligns common blocks on (at most) 4-byte boundaries. Looking at the gas source, .lcomm stuff is aligned on at most 8-byte boundaries based on the size of the common, which is at least reasonable (although it would be nice to control). Many assemblers accept a third argument to .comm . On linux, I think gas accepts: .comm stuff,400,8 saying 'stuff is 400 bytes long, and align it on an 8-byte boundary'. Does anyone have any idea if the vanilla cygwin32 toolkit has a way to align .comm blocks (even brute force option to the linker)? Otherwise, I guess it is time to hack source code. And, any info on .lcomm. Finally, another question -- what is the default alignment of the stack? This is all important to get good performance on double precision stuff on pentium, pentium pro, and later, as well as to allow users control based on cache line size. Regards, Larry Meadows lfm@pgroup.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ebritten@uci.edu Fri Aug 8 18:37:00 1997 From: ebritten@uci.edu (Eric Britten) Date: Fri, 08 Aug 1997 18:37:00 -0000 Subject: Console GUI Utilities Beta 4 Message-ID: <33EBBE4A.B2AEF142@uci.edu> I have updated the Console GUI Utilities. So far I have one for nm. I also wrote a generic application to handle the output of any console application and put the output in a window so you can scroll through it, copy it to the clipboard, save it to a file. You can also paste files from the clipboard and drop files to be arguments of the console program that the front end handles. Previous versions used a DLL which worked on Windows 95, but not Windows NT. For now Console GUI does not use a DLL. This version has been tested on Windows NT 4.0 and it works. Be warned, though. Bugs could exist, though most functions operate correctly. There are some glitches in the gnm program. Originally it was my plan to build a DLL or EXE for every command(ie ls size strings nm). Each program could call its own command and override default behavior of the generic front end. For instance nm has a Find function that allows you to search for symbols in address ranges or nearest to a certain address that's in a certain file. The generic front end contains a find function that knows nothing about the format of the output of the command. Then I realized that I could save a lot of time if I added a configuration file to Console GUI. It uses INI file format and with it you can specify some of the behavior of the generic front end when a specific command is being used. This way to customize the generic front end I need only add a few lines to a text file rather than writing hundreds of lines of code and building another DLL or EXE. You can specify the type of window that the output will be presented(RichEdit, ListBox, Edit), whether or not the command uses filenames, or whether or not to resolve shortcut files, etc. As a result of this configuration file, I have spent more time on the generic front end program than the front end for nm. They share a lot of code, but the code that they don't share that sort of makes everything work is not the same. As a result the nm front end does not work the same as generic front end. This will be fixed later. On another note Console GUI satisfies over 80% of the Windows 95 logo requirements and was built using gnuwin32 b18. The files are at http://pages.prodigy.net/britten/gnugui and they are free. Any feedback is welcome. Eric Britten ebritten@uci.edu - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mbenz@sabre.com Fri Aug 8 18:37:00 1997 From: mbenz@sabre.com (Micheal A. Benzinger) Date: Fri, 08 Aug 1997 18:37:00 -0000 Subject: less 3.32 and RPC Message-ID: <3.0.3.32.19970808203935.00918750@pop.flash.net> Hi all, Some time ago someone mentioned that the search characters "/" and "?" did not work with less. I don't know why it doesn't work but I do have a solution. I used the regexp.c that is in the distribution and the search started working. It would appear that the RegExp implementation found in the Cygnus Beta 18 distribution does not work correctly. I do not have the source so I cannot verify this. To make the change, edit the Makefile once the software has been con- figured and add "regexp.${O}" to the OBJ variable and remake. The one problem that still remains, however, is if there is a line that must screen wrap, then the search will be off by one line. Using the "-S" option to chop long lines doesn't help either. The other thing, has anybody ported or know of a port of RPC for GNU-Win32? Is it possible to port this or does this require modifications to the Cygnus DLL or worse need more cooperation from Windows? Any information on this would be appreciated. I know that DCE threads or pthreads are also necessary for this. I am not sure if this can be ported either. Regards, Mike Benzinger - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From iverson@cisco.com Fri Aug 8 23:00:00 1997 From: iverson@cisco.com (Tim Iverson) Date: Fri, 08 Aug 1997 23:00:00 -0000 Subject: alignment of common blocks References: <199708090054.RAA24261@pacific.pgroup.com> Message-ID: <199708090556.WAA11664@stars.cisco.com> You can align an entire section using various ldscript options. Also, -malign-double can be used to set the default alignment to 8 bytes for double, long double, and long long. I don't know whether Cygnus compiled this option into the default release. The stack pointer is almost always kept aligned on a 4 byte boundary unless someone makes a mistake (3,2,&1 do work, but slow the system). I don't know what effect, if any, -malign-double has on automatic doubles. If you do recompile from scratch, changing default alignments for the various blocks is just a header change. Don't remember which one off the top, though ... somewhere in gcc/config. Hope this helps, - Tim Iverson iverson@cisco.com +---------------- | Subject: alignment of common blocks | To: gnu-win32@cygnus.com | Date: Fri, 8 Aug 1997 17:54:36 -0700 (PDT) | From: Larry Meadows | | For performance reasons, we would like to be able to control the | alignment of common blocks (.comm name,size). We'd probably like to | be able to control the alignment of local common blocks (.lcomm name,size) ... | Finally, another question -- what is the default alignment of the stack? | | This is all important to get good performance on double precision stuff | on pentium, pentium pro, and later, as well as to allow users control | based on cache line size. | | Regards, | | Larry Meadows | lfm@pgroup.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From anomaly@jet.laker.net Fri Aug 8 23:56:00 1997 From: anomaly@jet.laker.net (anomaly@jet.laker.net) Date: Fri, 08 Aug 1997 23:56:00 -0000 Subject: How to get tcl scripts executable? Message-ID: <33EC15E7.2B98AA87@laker.net> Greetings, How do i make tcl scripts, both bare script and wish to be executable? Meaning what goes in the header of the script, the paths, best sub dirs to store the scripts, etc etc. Thanx. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ferret@seqeb.gov.au Fri Aug 8 23:56:00 1997 From: ferret@seqeb.gov.au (Graeme Wightman) Date: Fri, 08 Aug 1997 23:56:00 -0000 Subject: Port of JOE ? Message-ID: <33EC127B.74F58481@seqeb.gov.au> I have heard that someone ported JOE. Rather than reinvent the wheel, could anyone point me at a site where I can get a copy? Thanks, Graeme Wightman -- Graeme Wightman - System Support ____ Email: ferret@seqeb.gov.au SEQEB, GPO Box 1461, (/ .. \) or ferret@null.net Brisbane 4001 AUSTRALIA \__,_/ Phone: +61 7 34074150 / ^/ <<< GOD is REAL ... unless declared INTEGER. >>> - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From noer@cygnus.com Sat Aug 9 00:28:00 1997 From: noer@cygnus.com (Geoffrey Noer) Date: Sat, 09 Aug 1997 00:28:00 -0000 Subject: Ctrl-Z References: Message-ID: <199708090727.AAA26697@rtl.cygnus.com> [...] > Ctrl-Z IS NOT USED FOR MARKING THE END OF FILE SINCE MS-DOS 2.0!!!!!!!!!!!! > > Does gnu-win32 want to be compatible with DOS 1.0? > > The real problem behind this, and that is why it makes me mad, is that nobody > at Cygnus is reading this bug reports, so the same unfixed bugs remain [...] Nope, I read them. :-) Believe it or not, we have customers who have huge source bases where all their files end in ^Zs. Even if we didn't have ^Z as EOF, people would still run into the carriage return/line feed conversion part of the text mode != binary mode picture whenever binary files are read in text mode. Hopefully adding binary flags to some of the utilities will help... Cheers, Geoffrey Noer noer@cygnus.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sat Aug 9 02:38:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sat, 09 Aug 1997 02:38:00 -0000 Subject: Ctrl-Z Message-ID: <01BCA4C9.B5AD8180@sos> John Cook wrote: > files. NT's CMD.EXE even appears quite happy with text files > which use only LF instead of CR-LF as line separators (I don't > know if win95 is as tolerant). Try to remove 's from NT's boot.ini :-) Be careful... -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sat Aug 9 02:38:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sat, 09 Aug 1997 02:38:00 -0000 Subject: (Fwd) Mail Delivery Failure. Message-ID: <01BCA4C9.5F1143C0@sos> Jeremy Blackman wrote: > > The command > > tar -cvf t.tar p6_das2bin.c p6_das2bin.h Makefile changelog > > works fine, but > > tar -cvf t08081997-10:46.tar p6_das2bin.c p6_das2bin.h Makefile changelog > > I think this is because DOS/Windows can't use ':' in filenames that way. Man tar :-) The colon in filename separates remote host name from filename :-) Add --force-local option to tar command line. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mh@mike.franken.de Sat Aug 9 05:38:00 1997 From: mh@mike.franken.de (Michael Hirmke) Date: Sat, 09 Aug 1997 05:38:00 -0000 Subject: less 3.32 and RPC References: <3.0.3.32.19970808203935.00918750@pop.flash.net> Message-ID: <6bZwZPappfB@mike.franken.de> Hi Micheal, > Hi all, > > Some time ago someone mentioned that the search characters > "/" and "?" did not work with less. I don't know why it > doesn't work but I do have a solution. I used the regexp.c No, there only was a small bug in the Cygnus distribution, as Sergey stated a few months before: ------------------------< snip snip snip >----------------------------- Cdk install does not copies librx/rx.h to /usr/include so less's configure fails to check regexp type. It's a bug in cdk installation script. Also, while compiling newlib gcc fails to find "glue.h" include file. I had to create empty file by hands. ------------------------< snip snip snip >----------------------------- [...] > Regards, > > Mike Benzinger Bye. Michael. -- Michael Hirmke | Telefon +49 (911) 557999 Georg-Strobel-Strasse 81 | FAX +49 (911) 557664 90489 Nuernberg | E-Mail mh@mike.franken.de | WWW http://minimike.franken.de/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mh@mike.franken.de Sat Aug 9 05:45:00 1997 From: mh@mike.franken.de (Michael Hirmke) Date: Sat, 09 Aug 1997 05:45:00 -0000 Subject: Ineradicable file %_ revisited References: <3.0.2.32.19970808105139.0069ef90@dingo.cc.uq.edu.au> Message-ID: <6bZwYpd$pfB@mike.franken.de> Hi Peter, [...] > All (well-meaning) suggestions gratefully accepted. Did you try to get rid of it with *chkdsk* ? Perhaps chkdsk will recognize a filesystem error. > > Yours faithfully, > Peter Bye. Michael. -- Michael Hirmke | Telefon +49 (911) 557999 Georg-Strobel-Strasse 81 | FAX +49 (911) 557664 90489 Nuernberg | E-Mail mh@mike.franken.de | WWW http://minimike.franken.de/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From garbanzo@hooked.net Sat Aug 9 05:45:00 1997 From: garbanzo@hooked.net (Alex) Date: Sat, 09 Aug 1997 05:45:00 -0000 Subject: How to get tcl scripts executable? References: <33EC15E7.2B98AA87@laker.net> Message-ID: On Sat, 9 Aug 1997 anomaly@jet.laker.net wrote: > Greetings, > How do i make tcl scripts, both bare script and wish to be > executable? Meaning what goes in the header of the script, the paths, > best sub dirs to store the scripts, etc etc. > Thanx. If you're on nt or the like chmod +x the file, then add #!/path/to/tclsh at the top of the file where /path/to/tclsh is the correct location of tclsh. - alex - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mchase@ix.netcom.com Sat Aug 9 09:54:00 1997 From: mchase@ix.netcom.com (Michael A. Chase) Date: Sat, 09 Aug 1997 09:54:00 -0000 Subject: Ineradicable file %_ revisited Message-ID: <199708091654.LAA22616@dfw-ix4.ix.netcom.com> If all else fails, move all other files and directories out of the parent directory and then delete the parent in Explorer. If that doesn't work, try running disk diagnostics from your favorite utility package and see if it catches the invalid file name. BTW, does this mean you couldn't rename the file in Explorer? Mac :}) -----Original Message----- From: Peter B. West To: gnu-win32@cygnus.com Date: Friday, August 08, 1997 11:35 Subject: Ineradicable file %_ revisited >Fellow mailing list masochists, > >Thanks for all of the suggestions regarding file %_, which will not go >away. I fear some more information is required. > >Bare > ls >in the containing directory yields: >ls: %_: No such file or directory > >If there are other files in the directory, the preceding message is output >first, followed by the listing of the "real" files. > >Similarly, rm -i * yields >rm: %_: No such file or directory > >Reverting to Explorer, and examining properties of %_ gives: > >Size 0 bytes (0 bytes) >MS-DOS name: > Created unknown > Modified unknown > Accessed unknown > >Attempting to delete the little limpet gives: >Cannot delete %_: Cannot find the specified file etc.. > >I mention this because of the possibility that this anomaly is GNU created. > This is not necessarily the case, of course. > >All (well-meaning) suggestions gratefully accepted. > >Yours faithfully, >Peter > __ /__ Peter B. West > / p.west@mailbox.uq.edu.au > / "Master, to whom shall we go?" >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From cgf@bbc.com Sat Aug 9 16:02:00 1997 From: cgf@bbc.com (Chris Faylor) Date: Sat, 09 Aug 1997 16:02:00 -0000 Subject: Ctrl-Z References: <199708090727.AAA26697@rtl.cygnus.com> Message-ID: In article < 199708090727.AAA26697@rtl.cygnus.com >, Geoffrey Noer wrote: >[...] >>Ctrl-Z IS NOT USED FOR MARKING THE END OF FILE SINCE MS-DOS >>2.0!!!!!!!!!!!! >> >>Does gnu-win32 want to be compatible with DOS 1.0? >> >>The real problem behind this, and that is why it makes me mad, is that >>nobody at Cygnus is reading this bug reports, so the same unfixed bugs >>remain >[...] > >Nope, I read them. :-) > >Believe it or not, we have customers who have huge source bases where >all their files end in ^Zs. Even if we didn't have ^Z as EOF, people >would still run into the carriage return/line feed conversion part of >the text mode != binary mode picture whenever binary files are read in >text mode. > >Hopefully adding binary flags to some of the utilities will help... I have submitted a patch to Sergey for cygwin.dll which incorporated a new environment variable, CYGWIN32, to honor the following settable options: [no]title [no]tty [no]binmode So, you could do the following: set CYGWIN32=notitle binmode bash and bash would not update the title bars and would use binmode as the default for opens. I had intended to extend this to allow registry entries to be used which would would be settable based on the path of the executing program. That means that you could tell 'cat' to always operate in default binmode and 'tar' to always operate in binmode. Would this solve any problems? -- http://www.bbc.com/ cgf@bbc.com "Strange how unreal VMS=>UNIX Solutions Boston Business Computing the real can be." - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fjh@cs.mu.OZ.AU Sat Aug 9 21:06:00 1997 From: fjh@cs.mu.OZ.AU (Fergus Henderson) Date: Sat, 09 Aug 1997 21:06:00 -0000 Subject: Debian/Win32 References: <199708100321.UAA22814@fleming.jimpick.com> Message-ID: <199708100405.OAA15240@mundook.cs.mu.OZ.AU> [This discussion originated on win32@lists.debian.org, but I'm sending this to both that list and gnu-win32@cygnus.com]. Jim Pick, you wrote: > > > 2) The Cygnus distribution is not FSSTND based. I think we > > > can just juggle the files around into their proper places without > > > causing too much grief. > > > > Yes. Maybe we can persuade Cygnus to make their distribution FSSTND based. > > Or maybe we can convince them to concentrate on building the winsup/newlib > "kernel" layer, and leave the distribution building to us. That's > exactly what we do with Linux. That would be a good idea. > > > I don't like how the Cygnus distribution defaults to mounting the > > > \ directory at C:\ - it would be much nicer to mount this directory > > > at a different (user specified) location (ie. C:\debw32). That > > > way, all the standard FSSTND directories (/bin /dev /etc /home /lib > > > /proc /root /sbin /tmp /usr /var) wouldn't pollute the C:\ > > > directory. That's fairly important if we want to put it into > > > widespread use. > > > > This is also worth suggesting to Cygnus. > > Their InstallShield installer should have a dialogue box so > > that the user can choose where to mount the root directory. > > The default ought to be something other than \. > > Well, we can do it ourselves, and Cygnus can check it out. I'm > sure they'd be interested. > > InstallShield basically does the same job as dpkg, but not in as > good a manner. And it isn't a "free" tool. Cygnus would be ill-advised > to attempt to build a distribution with it. Frankly, if they are > going to build a distribution, they should using either rpm, dpkg, > or the FreeBSD package tools. Or stay out of the distribution building > business altogether (like what Linus does). Cygwin32 is really a base system (or at least the winsup/newlib parts are). Installing the base system is a bit different to installing additional packages. Is there a free tool that does as well as InstallShield does with regard to the Win32 user interface? > > > 7) I'm still interested in building a glibc-based layer, since that > > > would be an opportunity to do things in a more standard way. > > > > Yes, this is still a good idea. > > I was also wondering (and perhaps someone else would have more knowledge > in this area) - wouldn't it also be possible, down the road, to run > Linux binaries in ELF format on Win32 using an emulation layer? I know > FreeBSD can do this, and Linux does this with the iBCS package for > other binaries. It ought to be technically possible. It would certainly be very neat. > Another weird idea - I wonder if cygwin32.dll would operate under WINE > or TWIN? That might be nice for development purposes. :-) -- Fergus Henderson | "I have always known that the pursuit WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mphalpin@iag.net Sat Aug 9 21:24:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Sat, 09 Aug 1997 21:24:00 -0000 Subject: sh crashes on 95 (also screen) Message-ID: <199708100420.AAA28254@iag.net> Okay... when I'm doing make, sh has a habit of crashing. It's got bad recently. Where can I get da patch? Ack! SJ P.S. Has anyone ported a screen command? We could use one ;) P.P.S. Anyone tried out talk/talkd? Tell me comments... does it display the "ring"? .etc - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sun Aug 10 00:01:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sun, 10 Aug 1997 00:01:00 -0000 Subject: sh crashes on 95 (also screen) Message-ID: <01BCA57C.E6B2DA60@sos> Sonic Junior wrote: > > P.S. Has anyone ported a screen command? We could use one ;) > It isn't possible now. Screen requires UNIX domain sockets or FIFOes not implemented in winsup. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sun Aug 10 00:10:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sun, 10 Aug 1997 00:10:00 -0000 Subject: Debian/Win32 Message-ID: <01BCA57E.35C88F90@sos> Fergus Henderson wrote: > Is there a free tool that does as well as InstallShield does > with regard to the Win32 user interface? It's possible to write Tk-based installation script. Anyway Tcl/Tk is a part of gnu-win32 package. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From p.west@mailbox.uq.edu.au Sun Aug 10 04:41:00 1997 From: p.west@mailbox.uq.edu.au (Peter B. West) Date: Sun, 10 Aug 1997 04:41:00 -0000 Subject: Ineradicable file %_ revisited References: <3.0.2.32.19970808105139.0069ef90@dingo.cc.uq.edu.au> <33eb9645.1251683@smtp.netzone.com> Message-ID: <3.0.2.32.19970810031313.006a1430@dingo.cc.uq.edu.au> At 10:00 pm 8/08/97 GMT, you wrote: >I know this thing is annoying, but since it's 0 bytes, >it's just a directory entry, you can probably just ignore it. > >otherwise have you tried shelling out to dos and >using deltree /y dir_with_bad_file? > >else shell out to dos, and use a diskeditor >and delete it manually. > > Mikey > A thousand thanks, Mikey. 'deltree -y' did the deed, indeed. What a useful little program that is. Shows my ignorance of Win95 & DOS. To Michael Chase: yes, I did try to rename in Explorer - no go. To Michael Hirmke: I ran a full disk check, autocorrect errors, in Win95 with no result. Thanks to everyone who replied. Yours faithfully, Peter __ /__ Peter B. West / p.west@mailbox.uq.edu.au / "Master, to whom shall we go?" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sun Aug 10 05:24:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sun, 10 Aug 1997 05:24:00 -0000 Subject: New patch Message-ID: <01BCA5AA.00AB4920@sos> - Chris Faylor's (cfg@bbc.com) waiting code temporary removed. The code is right, works fine on Windows NT and is neccessary to support a correct semantics of wait syscalls, but it have unexpected behavior on Windows 95. No workaround found. - Some syscall semantics-related changes. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From adg@dibe.unige.it Sun Aug 10 05:24:00 1997 From: adg@dibe.unige.it (alessandro de gloria) Date: Sun, 10 Aug 1997 05:24:00 -0000 Subject: two questions Message-ID: <33EDA72D.7A51E76C@dibe.unige.it> Dear Sir, I want to use gnuwin32 for cross-compiling. I only use gcc and my assembler and linker. How can I specify, during configuretion, the directory where the cross-assembler and linker are. Does gnuwin32 support glibc-2.0? I tried to cross compile the libray but I got some problem in the makefile Best Regards, Alessandro De Gloria - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mike@mbsun.mlb.org Sun Aug 10 18:08:00 1997 From: mike@mbsun.mlb.org (Mike Bernson) Date: Sun, 10 Aug 1997 18:08:00 -0000 Subject: Serial Ports Message-ID: <33EE6608.695678E2@mbsun.mlb.org> I am modify the TTY driver for 4.4BSD to run under cygwin stuff. I can not find any easy to known that current state of buffers windows in using. How many character are in the input buffer or how many character are in the output buffer. It would also be nice be not need to known the current state of the Xon/Xoff flow state, CTS, RTS, DCD. Once this code is working I will release it to a number of beta tester. At current I plan on have serial port, console, and pty support for BSD style stuff. This code is very different for the current TTY code in that I do not plan on have a number of tasks per tty. I am looking at doing it with async I/O. I will then fix select code to handle the tty interface and remove kluges for console. In doing the ported of tty driver I have added BSD sleep/wakeup interface which is using events to do the sleep/wakeup stuff. The Big change is it does not take address but handles of events. I have also removed the binary stuff tty devices. It does not fit in very well. (Makes to many function handling nl/cr stuff). I have removed a number of flags form fhandler_base. The flags exists in more then one place which is very bad. Flags that I have removed are access_, w_binary_, r_binary I am looking at removing the following flags because they belong in open_flags. close_exec_p_, append_p_, async_, All of the above flags are normal in 1 spot and bitmap. This is the way unix handles it and I will be doing it this way.  - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From MCzapski@stc.com Sun Aug 10 18:35:00 1997 From: MCzapski@stc.com (Michael Czapski) Date: Sun, 10 Aug 1997 18:35:00 -0000 Subject: gcc: Internal compiler error: program cc1 got fatal signal 33 Message-ID: <01BCA64A.7EED4180.MCzapski@stc.com> Hello, Trying to make a reasnably siple piece of code under cygwin32, b18 with Sergey's patches/mods, which may or may not be releveant, I get the following error : gcc: Internal compiler error: program cc1 got fatal signal 33 Can anyone shed some light ? I am quite new to cygwin32 so I am not much in a position to help myself yet. TIA Michael Czapski MCzapski@stc.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From colin@bird.fu.is.saga-u.ac.jp Sun Aug 10 19:43:00 1997 From: colin@bird.fu.is.saga-u.ac.jp (Colin Peters) Date: Sun, 10 Aug 1997 19:43:00 -0000 Subject: Error : Ld : cannot load '-lgcc' no such file or directory Message-ID: <01BCA64B.6F15F800@gbird0> Jon Thackray[SMTP:jont@harlequin.co.uk] wrote: >Greg Maynard writes: [snip] > > properly ;).. Anyhow, when I run gcc it gives me the error 'Ld : cannot > > load '-lgcc' no such file or directory' and I don't know where the heck [snip] >-lgcc means it is looking for libgcc.a. That can be found in >gcc-lib/i386-cygwin32/cygnus-2.7.2-961023 (at least that's where it is >for beta 17.1). The real question though, is why are you requiring >libgcc.a at all? You might try nm libgcc.a to see if any of the >symbols exported (those marked T or ?) are in your code. Among other things libgcc.a contains __main, a function automatically called at the beginning of any function called 'main' in your compiled code. This function does some initialization, like calling do_ctors (or something like that) to initialize the global C++ objects in your program and so on. Some versions, perhaps all, also contain _alloca, which is used to allocate large arrays (rather than taking them off the stack). In short, the gcc library contains a number of calls made implicitly by the compiler, and is usually included in any program you write. This is why you don't need to link it explicitly: gcc will always link it for you. Now that you know where it is and why you need it; how do you get rid of the error? Well, normally gcc (or ld, or someone) will search the path mentioned above, at least if your GCC_EXEC_PREFIX is set correctly (as indicated in the README file). If that environment variable wasn't set correctly I would expect you to be getting the far more common 'cannot exec cpp' error. But perhaps you fixed that by moving cpp.exe to the same directory as gcc (or somewhere else on your path)? Another solution would be to put the directory containing libgcc.a in your LIBRARY_PATH environment variable. Hope this helps, Colin. -- Colin Peters - Saga Univ. Dept. of Information Science -- colin@bird.fu.is.saga-u.ac.jp - finger for PGP public key -- http://www.fu.is.saga-u.ac.jp/~colin/index.html -- http://www.geocities.com/Tokyo/Towers/6162/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From webbb@aries.tucson.saic.com Mon Aug 11 08:29:00 1997 From: webbb@aries.tucson.saic.com (Brian Webb) Date: Mon, 11 Aug 1997 08:29:00 -0000 Subject: Unix version of Tcl/Tk Message-ID: <33EF2FC8.69A8@aries.tucson.saic.com> Has anybody tried compiling a version of Tcl/Tk from the unix source tree? I'm trying to build an X Windows version of Tcl/Tk rather than the windows version that comes with b18. I can get Tcl to compile without too much difficulty, but it has trouble parsing init.tcl... Thanks, Brian Webb webbb@aries.tucson.saic.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From tom.rodman@pounder.sol.net Mon Aug 11 10:27:00 1997 From: tom.rodman@pounder.sol.net (Tom Rodman) Date: Mon, 11 Aug 1997 10:27:00 -0000 Subject: where are binaries for RCS that work with WIN95? Message-ID: <9708111728.AA05136@mke100.mke.spd.jci.com> I'm an "end-user" type, running beta 18 "usertools". Will someone pls point me to a public domain RCS package that works under Windows 95? I RCS V5.6 working under NT, but not Windows 95. Please respond via e-mail also. thanks!, Tom Rodman M67 Johnson Controls/ Controls Group 507 E Michigan St Milwaukee,WI 53202 1-414-274-5041/fax:4400 tom.rodman@pounder.sol.net - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From abraham@dina.kvl.dk Mon Aug 11 11:50:00 1997 From: abraham@dina.kvl.dk (Per Abrahamsen) Date: Mon, 11 Aug 1997 11:50:00 -0000 Subject: cugwin32 b18: c++ blocks with `In cygwin_except_handler' message Message-ID: I'm using cygwin32 b18 with Windows NT 4.0 ServicePack 1. The c++ command often blocks with the following message In cygwin_except_handler when trying to compile some code that worked with g++ 2.7.2 under sparc-solaris-2.5.1. Nothing more happens. The process doesn't exit. Any suggestions for how to debug this? An small example $ cat test.C void foo () { int a = 0; int b = a ?: 1; } $ c++ -c test.c (long/path/cc1plus.exe 1038) In cygwin_except_handler This example was easy to fix, just don't use the GNU `?:' extension. But it happens in many places in my code, also places that doesn't use GNU extensions, so I suspect there is something wrong with my setup. Any ideas? PS: Where should I report errors that doesn't look win32 specific in the g++ bundled with cygwin? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Gilles.Bourque@nrc.ca Mon Aug 11 14:01:00 1997 From: Gilles.Bourque@nrc.ca (Bourque, Gilles) Date: Mon, 11 Aug 1997 14:01:00 -0000 Subject: V + GNUWIN32 (b18) Message-ID: Hello, Did anybody compiled example codes of V (GUI) with GNUWIN32 beta18 from CYGNUS? If yes, could you please post your makefiles. Thanks in advance, Gilles Bourque Gilles.Bourque@nrc.ca - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From iko@crec.mipt.ru Mon Aug 11 14:01:00 1997 From: iko@crec.mipt.ru (Igor V. Kovalenko) Date: Mon, 11 Aug 1997 14:01:00 -0000 Subject: rm -rf -> Circular directory structure Message-ID: <33EF7405.AADDD43E@crec.mipt.ru> Looking at ...include/sys/types.h I found that ino_t is actually unsigned short, that is only 16 bit in size. Thus, we (in 'rm' etc..) do compare the lower 16 bits of prepared by hash_path_name() 32-bit integer. I think this is a main cause for $subj$, though the hashing function itself is not as perfect as it could be (in theory). PS: Or do I have incorrect header file? -- Igor V. Kovalenko mailto:iko@crec.mipt.ru - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From si10647@ci.uminho.pt Mon Aug 11 14:01:00 1997 From: si10647@ci.uminho.pt (Daniel P.C. da Silva) Date: Mon, 11 Aug 1997 14:01:00 -0000 Subject: filename expansion Message-ID: <311E5922.9F864639@ci.uminho.pt> I don't know how to pass a pattern (like find . -name "*.pl") without doing filename expansion. Under UNIX, I would use \* , "*" or '*' to protect the pattern from the shell. None of this methods work with gnu-win32 commands running from a command prompt (not bash). Is there another way of doing this? -- ------------------------------------------- Daniel P.C. Silva Universidade do Minho si10647@ci.uminho.pt ------------------------------------------- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dzhou@std.saic.com Mon Aug 11 14:01:00 1997 From: dzhou@std.saic.com (Dong Zhou) Date: Mon, 11 Aug 1997 14:01:00 -0000 Subject: socket binding Message-ID: Hi all, A question about sockets. It seems that my program can't correctly bind a socket. The code is like following: if ((fd = socket(AF_INET, udp?SOCK_DGRAM:SOCK_STREAM, 0)) < 0) err("socket"); if (bind(fd, &sinme, sizeof(sinme)) < 0) err("bind"); The socket() call is successful but the bind() fails, and the error number it returned is 106, which, from "errno.h", is EAFNOSUPPORT. and, TCP/IP stuff seems are working well (at least I can use Netscape and ftp...) and the program actually works on Solaris. (The port number assigned is 5001) Can somebody tell me: what does the error mean? and what is the problem behind it? (something wrong with my NT system?) Thanks a lot. Dong == Dong Zhou (703)907-2526(O) SAIC - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From danish-corner@vip.cybercity.dk Mon Aug 11 15:07:00 1997 From: danish-corner@vip.cybercity.dk (Danish Corner) Date: Mon, 11 Aug 1997 15:07:00 -0000 Subject: Not possible to hide console window Message-ID: <33EF85E2.399A072D@vip.cybercity.dk> According to the gnu-win32 faq it is possible to prevent the console window from showing up by adding the string "-Wl,-subsystem,windows" to the GCC command line. However, I get the warning from LD, "cannot find entry symbol _WinMainCRTStartup; defaulting to 00401000", and the console window keeps popping up when running the executable from Explorer. Is this a known bug? Is there a workaround? Hope that someone can help. /Steen - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From hop@elwood.pionet.net Mon Aug 11 18:34:00 1997 From: hop@elwood.pionet.net (Justin Hopkins) Date: Mon, 11 Aug 1997 18:34:00 -0000 Subject: termcap interface code Message-ID: <3.0.1.32.19970811203638.0069123c@elwood.pionet.net> Hi, I'm trying to compile texinfo 3.9 and I'm getting some undefined references at link-time. These are the functions: extern int tgetnum (), tgetflag (), tgetent (); extern char *tgetstr (), *tgoto (); extern char *getenv (); extern void tputs (); The comment nearby says these are Unix termcap interface code, so does anyone know where I can get these? Or if they're already in the cygwin32 b18 distribution..? Thanks. Justin Hopkins - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From hchang@ee.usyd.edu.au Mon Aug 11 19:37:00 1997 From: hchang@ee.usyd.edu.au (Hung Hsin Chang) Date: Mon, 11 Aug 1997 19:37:00 -0000 Subject: Bugs of "sed" or "|" ? Message-ID: I have probelms to make the makefiles contain "| sed" under win95. Any idea? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dbe@wgn.net Mon Aug 11 20:11:00 1997 From: dbe@wgn.net ($Bill Luebkert) Date: Mon, 11 Aug 1997 20:11:00 -0000 Subject: filename expansion References: <311E5922.9F864639@ci.uminho.pt> Message-ID: <33EFD432.380D@wgn.net> Daniel P.C. da Silva wrote: > > I don't know how to pass a pattern (like find . -name "*.pl") without > doing filename expansion. Under UNIX, I would use \* , "*" or '*' to > protect the pattern from the shell. None of this methods work with > gnu-win32 commands running from a command prompt (not bash). > Is there another way of doing this? There is a bug (IMHO) in the argument passing of non-cygwin shells. Cygwin assumes that the shell has no globbing ability and globs the cmdline. Furthermore, it ignores ' and ", so you can't escape it. I think I may have found a kludge around it though. Try: xargs find . -name *.pl -print ^Z or xargs find . -name *.pl -print ^Z -- ,-/- __ _ _ $Bill Luebkert (_/ / ) // // DBE Collectibles / ) /--< o // // http://www.wgn.net/~dbe/ -/-' /___/_<_ I am trying to build the perl5.004_01 libwin32-0.08 module using gnuwin32 b18 . The README says nothing about using gnuwin32, and my first error is trying to find header files winsock.h & tchar.h . Has anyone succesfully built these modules, and have any hints to share? -- Danny Aldham SCO Ace , MCSE , JAPH , DAD I don't need to hide my e-mail address, I broke my sendmail. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dbe@wgn.net Mon Aug 11 20:59:00 1997 From: dbe@wgn.net ($Bill Luebkert) Date: Mon, 11 Aug 1997 20:59:00 -0000 Subject: socket binding References: Message-ID: <33EFDF9C.52A1@wgn.net> Dong Zhou wrote: > > Hi all, > > A question about sockets. > > It seems that my program can't correctly bind a socket. > The code is like following: > > if ((fd = socket(AF_INET, udp?SOCK_DGRAM:SOCK_STREAM, 0)) < 0) > err("socket"); > if (bind(fd, &sinme, sizeof(sinme)) < 0) > err("bind"); > > The socket() call is successful but the bind() fails, and the error > number it returned is 106, which, from "errno.h", is EAFNOSUPPORT. That means address family not supported. So what did you set the address family in sinme to? > and, TCP/IP stuff seems are working well (at least I can use Netscape and > ftp...) and the program actually works on Solaris. (The port number > assigned is 5001) > > Can somebody tell me: what does the error mean? and what is the problem > behind it? (something wrong with my NT system?) Where's the rest of the pertinent code, like setting up sinme? -- ,-/- __ _ _ $Bill Luebkert (_/ / ) // // DBE Collectibles / ) /--< o // // http://www.wgn.net/~dbe/ -/-' /___/_<_ Hello, Can someone tell me if gnuwin32 supports passing of open file descriptors via sendmsg or ioctl? Thanks, Yung Hsien yung@connectinc.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fjh@cs.mu.OZ.AU Tue Aug 12 00:35:00 1997 From: fjh@cs.mu.OZ.AU (Fergus Henderson) Date: Tue, 12 Aug 1997 00:35:00 -0000 Subject: rm -rf -> Circular directory structure References: <33EF7405.AADDD43E@crec.mipt.ru> Message-ID: <199708120733.RAA29256@mundook.cs.mu.OZ.AU> "Igor V. Kovalenko" writes: >Looking at ...include/sys/types.h I found that ino_t is actually >unsigned short, that is only 16 bit in size. Thus, we (in 'rm' etc..) >do compare the lower 16 bits of prepared by hash_path_name() >32-bit integer. > I think this is a main cause for $subj$, though the hashing >function itself is not as perfect as it could be (in theory). The whole idea of using a hash function to compute inode numbers is fundamentally broken. Perhaps it would be better to think about a solution that works rather than making a fundamentally broken one slightly less broken. -- Fergus Henderson | "I have always known that the pursuit WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bridge@dial-up.net Tue Aug 12 09:18:00 1997 From: bridge@dial-up.net (Leonard Weincier) Date: Tue, 12 Aug 1997 09:18:00 -0000 Subject: SSH Message-ID: Hi I am trying to compile SSH, the 1.2.20 dist, and there are a lot of problems. Has anyone else had any success ? Thanks Len ----------------------------------------------------------------------- Leonard Weincier email : - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bridge@dial-up.net Tue Aug 12 09:18:00 1997 From: bridge@dial-up.net (Leonard Weincier) Date: Tue, 12 Aug 1997 09:18:00 -0000 Subject: bsearch problem Message-ID: Hi All There appears to be a problem in the implementation of bsearch. If you pass the bsearch function an array with only one element, it will bomb because of it's handling of the bounadary condition on the binary search. I looked at the implementation of bsear ch under linux and it appears to be solid. Has anyone else found this ? TIA Len ----------------------------------------------------------------------- Leonard Weincier email : - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kimball@sgrail.com Tue Aug 12 10:22:00 1997 From: kimball@sgrail.com (Kimball Thurston) Date: Tue, 12 Aug 1997 10:22:00 -0000 Subject: ld exception c0000005 Message-ID: <33F0660A.386C98E2@sgrail.com> Hey all, If I compile the attached piece of code, call it foo.c, and I am trying to make a relocatable dll out of it using the following commands within a batch file, I get a c0000005 exception every time it tries to do the second ld... any help would be appreciated.... Thanks, Kimball -----------------------commands (foo.bat)--------------------- gcc -O2 -c -o foo.o foo.c echo EXPORTS > foo.def nm foo.o | grep "^........ T _" | sed "s/^........ T _//" >> foo.def ld --base-file __fooTemp.base --dll -o foo.so foo.o -e _dll_entry@12 dlltool --as=as --dllname foo.so --def foo.def --base-file __fooTemp.base --output-exp __fooTemp.exp REM CRASHES AFTER HERE ld --base-file __fooTemp.base __fooTemp.exp --dll -o foo.so foo.o -lgcc -lcygwin -lkernel32 -e _dll_entry@12 dlltool --as=as --dllname foo.so --def foo.def --base-file __fooTemp.base --output-exp __fooTemp.exp ld __fooTemp.exp --dll -o foo.so foo.o -e _dll_entry@12 -----------------------exception message------------------- (ld.exe 1000) In cygwin_except_handler (ld.exe 1000) Exception trapped! (ld.exe 1000) exception C0000005 at 42D792 (ld.exe 1000) exception: ax 0 bx 2ECA068 cx 2ECAC9C dx 28 (ld.exe 1000) exception: si 4000 di FFFFFE00 bp 244EEB0 sp 244EEA0 (ld.exe 1000) exception is: STATUS_ACCESS_VIOLATION (ld.exe 1000) Stack trace: (ld.exe 1000) frame 0: sp = 0x244ECD4, pc = 0x1000CEC2 (ld.exe 1000) frame 1: sp = 0x244ECF0, pc = 0x77F94512 (ld.exe 1000) frame 2: sp = 0x244ED14, pc = 0x77F88EEB (ld.exe 1000) frame 3: sp = 0x244EDA0, pc = 0x77F76266 (ld.exe 1000) frame 4: sp = 0x244EEB0, pc = 0x42D895 (ld.exe 1000) frame 5: sp = 0x244EF04, pc = 0x43094E (ld.exe 1000) frame 6: sp = 0x244F31C, pc = 0x41FA39 (ld.exe 1000) frame 7: sp = 0x244F374, pc = 0x416DCF (ld.exe 1000) frame 8: sp = 0x244F3A4, pc = 0x1000C102 (ld.exe 1000) frame 9: sp = 0x244FF94, pc = 0x1000C113 (ld.exe 1000) frame 10: sp = 0x244FFA0, pc = 0x4408A9 (ld.exe 1000) frame 11: sp = 0x244FFB0, pc = 0x40103B (ld.exe 1000) frame 12: sp = 0x244FFC0, pc = 0x77F1B304 (ld.exe 1000) frame 13: sp = 0x244FFF0, pc = 0x0 (ld.exe 1000) End of stack trace ------------------------begin code-------------------------- #define INT8_WEIGHTR 85 #define INT8_WEIGHTG 85 #define INT8_WEIGHTB 85 #define INT16_WEIGHTR 21845 #define INT16_WEIGHTG 21845 #define INT16_WEIGHTB 21845 #define FLT_WEIGHTR 0.333333 #define FLT_WEIGHTG 0.333333 #define FLT_WEIGHTB 0.333333 #define Uint8_WHITE 255 #define Uint16_WHITE 65535 #define UINT8_BITS 8 #define UINT16_BITS 16 #define UINT8_POINT_FIVE 0x80 #define UINT16_POINT_FIVE 0x8000 #define Float32_WHITE 1.0F #define ALPHAMULT8(v, a, b)\ {\ Uint16 i = (Uint16)(a) * (Uint16)(b) + UINT8_POINT_FIVE;\ (v) = (i + (i>>UINT8_BITS)) >> UINT8_BITS;\ } #define ALPHAMULT16(v, a, b)\ {\ Uint32 i = (Uint32)(a) * (Uint32)(b) + UINT16_POINT_FIVE;\ (v) = (i + (i>>UINT16_BITS)) >> UINT16_BITS;\ } #define Uint8_LUM(v, r, g, b)\ {\ Uint8 rs, gs, bs;\ ALPHAMULT8(rs, (r), INT8_WEIGHTR);\ ALPHAMULT8(gs, (g), INT8_WEIGHTG);\ ALPHAMULT8(bs, (b), INT8_WEIGHTB);\ v = rs + gs + bs;\ } #define Uint16_LUM(v, r, g, b)\ {\ Uint16 rs, gs, bs;\ ALPHAMULT16(rs, (r), INT16_WEIGHTR);\ ALPHAMULT16(gs, (g), INT16_WEIGHTG);\ ALPHAMULT16(bs, (b), INT16_WEIGHTB);\ v = rs + gs + bs;\ } #define Float32_LUM(v, r, g, b) v = (((r) * FLT_WEIGHTR) +\ ((g) * FLT_WEIGHTG) +\ ((b) * FLT_WEIGHTB)) typedef signed int Int; typedef signed char Int8; typedef signed short Int16; typedef signed int Int32; typedef signed long Int64; typedef unsigned int Uint; typedef unsigned char Uint8; typedef unsigned short Uint16; typedef unsigned int Uint32; typedef unsigned long Uint64; typedef float Float; typedef float Float32; typedef double Float64; void computeUint8_E7A160(const Uint8 *ptrA, const Uint8 *ptrB, Uint8 *ptrR,Uint aChan, Uint bChan, Uint count) { Uint8 *endR = ptrR + count; while( ptrR < endR ) { *ptrR++ = ptrA[0]; *ptrR++ = ptrA[1]; *ptrR++ = ptrA[2]; *ptrR++ = Uint8_WHITE; ptrA += aChan; ptrB += bChan; } } void computeInUint8_E7A160(const Uint8 *ptrA, const Uint8 *ptrB, Uint8 *ptrR, Uint8 *ptrT,Uint aChan, Uint bChan, Uint count) { Uint8 *endR = ptrR + count; while( ptrR < endR ) { Uint8 *ptrTR; ptrTR = ptrT; *ptrTR++ = ptrA[0]; *ptrTR++ = ptrA[1]; *ptrTR++ = ptrA[2]; *ptrTR++ = Uint8_WHITE; ptrTR = ptrT; *ptrR++ = *ptrTR++; *ptrR++ = *ptrTR++; *ptrR++ = *ptrTR++; *ptrR++ = *ptrTR++; ptrA += aChan; ptrB += bChan; } } void computeUint16_E7A160(const Uint16 *ptrA, const Uint16 *ptrB, Uint16 *ptrR,Uint aChan, Uint bChan, Uint count) { Uint16 *endR = ptrR + count; while( ptrR < endR ) { *ptrR++ = ptrA[0]; *ptrR++ = ptrA[1]; *ptrR++ = ptrA[2]; *ptrR++ = Uint16_WHITE; ptrA += aChan; ptrB += bChan; } } void computeInUint16_E7A160(const Uint16 *ptrA, const Uint16 *ptrB, Uint16 *ptrR, Uint16 *ptrT,Uint aChan, Uint bChan, Uint count) { Uint16 *endR = ptrR + count; while( ptrR < endR ) { Uint16 *ptrTR; ptrTR = ptrT; *ptrTR++ = ptrA[0]; *ptrTR++ = ptrA[1]; *ptrTR++ = ptrA[2]; *ptrTR++ = Uint16_WHITE; ptrTR = ptrT; *ptrR++ = *ptrTR++; *ptrR++ = *ptrTR++; *ptrR++ = *ptrTR++; *ptrR++ = *ptrTR++; ptrA += aChan; ptrB += bChan; } } void computeFloat32_E7A160(const Float32 *ptrA, const Float32 *ptrB, Float32 *ptrR,Uint aChan, Uint bChan, Uint count) { Float32 *endR = ptrR + count; while( ptrR < endR ) { *ptrR++ = ptrA[0]; *ptrR++ = ptrA[1]; *ptrR++ = ptrA[2]; *ptrR++ = Float32_WHITE; ptrA += aChan; ptrB += bChan; } } void computeInFloat32_E7A160(const Float32 *ptrA, const Float32 *ptrB, Float32 *ptrR, Float32 *ptrT,Uint aChan, Uint bChan, Uint count) { Float32 *endR = ptrR + count; while( ptrR < endR ) { Float32 *ptrTR; ptrTR = ptrT; *ptrTR++ = ptrA[0]; *ptrTR++ = ptrA[1]; *ptrTR++ = ptrA[2]; *ptrTR++ = Float32_WHITE; ptrTR = ptrT; *ptrR++ = *ptrTR++; *ptrR++ = *ptrTR++; *ptrR++ = *ptrTR++; *ptrR++ = *ptrTR++; ptrA += aChan; ptrB += bChan; } } extern int __stdcall dll_entry (void * h, unsigned long reason, void *ptr); int __stdcall dll_entry (void *h, unsigned long reason, void *ptr) { h; reason; ptr; switch (reason) { case 1: break; case 2: break; case 3: break; case 0: break; } return 1; } asm(".section .idata$3\n" ".long 0,0,0,0,0"); - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From MAILER-DAEMON@ccc.edu Tue Aug 12 10:22:00 1997 From: MAILER-DAEMON@ccc.edu (Mail Delivery Subsystem) Date: Tue, 12 Aug 1997 10:22:00 -0000 Subject: Returned mail: Cannot send message for 3 days Message-ID: <9708121722.DU07967@coefw05.ccc.edu> ----- Transcript of session follows ----- 421 unix1.ccc.edu (blue)... Deferred: Connection refused by unix1.ccc.edu ----- Unsent message follows ----- Received: by coefw05.ccc.edu; (5.65v3.2/1.3/10May95) id AA26812; Sat, 9 Aug 1997 11:22:12 -0500 Received: from cc:Mail by ccgate.ccc.edu id AA871150691; Sat, 09 Aug 97 11:02:16 cst Date: Sat, 09 Aug 97 11:02:16 cst From: gnu-win32@cygnus.com Message-Id: <9707098711.AA871150691@ccgate.ccc.edu> To: gnu-win32@cygnus.com Subject: Re: latest Sergey's update (08/01) crashes on Win95 dmeans@bellsouth.net (David Means) wrote: > Ditto me too. I've got a valid /etc/passwd and /etc/group. Here too (Win95 OSR2). > If I execute a command at the prompt, the command will run, i.e., mount > or /bin/ls, but the system will hang afterwards. The mouse still works, > but nothing will get the system back save a three-finger-salute ( > ctrl-alt-del ). Sometimes it will take a moment or two before it hangs, > at least it did after the mount command. Exactly the same symptoms. Pete Jordan = = = = = = = = = = = = = = = = = = = = = = = Horus Communications http://www.horus.cix.co.uk/ = = = = = = = = = = = = = = = = = = = = = = = "'Not twisted,' Salzy once said of her own passion, 'it is helical. That sounds better.'" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From MAILER-DAEMON@ccc.edu Tue Aug 12 10:25:00 1997 From: MAILER-DAEMON@ccc.edu (Mail Delivery Subsystem) Date: Tue, 12 Aug 1997 10:25:00 -0000 Subject: Returned mail: Cannot send message for 3 days Message-ID: <9708121726.EG07967@coefw05.ccc.edu> ----- Transcript of session follows ----- 421 unix1.ccc.edu (blue)... Deferred: Connection refused by unix1.ccc.edu ----- Unsent message follows ----- Received: by coefw05.ccc.edu; (5.65v3.2/1.3/10May95) id AA20826; Sat, 9 Aug 1997 11:22:50 -0500 Received: from cc:Mail by ccgate.ccc.edu id AA871150633; Sat, 09 Aug 97 10:49:57 cst Date: Sat, 09 Aug 97 10:49:57 cst From: gnu-win32@cygnus.com Message-Id: <9707098711.AA871150633@ccgate.ccc.edu> To: gnu-win32@cygnus.com Subject: Re: Perl5.004_01 patch for b18 cgf@bbc.com (Chris Faylor) wrote: > Actually since, to the best of my knowledge, the patch that is pending > for perl5.004_02 is actually the patch on my web site, then I wouldn't > be too certain that things will work much better at all. Well, I didn't have serious problems building your patched code - the biggest hassle was with patch itself and that won't affect the 5.004_02 distribution. If the niggles I mentioned in my mails to you are fixed ($perlpath problem with the inconsistency in which directory config.sh is called from, "read answ" kludge around for Win95, path delimiter inconsistency in built GNUmakefile and MakeMaker generating filenames with "::" in for man files), all should be well AFAICS. Pete Jordan = = = = = = = = = = = = = = = = = = = = = = = Horus Communications http://www.horus.cix.co.uk/ = = = = = = = = = = = = = = = = = = = = = = = "'Not twisted,' Salzy once said of her own passion, 'it is helical. That sounds better.'" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From cowan@pswtech.com Tue Aug 12 14:21:00 1997 From: cowan@pswtech.com (Chris Cowan) Date: Tue, 12 Aug 1997 14:21:00 -0000 Subject: perldb under emacs Message-ID: <33F0D3A2.8CBDB397@pswtech.com> A while back, somebody posted the details for getting perldb to work proper under emacs with bash, under NT. Could someone repost it? Thanks, -- Chris Cowan cowan@pswtech.com (512) 342-3635 (512) 345-4976 - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From si10647@ci.uminho.pt Tue Aug 12 14:30:00 1997 From: si10647@ci.uminho.pt (Daniel P.C. da Silva) Date: Tue, 12 Aug 1997 14:30:00 -0000 Subject: filename expansion References: <311E5922.9F864639@ci.uminho.pt> <33EFD432.380D@wgn.net> Message-ID: <33F0C7A0.8CC00BFA@ci.uminho.pt> $Bill Luebkert wrote: > There is a bug (IMHO) in the argument passing of non-cygwin shells. > Cygwin assumes that the shell has no globbing ability and globs > the cmdline. Furthermore, it ignores ' and ", so you can't escape > it. I think I may have found a kludge around it though. > > Try: > xargs find . -name > *.pl -print > ^Z > or > xargs find . -name > *.pl > -print > ^Z Other solution is to write a program, not cygnus win32, that echoes his arguments to sh.Like this 4dos batch file: (called "sh-c.btm") @echo off echo %& | sh Now, I can call a cygnus win 32 command: sh-c find . -name "*.pl" -- ------------------------------------------- Daniel P.C. Silva Universidade do Minho si10647@ci.uminho.pt ------------------------------------------- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Ian@kiwiplan.co.nz Tue Aug 12 15:51:00 1997 From: Ian@kiwiplan.co.nz (Ian Collins) Date: Tue, 12 Aug 1997 15:51:00 -0000 Subject: Spooler Message-ID: <07551A47EAD6D01186430060B025C6F40A6C5B@modem1.kiwiplan2.co.nz> Does anyone know of a ported lp sub-system for gnu-win32? I can't seem to find any spooler related commands at all. Many Regards, Ian Collins - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From noer@cygnus.com Tue Aug 12 17:55:00 1997 From: noer@cygnus.com (Geoffrey Noer) Date: Tue, 12 Aug 1997 17:55:00 -0000 Subject: bsearch problem References: Message-ID: <199708130054.RAA21064@cirdan.cygnus.com> Leonard Weincier wrote: > > There appears to be a problem in the implementation of bsearch. This will be fixed in the next beta release. My current guess is that it will be out sometime in September. -- Geoffrey Noer noer@cygnus.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bzinger@iname.com Tue Aug 12 18:20:00 1997 From: bzinger@iname.com (Micheal A. Benzinger) Date: Tue, 12 Aug 1997 18:20:00 -0000 Subject: termcap interface code References: <3.0.1.32.19970811203638.0069123c@elwood.pionet.net> Message-ID: <3.0.3.32.19970812202007.0091a100@pop.flash.net> Hi Justin, You need to download termcap-1.3.tar.gz, untar, configure, make and install. This will make the termcap header and library available to the compiler/linker. If I remember correctly this required no changes to the installed files. Good luck, Mike Benzinger > >The comment nearby says these are Unix termcap interface code, >so does anyone know where I can get these? Or if they're >already in the cygwin32 b18 distribution..? > >Thanks. >Justin Hopkins - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bilbo@ameritech.net Tue Aug 12 21:32:00 1997 From: bilbo@ameritech.net (Bilbo Baggins) Date: Tue, 12 Aug 1997 21:32:00 -0000 Subject: SSH References: Message-ID: On Tue, 12 Aug 1997, Leonard Weincier wrote: > I am trying to compile SSH, the 1.2.20 dist, and there are a lot of problems. > Has anyone else had any success ? it basically went without a hitch.. cpl of minor complaints.. scp: doesnt seem to work.. havent had time to check into it (compiled fine) sshd: like above compiles fine but after authenticating the user complains about not being able to change the uids.. ssh: works as expected.. rest of the included tools seem to work as well.. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From khan@xraylith.wisc.edu Tue Aug 12 21:53:00 1997 From: khan@xraylith.wisc.edu (Mumit Khan) Date: Tue, 12 Aug 1997 21:53:00 -0000 Subject: bsearch problem References: Message-ID: <9708130354.AA05870@modi.xraylith.wisc.edu> Leonard Weincier writes: > Hi All > > There appears to be a problem in the implementation of bsearch. If you pass t > he bsearch function an array with only one element, it will bomb because of i > t's handling of the bounadary condition on the binary search. I looked at the > implementation of bsear > ch under linux and it appears to be solid. > > Has anyone else found this ? > I'm attaching a copy of my posting to gnu-win32 which includes a patch. Geoffrey Noer has rewritten the implementation for b19, but until then this will have to do. -- using template mhl.format -- Date: Tue, 25 Mar 1997 22:54:24 CST To: xxx cc: gnu-win32@cygnus.com From: Mumit Khan Subject: Re: nasty bug in bsearch() in b17.1 under Windows'95???? xxx writes: > > The bsearch.c in newlib contains a quite explicit and quite erroneous > comparison to the item one past the last range whenever the key isn't > found. Normally this is merely pointless extra work, but if the key is > greater than the last element, then it is an out of bounds reference, > just as you have discovered. The "easiest" fix is to simply remove the > lines > > if (compar (key, base) == 0) > return (_PTR) base; > Exactly. Just looked at the source, and it's quite obvious. As you mention, it also does an extra comparison for no reason whatsoever in the other cases. Here's the trivial patch, which took non-trivial number of hours to find ;-) debugging f771 on a '95 box is no fun. *** bsearch.c.~1 Tue Mar 25 22:47:42 1997 --- bsearch.c Tue Mar 25 22:47:46 1997 *************** _DEFUN (bsearch, (key, base, nmemb, size *** 94,100 **** } - if (compar (key, base) == 0) - return (_PTR) base; - return NULL; } --- 94,97 ---- Regards, Mumit -- khan@xraylith.wisc.edu http://www.xraylith.wisc.edu/~khan/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From David.Billinghurst@restech.cra.com.au Tue Aug 12 21:53:00 1997 From: David.Billinghurst@restech.cra.com.au (Billinghurst, David) Date: Tue, 12 Aug 1997 21:53:00 -0000 Subject: ftp locations Message-ID: I've built rcs-5.7 and cvs-1.9 using gnuwin32 b18, and plan to port some other stuff. A couple of people have asked me for binaries, but we do not allow inwards ftp. Any ideas on where I could put them. +++++++++++++++++++++++++++++++++++++++++ (Mr) David Billinghurst Comalco Research Centre PO Box 316, Thomastown, Vic, Australia, 3074 Phone: +61 3 9469 0642 FAX: +61 3 9462 2700 Email: David.Billinghurst@restech.cra.com.au It's not my speeling that's the trouble. It's just taht I can't tipe. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bilbo@ameritech.net Tue Aug 12 21:53:00 1997 From: bilbo@ameritech.net (Bilbo Baggins) Date: Tue, 12 Aug 1997 21:53:00 -0000 Subject: SSH References: Message-ID: > scp: doesnt seem to work.. havent had time to check into it (compiled > fine) > > sshd: like above compiles fine but after authenticating the user, > complains about not being able to change the uids.. ok the above problem cleared.. perms & ownership problem.. now the problem is that sshd is trying to allocate a pty and do a select() on it... - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gunther.ebert@ixos-leipzig.de Tue Aug 12 23:27:00 1997 From: gunther.ebert@ixos-leipzig.de (Gunther Ebert) Date: Tue, 12 Aug 1997 23:27:00 -0000 Subject: login and telnetd References: <9708060928.AA38659@hagar.aspentec.com> Message-ID: <33F16FCC.4EF7@ixos-leipzig.de> Jerome Gay wrote: > > telnetd : > when running 'inetd -d /etc/inetd.conf' in one bash session, and running > telnet (Sergey's telnet.exe) in another bash session, trying to connect on > localhost, I got : > > telnet client side : > Trying 127.0.0.1... > setuid called, and it's not done yet. > Connected to pathfinder.[...] > Escape character is '^]'. > Connection closed by foreign host. > > in bash session where inetd was started : > ... > someone wants telnet > accept, ctrl 3 > + Closing from 17 > 1022 execl /usr/local/libexec/in.telnetd > 1022 reaped, status 0x100 > > Is anyone using telnetd successfully ? > Any idea ? > Did anyone answer this question? I'm running exactly into this problem. -- Gunther Ebert iXOS Anwendungs-Software GmbH Angerstrasse 40-42 D-04177 Leipzig Phone : +49 341 48503-0 Fax : +49 341 48503-99 E-mail: mailto:gunther.ebert@ixos-leipzig.de www : http://www.ixos-leipzig.de - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bridge@dial-up.net Wed Aug 13 00:10:00 1997 From: bridge@dial-up.net (Leonard Weincier) Date: Wed, 13 Aug 1997 00:10:00 -0000 Subject: SSH References: Message-ID: Hi Bilbo A Couple of questions : Did you use b18 and the standard ssh make files? Are you using NT or 95 ? I am using 95 and when I ran ./configure it determined the sizeof int, double and short to be 0. I have subsequently been hacking #defines etc to get it to compile. When make tries to compile libz, I get a lot of errors about non-existent .o files ? This seems to be corrected when I remove the -pipe option from gcc ? Thanks for your help Len On Wed, 13 Aug 1997, Bilbo Baggins wrote: > > scp: doesnt seem to work.. havent had time to check into it (compiled > > fine) > > > > sshd: like above compiles fine but after authenticating the user, > > complains about not being able to change the uids.. > > ok the above problem cleared.. perms & ownership problem.. > > now the problem is that sshd is trying to allocate a pty and do a select() > on it... > > > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gay@aspentec.com Wed Aug 13 00:20:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Wed, 13 Aug 1997 00:20:00 -0000 Subject: Making telnetd and rxvt to work... Message-ID: <9708130719.AA28961@hagar.aspentec.com> Gunther- Yesterday, I evenually solved these problems by doing the following : 1- Ensure the user running inetd has the following rights (change them as appropriate using User Manager) : Act as part of the operating system, Create a token object, Increase quotas, Log on as a service, Replace a process level token 2- run 'inetd -d /etc/inetd.conf' ( -d in lowercase, especially when running from NT command line) 3- ensure /etc/inetd.conf is ok (changing the path of in.telnetd if needed, and replacing 'root' with the name of the account used to run inetd, sysroot in my case) : telnet stream tcp nowait sysroot /usr/local/libexec/in.telnetd 4- ensure you have valid /etc/passwd (and /etc/groups) : - LF and not CR/LF format, - valid home directory field, - /bin/bash and not /bin/sh as login shell exemple : sysroot::500:0::/home/sysroot:/bin/bash Hope this helps, Jerome "ZoSo -----Original Message----- From: Gunther Ebert To: Jerome Gay Cc: gnu-win32@cygnus.com Date: Wednesday, August 13, 1997 8:27 AM Subject: Re: login and telnetd >Jerome Gay wrote: >> >> telnetd : >> when running 'inetd -d /etc/inetd.conf' in one bash session, and running >> telnet (Sergey's telnet.exe) in another bash session, trying to connect on >> localhost, I got : >> >> telnet client side : >> Trying 127.0.0.1... >> setuid called, and it's not done yet. >> Connected to pathfinder.[...] >> Escape character is '^]'. >> Connection closed by foreign host. >> >> in bash session where inetd was started : >> ... >> someone wants telnet >> accept, ctrl 3 >> + Closing from 17 >> 1022 execl /usr/local/libexec/in.telnetd >> 1022 reaped, status 0x100 >> >> Is anyone using telnetd successfully ? >> Any idea ? >> > >Did anyone answer this question? I'm running exactly into this >problem. > >-- > >Gunther Ebert >iXOS Anwendungs-Software GmbH >Angerstrasse 40-42 >D-04177 Leipzig > >Phone : +49 341 48503-0 >Fax : +49 341 48503-99 >E-mail: mailto:gunther.ebert@ixos-leipzig.de >www : http://www.ixos-leipzig.de > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From david@coent.demon.co.uk Wed Aug 13 01:05:00 1997 From: david@coent.demon.co.uk (David Coe) Date: Wed, 13 Aug 1997 01:05:00 -0000 Subject: termcap interface code References: <3.0.1.32.19970811203638.0069123c@elwood.pionet.net> Message-ID: <33F1695D.337A@coent.demon.co.uk> Justin Hopkins wrote: > > Hi, I'm trying to compile texinfo 3.9 and I'm getting some > undefined references at link-time. These are the functions: > > extern int tgetnum (), tgetflag (), tgetent (); > extern char *tgetstr (), *tgoto (); > extern char *getenv (); > extern void tputs (); > > The comment nearby says these are Unix termcap interface code, > so does anyone know where I can get these? Or if they're > already in the cygwin32 b18 distribution..? You need to install termcap-1.3 first (small patch attached). Both then come up without problems with ./configure; make. Best wishes -- Dr David Coe \=\ 58 Fairlawn Drive, East Grinstead \=\ Tel +44 1342 326860 West Sussex, RH19 1NT, United Kingdom \=\ Fax +44 1342 316019 From gay@aspentec.com Wed Aug 13 01:08:00 1997 From: gay@aspentec.com (Jerome Gay) Date: Wed, 13 Aug 1997 01:08:00 -0000 Subject: an x environment Message-ID: <9708130807.AA30060@hagar.aspentec.com> Steve- After reading your messages, I tried to setup fvwm2 under gnu-win32, it compiled without any major problems, I also compiled and installed pixmap... fvwm2 works fine except I get the message (at the beginning) : [FVWM][executeModule]: <> No such module 'PixmapPath' in ModulePath '/usr/X11R6.3/lib/X11/fvwm2' and I don't get these nice buttons on the fvwm2 buttonbar... I found out that the pixmap distribution does not include any .pm files.... Where can I get some pm files to be used for buttons ? Where can I get some nice fvwm2rc examples ? Jerome "ZoSo" -----Original Message----- From: Steve o To: Michael Anderson Cc: gnu-win32@cygnus.com Date: Wednesday, August 06, 1997 4:25 AM Subject: Re: an x environment >Michael Anderson wrote: >> >> Steve o wrote: >> > >> > fvwm2 ported without a hitch. >> >> Hi Steve: >> >> I've been using bash windows and vi for my work and wondering just how I >> might get X and fvwm working. Can you tell me the steps needed to do >> this? Which X server are you using and what other software needs to be >> installed? Config files? If you posted the process to the newsgroup, it >> might help out others wondering how to do this. > >I am using Hummingbird's Exceed X Server, > http://www.hummingbird.com/products/exceed/ >I set the environment variable DISPLAY=:0 > >I installed the "X client libs for NT" from > http://www.cygnus.com/misc/gnu-win32/ > >Fvwm needs xpm so I got that from: > ftp://ftp.x.org/contrib/libraries/xpm-3.4j.tar.gz > >Fvwm can be obtained from: > http://www.hpc.uh.edu/fvwm/ > >The steps to install are: > tar -xvzf package.tar.gz > cd package > xmkmf -a > make > make install > >Sometimes the install target in the Makefile will have: > cp foo /usr/local/bin/foo >These need to be changed to: > cp foo.exe /usr/local/bin/foo.exe > >There are some sample .fvwm2rc files included in the distribution, >copy one of these to your home directory. > >If you run fvwm, you'll want to set the Exceed > Xconfig->Window Mode->window manager to X. > >You'll need Sergey's coolview and remote.tar.gz. >-steve >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jiten@meghadoot.hiso.honeywell.com Wed Aug 13 01:40:00 1997 From: jiten@meghadoot.hiso.honeywell.com (Jiten J Shah) Date: Wed, 13 Aug 1997 01:40:00 -0000 Subject: GNU C and MSVC++ libraries Message-ID: <33F171B6.6E35@hiso.honeywell.com> Hello !!! I am using GNU C available from ` ftp://ftp.cygnus.com/pub/gnu-win32/latest ' on NT. It is very very useful for me to port my UNIX application onto NT with that. I need some help. We have Hummingbird XDK on NT for developing/porting X client application. They provide libraries like xlib.lib, xt.lib on NT. From FAQ, I understand that mixing of GNU C and MSVC++ is possible provided we use link.exe from Microsoft. We tried generating object file with CYGNUS C compiler. And use link.exe to link it with xlib.lib. It generates executable but does not run properly. It gives following error : "The procedure entry point could not be located in the dynamic link library xlib.dll". Hummingbird tech support person got the same error and advised me to get some help from CYGNUS. Can you please give me some information on how to use MSVC++ libraries with GNU C? Thanks in advance. Jiten J. Shah Honeywell India Software Operation Bangalore INDIA - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pete@horus.cix.co.uk Wed Aug 13 01:57:00 1997 From: pete@horus.cix.co.uk (Pete Jordan) Date: Wed, 13 Aug 1997 01:57:00 -0000 Subject: A tool for executing Windows and DOS programs from bash References: Message-ID: pete@horus.cix.co.uk (Pete Jordan) wrote: > I've written a little launch program that translates pathnames from > GNU-Win32 format to Windows format so I can (for example) launch Word > from bash by typing: > > $word foo.doc > or: > $word /usr/baz.doc This is now available for download from http://www.horus.cix.co.uk/downloads.html Pete Jordan = = = = = = = = = = = = = = = = = = = = = = = Horus Communications http://www.horus.cix.co.uk/ = = = = = = = = = = = = = = = = = = = = = = = "As soon as we have defined something, we have ipso facto created three things: whatever is included in our definition, whatever is not included in our definition, and the relationship between them." - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From platsas@egnatia.ee.auth.gr Wed Aug 13 07:25:00 1997 From: platsas@egnatia.ee.auth.gr (Sakis Platsas) Date: Wed, 13 Aug 1997 07:25:00 -0000 Subject: Curses? Message-ID: Hi! I am trying out GNU-Win32 and up till now I found it really useful for people needing unix useability under Windows. However, I wanted to try compiling some programs that needed a form of 'curses' libs, like curses, ncurses etc. When I tried to compile PDCURSES, the too-many constant redefinitions seemed to 'swamp' the GCC compiler to a grinding halt with errors. Have you ever encountered such a difficulty? Is there ANY precompiled curses library available? Thanks in advance! _____________ Sakis Platsas platsas@egnatia.ee.auth.gr - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From raz@cmg.FCNBD.COM Wed Aug 13 07:25:00 1997 From: raz@cmg.FCNBD.COM (Robert Allan Zeh) Date: Wed, 13 Aug 1997 07:25:00 -0000 Subject: b18 and the Java JNI with jdk 1.1.3 Message-ID: <199708131408.JAA12703@tnd93.cmg.FCNBD.COM> I'm trying to use b18 to implement some native Java functions with jdk 1.1.3 using the Java Native Interface. Right now I'm at a point where I can write native java functions that return any of the primitive Java types, like jint or jchar, and everything works out fine. But I'm unable to return actual Java objects (other than NULL --- that works just fine). When I do return a non-null Java object the program crashes. I've tried returning unmodified objects that are passed in, and that doesn't work either. Has anyone else had any luck using the JNI with b18? Thanks, Robert Zeh razeh@balr.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From p.dalgaard@kubism.ku.dk Wed Aug 13 07:25:00 1997 From: p.dalgaard@kubism.ku.dk (Peter Dalgaard BSA) Date: Wed, 13 Aug 1997 07:25:00 -0000 Subject: Returned mail: Cannot send message for 3 days References: <9708121722.DU07967@coefw05.ccc.edu> Message-ID: Mail Delivery Subsystem writes: > > ----- Transcript of session follows ----- > 421 unix1.ccc.edu (blue)... Deferred: Connection refused by unix1.ccc.edu > ... > From: gnu-win32@cygnus.com ^^^^^^^^^^^^^^^^^^^^ > Message-Id: <9707098711.AA871150691@ccgate.ccc.edu> > To: gnu-win32@cygnus.com ^^^^^^^^^^^^^^^^^^^^ Uh-oh.... -- O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk) FAX: (+45) 35327907 - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bilbo@ameritech.net Wed Aug 13 07:25:00 1997 From: bilbo@ameritech.net (Dan Maynard) Date: Wed, 13 Aug 1997 07:25:00 -0000 Subject: SSH References: Message-ID: > Did you use b18 and the standard ssh make files? > Are you using NT or 95 ? NT4SP3 > I am using 95 and when I ran ./configure it determined the sizeof int, double and short to be 0. > I have subsequently been hacking #defines etc to get it to compile. possibility of mailing me (not the list) the output of configure? > > When make tries to compile libz, I get a lot of errors about non-existent .o files ? This seems to be corrected when I remove the -pipe option from gcc ? right, -pipe is broken - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From MCLSSAA2@fs2.mt.umist.ac.uk Wed Aug 13 08:26:00 1997 From: MCLSSAA2@fs2.mt.umist.ac.uk (Anthony.Appleyard) Date: Wed, 13 Aug 1997 08:26:00 -0000 Subject: A beginner's query Message-ID: <7790037C47@fs2.mt.umist.ac.uk> (0) About how many messages a day are there on gnu-win32? (1) I looked at the ..../gnu-win32/ FTP directory on Sunsite in UK, and I found a confused tree with no apparent index file telling me what is what. What should I load? (2) It would be useful if these files were packed by .zip instead of by .gz (3) How well advanced and debugged is gnu-win32 compared with djgpp C/C++ for DOS? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pete@horus.cix.co.uk Wed Aug 13 08:26:00 1997 From: pete@horus.cix.co.uk (Pete Jordan) Date: Wed, 13 Aug 1997 08:26:00 -0000 Subject: New patch References: <01BCA5AA.00AB4920@sos> Message-ID: sos@prospect.com.ru (Sergey Okhapkin) wrote: > - Chris Faylor's (cfg@bbc.com) waiting code temporary removed. The code > is right, works fine on Windows NT and is neccessary to support a > correct semantics of wait syscalls, but it have unexpected behavior on > Windows 95. > No workaround found. > - Some syscall semantics-related changes. Seems to work fine here (Win95 OSR2) Pete Jordan = = = = = = = = = = = = = = = = = = = = = = = Horus Communications http://www.horus.cix.co.uk/ = = = = = = = = = = = = = = = = = = = = = = = "'Not twisted,' Salzy once said of her own passion, 'it is helical. That sounds better.'" - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From phoenix@gower.net Wed Aug 13 11:06:00 1997 From: phoenix@gower.net (James A Leath) Date: Wed, 13 Aug 1997 11:06:00 -0000 Subject: Win95 Compiler Message-ID: <2.2.32.19970813180442.00673284@mail.gower.net> I have Windows 95, and I'm not very knowledgeable about compilers and the like, and I need to compile a C file on my computer. I was wondering if what was on your Cygnus page had what I needed, and I downloaded the usertools.exe (4.4 megs) and I'm not sure if this is what I needed. I notice the developers kit needed a compiler to compile it, and I thought that was odd since I don't have a compiler to compile it. Please help me. Thanks, :) James A. Leath phoenix@gower.net http://www.gower.net/phoenix/index.html - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ccurtin@trintech.com Wed Aug 13 11:50:00 1997 From: ccurtin@trintech.com (Colman Curtin) Date: Wed, 13 Aug 1997 11:50:00 -0000 Subject: sh crashes on 95 Message-ID: <199708131851.TAA30622@muzak.trintech.ie> > > Okay... when I'm doing make, sh has a habit of crashing. It's got bad >recently. Where can I get da patch? Ack! > > SJ > Yip I've run into the same problem. Running configure scripts it seems to choke at no particular point some times it get further than others. If I reboot under NT4 configure completes to the end. Looking at the processes while a configure is in progress I have seen up to six instances of sh running, I haven't seen more as it normally dies and more often than not needs the three fingered salute! Anyone got any ideas?? running w95,16MB with Sergery's latest release.(the one with the borken wait code on W95 removed) thanx Colman. ____________________________________________________________ Colman Curtin mailto:ccurtin@trintech.ie Software Engineer ------------------------------------------------------------ Trintech (Manufacturing) Ltd, http://www.trintech.com/ South County Business Park, Leopardstown, Dublin 18. Tel +353-1-2956766 Fax +353-1-2954735 ------------------------------------------------------------ Q: How many IBM CPU's does it take to execute a job? A: Four; three to hold it down, and one to rip its head off. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From veger@primenet.com Wed Aug 13 13:33:00 1997 From: veger@primenet.com (kD00D) Date: Wed, 13 Aug 1997 13:33:00 -0000 Subject: Makefiles Message-ID: <199708132033.NAA08530@usr03.primenet.com> D:\DIST\SRC> MAKE MAKE SMAUG MAKE: *** No rule to make target `act_com.o', needed by `smaug'. Stop. MAKE.EXE: *** [all] Error 2 {Actual screen capture} I am trying to get SMAUG for Win32 to compile. I am having problems. The SMAUG works fine under Win95 and WinNt here. However, I cannont get it to compile a make file. Aboce it the actual screen shot of the error I get. Any suggestions would be appreciated. Below is a a copy of the makefile I am using. The author uses the exact same file, and has it working on his system. If I manually compile the gcc -c to all the *.c files then make will work, however, I get errors in the final executable. Thanks for any help in advance. ----- MAKEFILE. ----- CC = gcc PROF = NOCRYPT = #Uncomment the next line if you want request support #DBUGFLG = -DREQUESTS C_FLAGS = -g3 -Wall $(PROF) $(NOCRYPT) $(DBUGFLG) L_FLAGS = $(PROF) O_FILES = act_comm.o act_info.o act_move.o act_obj.o act_wiz.o boards.o \ build.o clans.o comm.o comments.o const.o db.o deity.o fight.o \ handler.o hashstr.o ibuild.o id.o interp.o magic.o makeobjs.o \ mapout.o misc.o mpxset.o mud_comm.o mud_prog.o player.o requests.o \ reset.o save.o shops.o skills.o special.o tables.o track.o update.o \ grub.o C_FILES = act_comm.c act_info.c act_move.c act_obj.c act_wiz.c boards.c \ build.c clans.c comm.c comments.c const.c db.c deity.c fight.c \ handler.c hashstr.c ibuild.c id.c interp.c magic.c makeobjs.c \ mapout.c misc.c mpxset.c mud_comm.c mud_prog.c player.c requests.c \ reset.c save.c shops.c skills.c special.c tables.c track.c update.c \ grub.c H_FILES = mud.h bet.h all: # co $(H_FILES) make smaug # rm -f $(H_FILES) smaug: $(O_FILES) rm -f smaug.exe $(CC) $(L_FLAGS) -o smaug $(O_FILES) chmod g+w smaug.exe chmod g+w $(O_FILES) .c.o: mud.h $(CC) -c $(C_FLAGS) -DNOCRYPT $< clean: rm -f $(O_FILES) ----- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From startnet@nation.org Wed Aug 13 15:57:00 1997 From: startnet@nation.org (Derek Reynolds) Date: Wed, 13 Aug 1997 15:57:00 -0000 Subject: telnetd whereabouts? Message-ID: <1.5.4.32.19970813225315.0067657c@nation.org> Hi there, Can someone please let me know where the telnetd port can be found in bin format? thanks Derek Reynolds - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dahms@ifk20.mach.uni-karlsruhe.de Wed Aug 13 15:57:00 1997 From: dahms@ifk20.mach.uni-karlsruhe.de (dahms@ifk20.mach.uni-karlsruhe.de) Date: Wed, 13 Aug 1997 15:57:00 -0000 Subject: SSH Message-ID: <009B8BC8.1913CDC0.21318@ifk20.mach.uni-karlsruhe.de> Hi Leonard, you wrote: : I am using 95 and when I ran ./configure it determined the sizeof int, : double and short to be 0. : When make tries to compile libz, I get a lot of errors about non-existent : .o files ? : This seems to be corrected when I remove the -pipe option from gcc ? Could configure have been failed because of using -pipe, too? Then you might try again after "rm config.cache", "make distclean" or whatever this package supports/needs. Bye, Heribert (dahms@ifk20.mach.uni-karlsruhe.de) - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From veger@primenet.com Wed Aug 13 16:05:00 1997 From: veger@primenet.com (kD00D) Date: Wed, 13 Aug 1997 16:05:00 -0000 Subject: Makefiles Message-ID: <199708132305.QAA15842@usr02.primenet.com> D:\DIST\SRC> MAKE MAKE SMAUG MAKE: *** No rule to make target `act_com.o', needed by `smaug'. Stop. MAKE.EXE: *** [all] Error 2 {Actual screen capture} I am trying to get SMAUG for Win32 to compile. I am having problems. The SMAUG works fine under Win95 and WinNt here. However, I cannont get it to compile a make file. Aboce it the actual screen shot of the error I get. Any suggestions would be appreciated. Below is a a copy of the makefile I am using. The author uses the exact same file, and has it working on his system. If I manually compile the gcc -c to all the *.c files then make will work, however, I get errors in the final executable. Thanks for any help in advance. ----- MAKEFILE. ----- CC = gcc PROF = NOCRYPT = #Uncomment the next line if you want request support #DBUGFLG = -DREQUESTS C_FLAGS = -g3 -Wall $(PROF) $(NOCRYPT) $(DBUGFLG) L_FLAGS = $(PROF) O_FILES = act_comm.o act_info.o act_move.o act_obj.o act_wiz.o boards.o \ build.o clans.o comm.o comments.o const.o db.o deity.o fight.o \ handler.o hashstr.o ibuild.o id.o interp.o magic.o makeobjs.o \ mapout.o misc.o mpxset.o mud_comm.o mud_prog.o player.o requests.o \ reset.o save.o shops.o skills.o special.o tables.o track.o update.o \ grub.o C_FILES = act_comm.c act_info.c act_move.c act_obj.c act_wiz.c boards.c \ build.c clans.c comm.c comments.c const.c db.c deity.c fight.c \ handler.c hashstr.c ibuild.c id.c interp.c magic.c makeobjs.c \ mapout.c misc.c mpxset.c mud_comm.c mud_prog.c player.c requests.c \ reset.c save.c shops.c skills.c special.c tables.c track.c update.c \ grub.c H_FILES = mud.h bet.h all: # co $(H_FILES) make smaug # rm -f $(H_FILES) smaug: $(O_FILES) rm -f smaug.exe $(CC) $(L_FLAGS) -o smaug $(O_FILES) chmod g+w smaug.exe chmod g+w $(O_FILES) .c.o: mud.h $(CC) -c $(C_FLAGS) -DNOCRYPT $< clean: rm -f $(O_FILES) ----- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Fuehrer@seabase.com Wed Aug 13 16:05:00 1997 From: Fuehrer@seabase.com (Gary Fuehrer) Date: Wed, 13 Aug 1997 16:05:00 -0000 Subject: mmap(,,,MAP_PRIVATE,,) bug and fix Message-ID: <6CC63E2E4FC1D011A2A700609716117A266F05@seawolf> The B18 implementation of mmap() fails when MAP_PRIVATE is used. The problem is caused because of this pair of lines in mmap.cc: DWORD access = (prot & PROT_WRITE) ? FILE_MAP_WRITE : FILE_MAP_READ; access |= (flags & MAP_PRIVATE) ? FILE_MAP_COPY : 0; When MAP_PRIVATE is set, then "access" = "FILE_MAP_READ/WRITE | MAP_PRIVATE". When "access" is subsequently passed to MapViewOfFile(), it fails. The Microsoft documentation for MapViewOfFile() states the following concerning the access parameter: "This parameter can be one of the following values:...". So, the failure is because of trying to use more than one of the access types at the same time. Fix: Change the above lines of code to the following (only the second line is different): DWORD access = (prot & PROT_WRITE) ? FILE_MAP_WRITE : FILE_MAP_READ; if (flags & MAP_PRIVATE) access = FILE_MAP_COPY; I've tested this change and it seems to work correctly. Will this fix make it in for B19? Gary - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mphalpin@iag.net Wed Aug 13 19:01:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Wed, 13 Aug 1997 19:01:00 -0000 Subject: ircII? Anybody done it? (Also pico, pine, lynx) Message-ID: <199708140122.VAA15740@iag.net> Hi, I'm just curious if anyone has done a pico/pine/lynx or ircII port? I have trouble with them, maybe someone else has had better luck (btw, I'm looking for something that will run within a telnetd window.) I tried ircII, but it works, only problemmatically. I'm looking for full screen within a telnet session, (I don't like -d stuff). The ircII I wrote has a weird problem: Whenever you enter a weird, command, the command bar and the command scroll up!!! If you have a working one (or BitchX will do, I suppose. ;), please let me know ASAP! SJ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mphalpin@iag.net Wed Aug 13 19:01:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Wed, 13 Aug 1997 19:01:00 -0000 Subject: New patch Message-ID: <199708140124.VAA15972@iag.net> I have OSR/2, and I had the problem. Seemed to occur on "make" and "configure"s. Anyway, the new patch appears to fix it. - SJ ---------- > From: Pete Jordan > To: gnu-win32@cygnus.com > Subject: Re: New patch > Date: Wednesday, August 13, 1997 11:00 AM > > sos@prospect.com.ru (Sergey Okhapkin) wrote: > > > - Chris Faylor's (cfg@bbc.com) waiting code temporary removed. The code > > is right, works fine on Windows NT and is neccessary to support a > > correct semantics of wait syscalls, but it have unexpected behavior on > > Windows 95. > > No workaround found. > > - Some syscall semantics-related changes. > > Seems to work fine here (Win95 OSR2) > > Pete Jordan > = = = = = = = = = = = = = = = = = = = = = = = > Horus Communications > http://www.horus.cix.co.uk/ > = = = = = = = = = = = = = = = = = = = = = = = > "'Not twisted,' Salzy once said of her own > passion, 'it is helical. That sounds better.'" > - > For help on using this list (especially unsubscribing), send a message to > "gnu-win32-request@cygnus.com" with one line of text: "help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From khamsis@kmrmail.kmr.ll.mit.edu Wed Aug 13 19:01:00 1997 From: khamsis@kmrmail.kmr.ll.mit.edu (Sarir Khamsi) Date: Wed, 13 Aug 1997 19:01:00 -0000 Subject: Colors in bash Message-ID: <9708140204.AA03437@kmrmail.kmr.ll.mit.edu> Yo, Is there a way to get color happening in bash, like the output of ls having different colors for different types of files? I looked all through the FAQ and old messages but did not find anything. Thanks. Sarir -- Sarir "Riri" Khamsi Kwajalein, Marshall Islands email: khamsis@kmrmail.kmr.ll.mit.edu w: 805-355-1534 Time Zone: GMT +12 hours f: 805-355-3833 web: http://www.futureone.com/~sponge - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Ian@kiwiplan.co.nz Wed Aug 13 19:37:00 1997 From: Ian@kiwiplan.co.nz (Ian Collins) Date: Wed, 13 Aug 1997 19:37:00 -0000 Subject: FW: Spooler Message-ID: <07551A47EAD6D01186430060B025C6F40A6C68@modem1.kiwiplan2.co.nz> > Does anyone know of a ported lp sub-system for gnu-win32? > I can't seem to find any spooler related commands at all. > > Many Regards, > Ian Collins > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From suicide@freakz.org Wed Aug 13 22:58:00 1997 From: suicide@freakz.org (justin.) Date: Wed, 13 Aug 1997 22:58:00 -0000 Subject: Has anyone ported Sendmail 8.8.7 under CYGNUS. Message-ID: <33F29F11.D5EFA9DA@freakz.org> I run CYGWIN32 under NT 4.0 and i just CANT seem to port Sendmail... if anyone has... let me know.. suicide@freakz.org Proffesional Sofa Testing Engineer. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bridge@dial-up.net Wed Aug 13 23:57:00 1997 From: bridge@dial-up.net (Leonard Weincier) Date: Wed, 13 Aug 1997 23:57:00 -0000 Subject: SSH References: <009B8BC8.1913CDC0.21318@ifk20.mach.uni-karlsruhe.de> Message-ID: > : This seems to be corrected when I remove the -pipe option from gcc ? > > Could configure have been failed because of using -pipe, too? > Then you might try again after "rm config.cache", "make distclean" or > whatever this package supports/needs. With no -pipe option i get "cannot find sys/un.h" ?? File is not on the system. Len - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From john.nichol@il.us.swissbank.com Thu Aug 14 01:54:00 1997 From: john.nichol@il.us.swissbank.com (John Nichol) Date: Thu, 14 Aug 1997 01:54:00 -0000 Subject: OBJC Runtime for gnuwin32 References: <9708140204.AA03437@kmrmail.kmr.ll.mit.edu> Message-ID: <33F2C4F1.1873FE51@swissbank.com> I have pulled the b18 version of the gnuwin32 stuff down off the webpage. I cant find any objective-C runtime libraries, the gcc compiler does parse objc code. Is objc supported as part of this project or just C/C++? Can I get hold of objc runtime anywhere? If objc isn't supprted has anyone used gnuwin32 to build gcc-2.7.2.2 on NT? Any advice would be welcomed. Cheers John. -- John Nichol SBC Warburg, 2 Finsbury, London, EC2M 2PP Tel: +44 171 5680390 Work: john.nichol@swissbank.com http://w2155791/john_nichol/ Home: john.nichol@virgin.net http://freespace.virgin.net/john.nichol/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From arifi_koseoglu@karmaint.com Thu Aug 14 04:33:00 1997 From: arifi_koseoglu@karmaint.com (arifi_koseoglu@karmaint.com) Date: Thu, 14 Aug 1997 04:33:00 -0000 Subject: Emacs RMAIL - Lotus cc:Mail Message-ID: <9707148715.AA871575114@karmaint.com> Hi ! I wonder whether anyone knows of an Emacs extension which would let me use RMAIL to read/send cc:Mail e-mail messages ? Thanks in advance, Arifi Koseoglu arifi_koseoglu@karmaint.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gjones@websa.com Thu Aug 14 05:10:00 1997 From: gjones@websa.com (Glenn Jones) Date: Thu, 14 Aug 1997 05:10:00 -0000 Subject: SSH Message-ID: <199708141520.NAA05345@nickel.websa.co.za> I have also tried compiling SSH on a windows 95 machine. I removed the -pipe from the ./configure, but received the same results (0 for sizeof int, long int etc.) When I compiled without -pipe, there was an error in includes.h that I traced down to a missing .h file. includes.h required either streams.h an ptem.h or un.h. Neither ptem.h or un.h seem to be in the distribution that I have (b18). Does anyone know where I can get general guidelines for porting to gnu-win32? Thanks Glenn - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gunther.ebert@ixos-leipzig.de Thu Aug 14 05:10:00 1997 From: gunther.ebert@ixos-leipzig.de (Gunther Ebert) Date: Thu, 14 Aug 1997 05:10:00 -0000 Subject: editor which works with Sergey's telnetd Message-ID: <33F311A4.2593@ixos-leipzig.de> Hi, is there any text editor which works with telnetd? Gunther -- Gunther Ebert iXOS Anwendungs-Software GmbH Angerstrasse 40-42 D-04177 Leipzig Phone : +49 341 48503-0 Fax : +49 341 48503-99 E-mail: mailto:gunther.ebert@ixos-leipzig.de www : http://www.ixos-leipzig.de - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mh@mike.franken.de Thu Aug 14 05:25:00 1997 From: mh@mike.franken.de (Michael Hirmke) Date: Thu, 14 Aug 1997 05:25:00 -0000 Subject: telnetd whereabouts? References: <1.5.4.32.19970813225315.0067657c@nation.org> Message-ID: <6bpQz1o$pfB@mike.franken.de> Hi Derek, > Hi there, Can someone please let me know where the telnetd port can be found > in bin format? http://miracle.geol.msu.ru/sos/ > > thanks > > Derek Reynolds Bye. Michael. -- Michael Hirmke | Telefon +49 (911) 557999 Georg-Strobel-Strasse 81 | FAX +49 (911) 557664 90489 Nuernberg | E-Mail mh@mike.franken.de | WWW http://minimike.franken.de/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jdelanoy@lan.tjhsst.edu Thu Aug 14 06:41:00 1997 From: jdelanoy@lan.tjhsst.edu (Jay DeLanoy) Date: Thu, 14 Aug 1997 06:41:00 -0000 Subject: ircII? Anybody done it? (Also pico, pine, lynx) References: <199708140122.VAA15740@iag.net> Message-ID: <3.0.2.32.19970814075647.0090de60@lan.tjhsst.edu> At 09:23 PM 8/13/97 -0400, you wrote: > > Hi, I'm just curious if anyone has done a pico/pine/lynx or ircII port? I >have trouble with them, maybe someone else has had better luck (btw, I'm >looking for something that will run within a telnetd window.) > I tried ircII, but it works, only problemmatically. I'm looking for full >screen within a telnet session, (I don't like -d stuff). The ircII I wrote >has a weird problem: Whenever you enter a weird, command, the command bar >and the command scroll up!!! If you have a working one (or BitchX will do, >I suppose. ;), please let me know ASAP! > > SJ >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > go to http://www.itribe.com/virtunix . I know that pine, lynx, and ircII are all there. I'm not sure about pico. jay de lanoy, jdelanoy@lan.tjhsst.edu Republicans understand the importance of bondage between a mother and child. -- Senator Dan Quayle, US News and World Report (10/10/88) I'd rather be rich than stupid. I hope life isn't a big joke, because I don't get it. Instead of having "answers" on a math test, they should just call them "impressions," and if you got a different "impression," so what, can't we all be brothers? I wonder if Dracula ever has ticks. orange skablah? wiederaosskaboom! - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mh@mike.franken.de Thu Aug 14 06:41:00 1997 From: mh@mike.franken.de (Michael Hirmke) Date: Thu, 14 Aug 1997 06:41:00 -0000 Subject: Curses? References: Message-ID: <6bpQyXO4pfB@mike.franken.de> Hi Sakis, > Hi! I am trying out GNU-Win32 and up till now I found it really useful for > people needing unix useability under Windows. However, I wanted to try > compiling some programs that needed a form of 'curses' libs, like curses, > ncurses etc. When I tried to compile PDCURSES, the too-many constant > redefinitions seemed to 'swamp' the GCC compiler to a grinding halt with > errors. Have you ever encountered such a difficulty? Is there ANY > precompiled curses library available? ncurses - even the newest version 4.1 - compiles out of the box with Cygnus' gcc ! > > Thanks in advance! Bye. Michael. -- Michael Hirmke | Telefon +49 (911) 557999 Georg-Strobel-Strasse 81 | FAX +49 (911) 557664 90489 Nuernberg | E-Mail mh@mike.franken.de | WWW http://minimike.franken.de/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From majordomo@cygnus.com Thu Aug 14 06:41:00 1997 From: majordomo@cygnus.com (majordomo@cygnus.com) Date: Thu, 14 Aug 1997 06:41:00 -0000 Subject: Welcome to gnu-win32 Message-ID: <199708141316.GAA06525@cygnus.com> -- Welcome to the gnu-win32 mailing list! If you ever want to remove yourself from this mailing list, you can send mail to "majordomo@cygnus.com" with the following command in the body of your email message: unsubscribe gnu-win32 gsstark@mit.edu gnu-win32 Here's the general information for the list you've subscribed to, in case you don't already have it: ************************************************************************ * * * T H E G N U - W I N 3 2 M A I L I N G L I S T * * * ************************************************************************ The Cygwin32 API allows easy ports of many Unix applications to Win32. Using this library, we have ported many of the popular GNU tools to Win32. These are the "gnu-win32" tools. This list is a place to discuss the gnu-win32 tools and the Cygwin32 API. Announcements pertaining to the project are sent to the list "gnu-win32-announce", which you can subscribe to by sending mail to majordomo@cygnus.com with the body: "subscribe gnu-win32-announce". For more information, please visit our WWW site (listed below). This list uses Majordomo. Please send all admin requests to gnu-win32-request@cygnus.com. Things you should know: The gnu-win32 WWW Home page http://www.cygnus.com/misc/gnu-win32/ The gnu-win32 ftp location ftp://ftp.cygnus.com/pub/gnu-win32/ The gnu-win32 mailing list archives http://www.cygnus.com/ml/gnu-win32/ The main Cygnus home page http://www.cygnus.com/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mh@mike.franken.de Thu Aug 14 06:41:00 1997 From: mh@mike.franken.de (Michael Hirmke) Date: Thu, 14 Aug 1997 06:41:00 -0000 Subject: Win95 Compiler References: <2.2.32.19970813180442.00673284@mail.gower.net> Message-ID: <6bpQyvP4pfB@mike.franken.de> Hi James, > I have Windows 95, and I'm not very knowledgeable about compilers and the > like, and I need to compile a C file on my computer. I was wondering if what > was on your Cygnus page had what I needed, and I downloaded the > usertools.exe (4.4 megs) and I'm not sure if this is what I needed. I notice > the developers kit needed a compiler to compile it, and I thought that was > odd since I don't have a compiler to compile it. Please help me. You have to get cdk.exe instead of usertools.exe. cdk.exe includes gcc as compiler and all necessary tools for compiling. > > Thanks, :) > > James A. Leath Bye. Michael. -- Michael Hirmke | Telefon +49 (911) 557999 Georg-Strobel-Strasse 81 | FAX +49 (911) 557664 90489 Nuernberg | E-Mail mh@mike.franken.de | WWW http://minimike.franken.de/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From andyp@parallax.co.uk Thu Aug 14 06:41:00 1997 From: andyp@parallax.co.uk (Andy Piper) Date: Thu, 14 Aug 1997 06:41:00 -0000 Subject: SSH References: <009B8BC8.1913CDC0.21318@ifk20.mach.uni-karlsruhe.de> Message-ID: <3.0.1.32.19970814144106.0096aea0@mailhost> At 08:57 14/08/97 +0200, Leonard Weincier wrote: >> : This seems to be corrected when I remove the -pipe option from gcc ? >> >> Could configure have been failed because of using -pipe, too? >> Then you might try again after "rm config.cache", "make distclean" or >> whatever this package supports/needs. > >With no -pipe option i get "cannot find sys/un.h" ?? File is not on the system. I have exactly the same problem. andy ___ ____ Dr Andy Piper / _ \___ ________ _/ / Solutions_ (require 'disclaimer) / ___/ _ `/ __/ _ `/ / / _ `/\ \ / andyp@parallax.co.uk /_/ \_,_/_/ \_,_/_/_/\_,_//_\_\ boot /vmemacs - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From acherman@inf.ethz.ch Thu Aug 14 06:41:00 1997 From: acherman@inf.ethz.ch (Erwin Achermann) Date: Thu, 14 Aug 1997 06:41:00 -0000 Subject: What Terminal are there? Message-ID: <33F2FC54.2F27A0EA@inf.ethz.ch> Howdy, forgive me if this is too lame a newby question. When starting a bash it opens a dos-Terminal with no scroll bar. Well that's not a big problem, i just need 'less'. Ok, there i went, compiled termcap and 'less' it all went fairly well, I think i slowly am back to my beloved Unix-Environment :) But heck, my Dos-Terminal doesn't let me enter the '|' , nor the []s, nor the '@' nor the '#' nor the '~' !!!! They all are entered by using the Alt-Gr Key, so you see I am sitting at a swiss-german Keyboard with these Symbols at very unpleasant places. It is very cumbersome with Cut'n'Paste especially when my Terminal doesn't seem to constantly react on Paste.... sigh! Can someone of you kind souls out there tell me, what decent Terminal SW for NT4 there are and where to get them? Thank for all pointers Erwin -- |Erwin Achermann phone: +41-1-632 74 40 | |Institute of Scientific Computing fax: +41-1-632 11 72 | |ETH Zurich, IFW C29.2 mailto:acherman@inf.ethz.ch | |8092 Zurich, Switzerland http://www.inf.ethz.ch/personal/~acherman | - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pfitz@pratique.fr Thu Aug 14 08:07:00 1997 From: pfitz@pratique.fr (Tony FITZPATRICK) Date: Thu, 14 Aug 1997 08:07:00 -0000 Subject: sh crashes on 95 References: <199708131851.TAA30622@muzak.trintech.ie> Message-ID: <3.0.2.32.19970814160456.007b0590@mail.pratique.fr> Sonic Junior wrote: >> >> Okay... when I'm doing make, sh has a habit of crashing. It's got bad >>recently. Where can I get da patch? Ack! >> >> SJ >> At 19:49 13/08/97 +0100, you replied >Yip I've run into the same problem. Running configure scripts it seems to >choke at no particular point some times it get further than others. If I >reboot under NT4 configure completes to the end. >Looking at the processes while a configure is in progress I have seen up to >six instances of sh running, I haven't seen more as it normally dies and more >often than not needs the three fingered salute! > >Anyone got any ideas?? > I posted this problem to the list some time ago and having received no replies thought it was specific to me, obviously it isn't. All of Sergeys cygwin.dlls exhibit this problem with "configure" on W95. By a process of elimination I have traced it the modification of "fork.cc" in the very first set of changes posted by Sergey (called "itimer.tar.gz" on his web page). I can include all the other modifs. from this set and the regenerated cygwin.dll executes "configure" faultlessly. If I then regenerate with the new "fork.cc" I get the problem with "sh" crashing the system. That's as far as I've got for the moment. The original patch to fork (see below) is considerably updated in subsequent patches from Sergey but none of them seem to solve this particular problem. Please feel free to comment if you have any ideas but don't expect a rapid reply as I leave on two weeks vacation tommorow :^) Tony (W95 SP1 P100 32Mb) diff -crN winsup.orig/fork.cc winsup/fork.cc *** winsup.orig/fork.cc Fri Apr 25 03:45:37 1997 --- winsup/fork.cc Thu May 15 18:49:29 1997 *************** *** 278,283 **** --- 278,284 ---- child->dwProcessId = pi.dwProcessId; child->uid = u->self->uid; child->gid = u->self->gid; + child->umask = u->self->umask; child->sigs = u->self->sigs; child->sig_mask = u->self->sig_mask; set_child_mmap_ptr (child); *************** *** 395,400 **** --- 396,405 ---- debug_printf ("fork_helper: child is running\n"); u->self = s->p[x]; + + /* Initialize hidden window */ + window_init(); + reent_data = u->self->reent_save; debug_printf ("fork child: self %p, pid %d, ppid %d\n", u->self, x, u->self->ppid); - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From papresco@technologist.com Thu Aug 14 08:07:00 1997 From: papresco@technologist.com (Paul Prescod) Date: Thu, 14 Aug 1997 08:07:00 -0000 Subject: SSH References: <009B8BC8.1913CDC0.21318@ifk20.mach.uni-karlsruhe.de> Message-ID: <33F31F5A.534DE68B@technologist.com> dahms@ifk20.mach.uni-karlsruhe.de wrote: > Could configure have been failed because of using -pipe, too? > Then you might try again after "rm config.cache", "make distclean" or > whatever this package supports/needs. Okay, has ANYONE got ssh to build on Windows 95? The configure totally messes up and everything goes downhill from there. "un.h" is easy to find on the Web, but with SIZEOF( int ) == 0 no compile is going to result in a useful binary. Paul Prescod - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From stanton@haas.berkeley.edu Thu Aug 14 19:42:00 1997 From: stanton@haas.berkeley.edu (Richard Stanton) Date: Thu, 14 Aug 1997 19:42:00 -0000 Subject: How to avoid GNU make case sensitivity? Message-ID: <199708142220.PAA11975@haas.berkeley.edu> I have lots of C files on a FAT (DOS style) formatted disk. Now I'm running NT, and when I try to use GNU make to compile stuff, it doesn't recognize the files' existence, since NT thinks all the filenames are upper case. For example, if I type make test.o I get an error message about not having a rule to make test.o, when a directory listing shows the file TEST.C existing on the drive. Renaming it manually to test.c allows me to compile it OK, but I'd rather avoid manually renaming every .C file on my system one at a time. Is there a way to tell GNU make to ignore case? Richard Stanton - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From paul@hydra.parasoft.com Thu Aug 14 19:42:00 1997 From: paul@hydra.parasoft.com (Paul Pluzhnikov) Date: Thu, 14 Aug 1997 19:42:00 -0000 Subject: missing GNUMake-3.75/Win32 pieces? Message-ID: <199708150242.TAA27457@elephant.parasoft.com> Hi, It appears that someone has ported the archive stuff, e.g. libfoo.lib(foo.obj) rules to Win32. It also appears that the source for this port is not in the standard distribution. Would you by any chance know where I can find that source? Thanks, -- Paul Pluzhnikov paul@parasoft.com 626/305-0041 x373 In theory there is no difference between practice and theory. But in practice there is. - Conference on Symbolic Logic. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From cgf@bbc.com Thu Aug 14 19:42:00 1997 From: cgf@bbc.com (Chris Faylor) Date: Thu, 14 Aug 1997 19:42:00 -0000 Subject: editor which works with Sergey's telnetd References: <33F311A4.2593@ixos-leipzig.de> Message-ID: In article < 33F311A4.2593@ixos-leipzig.de >, Gunther Ebert wrote: >is there any text editor which works with >telnetd? Any editor that you can compile using the cygwin tools should work correctly. I have a binary copy of vim on my web site that works with Sergey's telnetd: http://www.tiac.net/users/cgf -- http://www.bbc.com/ cgf@bbc.com "Strange how unreal VMS=>UNIX Solutions Boston Business Computing the real can be." - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From koi@ssa.crane.navy.mil Thu Aug 14 19:42:00 1997 From: koi@ssa.crane.navy.mil (Mark Koi) Date: Thu, 14 Aug 1997 19:42:00 -0000 Subject: Timers and less Message-ID: I just installed the binaries. I was using DJGPP before with ALLEGRO for timers. Those allegro library calls don't work, and can't seem to compile allegro with gnu-win32. DOES SOMEBODY HAVE A SIMPLE TIMER PROGRAM? NEED RESOLUTION OF 100ms, not a big request but a little lost. Also, can somebody mail me a uuencoded copy of less.exe for gnu-win32. Or is it just as easy to get source and compile. koi@ssa.crane.navy.mil - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Thu Aug 14 19:42:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Thu, 14 Aug 1997 19:42:00 -0000 Subject: editor which works with Sergey's telnetd Message-ID: <01BCA909.A6D5A3B0@sos> Gunther Ebert wrote: > is there any text editor which works with > telnetd? I use joe editor or MC 4.0 built-in editor. Vim works fine also, but I can't understand the logic of vi's key bindings :-) -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mskow@earthling.net Thu Aug 14 19:42:00 1997 From: mskow@earthling.net (Michael Skowronski) Date: Thu, 14 Aug 1997 19:42:00 -0000 Subject: errors while compiling Message-ID: i just downloaded b18 of the gnuwin32 tools and i get the following error while compiling simple hello world test: (unknown) error while dumping state (probably corrupt stack) in cygwin_except_handler Does anyone know what this means and how to fix it? Mike - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gaylano@aisd.org Thu Aug 14 19:42:00 1997 From: gaylano@aisd.org (Gaylan Ohlhausen) Date: Thu, 14 Aug 1997 19:42:00 -0000 Subject: Sergey, environment, elvis Message-ID: <33F35B19.52B0BF8E@aisd.org> Where can I find "Sergey's" stuff. Also, I have had problems installing programs under b18 on NT4 sp3. What NT environment settings and/or bash environment settings should I use to get xpm and fvwm-2 to compile? Also elvis 2.0 won't compile. It has problems compiling tcaphelp.c. Thanks - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mkraft@writeme.com Thu Aug 14 19:42:00 1997 From: mkraft@writeme.com (Michael Kraft) Date: Thu, 14 Aug 1997 19:42:00 -0000 Subject: X11 to GDI/DirectX? Message-ID: <33F33893.86BFE56F@writeme.com> I was wondering if someone made or is working on a lib that will allow X11 programs to be compilled so they don't need an X11 client program to run? I know another program "NuTCRACKER" by DataFocus does this (it also ports Unix programs to Win32 like Gnu-win32 does) but I don't think I'd be able to afford it since they don't even list the price. If not do you know of someplace that lists all the GUI API's for X11? -- Michael Kraft (mkraft@writeme.com) http://www.geocities.com/Hollywood/Hills/4984/index.html - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From hassan@cp10.es.xerox.com Thu Aug 14 19:42:00 1997 From: hassan@cp10.es.xerox.com (Hassan Blurfrushan) Date: Thu, 14 Aug 1997 19:42:00 -0000 Subject: problems with bash (fork, Permission denied erros) Message-ID: <199708150040.RAA01616@beluga> I have noticed 2 problems when using bash: The first problem happens on both beta18 and beta17.1 versions. When running shell scripts on my NT 4.0 workstation, intermittently I get a "Permission denied" message like this: /usr/bin/grep: Permission denied /usr/bin/tr: Permission denied This always happens when muliple pipes are used. e.g. ps | grep -v grep | grep abc | awk '{print $NF}' The second problem happens on beta18 only (works fine on beta17.1). If there are any long if statements like the following: if [ $a = $b ]; then (about 2000 lines of shell stuff here) fi Then bash fails with the following error: fork_helper: child died before initialization with win32 error 128 fork_helper: child died before initialization with win32 error 128 fork: No more processes H. Blurfrushan - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bzinger@iname.com Thu Aug 14 19:42:00 1997 From: bzinger@iname.com (Micheal A. Benzinger) Date: Thu, 14 Aug 1997 19:42:00 -0000 Subject: Colors in bash References: <9708140204.AA03437@kmrmail.kmr.ll.mit.edu> Message-ID: <3.0.3.32.19970814190808.0091add0@pop.flash.net> Sarir, You have two choices: ls --color dir --color For more information on these two commands, use the following: ls --help dir --help Mike Benzinger At 02:04 PM 8/14/97, Sarir Khamsi wrote: >Yo, > >Is there a way to get color happening in bash, like the output of ls >having different colors for different types of files? I looked all >through the FAQ and old messages but did not find anything. Thanks. > >Sarir > >-- >Sarir "Riri" Khamsi Kwajalein, Marshall Islands - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jtbowman@theshop.net Thu Aug 14 19:42:00 1997 From: jtbowman@theshop.net (Jeff Bowman) Date: Thu, 14 Aug 1997 19:42:00 -0000 Subject: Install of B18 does not work. Message-ID: <33F3B0EE.100816DC@theshop.net> Hi, I am having a bit of difficulty getting usertools.exe from the /latest directory to install on my machine. I have twice downloaded the file and tried multiple times to install, all to no avail. Here are the facts: 1) I use RedHat Linux (kernel 2.0.30) at home. This is what I have been using to ftp the usertools.exe file. 2) I use tar on my Linux machine to put the file onto 3 diskettes. 3) I use Windows NT 4.0 FixPack 3.0 at work. (Intel Pentium 166, 32Mb Ram, 1.6Gb HD with almost 600Mb free space. NTFS Filesystem) 4) I am using djtar (tar utility distributed with djgpp compiler since I can't find one for WinNT) to untar the file from the diskettes. -- To this point everything seems to be going ok -- 5) I run usertools.exe. The Install Shield dialog box appears telling me that it is copying the install files I need. 6) The progress bar reaches approximately 43% then stops and displays an error. 7) The error text is: "Failed to uncompress %s. There may not be enough free space in the TEMP directory" ^^^^^^^^^^^^^^^^^^^^ NOTE: I believe this to be the exact wording, however, the actual message my say "copy") As I noted in fact 3, I have approximately 600Mb of free space on my drive. To my knowledge, there isn't a way to allocate space to a specific directory. Even if your file is compressed 90% (ie is 10% of it's original size) the largest it should become is approximately 40Mb. (The download is approximately 4Mb). So I should have enough room (I think) to continue the install. I have read the read me concerning the installation of the tools. Is there anything else I can do or try either in configuration or use a different file (ie not B18) to install this?? Or is there a way I can get the files outside of your SFX file??? (ie, is there a way I can manually install this, not using your installation program???) Thanks for your help, I am looking forward to using these tools. Jeff Bowman jtbowman@theshop.net wolfjb@hotmail.com -- wolfjb@hotmail.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From broche@bensonlaw.com Thu Aug 14 19:42:00 1997 From: broche@bensonlaw.com (Bryon Roche) Date: Thu, 14 Aug 1997 19:42:00 -0000 Subject: Colors in bash References: <9708140204.AA03437@kmrmail.kmr.ll.mit.edu> Message-ID: <33F37BB9.756CD532@bensonlaw.com> Sarir Khamsi wrote: > > Yo, > > Is there a way to get color happening in bash, like the output of ls > having different colors for different types of files? I looked all > through the FAQ and old messages but did not find anything. Thanks. You can go and find the color-ls patch from the slackware distribution. I had a hack that worked with djgpp, but I don't anymore. (at least the last time I checked.) -- /---------------------------\ |Bryon Roche | | mailto:broche@bensonlaw.com | \---------------------------/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From marcusd@cathcart.sysc.pdx.edu Fri Aug 15 04:25:00 1997 From: marcusd@cathcart.sysc.pdx.edu (Marcus G. Daniels) Date: Fri, 15 Aug 1997 04:25:00 -0000 Subject: VxD drivers Message-ID: Has anyone had experience using the gnu-win32 to build device drivers for Windows 95? What would be the minimum set of resources needed to do this? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Fri Aug 15 05:04:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Fri, 15 Aug 1997 05:04:00 -0000 Subject: X11 to GDI/DirectX? Message-ID: <01BCA98A.2004FCB0@gater.krystalbank.msk.ru> Michael Kraft wrote: > I was wondering if someone made or is working on a lib that will allow X11 > programs to be compilled so they don't need an X11 client program to run? I Looks like I have some related ideas... -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kunglao@prairienet.org Fri Aug 15 05:04:00 1997 From: kunglao@prairienet.org (Gilligan) Date: Fri, 15 Aug 1997 05:04:00 -0000 Subject: editor which works with Sergey's telnetd References: <33F311A4.2593@ixos-leipzig.de> Message-ID: <33F4E182.6C989113@prairienet.org> Gunther Ebert wrote: > > Hi, > > is there any text editor which works with > telnetd? > > Gunther Jed seems to work fine. Vim had some complaint about the term, then only ran in a dumb mode. Jed is even doing colors with C programs through telnetd (running Netterm to telnet to my Linux box, then telnet back to Win95). It compiled fairly straightforward. I had to tweak some of the .sl files, I recall (but it has been a while since then). If anyone wants a binary distribution of it, I'll see what I can do. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From gunther.ebert@ixos-leipzig.de Fri Aug 15 05:04:00 1997 From: gunther.ebert@ixos-leipzig.de (Gunther Ebert) Date: Fri, 15 Aug 1997 05:04:00 -0000 Subject: editor which works with Sergey's telnetd References: <33F311A4.2593@ixos-leipzig.de> Message-ID: <33F461BE.ABB@ixos-leipzig.de> Chris Faylor wrote: > Any editor that you can compile using the cygwin tools should work correctly. > > I have a binary copy of vim on my web site that works with Sergey's telnetd: > > http://www.tiac.net/users/cgf Thanks, it works fine. Gunther -- Gunther Ebert iXOS Anwendungs-Software GmbH Angerstrasse 40-42 D-04177 Leipzig Phone : +49 341 48503-0 Fax : +49 341 48503-99 E-mail: mailto:gunther.ebert@ixos-leipzig.de www : http://www.ixos-leipzig.de - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Keetnet@wilmington.net Fri Aug 15 05:04:00 1997 From: Keetnet@wilmington.net (Keet / Foxbird) Date: Fri, 15 Aug 1997 05:04:00 -0000 Subject: DIR From Within BASH... Message-ID: <3.0.32.19970815065804.00694970@wilmington.net> On a side note, which I'm not sure is a bug, feature, new way of doing things, or a screw-up by microsoft, i've noticed that DIR from within BASH does not handle files or directories with spaces in them all that well. It reports one such (for example 'New Folder') to be: New\ Folder LS seems to work quite well and handles them just fine, just DIR causes the strange printing, any ideas? - Greg N. keetnet@wilmington.net - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mchase@ix.netcom.com Fri Aug 15 12:35:00 1997 From: mchase@ix.netcom.com (Michael A. Chase) Date: Fri, 15 Aug 1997 12:35:00 -0000 Subject: BUG: Pipes interfere with terminal output under coolview 970810 Message-ID: <199708151252.HAA19155@dfw-ix10.ix.netcom.com> When I use a pipe on the command line, the results from the final command do not appear on the display. I can send the output to a file. I'm using WinNt 4.0 SP3, gnuwin32 b18, and coolview dated 970810. The problem goes away when I reinstall the cgywin.dll dated 970731. Thanks, Mac :}) mchase@ix.netcom.com -----BEGIN GEEK CODE BLOCK----- Version: 3.1 GCS/CC/E/IT/O d s+:+ a+ C++$ ULUHSC*++$ P++$ L+(++) E- W++ N++ o? K-? w+ O M- V- PS+@ PE Y+ PGP+ t+ 5++ X R tv b+++(++++) DI++(+++) D? G e++ h--- r+++ y+ ------END GEEK CODE BLOCK------ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bridge@dial-up.net Fri Aug 15 12:35:00 1997 From: bridge@dial-up.net (Leonard Weincier) Date: Fri, 15 Aug 1997 12:35:00 -0000 Subject: SSH References: <33F31F5A.534DE68B@technologist.com> Message-ID: > Okay, has ANYONE got ssh to build on Windows 95? The configure totally I don't think anyone has, it failed with me. I have also failed under NT Len - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From stanton@haas.berkeley.edu Fri Aug 15 14:20:00 1997 From: stanton@haas.berkeley.edu (Richard Stanton) Date: Fri, 15 Aug 1997 14:20:00 -0000 Subject: Problem compiling/linking large number of object files. Message-ID: <199708152119.OAA13463@haas.berkeley.edu> The problem actually seems to be related to using the emx C compiler under NT, rather than being related to GNU make. It's thus not exactly 100% relevant to this list, for which I apologise, but maybe someone out there has an idea what's wrong: At a command prompt, I tried the command gcc -o np2f.exe -Le:/lib np2f.o parse.o matrix.o nonpar.o nrutil.o matstat.o constant.o sort.o locate.o quantile.o powi.o indsort.o brent.o dfpmin.o gamma.o linmin.o gradcd.o mnbrak.o ranlib.o intutils.o ols.o chol.o cholbksb.o invpd.o t.o f.o beta.o gammln.o mthutl.o -lm Nothing happened - I quickly got a command prompt back, but no .exe file was created. The problem seems to be something to do with the number of files, or length of the command line, as shorter lists of .o files work fine. Any idea what's wrong? Thanks. Richard Stanton - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kunglao@prairienet.org Fri Aug 15 15:12:00 1997 From: kunglao@prairienet.org (Gilligan) Date: Fri, 15 Aug 1997 15:12:00 -0000 Subject: Colors in bash References: <9708140204.AA03437@kmrmail.kmr.ll.mit.edu> <33F37BB9.756CD532@bensonlaw.com> Message-ID: <33F4E22C.1CB71C1E@prairienet.org> Bryon Roche wrote: > > Sarir Khamsi wrote: > > > > Yo, > > > > Is there a way to get color happening in bash, like the output of ls > > having different colors for different types of files? I looked all > > through the FAQ and old messages but did not find anything. Thanks. > You can go and find the color-ls patch from the slackware > distribution. > I had a hack that worked with > djgpp, but I don't anymore. (at least the last time I checked.) The stock b18 ls allows color, you just have to feed it the --color option and add the output of the dircolors command to your .bashrc to setup the LS_COLORS env var. -- Due to the abundance of spamming, action has been taken. Please remove the underscore (_) from my address when replying to me. Spammers, please leave the underscore in. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From syring@pollux.edv.agrar.tu-muenchen.de Fri Aug 15 15:12:00 1997 From: syring@pollux.edv.agrar.tu-muenchen.de (Karl Martin Syring) Date: Fri, 15 Aug 1997 15:12:00 -0000 Subject: AW: What Terminal are there? Message-ID: <01BCA9D8.FB78CF80@syring@pollux.edv.agrar.tu-muenchen.de> You need Sergey's cygwin.dll (and may be bash, too) from http://www.lexa.ru/sos Regards Karl But heck, my Dos-Terminal doesn't let me enter the '|' , nor the []s, nor the '@' nor the '#' nor the '~' !!!! They all are entered by using the Alt-Gr Key, so you see I am sitting at a swiss-german Keyboard with these Symbols at very unpleasant places. It is very cumbersome with Cut'n'Paste especially when my Terminal doesn't seem to constantly react on Paste.... sigh! - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mchase@ix.netcom.com Fri Aug 15 15:12:00 1997 From: mchase@ix.netcom.com (Michael A. Chase) Date: Fri, 15 Aug 1997 15:12:00 -0000 Subject: Undeliverable Mail Message-ID: <199708151252.HAA19158@dfw-ix10.ix.netcom.com> Sergey, Several messages I tried send directly to you over the last few days have been returned with responses like the one after my signature. Was there a temporary problem at your mail host, or am I trying the wrong address? Over the same period, you appear to have responded to messages in the gnu-win32 mail list, so I'm trying to email you through there. Mac :}) mchase@ix.netcom.com -----BEGIN GEEK CODE BLOCK----- Version: 3.1 GCS/CC/E/IT/O d s+:+ a+ C++$ ULUHSC*++$ P++$ L+(++) E- W++ N++ o? K-? w+ O M- V- PS+@ PE Y+ PGP+ t+ 5++ X R tv b+++(++++) DI++(+++) D? G e++ h--- r+++ y+ ------END GEEK CODE BLOCK------ The original message was received at Thu, 7 Aug 1997 08:30:17 -0500 (CDT) from smap@localhost ----- The following addresses had permanent fatal errors ----- ----- Transcript of session follows ----- ... Deferred: Connection timed out with ns.co.ru. Message could not be delivered for 3 days Message will be deleted from queue - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kunglao@prairienet.org Fri Aug 15 15:12:00 1997 From: kunglao@prairienet.org (Gilligan) Date: Fri, 15 Aug 1997 15:12:00 -0000 Subject: login and telnetd References: <9708060928.AA38659@hagar.aspentec.com> <33F16FCC.4EF7@ixos-leipzig.de> Message-ID: <33F4DDD6.E9F1D953@prairienet.org> Gunther Ebert wrote: > > Jerome Gay wrote: > > > > telnetd : > > when running 'inetd -d /etc/inetd.conf' in one bash session, and > running > > telnet (Sergey's telnet.exe) in another bash session, trying to > connect on > > localhost, I got : > > > > telnet client side : > > Trying 127.0.0.1... > > setuid called, and it's not done yet. > > Connected to pathfinder.[...] > > Escape character is '^]'. > > Connection closed by foreign host. > > > > in bash session where inetd was started : > > ... > > someone wants telnet > > accept, ctrl 3 > > + Closing from 17 > > 1022 execl /usr/local/libexec/in.telnetd > > 1022 reaped, status 0x100 > > > > Is anyone using telnetd successfully ? > > Any idea ? > > > > Did anyone answer this question? I'm running exactly into this > problem. I had a similar problem, it turned out to be just that I didn't have root defined in my /etc/passwd. I changed the inetd.conf telnet entry from root to my username, and it worked. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From menze@spud.dynatekusa.com Fri Aug 15 19:23:00 1997 From: menze@spud.dynatekusa.com (Ed Menze) Date: Fri, 15 Aug 1997 19:23:00 -0000 Subject: case sensitivity in pwd program Message-ID: Hello, I'm running the gnu-win32 beta 18 release on NT 4.0, and I have a question regarding the pwd program (rather than the bash builtin) and its interaction with filename globbing. Is there a way to cause pwd to display the case of directory components in a way that is consistent with file globbing? For example: bash$ mkdir -p a/B/c bash$ cd a/b/C bash$ d:/local-nt/cygnus/H-i386-cygwin32/bin/pwd //d/menze/tmp/a/b/C The result of the pwd program displays the case of the path components as they were given to 'cd', not as they were created bash$ cd ../../.. bash$ echo */b/C */b/C Using the path components as reported by pwd, globbing fails bash$ echo */B/c a/B/c I have to use the same case with which the filenames were created in order to get the right expansion. I need to be able to get the current directory with components that have the proper case for later use in file globbing, as in the example above. Any ideas? Thanks, Ed Menze Dynatek Automation Systems menze@dynatekusa.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jeffdbREMOVETHIS@netzone.com Fri Aug 15 19:23:00 1997 From: jeffdbREMOVETHIS@netzone.com (Mikey) Date: Fri, 15 Aug 1997 19:23:00 -0000 Subject: 100ms timer function. Message-ID: <33f60a34.2663029@smtp.netzone.com> I remember the question, but not who posted it, this looks like it could be an answer, if not THE answer ;^). See the example for RDTSC, on http://www.rt66.com/~brennan/djgpp/djgpp_asm.html (jeffdbREMOVETHIS@netzone.com) delete REMOVETHIS from the above to reply Mikey - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From flips@eecs.umich.edu Fri Aug 15 19:23:00 1997 From: flips@eecs.umich.edu (Ryan de Jonckheere) Date: Fri, 15 Aug 1997 19:23:00 -0000 Subject: bash hangs on startup Message-ID: I actually have two problems. The first is bash 2.01 seems to hang when it starts up, or at least wait a really long time with the latest cygwin.dll from the coolview package (and what a great package that is!). If I replace cygwin.dll with one from mid july (the first one where remote telnetd would run) bash runs fine without pausing. I checked my /etc/passwd and /etc/group files (no CR/LF, just LF, fields appear fine) I tried setting my $HOME variable int the system control panel, unsetting it, but all to no avail... Finally I give up, it is more of an annoyance really, because after a while bash does start running, but then every 5 commands or so it pauses for a lng time again. I've actually started using Cmd.exe (gasp!) because of this. My other problem is with telnetd. I managed to set it up correctly on one computer, and it works fine there, but on the machine I spoke of above, I always get "Login Incorrect" message. I checked the User Rights, the user inetd is running under is a domain account, and is the same account on both computers. Each machine is windows nt 4.0 sp3. The problematic computer is server and the working one is workstation. If any one can help, I will be forever indebted to him/her. Ryan - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From venkman1@ix.netcom.com Fri Aug 15 19:29:00 1997 From: venkman1@ix.netcom.com (Narayan Natarajan) Date: Fri, 15 Aug 1997 19:29:00 -0000 Subject: Running winNT/95 programs... Message-ID: <3.0.1.32.19970815214435.00910ca0@popd.ix.netcom.com> Hi, I read a previous thread on running wNT/95 programs through an intermediate program. I am just a little confused. Why go through all the trouble when there is a "start" command in both w95 and NT? I have had no problems running any windows stuff through "start". For example: my alias for start on winnt: alias start='cmd /c start ' and on win95 there is an actual executable start.exe, so no alias reqd. once set the following works just fine. start --- starts associated program in windows. start --- works so long as path is visible I also have vi (elvis) aliased as: alias vi="start vi" so that it comes up in a separate window. Narayan | Narayan _/~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~/ |Natarajan _/ venkman1@ix.netcom.com http://www.netcom.com/~venkman1 _/ | _/ (810)-475-0667 . .____ _ _/ | _/ "The heavens call to you, and * . x -(____/ | _ / | _/ circle about you, displaying to ________/____/ \ ( )----- | _/ you their eternal splendors, and ,/ _ ( ------- | _/ your eye gazes only to earth." <________\ // (_)----- | _/ -- Dante --<== * |_____/ _ |/ -(_____) (_)----- |~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From marcus@bighorn.dr.lucent.com Fri Aug 15 21:41:00 1997 From: marcus@bighorn.dr.lucent.com (marcus@bighorn.dr.lucent.com) Date: Fri, 15 Aug 1997 21:41:00 -0000 Subject: bfd, ld, and dlltool patches Message-ID: <199708151502.JAA18365@chorus.dr.lucent.com> OK, well, several people have asked about this.. I had been trying to get through the layers of red tape around here and get the GNU copyright assignment straight, but perhaps it is best to send this along anyhow. I think that all of the bfd and ld patches are short and simple enough to not cause any copyright concerns, but the dlltool changes are a little more substantial (although still not anything much worth all the hassle about, but legal is legal, I guess).. Anyhow, with these changes I think that I can link .lib files with gcc produced binaries except that some .lib modules have segments with communal data flags that don't seem to be interpreted by ld the same way that Microsoft must be intending them. More work needs to be done here... The patched dlltool produces dlls that appear to be to be very similar to the Microsoft dlls, except that Microsoft dlls have all internal filenames the same (the name of the dll itself) and Microsoft LINK seems to depend on it. Unfortunately, bfd makes it difficult to produce an archive with internal files named the same, so I have to figure out something here (setting the in-memory flag, perhaps?) At any rate, I don't believe that this solves all of the inter-operability problems, but it does solve some, I think. In particular, the dlltool change does remove the need for the fixup.c kludge to insert the import table terminator, and it does it the way that the Microsoft .lib files do it, so I think that that is a win, at least. So, for what it's worth, here are my patches. Lucent, being as paranoid as it is, doesn't have any method for me to make them available as a URL, so I'll post them to the list once (I know they're large). If someone feels this is too large for the list, I'm sorry, but several people have asked for them and there may be others out there as well, so I think that this is appropriate. I would suggest looking at all of the changes and make sure that they make sense to you before applying them, since I was learning how things worked as I went. Certainly if somebody sees a bug or a better way to do something, I'd really appreciate the insight! marcus hall ----- *** bfd/orig/coffcode.h Wed May 7 04:44:57 1997 --- bfd/coffcode.h Wed Jul 16 08:59:48 1997 *************** *** 390,396 } else if (!strncmp (sec_name, ".stab", 5)) { ! styp_flags = STYP_INFO; } #ifdef COFF_WITH_PE else if (!strcmp (sec_name, ".edata")) --- 390,396 ----- } else if (!strncmp (sec_name, ".stab", 5)) { ! styp_flags = STYP_INFO | STYP_NOLOAD; } #ifdef COFF_WITH_PE else if (!strcmp (sec_name, ".edata")) *************** *** 541,547 || strncmp (name, ".stab", 5) == 0) { #ifdef COFF_PAGE_SIZE ! sec_flags |= SEC_DEBUGGING; #endif } #ifdef _LIB --- 541,547 ----- || strncmp (name, ".stab", 5) == 0) { #ifdef COFF_PAGE_SIZE ! sec_flags |= SEC_DEBUGGING | SEC_NEVER_LOAD; #endif } #ifdef _LIB *************** *** 2610,2615 #endif #ifdef COFF_IMAGE_WITH_PE /* suppress output of the sections if they are null. ld includes the bss and data sections even if there is no size assigned --- 2610,2635 ----- #endif + #ifdef COFF_WITH_PE + /* move alignment into s_flags */ + switch (current->alignment_power) + { + case 1: section.s_flags |= IMAGE_SCN_ALIGN_1BYTES; + break; + case 2: section.s_flags |= IMAGE_SCN_ALIGN_2BYTES; + break; + case 3: section.s_flags |= IMAGE_SCN_ALIGN_4BYTES; + break; + case 4: section.s_flags |= IMAGE_SCN_ALIGN_8BYTES; + break; + case 5: section.s_flags |= IMAGE_SCN_ALIGN_16BYTES; + break; + case 6: section.s_flags |= IMAGE_SCN_ALIGN_32BYTES; + break; + case 7: section.s_flags |= IMAGE_SCN_ALIGN_64BYTES; + break; + } + #ifdef COFF_IMAGE_WITH_PE /* suppress output of the sections if they are null. ld includes the bss and data sections even if there is no size assigned *************** *** 2619,2624 internal_f.f_nscns--; else #endif { SCNHDR buff; if (coff_swap_scnhdr_out (abfd, §ion, &buff) == 0 --- 2639,2645 ----- internal_f.f_nscns--; else #endif + #endif { SCNHDR buff; if (coff_swap_scnhdr_out (abfd, §ion, &buff) == 0 *** bfd/orig/coffgen.c Tue Apr 15 02:21:58 1997 --- bfd/coffgen.c Tue Jun 24 10:23:17 1997 *************** *** 911,916 if (native->u.syment.n_sclass == C_FILE) symbol->flags |= BSF_DEBUGGING; if (symbol->flags & BSF_DEBUGGING && bfd_is_abs_section (symbol->section)) { --- 911,922 ----- if (native->u.syment.n_sclass == C_FILE) symbol->flags |= BSF_DEBUGGING; + if (symbol->flags & BSF_SECTION_SYM && coff_data(abfd)->pe) + { + /* Writing section symbol to PE file */ + native->u.syment.n_sclass = C_SECTION; + } + if (symbol->flags & BSF_DEBUGGING && bfd_is_abs_section (symbol->section)) { *** bfd/orig/cofflink.c Tue Apr 15 02:21:58 1997 --- bfd/cofflink.c Wed Jul 16 09:00:14 1997 *************** *** 34,39 static boolean coff_link_check_ar_symbols PARAMS ((bfd *, struct bfd_link_info *, boolean *)); static boolean coff_link_add_symbols PARAMS ((bfd *, struct bfd_link_info *)); static char *dores_com PARAMS ((char *, bfd *, int)); static char *get_name PARAMS ((char *, char **)); static int process_embedded_commands --- 34,41 ----- static boolean coff_link_check_ar_symbols PARAMS ((bfd *, struct bfd_link_info *, boolean *)); static boolean coff_link_add_symbols PARAMS ((bfd *, struct bfd_link_info *)); + static asection *coff_get_first_section_of_archive + PARAMS ((struct bfd_link_info *, bfd *, CONST char *)); static char *dores_com PARAMS ((char *, bfd *, int)); static char *get_name PARAMS ((char *, char **)); static int process_embedded_commands *************** *** 2482,2487 + sec->output_offset); } else if (! info->relocateable) { if (! ((*info->callbacks->undefined_symbol) --- 2484,2510 ----- + sec->output_offset); } + else if (coff_data(input_bfd)->pe + && h->class == C_SECTION + && h->root.type == bfd_link_hash_undefined) + + { + /* Reference to a section label, currently undefined */ + asection *sec; + + sec = coff_get_first_section_of_archive(info, + h->root.u.undef.abfd, + h->root.root.string); + if (sec == NULL) + goto undef_sym; + + val = (sec->output_section->vma + + sec->output_offset + + sym->n_value + - sec->vma); + + } + else if (! info->relocateable) { undef_sym: *************** *** 2484,2489 else if (! info->relocateable) { if (! ((*info->callbacks->undefined_symbol) (info, h->root.root.string, input_bfd, input_section, rel->r_vaddr - input_section->vma))) --- 2507,2513 ----- else if (! info->relocateable) { + undef_sym: if (! ((*info->callbacks->undefined_symbol) (info, h->root.root.string, input_bfd, input_section, rel->r_vaddr - input_section->vma))) *************** *** 2494,2500 if (info->base_file) { /* Emit a reloc if the backend thinks it needs it. */ ! if (sym && pe_data(output_bfd)->in_reloc_p(output_bfd, howto)) { /* relocation to a symbol in a section which isn't absolute - we output the address here --- 2518,2525 ----- if (info->base_file) { /* Emit a reloc if the backend thinks it needs it. */ ! if (sym && pe_data(output_bfd)->in_reloc_p(output_bfd, howto) ! && !(input_section->flags & SEC_NEVER_LOAD)) { /* relocation to a symbol in a section which isn't absolute - we output the address here *************** *** 2554,2556 return true; } --- 2579,2612 ----- return true; } + static asection * + coff_get_first_section_of_archive(info, abfd, name) + struct bfd_link_info *info; + bfd *abfd; + CONST char *name; + { + bfd *sub; + asection *sect; + asection *rtn = NULL; + bfd_vma offset; + + /* move abfd to containing archive */ + if (abfd->my_archive == NULL) + return NULL; + abfd = abfd->my_archive; + + for (sub = info->input_bfds; sub != NULL; sub = sub->link_next) { + /* make sure that the bfd comes from the same archive */ + if (sub->my_archive == NULL + || strcmp(abfd->filename, sub->my_archive->filename) ) + continue; + + for (sect = sub->sections; sect != NULL; sect = sect->next) { + if (strcmp (sect->name, name)) + continue; + if (rtn == NULL || offset > sect->output_offset) + rtn = sect; + } + } + return rtn; + } *** bfd/orig/peicode.h Tue Apr 15 02:22:16 1997 --- bfd/peicode.h Wed Jul 16 09:00:27 1997 *************** *** 493,499 rather than something useful. When these symbols are encountered, change the value to 0 and the section number to 1 so that they will be handled somewhat correctly in the bfd code. */ ! if (in->n_sclass == 0x68) { in->n_value = 0x0; in->n_scnum = 1; /* I have tried setting the class to 3 and using the following to set --- 493,499 ----- rather than something useful. When these symbols are encountered, change the value to 0 and the section number to 1 so that they will be handled somewhat correctly in the bfd code. */ ! if (in->n_sclass == C_SECTION) { in->n_value = 0x0; /* in->n_scnum = 1;*/ /*!*/ /* I have tried setting the class to 3 and using the following to set *************** *** 495,501 somewhat correctly in the bfd code. */ if (in->n_sclass == 0x68) { in->n_value = 0x0; ! in->n_scnum = 1; /* I have tried setting the class to 3 and using the following to set the section number. This will put the address of the pointer to the string kernel32.dll at addresses 0 and 0x10 off start of idata section --- 495,501 ----- somewhat correctly in the bfd code. */ if (in->n_sclass == C_SECTION) { in->n_value = 0x0; ! /* in->n_scnum = 1;*/ /*!*/ /* I have tried setting the class to 3 and using the following to set the section number. This will put the address of the pointer to the string kernel32.dll at addresses 0 and 0x10 off start of idata section *************** *** 844,850 if (sec != NULL) { aout->DataDirectory[idx].VirtualAddress = sec->vma - base; ! aout->DataDirectory[idx].Size = pei_section_data (abfd, sec)->virt_size; sec->flags |= SEC_DATA; } } --- 844,852 ----- if (sec != NULL) { aout->DataDirectory[idx].VirtualAddress = sec->vma - base; ! if (coff_section_data(abfd, sec) != NULL ! && pei_section_data(abfd, sec) != NULL) ! aout->DataDirectory[idx].Size = pei_section_data (abfd, sec)->virt_size; sec->flags |= SEC_DATA; } } *************** *** 1123,1128 flags = SEC_DATA| IMAGE_SCN_MEM_READ | IMAGE_SCN_MEM_DISCARDABLE; else if (strcmp (scnhdr_int->s_name, ".idata") == 0) flags = IMAGE_SCN_MEM_READ | IMAGE_SCN_MEM_WRITE | SEC_DATA; else if (strcmp (scnhdr_int->s_name, ".rdata") == 0 || strcmp (scnhdr_int->s_name, ".edata") == 0) flags = IMAGE_SCN_MEM_READ | SEC_DATA; --- 1125,1132 ----- flags = SEC_DATA| IMAGE_SCN_MEM_READ | IMAGE_SCN_MEM_DISCARDABLE; else if (strcmp (scnhdr_int->s_name, ".idata") == 0) flags = IMAGE_SCN_MEM_READ | IMAGE_SCN_MEM_WRITE | SEC_DATA; + else if (strncmp (scnhdr_int->s_name, ".idata$", strlen(".idata$")) == 0) + flags = IMAGE_SCN_MEM_READ | IMAGE_SCN_MEM_WRITE | SEC_DATA; else if (strcmp (scnhdr_int->s_name, ".rdata") == 0 || strcmp (scnhdr_int->s_name, ".edata") == 0) flags = IMAGE_SCN_MEM_READ | SEC_DATA; *************** *** 1150,1155 } #endif bfd_h_put_32(abfd, flags, (bfd_byte *) scnhdr_ext->s_flags); } --- 1154,1162 ----- } #endif + /* put back alignment info */ + flags |= scnhdr_int->s_flags & IMAGE_SCN_ALIGN_64BYTES; + bfd_h_put_32(abfd, flags, (bfd_byte *) scnhdr_ext->s_flags); } *** binutils/orig/dlltool.c Fri May 2 02:54:19 1997 --- binutils/dlltool.c Wed Jul 16 09:01:59 1997 *************** *** 1437,1442 int i; export_type *exp; dlist_type *dl; sprintf (outfile, "t%s", exp_name); --- 1437,1443 ----- int i; export_type *exp; dlist_type *dl; + static char *xlate (char *); sprintf (outfile, "t%s", exp_name); *************** *** 1532,1538 fprintf(f,"%s Export Name Table\n", ASM_C); for (i = 0; (exp = d_exports_lexically[i]); i++) if (!exp->noname || show_allnames) ! fprintf (f, "n%d: %s \"%s\"\n", exp->ordinal, ASM_TEXT, exp->name); if (a_list) { --- 1533,1539 ----- fprintf(f,"%s Export Name Table\n", ASM_C); for (i = 0; (exp = d_exports_lexically[i]); i++) if (!exp->noname || show_allnames) ! fprintf (f, "n%d: %s \"%s\"\n", exp->ordinal, ASM_TEXT, xlate(exp->name)); if (a_list) { *************** *** 1722,1735 #ifndef DLLTOOL_PPC #define TEXT 0 ! #define DATA 1 ! #define BSS 2 ! #define IDATA7 3 ! #define IDATA5 4 ! #define IDATA4 5 ! #define IDATA6 6 ! #define PDATA 7 ! #define RDATA 8 #define NSECS 7 --- 1723,1731 ----- #ifndef DLLTOOL_PPC #define TEXT 0 ! #define IDATA5 1 ! #define IDATA4 2 ! #define IDATA6 3 #define NSECS 4 *************** *** 1731,1737 #define PDATA 7 #define RDATA 8 ! #define NSECS 7 static sinfo secdata[NSECS] = --- 1727,1733 ----- #define IDATA4 2 #define IDATA6 3 ! #define NSECS 4 static sinfo secdata[NSECS] = *************** *** 1737,1745 static sinfo secdata[NSECS] = { { TEXT, ".text", SEC_CODE | SEC_HAS_CONTENTS, 2}, - { DATA, ".data", SEC_DATA, 2}, - { BSS, ".bss", 0, 2}, - { IDATA7, ".idata$7", SEC_HAS_CONTENTS, 2}, { IDATA5, ".idata$5", SEC_HAS_CONTENTS, 2}, { IDATA4, ".idata$4", SEC_HAS_CONTENTS, 2}, { IDATA6, ".idata$6", SEC_HAS_CONTENTS, 1} --- 1733,1738 ----- static sinfo secdata[NSECS] = { { TEXT, ".text", SEC_CODE | SEC_HAS_CONTENTS, 2}, { IDATA5, ".idata$5", SEC_HAS_CONTENTS, 2}, { IDATA4, ".idata$4", SEC_HAS_CONTENTS, 2}, { IDATA6, ".idata$6", SEC_HAS_CONTENTS, 1} *************** *** 1832,1884 export_type *exp; int i; { - if (0) - { - FILE *f; - char *prefix="d"; - sprintf (outfile, "%ss%d.s", prefix, i); - f = fopen (outfile, FOPEN_WT); - fprintf (f, "\t.text\n"); - fprintf (f, "\t%s\t%s%s\n", ASM_GLOBAL, ASM_PREFIX, exp->name); - fprintf (f, "\t%s\t__imp_%s\n", ASM_GLOBAL, exp->name); - fprintf (f, "%s%s:\n\t%s\t__imp_%s\n", ASM_PREFIX, - exp->name, ASM_JUMP, exp->name); - - fprintf (f, "\t.section\t.idata$7\t%s To force loading of head\n", ASM_C); - fprintf (f, "\t%s\t%s\n", ASM_LONG, head_label); - - - fprintf (f,"%s Import Address Table\n", ASM_C); - - fprintf (f, "\t.section .idata$5\n"); - fprintf (f, "__imp_%s:\n", exp->name); - - dump_iat (f, exp); - - fprintf (f, "\n%s Import Lookup Table\n", ASM_C); - fprintf (f, "\t.section .idata$4\n"); - - dump_iat (f, exp); - - if(!exp->noname || show_allnames) - { - fprintf (f, "%s Hint/Name table\n", ASM_C); - fprintf (f, "\t.section .idata$6\n"); - fprintf (f, "ID%d:\t%s\t%d\n", exp->ordinal, ASM_SHORT, exp->hint); - fprintf (f, "\t%s\t\"%s\"\n", ASM_TEXT, xlate (exp->name)); - } - - fclose (f); - - - sprintf (outfile, "-o %ss%d.o %ss%d.s", prefix, i, prefix, i); - - run (as_name, outfile); - - } - else /* fnord */ - { - bfd *abfd; asymbol *exp_label; --- 1825,1830 ----- export_type *exp; int i; { bfd *abfd; asymbol *exp_label; *************** *** 1948,1957 rdata section, the first element of which is a pointer to the code (..function_name), and the second points to the .toc */ ! if (machine == MPPC) ! exp_label->section = secdata[RDATA].sec; ! else ! exp_label->section = secdata[TEXT].sec; exp_label->flags = BSF_GLOBAL; exp_label->value = 0; --- 1894,1904 ----- rdata section, the first element of which is a pointer to the code (..function_name), and the second points to the .toc */ ! #ifdef DLLTOOL_PPC ! exp_label->section = secdata[RDATA].sec; ! #else ! exp_label->section = secdata[TEXT].sec; ! #endif exp_label->flags = BSF_GLOBAL; exp_label->value = 0; *************** *** 1959,1965 ptrs[oidx++] = exp_label; iname = bfd_make_empty_symbol(abfd); ! iname->name = make_label ("__imp_", exp->name); iname->section = secdata[IDATA5].sec; iname->flags = BSF_GLOBAL; iname->value = 0; --- 1906,1912 ----- ptrs[oidx++] = exp_label; iname = bfd_make_empty_symbol(abfd); ! iname->name = make_label ("_imp__", exp->name); iname->section = secdata[IDATA5].sec; iname->flags = BSF_GLOBAL; iname->value = 0; *************** *** 2029,2035 { rel->howto = bfd_reloc_type_lookup (abfd, BFD_RELOC_16_GOTOFF); - rel->sym_ptr_ptr = iname_pp; } else { --- 1976,1981 ----- { rel->howto = bfd_reloc_type_lookup (abfd, BFD_RELOC_16_GOTOFF); } else { *************** *** 2034,2040 else { rel->howto = bfd_reloc_type_lookup (abfd, BFD_RELOC_32); - rel->sym_ptr_ptr = secdata[IDATA5].sympp; } sec->orelocation = rpp; sec->reloc_count = 1; --- 1980,1985 ----- else { rel->howto = bfd_reloc_type_lookup (abfd, BFD_RELOC_32); } rel->sym_ptr_ptr = iname_pp; sec->orelocation = rpp; *************** *** 2036,2041 rel->howto = bfd_reloc_type_lookup (abfd, BFD_RELOC_32); rel->sym_ptr_ptr = secdata[IDATA5].sympp; } sec->orelocation = rpp; sec->reloc_count = 1; break; --- 1981,1987 ----- { rel->howto = bfd_reloc_type_lookup (abfd, BFD_RELOC_32); } + rel->sym_ptr_ptr = iname_pp; sec->orelocation = rpp; sec->reloc_count = 1; break; *************** *** 2082,2087 strcpy (si->data + 2, xlate (exp->name)); } break; case IDATA7: si->size = 4; si->data =xmalloc(4); --- 2028,2034 ----- strcpy (si->data + 2, xlate (exp->name)); } break; + #ifdef DLLTOOL_PPC case IDATA7: si->size = 4; si->data =xmalloc(4); *************** *** 2205,2210 sec->orelocation = rpp; sec->reloc_count = 2; break; } } --- 2152,2158 ----- sec->orelocation = rpp; sec->reloc_count = 2; break; + #endif /*DLLTOOL_PPC*/ } } *************** *** 2243,2249 return abfd; } - } static bfd * --- 2191,2196 ----- return abfd; } char import_desc[20]; char null_thunk[4]; *************** *** 2245,2250 } static bfd * make_head() --- 2192,2199 ----- } + char import_desc[20]; + char null_thunk[4]; static bfd * make_head() *************** *** 2249,2255 static bfd * make_head() { ! FILE * f = fopen ("dh.s", FOPEN_WT); fprintf (f, "%s IMAGE_IMPORT_DESCRIPTOR\n", ASM_C); fprintf (f, "\t.section .idata$2\n"); --- 2198,2210 ----- static bfd * make_head() { ! bfd *abfd; ! asection *idata2, *idata6; ! asymbol *symarray[7]; ! asymbol **sym = symarray; ! asymbol *sym_idata4, *sym_idata5, *sym_idata6; ! arelent **rpp; ! char *str; abfd = bfd_openw ("dh.o", HOW_BFD_TARGET); *************** *** 2251,2258 { FILE * f = fopen ("dh.s", FOPEN_WT); ! fprintf (f, "%s IMAGE_IMPORT_DESCRIPTOR\n", ASM_C); ! fprintf (f, "\t.section .idata$2\n"); fprintf(f,"\t%s\t%s\n", ASM_GLOBAL,head_label); --- 2206,2212 ----- arelent **rpp; char *str; ! abfd = bfd_openw ("dh.o", HOW_BFD_TARGET); if (!abfd) { *************** *** 2254,2260 fprintf (f, "%s IMAGE_IMPORT_DESCRIPTOR\n", ASM_C); fprintf (f, "\t.section .idata$2\n"); ! fprintf(f,"\t%s\t%s\n", ASM_GLOBAL,head_label); fprintf (f, "%s:\n", head_label); --- 2208,2221 ----- abfd = bfd_openw ("dh.o", HOW_BFD_TARGET); ! if (!abfd) ! { ! fprintf (stderr, "%s: bfd_open failed open output file dh.o\n", ! program_name); ! exit(1); ! } ! bfd_set_format (abfd, bfd_object); ! bfd_set_arch_mach (abfd, HOW_BFD_ARCH, 0); idata2 = bfd_make_section (abfd, ".idata$2"); bfd_set_section_flags (abfd, idata2, SEC_HAS_CONTENTS); *************** *** 2256,2262 fprintf(f,"\t%s\t%s\n", ASM_GLOBAL,head_label); ! fprintf (f, "%s:\n", head_label); fprintf (f, "\t%shname%s\t%sPtr to image import by name list\n", ASM_RVA_BEFORE, ASM_RVA_AFTER, ASM_C); --- 2217,2226 ----- bfd_set_format (abfd, bfd_object); bfd_set_arch_mach (abfd, HOW_BFD_ARCH, 0); ! idata2 = bfd_make_section (abfd, ".idata$2"); ! bfd_set_section_flags (abfd, idata2, SEC_HAS_CONTENTS); ! bfd_set_section_alignment (abfd, idata2, 2); ! idata2->output_section = idata2; /* label for import descriptor */ *sym = bfd_make_empty_symbol (abfd); *************** *** 2258,2265 fprintf (f, "%s:\n", head_label); ! fprintf (f, "\t%shname%s\t%sPtr to image import by name list\n", ! ASM_RVA_BEFORE, ASM_RVA_AFTER, ASM_C); fprintf (f, "\t%sthis should be the timestamp, but NT sometimes\n", ASM_C); fprintf (f, "\t%sdoesn't load DLLs when this is set.\n", ASM_C); --- 2222,2234 ----- bfd_set_section_alignment (abfd, idata2, 2); idata2->output_section = idata2; ! /* label for import descriptor */ ! *sym = bfd_make_empty_symbol (abfd); ! (*sym)->name = head_label; ! (*sym)->section = idata2; ! (*sym)->flags = BSF_GLOBAL; ! (*sym)->value = 0; ! sym++; *sym = sym_idata4 = bfd_make_empty_symbol (abfd); (*sym)->name = ".idata$4"; *************** *** 2261,2278 fprintf (f, "\t%shname%s\t%sPtr to image import by name list\n", ASM_RVA_BEFORE, ASM_RVA_AFTER, ASM_C); ! fprintf (f, "\t%sthis should be the timestamp, but NT sometimes\n", ASM_C); ! fprintf (f, "\t%sdoesn't load DLLs when this is set.\n", ASM_C); ! fprintf (f, "\t%s\t0\t%s loaded time\n", ASM_LONG, ASM_C); ! fprintf (f, "\t%s\t0\t%s Forwarder chain\n", ASM_LONG, ASM_C); ! fprintf (f, "\t%s__%s_iname%s\t%s imported dll's name\n", ! ASM_RVA_BEFORE, ! imp_name_lab, ! ASM_RVA_AFTER, ! ASM_C); ! fprintf (f, "\t%sfthunk%s\t%s pointer to firstthunk\n", ! ASM_RVA_BEFORE, ! ASM_RVA_AFTER, ASM_C); fprintf (f, "%sStuff for compatibility\n", ASM_C); --- 2230,2241 ----- (*sym)->value = 0; sym++; ! *sym = sym_idata4 = bfd_make_empty_symbol (abfd); ! (*sym)->name = ".idata$4"; ! (*sym)->section = (asection *)&bfd_und_section; ! (*sym)->flags = BSF_SECTION_SYM; ! (*sym)->value = 0xc0000040; ! sym++; *sym = sym_idata5 = bfd_make_empty_symbol (abfd); (*sym)->name = ".idata$5"; *************** *** 2274,2280 ASM_RVA_BEFORE, ASM_RVA_AFTER, ASM_C); ! fprintf (f, "%sStuff for compatibility\n", ASM_C); if (!no_idata5) { --- 2237,2248 ----- (*sym)->value = 0xc0000040; sym++; ! *sym = sym_idata5 = bfd_make_empty_symbol (abfd); ! (*sym)->name = ".idata$5"; ! (*sym)->section = (asection *)&bfd_und_section; ! (*sym)->flags = BSF_SECTION_SYM; ! (*sym)->value = 0xc0000040; ! sym++; /* label for NULL_IMPORT_DESCRIPTOR */ *sym = bfd_make_empty_symbol (abfd); *************** *** 2276,2282 fprintf (f, "%sStuff for compatibility\n", ASM_C); ! if (!no_idata5) { fprintf (f, "\t.section\t.idata$5\n"); fprintf (f, "\t%s\t0\n", ASM_LONG); --- 2244,2335 ----- (*sym)->value = 0xc0000040; sym++; ! /* label for NULL_IMPORT_DESCRIPTOR */ ! *sym = bfd_make_empty_symbol (abfd); ! (*sym)->name = "__NULL_IMPORT_DESCRIPTOR"; ! (*sym)->section = (asection *)&bfd_und_section; ! (*sym)->flags = BSF_NO_FLAGS; ! (*sym)->value = 0; ! sym++; ! ! /* label for NULL_THUNK */ ! str = xmalloc (strlen(imp_name_lab) + 18); ! str[0] = '\177'; ! strcpy (str+1, imp_name_lab); ! strcat (str, "_NULL_THUNK_DATA"); ! *sym = bfd_make_empty_symbol (abfd); ! (*sym)->name = str; ! (*sym)->section = (asection *)&bfd_und_section; ! (*sym)->flags = BSF_NO_FLAGS; ! (*sym)->value = 0; ! sym++; ! ! *sym = sym_idata6 = bfd_make_empty_symbol (abfd); ! /* name and section filled in later */ ! (*sym)->flags = BSF_LOCAL /*| BSF_SECTION_SYM*/; ! (*sym)->value = 0; ! sym++; ! ! memset (&import_desc, 0, sizeof (import_desc)); ! bfd_set_section_size (abfd, idata2, sizeof (import_desc)); ! bfd_set_section_vma (abfd, idata2, 0); ! ! rpp = xmalloc (sizeof (arelent *) * 4); ! rpp[0] = xmalloc (sizeof (arelent) * 3); ! rpp[1] = rpp[0]+1; ! rpp[2] = rpp[0]+2; ! rpp[3] = 0; ! ! /* Import list pointer */ ! rpp[0]->address = 0; ! rpp[0]->addend = 0; ! rpp[0]->howto = bfd_reloc_type_lookup (abfd, BFD_RELOC_RVA); ! rpp[0]->sym_ptr_ptr = &sym_idata4; ! ! /* library name pointer */ ! rpp[1]->address = 12; ! rpp[1]->addend = 0; ! rpp[1]->howto = bfd_reloc_type_lookup (abfd, BFD_RELOC_RVA); ! rpp[1]->sym_ptr_ptr = &sym_idata6; ! ! /* Thunk list pointer */ ! rpp[2]->address = 16; ! rpp[2]->addend = 0; ! rpp[2]->howto = bfd_reloc_type_lookup (abfd, BFD_RELOC_RVA); ! rpp[2]->sym_ptr_ptr = &sym_idata5; ! ! idata2->reloc_count = 3; ! idata2->orelocation = rpp; ! ! idata6 = bfd_make_section (abfd, ".idata$6"); ! bfd_set_section_flags (abfd, idata6, SEC_HAS_CONTENTS); ! bfd_set_section_alignment (abfd, idata6, 1); ! bfd_set_section_size (abfd, idata6, strlen (dll_name)+1); ! bfd_set_section_vma (abfd, idata6, 0); ! idata6->output_section = idata6; ! sym_idata6->name = idata6->name; ! sym_idata6->section = idata6; ! ! /* generate output */ ! bfd_set_section_contents (abfd, idata2, &import_desc, 0, ! sizeof (import_desc)); ! bfd_set_section_contents (abfd, idata6, dll_name, 0, strlen(dll_name)+1); ! ! bfd_set_symtab (abfd, symarray, sym-symarray); ! bfd_close (abfd); ! ! return bfd_openr ("dh.o", HOW_BFD_TARGET); ! } ! ! static bfd * ! make_hterm() ! { ! bfd *abfd; ! asection *idata3; ! asymbol *sym; ! ! abfd = bfd_openw ("dht.o", HOW_BFD_TARGET); ! if (!abfd) { fprintf (stderr, "%s: bfd_open failed open output file dht.o\n", program_name); *************** *** 2278,2286 if (!no_idata5) { ! fprintf (f, "\t.section\t.idata$5\n"); ! fprintf (f, "\t%s\t0\n", ASM_LONG); ! fprintf (f, "fthunk:\n"); } if (!no_idata4) { --- 2331,2339 ----- abfd = bfd_openw ("dht.o", HOW_BFD_TARGET); if (!abfd) { ! fprintf (stderr, "%s: bfd_open failed open output file dht.o\n", ! program_name); ! exit(1); } bfd_set_format (abfd, bfd_object); bfd_set_arch_mach (abfd, HOW_BFD_ARCH, 0); *************** *** 2282,2290 fprintf (f, "\t%s\t0\n", ASM_LONG); fprintf (f, "fthunk:\n"); } ! if (!no_idata4) ! { ! fprintf (f, "\t.section\t.idata$4\n"); fprintf (f, "\t%s\t0\n", ASM_LONG); fprintf (f, "\t.section .idata$4\n"); --- 2335,2342 ----- program_name); exit(1); } ! bfd_set_format (abfd, bfd_object); ! bfd_set_arch_mach (abfd, HOW_BFD_ARCH, 0); idata3 = bfd_make_section (abfd, ".idata$3"); bfd_set_section_flags (abfd, idata3, SEC_HAS_CONTENTS); *************** *** 2286,2296 { fprintf (f, "\t.section\t.idata$4\n"); ! fprintf (f, "\t%s\t0\n", ASM_LONG); ! fprintf (f, "\t.section .idata$4\n"); ! fprintf (f, "hname:\n"); ! } ! fclose (f); sprintf (outfile, "-o dh.o dh.s"); run (as_name, outfile); --- 2338,2347 ----- bfd_set_format (abfd, bfd_object); bfd_set_arch_mach (abfd, HOW_BFD_ARCH, 0); ! idata3 = bfd_make_section (abfd, ".idata$3"); ! bfd_set_section_flags (abfd, idata3, SEC_HAS_CONTENTS); ! bfd_set_section_alignment (abfd, idata3, 2); ! idata3->output_section = idata3; /* label for NULL_IMPORT_DESCRIPTOR */ sym = bfd_make_empty_symbol (abfd); *************** *** 2292,2299 } fclose (f); ! sprintf (outfile, "-o dh.o dh.s"); ! run (as_name, outfile); return bfd_openr ("dh.o", HOW_BFD_TARGET); } --- 2343,2354 ----- bfd_set_section_alignment (abfd, idata3, 2); idata3->output_section = idata3; ! /* label for NULL_IMPORT_DESCRIPTOR */ ! sym = bfd_make_empty_symbol (abfd); ! sym->name = "__NULL_IMPORT_DESCRIPTOR"; ! sym->section = idata3; ! sym->flags = BSF_GLOBAL; ! sym->value = 0; memset (&import_desc, 0, sizeof (import_desc)); bfd_set_section_size (abfd, idata3, sizeof (import_desc)); *************** *** 2295,2301 sprintf (outfile, "-o dh.o dh.s"); run (as_name, outfile); ! return bfd_openr ("dh.o", HOW_BFD_TARGET); } static bfd * --- 2350,2364 ----- sym->flags = BSF_GLOBAL; sym->value = 0; ! memset (&import_desc, 0, sizeof (import_desc)); ! bfd_set_section_size (abfd, idata3, sizeof (import_desc)); ! bfd_set_section_vma (abfd, idata3, 0); ! bfd_set_section_contents (abfd, idata3, &import_desc, 0, ! sizeof (import_desc)); ! bfd_set_symtab (abfd, &sym, 1); ! bfd_close (abfd); ! ! return bfd_openr ("dht.o", HOW_BFD_TARGET); } static bfd * *************** *** 2301,2307 static bfd * make_tail() { ! FILE * f; f = fopen ("dt.s", FOPEN_WT); --- 2364,2373 ----- static bfd * make_tail() { ! bfd *abfd; ! asection *idata4, *idata5; ! asymbol *sym; ! char *str; abfd = bfd_openw ("dt.o", HOW_BFD_TARGET); if (!abfd) *************** *** 2303,2311 { FILE * f; ! f = fopen ("dt.s", FOPEN_WT); ! ! if (!no_idata4) { fprintf (f, "\t.section .idata$4\n"); fprintf (f, "\t%s\t0\n", ASM_LONG); --- 2369,2376 ----- asymbol *sym; char *str; ! abfd = bfd_openw ("dt.o", HOW_BFD_TARGET); ! if (!abfd) { fprintf (stderr, "%s: bfd_open failed open output file dt.o\n", program_name); *************** *** 2307,2314 if (!no_idata4) { ! fprintf (f, "\t.section .idata$4\n"); ! fprintf (f, "\t%s\t0\n", ASM_LONG); } if (!no_idata5) { --- 2372,2380 ----- abfd = bfd_openw ("dt.o", HOW_BFD_TARGET); if (!abfd) { ! fprintf (stderr, "%s: bfd_open failed open output file dt.o\n", ! program_name); ! exit(1); } bfd_set_format (abfd, bfd_object); bfd_set_arch_mach (abfd, HOW_BFD_ARCH, 0); *************** *** 2310,2320 fprintf (f, "\t.section .idata$4\n"); fprintf (f, "\t%s\t0\n", ASM_LONG); } ! if (!no_idata5) ! { ! fprintf (f, "\t.section .idata$5\n"); ! fprintf (f, "\t%s\t0\n", ASM_LONG); ! } #ifdef DLLTOOL_PPC /* Normally, we need to see a null descriptor built in idata$3 to --- 2376,2383 ----- program_name); exit(1); } ! bfd_set_format (abfd, bfd_object); ! bfd_set_arch_mach (abfd, HOW_BFD_ARCH, 0); idata4 = bfd_make_section (abfd, ".idata$4"); bfd_set_section_flags (abfd, idata4, SEC_HAS_CONTENTS); *************** *** 2316,2329 fprintf (f, "\t%s\t0\n", ASM_LONG); } ! #ifdef DLLTOOL_PPC ! /* Normally, we need to see a null descriptor built in idata$3 to ! act as the terminator for the list. The ideal way, I suppose, ! would be to mark this section as a comdat type 2 section, so ! only one would appear in the final .exe (if our linker supported ! comdat, that is) or cause it to be inserted by something else (say ! crt0) ! */ fprintf (f, "\t.section .idata$3\n"); fprintf (f, "\t%s\t0\n", ASM_LONG); --- 2379,2390 ----- bfd_set_format (abfd, bfd_object); bfd_set_arch_mach (abfd, HOW_BFD_ARCH, 0); ! idata4 = bfd_make_section (abfd, ".idata$4"); ! bfd_set_section_flags (abfd, idata4, SEC_HAS_CONTENTS); ! bfd_set_section_alignment (abfd, idata4, 2); ! idata4->output_section = idata4; ! bfd_set_section_size (abfd, idata4, sizeof (null_thunk)); ! bfd_set_section_vma (abfd, idata4, 0); idata5 = bfd_make_section (abfd, ".idata$5"); bfd_set_section_flags (abfd, idata5, SEC_HAS_CONTENTS); *************** *** 2325,2337 crt0) */ ! fprintf (f, "\t.section .idata$3\n"); ! fprintf (f, "\t%s\t0\n", ASM_LONG); ! fprintf (f, "\t%s\t0\n", ASM_LONG); ! fprintf (f, "\t%s\t0\n", ASM_LONG); ! fprintf (f, "\t%s\t0\n", ASM_LONG); ! fprintf (f, "\t%s\t0\n", ASM_LONG); ! #endif #ifdef DLLTOOL_PPC /* Other PowerPC NT compilers use idata$6 for the dllname, so I --- 2386,2397 ----- bfd_set_section_size (abfd, idata4, sizeof (null_thunk)); bfd_set_section_vma (abfd, idata4, 0); ! idata5 = bfd_make_section (abfd, ".idata$5"); ! bfd_set_section_flags (abfd, idata5, SEC_HAS_CONTENTS); ! bfd_set_section_alignment (abfd, idata5, 2); ! idata5->output_section = idata5; ! bfd_set_section_size (abfd, idata5, sizeof (null_thunk)); ! bfd_set_section_vma (abfd, idata5, 0); /* label for NULL_THUNK */ str = xmalloc (strlen(imp_name_lab) + 18); *************** *** 2333,2345 fprintf (f, "\t%s\t0\n", ASM_LONG); #endif ! #ifdef DLLTOOL_PPC ! /* Other PowerPC NT compilers use idata$6 for the dllname, so I ! do too. Original, huh? */ ! fprintf (f, "\t.section .idata$6\n"); ! #else ! fprintf (f, "\t.section .idata$7\n"); ! #endif fprintf (f, "\t%s\t__%s_iname\n", ASM_GLOBAL, imp_name_lab); fprintf (f, "__%s_iname:\t%s\t\"%s\"\n", imp_name_lab, ASM_TEXT, dll_name); --- 2393,2408 ----- bfd_set_section_size (abfd, idata5, sizeof (null_thunk)); bfd_set_section_vma (abfd, idata5, 0); ! /* label for NULL_THUNK */ ! str = xmalloc (strlen(imp_name_lab) + 18); ! str[0] = '\177'; ! strcpy (str+1, imp_name_lab); ! strcat (str, "_NULL_THUNK_DATA"); ! sym = bfd_make_empty_symbol (abfd); ! sym->name = str; ! sym->section = idata5; ! sym->flags = BSF_GLOBAL; ! sym->value = 0; bfd_set_section_contents (abfd, idata4, &null_thunk, 0, sizeof (null_thunk)); bfd_set_section_contents (abfd, idata5, &null_thunk, 0, sizeof (null_thunk)); *************** *** 2341,2349 fprintf (f, "\t.section .idata$7\n"); #endif ! fprintf (f, "\t%s\t__%s_iname\n", ASM_GLOBAL, imp_name_lab); ! fprintf (f, "__%s_iname:\t%s\t\"%s\"\n", imp_name_lab, ASM_TEXT, dll_name); ! fclose (f); sprintf (outfile, "-o dt.o dt.s"); run (as_name, outfile); --- 2404,2413 ----- sym->flags = BSF_GLOBAL; sym->value = 0; ! bfd_set_section_contents (abfd, idata4, &null_thunk, 0, sizeof (null_thunk)); ! bfd_set_section_contents (abfd, idata5, &null_thunk, 0, sizeof (null_thunk)); ! bfd_set_symtab (abfd, &sym, 1); ! bfd_close (abfd); return bfd_openr ("dt.o", HOW_BFD_TARGET); } *************** *** 2345,2352 fprintf (f, "__%s_iname:\t%s\t\"%s\"\n", imp_name_lab, ASM_TEXT, dll_name); fclose (f); - sprintf (outfile, "-o dt.o dt.s"); - run (as_name, outfile); return bfd_openr ("dt.o", HOW_BFD_TARGET); } --- 2409,2414 ----- bfd_set_symtab (abfd, &sym, 1); bfd_close (abfd); return bfd_openr ("dt.o", HOW_BFD_TARGET); } *************** *** 2356,2361 int i; export_type *exp; bfd *ar_head; bfd *ar_tail; bfd *outarch; bfd * head = 0; --- 2418,2424 ----- int i; export_type *exp; bfd *ar_head; + bfd *ar_hterm; bfd *ar_tail; bfd *outarch; bfd * head = 0; *************** *** 2377,2384 /* Work out a reasonable size of things to put onto one line. */ ! ar_head = make_head(); ! ar_tail = make_tail(); for (i = 0; (exp = d_exports_lexically[i]); i++) { --- 2440,2448 ----- /* Work out a reasonable size of things to put onto one line. */ ! ar_head = make_head (outarch); ! ar_hterm = make_hterm (outarch); ! ar_tail = make_tail(outarch); for (i = 0; (exp = d_exports_lexically[i]); i++) { *************** *** 2382,2388 for (i = 0; (exp = d_exports_lexically[i]); i++) { ! bfd *n = make_one_lib_file (exp, i); /* fnord */ n->next = head; head = n; } --- 2446,2452 ----- for (i = 0; (exp = d_exports_lexically[i]); i++) { ! bfd *n = make_one_lib_file (exp, i, outarch); n->next = head; head = n; } *************** *** 2390,2396 /* Now stick them all into the archive */ ar_head->next = head; ! ar_tail->next = ar_head; head = ar_tail; bfd_set_archive_head (outarch, head); --- 2454,2461 ----- /* Now stick them all into the archive */ ar_head->next = head; ! ar_hterm->next = ar_head; ! ar_tail->next = ar_hterm; head = ar_tail; bfd_set_archive_head (outarch, head); *************** *** 2399,2412 /* Delete all the temp files */ if (dontdeltemps == 0) { ! sprintf (outfile, "dh.o"); ! unlink (outfile); ! sprintf (outfile, "dh.s"); ! unlink (outfile); ! sprintf (outfile, "dt.o"); ! unlink (outfile); ! sprintf (outfile, "dt.s"); ! unlink (outfile); } if (dontdeltemps < 2) --- 2464,2472 ----- /* Delete all the temp files */ if (dontdeltemps == 0) { ! unlink ("dh.o"); ! unlink ("dht.o"); ! unlink ("dt.o"); } if (dontdeltemps < 2) *************** *** 2878,2884 if (!isalpha (*p) && !isdigit (*p)) *p = '_'; } ! head_label = make_label("_head_", imp_name_lab); gen_lib_file (); } --- 2938,2944 ----- if (!isalpha (*p) && !isdigit (*p)) *p = '_'; } ! head_label = make_label ("_IMPORT_DESCRIPTOR_", imp_name_lab); gen_lib_file (); } *************** *** 2888,2891 return 0; } ! --- 2948,2954 ----- return 0; } ! wwww(char *fmt, int a1, int a2, int a3, int a4, int a5, int a6) ! { ! fprintf(stderr, fmt, a1, a2, a3, a4, a5, a6); ! } - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From stanton@haas.berkeley.edu Fri Aug 15 21:41:00 1997 From: stanton@haas.berkeley.edu (Richard Stanton) Date: Fri, 15 Aug 1997 21:41:00 -0000 Subject: Problem compiling/linking large number of object files. Message-ID: <199708152052.NAA09940@haas.berkeley.edu> I'm using GNU make with the emx C compiler under NT (I've just switched from OS/2, and like having executables that will run under either OS). This may not be make related, and may be a generic NT problem, but it looks like there's a maximum command line limit or something that I'm running into. For example, while all the .o files compile OK, here's what happens when I try and link them all together: cd c:/projects/sbir/ make -k np2f.exe gcc -o np2f.exe -Le:/lib np2f.o parse.o matrix.o nonpar.o nrutil.o matstat.o constant.o sort.o locate.o quantile.o powi.o indsort.o make: execvp: gcc: Invalid argument make: *** [np2f.exe] Error 127 If I just delete the last file in the list (actually, there were originally lots more than this, but I deleted them until I got to what seems to be a critical point), it compiles fine (except for the missing symbols...): cd c:/projects/sbir/ make -k np2f.exe gcc -o np2f.exe -Le:/lib np2f.o parse.o matrix.o nonpar.o nrutil.o matstat.o constant.o sort.o locate.o quantile.o powi.o np2f.o: Undefined symbol _phrtsd referenced from text segment np2f.o: Undefined symbol _setall referenced from text segment etc. Any idea what the problem is here? It compiles fine under OS/2, if that makes any difference. Thanks for any suggestions. Richard Stanton - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sat Aug 16 01:07:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sat, 16 Aug 1997 01:07:00 -0000 Subject: What Terminal are there? Message-ID: <01BCAA3D.2122D870@sos> Karl Martin Syring wrote: > But heck, my Dos-Terminal doesn't let me enter the '|' , nor the []s, > nor the '@' nor the '#' nor the '~' !!!! They all are entered by using > the Alt-Gr Key, so you see I am sitting at a swiss-german Keyboard with Use the right Alt key, but not the left one. The left Alt key generates metachar esc prefix. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jasiu@elkomtech.com.pl Sat Aug 16 02:27:00 1997 From: jasiu@elkomtech.com.pl (Janusz Gorecki) Date: Sat, 16 Aug 1997 02:27:00 -0000 Subject: errors while compiling Message-ID: <01BCAA36.C8C8F3E0@jasiu> Mike wrote: > i just downloaded b18 of the gnuwin32 tools and i get the following error> > while compiling simple hello world test: > > (unknown) error while dumping state (probably corrupt stack) > in cygwin_except_handler > > Does anyone know what this means and how to fix it? > Mike Just make sure you don't have an old cygwin.dll in /windows/system directory or so. Janusz - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sat Aug 16 05:21:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sat, 16 Aug 1997 05:21:00 -0000 Subject: Pipes interfere with terminal output under coolview 970810 Message-ID: <01BCAA5E.5DF227D0@sos> Michael A. Chase wrote: > When I use a pipe on the command line, the results from the final command > do not appear on the display. I can send the output to a file. > > I'm using WinNt 4.0 SP3, gnuwin32 b18, and coolview dated 970810. The > problem goes away when I reinstall the cgywin.dll dated 970731. Fixed. diff -ur /winsup/dcrt0.cc winsup/dcrt0.cc --- /winsup/dcrt0.cc Tue Aug 12 23:46:29 1997 +++ winsup/dcrt0.cc Sat Aug 16 12:45:35 1997 @@ -671,7 +671,7 @@ } /* Kill the foreground process group on session leader exit */ - if (getpid () == u->self->sid && u->self->ctty != -1) + if (getpgrp () > 0 && getpid () == u->self->sid && u->self->ctty != -1) { tty *ttyp = s->t.gettty (u->self->ctty); -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sat Aug 16 05:21:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sat, 16 Aug 1997 05:21:00 -0000 Subject: Problem compiling/linking large number of object files. Message-ID: <01BCAA3C.A53F9AE0@sos> Richard Stanton wrote: > The problem actually seems to be related to using the emx C compiler > under NT, rather than being related to GNU make. It's thus not exactly Is emx C native win32 task or it is a dos program? The length of dos program's command line is limited to 128 chars (remember the PSP format). > At a command prompt, I tried the command > > gcc -o np2f.exe -Le:/lib np2f.o parse.o matrix.o nonpar.o nrutil.o matstat.o constant.o sort.o locate.o quantile.o powi.o indsort.o brent.o dfpmin.o gamma.o linmin.o gradcd.o mnbrak.o ranlib.o intutils.o ols.o chol.o cholbksb.o invpd.o t.o f.o beta.o gammln.o mthutl.o -lm > > Nothing happened - I quickly got a command prompt back, but no .exe > file was created. The problem seems to be something to do with the > number of files, or length of the command line, as shorter lists of .o > files work fine. While building libX11.dll the ld's command line contains about 400 filenames and wraps to ~80 screen lines :-) But ld is win32 task. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From stanton@haas.berkeley.edu Sat Aug 16 06:31:00 1997 From: stanton@haas.berkeley.edu (Richard Stanton) Date: Sat, 16 Aug 1997 06:31:00 -0000 Subject: How to avoid GNU make case sensitivity? References: <01BCA982.087C55A0@gater.krystalbank.msk.ru> Message-ID: <199708151739.KAA11732@haas.berkeley.edu> Sergey> #!/bin/sh Sergey> for i in *; do Sergey> new=`echo $i|tr [A-Z] [a-z]`; Sergey> mv $i $new Sergey> done Or, using 4NT, in the topmost directory, where I want to "downcase" everything in that directory or below: for /R . %%f in (*) lower %f where the batch file lower.cmd contains the line ren %1 %@lower[%1] a. This could all be done in one line. b. I don't know if I need the double %%, but it doesn't hurt. Richard Stanton - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From martin@mrrl.lut.ac.uk Sat Aug 16 06:31:00 1997 From: martin@mrrl.lut.ac.uk (Martin Hamilton) Date: Sat, 16 Aug 1997 06:31:00 -0000 Subject: The Win32 binaries on the FTP sites References: <33f1adc8.0@epflnews.epfl.ch> Message-ID: -----BEGIN PGP SIGNED MESSAGE----- Martin Ouwehand writes: | I know of three such ports to Win32 (with source/patch). | In no particular order: | | ftp://ftp.net.lut.ac.uk/martin/ssh-win32/ | http://guardian.htu.tuwien.ac.at/therapy/ssh/ | http://bmrc.berkeley.edu/people/chaffee/winntutil.html [..] Just to note that a new release of the GNU-Win32 development environment has come out since I did this, and several people have reported varying degrees of success building ssh on its mailing list - cc'd. I've not yet managed to find the time to sit down and go through the process again... :-) Could I suggest that the people who are interested in seeing a GNU-Win32 based ssh port get together somewhere so as to avoid duplicating effort ? Cheerio, Martin -----BEGIN PGP SIGNATURE----- Version: 2.6.3i Charset: noconv iQCVAwUBM/QY+9ZdpXZXTSjhAQFZVgP+MqEnBvh2Pf3IFf/bE8Ow8+GNqZY5+hg2 ctDsX+h4cg/2Yind8Hy3bkjChg2er0IiL5RjpZ3j1/mozuTBWD7y0GsAnUUNj0Sf ItaUbSIll1brS7fZbkTwdZsq2f6xYgePegxlkU0KDormYiN7OB9u2UEq/zHFTqIV l8RJ7iYCw5I= =g885 -----END PGP SIGNATURE----- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mh@mike.franken.de Sat Aug 16 06:31:00 1997 From: mh@mike.franken.de (Michael Hirmke) Date: Sat, 16 Aug 1997 06:31:00 -0000 Subject: Running winNT/95 programs... References: <3.0.1.32.19970815214435.00910ca0@popd.ix.netcom.com> Message-ID: <6bxWwk64pfB@mike.franken.de> Hi Narayan, > Hi, > > I read a previous thread on running wNT/95 programs through an intermediate > program. I am just a little confused. Why go through all the trouble when > there is a "start" command in both w95 and NT? I have had no problems > running any windows stuff through "start". > > For example: my alias for start on winnt: > > alias start='cmd /c start ' Sorry to ask, this might be a dumb question: Why would one want to use "start" or something like that ? Why not just run the desired application form the command line as usual ? If I want to run explorer, I just type "explorer" from the bash command line and it comes up as usual. [...] > I also have vi (elvis) aliased as: > alias vi="start vi" > so that it comes up in a separate window. Again - why ?? Bye. Michael. -- Michael Hirmke | Telefon +49 (911) 557999 Georg-Strobel-Strasse 81 | FAX +49 (911) 557664 90489 Nuernberg | E-Mail mh@mike.franken.de | WWW http://minimike.franken.de/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sat Aug 16 06:31:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sat, 16 Aug 1997 06:31:00 -0000 Subject: How to avoid GNU make case sensitivity? Message-ID: <01BCA982.087C55A0@gater.krystalbank.msk.ru> Richard Stanton wrote: > I have lots of C files on a FAT (DOS style) formatted disk. Now I'm > running NT, and when I try to use GNU make to compile stuff, it > doesn't recognize the files' existence, since NT thinks all the > filenames are upper case. For example, if I type > > make test.o > > I get an error message about not having a rule to make test.o, when a > directory listing shows the file TEST.C existing on the > drive. Renaming it manually to test.c allows me to compile it OK, but > I'd rather avoid manually renaming every .C file on my system one at a > time. #!/bin/sh for i in *; do new=`echo $i|tr [A-Z] [a-z]`; mv $i $new done -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From lfm@pgroup.com Sat Aug 16 06:31:00 1997 From: lfm@pgroup.com (Larry Meadows) Date: Sat, 16 Aug 1997 06:31:00 -0000 Subject: Timers and less (fwd) Message-ID: <199708151628.JAA17798@pacific.pgroup.com> The following code works well if you are using a pentium, ppro, or pII; it is an interface to the on-chip cycle counter: .file "timer86a.s" .text .align 16 .globl _iclock _iclock: .byte 0x0f, 0x31 ret In C, declare iclock as a long long int, it returns a 64-bit integer that has resolution equal to the clock rate of the chip. The return value is in %eax, %edx (low, high) which matches the C calling convention. If you want time as a double you just divide the return value by the clock frequency: extern long long iclock() time = iclock() / 200.0e6; I guess you can use inline assembler in gcc, I haven't tried. ----- Forwarded message from Mark Koi ----- >From gnu-win32-owner@cygnus.com Fri Aug 15 05:22:20 1997 Message-ID: < UnwpQpFz000140iExo@stoneaxe > Date: Thu, 14 Aug 1997 14:06:29 -0500 (EST) From: Mark Koi To: gnu-win32@cygnus.com Subject: Timers and less Sender: owner-gnu-win32@cygnus.com Precedence: bulk I just installed the binaries. I was using DJGPP before with ALLEGRO for timers. Those allegro library calls don't work, and can't seem to compile allegro with gnu-win32. DOES SOMEBODY HAVE A SIMPLE TIMER PROGRAM? NEED RESOLUTION OF 100ms, not a big request but a little lost. Also, can somebody mail me a uuencoded copy of less.exe for gnu-win32. Or is it just as easy to get source and compile. koi@ssa.crane.navy.mil - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". ----- End of forwarded message from Mark Koi ----- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From richard@lwts.org.uk Sat Aug 16 06:31:00 1997 From: richard@lwts.org.uk (Richard Thorton) Date: Sat, 16 Aug 1997 06:31:00 -0000 Subject: Starter C compiler - Win95 Message-ID: <199708151024.KAA29351@relay.technocom.net> Im looking to find a C compiler to learn C. Is the GNU Win32 compiler a sound one to start with. I havent used any other program apart from basic BASIC at college many years ago. I thought C might be the place to start. Any help appreciated. PS C for Dummies has excited me to learn C Many thanks Richard - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mchase@ix.netcom.com Sat Aug 16 11:19:00 1997 From: mchase@ix.netcom.com (Michael A. Chase) Date: Sat, 16 Aug 1997 11:19:00 -0000 Subject: How to avoid GNU make case sensitivity? Message-ID: <199708161818.NAA17846@dfw-ix9.ix.netcom.com> The following script works under sh, ksh, and bash. One caution, it doesn't deal with spaces in filenames. : # llcase, lucase, lxcase, ulcase, uucase, uxcase, xlcase, xucase, xxcase # Convert file name case # The first character of the script name indicates the base name change # The second character indicates the extension name change # u = force to upper case # l = force to lower case # [^ul] = leave alone upper_name=0 lower_name=0 upper_type=0 lower_type=0 base=`basename $0` case $base in u?*) upper_name=1 ;; l?*) lower_name=1 ;; esac case $base in ?u*) upper_type=1 ;; ?l*) lower_type=1 ;; esac for file in $* do [ ! -f $file ] && continue # skip non-files filepath=`dirname $file` filename=`basename $file` basename=$filename filetype=`echo $filename | sed -ne 's/^.*\(\.[^\.]*\)$/\1/p'` filename=`echo $filename | sed -e 's/^\(.*\)\.[^.]*$/\1/'` [ $upper_name = 1 ] && filename=`echo $filename | tr '[a-z]' '[A-Z]'` [ $lower_name = 1 ] && filename=`echo $filename | tr '[A-Z]' '[a-z]'` [ $upper_type = 1 ] && filetype=`echo $filetype | tr '[a-z]' '[A-Z]'` [ $lower_type = 1 ] && filetype=`echo $filetype | tr '[A-Z]' '[a-z]'` [ "X$basename" != "X${filename}${filetype}" ] && \ mv $file ${filepath}/${filename}${filetype} # skip if name won't change done -----Original Message----- From: Richard Stanton To: gnu-win32@cygnus.com Date: Friday, August 15, 1997 01:08 Subject: How to avoid GNU make case sensitivity? >I have lots of C files on a FAT (DOS style) formatted disk. Now I'm >running NT, and when I try to use GNU make to compile stuff, it >doesn't recognize the files' existence, since NT thinks all the >filenames are upper case. For example, if I type > >make test.o > >I get an error message about not having a rule to make test.o, when a >directory listing shows the file TEST.C existing on the >drive. Renaming it manually to test.c allows me to compile it OK, but >I'd rather avoid manually renaming every .C file on my system one at a >time. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From papresco@technologist.com Sat Aug 16 15:16:00 1997 From: papresco@technologist.com (Paul Prescod) Date: Sat, 16 Aug 1997 15:16:00 -0000 Subject: Starter C compiler - Win95 References: <199708151024.KAA29351@relay.technocom.net> Message-ID: <33F626CA.1966F70C@technologist.com> Richard Thorton wrote: > > Im looking to find a C compiler to learn C. > > Is the GNU Win32 compiler a sound one to start with. I havent used any > other program apart from basic BASIC at college many years ago. I thought C > might be the place to start. With all due respect, I think that Java or JavaScript is a better place to revive your programming skills. C and C++ are the two most difficult languages in common use. Paul Prescod - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ramon@jl1.quim.ucm.es Sat Aug 16 15:44:00 1997 From: ramon@jl1.quim.ucm.es (Ramsn Garcma Fernandez) Date: Sat, 16 Aug 1997 15:44:00 -0000 Subject: Header files. Message-ID: I have seen that Cygwin32 support of Windows API is rather limited in header files. Why don't you take the headers from Willows toolkit, a free emulator of the Windows API under Unix. It is available at http://www.willows.com . It is licenced under GNU LGPL. Regards, Ramon - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From venkman1@ix.netcom.com Sat Aug 16 17:50:00 1997 From: venkman1@ix.netcom.com (Narayan Natarajan) Date: Sat, 16 Aug 1997 17:50:00 -0000 Subject: Running winNT/95 programs... References: <3.0.1.32.19970815214435.00910ca0@popd.ix.netcom.com> <6bxWwk64pfB@mike.franken.de> Message-ID: <3.0.1.32.19970816201809.0091ce20@popd.ix.netcom.com> Hi Mike, >> For example: my alias for start on winnt: >> >> alias start='cmd /c start ' > >Sorry to ask, this might be a dumb question: Why would one want to use >"start" or something like that ? Why not just run the desired >application form the command line as usual ? >If I want to run explorer, I just type "explorer" from the bash command >line and it comes up as usual. The only reason I do this is it lets me have my command line back. "start" sort of backgrounds the task. I guess you are correct when you say run it from the command line for applications such as explorer. Explorer seems to give the command line back but netscape, for instance, does not. >> alias vi="start vi" >> so that it comes up in a separate window. > >Again - why ?? Thaaats why !!! :-) Narayan | Narayan _/~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~/ |Natarajan _/ venkman1@ix.netcom.com http://www.netcom.com/~venkman1 _/ | _/ (810)-475-0667 . .____ _ _/ | _/ "The heavens call to you, and * . x -(____/ | _ / | _/ circle about you, displaying to ________/____/ \ ( )----- | _/ you their eternal splendors, and ,/ _ ( ------- | _/ your eye gazes only to earth." <________\ // (_)----- | _/ -- Dante --<== * |_____/ _ |/ -(_____) (_)----- |~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dmeans@bellsouth.net Sat Aug 16 17:50:00 1997 From: dmeans@bellsouth.net (David Means) Date: Sat, 16 Aug 1997 17:50:00 -0000 Subject: Win95 Compiler References: <2.2.32.19970813180442.00673284@mail.gower.net> Message-ID: <33F63437.22053425@bellsouth.net> James A Leath wrote: > I have Windows 95, and I'm not very knowledgeable about compilers and > the > like, and I need to compile a C file on my computer. I was wondering > if what > was on your Cygnus page had what I needed, and I downloaded the > usertools.exe (4.4 megs) and I'm not sure if this is what I needed. I > notice > the developers kit needed a compiler to compile it, and I thought that > was > odd since I don't have a compiler to compile it. Please help me. > > Thanks, :) > > James A. Leath > > phoenix@gower.net > http://www.gower.net/phoenix/index.html > > - > For help on using this list (especially unsubscribing), send a message > to > "gnu-win32-request@cygnus.com" with one line of text: "help". Get cdk.exe from ftp://ftp.cygnus.com/pub/gnu-win32/gnu-win32-b18/ -- David Means mailto:dmeans@bellsouth.net -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/octet-stream Size: 4241 bytes Desc: not available URL: From dmeans@bellsouth.net Sat Aug 16 17:50:00 1997 From: dmeans@bellsouth.net (David Means) Date: Sat, 16 Aug 1997 17:50:00 -0000 Subject: A beginner's query References: <7790037C47@fs2.mt.umist.ac.uk> Message-ID: <33F6329B.B7906AB6@bellsouth.net> Anthony.Appleyard wrote: > (0) About how many messages a day are there on gnu-win32? If recollection serves right, anywhere from a low of about 10 or so, to about a 100, just depends on issues being discussed, and the convictions of the readers. ;-) > (1) I looked at the ..../gnu-win32/ FTP directory on Sunsite in UK, > and I > found a confused tree with no apparent index file telling me what is > what. > What should I load? cdk.exe Check out ftp://ftp.cygnus.com/pub/gnu-win32/gnu-win32-b18/ > (2) It would be useful if these files were packed by .zip instead of > by .gz Winzip (at least version 6.2) will unzip and un-tar unix archives. > (3) How well advanced and debugged is gnu-win32 compared with djgpp > C/C++ > for DOS? Hum... can't help you on that one! :-) > - > For help on using this list (especially unsubscribing), send a message > to > "gnu-win32-request@cygnus.com" with one line of text: "help". Have a good day! David dmeans@bellsouth.net -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/octet-stream Size: 4241 bytes Desc: not available URL: From warnstam@usa.net Sat Aug 16 17:50:00 1997 From: warnstam@usa.net (Jan Warnstam) Date: Sat, 16 Aug 1997 17:50:00 -0000 Subject: wish42 - console trouble!!! Message-ID: <33F64668.D6E@usa.net> I just installed the package (cdk.exe) and now I try to run the wish42 executable. It opens a shell window that looks terrific, but when I try to type 'ls' it just says: channel "console1" wasn't opened for writing What is the error? Have I done something wrong, or is this a bug? Please help me, as I do not seem to solve this problem on my own! -- [----------------------------------------------------] Jan Warnstam * SWEDEN * E-mail: warnstam@usa.net Homepage: http://home1.swipnet.se/~w-19502/index.htm [----------------------------------------------------] - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From cxh@eecs.berkeley.edu Sat Aug 16 17:50:00 1997 From: cxh@eecs.berkeley.edu (Christopher Hylands) Date: Sat, 16 Aug 1997 17:50:00 -0000 Subject: glimpse and ispell Windows (cygwin b18) binaries available Message-ID: <199708162325.QAA15218@kahn.eecs.berkeley.edu> I've created glimpse and ispell binaries that run under NT4.0. These binaries require Cygnus GNU-win b18, available at: http://www.cygnus.com/misc/gnu-win32/ The glimpse and ispell binaries and sources are available at: http://ptolemy.eecs.berkeley.edu/tycho/tychoTools/tychoTools.html Ispell and glimpse are used by Tycho, the Itcl graphical development environment. For more information, see http://ptolemy.eecs.berkeley.edu/tycho The only purpose of the Ispell port is to allow spell checking from within the Tycho editor under NT with the ispell -u option. If you are looking for a complete ispell4.0 for windows, try http://www.itribe.net/virtunix/files/ispell-4.0-nt.zip The reason that we are not using that version is because under NT, the -u option causes a core dump.her win The glimpse port seems to work ok, except glimpseindex produces messages about whereis and sync. Merging the glimpse changes into the glimpse sources is doable, but I have no plans to do it myself. Both of these ports are a bit of a rush job, but I needed the functionality under NT, so I hacked up the binaries. If the Glimpse and Cygwin maintainers want to setup links to http://ptolemy.eecs.berkeley.edu/tycho/tychoTools/tychoTools.html or copy the files to their local websites, then that's fine by me. -Christopher Christopher Hylands, Ptolemy Project Manager University of California cxh@eecs.berkeley.edu US Mail: 558 Cory Hall #1770 ph: (510)643-9841 fax:(510)642-2739 Berkeley, CA 94720-1770 home: (510)526-4010 (if busy -4068) (Office: 493 Cory) - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From grapeape@scott.net Sat Aug 16 20:11:00 1997 From: grapeape@scott.net (Chris Green) Date: Sat, 16 Aug 1997 20:11:00 -0000 Subject: wish42 - console trouble!!! References: <33F64668.D6E@usa.net> Message-ID: wish4.2 is the tcl/tk command intrepreter shell. Its not a basic command shell. What this really means : You write or run tcl/tk programs using wish instead of directly interacting with it. bash is the command shell. Chris On 17-Aug-97 Jan Warnstam wrote: >I just installed the package (cdk.exe) and now I try to >run the >wish42 executable. It opens a shell window that looks >terrific, >but when I try to type 'ls' it just says: > >channel "console1" wasn't opened for writing > >What is the error? Have I done something wrong, or is this >a bug? >Please help me, as I do not seem to solve this problem on >my own! > >-- >[----------------------------------------------------] > Jan Warnstam * SWEDEN * E-mail: warnstam@usa.net > Homepage: http://home1.swipnet.se/~w-19502/index.htm >[----------------------------------------------------] > > >- >For help on using this list (especially unsubscribing), >send a message to >"gnu-win32-request@cygnus.com" with one line of text: >"help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From room_cygnus@uncnsrd.mt-kisco.ny.us Sat Aug 16 20:30:00 1997 From: room_cygnus@uncnsrd.mt-kisco.ny.us (IGnatius T Foobar) Date: Sat, 16 Aug 1997 20:30:00 -0000 Subject: No Subject Message-ID: <336132@uncnsrd.mt-kisco.ny.us> I noticed that a lot of people are talking about "Sergey's telnetd". Someone please tell me more about this program. I assume it's been configured to allow inbound telnet to either bash or some other program running on an NT box? I would also presume that someone has an inetd of some sort working? Please direct me to these programs, if they are available. I would also like to know how one goes about configuring a program like inetd to automatically start up when NT is booted ... in essence, to set it up as a "service", I guess. Thanks in advance! - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mphalpin@iag.net Sat Aug 16 20:49:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Sat, 16 Aug 1997 20:49:00 -0000 Subject: Problem with select() Message-ID: <199708170336.XAA14550@iag.net> This is going to sound weird, but I have apps that seemed to run fine before. But when I recompiled them (all of them), select() no longer worked!!! It would return 0!!! Why?????? What's wrong here?? - SJ ---------- > From: Sergey Okhapkin > To: 'Paul J. Metzger' > Cc: 'gnu-win32' > Subject: RE: GNU Win32 nonblocking I/O > Date: Saturday, June 28, 1997 5:02 AM > > Paul J. Metzger wrote: > > Hello, gentlemen. We've just recently started some development > > under GNU Win32 and I'm having a problem with non-blocking I/O > > of a console window under NT. I looked at the mailing list archives, > > Nonblocking I/O is supported now for sockets only. I'm working now on > redesign of tty support in cygwin.dll (including job control features). New > tty subsytem will support nonblocking I/O too. > > BTW, with my current (really buggy :-) cygwin.dll it is possible now to > start several bash sessions and run something like "ls >/dev/tty1" or "cat > /dev/tty1" :-) > > -- > Sergey Okhapkin, http://www.lexa.ru/sos > Moscow, Russia > Looking for a job > > - > For help on using this list (especially unsubscribing), send a message to > "gnu-win32-request@cygnus.com" with one line of text: "help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From noer@cygnus.com Sat Aug 16 23:29:00 1997 From: noer@cygnus.com (Geoffrey Noer) Date: Sat, 16 Aug 1997 23:29:00 -0000 Subject: Beta 19 status Message-ID: <199708170629.XAA01336@rtl.cygnus.com> We just imported bash 2.01 into our local sources so it will be included in beta 19. I haven't seen any problems with it yet and having working job control is great. I've appended the beginnings of the history doc with changes since b18. It isn't all there yet, but for those of you who are curious... Also my current best guess is that beta 19 will be out in September. A fair amount of work is being done on the Win32 side of things right now (converting gcc to default to not linking in the cygwin.dll, handling .exe extensions better during the build process, etc...). So beta 19 will happen as soon as everything settles down a little. -- Geoffrey Noer noer@cygnus.com -------------------------------------- @item Release Beta 19 (Probably September 1997) This is a major release. The new cygwin.dll is still backwards-compatible with previously linked applications but contains significant changes. The PE format definition now matches Microsoft's more closely. This should allow better interoperability with other vendors' development tools. This change invalidates all previously built object (.o) and static library (.a) files so be sure to delete/rebuild old .o and .a files you are using! @subsection Changes in specific tools: We have upgraded some of the utilities. GNU-Win32 now includes bash 2.01, textutils 1.22, and gawk 3.0.3. Bash under Cygwin32 now includes working job control among other improvements. Ian Lance Taylor wrote a resource compiler called "windres". It can be used to compile windows resources from a textual rc file into a COFF file. The sources for it are part of the binutils package. @subsection Changes in the Cygwin32 API (cygwin.dll): Old symlinks are also invalidated by this release. The "system" attribute is now used to mark symlinks which significantly speeds up fstat and other file related calls. Either recreate old ones or set their "system" attribute flag so they will be recognized properly. The following is now supported: * pseudo-ttys The following calls have been added: * ptsname, grantpt, unlockpt * login, logout, ttyslot * wait3, wait4 * setitimer, getitimer * cfg(s)eti(o)speed * initgroups (stub), getgroups * sexecve, sexecl, sexecle, sexeclp, sexeclpe, sexecv, sexecp, sexecvpe * pause * getpgid, killpg * strlwr, strupr * ftime * fpathconf The bsearch call has been rewritten. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fjh@cs.mu.OZ.AU Sun Aug 17 02:19:00 1997 From: fjh@cs.mu.OZ.AU (Fergus Henderson) Date: Sun, 17 Aug 1997 02:19:00 -0000 Subject: Handling of `.exe' files (was: Beta 19 status) References: <199708170629.XAA01336@rtl.cygnus.com> Message-ID: <199708170919.TAA32085@murlibobo.cs.mu.OZ.AU> Geoffrey Noer writes: >A fair amount of work is being done on the Win32 side of things right now >(converting gcc to default to not linking in the cygwin.dll, handling >.exe extensions better during the build process, etc...). How do you plan to handle .exe extensions? -- Fergus Henderson | "I have always known that the pursuit WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pyeh@jungfrau.usc.edu Sun Aug 17 02:19:00 1997 From: pyeh@jungfrau.usc.edu (Peter Yeh) Date: Sun, 17 Aug 1997 02:19:00 -0000 Subject: OpenNT question again Message-ID: <199708170902.CAA05441@jungfrau.usc.edu> Hi All, Last month, I asked about the comparison between OpenNT by Softway and GNU-Win32 and thank you I received many useful replies. Right now, I would like to ask those of you have OpenNT to help me once more. Softway has post their version of GCC on their web page. It seems that it is related to Cygnus. According to their web page, you don't need MSVC any more to compile programs. Thus, I hope some of you are kind enough to install the package and do a honest comparison between OpenNT with GCC and GNU-Win32. My main considerations are compatibility/portability (from Unix programs) and performance (compilation and execution). Thank you in advance for your help. --Peter - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From tage.westlund@stockholm.mail.telia.com Sun Aug 17 05:12:00 1997 From: tage.westlund@stockholm.mail.telia.com (Tage Westlund) Date: Sun, 17 Aug 1997 05:12:00 -0000 Subject: Gnuwin32 b18 "make from source" bugs and fixes Message-ID: <33F70434.3145@stockholm.mail.telia.com> To users and designers! SOME REASONS WHY "make" gnuwin32 FROM SOURCE STOPS. =============================================================== I have spent some time building a gnuwin32 b18 system for win95 from source code. This was downloaded on july 1:st from: ftp://sunsite.fri.uni-lj.si/pub/gnu-win32/gnu-win32-b18/cdk-src/* in pieces xca - xcj. My goal is to keep three versions of gnuwin32 available: b17.1 in directory /cygnus/H-i386-cygwin32/bin b18 from binaries xaa - xah in /gnuwin32/b18/H-i386-cygwin32/bin b18 from source and with needed patches in /usr/local/bin I simply intend to change autoexec.bat to change the version. During system generation autoexec pointed at the binary b18. I first downloaded the binary b18 according to the README.txt =============================================================== I have not yet run any imported patches. Maybe I should have done that? I have not yet run the user-tools "make". Unwise? I will now try to describe the errors that stopped the "bash$ make" and "bash$ make install" processes and those more or less deterministic hacks that brougt me to the full installa- tion. Since this mail is too voluminous, I will first give the most unique error text for every "make" stop so you can decide if the rest is of any interest. The patch details follow later. I am rather convinced that parts of the cases 1,8,10,11, and 12 contain ***TRUE BUGS***. The rest could be due to my lack of experience and of more complete README texts. =============================================================== Executing "bash$ make" I got error stops 1-7 1. make: DLLTOOL@: Command not found make: *** [libtcl7.6.a] Error 127 make: *** [all] Error 2 make: *** [all-tcl] Error 2 bash$ ( make exits ) Code that was adjusted to avoid stop at next "make" restart: .../cdk/tcl/win/Makefile .../cdk/tk/win/Makefile ( below I use the following shorthand for the text above: make: DLLTOOL@: Command not found *** libtcl7.6.a all all-tcl 127 2 2 *** Code: .../cdk/tcl/win/Makefile .../cdk/tk/win/Makefile ) 2. In file included from ./gdbtk.c:31: .../cdk/tk/generic/tk.h:72: X11/Xlib.h: No such file or dir. gcc: Internal compiler error: program cpp got fatal signal 33 *** gdbtk.o all-gdb 1 2 *** Code: .../cdk/tk/generic/tk.h .../cdk/tk/xlib/X11/Xlib.h 3. At gcc ... -o gdb linking ld.exe: cannot open -lX11: No such file or dir. gcc: Internal compiler error: program got fatal signal 1 Code: .../cdk/gdb/Makefile 4. At gcc ... -o gperf linking ld: cannot open crt0.o: No such file or dir. xgcc: Internal compiler error: program ld got fatal signal 1 *** gperf src-all subdir-do all all-target-libg++ 1 2 1 2 2 *** Code: .../cdk/libg++/gperf/src/crt0.o 5. At gcc ... -o gperf linking ld: cannot open -lm: No such file or dir. xgcc: Internal compiler error: program ld got fatal signal 1 *** gperf src-all subdir-do all all-target-libg++ 1 2 1 2 2 *** Code: .../cdk/libg++/gperf/src/Makefile 6. At gcc ... -o gperf linking ld; cannot open -lcygwin: No such file or dir. Code: .../cdk/libg++/gperf/src/Makefile .../cdk/newlib/libcygwin.a 7. At gcc ... -o gperf linking ld; cannot open -lkernel32: No such file or dir. Code: .../cdk/libg++/gperf/src/Makefile .../cdk/newlib/libkernel32.a ============================================================ Executing "bash$ make install" I got error stops 8-12 8. Making directory /lib Installing libtcl7.6.a ranlib: libtcl7.6.a: No such file or dir. *** install-binaries install install-tcl 9 2 2 *** Code: .../cdk/tcl/win/Makefile.in and Makefile .../cdk/tk/win/Makefile.in and Makefile 9. Installing libtk4.2.a make: ./install-sh: Command not found *** install-binaries install install-tk 127 2 2 *** Code: .../cdk/tk/win/install-sh 10. Installing ./library/../prolog.ps cp: ./library/../prolog.ps: No such file or dir. Installing doc/*.1 /bin/sh: /usr/local/man/man1/*.1: No such file or dir. chmod: /usr/local/man/man1/*.1: No such file or dir. *** install-man install install-tk 1 2 2 *** Code: .../cdk/tk/win/Makefile 11. cd /usr/local/bin && ln flex flex++ (C:\BIN\SH.EXE 15076) In cygwin_except_handler The system hangs. One has to make power off and on. Code: .../cdk/flex/Makefile 12. ( cd include; \ tar -cf - .; exit 0 ) | ( cd /usr/local/lib/gcc-lib/ i386-pc-cygwin32/cygnus-2.7.2-970404/include ; tar xpBf - ) (no new lines in original) tar: Could not change access and modification time of objc: Permission denied tar: Could not change access and modification time of .: Permission denied (C:\BIN\SH.EXE 60373) In cygwin_except_handler (C:\BIN\SH.EXE 60373) Exception trapped - - - - the usual stack tracing - - - - *** install-headers-tar install-gcc 2 2 *** Code: .../gcc/Makefile ============================================================== WORKAROUNDS for 1-12. Some are obvious minor bugs, some are only corrections to secondary errors where I have no deep under- standing, some are due to tunneling effects, cosmic radiation or what? 1. Thanks to Benjamin Sean Scarlet I found in .../cdk/tcl/win/Makefile and .../cdk/tk/win/Makefile DLLTOOL = @DLLTOOL@ ( configure has left this line un- changed from Makefile.in ) DLLTOOL = dlltool <=== 1 is the patch in .../cdk/tk/win/Makefile there is also: COMPAT_OBJS = @LIBOBJS@ but it is unused 2. I modified the source. Probably not the utmost method. in .../cdk/tk/generic/tk.h line 72: #include altered: #include <../tk/xlib/X11/Xlib.h> <=== 1 in .../cdk/tk/xlib/X11/Xlib.h lines 35 and 42 #include altered: #include <../tk/xlib/X11/X.h> <=== 2 #include altered: #include <../tk/xlib/X11/Xfuncproto.h> <=== 3 3. In .../cdk/gdb/Makefile two variables INSTALLED_LIBS= and CLIBS= both containing the line: $(TERMCAP) $(XM_CLIBS) $(TM_CLIBS) $(NAT_CLIBS) -lm \ $(TK) $(TCL) $(X11_LDFLAGS) $(X11_LIBS) \ -lX11 \ altered: $(TERMCAP) $(XM_CLIBS) $(TM_CLIBS) $(NAT_CLIBS) -lm \ $(TK) $(TCL) $(X11_LDFLAGS) $(X11_LIBS) \ -L/.../cdk/tk/xlib -lX11 <=== 1 + 2 ( .../ depicts for instance in my case /gnuwin32/b18src ) ( original: no "new lines" in $(TERMCAP) above ) 4. I copied the files .../cdk/newlib/crt0.c and crt0.o to .../cdk/libg++/gperf/src/crt0.c and crt0.o <=== 1 + 2 5-7. In .../cdk/libg++/gperf/src/Makefile at the end: $(TARGETPROG): $(OBJECTS) $(CXX) $(LDFLAGS) -o $@ $(OBJECTS) $(LIBS) -lm was altered to: $(TARGETPROG): $(OBJECTS) <=== 1 $(CXX) $(LDFLAGS) -o $@ $(OBJECTS) $(LIBS) \ -L/gnuwin32/b18src/cdk/newlib -lm -lcygwin -lkernel32 From .../b18/H-i386-cygwin32/i386-cygwin32/lib/ directory the files libcygwin.a and libkernel32.a were copied to .../cdk/newlib/ directory <=== 2 + 3 ( .../b18 denotes the binary release directories ) 8. I found a really hidden error in: .../cdk/tcl/win/Makefile.in It showed up that in the line: exec_prefix = @exec_prefix@ there was an extra blank after the word "prefix@". This was transformed to an extra blank after the word "local " in Makefile. This extra blank had to be removed. <=== 1 Exactly the same defect is seen in: .../cdk/tk/win/Makefile.in and in the corresponding .../cdk/tk/win/Makefile <=== 2 9. The script install-sh had to be brought also to the directory .../cdk/th/win/ for instance from .../cdk/tcl/win/install-sh <=== 1 10. In .../cdk/tk/win/Makefile the line for i in $(SRC_DIR)/../library/*.tcl $(SRC_DIR)/../library/*.gif $(SRC_DIR)/../library/*.xbm $(SRC_DIR)/../library/tclIndex $(SRC_DIR)/library/../prolog.ps $(SRC_DIR)/winMain.c; \ ( original: no new line ) should be spelt: <=== 1 $(SRC_DIR)/../library/prolog.ps $(SRC_DIR)/winMain.c; \ On three locations in the same file at install-man: cd $(SRC_DIR)/doc; for i in *.1; altered: cd $(SRC_DIR)/../doc; for i in *.1; \ <=== 2 + 3 + 4 11. In .../cdk/flex/Makefile the line -cd $(libdir) && $(RANLIB) $(FLEXLIB) is altered: cd $(libdir) && $(RANLIB) $(FLEXLIB) <=== 1 Reference to an older release: INSTALL = /CYGNUS/H-I386-CYGWIN32/BIN/install -c is altered: INSTALL = .../cdk/install-sh -c <=== 2 Setting the link: cd $(bindir) && ln $(FLEX) $(FLEX)++ gives the hang of the system. Alter to: cd $(bindir) <=== 3 But then the link has to be done manually. There was no problem. Or simply cp flex.exe flex++.exe <=== 4 12. The problem is when copying the contents of: .../cdk/gcc/include as a part of the installation to: /usr/local/lib/gcc-lib/i386-pc-cygwin32/ (no new line) cygnus-2.7.2-970404/include Because of Permission denied error I had to discard the installation method of piping between two tar processes: # (cd include; \ <=== 1 # tar -cf - .; exit 0) | (cd $(libsubdir)/include; \ # tar $(TAROUTOPTS) - ) Therefore I had to copy between the "include" directories manually. <=== 2 ============================================================== main(){printf("Hello World\n\"Little keystrokes fell great oaks\"\n");} Tage - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mh@mike.franken.de Sun Aug 17 07:17:00 1997 From: mh@mike.franken.de (Michael Hirmke) Date: Sun, 17 Aug 1997 07:17:00 -0000 Subject: References: <336132@uncnsrd.mt-kisco.ny.us> Message-ID: <6c0jIJvKpfB@mike.franken.de> Hi IGnatius, [...] > Please direct me to these programs, if they are available. > http://miracle.geol.msu.ru/sos/ > I would also like to know how one goes about configuring a program like > inetd to automatically start up when NT is booted ... in essence, to set > it up as a "service", I guess. See description on that page. > > Thanks in advance! Bye. Michael. -- Michael Hirmke | Telefon +49 (911) 557999 Georg-Strobel-Strasse 81 | FAX +49 (911) 557664 90489 Nuernberg | E-Mail mh@mike.franken.de | WWW http://minimike.franken.de/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sun Aug 17 09:48:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sun, 17 Aug 1997 09:48:00 -0000 Subject: Beta 19 status Message-ID: <01BCAB4F.01D65100@sos> Geoffrey Noer wrote: > .exe extensions better during the build process, etc...). So beta 19 > will happen as soon as everything settles down a little. Signals handling is incompatible with Windows95 :-( The only way I know for a now to run cygnus tools stable on W95 is to disable all the signal (and job control) support. Any ideas/suggestions? window.cc: void window_init () { DWORD tid; HANDLE hThread; u->self->hwnd = NULL; > if (get_os_type () == win95) > return; window_started = CreateEvent (NULL, TRUE, FALSE, NULL); hThread = CreateThread (NULL, 0, (LPTHREAD_START_ROUTINE) Winmain, NULL, 0, &tid); Sorry, I have no ideas how to avoid random crashes with the diagnostics like "sh.exe causes the general protection error in module USER.EXE"... sigh :-) -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mphalpin@iag.net Sun Aug 17 10:15:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Sun, 17 Aug 1997 10:15:00 -0000 Subject: cygwin.dll Source Code Message-ID: <199708171707.NAA08145@iag.net> Does anyone have the cygwin.dll source with all the patches installed? If so, please tell me to obtain it. Thanks. - SJ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fine@mail.cern.ch Sun Aug 17 12:49:00 1997 From: fine@mail.cern.ch (Valery Fine) Date: Sun, 17 Aug 1997 12:49:00 -0000 Subject: Starter C compiler - Win95 References: <199708151024.KAA29351@relay.technocom.net> Message-ID: <199708171948.VAA137866@sp068.cern.ch> On 15 Aug 97 at 10:18, Richard Thorton wrote: > Im looking to find a C compiler to learn C. > > Is the GNU Win32 compiler a sound one to start with. I havent used > any other program apart from basic BASIC at college many years ago. > I thought C might be the place to start. > > Any help appreciated. > Try ROOT/Cint package. It includes C/C++ interpreter, plenty of graphics and Network facilities. It does make C/C++ as easy as BASIC. You welcome to visit our Web site http://root.cern.ch I've attached an abstract of CINT for your info too. With my best regards, Valery ============ Attachment follows ======================= Date sent: Sat, 09 Aug 1997 13:59:00 +0900 From: Subject: CINT book abstract To: roottalk@root.cern.ch Here is a crude translation of my CINT book's preface. Maaharu Goto ======================================================================== PREFACE What is CINT? Computer software is used everywhere and its' importance is decisively increasing. Not only products like Personal Computers and Workstations, which are evidently sold as computers, many of the electronic devices we use in our life are controlled by microprocessors and software operating on them. Application spreads all around scientific and engineering field as well as economy, management, etc... Software is already a fundamental component in today's society. With the increasing importance of software, productivity of its' development becomes an important issue. As software is used for wider range of application, there are increasing number of occasions that software professional has to work with expert in other professional area. Excellent results can only be obtained by making the best use of both software and non-software expertise. Hence, a good programming environment to support cooperative work between software and non-software professionals is demanded. (Fig 0.1 software and non-software professionals working together) Thanks to price decrease and proliferation of personal computers, writing a small computer program is not a special skill anymore. Most of the non-software expert can write some sort of computer program. It is feasible enough that non-software professional communicates with software professional directly using computer language. It is especially important to privide such environment for C/C++ because it is the most popular programming language today. # C++ is made for the serious programmer In "The C++ Programming Language Second Edition", Dr. Bjarn Stroustrup states that "C++ is a general purpose programming language designed to make programming more enjoyable for the serious programmer." As represented in his words, existing C++ development environment focuses on professional programmer who develops serious application. Such environment is often inadequate for non-software professional who writes program only part-time. Complexity of C++ is an obstacle for a beginner, however, there are number of good books about learning C++ so that non-software professinal can learn C/C++ to his needs. What bothers us in a real-world situation is a compiler/linker problem. Long turn-around-time disturbs creative thinking. You may spend a couple of days just digging out unresolved symbols in a huge software library. It is not desirable for non-software professional to spend time on such thing. (Fig 0.2 Compiler vs Interpreter) # We need an everybody-can-use interpreter BASIC interpreter has been widely accepted by computer beginners. It is very nice if we have good interpreter environment for C/C++ too. Ease of use and short turn-around-time of the interpreter is highly appreciated by the non-software professional who makes many experimental changes in a small program. I needed C++ interpreter myself in 1990. C/C++ interpreter would have perfectly fit to my need as a bridge between software and non-software professional. There was a commercial C/C++ interpreter, but the license fee was expensive and it had substantial limitation. We also tried script languages such as Perl, but nothing fulfilled our need. I wanted to avoid making C/C++ interpreter by myself. But, without having alternative, I finally decided to undertake development of CINT C++ interpreter. The first version is created in March 1991. We immediately applied it to a running project, where we saw significant improvement on productivity and creativity. After many enhancments, in February 1996, I started to expose CINT as a free software from CERN, The European High Energy Physics Laboratory. # CINT combines benefit of interpreter and compiler CINT consists of "cint" , the C/C++ interpreter, and "makecint" , interpreter-compiler. "cint" can be used standalone as a prototyping tool or for educational purposes. But, the true merit can be obtained by using "makecint". Makecint encapsurates arbitrary C/C++ object into the interpreter as precompiled library. In existing script language environment, script and compiled code have to be written in different language and interface method between them has to be manually defined. CINT automatically makes interface method so that the interpreter and precompiled library can access each other. It provides a single language solution combining benefit of interpreter and compiler. A script which is originally developped for interpreter can later be compiled and embedded into the interpreter. Or precompiled library can temporarily be interpreted. Precompiled library can be linked together with the interpreter as archived library, or can be configured as DLL(Dynamic Link Library) which can be loaded at run time. With this feature, huge library written by software professional can be embedded into the interpreter and used easily by non-software professional. (Fig 0.3 Concept of cint/makecint) CINT operates with arbitrary ANSI C/C++ compiler. It currently works on HP-UX, Linux, Solaris, SunOS, AIX, IRIX, ALPHA-OSF, FreeBSD, NetBSD, BeBox, Windows-NT/95, MS-DOS, MacOS, Convex, etc... If you feel C++ is too difficult, you can use CINT with C compiler only. CINT has other features. Interpreter should inherently be a dynamic environment. C/C++ is defined as a compiler language, however, CINT extends its' capability without changing language definition. Source code and DLL can be loaded/unloaded at run time. C/C++ statement can be evaluated from command prompt. Security mode is introduced as an option to improve robustness issue such as memory leak and access through bad pointer. CINT is already used in many research institutes and companies. User community is rapidly growing. # How to use this book I tried to explain everything about CINT in this book. Detail usage and advanced technique will be described with many examples. Reference of all special commands and APIs are descrived so that you can make the best use of CINT. Basic information: Chapter 1. Getting started: cint, the interpreter Chapter 2. Concept and background Chapter 3. Getting started: makecint, the interpreter-compiler Advanced information: Chapter 4. Cint advanced information Chapter 5. Makecint advanced information Chapter 6. Concept of C++ script Chapter 7. Example of real-world application Reference: Appendix1. CINT keyword reference Appendix2. cint command line option Appendix3. makecint command line option Appendix4. cint debugger interface command Appendix5. Syntax limitation The first part explains basic information of using CINT. Chapter 1 describes basic information about cint interpreter Chapter 2 explains feature and concept behind it. Chapter 3 describes basic information about makecint interpreter-compiler Second part explains advanced information. Chapter 4 and 5 describes advanced feature and usage of cint and makecint. Chapter 6 explains concept of C++ script which is an idea of easy-to-use class library. Chapter 7 describes how CINT is used in real-world applications. ROOT/CINT, which is jointly developped with CERN, is a remarkable achievement. It is a new generation Object Oriented Framework. You don't need to read through every chapter. Please use "Readers hint" at the beginning of each chapter as a guide to get information quickly. # Acknowledgement I'd like to say thank you to all people who supported CINT and my activity. Junichi Mizoguchi in Hewlett-Packard Japan invented basic concept of makecint. Shirou Tsuruda has deep understanding about advanced technology and promoted CINT as a standarized productivity tool in Hewlett-Packard Japan Hachioji division. Tatsuya Saito and Hiroshi Nada gave me chances to publicly introduce CINT. Dr. Fons Rademakers, Dr. Rene Brun, Dr. Valery Fine and Dr. Andres Sandoval in CERN made great contribution to the quality and evolution of CINT. Osamu Kotanigawa in Fuji Xerox, Mr. Tomina in Nihon-Denshi-Keisan, and Dr. Otofuji in Akita University contributed porting of CINT to various computer platforms. Kiyoshi Yamamoto in CQ publishing gave me opportunity to write this book and articles in Interface magazine. Spring 1997 Masaharu Goto (gotom@jpn.hp.com , MXJ02154@niftyserve.or.jp ) ================================================================= Dr. Valery Fine Telex : 911621 dubna su ----------- LCTA/Joint Inst.for NuclearRes Phone : +7 09621 6 40 80 141980 Dubna, Moscow region Fax : +7 09621 6 51 45 Russia mailto:fine@main1.jinr.dubna.su Dr. Valeri Faine ------------ Phone: +41 22 767 6468 CERN FAX : +41 22 767 7910 CH-1211 Geneva, 23 mailto:fine@mail.cern.ch Switzerland http://nicewww.cern.ch/~fine - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mphalpin@iag.net Sun Aug 17 13:53:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Sun, 17 Aug 1997 13:53:00 -0000 Subject: Lynx Problem Message-ID: <199708172045.QAA29110@iag.net> Has anyone else had this problem? Lynx appears to start fine, but then I get "Alert! Unable to connect to remote host" Any idea? I got it from the URL Sergey gave. - SJ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From koeritz@twain.com Sun Aug 17 13:53:00 1997 From: koeritz@twain.com (Christopher A. Koeritz) Date: Sun, 17 Aug 1997 13:53:00 -0000 Subject: regarding cygnus win32 gnu tools... Message-ID: <33F76471.87CB7363@twain.com> THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) THANK YOU (thank you) having found them, i have not tried them, and yet i have the highest hopes. -- ___________________ chosen by the Nechung Oracle program... ___________________ After finding no qualified candidates for the position of principal, the school board is extremely pleased to announce the appointment of David Steele to the post. -- Philip Streifer, Superintendent of Schools, Barrington, Rhode Island ______________ not necessarily my opinions, not necessarily not. ______________ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Robert_Seeger@baynetworks.com Sun Aug 17 14:56:00 1997 From: Robert_Seeger@baynetworks.com (Robert Seeger) Date: Sun, 17 Aug 1997 14:56:00 -0000 Subject: Tcl/Tk with extensions under cyg-win32? Message-ID: <3.0.32.19970817175328.009554b0@bl-mail2.corpeast.baynetworks.com> I was hoping somebody has had some experience with what I want to do. I'd like to compile Tk/Tcl 4.2/7.6 with several extensions, including itcl 2.2, TclX, Expect, and Scotty (Tnm). I'd like to compile them as shared libraries/loadable packages, if that is possible under cyg-win32. If not, I suppose I can just link them in statically, it's just that I'd rather not. I was wondering if anyone had any experience in doing this and, if so, what problems were run into. How much had to be changed to compile under cyg-win32? What target opsys did you use, linux? What other things should I be concerned about when I go about doing this. Thanks for any help, Robert Seeger ======================================== Robert Seeger Network Engineer Bay Networks Telephone: (518)237-2087 Pager: (800)SKY-8888 Pin#1264792 Fax: (518)237-4190 Email: rseeger@baynetworks.com Address: 224 5th Ave, Apt#2 Lansingburgh, NY 12182 ======================================== - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Scotbrad1@aol.com Sun Aug 17 16:26:00 1997 From: Scotbrad1@aol.com (Scotbrad1@aol.com) Date: Sun, 17 Aug 1997 16:26:00 -0000 Subject: Question about GCC & G++ Message-ID: <970817192530_-968740337@emout17.mail.aol.com> I just downloaded GNU-Win32 yesterday and the whole virtual UNIX is very confusing to me. All I want to use are G++ (C++?) and GCC (C?), and I was wondering, basically, how do I compile programs with these? Do I make the source in a text file with a certain extension? I have only programmed in DOS C++ so I have never used UNIX compilers before. Plus, I was reading the FAQ and saw something about V. I looked through the GNU tools to find this but could not so I went to the V web page and downloaded it since I want to try programming in a Visual C++ type of environment and to try making windows applications. The install directions, though, said nothing about how to set it up under GNU-Win32. How do I set up V with GNU-Win32? Thanks, Scott Bradley - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ian@cygnus.com Sun Aug 17 16:54:00 1997 From: ian@cygnus.com (ian@cygnus.com) Date: Sun, 17 Aug 1997 16:54:00 -0000 Subject: Tcl/Tk with extensions under cyg-win32? References: <3.0.32.19970817175328.009554b0@bl-mail2.corpeast.baynetworks.com> Message-ID: <199708172321.TAA03630@tweedledumb.cygnus.com> >I was hoping somebody has had some experience with what I want to do. I'd >like to compile Tk/Tcl 4.2/7.6 with several extensions, including itcl 2.2, >TclX, Expect, and Scotty (Tnm). I'd like to compile them as shared >libraries/loadable packages, if that is possible under cyg-win32. If not, I >suppose I can just link them in statically, it's just that I'd rather not. >I was wondering if anyone had any experience in doing this and, if so, what >problems were run into. How much had to be changed to compile under >cyg-win32? What target opsys did you use, linux? What other things should I >be concerned about when I go about doing this. I was able to get Tcl, Tk, Itcl and Tix working. Probably the Tcl and Tk work, at least, will be in the beta 19 release. I started with the makefile.vc file in each case, and then hacked on it until it looked like a Makefile.in file. I wrote a little configure.in script to go along with it, and then configured and built just like any GNU package. For example, I've appended the Tcl configure.in and Makefile.in that I used. Ian ================================================== # Copyright (c) 1995-1996 Sun Microsystems, Inc. # This file is CYGNUS LOCAL. It is a copy of makefile.vc from the # standard tcl distribution, modified to work with cygwin32 and an # autoconf configure script. I have chosen to minimize the number of # changes, so the comments continue to refer to Visual C++ and the # like. This should make it easier to merge in a new version if that # is necessary. prefix = @prefix@ exec_prefix = @exec_prefix@ VPATH = @srcdir@:@srcdir@/../generic:@srcdir@/../compat srcdir = @srcdir@ INSTALL = @INSTALL@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_DATA = @INSTALL_DATA@ CC = @CC@ CFLAGS = @CFLAGS@ NM = @NM@ AS = @AS@ LD = @LD@ DLLTOOL = @DLLTOOL@ WINDRES = @WINDRES@ # Current Tcl version; used in various names. DIRVERSION = @TCL_VERSION@ # The following definition can be set to non-null for special systems # like AFS with replication. It allows the pathnames used for installation # to be different than those used for actually reference files at # run-time. INSTALL_ROOT is prepended to $prefix and $exec_prefix # when installing files. INSTALL_ROOT = # Directory from which applications will reference the library of Tcl # scripts (note: you can set the TCL_LIBRARY environment variable at # run-time to override this value): TCL_LIBRARY = @datadir@/tcl$(DIRVERSION) # Path name to use when installing library scripts: SCRIPT_INSTALL_DIR = $(INSTALL_ROOT)$(TCL_LIBRARY) # Directory in which to install libtcl.so or libtcl.a: LIB_INSTALL_DIR = $(INSTALL_ROOT)@libdir@ # Directory in which to install the program tclsh: BIN_INSTALL_DIR = $(INSTALL_ROOT)@bindir@ # Directory in which to install the include file tcl.h: INCLUDE_INSTALL_DIR = $(INSTALL_ROOT)@includedir@ # # Visual C++ 2.x and 4.0 makefile # # Does not depend on the presence of any environment variables in # order to compile tcl; all needed information is derived from # location of the compiler directories. # # Project directories # # ROOT = top of source tree # # TMPDIR = location where .obj files should be stored during build # # TOOLS32 = location of VC++ 32-bit development tools. Note that the # VC++ 2.0 header files are broken, so you need to use the # ones that come with the developer network CD's, or later # versions of VC++. # # TOOLS16 = location of VC++ 1.5 16-bit tools, needed to build thunking # library. This information is optional; if the 16-bit compiler # is not available, then the 16-bit code will not be built. # Tcl will still run without the 16-bit code, but... # A. Under Windows 3.X you will any calls to the exec command # will return an error. # B. A 16-bit program to test the behavior of the exec # command under NT and 95 will not be built. # ROOT = $(srcdir)/.. TMPDIR = . TOOLS32 = c:\msdev TOOLS16 = c:\msvc # Set this to the appropriate value of /MACHINE: for your platform MACHINE = IX86 # Comment the following line to compile with symbols NODEBUG=1 # uncomment one of the following lines to compile with TCL_MEM_DEBUG, # TCL_MEM_DEBUG, or TCL_COMPILE_DEBUG #DEBUGDEFINES = -DTCL_MEM_DEBUG #DEBUGDEFINES = -DTCL_MEM_DEBUG -DTCL_COMPILE_DEBUG #DEBUGDEFINES = -DTCL_MEM_DEBUG -DTCL_COMPILE_STATS #DEBUGDEFINES = -DTCL_MEM_DEBUG -DTCL_COMPILE_DEBUG -DTCL_COMPILE_STATS ###################################################################### # Do not modify below this line ###################################################################### VERSION = 80 TCLLIB = libtcl$(VERSION).a TCLDLL = cygtcl$(VERSION).dll TCL16DLL = # tcl16$(VERSION).dll TCLSH = cygtclsh$(VERSION).exe TCLTEST = tcltest.exe DUMPEXTS = # $(TMPDIR)\dumpexts.exe TCLPIPEDLL = cygtclpip$(VERSION).dll TCLREGDLL = cygtclreg$(VERSION).dll CAT16 = # cat16.exe CAT32 = cat32.exe TCLSHOBJS = \ $(TMPDIR)/tclAppInit.o TCLTESTOBJS = \ $(TMPDIR)/tclTest.o \ $(TMPDIR)/tclTestObj.o \ $(TMPDIR)/tclWinTest.o \ $(TMPDIR)/testMain.o TCLOBJS = \ $(TMPDIR)/panic.o \ $(TMPDIR)/regexp.o \ $(TMPDIR)/strftime.o \ $(TMPDIR)/tclAsync.o \ $(TMPDIR)/tclBasic.o \ $(TMPDIR)/tclBinary.o \ $(TMPDIR)/tclCkalloc.o \ $(TMPDIR)/tclClock.o \ $(TMPDIR)/tclCmdAH.o \ $(TMPDIR)/tclCmdIL.o \ $(TMPDIR)/tclCmdMZ.o \ $(TMPDIR)/tclCompExpr.o \ $(TMPDIR)/tclCompile.o \ $(TMPDIR)/tclDate.o \ $(TMPDIR)/tclEnv.o \ $(TMPDIR)/tclEvent.o \ $(TMPDIR)/tclExecute.o \ $(TMPDIR)/tclFCmd.o \ $(TMPDIR)/tclFileName.o \ $(TMPDIR)/tclGet.o \ $(TMPDIR)/tclHash.o \ $(TMPDIR)/tclHistory.o \ $(TMPDIR)/tclIndexObj.o \ $(TMPDIR)/tclInterp.o \ $(TMPDIR)/tclIO.o \ $(TMPDIR)/tclIOCmd.o \ $(TMPDIR)/tclIOSock.o \ $(TMPDIR)/tclIOUtil.o \ $(TMPDIR)/tclLink.o \ $(TMPDIR)/tclListObj.o \ $(TMPDIR)/tclLoad.o \ $(TMPDIR)/tclMain.o \ $(TMPDIR)/tclNamesp.o \ $(TMPDIR)/tclNotify.o \ $(TMPDIR)/tclObj.o \ $(TMPDIR)/tclParse.o \ $(TMPDIR)/tclPipe.o \ $(TMPDIR)/tclPkg.o \ $(TMPDIR)/tclPosixStr.o \ $(TMPDIR)/tclPreserve.o \ $(TMPDIR)/tclProc.o \ $(TMPDIR)/tclStringObj.o \ $(TMPDIR)/tclTimer.o \ $(TMPDIR)/tclUtil.o \ $(TMPDIR)/tclVar.o \ $(TMPDIR)/tclWin32Dll.o \ $(TMPDIR)/tclWinChan.o \ $(TMPDIR)/tclWinError.o \ $(TMPDIR)/tclWinFCmd.o \ $(TMPDIR)/tclWinFile.o \ $(TMPDIR)/tclWinInit.o \ $(TMPDIR)/tclWinLoad.o \ $(TMPDIR)/tclWinMtherr.o \ $(TMPDIR)/tclWinNotify.o \ $(TMPDIR)/tclWinPipe.o \ $(TMPDIR)/tclWinSock.o \ $(TMPDIR)/tclWinTime.o cc32 = $(TOOLS32)\bin\cl.exe link32 = $(TOOLS32)\bin\link.exe rc32 = $(TOOLS32)\bin\rc.exe include32 = -I$(TOOLS32)\include cc16 = $(TOOLS16)\bin\cl.exe link16 = $(TOOLS16)\bin\link.exe rc16 = $(TOOLS16)\bin\rc.exe include16 = -I$(TOOLS16)\include WINDIR = $(ROOT)/win GENERICDIR = $(ROOT)/generic TCL_INCLUDES = -I$(WINDIR) -I$(GENERICDIR) TCL_DEFINES = -D__WIN32__ -DUSE_TCLALLOC=0 $(DEBUGDEFINES) TCL_CFLAGS = $(cdebug) $(cflags) $(cvarsdll) \ $(TCL_INCLUDES) $(TCL_DEFINES) $(CFLAGS) CON_CFLAGS = $(cdebug) $(cflags) $(cvars) -DCONSOLE DOS_CFLAGS = $(cdebug) $(cflags) $(include16) -AL DLL16_CFLAGS = $(cdebug) $(cflags) $(include16) -ALw linkdebug = conlflags = -Wl,--subsystem,console -mwindows conlibsdll = ###################################################################### # Link flags ###################################################################### #!IFDEF NODEBUG #ldebug = /RELEASE #!ELSE #ldebug = -debug:full -debugtype:cv #!ENDIF # declarations common to all linker options lcommon = /NODEFAULTLIB /RELEASE /NOLOGO # declarations for use on Intel i386, i486, and Pentium systems #!IF "$(MACHINE)" == "IX86" #DLLENTRY = @12 #lflags = $(lcommon) -align:0x1000 /MACHINE:$(MACHINE) #!ELSE #lflags = $(lcommon) /MACHINE:$(MACHINE) #!ENDIF conlflags = -Wl,--subsystem,console -mwindows guilflags = -mwindows dlllflags = #!IF "$(MACHINE)" == "PPC" #libc = libc.lib #libcdll = crtdll.lib #!ELSE #libc = libc.lib oldnames.lib #libcdll = msvcrt.lib oldnames.lib #!ENDIF baselibs = -lkernel32 $(optlibs) -ladvapi32 -luser32 winlibs = $(baselibs) -lgdi32 -lcomdlg32 -lwinspool guilibs = $(libc) $(winlibs) conlibs = $(libc) $(baselibs) guilibsdll = $(libcdll) $(winlibs) conlibsdll = $(libcdll) $(baselibs) ###################################################################### # Compile flags ###################################################################### #!IFDEF NODEBUG #cdebug = -Ox #!ELSE #cdebug = -Z7 -Od -WX #!ENDIF # declarations common to all compiler options ccommon = -c -W3 -nologo -YX -Dtry=__try -Dexcept=__except #!IF "$(MACHINE)" == "IX86" #cflags = $(ccommon) -D_X86_=1 #!ELSE #!IF "$(MACHINE)" == "MIPS" #cflags = $(ccommon) -D_MIPS_=1 #!ELSE #!IF "$(MACHINE)" == "PPC" #cflags = $(ccommon) -D_PPC_=1 #!ELSE #!IF "$(MACHINE)" == "ALPHA" #cflags = $(ccommon) -D_ALPHA_=1 #!ENDIF #!ENDIF #!ENDIF #!ENDIF cvars = -DWIN32 -D_WIN32 cvarsmt = $(cvars) -D_MT cvarsdll = $(cvarsmt) -D_DLL ###################################################################### # Project specific targets ###################################################################### release: $(TCLSH) dlls $(TCLLIB) dlls: $(TCLDLL) $(TCLPIPEDLL) $(TCLREGDLL) all: $(TCLSH) dlls $(CAT16) $(CAT32) $(TCLLIB) tcltest: $(TCLTEST) dlls $(CAT16) $(CAT32) test: $(TCLTEST) dlls $(CAT16) $(CAT32) ( echo cd $(ROOT)/tests\; source all ) | $(TCLTEST) install: install-binaries install-libraries install-binaries: $(TCLDLL) $(TCLLIB) $(TCLSH) $(TCLPIPEDLL) @for i in $(LIB_INSTALL_DIR) $(BIN_INSTALL_DIR) ; \ do \ if [ ! -d $$i ] ; then \ echo "Making directory $$i"; \ mkdir $$i; \ chmod 755 $$i; \ else true; \ fi; \ done; @echo "Installing $(TCLLIB)" @$(INSTALL_DATA) $(TCLLIB) $(LIB_INSTALL_DIR)/$(TCLLIB) @chmod 555 $(LIB_INSTALL_DIR)/$(TCLLIB) @echo "Installing tclsh" @$(INSTALL_PROGRAM) $(TCLSH) $(BIN_INSTALL_DIR)/$(TCLSH) @echo "Installing tclConfig.sh" @$(INSTALL_DATA) ../unix/tclConfig.sh $(LIB_INSTALL_DIR)/tclConfig.sh install-libraries: @echo "Installing DLL" @$(INSTALL_DATA) $(TCLDLL) $(BIN_INSTALL_DIR)/$(TCLDLL) @$(INSTALL_DATA) $(TCLPIPEDLL) $(BIN_INSTALL_DIR)/$(TCLPIPEDLL) @$(INSTALL_DATA) $(TCLREGDLL) $(BIN_INSTALL_DIR)/$(TCLREGDLL) @for i in $(INSTALL_ROOT)@datadir@ $(INCLUDE_INSTALL_DIR) \ $(SCRIPT_INSTALL_DIR) ; \ do \ if [ ! -d $$i ] ; then \ echo "Making directory $$i"; \ mkdir $$i; \ chmod 755 $$i; \ else true; \ fi; \ done; @for i in http1.0 ; \ do \ if [ ! -d $(SCRIPT_INSTALL_DIR)/$$i ] ; then \ echo "Making directory $(SCRIPT_INSTALL_DIR)/$$i"; \ mkdir $(SCRIPT_INSTALL_DIR)/$$i; \ chmod 755 $(SCRIPT_INSTALL_DIR)/$$i; \ else true; \ fi; \ done; @echo "Installing tcl.h" @$(INSTALL_DATA) $(GENERICDIR)/tcl.h $(INCLUDE_INSTALL_DIR)/tcl.h @for i in $(ROOT)/library/*.tcl $(ROOT)/library/tclIndex $(WINDIR)/tclAppInit.c; \ do \ echo "Installing $$i"; \ $(INSTALL_DATA) $$i $(SCRIPT_INSTALL_DIR); \ done; @for i in http1.0 ; \ do \ for j in $(ROOT)/library/$$i/*.tcl ; \ do \ echo "Installing $$j"; \ $(INSTALL_DATA) $$j $(SCRIPT_INSTALL_DIR)/$$i; \ done; \ done; # $(DUMPEXTS): $(WINDIR)/winDumpExts.c # $(CC) -c $(CON_CFLAGS) $? # $(CC) $(linkdebug) $(conlflags) -o $@ $(TMPDIR)/winDumpExts.obj $(guilibs) $(TCLDLL): $(TCLOBJS) tclres.o $(TMPDIR)/tclcyg.def $(LD) --base-file tcl.base --dll -o $(TCLDLL) $(TCLOBJS) tclres.o -lcygwin -ladvapi32 -luser32 -lgdi32 -lcomdlg32 -lkernel32 --subsystem windows -e _DllMain@12 $(DLLTOOL) --as=$(AS) --dllname $(TCLDLL) --def $(TMPDIR)/tclcyg.def --base-file tcl.base --output-exp tcl.exp $(LD) --base-file tcl.base tcl.exp -dll -o $(TCLDLL) $(TCLOBJS) tclres.o -lcygwin -ladvapi32 -luser32 -lgdi32 -lcomdlg32 -lkernel32 --subsystem windows -e _DllMain@12 $(DLLTOOL) --as=$(AS) --dllname $(TCLDLL) --def $(TMPDIR)/tclcyg.def --base-file tcl.base --output-exp tcl.exp $(LD) tcl.exp --dll -o $(TCLDLL) $(TCLOBJS) tclres.o -lcygwin -ladvapi32 -luser32 -lgdi32 -lcomdlg32 -lkernel32 --subsystem windows -e _DllMain@12 $(TCLLIB): $(TMPDIR)/tclcyg.def $(DLLTOOL) --as=$(AS) --dllname $(TCLDLL) --def $(TMPDIR)/tclcyg.def --output-lib $(TCLLIB) $(TCLSH): $(TCLSHOBJS) tclshres.o $(TCLLIB) $(CC) $(linkdebug) $(conlflags) \ -o $@ $(conlibsdll) $(TCLSHOBJS) tclshres.o $(TCLLIB) $(TCLTEST): $(TCLTESTOBJS) tclshres.o $(TCLLIB) $(CC) $(linkdebug) $(conlflags) -Wl,--stack=1572864 \ -o $@ $(conlibsdll) $(TCLTESTOBJS) tclshres.o $(TCLLIB) # $(TCL16DLL): $(WINDIR)\tcl16.rc $(WINDIR)\tclWin16.c # if exist $(cc16) $(cc16) @<< # $(DLL16_CFLAGS) -Fo$(TMPDIR)\ $(WINDIR)\tclWin16.c # << # @copy << $(TMPDIR)\tclWin16.def > nul # LIBRARY $(@B);dll # EXETYPE WINDOWS # CODE PRELOAD MOVEABLE DISCARDABLE # DATA PRELOAD MOVEABLE SINGLE # HEAPSIZE 1024 # EXPORTS # WEP @1 RESIDENTNAME # UTPROC @2 # << # if exist $(cc16) $(link16) /NOLOGO /ONERROR:NOEXE /NOE @<< # $(TMPDIR)\tclWin16.obj # $@ # nul # $(TOOLS16)\lib\ ldllcew oldnames libw toolhelp # $(TMPDIR)\tclWin16.def # << # if exist $(cc16) $(rc16) -i $(GENERICDIR) $(TCL_DEFINES) $(WINDIR)\tcl16.rc $@ $(TCLPIPEDLL): $(WINDIR)/stub16.c $(CC) -c $(CON_CFLAGS) $(WINDIR)/stub16.c $(CC) $(linkdebug) $(conlflags) -o $@ $(TMPDIR)/stub16.o $(guilibs) $(TCLREGDLL): $(TMPDIR)/tclWinReg.o $(TCLLIB) echo EXPORTS > $(TMPDIR)/tclreg.def echo Registry_Init >> $(TMPDIR)/tclreg.def $(LD) --base-file tclreg.base --dll -o $(TCLREGDLL) $(TMPDIR)/tclWinReg.o $(TCLLIB) -lcygwin -ladvapi32 -luser32 -lgdi32 -lcomdlg32 -lkernel32 --subsystem windows -e _DllEntryPoint@12 $(DLLTOOL) --as=$(AS) --dllname $(TCLREGDLL) --def $(TMPDIR)/tclreg.def --base-file tclreg.base --output-exp tclreg.exp $(LD) --base-file tclreg.base tclreg.exp -dll -o $(TCLREGDLL) $(TMPDIR)/tclWinReg.o $(TCLLIB) -lcygwin -ladvapi32 -luser32 -lgdi32 -lcomdlg32 -lkernel32 --subsystem windows -e _DllEntryPoint@12 $(DLLTOOL) --as=$(AS) --dllname $(TCLREGDLL) --def $(TMPDIR)/tclreg.def --base-file tclreg.base --output-exp tclreg.exp $(LD) tclreg.exp --dll -o $(TCLREGDLL) $(TMPDIR)/tclWinReg.o $(TCLLIB) -lcygwin -ladvapi32 -luser32 -lgdi32 -lcomdlg32 -lkernel32 --subsystem windows -e _DllEntryPoint@12 rm -f $(TMPDIR)/tclreg.def $(CAT32): $(WINDIR)/cat.c $(CC) -c $(CON_CFLAGS) $? $(CC) $(conlflags) -o $@ -Wl,-stack,16384 $(TMPDIR)/cat.o $(conlibs) # $(CAT16): $(WINDIR)\cat.c # if exist $(cc16) $(cc16) $(DOS_CFLAGS) -Fo$(TMPDIR)\ $? # set LIB=$(TOOLS16)\lib # if exist $(cc16) $(link16) /NOLOGO /ONERROR:NOEXE /NOI /STACK:16384 \ # $(TMPDIR)\cat.obj,$@,nul,llibce.lib,nul $(TMPDIR)/tclcyg.def: $(DUMPEXTS) $(TCLOBJS) echo 'EXPORTS' > tmp.def for o in $(TCLOBJS); do \ $(NM) --extern-only --defined-only $$o | sed -e 's/[^ ]* [^ ]* //' -e 's/^_//' | fgrep -v DllEntryPoint | fgrep -v DllMain | fgrep -v impure_ptr >> tmp.def; \ done mv tmp.def $(TMPDIR)/tclcyg.def # # Special case object file targets # $(TMPDIR)/testMain.o: $(WINDIR)/tclAppInit.c $(CC) -c -o $(TMPDIR)/testMain.o $(TCL_CFLAGS) -DTCL_TEST $? # # Implicit rules # .c.o: $(CC) -c $(TCL_CFLAGS) $< tclres.o: $(srcdir)/tcl.rc $(WINDRES) --include $(GENERICDIR) --include $(WINDIR) --define __WIN32__ --define VS_VERSION_INFO=1 $(srcdir)/tcl.rc tclres.o tclshres.o: $(srcdir)/tclsh.rc $(WINDRES) --include $(GENERICDIR) --include $(WINDIR) --define __WIN32__ --define VS_VERSION_INFO=1 $(srcdir)/tclsh.rc tclshres.o #{$(WINDIR)}.rc{$(TMPDIR)}.res: # $(rc32) -fo $@ -r -i $(GENERICDIR) -i $(WINDIR) -D__WIN32__ \ # $(TCL_DEFINES) $< clean: rm -f *.exp *.a *.dll *.exe $(TMPDIR)/*.o *.res *.def rm -f tcl.base tclreg.base Makefile: $(WINDIR)/Makefile.in config.status $(SHELL) config.status config.status: $(WINDIR)/configure ./config.status --recheck ================================================== ================================================== dnl The file is CYGNUS LOCAL. It is used for cygwin32. dnl This file is an input file used by the GNU "autoconf" program to dnl generate the file "configure", which is run during Tcl installation dnl to configure the system for the local environment. AC_PREREQ(2.5) AC_INIT(../generic/tcl.h) if test "${prefix}" = "NONE"; then prefix=/usr/local fi if test "${exec_prefix}" = "NONE"; then exec_prefix=$prefix fi AC_PROG_CC NM=${NM-nm} AC_SUBST(NM) AS=${AS-as} AC_SUBST(AS) LD=${LD-ld} AC_SUBST(LD) DLLTOOL=${DLLTOOL-dlltool} AC_SUBST(DLLTOOL) WINDRES=${WINDRES-windres} AC_SUBST(WINDRES) AC_PROG_INSTALL # The following variables are just for tclConfig.sh, not for Makefile. LIBOBJS= AC_SUBST(LIBOBJS) TCL_VERSION=8.0 TCL_MAJOR_VERSION=8 TCL_MINOR_VERSION=0 TCL_PATCH_LEVEL=b2 VERSION=${TCL_VERSION} DL_LIBS= MATH_LIBS=-lm SHLIB_CFLAGS= SHLIB_LD= SHLIB_LD_LIBS= SHLIB_SUFFIX= LD_FLAGS= TCL_LD_SEARCH_FLAGS= TCL_BUILD_LIB_SPEC="-L`pwd` -ltcl`echo ${VERSION} | tr -d .`" TCL_LIB_SPEC="-L${exec_prefix}/lib -ltcl`echo ${VERSION} | tr -d .`" TCL_LIB_VERSIONS_OK=nodots TCL_SHARED_LIB_SUFFIX= TCL_UNSHARED_LIB_SUFFIX="`echo ${VERSION} | tr -d .`.a" eval "TCL_LIB_FILE=libtcl${TCL_UNSHARED_LIB_SUFFIX}" TCL_SRC_DIR=`cd $srcdir/..; pwd` if test "$prefix" != "$exec_prefix"; then TCL_PACKAGE_PATH="${exec_prefix}/lib ${prefix}/lib" else TCL_PACKAGE_PATH="${prefix}/lib" fi AC_SUBST(DL_LIBS) AC_SUBST(LD_FLAGS) AC_SUBST(MATH_LIBS) AC_SUBST(SHLIB_CFLAGS) AC_SUBST(SHLIB_LD) AC_SUBST(SHLIB_LD_LIBS) AC_SUBST(SHLIB_SUFFIX) AC_SUBST(TCL_BUILD_LIB_SPEC) AC_SUBST(TCL_LD_SEARCH_FLAGS) AC_SUBST(TCL_LIB_FILE) AC_SUBST(TCL_LIB_SPEC) AC_SUBST(TCL_LIB_VERSIONS_OK) AC_SUBST(TCL_MAJOR_VERSION) AC_SUBST(TCL_MINOR_VERSION) AC_SUBST(TCL_PACKAGE_PATH) AC_SUBST(TCL_PATCH_LEVEL) AC_SUBST(TCL_SHARED_LIB_SUFFIX) AC_SUBST(TCL_SHLIB_CFLAGS) AC_SUBST(TCL_SRC_DIR) AC_SUBST(TCL_UNSHARED_LIB_SUFFIX) AC_SUBST(TCL_VERSION) AC_OUTPUT(Makefile ../unix/tclConfig.sh) ================================================== - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fjh@cs.mu.OZ.AU Sun Aug 17 16:54:00 1997 From: fjh@cs.mu.OZ.AU (Fergus Henderson) Date: Sun, 17 Aug 1997 16:54:00 -0000 Subject: Tcl/Tk with extensions under cyg-win32? References: <3.0.32.19970817175328.009554b0@bl-mail2.corpeast.baynetworks.com> Message-ID: <199708172342.JAA03047@mundook.cs.mu.OZ.AU> You write: >I was hoping somebody has had some experience with what I want to do. I'd >like to compile Tk/Tcl 4.2/7.6 with several extensions, including itcl 2.2, >TclX, Expect, and Scotty (Tnm). I'd like to compile them as shared >libraries/loadable packages, if that is possible under cyg-win32. If not, I >suppose I can just link them in statically, it's just that I'd rather not. > >I was wondering if anyone had any experience in doing this and, if so, what >problems were run into. How much had to be changed to compile under >cyg-win32? What target opsys did you use, linux? What other things should I >be concerned about when I go about doing this. Check < http://www.cs.mu.oz.au/~fjh/gnu-win32/gnu-win32.html > for some instructions regarding an easy way to build DLLs using gnu-win32 b18. If you use the Makefile.DLLs listed on that page, then you need to make only minor changes. If you want loadable packages (i.e. dlopen() etc.) then b18 is not sufficient. However, someone recently posted a patch to support dlopen() and friends, so check the mail archives. -- Fergus Henderson | "I have always known that the pursuit WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From spong@usa.net Sun Aug 17 17:38:00 1997 From: spong@usa.net (Matt Spong) Date: Sun, 17 Aug 1997 17:38:00 -0000 Subject: Microsoft Foundation Classes Message-ID: Is it possible to use MFC 4.1 using gnu-win32? If so, how? ____________________________________________________________________ Get free e-mail and a permanent address at http://www.netaddress.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Ian@kiwiplan.co.nz Sun Aug 17 17:50:00 1997 From: Ian@kiwiplan.co.nz (Ian Collins) Date: Sun, 17 Aug 1997 17:50:00 -0000 Subject: Unix spooler Message-ID: <07551A47EAD6D01186430060B025C6F40A6C78@modem1.kiwiplan2.co.nz> How would I access my network/local windows 95/nt printers from a bash prompt? There doesn't seem to be any spooler implementation in gnu-win32 (or is there). Many Regards, Ian Collins - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From room_cygnus@uncnsrd.mt-kisco.ny.us Sun Aug 17 18:30:00 1997 From: room_cygnus@uncnsrd.mt-kisco.ny.us (IGnatius T Foobar) Date: Sun, 17 Aug 1997 18:30:00 -0000 Subject: Thanks! Message-ID: <336214@uncnsrd.mt-kisco.ny.us> Michael Hirmke: thank you! The discovery of "autoexnt" on the web page you directed me to will be _very_ helpful. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sun Aug 17 21:10:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sun, 17 Aug 1997 21:10:00 -0000 Subject: Unix spooler Message-ID: <01BCABAE.5C110610@sos> Ian Collins wrote: > How would I access my network/local windows 95/nt printers from a bash > prompt? > There doesn't seem to be any spooler implementation in gnu-win32 (or is > there). Just try "ls >prn" :-) -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From geier@forwiss.uni-passau.de Sun Aug 17 22:29:00 1997 From: geier@forwiss.uni-passau.de (Roland Geier) Date: Sun, 17 Aug 1997 22:29:00 -0000 Subject: Tcl/Tk with extensions under cyg-win32? Message-ID: <199708180528.HAA05818@dior.forwiss.uni-passau.de> > I was hoping somebody has had some experience with what I want to do. I'd > like to compile Tk/Tcl 4.2/7.6 with several extensions, including itcl 2.2, > TclX, Expect, and Scotty (Tnm). I'd like to compile them as shared > libraries/loadable packages, if that is possible under cyg-win32. If not, I > suppose I can just link them in statically, it's just that I'd rather not. > > I was wondering if anyone had any experience in doing this and, if so, what > problems were run into. How much had to be changed to compile under > cyg-win32? What target opsys did you use, linux? What other things should I > be concerned about when I go about doing this. With tcl's dynamic package-concept you don't have to rebuild tcl/tk at all but use the provided b18-tcl/tk-binaries instead. Just make dlls from the extensions and use the package-mechanism to load them dynamically, that's all. I did that for itcl1.5 and my own extensions. If the generation of relocatable dlls won't work for you (perhaps due to some weird c++-code in your extensions) make static-packages ('.a'-archives) and use the tcl-package-mechanism for static packages. Then you won't have to recompile tcl/tk either but only extend the interpreter. Assume you have an gnuwin32-compiled archive 'itcl-2.2.a' you want to use as a static package in tcl. Then build your own interpreter that has the package included: : #include : : static int : loadStaticPackages( Tcl_Interp* interp ) { : char loadCmd[256]; : : Tcl_StaticPackage( NULL, "itcl" , Itcl_Init, NULL ); : strcpy( loadCmd, "load \"\" itcl" ); : Tcl_Eval( interp, loadCmd ); : } : : : int : Tcl_AppInit(Tcl_Interp* interp) { : if (Tcl_Init(interp) == TCL_ERROR) { : return TCL_ERROR; : } : return ( loadStaticPackages( interp ) ); : } : : int : main(int argc, char** argv) { : Tcl_Main(argc, argv, Tcl_AppInit); : return 0; : } Now you can use "package require Itcl" in your tcl-code in the same manner as if you wanted to load the package dynamically. So if you succeed to build dll's from your extensions in the future (perhaps with oncoming b19) you wont't have to change your tcl-code. Hth, --- Roland. --- Roland Geier Phone: ++49 (8505) 92182 Universitaet Passau, GER Fax: ++49 (8505) 922311 Lehrstuhl Prof. Dr. Donner geier@forwiss.uni-passau.de D-94032 Passau http://www.uni-passau.de/~geier/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dbrci@giasbg01.vsnl.net.in Sun Aug 17 23:51:00 1997 From: dbrci@giasbg01.vsnl.net.in (DaimlerBenz India) Date: Sun, 17 Aug 1997 23:51:00 -0000 Subject: GNU on NT Message-ID: <1.5.4.16.19970817043422.18a7f9c0@giasbg01.vsnl.net.in> Sir, Is make.exe (GNU) is available on NT? Can I make use of the same ? For using, Do I need to pay? Gupta - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kuku@gilberto.physik.RWTH-Aachen.DE Mon Aug 18 00:22:00 1997 From: kuku@gilberto.physik.RWTH-Aachen.DE (Christoph Kukulies) Date: Mon, 18 Aug 1997 00:22:00 -0000 Subject: Tcl/Tk with extensions under cyg-win32? References: <3.0.32.19970817175328.009554b0@bl-mail2.corpeast.baynetworks.com> <199708172342.JAA03047@mundook.cs.mu.OZ.AU> Message-ID: <19970818092443.29230@gil.physik.rwth-aachen.de> On Mon, Aug 18, 1997 at 09:42:22AM +1000, Fergus Henderson wrote: > You write: > > >I was hoping somebody has had some experience with what I want to do. I'd > >like to compile Tk/Tcl 4.2/7.6 with several extensions, including itcl 2.2, [--] > Check < http://www.cs.mu.oz.au/~fjh/gnu-win32/gnu-win32.html > for some That URL doesn't seem to exist. Could you check? > instructions regarding an easy way to build DLLs using gnu-win32 b18. > If you use the Makefile.DLLs listed on that page, then you need to make > only minor changes. [--] > > -- > Fergus Henderson | "I have always known that the pursuit > WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" > PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. > - > For help on using this list (especially unsubscribing), send a message to > "gnu-win32-request@cygnus.com" with one line of text: "help". -- Chris Christoph P. U. Kukulies kuku@gil.physik.rwth-aachen.de - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From arifi_koseoglu@karmaint.com Mon Aug 18 05:26:00 1997 From: arifi_koseoglu@karmaint.com (arifi_koseoglu@karmaint.com) Date: Mon, 18 Aug 1997 05:26:00 -0000 Subject: Unix spooler Message-ID: <9707188719.AA871940283@karmaint.com> How'bout if the printer is not on te Parallel port but on the network ? -Arifi ______________________________ Reply Separator _________________________________ Subject: RE: Unix spooler Author: Sergey Okhapkin at INTERNET Date: 8/18/97 11:41 AM Ian Collins wrote: > How would I access my network/local windows 95/nt printers from a bash > prompt? > There doesn't seem to be any spooler implementation in gnu-win32 (or is > there). Just try "ls >prn" :-) -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ahatcher@chromatic.com Mon Aug 18 10:58:00 1997 From: ahatcher@chromatic.com (Andy Hatcher) Date: Mon, 18 Aug 1997 10:58:00 -0000 Subject: b18 "env -" bug with WIN95 Message-ID: <33F88D3C.25A@chromatic.com> env - [command] env --ignore-environment [command] Should remove all of the environment before executing the command. Neither works. Test case: env - env should print nothing env - prints nothing, but the effect doesn't last to the next executed command. -- Andy Hatcher ahatcher@chromatic.com (408)752-9387 - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From wleigh@xionics.com Mon Aug 18 12:31:00 1997 From: wleigh@xionics.com (Bill Leigh) Date: Mon, 18 Aug 1997 12:31:00 -0000 Subject: Cygwin32 API Message-ID: <33F8A307.63DECDAD@xionics.com> Where would I find documentation about the use of the API? -- Bill Leigh wleigh@xionics.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Daviid.Smethurst@pancredit.com Mon Aug 18 12:31:00 1997 From: Daviid.Smethurst@pancredit.com (David Smethurst) Date: Mon, 18 Aug 1997 12:31:00 -0000 Subject: Newbie ! Message-ID: Hi, Hope this isn't a silly question (and the right place to post to !) I am trying to get a normal 'Window' to open but I don't seem to be having any luck. Can anyone give me some pointers on where to start/what to do ? Thanks Dave --- D.Smethurst Developer - PanCredit Systems Ltd. #include - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ebritten@uci.edu Mon Aug 18 21:16:00 1997 From: ebritten@uci.edu (Eric Britten) Date: Mon, 18 Aug 1997 21:16:00 -0000 Subject: Microsoft Foundation Classes References: Message-ID: <33F893CD.13056BBE@uci.edu> > Is it possible to use MFC 4.1 using gnu-win32? If so, how? > If you can get the source and compile it, I would say yes. I compiled OWL 2.0 a few months ago and wrote a simple program that worked. OWL is a class library that is different from MFC, however I did cut out the multithreaded functionallity. OWL does not use any compiler specific keywords. You only need templates and exception handling. As long as GCC supports all the syntax conventions of MFC you can compile it. I know is uses a special keyword called afx_msg; it is used for window message handling. I am not sure if there is a work around for this keyword. This is your biggest problem to overcome. You need the source because VC++ and Borland C++ use a different name mangling conventions. If you can get the source to compile, you can then create static and dynamic link MFC libraries. DLLs with GCC are hard to create, I have had little luck with them. > ____________________________________________________________________ > Get free e-mail and a permanent address at http://www.netaddress.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From garbanzo@hooked.net Mon Aug 18 21:16:00 1997 From: garbanzo@hooked.net (Alex) Date: Mon, 18 Aug 1997 21:16:00 -0000 Subject: Unix spooler Message-ID: On Mon, 18 Aug 1997 arifi_koseoglu@karmaint.com wrote: > > > How'bout if the printer is not on te Parallel port but on the network ? > > -Arifi Perhaps use the UNC pathname for it? - alex - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pan@hol.gr Mon Aug 18 21:16:00 1997 From: pan@hol.gr (Panos Platon Tsapralis) Date: Mon, 18 Aug 1997 21:16:00 -0000 Subject: ncurses, gdbm & ps.exe Message-ID: <01BCAC24.AF2B9380@panos> Can someone please tell me where I can find a, preferably precompiled, version of the "(n)curses" library, that can be used with Beta18 of GNU_Win32 ? Can I use the sources of the SlackWare Linux, without making any modifications ? The same question applies for the "gdbm" library. And I still can't find a documentation for the built-in ps.exe command ( which, I am pretty sure, bears no resemblence to the original Unix command ). +----------------------------------------------+ | Panos Platon Tsapralis,Software Engineer, | | M.I.S.Applications Manager - TeleStet, | +----------------------------------------------+ | E-mail Address = pan@hol.gr, | | Home Address = 3 Aishylou Str., 15234, | | Halandri, Athens, | | Attica, GREECE, | | Home Phone No. = (00301)68.58.298, | | Bysiness Address = 60 Kifissias Ave., 15125, | | Amaroussion, Athens, | | Attica, GREECE, | | Office Phone No. = (00301)61.86.013, | | Mobile Phone No. = (003093)23.90.13. | +----------------------------------------------+ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From garbanzo@hooked.net Mon Aug 18 21:16:00 1997 From: garbanzo@hooked.net (Alex) Date: Mon, 18 Aug 1997 21:16:00 -0000 Subject: Unix spooler References: <9707188719.AA871940283@karmaint.com> Message-ID: On Mon, 18 Aug 1997 arifi_koseoglu@karmaint.com wrote: > > > How'bout if the printer is not on te Parallel port but on the network ? > > -Arifi Perhaps use the UNC pathname for it? - alex - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From venkman1@ix.netcom.com Mon Aug 18 21:16:00 1997 From: venkman1@ix.netcom.com (Narayan Natarajan) Date: Mon, 18 Aug 1997 21:16:00 -0000 Subject: Lynx Problem References: <199708172045.QAA29110@iag.net> Message-ID: <3.0.1.32.19970818175835.0091a320@popd.ix.netcom.com> Yes, I got hold of that ver of lynx too. There is a solution: I found another which worked... I don't remember quite exactly where but found while nosing around the virtually Un*x page: http://www.itribe.net/virtunix/otherstuff.html If you can't find it, I still have the zip file and can mail it to you (520Kb). Narayan At 04:46 PM 8/17/97 -0400, you wrote: > > Has anyone else had this problem? Lynx appears to start fine, but then I >get "Alert! Unable to connect to remote host" > Any idea? I got it from the URL Sergey gave. > > - SJ >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > > | Narayan _/~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~/ |Natarajan _/ venkman1@ix.netcom.com http://www.netcom.com/~venkman1 _/ | _/ (810)-475-0667 . .____ _ _/ | _/ "The heavens call to you, and * . x -(____/ | _ / | _/ circle about you, displaying to ________/____/ \ ( )----- | _/ you their eternal splendors, and ,/ _ ( ------- | _/ your eye gazes only to earth." <________\ // (_)----- | _/ -- Dante --<== * |_____/ _ |/ -(_____) (_)----- |~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pan@hol.gr Mon Aug 18 21:16:00 1997 From: pan@hol.gr (Panos Platon Tsapralis) Date: Mon, 18 Aug 1997 21:16:00 -0000 Subject: df.exe gives me an error message... Message-ID: <01BCAC23.D7422380@panos> I tried to use the df.exe file that you sent to me, attached to your last message, and it gives me the following message in the command line of BASH : bash.exe: /GNUWIN32/B18/H-I386-CYGWIN32/BIN/df.exe: (null) while, at the same time a dialog box appears on-screen, with a caption "Error Starting Program", a "OK" button and a message that says : "The DF.EXE file is linked to missing export KERNEL32.DLL:GetDiskFreeSpaceExA. What am I missing ? +----------------------------------------------+ | Panos Platon Tsapralis,Software Engineer, | | M.I.S.Applications Manager - TeleStet, | +----------------------------------------------+ | E-mail Address = pan@hol.gr, | | Home Address = 3 Aishylou Str., 15234, | | Halandri, Athens, | | Attica, GREECE, | | Home Phone No. = (00301)68.58.298, | | Bysiness Address = 60 Kifissias Ave., 15125, | | Amaroussion, Athens, | | Attica, GREECE, | | Office Phone No. = (00301)61.86.013, | | Mobile Phone No. = (003093)23.90.13. | +----------------------------------------------+ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pan@hol.gr Mon Aug 18 21:16:00 1997 From: pan@hol.gr (Panos Platon Tsapralis) Date: Mon, 18 Aug 1997 21:16:00 -0000 Subject: telnetd etc. Message-ID: <01BCAC26.6D8E2C60@panos> I have been reading messages in the list about telnetd and various other Unix connectivity options. Should I understand that all these goodies are available on the GNU_Win32 platform ? If so, can they be used in Windows95 with the standard TCP/IP facility ? Where can I find / download them ? +----------------------------------------------+ | Panos Platon Tsapralis,Software Engineer, | | M.I.S.Applications Manager - TeleStet, | +----------------------------------------------+ | E-mail Address = pan@hol.gr, | | Home Address = 3 Aishylou Str., 15234, | | Halandri, Athens, | | Attica, GREECE, | | Home Phone No. = (00301)68.58.298, | | Bysiness Address = 60 Kifissias Ave., 15125, | | Amaroussion, Athens, | | Attica, GREECE, | | Office Phone No. = (00301)61.86.013, | | Mobile Phone No. = (003093)23.90.13. | +----------------------------------------------+ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mphalpin@iag.net Tue Aug 19 16:29:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Tue, 19 Aug 1997 16:29:00 -0000 Subject: Changing Root (/) Message-ID: <199708192236.SAA15064@iag.net> Anybody here happen to know how I would change the root (/) directory? It's current c:, but I want it to be e:\gnuwin32\unix If you know, please contact me! Thanks! - SJ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From tcarson@cs.ucsd.edu Tue Aug 19 16:29:00 1997 From: tcarson@cs.ucsd.edu (Ted Carson) Date: Tue, 19 Aug 1997 16:29:00 -0000 Subject: profiling Message-ID: <33FA2C51.A2DA5822@cs.ucsd.edu> Hi All, I'm using the GNU-WIN32 gcc compiler under NT and was wondering if a port of any profiling tools existed. For example 'gprof' would be great. There is no mention of profiling in the FAQ. thanks, t - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From cgf@bbc.com Tue Aug 19 18:47:00 1997 From: cgf@bbc.com (Chris Faylor) Date: Tue, 19 Aug 1997 18:47:00 -0000 Subject: b18 "env -" bug with WIN95 References: <33F88D3C.25A@chromatic.com> Message-ID: In article < 33F88D3C.25A@chromatic.com >, Andy Hatcher wrote: >env - [command] >env --ignore-environment [command] >Should remove all of the environment before executing the >command. Neither works. This should be fixed in Sergey's version of cygwin.dll. http://www.lexa.ru/sos -- http://www.bbc.com/ cgf@bbc.com "Strange how unreal VMS=>UNIX Solutions Boston Business Computing the real can be." - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ebritten@uci.edu Tue Aug 19 20:07:00 1997 From: ebritten@uci.edu (Eric Britten) Date: Tue, 19 Aug 1997 20:07:00 -0000 Subject: df.exe gives me an error message... References: <01BCAC23.D7422380@panos> Message-ID: <33F9D7B5.AD404D33@uci.edu> Panos Platon Tsapralis wrote: > I tried to use the df.exe file that you sent to me, attached > to your last message, and it gives me the following message in the > command line of BASH : > > bash.exe: /GNUWIN32/B18/H-I386-CYGWIN32/BIN/df.exe: (null) > > while, at the same time a dialog box appears on-screen, with a caption > "Error Starting Program", a "OK" button and a message that says : > > "The DF.EXE file is linked to missing export > KERNEL32.DLL:GetDiskFreeSpaceExA. > > What am I missing ? It looks like you are using the wrong version of Windows or df should have been written to use GetDiskFreeSpaceA if GetDiskFreeSpaceExA doesn't exist. GetDiskFreeSpaceExA is available in Windows 95 OSR2(System Properties will say "Microsoft Windows 95" "4.00.950 B" if the B is not there you do not have OSR2). Windows NT 4.0 does support the GetDiskFreeSpaceExA function. Thus your OS must be Windows 95. The df utility should be rewritten to use GetDiskFreeSpaceA if GetDiskFreeSpaceExA doesn't exist. Its not too hard. Use LoadLibrary/LoadLibraryEx to load Kernel32.dll, then GetProcAddress to find GetDiskFreeSpaceExA. If it is not present, use GetDiskFreeSpaceA. Unfortunately GetDiskFreeSpaceA may return incorrect information on the size of the disks. Eric Britten > > > +----------------------------------------------+ > | Panos Platon Tsapralis,Software Engineer, | > | M.I.S.Applications Manager - TeleStet, | > +----------------------------------------------+ > | E-mail Address = pan@hol.gr, | > | Home Address = 3 Aishylou Str., 15234, | > | Halandri, Athens, | > | Attica, GREECE, | > | Home Phone No. = (00301)68.58.298, | > | Bysiness Address = 60 Kifissias Ave., 15125, | > | Amaroussion, Athens, | > | Attica, GREECE, | > | Office Phone No. = (00301)61.86.013, | > | Mobile Phone No. = (003093)23.90.13. | > +----------------------------------------------+ > - > For help on using this list (especially unsubscribing), send a message > to > "gnu-win32-request@cygnus.com" with one line of text: "help". - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From alto-armina@pplus.shell2.ba.best.com Tue Aug 19 20:07:00 1997 From: alto-armina@pplus.shell2.ba.best.com (Armina Badalian) Date: Tue, 19 Aug 1997 20:07:00 -0000 Subject: Partial linking with Cygwin32 Message-ID: <33F99D32.7750@pplus.shell2.ba.best.com> Hello, I am trying to use Cygwin32 to cross compile win32 static library on SGI workstation running Irix5. Installation went on quite smoothly, but I hit a snag when I tried to partially link the object files. The first linking went ok, but when I tried to link the resulting object file again, linker complained about unrelocatable symbols inside .text section. Could you please advise me if there a way to do partial linking using Cygwin32 tools? Thank you very much, Mark Gonikberg, AltoCom, Inc. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jazz@softway.com Tue Aug 19 20:07:00 1997 From: jazz@softway.com (Jason Zions) Date: Tue, 19 Aug 1997 20:07:00 -0000 Subject: OpenNT question again References: <199708170902.CAA05441@jungfrau.usc.edu> Message-ID: <33F939FE.2560@softway.com> The alpha-test version of gcc available on Softway's web site is indeed a port of the FSF's gcc; I don't know if that's what you mean by "related to Cygnus". Once we get the last of the bugs out, the diffs to the source will be made available in accordance with the GPL; no sense in doing it yet, since we're still cleaning nits. Jason Zions Softway Systems - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From noer@cygnus.com Tue Aug 19 20:07:00 1997 From: noer@cygnus.com (Geoffrey Noer) Date: Tue, 19 Aug 1997 20:07:00 -0000 Subject: cdk.exe install problem References: Message-ID: <199708200306.UAA26680@cirdan.cygnus.com> Jeremy Blackman wrote: [...] > Clearing out your TEMP directory entirely should do it - that will get rid > of both any files with conflicting names, and the potential 'too many > files' problem. :) [...] Added the installshield problem to the gnu-win32 FAQ. Thanks! -- Geoffrey Noer noer@cygnus.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ccurtin@trintech.com Tue Aug 19 20:07:00 1997 From: ccurtin@trintech.com (Colman Curtin) Date: Tue, 19 Aug 1997 20:07:00 -0000 Subject: lynx compilation Message-ID: <199708181006.LAA47361@muzak.trintech.ie> hi, did you use the attached patch? Sergey posted this to the list in June. said lynx2.5* and 2.7* compiled using make generic. hope this helps. Colman. P.s. I have the same problem with lynx as you but haven't had time to check it out. are you running 95/NT? ____________________________________________________________ Colman Curtin mailto:ccurtin@trintech.ie Software Engineer ------------------------------------------------------------ Trintech (Manufacturing) Ltd, http://www.trintech.com/ South County Business Park, Leopardstown, Dublin 18. Tel +353-1-2956766 Fax +353-1-2954735 ------------------------------------------------------------ It is impossible to make anything foolproof because fools are so ingenious. Edsel Murphy, dec. begin 666 lynx.diff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essage-ID: <9708191940.AA04238@kmrmail.kmr.ll.mit.edu> >>>>> "arifi" == arifi koseoglu writes: > How'bout if the printer is not on te Parallel port but on the network ? > -Arifi > ______________________________ Reply Separator _____________________________ > Subject: RE: Unix spooler > Author: Sergey Okhapkin at INTERNET > Date: 8/18/97 11:41 AM > Ian Collins wrote: >> How would I access my network/local windows 95/nt printers from a bash >> prompt? >> There doesn't seem to be any spooler implementation in gnu-win32 (or is >> there). > Just try "ls >prn" :-) > -- > Sergey Okhapkin, http://www.lexa.ru/sos > Moscow, Russia > Looking for a job I just learned this trick yesterday: net use /persistent:yes net use lpt1: \\YourPrintServer\YourPrintername This was to get enscript for 95/NT to work with a network printer, but I have not tried to redirect from the command line. Hope it helps. Sarir -- Sarir "Riri" Khamsi Kwajalein, Marshall Islands email: khamsis@kmrmail.kmr.ll.mit.edu w: 805-355-1534 Time Zone: GMT +12 hours f: 805-355-3833 web: http://www.futureone.com/~sponge - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From noer@cygnus.com Tue Aug 19 20:07:00 1997 From: noer@cygnus.com (Geoffrey Noer) Date: Tue, 19 Aug 1997 20:07:00 -0000 Subject: Header files. References: Message-ID: <199708200253.TAA26633@cirdan.cygnus.com> Ramsn Garcma Fernandez wrote: > > I have seen that Cygwin32 support of Windows API is rather limited in > header files. > > Why don't you take the headers from Willows toolkit, a free emulator > of the Windows API under Unix. A very good suggestion. I am investigating this possibility... -- Geoffrey Noer noer@cygnus.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mikew@harlequin.co.uk Wed Aug 20 09:06:00 1997 From: mikew@harlequin.co.uk (Mike Williams) Date: Wed, 20 Aug 1997 09:06:00 -0000 Subject: Changing Root (/) References: <199708192236.SAA15064@iag.net> Message-ID: <199708200915.KAA23530@spin.man.harlequin.co.uk> On 19 Aug 97 at 18:37, Sonic Junior wrote: > Anybody here happen to know how I would change the root (/) directory? > It's current c:, but I want it to be e:\gnuwin32\unix > If you know, please contact me! Thanks! Hi, I did this just yesterday - I had to edit the registry settings directly since you can't umount /. The downside is that if you ever do mount --reset you have to re-edit the registry settings. Unless there is a better way I don't know yet ... Mike. -- I wish I were what I was when I wished I were what I am now. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kh@wg.icl.co.uk Wed Aug 20 09:06:00 1997 From: kh@wg.icl.co.uk (Kevin Hughes) Date: Wed, 20 Aug 1997 09:06:00 -0000 Subject: Link behaviour when accessing REAL unix via NFS Message-ID: <9708201326.AA14118@ss38.wg.icl.co.uk> I hanvnt got the equipment to try this (yet) but am interested in how bash under win95 sees a remotely mounted UNIX filesystem when using NFS. I envisage having an NT box with NFS access to a SUN system. I want to run bash etc. on the NT platform with access to the filestore on the SUN. I have many scripts which do things like if test -h $file i.e I am tring to determine if $file is a link or not. How will this behave on the bash under NT? I can think of at least two possibilities it will work it fail because the signature in $file which cygwin32 recognises as a link indicator is not there the NFS filestore automatically follows the UNIX link when cygwin opens/read the file to get at the signature Will the patches from Sergey make any difference? Any info much appreciated Kevin - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From alien@bra01.icl.co.uk Wed Aug 20 09:06:00 1997 From: alien@bra01.icl.co.uk (alien) Date: Wed, 20 Aug 1997 09:06:00 -0000 Subject: FW: Spooler References: <07551A47EAD6D01186430060B025C6F40A6C68@modem1.kiwiplan2.co.nz> Message-ID: <199708201421.OAA07403@vpm.bra01.icl.co.uk> Hey you guys, whats wrong with the corvette is that not America's muscle car ?? Myself I would go for a Lotus Esprit V8 Turbo, 2nd a Corvette I love me Starion & as the man from Oz said there ain't may around I have only seen 4 in the last 3 years, 2 std & 2x wide body (1 of them happens to live on the same estate as me :( ) Laters Rob > > At 05:47 PM 8/13/97 -0700, you wrote: > >What's wrong with us? Why are we so hooked to the Starquest? > >Can Mitsubishi Elipse GSX or 3000GT VR4 (GTO) give us the same rush! > >Or are we just too cheap to buy a newer, more reliable car? > >The one sitting on my drive way is my second Conquest. I know a few > >people had owned 2 Conquests. Buy Why?? > >Had you ask yourself why? Why so crazy!?! Don't get me wrong! I don't > >hate my Conquests, if I hate it, I wouldn't buy the second one. > >Anything that can replace our Conquest? > >Miata? CRX? Corrado? 300ZX? 944? Tell me what you want? > > > >--Clemson > > > > > _________________ > Phil Engert > "important quote here" > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From noer@cygnus.com Wed Aug 20 09:06:00 1997 From: noer@cygnus.com (Geoffrey Noer) Date: Wed, 20 Aug 1997 09:06:00 -0000 Subject: Cygwin32 API References: <33F8A307.63DECDAD@xionics.com> Message-ID: <199708200251.TAA26624@cirdan.cygnus.com> Bill Leigh wrote: > > Where would I find documentation about the use of the API? Currently, the documentation == the source code in most cases. src/winsup and src/newlib contain most of the code that goes into the cygwin.dll. You should also read the FAQ although that doesn't say too much about the API. There are also the ChangeLog files in the source directories -- search for the routine you want to know about. If someone wants to help me document the Cygwin32 API, please let me know. This is really something that needs to be done. (Defining the scope and listing limitations in terms of which functions are only supported under Win NT and which work on Win NT and 95). -- Geoffrey Noer noer@cygnus.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From stf@hugo.int-evry.fr Wed Aug 20 09:06:00 1997 From: stf@hugo.int-evry.fr (Stephane Mariel) Date: Wed, 20 Aug 1997 09:06:00 -0000 Subject: DLL Problem Message-ID: <199708201219.OAA01103@hugo.int-evry.fr> [this is a repost, the first one should be lost, hope you will not recieve it twice] Hi, I'm trying to port a tcl module from unix to win95 as a dll to load within wish. So i tried to began with the example provided on cygnus.com , it works fine, i got my foo.dll, main.exe, but moving to my real software when I run ld I got: c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libccrt0.o)(.t ext+0x8a):libccrt0.cc: undefined reference to `GetModuleHandleA@4' c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t ext+0xc):libcmain.cc: undefined reference to `GetModuleHandleA@4' c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t ext+0x15):libcmain.cc: undefined reference to `GetCommandLineA@0' c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t ext+0x1e):libcmain.cc: undefined reference to `WinMain@16' This seems to be a reentrancy problem, since a simple call like : fprintf(stderr, ...) makes use of _impure_ptr which is a struct _reent* ptr that is provided by any process using the dll. So , probably the exact question is : how to say to ld: this symbol must not be search in libc, (libccrto.o, libcmain.o) but must be mapped to a symbol defined in the calling process? Any idea? Thanks, stf - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kcantrel@dg-rtp.dg.com Wed Aug 20 09:06:00 1997 From: kcantrel@dg-rtp.dg.com (Ken Cantrell) Date: Wed, 20 Aug 1997 09:06:00 -0000 Subject: Changing Root (/) Message-ID: <9708201319.AA07744@dg-rtp.dg.com> From: Sonic Junior To: GNU Win 32 Date: Wednesday, August 20, 1997 4:21 AM Subject: Changing Root (/) > > Anybody here happen to know how I would change the root (/) directory? >It's current c:, but I want it to be e:\gnuwin32\unix > If you know, please contact me! Thanks! > a) change into your cygwin-32 bin directory so that you'll still be able access mount after doing (b) b) "umount /" c) "mount e:\gnuwin32\unix /" (or add the switch for binary if you want) > - SJ >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Ian@kiwiplan.co.nz Wed Aug 20 09:06:00 1997 From: Ian@kiwiplan.co.nz (Ian Collins) Date: Wed, 20 Aug 1997 09:06:00 -0000 Subject: winnt/95 authentication Message-ID: <07551A47EAD6D01186430060B025C6F40A6C86@modem1.kiwiplan2.co.nz> Is there any way of authenticating a password against windows nt server domain security in gnu-win32? I have written a network service (similiar to telnetd) that I want to have authenticated before it will accept a connection. The PC it will run on will be either winnt or win95, and I want the program to authenticate a supplied password against domain security. Any ideas anyone? Many Regards, Ian Collins - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From tzeruch@ceddec.com Wed Aug 20 09:06:00 1997 From: tzeruch@ceddec.com (root) Date: Wed, 20 Aug 1997 09:06:00 -0000 Subject: Something for the FAQ - How do I cross compile... Message-ID: <97Aug20.120611edt.32257@brickwall.ceddec.com> If I want to build the Gnu-win32 so that I can produce NT programs under Linux or another host/build machine, what do I use for the --target entry? i686-cygwin32-gnu says fcntl.h is missing, and I will try some other things, but it would really help if you would say somewhere in the faq how to do this. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From galliano@athena.polito.it Wed Aug 20 09:06:00 1997 From: galliano@athena.polito.it (Daniele Alberto GALLIANO) Date: Wed, 20 Aug 1997 09:06:00 -0000 Subject: GNU-C compiler failure Message-ID: <33E461D4.37E9@athena.polito.it> Dear sirs, I tried Your latest release of Cygnus-Win32 (b18), and I encountered a problem. I followed carefully Your installation instructions, environment settings included. When I tried to compile the most trivial C program, I get a crash of the compiler, while an older version (I changed, because I needed mmap()) is able to compile on the same machine (960712). I include a few details in the following. Looking forward for Your kind reply, Daniele Alberto GALLIANO >uname -a CYGWIN32/NT PC-BORCHIA_4 4.0 17.5 i586 >type hello.c #include main () { printf("Hello, world!!\n"); exit(0); } >gcc -v hello.c Reading specs from C:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32\cyg nus-2.7.2-970404\specs gcc version cygnus-2.7.2-970404 C:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32\cygnus-2.7.2-970404\c pp.exe -lang-c -v -iprefix C:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\i386-cygw in32\cygnus-2.7.2-970404\ -undef -D__GNUC__=2 -D__GNUC_MINOR__=7 -Di386 -D_WIN32 -DPOSIX -D__CYGWIN32__ -DWINNT -D_X86_=1 -D__STDC__=1 -D__stdcall=__attribute__ ((__stdcall__)) -D__cdecl=__attribute__((__cdecl__)) -D__i386__ -D_WIN32 -D__POS IX__ -D__CYGWIN32__ -D__WINNT__ -D_X86_=1 -D__STDC__=1 -D__stdcall=__attribute__ ((__stdcall__)) -D__cdecl=__attribute__((__cdecl__)) -D__i386 -D__POSIX -D__WINN T -Asystem(winnt) -Acpu(i386) -Amachine(i386) -Di386 -Asystem(unix) -Acpu(i386) -Amachine(i386) -D__i386__ -Asystem(unix) -Acpu(i386) -Amachine(i386) hello.c C: \temp\cc001000.i (unknown) In cygwin_except_handler (unknown) Exception trapped! (unknown) exception C0000005 at 1000C981 (unknown) exception: ax 418BA0 bx 241FB4C cx A90B0 dx 0 (unknown) exception: si 11 di 36383369 bp 241FB60 sp 241FB3C (unknown) exception is: STATUS_ACCESS_VIOLATION (unknown) Stack trace: (unknown) frame 0: sp = 0x241F970, pc = 0x10002A68 (unknown) frame 1: sp = 0x241F98C, pc = 0x77F94072 (unknown) frame 2: sp = 0x241F9B0, pc = 0x77F88A53 (unknown) frame 3: sp = 0x241FA3C, pc = 0x77F75E82 (unknown) frame 4: sp = 0x241FB60, pc = 0x10001958 (unknown) frame 5: sp = 0x241FF94, pc = 0x10001ECF (unknown) frame 6: sp = 0x241FFA0, pc = 0x4118D5 (unknown) frame 7: sp = 0x241FFB0, pc = 0x40103B (unknown) frame 8: sp = 0x241FFC0, pc = 0x77F1AFC1 (unknown) frame 9: sp = 0x241FFF0, pc = 0x0 (unknown) End of stack trace (unknown) In cygwin_except_handler (unknown) Error while dumping state (probably corrupted stack) gcc: Internal compiler error: program cpp got fatal signal 2 This is what I got from the older compiler: >gcc -v hello.c Reading specs from C:\gcc\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32/cygnus-2.7.2 -960712/specs gcc version cygnus-2.7.2-960712 C:\gcc\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32/cygnus-2.7.2-960712/cpp.exe -l ang-c -v -iprefix C:\gcc\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32/cygnus-2.7.2- 960712/ -undef -D__GNUC__=2 -D__GNUC_MINOR__=7 -Di386 -D_WIN32 -DPOSIX -D__CYGWI N32__ -DWINNT -D_X86_=1 -D__STDC__=1 -D__stdcall=__attribute__((__stdcall__)) -D __cdecl=__attribute__((__cdecl__)) -D__i386__ -D_WIN32 -D__POSIX__ -D__CYGWIN32_ _ -D__WINNT__ -D_X86_=1 -D__STDC__=1 -D__stdcall=__attribute__((__stdcall__)) -D __cdecl=__attribute__((__cdecl__)) -D__i386 -D__POSIX -D__WINNT -Asystem(winnt) -Acpu(i386) -Amachine(i386) hello.c C:\temp/cc001000.i GNU CPP version cygnus-2.7.2-960712 (80386, BSD syntax) #include "..." search starts here: #include <...> search starts here: C:\gcc\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32/cygnus-2.7.2-960712/../../../. ./../include C:\gcc\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32/cygnus-2.7.2-960712/../../../. ./i386-cygwin32/include C:\gcc\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32/cygnus-2.7.2-960712/include End of search list. C:\gcc\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32/cygnus-2.7.2-960712/cc1.exe C: \temp/cc001000.i -quiet -dumpbase hello.c -version -o C:\temp/cc001000.s GNU C version cygnus-2.7.2-960712 using the Haifa scheduler (i386-cygwin32) comp iled by GNU C version cygnus-2.7.2-960712. as -o C:\temp/cc0010001.o C:\temp/cc001000.s ld C:\gcc\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32/cygnus-2.7.2-960712/../../. ./crt0.o -LC:\gcc\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32/cygnus-2.7.2-960712 -LC:\gcc\H-i386-cygwin32\lib\gcc-lib\ -LC:\gcc\H-i386-cygwin32\lib\gcc-lib\i386- cygwin32/cygnus-2.7.2-960712/../../.. C:\temp/cc0010001.o -lcygwin - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kshaw@plight.lbin.com Wed Aug 20 13:45:00 1997 From: kshaw@plight.lbin.com (kendall shaw) Date: Wed, 20 Aug 1997 13:45:00 -0000 Subject: setting HOME Message-ID: <199708202039.NAA19485@plight.lbin.com> I've got a couple of problems. 1. Emacs wants HOME to have DOS filename conventions. If I set HOME to c:/, then when I run bash with PS1='\w ' I get no prompt at root, and "./usr" in /usr. Do you know a solution to this? 2. I have multiple accounts on my nt box. I which I could do something like HOME=c:/users/%username%, but bash doesn't know what %username% is, and cmd bash just runs cmd. Any suggestions? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From 3djokic@informatik.uni-hamburg.de Wed Aug 20 13:45:00 1997 From: 3djokic@informatik.uni-hamburg.de (Dejan) Date: Wed, 20 Aug 1997 13:45:00 -0000 Subject: bash: non English keyboard support Message-ID: <33FB560D.638425E1@informatik.uni-hamburg.de> Hello, I have tested your bash for windows, and it works really fine,but my keyboard is a german one, an some special characters like the backslash the pipe symbol and other ones are usually typed with the right ALT key pressed down, but this doesn't work with your bash. Have you any idea how to solve this problem?? -- D.D. Hamburg - Germany email: 3djokic@informatik.uni-hamburg.de or: dejan@stollmann.de - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From rjhe@oce.nl Wed Aug 20 14:12:00 1997 From: rjhe@oce.nl (Rob Hegt) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: Unix spooler References: Message-ID: <420CC47239E@fs-nw-ps.oce.nl> > Date: Mon, 18 Aug 1997 21:16:13 -0700 (PDT) > From: Alex > To: arifi_koseoglu@karmaint.com > Cc: gnu-win32@cygnus.com > Subject: Re: Re[2]: Unix spooler > On Mon, 18 Aug 1997 arifi_koseoglu@karmaint.com wrote: > > > > > > > How'bout if the printer is not on te Parallel port but on the network ? > > > > -Arifi > > Perhaps use the UNC pathname for it? > > - alex > > > - > For help on using this list (especially unsubscribing), send a message to > "gnu-win32-request@cygnus.com" with one line of text: "help". > This may all be true, but I guess that what was initially intended was an lpr/lpd suite for win32. So, what is the status on that? Rob Hegt - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Wed Aug 20 14:12:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: Unix spooler Message-ID: <01BCAC02.15865B10@gater.krystalbank.msk.ru> arifi_koseoglu@karmaint.com wrote: > > > How'bout if the printer is not on te Parallel port but on the network ? "net use lpt1 \\server\printer". -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From acherman@inf.ethz.ch Wed Aug 20 14:12:00 1997 From: acherman@inf.ethz.ch (Erwin Achermann) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: VT-100 Message-ID: <33F99C27.CE90E77D@inf.ethz.ch> Howdy all, Finally i'am as far as it is really cute to work on my NT4sp3 emacs, bash and less are all working hand in hand, handle CR LF correctly. NICE JOB you all did! But still there is a little wich i have, My $TERM is set tu "dumb" Is there a VT100 (or even better emulation arround somewhere? thanks Erwin -- |Erwin Achermann phone: +41-1-632 74 40 | |Institute of Scientific Computing fax: +41-1-632 11 72 | |ETH Zurich, IFW C29.2 mailto:acherman@inf.ethz.ch | |8092 Zurich, Switzerland http://www.inf.ethz.ch/personal/acherman | - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Fuehrer@seabase.com Wed Aug 20 14:12:00 1997 From: Fuehrer@seabase.com (Gary Fuehrer) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: SML-NJ for CygWin32 Message-ID: <6CC63E2E4FC1D011A2A700609716117A266F0A@seawolf> I have now successfully gotten SML-NJ 109.30 to build and run under CygWin32. To accomplish this, changes had to be made to both cygwin.dll (the winsup project) as well as to the SML-NJ runtime source code. The changes to SML-NJ are just to add the necessary settings for the CygWin32 platform. I am now attempting to get the SML-NJ team to incorporate those changes. Because they are just additional settings to support another platform, I anticipate the SML-NJ team will gladly accept them. The changes in winsup included a bug fix in mmap.cc and a patch to exception.cc which adds functionality. Another bug, the one I reported below in gcc's preprocessor, has a workaround for the purposes of this porting effort, but it would be very nice to get it fixed also. Below is my original posting to this news group and how I resolved or worked around each one. Here is my current list of issues and questions concerning CygWin32: 1. Will the preprocessor bug in the CygWin32 build of gcc be fixed in beta 19? (see description of bug in point 1. of previous message below). 2. Will my fix for the mmap bug I described in my posting to gnu-win32 on 8/12/97 entitled "mmap(,,,MAP_PRIVATE,,) bug and fix" be incorporated in beta 19? 3. What can we do to add my patch in exception handling that provides the sig_info and sigcontext optional arguments to signal handlers like is done in every other flavor of Unix? Without intending any offense to anyone at all, I might say that the exception handling code in the beta 18 exception.cc is rather poor and lacks the usual optional argument I need (see point 2. of previous message below). I know exception handling has changed (improved) in beta 19. I also know that Sergey Okhapkin has some patches to the exception code in beta 18. And now I have a patch that adds the optional arguments and fixes another bug (which I haven't yet reported). Is there some way that the various knowledgeable parties involved in the changes to exception.cc can coordinate in order to get a "much better" version checked into beta 19? At the very least, could we see the current beta 19 version of exception.cc? Thank you, Gary > ---------- > From: Geoffrey Noer[SMTP:noer@cygnus.com] > Sent: Thursday, July 24, 1997 2:03 AM > To: Gary Fuehrer > Cc: noer@cygnus.com > Subject: Re: FW: SML-NJ for CygWin32 > > I'm really sorry, but I'm not going to be able to look into this > in detail. Beta 19 when it comes out will fix a lot of > signals-related > issues so that may help. But several of the calls you are missing > still > haven't been implemented yet so it won't solve all of your problems... > -gjn > > Gary Fuehrer wrote: > > > > I am attempting to get Standard ML of New Jersey (version 109.29) > built > > and working under CygWin32 (version b18). CygWin32 is excellent, > and > > I'm having only minor difficulties getting this to happen. Would > > someone consider the following problems I'm having and give me some > > feedback as to how resolvable/impossible they might be? > > > > 1. SML-NJ builds an assembler source file by passing a bunch of > macro > > definitions and statements through gcc's preprocessor. For x86, > certain > > assembly statements need to have a dollar sign before a numeric > constant > > (e.g. movl $3, %eax). Combined with C language macros, the > > following kind of statements get passed to gcc when building the > SML-NJ > > runtime executable: > > #define REQ_CALLC 3 > > : > > movl $REQ_CALLC, %eax > > > > The expected output of the preprocessor is: > > movl $3, %eax > > > > With the CygWin32 version of gcc, I get the following: > > movl $REQ_CALLC, %eax > > > > It looks as if the dollar sign is being interpreted as part of the > name > > of a symbol, which is then not defined as a macro and hence remains > > un-expanded. In demonstration of this, I can add a space separator > > between the dollar sign and the macro and things work: > > movl $ REQ_CALLC, %eax ==> preprocessor ==> movl $ 3, %eax > > > > This is clearly incorrect behavior on the part of CygWin's gcc. > Dollar > > signs are not valid leading characters of C (or C++) identifiers. > > Until this is fixed in the preprocessor, the work-around for building SML-NJ is to add a space separator between the "$" and the constant. I will NOT submit this change to the SML-NJ team. They would likely reject it anyhow (I would). > > > > 2. The signal handling in CygWin32 doesn't let me do the following: > > a. Get and Set the "eip" register (or any register) of the > > excepting thread. > > b. Ascertain the kind of floating point exception that occurred. > > The need is for "siginfo_t" and "sigcontext" parameters that can be > > optionally received and modified by signal handlers. I have a patch for exception.cc in winsup to add this functionality and fix a related bug. > > > > 3. There is no SIGVTALRM signal. SML-NJ doesn't seem to need this in order to function. For now, I'm providing a dummy #define. > > > > 4. "L_ctermid" is not defined (but "L_cuserid" is). I don't remember why SML-NJ needs this. For now, I'm providing a dummy #define. It's not critical. > > > > 5. The following definitions are not in the Posix sockets include > > files (only the Windows socket stuff) > > # define SO_SNDBUF 0x1001 /* send buffer size > > */ > > # define SO_RCVBUF 0x1002 /* receive buffer > > size */ > > # define SO_SNDLOWAT 0x1003 /* send low-water > > mark */ > > # define SO_RCVLOWAT 0x1004 /* receive low-water > > mark */ > > # define SO_SNDTIMEO 0x1005 /* send timeout */ > > # define SO_RCVTIMEO 0x1006 /* receive timeout > > */ > > # define SO_ERROR 0x1007 /* get error status > > and clear */ > > # define SO_TYPE 0x1008 /* get socket type > > */ I don't remember why SML-NJ needs this. For now, I'm providing dummy #defines. It's not critical. > > > > 6. This header file doesn't exist: > > #include I don't remember why SML-NJ needs this. For now I'm leaving it out. It's not critical. > > > > 7. The following functions, declared in CygWin's header files, fail > > to link: > > getnetbyname > > getnetbyaddr > > fpgetround > > fpsetround > > pause > > getgroups > > ctermid > > mkfifo > > fchown > > fpathconf > SML-NJ doesn't seem to need these to funcion. I made a mistake about > needing fpsetround and fpgetround. Some others are in beta 19 (ie > getgroups, pause, fpathconf). Posix library funcionality in SML-NJ > for CygWin32 is reduced by what remains unimplemented (getnetbyname, > getnetbyaddr, mkfifo, fchown, ctermid). > > > > > > > I would appreciate any information you can give me on how these > problems > > will be resolved in future releases of CygWin32. > > > > Thanks, > > Gary > > > > > -- > Geoffrey Noer > noer@cygnus.com > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mark.n@usa.net Wed Aug 20 14:12:00 1997 From: mark.n@usa.net (Mark Newnham) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: Unix spooler Message-ID: <95DE24D1FCFED0118B910060940A33400462@mailhub.adacorp.com> > ---------- > From: Alex[SMTP:garbanzo@hooked.net] > Sent: Monday, August 18, 1997 10:14 PM > To: mark@adacorp.com > Cc: gnu-win32@cygnus.com; Ian@kiwiplan.co.nz; Sergey Okhapkin > Subject: Re: Re[2]: Unix spooler > > > > On Mon, 18 Aug 1997 arifi_koseoglu@karmaint.com wrote: > > > > > > > How'bout if the printer is not on te Parallel port but on the > network ? > > > > -Arifi > > Perhaps use the UNC pathname for it? > > - alex > in simple terms, if you are on an NT network, try "net use lpt1 etc etc....." , if on a novell network, "capture l1 etc etc....." before you attempt redirection to lpt1. Mark > ################################################## > # Application Design Associates, Inc # > # 6021 S.Syracuse Way #302 # > # Englewood CO 80111 Tel +1 (303)779 8958 # > # USA Fax +1 (303)779 0949 # > ################################################## - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mphalpin@iag.net Wed Aug 20 14:12:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: lynx compilation Message-ID: <199708182056.QAA28886@iag.net> I had that problem. I'm using a copy of lynx I got precompiled from Sergey. I am using win95, and hope to test it on NT soon. Maybe Sergey knows why it does that? It just keeps saying "Alert! Unable to connect to host." I've tried multiple hosts, including myself. - Sonic Junior ---------- > From: Colman Curtin > To: Sonic Junior > Cc: GNU Win 32 > Subject: lynx compilation > Date: Monday, August 18, 1997 6:04 AM > > hi, > did you use the attached patch? > Sergey posted this to the list in June. said lynx2.5* and 2.7* compiled using > make generic. > hope this helps. > > Colman. > > P.s. > I have the same problem with lynx as you but haven't had time to check it out. > are you running 95/NT? > ____________________________________________________________ > Colman Curtin mailto:ccurtin@trintech.ie > Software Engineer > ------------------------------------------------------------ > Trintech (Manufacturing) Ltd, http://www.trintech.com/ > South County Business Park, > Leopardstown, > Dublin 18. > Tel +353-1-2956766 Fax +353-1-2954735 > ------------------------------------------------------------ > > It is impossible to make anything foolproof because fools > are so ingenious. > > Edsel Murphy, dec. > > > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fox@cat.nyu.edu Wed Aug 20 14:12:00 1997 From: fox@cat.nyu.edu (David Fox) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: Microsoft Foundation Classes References: <33F893CD.13056BBE@uci.edu> Message-ID: Eric Britten writes: > > Is it possible to use MFC 4.1 using gnu-win32? If so, how? > > You need the source because VC++ and Borland C++ use a different name > mangling conventions. If you can get the source to compile, you can > then create static and dynamic link MFC libraries. DLLs with GCC are > hard to create, I have had little luck with them. How about using GCC's "ASM labels" mechanism to make inline functions that invoke the MFC procedures? Is the calling convention compatible? ] Controlling Names Used in Assembler Code ] ======================================== ] ] You can specify the name to be used in the assembler code for a C ] function or variable by writing the `asm' (or `__asm__') keyword after ] the declarator as follows: ] ] int foo asm ("myfoo") = 2; ] ] This specifies that the name to be used for the variable `foo' in the ] assembler code should be `myfoo' rather than the usual `_foo'. ] ] On systems where an underscore is normally prepended to the name of ] a C function or variable, this feature allows you to define names for ] the linker that do not start with an underscore. ] ] You cannot use `asm' in this way in a function *definition*; but you ] can get the same effect by writing a declaration for the function ] before its definition and putting `asm' there, like this: ] ] extern func () asm ("FUNC"); ] ] func (x, y) ] int x, y; ] ... ] ] It is up to you to make sure that the assembler names you choose do ] not conflict with any other assembler symbols. Also, you must not use a ] register name; that would produce completely invalid assembler code. ] GNU CC does not as yet have the ability to store static variables in ] registers. Perhaps that will be added. -- David Fox http://www.cat.nyu.edu/fox xoF divaD NYU Media Research Lab fox@cat.nyu.edu baL hcraeseR aideM UYN - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From koi@ssa.crane.navy.mil Wed Aug 20 14:12:00 1997 From: koi@ssa.crane.navy.mil (mark koi) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: Serial IO focusing Message-ID: <33FB542A.8271A54F@ssa.crane.navy.mil> I had a program under DOS which controlled a serial port using a polling method. Now I am trying to use gnu-win32, and I see no libraries for things like outp(), inp(). Can someone steer me in the right direction for examples/how-tos for controlling a serial port in WIN95. The application does not have to be windows based. Thanks and I would appreciate help cause I am lost. koi@ssa.crane.navy.mil - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From stf@hugo.int-evry.fr Wed Aug 20 14:12:00 1997 From: stf@hugo.int-evry.fr (Stephane Mariel) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: DLL problem Message-ID: <199708190947.LAA20111@hugo.int-evry.fr> Hi, I'm trying to port a unix network tool (at least the client) to Win95/NT , since there is a DLL to create, I tested the example provided on www.cygnus.com. It works fine, i got my foo.dll, main.exe, but moving to my real software when I run ld for the first time I got c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libccrt0.o)(.t ext+0x8a):libccrt0.cc: undefined reference to `GetModuleHandleA@4' c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t ext+0xc):libcmain.cc: undefined reference to `GetModuleHandleA@4' c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t ext+0x15):libcmain.cc: undefined reference to `GetCommandLineA@0' c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t ext+0x1e):libcmain.cc: undefined reference to `WinMain@16' I probably should say to ld something like, there would be a winmain elsewhere, and I will use the right winapi.dll when running the program? It's probably not very difficult, but i'm not very familiar with all the Win stuff, so any idea? Thanks, stf - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From weiku@UTKUX1.UTK.EDU Wed Aug 20 14:12:00 1997 From: weiku@UTKUX1.UTK.EDU (Wei Ku) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: Changing Root (/) Message-ID: <199708201920.PAA38476@solid.ssd.ornl.gov> umount / mount e:\\gnuwin32\\unix / Sincerely, Wei Ku *************************************** Department of Physics and Astronomy The University of Tennessee 1408 Circle Drive Knoxville, Tennessee 37996-1200 weiku@utkux.utcc.utk.edu --------------------------------------- Solid State Division Oak Ridge National Laboratory P.O.Box 2008 Oak Ridge, TN 37831-6032 Phone: (423) 574-5795 Fax: (423) 574-4143 weiku@solid.ssd.ornl.gov *************************************** -----Original Message----- From: Sonic Junior To: GNU Win 32 Date: Wednesday, August 20, 1997 2:58 AM Subject: Changing Root (/) > > Anybody here happen to know how I would change the root (/) directory? >It's current c:, but I want it to be e:\gnuwin32\unix > If you know, please contact me! Thanks! > > - SJ >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ggp@informix.com Wed Aug 20 14:12:00 1997 From: ggp@informix.com (Guy Gascoigne - Piggford) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: Microsoft Foundation Classes Message-ID: <3.0.32.19970819091602.00a51820@pop.pdx.informix.com> At 11:26 AM 8/18/97 -0700, Eric Britten wrote: >As long as GCC supports all the syntax conventions of MFC you can >compile it. I know is uses a special keyword called afx_msg; it is used >for window message handling. I am not sure if there is a work around >for this keyword. This is your biggest problem to overcome. Just in case you havn't found out, afx_msg isn't a keyword, it's defined in afxwin.h and is simply used as a place holder for message handlers. >From the header file. // Type modifier for message handlers #ifndef afx_msg #define afx_msg // intentional placeholder #endif I havn't tried compiling this stuff with gcc though, I'd be interested in knowing how it goes. Guy - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From wstewart@aa.net Wed Aug 20 14:12:00 1997 From: wstewart@aa.net (Wayne Stewart) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: expect/dejagnu for cygwin32? Message-ID: <199708182042.NAA08246@big.aa.net> Have tcl, expect, and/or dejagnu been ported to cygwin32? Thanks, Wayne Stewart wstewart@aa.net - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fjh@cs.mu.OZ.AU Wed Aug 20 14:12:00 1997 From: fjh@cs.mu.OZ.AU (Fergus Henderson) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: Tcl/Tk with extensions under cyg-win32? References: <3.0.32.19970817175328.009554b0@bl-mail2.corpeast.baynetworks.com> <199708172342.JAA03047@mundook.cs.mu.OZ.AU> <19970818092443.29230@gil.physik.rwth-aachen.de> Message-ID: <199708181650.CAA14066@murlibobo.cs.mu.OZ.AU> Christoph Kukulies wrote: >On Mon, Aug 18, 1997 at 09:42:22AM +1000, Fergus Henderson wrote: >> Check < http://www.cs.mu.oz.au/~fjh/gnu-win32/gnu-win32.html > for some > >That URL doesn't seem to exist. Could you check? Sorry, I meant < http://www.cs.mu.oz.au/~fjh/gnu-win32/ >. -- Fergus Henderson | "I have always known that the pursuit WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From spong@usa.net Wed Aug 20 14:12:00 1997 From: spong@usa.net (Matt Spong) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: MFC Message-ID: <33F9CB31.41C67EA6@usa.net> Is it possible to use MFC with cygwin32? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From demircio@boun.edu.tr Wed Aug 20 14:12:00 1997 From: demircio@boun.edu.tr (Murat Demircioglu) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: NFS Server Message-ID: <33F87C9C.4084@boun.edu.tr> Is there a free NFS server for NT ? Is it possible to port Unix NFS servers to NT with gnu-32. ? Murat Demircioglu - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From blroberts@xactware.com Wed Aug 20 14:12:00 1997 From: blroberts@xactware.com (Barry Roberts) Date: Wed, 20 Aug 1997 14:12:00 -0000 Subject: Running winNT/95 programs... Message-ID: Umm, isn't that a little like asking why would anybody ever use '&' to start an asynchronous process? Because you don't want to lose your shell while the process runs, and with non-gui apps, start is the only way to do it? I don't see what the explorer has to do with it, this IS a command line thing. The other reason is that you can get some of the explorer's smarts (like file associations) from the command line. For example, 'start refcard.ps' opens ghostview for me. Barry Roberts >>> Michael Hirmke 08/16 4:41 AM >>> Hi Narayan, > Hi, > > I read a previous thread on running wNT/95 programs through an intermediate > program. I am just a little confused. Why go through all the trouble when > there is a "start" command in both w95 and NT? I have had no problems > running any windows stuff through "start". > > For example: my alias for start on winnt: > > alias start='cmd /c start ' Sorry to ask, this might be a dumb question: Why would one want to use "start" or something like that ? Why not just run the desired application form the command line as usual ? If I want to run explorer, I just type "explorer" from the bash command line and it comes up as usual. [...] > I also have vi (elvis) aliased as: > alias vi="start vi" > so that it comes up in a separate window. Again - why ?? Bye. Michael. ! - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From judd@texcelus.com Wed Aug 20 15:21:00 1997 From: judd@texcelus.com (Judd Rogers) Date: Wed, 20 Aug 1997 15:21:00 -0000 Subject: Gnu-win32 mailing list archives: Re: bash window banner Message-ID: <33FB6DB3.D521221F@texcelus.com> Is there a way, on winNT, to get bash to set the title/banner? Like Tom (see note attached) I like to set the title to $PWD. http://www.cygnus.com/ml/gnu-win32/1997-Aug/0038.html From hartroth@telematik.informatik.uni-karlsruhe.de Wed Aug 20 15:21:00 1997 From: hartroth@telematik.informatik.uni-karlsruhe.de (Jörn Hartroth) Date: Wed, 20 Aug 1997 15:21:00 -0000 Subject: ispell port on GNU-Win32? Message-ID: <33F97112.E81F02E5@telematik.informatik.uni-karlsruhe.de> Hi everybody! I'm wondering if ispell 3.1 has been ported to Win32 using the Cygnus GNU-Win32 environment. I am aware of at least two other ports of ispell 3.1 to the DOS/Windows domain, one based on DJGPP and the other on EMX/RSX. However, neither of these worked well for me in combination with NTEmacs on Windows NT 4.0. There is also a standalone port of ispell 4.0 to Win32 but it does not offer an 8bit option required for non-english dictionaries. Any comments are appreciated. Cheers, Joern -- ======================================================================== Joern Hartroth Address: Institut fuer Telematik Tel. : (++49)(0)721 - 608 6414 Universitaet Karlsruhe (TH) Fax : (++49)(0)721 - 388097 76128 Karlsruhe Wiss. Mitarbeiter - Research Assistant Germany eMail: hartroth@ira.uka.de hartroth@telematik.informatik.uni-karlsruhe.de ======================================================================== S/MIME Cryptographic Signature -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/octet-stream Size: 2925 bytes Desc: not available URL: From fred.cox@gtri.gatech.edu Wed Aug 20 16:52:00 1997 From: fred.cox@gtri.gatech.edu (Fred Cox) Date: Wed, 20 Aug 1997 16:52:00 -0000 Subject: Installation hangs Message-ID: <3.0.32.19970820195106.00940100@globe.gtri.gatech.edu> When I try to execute cdk.exe from the latest directory, the installation program appears to hang at 99% complete. I made sure the file was ftp'ed in binary mode, so that should not be the problem. I also tried multiple downloads using different programs in case of a corrupted download. Any suggestions? Fred - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From madar@lightlink.com Wed Aug 20 18:50:00 1997 From: madar@lightlink.com (Lucas Madar) Date: Wed, 20 Aug 1997 18:50:00 -0000 Subject: Resource Compiler Message-ID: <3.0.2.32.19970820192621.00839240@light.lightlink.com> Are there any freeware resource compilers/builders out there? I've tried rcl, which gives me tons and tons of errors no matter how I try running the preprocessor.. ----- Lucas Madar / darkrot IRC Administrator, hebron.in.us.dal.net email: madar@lightlink.com / darkrot@dal.net whois: LM1691 Warning labels thwart the process of natural selection. If at first you don't succeed, well that's pretty much it for skydiving. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From tci@netcom.com Wed Aug 20 18:50:00 1997 From: tci@netcom.com (Jack Tolson) Date: Wed, 20 Aug 1997 18:50:00 -0000 Subject: DLL NT4.0 Can I be done..."Sure! With enough money and time... :)" Message-ID: <33FB9DC5.27366B45@netcom.com> Once again I am writing this mailing list to find out,???? "can you create a relocatable dll in gnuwin32 under WINNT4.0.???? I know someone from http://www.cygnus.com has tried this with the examples that are on there web site.???? I can compile the code but it will not run. The code Dr. Watson, (UNIX guys CORE).???? Is it me or is it the code? Can it be done????? I have written the new group gnu.g++.help and others but no help.???? I would have thought that this would be a common thing on a windows platform.???? I could be wrong. "NOT..." Any of you guys/gals from cygnus have an???? answer, or do I just need to buy cygnus distribution and it will work.???? I thought so... Just kidding.???? In my frustration I get sarcastic... I really need to get this working. I am using gnu on a Sun box and would love to compile the same code on my PC. HELP HELP HELP.... I know many would like this answer..... Jack Tolson Technology Concepts, Inc. tci@netcom.com Live is short, what are???? you waiting on.... From mphalpin@iag.net Wed Aug 20 18:50:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Wed, 20 Aug 1997 18:50:00 -0000 Subject: Select problem (more info) Message-ID: <199708210146.VAA20925@iag.net> Select returns -1. A perror on select() said: select: : Bad file number (That's a copy and paste. perror("select: "); ) Any idea? - SJ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From elia@itatlanta.com Wed Aug 20 18:50:00 1997 From: elia@itatlanta.com (Paul Vincent Elia) Date: Wed, 20 Aug 1997 18:50:00 -0000 Subject: Changing Root (/) References: <199708192236.SAA15064@iag.net> Message-ID: <3.0.3.32.19970820200625.006a9148@pop.mindspring.com> At 06:37 PM 8/19/97 -0400, Sonic Junior wrote: > > Anybody here happen to know how I would change the root (/) directory? >It's current c:, but I want it to be e:\gnuwin32\unix You use the "mount" utility to map drives to paths. But I have noticed that you can't redefine "c:" from the command line. Here's what works -- use regedit.exe, go to "My Computer\HKEY_USERS\.Default\Software\Cygnus Solutions\CYGWIN.DLL setup\b15.0\mounts\" and hunt for the "0" mount point that corresponds to "c:" and "/". Change the "c:" to "e:\gnuwin32\unix" and everything should work. You can then run mount from a command prompt to be sure. Note that if you have multiple logins on the same PC, you may need to hunt for other "Cygnus Solutions" settings, too. Paul V. Elia President and Senior Consultant IT Atlanta, Inc., Information Technology Consultants 4060 Peachtree Road, Suite D-304, Atlanta, GA 30319-3020 main: (404) 467-8485 fax: (404) 846-9124 Temporary WWW: http://www.mindspring.com/~elia/itatlanta.html Coming Soon: http://www.itatlanta.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Ian@kiwiplan.co.nz Wed Aug 20 22:03:00 1997 From: Ian@kiwiplan.co.nz (Ian Collins) Date: Wed, 20 Aug 1997 22:03:00 -0000 Subject: Winnt/95 authentication Message-ID: <07551A47EAD6D01186430060B025C6F40A6C8D@ntserver.kiwiplan2.co.nz> Is there any way of authenticating a password against windows nt server domain security in gnu-win32? I have written a network service (similiar to telnetd) that I want to have authenticated before it will accept a connection. The PC it will run on will be either winnt or win95, and I want the program to authenticate a supplied password against domain security. Any ideas anyone? Many Regards, Ian Collins - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dgold@brown.edu Wed Aug 20 22:04:00 1997 From: dgold@brown.edu (Dan Goldwater) Date: Wed, 20 Aug 1997 22:04:00 -0000 Subject: how to build a dll that uses other dll's? Message-ID: <33FBCC2D.98CBE385@brown.edu> Hi! I am trying to build a DLL which has function calls which must be resolved into existing win32 DLL's. Following the instructions for building a relocatable dll, i get an output but each of the three ld steps reports all the undefined references for the functions which are in the existing win32 dll's. i had to add the --noinhibit-exec option to ld so that it wouldn't delete the output dll due to the errors. when i build this project on solaris, the linker doesn't complain about the undefined references, and they are resolved to appropriate existing .so files at runtime. is there some option i need to give ld so it knows i want to resolve the references at runtime to the existing dll's? or am i clueless? thanks! -- --------------------------------------------------------------------- Dan Goldwater dgold@brown.edu http://fatmac.ee.cornell.edu/~goldwada --------------------------------------------------------------------- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From khan@xraylith.wisc.edu Thu Aug 21 00:42:00 1997 From: khan@xraylith.wisc.edu (Mumit Khan) Date: Thu, 21 Aug 1997 00:42:00 -0000 Subject: Something for the FAQ - How do I cross compile... References: <97Aug20.120611edt.32257@brickwall.ceddec.com> Message-ID: <199708210648.BAA04701@eh_pc11.xraylith.wisc.edu> root writes: > If I want to build the Gnu-win32 so that I can produce NT programs > under Linux or another host/build machine, what do I use for the > --target entry? i686-cygwin32-gnu says fcntl.h is missing, and I > will try some other things, but it would really help if you would > say somewhere in the faq how to do this. > - I remember a post by Geoffrey Noer a long time back with the essentials, and I've successfully used that since b17 to build everything under i386-linux. I'll use i386-linux instead of the more specific i[456] etc. I do this regularly and haven't had any problems. Here's an excerpt from a message that I had sent someone else (replace b17 with 18.1): -- using template mhl.format -- From: Mumit Khan Subject: BUILDING CROSS COMPILER You really need to start with getting the whole source tree for cygnus gnu-win32 b17.1 from cygnus's ftp site. You of course do need a C compiler already installed on your machine (eg., gcc) to bootstrap the cross compiler. GCC 2.7.2* work quite well when building the cross compiler. You don't need the source for gcc other than what comes with gnu-win32 source distribution (there is gcc in there already, as well as a whole bunch of other things). Here's what I've done: Part 1: Building the cross-compiler 1. Get Cygnus b17.1 source tree for the compiler toolchain ftp://ftp.cygnus.com/pub/gnu-win32/cdk-src/* Comes as a bunch of files xaa, xab, etc. 2. Unzip and untar on your Unix machine (say a Linux machine). Let's assume you untar everything under $srcdir (where $srcdir in my case is /src/gnu/gnu-win32-b17.1) % mkdir gnu-win32-b17.1 % cd gnu-win32-b17.1 % cat ../xa* | tar zxvf - 3. Now create a build tree. % mkdir cross-build % cd cross-build 4. Configure for cross compiler % $srcdir/configure --target=i386-cygwin32 --host=i386-linux \ -v --prefix=/scratch/gnu Replace /scratch/gnu above with what you want. This is where the cross-compilers will eventually end up! 5. Make the cross-compiler now % make >& make.log % make install Part 2: [OPTIONAL] Building the native tools 1. Add $prefix/bin to your path <<<<<<<< Make sure you do this! 2. Create a new build tree % mkdir cygwin32 % cd cygwin32 3. Configure for native tools % $srcdir/configure --prefix=/scratch/gnu-win32 \ --exec-prefix=/scratch/gnu-win32/H-i386-cygwin32 \ --target=i386-cygwin32 --host=i386-cygwin32 \ --build=i386-linux -v 4. ``make'' and ``make install'' Part 3: [OPTIONAL] Build native user tools. [... same as Part 2 ... ] -- regards, Mumit -- khan@xraylith.wisc.edu http://www.xraylith.wisc.edu/~khan/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From hartroth@telematik.informatik.uni-karlsruhe.de Thu Aug 21 00:42:00 1997 From: hartroth@telematik.informatik.uni-karlsruhe.de (Jörn Hartroth) Date: Thu, 21 Aug 1997 00:42:00 -0000 Subject: ispell 3.1 ported to GNU-Win32? Message-ID: <33FBF1A3.FB050D@telematik.informatik.uni-karlsruhe.de> Hi everybody! I'm wondering if ispell 3.1 has been ported to WinNT 4.0 using the GNU-Win32 library. I am aware of two other ports based on EMX/RSX and DJGPP, but I haven't gotten either to work with NTEmacs apparently due to problems in the process spawning mechanism. Any comments are appreciated. Cheers, Joern -- ======================================================================== Joern Hartroth Address: Institut fuer Telematik Tel. : (++49)(0)721 - 608 6414 Universitaet Karlsruhe (TH) Fax : (++49)(0)721 - 388097 76128 Karlsruhe Wiss. Mitarbeiter - Research Assistant Germany eMail: hartroth@ira.uka.de hartroth@telematik.informatik.uni-karlsruhe.de ======================================================================== - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From hubicka@horac.ta.jcu.cz Thu Aug 21 02:33:00 1997 From: hubicka@horac.ta.jcu.cz (Jan Hubicka) Date: Thu, 21 Aug 1997 02:33:00 -0000 Subject: Small czech article about GNU-WIN32 Message-ID: <19970821114502.36419@horac.ta.jcu.cz> Hi I am maitainer of czech GNU homepage and I think GNU-win32 project of one of thinks that worth to note there. But I don't have it (I even don't have windows) so I can't test it and I don't want to write about something I don't know. Please if you are able to write small article in Czech about this project (like one that is already available about djgpp, Linux, hurd etc) please contact me at my email(hubicka@paru.cas.cz) Homepage is at http://www.ta.jcu.cz/gnu/gnu.html Thanks Honza -- ------------------------------------------------------------------------------ Have you browsed my www pages? Look at: http://www.paru.cas.cz/~hubicka Koules-the game for Svgalib,X11 and OS/2, Xonix-the game for X11 czech documentation for linux index, original 2D computer art and funny 100 years old photos and articles are there! - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fjh@cs.mu.OZ.AU Thu Aug 21 02:33:00 1997 From: fjh@cs.mu.OZ.AU (Fergus Henderson) Date: Thu, 21 Aug 1997 02:33:00 -0000 Subject: DLL Problem References: <199708201219.OAA01103@hugo.int-evry.fr> Message-ID: <199708210837.SAA12023@mundook.cs.mu.OZ.AU> "Stephane Mariel" writes: >I'm trying to port a tcl module from unix to win95 as a dll to load within >wish. > >ext+0x1e):libcmain.cc: undefined reference to `WinMain@16' > >This seems to be a reentrancy problem, since a simple call like : >fprintf(stderr, ...) >makes use of _impure_ptr which is a struct _reent* ptr that is provided by >any process using the dll. So , probably the exact question is : how to say >to ld: this symbol must not be search in libc, (libccrto.o, libcmain.o) but >must be mapped to a symbol defined in the calling process? With beta-18, you need to define _impure_ptr in your DLL. See < http://www.cs.mu.oz.au/~fjh/gnu-win32/how-to-build-dlls.html >. An extract: # The DLL may also contain some references to _impure_ptr # (e.g. stdin is defined as a macro which expands to _impure_ptr.stdin). # We need to provide a definition for this (otherwise it will link in # the definition in libccrt.o, which causes lots of problems, # eventually leading to undefined symbol `WinMain'). # The main program needs to initialize all the _impure_ptr variables # for the DLLs with its _impure_ptr. -- Fergus Henderson | "I have always known that the pursuit WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bridge@dial-up.net Thu Aug 21 04:09:00 1997 From: bridge@dial-up.net (Leonard Weincier) Date: Thu, 21 Aug 1997 04:09:00 -0000 Subject: SLIP Server ??? Message-ID: Hi I am looking for a SLIP server for NT and I was wondering if anyone on this list knows of one. TIA Len - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Minh.Le_Cong@spacebel.be Thu Aug 21 04:09:00 1997 From: Minh.Le_Cong@spacebel.be (Minh Le Cong) Date: Thu, 21 Aug 1997 04:09:00 -0000 Subject: CYGNUS-WIN32 and Oracle OCI7.2 Library Message-ID: <01BCAE2D.BD3771A0@pcmlg> Hello all, I'm new to CYGNUS-WIN32 and I have one question involved CYGNUS-WIN32. As you may know, Personal Oracle 7.2 for NT include an API called OCI (Oracle Call Interface). This is used to access the Oracle database. This is a C API . It consists of some .h files , a DLL (ociw32.dll) and OCIw32.lib The ociw32.lib is intended for MSVC users. So my question is that how can I link my program with OCI library if I use CYGNUS-WIN32 ? how gcc and ld deal with .LIB libary files ( not .a library file)? My email address is mlg@spacebel.be Thank you very much for any help, Regards, mlg - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mreinelt@sime.com Thu Aug 21 04:09:00 1997 From: mreinelt@sime.com (Michael Reinelt) Date: Thu, 21 Aug 1997 04:09:00 -0000 Subject: tar from/to floppy disk Message-ID: <199708210948.LAA00373@artus.avalon> I'm trying to get tar to read/write from floppy disks. (It's the only NT tar which _should_ work with floppies). As I said, it only should. With NT4 Server, the floppy starts up, and after a few seconds, I get the following: tar: Cannot open /dev/fd0: Not super-user (Win95 refuses to know about /dev/fd0.) Already working as administrator, I tried to become super-user :-) I created /etc/passwd, and changed the UID of administrator to 0. I added a new user "root", again with UID 0. Doesn't help. Any hints? bye, Michael -- Michael Reinelt vox: +43 316 577288 Algersdorferstr. 66a fax: +43 316 577852 A-8020 Graz, Austria email: mreinelt@sime.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kuku@gilberto.physik.RWTH-Aachen.DE Thu Aug 21 04:14:00 1997 From: kuku@gilberto.physik.RWTH-Aachen.DE (Christoph Kukulies) Date: Thu, 21 Aug 1997 04:14:00 -0000 Subject: building winsup (cross) Message-ID: <199708211117.NAA08246@gil.physik.rwth-aachen.de> Assuming a resource compiler for cross building of res files is in the subdirectory cdk/winsup I'm trying to build --target=i386-cygwin32 and compiling it on the cross build machine yields: gcc -c -g -O2 -I./msinclude -I./include -I. -I./../libstdc++/stl -I./../libio -I../libio -I./../newlib/libc/include -nostdinc++ ../librx/rx.c In file included from ../librx/rx.c:41: ../newlib/libc/include/stdio.h:144: parse error before `__gnuc_va_list' ../newlib/libc/include/stdio.h:145: parse error before `__gnuc_va_list' ../newlib/libc/include/stdio.h:146: parse error before `__gnuc_va_list' ../newlib/libc/include/stdio.h:174: parse error before `__gnuc_va_list' ../newlib/libc/include/stdio.h:219: parse error before `__gnuc_va_list' ../newlib/libc/include/stdio.h:220: parse error before `__gnuc_va_list' ../newlib/libc/include/stdio.h:221: parse error before `__gnuc_va_list' ../librx/rx.h: In function `rx_search': In file included from ../librx/rx.c:111: ../librx/rx.h:3635: warning: passing arg 1 of `bcopy' from incompatible pointer type ../librx/rx.h:3635: warning: passing arg 2 of `bcopy' from incompatible pointer type ../librx/rx.h:3637: warning: passing arg 1 of `bcopy' from incompatible pointer type ../librx/rx.h:3637: warning: passing arg 2 of `bcopy' from incompatible pointer type ^Cgmake: *** [../librx/rx.o] Interrupt Anyone knowing what the problem is here? ../newlib/libc/include/stdio.h has: #ifdef __GNUC__ #define __VALIST __gnuc_va_list #else #define __VALIST char* #endif -- Chris Christoph P. U. Kukulies kuku@gil.physik.rwth-aachen.de - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From koi@ccmail.ssa.crane.navy.mil Thu Aug 21 07:17:00 1997 From: koi@ccmail.ssa.crane.navy.mil (Mark Koi) Date: Thu, 21 Aug 1997 07:17:00 -0000 Subject: Serial IO drivers Message-ID: <33FC4DC3.20E29B81@ccmail.ssa.crane.navy.mil> Hi, I had a working serial IO driver working with a polling method using DJGPP. Porting to gnuwin32 is not straight forward. Are there similar library calls for outp and inp? Were can I go to learn how to write a serial driver for gnuwin32 or does somebody have an example? Should I write a device driver under linux for serial control, and then would it port painlessly to gnuwin32? My brain is scrambled, and I am new to gnuwin32. Any help is appreciated. -- Mail: ATTN: Mark Koi (Code: 7027, Bldg: 2036) NAVSURFWARCENDIV CRANE Voice:812-854-1167 300 HIGHWAY 361 Fax: 812-854-3437 Crane, Indiana 47522-5001 Email: Mark.Koi@ssa,crane,navy,mil URL : http://www,ssa,crane,navy,mil/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dan.rearick@cdmspa.com Thu Aug 21 09:04:00 1997 From: dan.rearick@cdmspa.com (Dan Rearick) Date: Thu, 21 Aug 1997 09:04:00 -0000 Subject: Include Path Error Message-ID: <01BCAE2B.39591F20@drearick> Problem: I am attempting to compile my first GNUWin32 program, the infamous HELLO.C, to verify the installation of GNUWin32 under Windows 95 and I am receiving the following error: Entered: gcc hello.c Results: hello.c:1: No include path in which to find stdio.h GCC.EXE: Internal compiler error: program cpp got fatal signal 33 Below is a listing of my environment parameters obtained with the Win95/DOS SET command: /* Begin parameters list */ TMP=C:\WINDOWS\TEMP TEMP=C:\WINDOWS\TEMP PROMPT=$p$g winbootdir=C:\WINDOWS COMSPEC=C:\WINDOWS\COMMAND.COM NESTORPATH=C:\OCRFF\NESTOR30 GCC_EXEC_PREFIX = C:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\ TCL_LIBRARY = C:/gnuwin32/b18/tcl/lib/tcl7.6 GDBTK_LIBRARY = C:/gnuwin32/b18/share/gdbtcl PATH=C:\WINDOWS;C:\WINDOWS\COMMAND;C:\MSSQL\BIN;C:\UTL;C:\PERL\BIN;C:\OPTIALL;C: \KIPP\BIN;C:\PROGRAMF\MTI\OCRFF_16;C:\OCRFF\OCR\WORDSCAN.30;C:\OCRFF\KIPP\BIN;C: \GNUWIN32\B18\H-I386-CYGWIN32\BIN;C:\GNUWIN32\B18\TCL\BIN;C:\GNUWIN32\B18\H-I386 -CYGWIN32\LIB\GCC-LIB\I386-CYGWIN32\CYGNUS-2.7.2-970404;C:\GNUWIN32\B18\H-I386-C YGWIN32\I386-CYGWIN32\INCLUDE windir=C:\WINDOWS SNDSCAPE=C:\WINDOWS BLASTER=A220 I5 D1 T1 CMDLINE=gcc hello.c /* End parameters list */ Would someone please enlighten me as to what is missing. The software was installed from a full version of GNUWin32 (CDK.EXE, size 14MB) and the installation completed error free. Thanks. DBR... - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jla@arceneaux.com Thu Aug 21 17:22:00 1997 From: jla@arceneaux.com (Joseph Arceneaux) Date: Thu, 21 Aug 1997 17:22:00 -0000 Subject: Off topic filesystem question Message-ID: <199708220021.RAA13778@emptiness.arceneaux.com> This is not really a cygwin question, so I must apologize in advance. I have a Unix file system which I'm trying to render onto a CD. The CD burner only runs off a Windows system (or potentially, NT). The Question is how to preserve the many softlinks in the file system so they appear properly on the CD. Any suggestions would be appreciated. Joe - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From colin@bird.fu.is.saga-u.ac.jp Thu Aug 21 21:41:00 1997 From: colin@bird.fu.is.saga-u.ac.jp (Colin Peters) Date: Thu, 21 Aug 1997 21:41:00 -0000 Subject: GNU-C compiler failure Message-ID: <01BCAF00.B9C6B700@gbird0> Daniele Alberto GALLIANO[SMTP:galliano@athena.polito.it] wrote: [snip] >>type hello.c >#include > >main () >{ > printf("Hello, world!!\n"); > exit(0); >} > [snip] >>gcc -v hello.c >(unknown) In cygwin_except_handler Just a suggestion, but I think you should check and see if you might have any old copies of cygwin.dll lying around on your PATH or in your windows or windows/system directories. Colin. -- Colin Peters - Saga Univ. Dept. of Information Science -- colin@bird.fu.is.saga-u.ac.jp - finger for PGP public key -- http://www.fu.is.saga-u.ac.jp/~colin/index.html -- http://www.geocities.com/Tokyo/Towers/6162/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From elia@itatlanta.com Thu Aug 21 21:41:00 1997 From: elia@itatlanta.com (Paul Vincent Elia) Date: Thu, 21 Aug 1997 21:41:00 -0000 Subject: winnt/95 authentication References: <07551A47EAD6D01186430060B025C6F40A6C86@modem1.kiwiplan2.co.nz> Message-ID: <3.0.3.32.19970821192623.0069e5d8@pop.mindspring.com> At 05:33 PM 8/20/97 +1200, Ian Collins wrote: >Is there any way of authenticating a password against windows nt server >domain security in gnu-win32? > >I have written a network service (similiar to telnetd) that I want to >have authenticated before it will accept a connection. > >The PC it will run on will be either winnt or win95, and I want the >program to authenticate a supplied password against domain security. > >Any ideas anyone? > >Many Regards, >Ian Collins > I don't know the answer, but I've got two cents to put in on this one. It would be great if getpwnam() and its related functions worked. This would be a great path towards authenticating on Win95/WinNT while still keeping Unix-sh ties. - Paul Elia - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From geier@forwiss.uni-passau.de Thu Aug 21 21:41:00 1997 From: geier@forwiss.uni-passau.de (Roland Geier) Date: Thu, 21 Aug 1997 21:41:00 -0000 Subject: DLL Problem References: <199708201219.OAA01103@hugo.int-evry.fr> Message-ID: <199708212048.WAA07599@bombadill.forwiss.uni-passau.de> > I'm trying to port a tcl module from unix to win95 as a dll to load within > wish. > > So i tried to began with the example provided on cygnus.com , it works > fine, i got my foo.dll, main.exe, but moving to my real software when I run > ld I got: > c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libccrt0.o)(.t > > ext+0x8a):libccrt0.cc: undefined reference to `GetModuleHandleA@4' > c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t > > ext+0xc):libcmain.cc: undefined reference to `GetModuleHandleA@4' > c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t > > ext+0x15):libcmain.cc: undefined reference to `GetCommandLineA@0' > c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t > > ext+0x1e):libcmain.cc: undefined reference to `WinMain@16' > > This seems to be a reentrancy problem, since a simple call like : > fprintf(stderr, ...) ...for me this seems to be a forgotten '-lkernel32' at link-time :). The undefined ref to 'WinMain' can be avoided with: echo 'int main() { return 0; }' > fixup.c; echo 'asm (".section .idata$3\n" ".long 0,0,0,0,0,0,0,0");' >> fixup.c; gcc -c fixup.c # link fixup.o with your dll... hth, --- Roland. --- Roland Geier Phone: ++49 (8505) 92182 Universitaet Passau, GER Fax: ++49 (8505) 922311 Lehrstuhl Prof. Dr. Donner geier@forwiss.uni-passau.de D-94032 Passau http://www.uni-passau.de/~geier/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From iverson@cisco.com Thu Aug 21 22:56:00 1997 From: iverson@cisco.com (Tim Iverson) Date: Thu, 21 Aug 1997 22:56:00 -0000 Subject: Serial IO focusing References: <33FB542A.8271A54F@ssa.crane.navy.mil> Message-ID: <199708220548.WAA17194@stars.cisco.com> Cygnus may have some of these already coded in their machine.h file, but if you don't find it there, most folks usually just write their own using GCC's inline asm; eg. in your very own machine.h file ... static unsigned char inline inb (const unsigned short port) { register unsigned char val; __asm __volatile ("inb %%dx,%%al" : "=a" (val) : "d" (port)); return val; } If I managed to remember the gas syntax for inb correctly, { int x = inb(0x3f8); } Will produce the following assembly movw %dx, $0x3f8 inb %dx, %al movzbl %al, -4(%epb) Or something quite similar -- GCC is smart enough to keep 0x3f8 in a register and not reload it every time if you're in a loop. It will also skip the move to automatic storage if it can and just use the value in AL. There is some cryptic documentation on inline asm for GCC in the info files. For the details, you will need to read the source -- particularly the machine description file in gcc/config/i386/i386.md. This is the only place you can find the descriptions for the various register fields (eg. the "=a") to use in the asm lines. - Tim +---------------- | Date: Wed, 20 Aug 1997 16:31:39 -0400 | From: mark koi | To: gnu-win32@cygnus.com | Subject: Serial IO focusing | | method. Now I am trying to use gnu-win32, and I see no libraries for | things like outp(), inp(). Can someone steer me in the right direction | for examples/how-tos for controlling a serial port in WIN95. The - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From seibold@ind.uni-stuttgart.de Thu Aug 21 22:56:00 1997 From: seibold@ind.uni-stuttgart.de (Wolfram Seibold) Date: Thu, 21 Aug 1997 22:56:00 -0000 Subject: fseek bug in b18 Message-ID: <33FD2A11.4A31@ind.uni-stuttgart.de> Hi, I had no success in using the fseek function under Windows NT 4.0 on NTFS file system. My workaround is reading n bytes from the beginning of the file, but this is very slow. Any ideas? Thanks. Wolfram Seibold -- ----------------------------------------------------------------------------- !ACHTUNG ADRESSAENDERUNG! !NEW ADDRESS! ----------------------------------------------------------------------------- Universitaet Stuttgart University of Stuttgart Institut fuer Nachrichtenvermittlung Institute of Communication und Datenverarbeitung Networks and Computer Engineering Prof. Dr.-Ing. Dr. h. c. P. J. Kuehn Prof. Dr.-Ing. Dr. h. c. P. J. Kuehn Pfaffenwaldring 47 Pfaffenwaldring 47 70569 Stuttgart 70569 Stuttgart Germany Wolfram Seibold Tel. : +49 711 685 7965 Fax : +49 711 685 7983 EMail: seibold@ind.uni-stuttgart.de WWW : http://www.ind.uni-stuttgart.de/IND/MA/Se/ ----------------------------------------------------------------------------- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From noer@cygnus.com Fri Aug 22 01:02:00 1997 From: noer@cygnus.com (Geoffrey Noer) Date: Fri, 22 Aug 1997 01:02:00 -0000 Subject: expect/dejagnu for cygwin32? References: <199708182042.NAA08246@big.aa.net> Message-ID: <199708220802.BAA28089@rtl.cygnus.com> Wayne Stewart wrote: > > Have tcl, Yes, Unix-style configury for building the Win port of tcl will be present in the next beta release. > expect, and/or dejagnu been ported to cygwin32? With the pty support that will be in the next release, we'll be almost there. I tried briefly to get expect working but there are still bugs to fix before it'll work correctly. -- Geoffrey Noer noer@cygnus.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sl9y3@cc.usu.edu Fri Aug 22 02:58:00 1997 From: sl9y3@cc.usu.edu (Sandeep Neema) Date: Fri, 22 Aug 1997 02:58:00 -0000 Subject: make cacking! Message-ID: <33FD61E2.77D0341E@cc.usu.edu> hi all, excuse my ignorance but the make that comes with b18 distribution always cacks on all my makefiles with error: Makefile:: *** missing seperator. Stop. I am lost. please help!! - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ml@ddci.dk Fri Aug 22 04:06:00 1997 From: ml@ddci.dk (Mikkel Lauritsen) Date: Fri, 22 Aug 1997 04:06:00 -0000 Subject: Q: gnu-win32 retargetability Message-ID: <33FD70C7.42877E5C@ddci.dk> Hi all, I'm new to the gnu-win32 tools, so this might be a somewhat stupid question. Any pointers to information will be much appreciated; I've been going through the FAQ and some of the mailing list archives to no avail. What I would like to know is if the gnu-win32 tools support cross development, that is handling non-native file formats on NT? One of the advantages of the gnu binutils is that they are very easy to retarget - just compile with --target=whatever, and you're all set. Do the gnu-win32 tools also support this in a simple way, or does retargeting require a larger effort? TIA Mikkel Lauritsen -- Mikkel Lauritsen E-mail: ml@ddci.dk DDC-I A/S Phone: +45 4587 2077 ext. 275 Gl. Lundtoftevej 1B Fax: +45 4587 2217 DK-2800 Lyngby, Denmark - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fjh@cs.mu.OZ.AU Fri Aug 22 04:06:00 1997 From: fjh@cs.mu.OZ.AU (Fergus Henderson) Date: Fri, 22 Aug 1997 04:06:00 -0000 Subject: Serial IO focusing References: <33FB542A.8271A54F@ssa.crane.navy.mil> Message-ID: <199708220915.TAA03323@murlibobo.cs.mu.OZ.AU> mark koi, you wrote: > I had a program under DOS which controlled a serial port using a polling > > method. Now I am trying to use gnu-win32, and I see no libraries for > things like outp(), inp(). I think you probably just need to define these using inline assembler. #define outb(value, port) \ ({ __asm__ __volatile__ ("outb %b0,%w1"::"a" (value), "d" (port)); }) #define inb(port) \ ({ char _v; \ __asm__ __volatile__ ("inb %w1,%0":"=a" (_v):"d" (port)); \ _v; \ }) However, I haven't tried this... I don't know if Windows allows direct access to I/O ports like this. You might have to write a Windows device driver (VxD). -- Fergus Henderson | "I have always known that the pursuit WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From newsham@aloha.net Fri Aug 22 04:06:00 1997 From: newsham@aloha.net (Tim Newsham) Date: Fri, 22 Aug 1997 04:06:00 -0000 Subject: system() bug Message-ID: <199708212115.LAA20858@haleakala.aloha.net> Hi, I've noticed occasional problems with system() before. Well at the moment the problem seems to be happening every time so I figured I'd take the occasion to get some debugging info. The test program I used is: main() { system("/winnt/system32/arp -a"); printf("done!\n"); } At the moment, every time I run this program I get a crash. (Normally crashes are very rare, I don't know what triggers it or why all of the sudden this crashes every time). The output when run is: Interface: 10.200.200.4 on Interface 2 Internet Address Physical Address Type 10.200.200.3 00-60-97-da-7d-2b dynamic (c:\a.exe 1958) In cygwin_except_handler (c:\a.exe 1958) Exception trapped! (c:\a.exe 1958) exception C0000008 at 77F75EF8 (c:\a.exe 1958) exception: ax C0000008 bx 6 cx A01 dx FFFFFFFF (c:\a.exe 1958) exception: si 401060 di 240EF44 bp 240E610 sp 240E5C0 (c:\a.exe 1958) Stack trace: (c:\a.exe 1958) frame 0: sp = 0x240E400, pc = 0x1000F2B1 (c:\a.exe 1958) frame 1: sp = 0x240E418, pc = 0x77F94072 (c:\a.exe 1958) frame 2: sp = 0x240E43C, pc = 0x77F88A53 (c:\a.exe 1958) frame 3: sp = 0x240E4C8, pc = 0x77F75E82 (c:\a.exe 1958) frame 4: sp = 0x240E610, pc = 0x77F6766F (c:\a.exe 1958) frame 5: sp = 0x240EF60, pc = 0x1002B07D (c:\a.exe 1958) frame 6: sp = 0x240EF9C, pc = 0x1002B50D (c:\a.exe 1958) frame 7: sp = 0x240F3C0, pc = 0x1002B32F (c:\a.exe 1958) frame 8: sp = 0x240F3D8, pc = 0x1002FD47 (c:\a.exe 1958) frame 9: sp = 0x240F404, pc = 0x401072 (c:\a.exe 1958) frame 10: sp = 0x240F410, pc = 0x1000DF08 (c:\a.exe 1958) frame 11: sp = 0x240FF94, pc = 0x1000DF38 (c:\a.exe 1958) frame 12: sp = 0x240FFA0, pc = 0x4011DF (c:\a.exe 1958) frame 13: sp = 0x240FFB0, pc = 0x40103B (c:\a.exe 1958) frame 14: sp = 0x240FFC0, pc = 0x77F1AFC1 (c:\a.exe 1958) frame 15: sp = 0x240FFF0, pc = 0x0 (c:\a.exe 1958) End of stack trace The output of STRACE is included below. The most relevant lines seem to be: (c:\a.exe 1958) wait_for_single: 0 = WFMO (104, -1) (c:\a.exe 1958) lock_pinfo_for update : locking handle 44, timeout -1 (c:\a.exe 1958) wait_found: result = 100000 (c:\a.exe 1958) unlock_pinfo : handle 44 (c:\a.exe 1958) wait_found: reparenting (c:\a.exe 1958) lock_pinfo_for update : locking handle 44, timeout -1 (c:\a.exe 1958) unlock_pinfo : handle 44 (c:\a.exe 1958) wait_for_single: child pid 1959 (c:\a.exe 1958) wait_for_single: 0 = WFMO (160, -1) (c:\a.exe 1958) lock_pinfo_for update : locking handle 44, timeout -1 (c:\a.exe 1958) wait_found: result = 0 (c:\a.exe 1958) unlock_pinfo : handle 44 (c:\a.exe 1958) wait_found: 1959 = waitpid (-1, 0x240EF44, 0) (0x0) (c:\a.exe 1958) In cygwin_except_handler sig = 7 (c:\a.exe 1958) _exit (2) Tim N. ---- (unknown) open_shared_file_map: s = 0xA000000, h = 24 (unknown) lock_pinfo_for update : locking handle 0, timeout -1 (unknown) lock_pinfo_for update : creating mutex = 44 (C:\bin\sh.exe 1958) copy_shared_fd_table: pid = 1958, name = cygwin32.fd_map.17.1958 (C:\bin\sh.exe 1958) copy_shared_fd_table: area = 3670000, old p->hmap.vec = 2ECA004 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: buf = 3670004 (C:\bin\sh.exe 1958) _sbrk (0) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E70000, end 0x3670000, size 8388608, avail 8388608 (C:\bin\sh.exe 1958) 0x2E70000 = sbrk (0x0) (total 0x0) (C:\bin\sh.exe 1958) _sbrk (8192) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E70000, end 0x3670000, size 8388608, avail 8388608 (C:\bin\sh.exe 1958) 0x2E70000 = sbrk (0x2000) (total 0x2000) (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: number_of_fds = 64, vec = 2E70004 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: open fd = 0, /dev/conin (C:\bin\sh.exe 1958) _sbrk (0) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E72000, end 0x3672000, size 8388608, avail 8380416 (C:\bin\sh.exe 1958) 0x2E72000 = sbrk (0x0) (total 0x2000) (C:\bin\sh.exe 1958) _sbrk (2048) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E72000, end 0x3672000, size 8388608, avail 8380416 (C:\bin\sh.exe 1958) 0x2E72000 = sbrk (0x800) (total 0x2800) (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: len = 113 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: open fd = 1, out (C:\bin\sh.exe 1958) _sbrk (0) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E72800, end 0x3672800, size 8388608, avail 8378368 (C:\bin\sh.exe 1958) 0x2E72800 = sbrk (0x0) (total 0x2800) (C:\bin\sh.exe 1958) _sbrk (2048) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E72800, end 0x3672800, size 8388608, avail 8378368 (C:\bin\sh.exe 1958) 0x2E72800 = sbrk (0x800) (total 0x3000) (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: len = 34 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: open fd = 2, out (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: len = 34 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 3 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 4 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 5 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 6 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 7 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 8 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 9 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 10 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 11 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 12 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 13 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 14 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 15 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 16 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 17 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 18 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 19 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 20 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 21 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 22 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 23 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 24 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 25 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 26 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 27 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 28 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 29 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 30 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 31 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 32 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 33 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 34 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 35 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 36 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 37 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 38 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 39 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 40 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 41 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 42 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 43 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 44 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 45 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 46 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 47 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 48 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 49 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 50 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 51 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 52 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 53 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 54 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 55 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 56 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 57 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 58 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 59 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 60 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 61 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 62 (C:\bin\sh.exe 1958) hinfo_vec::de_linearize_fd_array: closed fd = 63 (C:\bin\sh.exe 1958) copy_shared_fd_table: new p->hmap.vec = 2E70004 (C:\bin\sh.exe 1958) unlock_pinfo : handle 44 (C:\bin\sh.exe 1958) _sbrk (0) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E73000, end 0x3673000, size 8388608, avail 8376320 (C:\bin\sh.exe 1958) 0x2E73000 = sbrk (0x0) (total 0x3000) (C:\bin\sh.exe 1958) _sbrk (2048) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E73000, end 0x3673000, size 8388608, avail 8376320 (C:\bin\sh.exe 1958) 0x2E73000 = sbrk (0x800) (total 0x3800) (C:\bin\sh.exe 1958) open (/etc/passwd, 0x20000) (C:\bin\sh.exe 1958) fhandler_disk_file::open (/etc/passwd, 0x20000) (C:\bin\sh.exe 1958) conv_to_win32_path (/etc) (C:\bin\sh.exe 1958) /etc = normalize_posix_path (/etc) (C:\bin\sh.exe 1958) c:\etc = conv_to_win32_path (/etc) (C:\bin\sh.exe 1958) conv_to_win32_path (/etc) (C:\bin\sh.exe 1958) /etc = normalize_posix_path (/etc) (C:\bin\sh.exe 1958) c:\etc = conv_to_win32_path (/etc) (C:\bin\sh.exe 1958) conv_to_win32_path (/etc/passwd) (C:\bin\sh.exe 1958) /etc/passwd = normalize_posix_path (/etc/passwd) (C:\bin\sh.exe 1958) c:\etc\passwd = conv_to_win32_path (/etc/passwd) (C:\bin\sh.exe 1958) conv_to_win32_path (/etc) (C:\bin\sh.exe 1958) /etc = normalize_posix_path (/etc) (C:\bin\sh.exe 1958) c:\etc = conv_to_win32_path (/etc) (C:\bin\sh.exe 1958) 0 = symlink_follow (/etc/passwd) -> /etc/passwd (C:\bin\sh.exe 1958) fhandler_base::open (/etc/passwd, 0x20000) (C:\bin\sh.exe 1958) conv_to_win32_path (/etc/passwd) (C:\bin\sh.exe 1958) /etc/passwd = normalize_posix_path (/etc/passwd) (C:\bin\sh.exe 1958) c:\etc\passwd = conv_to_win32_path (/etc/passwd) (C:\bin\sh.exe 1958) 76 = CreateFileA (c:\etc\passwd, 0x80000000, 0x7, 0x240F780, 0x3, 0x80, 0) (C:\bin\sh.exe 1958) set handle to 76 (C:\bin\sh.exe 1958) 0x2E70170 = fhandler_base::open (/etc/passwd, 0x20000) (C:\bin\sh.exe 1958) 0x2E70170 = fhandler_disk_file::open (/etc/passwd, 0x20000) (C:\bin\sh.exe 1958) 3 = open (/etc/passwd, 0x20000) (C:\bin\sh.exe 1958) 1 = GetFileInformationByHandle (/etc/passwd, 76) (C:\bin\sh.exe 1958) conv_to_win32_path (/etc/passwd) (C:\bin\sh.exe 1958) /etc/passwd = normalize_posix_path (/etc/passwd) (C:\bin\sh.exe 1958) c:\etc\passwd = conv_to_win32_path (/etc/passwd) (C:\bin\sh.exe 1958) _sbrk (0) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E73800, end 0x3673800, size 8388608, avail 8374272 (C:\bin\sh.exe 1958) 0x2E73800 = sbrk (0x0) (total 0x3800) (C:\bin\sh.exe 1958) _sbrk (2048) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E73800, end 0x3673800, size 8388608, avail 8374272 (C:\bin\sh.exe 1958) 0x2E73800 = sbrk (0x800) (total 0x4000) (C:\bin\sh.exe 1958) 0 = fstat (, 0x240F8D8) st_atime=33FCAF52 st_size=316, st_mode=0x81B6, st_ino=142299037, sizeof=64 (C:\bin\sh.exe 1958) 0 = fstat (3, 240F8D8) (C:\bin\sh.exe 1958) _sbrk (0) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E74000, end 0x3674000, size 8388608, avail 8372224 (C:\bin\sh.exe 1958) 0x2E74000 = sbrk (0x0) (total 0x4000) (C:\bin\sh.exe 1958) _sbrk (2048) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E74000, end 0x3674000, size 8388608, avail 8372224 (C:\bin\sh.exe 1958) 0x2E74000 = sbrk (0x800) (total 0x4800) (C:\bin\sh.exe 1958) read (3, 0x2E74004, 1024) (C:\bin\sh.exe 1958) read 309 bytes ( A d m i n i s t r a t o r : : 5 ...) (C:\bin\sh.exe 1958) 309 = read (3, 0x2E74004, 1024) (C:\bin\sh.exe 1958) _sbrk (0) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E74800, end 0x3674800, size 8388608, avail 8370176 (C:\bin\sh.exe 1958) 0x2E74800 = sbrk (0x0) (total 0x4800) (C:\bin\sh.exe 1958) _sbrk (2048) (C:\bin\sh.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E74800, end 0x3674800, size 8388608, avail 8370176 (C:\bin\sh.exe 1958) 0x2E74800 = sbrk (0x800) (total 0x5000) (C:\bin\sh.exe 1958) read (3, 0x2E74004, 1024) (C:\bin\sh.exe 1958) 0 = read (3, 0x2E74004, 1024) (C:\bin\sh.exe 1958) close (3) (C:\bin\sh.exe 1958) fhandler_base::close (handle 76) (C:\bin\sh.exe 1958) set handle to -99 (C:\bin\sh.exe 1958) 0 = close (3) (C:\bin\sh.exe 1958) Running delqueue (C:\bin\sh.exe 1958) open (/etc/group, 0x20000) (C:\bin\sh.exe 1958) fhandler_disk_file::open (/etc/group, 0x20000) (C:\bin\sh.exe 1958) conv_to_win32_path (/etc) (C:\bin\sh.exe 1958) /etc = normalize_posix_path (/etc) (C:\bin\sh.exe 1958) c:\etc = conv_to_win32_path (/etc) (C:\bin\sh.exe 1958) conv_to_win32_path (/etc) (C:\bin\sh.exe 1958) /etc = normalize_posix_path (/etc) (C:\bin\sh.exe 1958) c:\etc = conv_to_win32_path (/etc) (C:\bin\sh.exe 1958) conv_to_win32_path (/etc/group) (C:\bin\sh.exe 1958) /etc/group = normalize_posix_path (/etc/group) (C:\bin\sh.exe 1958) c:\etc\group = conv_to_win32_path (/etc/group) (C:\bin\sh.exe 1958) conv_to_win32_path (/etc) (C:\bin\sh.exe 1958) /etc = normalize_posix_path (/etc) (C:\bin\sh.exe 1958) c:\etc = conv_to_win32_path (/etc) (C:\bin\sh.exe 1958) 0 = symlink_follow (/etc/group) -> /etc/group (C:\bin\sh.exe 1958) fhandler_base::open (/etc/group, 0x20000) (C:\bin\sh.exe 1958) conv_to_win32_path (/etc/group) (C:\bin\sh.exe 1958) /etc/group = normalize_posix_path (/etc/group) (C:\bin\sh.exe 1958) c:\etc\group = conv_to_win32_path (/etc/group) (C:\bin\sh.exe 1958) 128 = CreateFileA (c:\etc\group, 0x80000000, 0x7, 0x240F780, 0x3, 0x80, 0) (C:\bin\sh.exe 1958) set handle to 128 (C:\bin\sh.exe 1958) 0x2E70170 = fhandler_base::open (/etc/group, 0x20000) (C:\bin\sh.exe 1958) 0x2E70170 = fhandler_disk_file::open (/etc/group, 0x20000) (C:\bin\sh.exe 1958) 3 = open (/etc/group, 0x20000) (C:\bin\sh.exe 1958) 1 = GetFileInformationByHandle (/etc/group, 128) (C:\bin\sh.exe 1958) conv_to_win32_path (/etc/group) (C:\bin\sh.exe 1958) /etc/group = normalize_posix_path (/etc/group) (C:\bin\sh.exe 1958) c:\etc\group = conv_to_win32_path (/etc/group) (C:\bin\sh.exe 1958) 0 = fstat (, 0x240F8D8) st_atime=33FCAF52 st_size=222, st_mode=0x81B6, st_ino=79328046, sizeof=64 (C:\bin\sh.exe 1958) 0 = fstat (3, 240F8D8) (C:\bin\sh.exe 1958) read (3, 0x2E74004, 1024) (C:\bin\sh.exe 1958) read 213 bytes ( A c c o u n t 0x20 O p e r a t o r ...) (C:\bin\sh.exe 1958) 213 = read (3, 0x2E74004, 1024) (C:\bin\sh.exe 1958) read (3, 0x2E74004, 1024) (C:\bin\sh.exe 1958) 0 = read (3, 0x2E74004, 1024) (C:\bin\sh.exe 1958) close (3) (C:\bin\sh.exe 1958) fhandler_base::close (handle 128) (C:\bin\sh.exe 1958) set handle to -99 (C:\bin\sh.exe 1958) 0 = close (3) (C:\bin\sh.exe 1958) Running delqueue (C:\bin\sh.exe 1958) line from passwd file Administrator::500:513:::/bin/sh (C:\bin\sh.exe 1958) line from passwd file ahuger::1004:513:Alfred Huger://c/usr/home/ahuger:/bin/sh (C:\bin\sh.exe 1958) line from passwd file artwong::2009:513:art wong::/bin/sh (C:\bin\sh.exe 1958) line from passwd file Blas::1003:513:Blas Zappa://linux/blas:/bin/sh (C:\bin\sh.exe 1958) line from passwd file Guest::501:514:::/bin/sh (C:\bin\sh.exe 1958) line from passwd file IUSR_INFIDEL::1001:513:Internet Guest Account::/bin/sh (C:\bin\sh.exe 1958) line from passwd file newsham::2013:513:tim newsham://c/home/newsham:/bin/sh (C:\bin\sh.exe 1958) res 0 (C:\bin\sh.exe 1958) wVersion 514 (C:\bin\sh.exe 1958) wHighVersion 514 (C:\bin\sh.exe 1958) szDescription WinSock 2.0 (C:\bin\sh.exe 1958) szSystemStatus Running (duh) (C:\bin\sh.exe 1958) iMaxSockets 0 (C:\bin\sh.exe 1958) iMaxUdpDg 0 (C:\bin\sh.exe 1958) lpVendorInfo 268780192 (C:\bin\sh.exe 1958) Application CYGWIN version: 17.6 (C:\bin\sh.exe 1958) CYGWIN DLL version : 17.6 (C:\bin\sh.exe 1958) compute_argc (a ) (C:\bin\sh.exe 1958) compute_argc: src is now a (C:\bin\sh.exe 1958) compute_argc: count is 1 (C:\bin\sh.exe 1958) build_argv: fill iteration a (C:\bin\sh.exe 1958) build_argv: argv[0]=a (C:\bin\sh.exe 1958) Processing 36 env vars ... (C:\bin\sh.exe 1958) need to convert 15 PATH=c:\bin;c:\gnuwin32\b18\H-i386-cygwin32\bin;c:\gnuwin32\b18\tcl\bin;c:\winnt\system32;c:\winnt;c:\perl5\bin;c:\usr\local\bin;c:\usr\X11\bin;. (C:\bin\sh.exe 1958) conv_to_posix_path (c:\bin, keep-rel) (C:\bin\sh.exe 1958) c:\bin = normalize_win32_path (c:\bin) (C:\bin\sh.exe 1958) /bin = conv_to_posix_path (c:\bin) (C:\bin\sh.exe 1958) conv_to_posix_path (c:\gnuwin32\b18\H-i386-cygwin32\bin, keep-rel) (C:\bin\sh.exe 1958) c:\gnuwin32\b18\H-i386-cygwin32\bin = normalize_win32_path (c:\gnuwin32\b18\H-i386-cygwin32\bin) (C:\bin\sh.exe 1958) /gnuwin32/b18/H-i386-cygwin32/bin = conv_to_posix_path (c:\gnuwin32\b18\H-i386-cygwin32\bin) (C:\bin\sh.exe 1958) conv_to_posix_path (c:\gnuwin32\b18\tcl\bin, keep-rel) (C:\bin\sh.exe 1958) c:\gnuwin32\b18\tcl\bin = normalize_win32_path (c:\gnuwin32\b18\tcl\bin) (C:\bin\sh.exe 1958) /gnuwin32/b18/tcl/bin = conv_to_posix_path (c:\gnuwin32\b18\tcl\bin) (C:\bin\sh.exe 1958) conv_to_posix_path (c:\winnt\system32, keep-rel) (C:\bin\sh.exe 1958) c:\winnt\system32 = normalize_win32_path (c:\winnt\system32) (C:\bin\sh.exe 1958) /winnt/system32 = conv_to_posix_path (c:\winnt\system32) (C:\bin\sh.exe 1958) conv_to_posix_path (c:\winnt, keep-rel) (C:\bin\sh.exe 1958) c:\winnt = normalize_win32_path (c:\winnt) (C:\bin\sh.exe 1958) /winnt = conv_to_posix_path (c:\winnt) (C:\bin\sh.exe 1958) conv_to_posix_path (c:\perl5\bin, keep-rel) (C:\bin\sh.exe 1958) c:\perl5\bin = normalize_win32_path (c:\perl5\bin) (C:\bin\sh.exe 1958) /perl5/bin = conv_to_posix_path (c:\perl5\bin) (C:\bin\sh.exe 1958) conv_to_posix_path (c:\usr\local\bin, keep-rel) (C:\bin\sh.exe 1958) c:\usr\local\bin = normalize_win32_path (c:\usr\local\bin) (C:\bin\sh.exe 1958) /usr/local/bin = conv_to_posix_path (c:\usr\local\bin) (C:\bin\sh.exe 1958) conv_to_posix_path (c:\usr\X11\bin, keep-rel) (C:\bin\sh.exe 1958) c:\usr\X11\bin = normalize_win32_path (c:\usr\X11\bin) (C:\bin\sh.exe 1958) /usr/X11/bin = conv_to_posix_path (c:\usr\X11\bin) (C:\bin\sh.exe 1958) conv_to_posix_path (., keep-rel) (C:\bin\sh.exe 1958) . = conv_to_posix_path (.) (C:\bin\sh.exe 1958) env var converted to PATH=/bin:/gnuwin32/b18/H-i386-cygwin32/bin:/gnuwin32/b18/tcl/bin:/winnt/system32:/winnt:/perl5/bin:/usr/local/bin:/usr/X11/bin:. (c:\a.exe 1958) argv[0] = a (c:\a.exe 1958) envp[0] = 240F988 BASEDIR=C:\DDK (c:\a.exe 1958) envp[1] = 240F974 COMPUTERNAME=SPOON (c:\a.exe 1958) envp[2] = 240F950 COMSPEC=C:\WINNT\system32\cmd.exe (c:\a.exe 1958) envp[3] = 240F93C CYGWIN_NOTITLE=1 (c:\a.exe 1958) envp[4] = 240F930 DDKDRIVE=C: (c:\a.exe 1958) envp[5] = 240F8F0 GCC_EXEC_PREFIX=c:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\ (c:\a.exe 1958) envp[6] = 240F8C4 GDBTK_LIBRARY=c:/gnuwin32/b18/share/gdbtcl (c:\a.exe 1958) envp[7] = 240F8B0 HOME=/home/newsham (c:\a.exe 1958) envp[8] = 240F8A0 HOMEDRIVE=c: (c:\a.exe 1958) envp[9] = 240F888 HOMEPATH=\home\newsham (c:\a.exe 1958) envp[10] = 240F878 HOSTTYPE=i386 (c:\a.exe 1958) envp[11] = 240F850 LIB=c:\msdev\lib;c:\msdev\mfc\lib;%lib% (c:\a.exe 1958) envp[12] = 240F838 NUMBER_OF_PROCESSORS=1 (c:\a.exe 1958) envp[13] = 240F828 OS=Windows_NT (c:\a.exe 1958) envp[14] = 240F818 OSTYPE=win32 (c:\a.exe 1958) envp[15] = 240F424 PATH=/bin:/gnuwin32/b18/H-i386-cygwin32/bin:/gnuwin32/b18/tcl/bin:/winnt/system32:/winnt:/perl5/bin:/usr/local/bin:/usr/X11/bin:. (c:\a.exe 1958) envp[16] = 240F768 PROCESSOR_ARCHITECTURE=x86 (c:\a.exe 1958) envp[17] = 240F724 PROCESSOR_IDENTIFIER=x86 Family 5 Model 2 Stepping 12, GenuineIntel (c:\a.exe 1958) envp[18] = 240F710 PROCESSOR_LEVEL=5 (c:\a.exe 1958) envp[19] = 240F6F8 PROCESSOR_REVISION=020c (c:\a.exe 1958) envp[20] = 240F6E8 SHELL=/bin/sh (c:\a.exe 1958) envp[21] = 240F6E0 SHLVL=1 (c:\a.exe 1958) envp[22] = 240F6D0 STRACE=1,trace (c:\a.exe 1958) envp[23] = 240F6C0 SYSTEMDRIVE=C: (c:\a.exe 1958) envp[24] = 240F6AC SYSTEMROOT=C:\WINNT (c:\a.exe 1958) envp[25] = 240F680 TCL_LIBRARY=c:/gnuwin32/b18/tcl/lib/tcl7.6 (c:\a.exe 1958) envp[26] = 240F670 TEMP=C:\TEMP (c:\a.exe 1958) envp[27] = 240F664 TERM=linux (c:\a.exe 1958) envp[28] = 240F658 TMP=C:\TEMP (c:\a.exe 1958) envp[29] = 240F648 USERDOMAIN=X0D (c:\a.exe 1958) envp[30] = 240F634 USERNAME=newsham (c:\a.exe 1958) envp[31] = 240F60C USERPROFILE=C:\WINNT\Profiles\newsham (c:\a.exe 1958) envp[32] = 240F5F8 VIM=/usr/share/vim (c:\a.exe 1958) envp[33] = 240F5E8 WINDIR=C:\WINNT (c:\a.exe 1958) envp[34] = 240F5E0 _=./a (c:\a.exe 1958) envp[35] = 240F5D4 PID=1958 (c:\a.exe 1958) CYGWIN Release 17.6, compiled Aug 15 1997 11:21:45 (c:\a.exe 1958) _spawnve (/bin/sh, sh, 240F998) (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) conv_to_win32_path (/bin/sh) (c:\a.exe 1958) /bin/sh = normalize_posix_path (/bin/sh) (c:\a.exe 1958) c:\bin\sh = conv_to_win32_path (/bin/sh) (c:\a.exe 1958) /src/current/cdk/winsup/path.cc:1123 seterrno: 2 (ERROR_FILE_NOT_FOUND) -> 2 (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) 0 = symlink_follow (/bin/sh) -> /bin/sh (c:\a.exe 1958) conv_to_win32_path (/bin/sh) (c:\a.exe 1958) /bin/sh = normalize_posix_path (/bin/sh) (c:\a.exe 1958) c:\bin\sh = conv_to_win32_path (/bin/sh) (c:\a.exe 1958) _spawnve: can't find /bin/sh, trying with .exe appended (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) conv_to_win32_path (/bin/sh.exe) (c:\a.exe 1958) /bin/sh.exe = normalize_posix_path (/bin/sh.exe) (c:\a.exe 1958) c:\bin\sh.exe = conv_to_win32_path (/bin/sh.exe) (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) 0 = symlink_follow (/bin/sh.exe) -> /bin/sh.exe (c:\a.exe 1958) conv_to_win32_path (/bin/sh.exe) (c:\a.exe 1958) /bin/sh.exe = normalize_posix_path (/bin/sh.exe) (c:\a.exe 1958) c:\bin\sh.exe = conv_to_win32_path (/bin/sh.exe) (c:\a.exe 1958) lock_pinfo_for update : locking handle 44, timeout -1 (c:\a.exe 1958) pinfo_list::get_empty_pinfo: found empty slot 5 (c:\a.exe 1958) pinfo_list::allocate_pid: pinfo->hmap.vec == 0 (c:\a.exe 1958) _sbrk (0) (c:\a.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E75000, end 0x3675000, size 8388608, avail 8368128 (c:\a.exe 1958) 0x2E75000 = sbrk (0x0) (total 0x5000) (c:\a.exe 1958) _sbrk (8192) (c:\a.exe 1958) _sbrk: base 0x2E70000, ptr 0x2E75000, end 0x3675000, size 8388608, avail 8368128 (c:\a.exe 1958) 0x2E75000 = sbrk (0x2000) (total 0x7000) (c:\a.exe 1958) unlock_pinfo : handle 44 (c:\a.exe 1958) spawn_guts (/bin/sh.exe) (c:\a.exe 1958) spawn_guts () argv[0] in `sh' (c:\a.exe 1958) spawn_guts () argv[0] out `sh' (c:\a.exe 1958) spawn_guts () argv[1] in `-c' (c:\a.exe 1958) spawn_guts () argv[1] out `-c' (c:\a.exe 1958) spawn_guts () argv[2] in `/winnt/system32/arp -a' (c:\a.exe 1958) spawn_guts () argv[2] out `/winnt/system32/arp -a' (c:\a.exe 1958) spawn_guts: translating PATH=/bin:/gnuwin32/b18/H-i386-cygwin32/bin:/gnuwin32/b18/tcl/bin:/winnt/system32:/winnt:/perl5/bin:/usr/local/bin:/usr/X11/bin:. (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\a.exe 1958) /gnuwin32/b18/H-i386-cygwin32/bin = normalize_posix_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\a.exe 1958) c:\gnuwin32\b18\H-i386-cygwin32\bin = conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\a.exe 1958) conv_to_win32_path (/gnuwin32/b18/tcl/bin) (c:\a.exe 1958) /gnuwin32/b18/tcl/bin = normalize_posix_path (/gnuwin32/b18/tcl/bin) (c:\a.exe 1958) c:\gnuwin32\b18\tcl\bin = conv_to_win32_path (/gnuwin32/b18/tcl/bin) (c:\a.exe 1958) conv_to_win32_path (/winnt/system32) (c:\a.exe 1958) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\a.exe 1958) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\a.exe 1958) conv_to_win32_path (/winnt) (c:\a.exe 1958) /winnt = normalize_posix_path (/winnt) (c:\a.exe 1958) c:\winnt = conv_to_win32_path (/winnt) (c:\a.exe 1958) conv_to_win32_path (/perl5/bin) (c:\a.exe 1958) /perl5/bin = normalize_posix_path (/perl5/bin) (c:\a.exe 1958) c:\perl5\bin = conv_to_win32_path (/perl5/bin) (c:\a.exe 1958) conv_to_win32_path (/usr/local/bin) (c:\a.exe 1958) /usr/local/bin = normalize_posix_path (/usr/local/bin) (c:\a.exe 1958) c:\usr\local\bin = conv_to_win32_path (/usr/local/bin) (c:\a.exe 1958) conv_to_win32_path (/usr/X11/bin) (c:\a.exe 1958) /usr/X11/bin = normalize_posix_path (/usr/X11/bin) (c:\a.exe 1958) c:\usr\X11\bin = conv_to_win32_path (/usr/X11/bin) (c:\a.exe 1958) conv_to_win32_path (.) (c:\a.exe 1958) conv_to_posix_path (c:\ballista, no-keep-rel) (c:\a.exe 1958) c:\ballista = normalize_win32_path (c:\ballista) (c:\a.exe 1958) /ballista = conv_to_posix_path (c:\ballista) (c:\a.exe 1958) 0x240E754 (/ballista) = getcwd_inner (0x240E754, 260, posix) (c:\a.exe 1958) /ballista = normalize_posix_path (.) (c:\a.exe 1958) . = conv_to_win32_path (.) (c:\a.exe 1958) spawn_guts: setting PATH=c:\bin;c:\gnuwin32\b18\H-i386-cygwin32\bin;c:\gnuwin32\b18\tcl\bin;c:\winnt\system32;c:\winnt;c:\perl5\bin;c:\usr\local\bin;c:\usr\X11\bin;. (c:\a.exe 1958) spawn_guts: env count=35 bytes=906 (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) conv_to_win32_path (/bin/sh.exe) (c:\a.exe 1958) /bin/sh.exe = normalize_posix_path (/bin/sh.exe) (c:\a.exe 1958) c:\bin\sh.exe = conv_to_win32_path (/bin/sh.exe) (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) 0 = symlink_follow (/bin/sh.exe) -> /bin/sh.exe (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) conv_to_win32_path (/bin/sh.exe) (c:\a.exe 1958) /bin/sh.exe = normalize_posix_path (/bin/sh.exe) (c:\a.exe 1958) c:\bin\sh.exe = conv_to_win32_path (/bin/sh.exe) (c:\a.exe 1958) conv_to_win32_path (/bin) (c:\a.exe 1958) /bin = normalize_posix_path (/bin) (c:\a.exe 1958) c:\bin = conv_to_win32_path (/bin) (c:\a.exe 1958) 0 = symlink_follow (/bin/sh.exe) -> /bin/sh.exe (c:\a.exe 1958) conv_to_win32_path (/bin/sh.exe) (c:\a.exe 1958) /bin/sh.exe = normalize_posix_path (/bin/sh.exe) (c:\a.exe 1958) c:\bin\sh.exe = conv_to_win32_path (/bin/sh.exe) (c:\a.exe 1958) spawn_guts (c:\bin\sh.exe, sh -c "/winnt/system32/arp -a" ) (c:\a.exe 1958) create_shared_fd_table: pid = 1959, no of fds = 64, map_name = cygwin32.fd_map.17.1959 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: buf = 0, buflen = 0 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing open fd 0:/dev/conin (c:\a.exe 1958) hinfo_vec::linearize_fd_array: len was 102 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing open fd 1:out (c:\a.exe 1958) hinfo_vec::linearize_fd_array: len was 30 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing open fd 2:out (c:\a.exe 1958) hinfo_vec::linearize_fd_array: len was 30 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 3 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 4 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 5 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 6 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 7 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 8 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 9 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 10 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 11 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 12 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 13 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 14 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 15 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 16 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 17 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 18 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 19 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 20 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 21 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 22 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 23 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 24 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 25 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 26 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 27 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 28 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 29 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 30 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 31 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 32 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 33 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 34 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 35 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 36 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 37 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 38 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 39 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 40 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 41 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 42 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 43 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 44 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 45 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 46 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 47 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 48 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 49 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 50 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 51 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 52 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 53 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 54 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 55 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 56 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 57 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 58 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 59 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 60 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 61 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 62 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 63 (c:\a.exe 1958) create_shared_fd_table: area = 5670000, u->self->hmap.vec = 2E70004, size = 246 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: buf = 5670004, buflen = 246 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing open fd 0:/dev/conin (c:\a.exe 1958) hinfo_vec::linearize_fd_array: len was 102 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing open fd 1:out (c:\a.exe 1958) hinfo_vec::linearize_fd_array: len was 30 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing open fd 2:out (c:\a.exe 1958) hinfo_vec::linearize_fd_array: len was 30 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 3 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 4 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 5 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 6 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 7 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 8 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 9 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 10 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 11 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 12 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 13 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 14 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 15 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 16 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 17 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 18 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 19 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 20 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 21 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 22 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 23 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 24 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 25 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 26 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 27 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 28 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 29 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 30 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 31 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 32 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 33 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 34 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 35 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 36 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 37 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 38 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 39 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 40 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 41 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 42 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 43 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 44 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 45 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 46 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 47 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 48 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 49 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 50 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 51 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 52 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 53 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 54 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 55 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 56 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 57 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 58 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 59 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 60 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 61 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 62 (c:\a.exe 1958) hinfo_vec::linearize_fd_array: linearizing closed fd 63 (c:\a.exe 1958) 65 = spawn_guts (/bin/sh.exe, sh -c "/winnt/system32/arp -a" ) (c:\a.exe 1958) lock_pinfo_for update : locking handle 44, timeout -1 (c:\a.exe 1958) unlock_pinfo : handle 44 (c:\a.exe 1958) wait_for_single: child pid 1959 (unknown) open_shared_file_map: s = 0xA000000, h = 24 (unknown) lock_pinfo_for update : locking handle 0, timeout -1 (unknown) lock_pinfo_for update : creating mutex = 44 (/bin/sh.exe 1959) copy_shared_fd_table: pid = 1959, name = cygwin32.fd_map.17.1959 (/bin/sh.exe 1959) copy_shared_fd_table: area = 36B0000, old p->hmap.vec = 2E75004 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: buf = 36B0004 (/bin/sh.exe 1959) _sbrk (0) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB0000, end 0x36B0000, size 8388608, avail 8388608 (/bin/sh.exe 1959) 0x2EB0000 = sbrk (0x0) (total 0x0) (/bin/sh.exe 1959) _sbrk (8192) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB0000, end 0x36B0000, size 8388608, avail 8388608 (/bin/sh.exe 1959) 0x2EB0000 = sbrk (0x2000) (total 0x2000) (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: number_of_fds = 64, vec = 2EB0004 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: open fd = 0, /dev/conin (/bin/sh.exe 1959) _sbrk (0) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB2000, end 0x36B2000, size 8388608, avail 8380416 (/bin/sh.exe 1959) 0x2EB2000 = sbrk (0x0) (total 0x2000) (/bin/sh.exe 1959) _sbrk (2048) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB2000, end 0x36B2000, size 8388608, avail 8380416 (/bin/sh.exe 1959) 0x2EB2000 = sbrk (0x800) (total 0x2800) (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: len = 113 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: open fd = 1, out (/bin/sh.exe 1959) _sbrk (0) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB2800, end 0x36B2800, size 8388608, avail 8378368 (/bin/sh.exe 1959) 0x2EB2800 = sbrk (0x0) (total 0x2800) (/bin/sh.exe 1959) _sbrk (2048) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB2800, end 0x36B2800, size 8388608, avail 8378368 (/bin/sh.exe 1959) 0x2EB2800 = sbrk (0x800) (total 0x3000) (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: len = 34 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: open fd = 2, out (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: len = 34 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 3 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 4 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 5 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 6 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 7 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 8 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 9 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 10 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 11 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 12 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 13 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 14 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 15 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 16 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 17 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 18 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 19 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 20 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 21 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 22 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 23 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 24 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 25 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 26 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 27 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 28 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 29 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 30 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 31 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 32 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 33 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 34 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 35 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 36 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 37 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 38 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 39 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 40 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 41 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 42 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 43 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 44 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 45 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 46 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 47 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 48 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 49 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 50 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 51 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 52 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 53 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 54 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 55 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 56 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 57 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 58 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 59 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 60 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 61 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 62 (/bin/sh.exe 1959) hinfo_vec::de_linearize_fd_array: closed fd = 63 (/bin/sh.exe 1959) copy_shared_fd_table: new p->hmap.vec = 2EB0004 (/bin/sh.exe 1959) unlock_pinfo : handle 44 (/bin/sh.exe 1959) _sbrk (0) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB3000, end 0x36B3000, size 8388608, avail 8376320 (/bin/sh.exe 1959) 0x2EB3000 = sbrk (0x0) (total 0x3000) (/bin/sh.exe 1959) _sbrk (2048) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB3000, end 0x36B3000, size 8388608, avail 8376320 (/bin/sh.exe 1959) 0x2EB3000 = sbrk (0x800) (total 0x3800) (/bin/sh.exe 1959) open (/etc/passwd, 0x20000) (/bin/sh.exe 1959) fhandler_disk_file::open (/etc/passwd, 0x20000) (/bin/sh.exe 1959) conv_to_win32_path (/etc) (/bin/sh.exe 1959) /etc = normalize_posix_path (/etc) (/bin/sh.exe 1959) c:\etc = conv_to_win32_path (/etc) (/bin/sh.exe 1959) conv_to_win32_path (/etc) (/bin/sh.exe 1959) /etc = normalize_posix_path (/etc) (/bin/sh.exe 1959) c:\etc = conv_to_win32_path (/etc) (/bin/sh.exe 1959) conv_to_win32_path (/etc/passwd) (/bin/sh.exe 1959) /etc/passwd = normalize_posix_path (/etc/passwd) (/bin/sh.exe 1959) c:\etc\passwd = conv_to_win32_path (/etc/passwd) (/bin/sh.exe 1959) conv_to_win32_path (/etc) (/bin/sh.exe 1959) /etc = normalize_posix_path (/etc) (/bin/sh.exe 1959) c:\etc = conv_to_win32_path (/etc) (/bin/sh.exe 1959) 0 = symlink_follow (/etc/passwd) -> /etc/passwd (/bin/sh.exe 1959) fhandler_base::open (/etc/passwd, 0x20000) (/bin/sh.exe 1959) conv_to_win32_path (/etc/passwd) (/bin/sh.exe 1959) /etc/passwd = normalize_posix_path (/etc/passwd) (/bin/sh.exe 1959) c:\etc\passwd = conv_to_win32_path (/etc/passwd) (/bin/sh.exe 1959) 76 = CreateFileA (c:\etc\passwd, 0x80000000, 0x7, 0x244F780, 0x3, 0x80, 0) (/bin/sh.exe 1959) set handle to 76 (/bin/sh.exe 1959) 0x2EB0170 = fhandler_base::open (/etc/passwd, 0x20000) (/bin/sh.exe 1959) 0x2EB0170 = fhandler_disk_file::open (/etc/passwd, 0x20000) (/bin/sh.exe 1959) 3 = open (/etc/passwd, 0x20000) (/bin/sh.exe 1959) 1 = GetFileInformationByHandle (/etc/passwd, 76) (/bin/sh.exe 1959) conv_to_win32_path (/etc/passwd) (/bin/sh.exe 1959) /etc/passwd = normalize_posix_path (/etc/passwd) (/bin/sh.exe 1959) c:\etc\passwd = conv_to_win32_path (/etc/passwd) (/bin/sh.exe 1959) _sbrk (0) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB3800, end 0x36B3800, size 8388608, avail 8374272 (/bin/sh.exe 1959) 0x2EB3800 = sbrk (0x0) (total 0x3800) (/bin/sh.exe 1959) _sbrk (2048) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB3800, end 0x36B3800, size 8388608, avail 8374272 (/bin/sh.exe 1959) 0x2EB3800 = sbrk (0x800) (total 0x4000) (/bin/sh.exe 1959) 0 = fstat (, 0x244F8D8) st_atime=33FCAF55 st_size=316, st_mode=0x81B6, st_ino=142299037, sizeof=64 (/bin/sh.exe 1959) 0 = fstat (3, 244F8D8) (/bin/sh.exe 1959) _sbrk (0) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB4000, end 0x36B4000, size 8388608, avail 8372224 (/bin/sh.exe 1959) 0x2EB4000 = sbrk (0x0) (total 0x4000) (/bin/sh.exe 1959) _sbrk (2048) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB4000, end 0x36B4000, size 8388608, avail 8372224 (/bin/sh.exe 1959) 0x2EB4000 = sbrk (0x800) (total 0x4800) (/bin/sh.exe 1959) read (3, 0x2EB4004, 1024) (/bin/sh.exe 1959) read 309 bytes ( A d m i n i s t r a t o r : : 5 ...) (/bin/sh.exe 1959) 309 = read (3, 0x2EB4004, 1024) (/bin/sh.exe 1959) _sbrk (0) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB4800, end 0x36B4800, size 8388608, avail 8370176 (/bin/sh.exe 1959) 0x2EB4800 = sbrk (0x0) (total 0x4800) (/bin/sh.exe 1959) _sbrk (2048) (/bin/sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB4800, end 0x36B4800, size 8388608, avail 8370176 (/bin/sh.exe 1959) 0x2EB4800 = sbrk (0x800) (total 0x5000) (/bin/sh.exe 1959) read (3, 0x2EB4004, 1024) (/bin/sh.exe 1959) 0 = read (3, 0x2EB4004, 1024) (/bin/sh.exe 1959) close (3) (/bin/sh.exe 1959) fhandler_base::close (handle 76) (/bin/sh.exe 1959) set handle to -99 (/bin/sh.exe 1959) 0 = close (3) (/bin/sh.exe 1959) Running delqueue (/bin/sh.exe 1959) open (/etc/group, 0x20000) (/bin/sh.exe 1959) fhandler_disk_file::open (/etc/group, 0x20000) (/bin/sh.exe 1959) conv_to_win32_path (/etc) (/bin/sh.exe 1959) /etc = normalize_posix_path (/etc) (/bin/sh.exe 1959) c:\etc = conv_to_win32_path (/etc) (/bin/sh.exe 1959) conv_to_win32_path (/etc) (/bin/sh.exe 1959) /etc = normalize_posix_path (/etc) (/bin/sh.exe 1959) c:\etc = conv_to_win32_path (/etc) (/bin/sh.exe 1959) conv_to_win32_path (/etc/group) (/bin/sh.exe 1959) /etc/group = normalize_posix_path (/etc/group) (/bin/sh.exe 1959) c:\etc\group = conv_to_win32_path (/etc/group) (/bin/sh.exe 1959) conv_to_win32_path (/etc) (/bin/sh.exe 1959) /etc = normalize_posix_path (/etc) (/bin/sh.exe 1959) c:\etc = conv_to_win32_path (/etc) (/bin/sh.exe 1959) 0 = symlink_follow (/etc/group) -> /etc/group (/bin/sh.exe 1959) fhandler_base::open (/etc/group, 0x20000) (/bin/sh.exe 1959) conv_to_win32_path (/etc/group) (/bin/sh.exe 1959) /etc/group = normalize_posix_path (/etc/group) (/bin/sh.exe 1959) c:\etc\group = conv_to_win32_path (/etc/group) (/bin/sh.exe 1959) 124 = CreateFileA (c:\etc\group, 0x80000000, 0x7, 0x244F780, 0x3, 0x80, 0) (/bin/sh.exe 1959) set handle to 124 (/bin/sh.exe 1959) 0x2EB0170 = fhandler_base::open (/etc/group, 0x20000) (/bin/sh.exe 1959) 0x2EB0170 = fhandler_disk_file::open (/etc/group, 0x20000) (/bin/sh.exe 1959) 3 = open (/etc/group, 0x20000) (/bin/sh.exe 1959) 1 = GetFileInformationByHandle (/etc/group, 124) (/bin/sh.exe 1959) conv_to_win32_path (/etc/group) (/bin/sh.exe 1959) /etc/group = normalize_posix_path (/etc/group) (/bin/sh.exe 1959) c:\etc\group = conv_to_win32_path (/etc/group) (/bin/sh.exe 1959) 0 = fstat (, 0x244F8D8) st_atime=33FCAF55 st_size=222, st_mode=0x81B6, st_ino=79328046, sizeof=64 (/bin/sh.exe 1959) 0 = fstat (3, 244F8D8) (/bin/sh.exe 1959) read (3, 0x2EB4004, 1024) (/bin/sh.exe 1959) read 213 bytes ( A c c o u n t 0x20 O p e r a t o r ...) (/bin/sh.exe 1959) 213 = read (3, 0x2EB4004, 1024) (/bin/sh.exe 1959) read (3, 0x2EB4004, 1024) (/bin/sh.exe 1959) 0 = read (3, 0x2EB4004, 1024) (/bin/sh.exe 1959) close (3) (/bin/sh.exe 1959) fhandler_base::close (handle 124) (/bin/sh.exe 1959) set handle to -99 (/bin/sh.exe 1959) 0 = close (3) (/bin/sh.exe 1959) Running delqueue (/bin/sh.exe 1959) line from passwd file Administrator::500:513:::/bin/sh (/bin/sh.exe 1959) line from passwd file ahuger::1004:513:Alfred Huger://c/usr/home/ahuger:/bin/sh (/bin/sh.exe 1959) line from passwd file artwong::2009:513:art wong::/bin/sh (/bin/sh.exe 1959) line from passwd file Blas::1003:513:Blas Zappa://linux/blas:/bin/sh (/bin/sh.exe 1959) line from passwd file Guest::501:514:::/bin/sh (/bin/sh.exe 1959) line from passwd file IUSR_INFIDEL::1001:513:Internet Guest Account::/bin/sh (/bin/sh.exe 1959) line from passwd file newsham::2013:513:tim newsham://c/home/newsham:/bin/sh (/bin/sh.exe 1959) res 0 (/bin/sh.exe 1959) wVersion 514 (/bin/sh.exe 1959) wHighVersion 514 (/bin/sh.exe 1959) szDescription WinSock 2.0 (/bin/sh.exe 1959) szSystemStatus Running (duh) (/bin/sh.exe 1959) iMaxSockets 0 (/bin/sh.exe 1959) iMaxUdpDg 0 (/bin/sh.exe 1959) lpVendorInfo 268780192 (/bin/sh.exe 1959) Application CYGWIN version: 17.5 (/bin/sh.exe 1959) CYGWIN DLL version : 17.6 (/bin/sh.exe 1959) compute_argc (sh -c "/winnt/system32/arp -a" ) (/bin/sh.exe 1959) compute_argc: src is now sh -c "/winnt/system32/arp -a" (/bin/sh.exe 1959) compute_argc: src is now -c "/winnt/system32/arp -a" (/bin/sh.exe 1959) compute_argc: src is now "/winnt/system32/arp -a" (/bin/sh.exe 1959) compute_argc: count is 3 (/bin/sh.exe 1959) build_argv: fill iteration sh -c "/winnt/system32/arp -a" (/bin/sh.exe 1959) build_argv: argv[0]=sh (/bin/sh.exe 1959) build_argv: fill iteration -c "/winnt/system32/arp -a" (/bin/sh.exe 1959) build_argv: argv[1]=-c (/bin/sh.exe 1959) build_argv: fill iteration "/winnt/system32/arp -a" (/bin/sh.exe 1959) build_argv: argv[2]=/winnt/system32/arp -a (/bin/sh.exe 1959) build_argv: fill iteration (/bin/sh.exe 1959) Processing 36 env vars ... (/bin/sh.exe 1959) need to convert 15 PATH=c:\bin;c:\gnuwin32\b18\H-i386-cygwin32\bin;c:\gnuwin32\b18\tcl\bin;c:\winnt\system32;c:\winnt;c:\perl5\bin;c:\usr\local\bin;c:\usr\X11\bin;. (/bin/sh.exe 1959) conv_to_posix_path (c:\bin, keep-rel) (/bin/sh.exe 1959) c:\bin = normalize_win32_path (c:\bin) (/bin/sh.exe 1959) /bin = conv_to_posix_path (c:\bin) (/bin/sh.exe 1959) conv_to_posix_path (c:\gnuwin32\b18\H-i386-cygwin32\bin, keep-rel) (/bin/sh.exe 1959) c:\gnuwin32\b18\H-i386-cygwin32\bin = normalize_win32_path (c:\gnuwin32\b18\H-i386-cygwin32\bin) (/bin/sh.exe 1959) /gnuwin32/b18/H-i386-cygwin32/bin = conv_to_posix_path (c:\gnuwin32\b18\H-i386-cygwin32\bin) (/bin/sh.exe 1959) conv_to_posix_path (c:\gnuwin32\b18\tcl\bin, keep-rel) (/bin/sh.exe 1959) c:\gnuwin32\b18\tcl\bin = normalize_win32_path (c:\gnuwin32\b18\tcl\bin) (/bin/sh.exe 1959) /gnuwin32/b18/tcl/bin = conv_to_posix_path (c:\gnuwin32\b18\tcl\bin) (/bin/sh.exe 1959) conv_to_posix_path (c:\winnt\system32, keep-rel) (/bin/sh.exe 1959) c:\winnt\system32 = normalize_win32_path (c:\winnt\system32) (/bin/sh.exe 1959) /winnt/system32 = conv_to_posix_path (c:\winnt\system32) (/bin/sh.exe 1959) conv_to_posix_path (c:\winnt, keep-rel) (/bin/sh.exe 1959) c:\winnt = normalize_win32_path (c:\winnt) (/bin/sh.exe 1959) /winnt = conv_to_posix_path (c:\winnt) (/bin/sh.exe 1959) conv_to_posix_path (c:\perl5\bin, keep-rel) (/bin/sh.exe 1959) c:\perl5\bin = normalize_win32_path (c:\perl5\bin) (/bin/sh.exe 1959) /perl5/bin = conv_to_posix_path (c:\perl5\bin) (/bin/sh.exe 1959) conv_to_posix_path (c:\usr\local\bin, keep-rel) (/bin/sh.exe 1959) c:\usr\local\bin = normalize_win32_path (c:\usr\local\bin) (/bin/sh.exe 1959) /usr/local/bin = conv_to_posix_path (c:\usr\local\bin) (/bin/sh.exe 1959) conv_to_posix_path (c:\usr\X11\bin, keep-rel) (/bin/sh.exe 1959) c:\usr\X11\bin = normalize_win32_path (c:\usr\X11\bin) (/bin/sh.exe 1959) /usr/X11/bin = conv_to_posix_path (c:\usr\X11\bin) (/bin/sh.exe 1959) conv_to_posix_path (., keep-rel) (/bin/sh.exe 1959) . = conv_to_posix_path (.) (/bin/sh.exe 1959) env var converted to PATH=/bin:/gnuwin32/b18/H-i386-cygwin32/bin:/gnuwin32/b18/tcl/bin:/winnt/system32:/winnt:/perl5/bin:/usr/local/bin:/usr/X11/bin:. (c:\bin\sh.exe 1959) argv[0] = sh (c:\bin\sh.exe 1959) argv[1] = -c (c:\bin\sh.exe 1959) argv[2] = /winnt/system32/arp -a (c:\bin\sh.exe 1959) envp[0] = 244F960 BASEDIR=C:\DDK (c:\bin\sh.exe 1959) envp[1] = 244F94C COMPUTERNAME=SPOON (c:\bin\sh.exe 1959) envp[2] = 244F928 COMSPEC=C:\WINNT\system32\cmd.exe (c:\bin\sh.exe 1959) envp[3] = 244F914 CYGWIN_NOTITLE=1 (c:\bin\sh.exe 1959) envp[4] = 244F908 DDKDRIVE=C: (c:\bin\sh.exe 1959) envp[5] = 244F8C8 GCC_EXEC_PREFIX=c:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\ (c:\bin\sh.exe 1959) envp[6] = 244F89C GDBTK_LIBRARY=c:/gnuwin32/b18/share/gdbtcl (c:\bin\sh.exe 1959) envp[7] = 244F888 HOME=/home/newsham (c:\bin\sh.exe 1959) envp[8] = 244F878 HOMEDRIVE=c: (c:\bin\sh.exe 1959) envp[9] = 244F860 HOMEPATH=\home\newsham (c:\bin\sh.exe 1959) envp[10] = 244F850 HOSTTYPE=i386 (c:\bin\sh.exe 1959) envp[11] = 244F828 LIB=c:\msdev\lib;c:\msdev\mfc\lib;%lib% (c:\bin\sh.exe 1959) envp[12] = 244F810 NUMBER_OF_PROCESSORS=1 (c:\bin\sh.exe 1959) envp[13] = 244F800 OS=Windows_NT (c:\bin\sh.exe 1959) envp[14] = 244F7F0 OSTYPE=win32 (c:\bin\sh.exe 1959) envp[15] = 244F3FC PATH=/bin:/gnuwin32/b18/H-i386-cygwin32/bin:/gnuwin32/b18/tcl/bin:/winnt/system32:/winnt:/perl5/bin:/usr/local/bin:/usr/X11/bin:. (c:\bin\sh.exe 1959) envp[16] = 244F740 PROCESSOR_ARCHITECTURE=x86 (c:\bin\sh.exe 1959) envp[17] = 244F6FC PROCESSOR_IDENTIFIER=x86 Family 5 Model 2 Stepping 12, GenuineIntel (c:\bin\sh.exe 1959) envp[18] = 244F6E8 PROCESSOR_LEVEL=5 (c:\bin\sh.exe 1959) envp[19] = 244F6D0 PROCESSOR_REVISION=020c (c:\bin\sh.exe 1959) envp[20] = 244F6C0 SHELL=/bin/sh (c:\bin\sh.exe 1959) envp[21] = 244F6B8 SHLVL=1 (c:\bin\sh.exe 1959) envp[22] = 244F6A8 STRACE=1,trace (c:\bin\sh.exe 1959) envp[23] = 244F698 SYSTEMDRIVE=C: (c:\bin\sh.exe 1959) envp[24] = 244F684 SYSTEMROOT=C:\WINNT (c:\bin\sh.exe 1959) envp[25] = 244F658 TCL_LIBRARY=c:/gnuwin32/b18/tcl/lib/tcl7.6 (c:\bin\sh.exe 1959) envp[26] = 244F648 TEMP=C:\TEMP (c:\bin\sh.exe 1959) envp[27] = 244F63C TERM=linux (c:\bin\sh.exe 1959) envp[28] = 244F630 TMP=C:\TEMP (c:\bin\sh.exe 1959) envp[29] = 244F620 USERDOMAIN=X0D (c:\bin\sh.exe 1959) envp[30] = 244F60C USERNAME=newsham (c:\bin\sh.exe 1959) envp[31] = 244F5E4 USERPROFILE=C:\WINNT\Profiles\newsham (c:\bin\sh.exe 1959) envp[32] = 244F5D0 VIM=/usr/share/vim (c:\bin\sh.exe 1959) envp[33] = 244F5C0 WINDIR=C:\WINNT (c:\bin\sh.exe 1959) envp[34] = 244F5B8 _=./a (c:\bin\sh.exe 1959) envp[35] = 244F5AC PID=1959 (c:\bin\sh.exe 1959) CYGWIN Release 17.6, compiled Aug 15 1997 11:21:45 (c:\bin\sh.exe 1959) stat (/tmp, 0x244F3A8) (c:\bin\sh.exe 1959) conv_to_win32_path (/tmp) (c:\bin\sh.exe 1959) /tmp = normalize_posix_path (/tmp) (c:\bin\sh.exe 1959) c:\tmp = conv_to_win32_path (/tmp) (c:\bin\sh.exe 1959) 0 = symlink_follow (/tmp) -> /tmp (c:\bin\sh.exe 1959) conv_to_win32_path (/tmp) (c:\bin\sh.exe 1959) /tmp = normalize_posix_path (/tmp) (c:\bin\sh.exe 1959) c:\tmp = conv_to_win32_path (/tmp) (c:\bin\sh.exe 1959) 16 = GetFileAttributesA (c:\tmp) (c:\bin\sh.exe 1959) 0 = stat (/tmp, 0x244F3A8) (c:\bin\sh.exe 1959) _sbrk (0) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB5000, end 0x36B5000, size 8388608, avail 8368128 (c:\bin\sh.exe 1959) 0x2EB5000 = sbrk (0x0) (total 0x5000) (c:\bin\sh.exe 1959) _sbrk (2048) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB5000, end 0x36B5000, size 8388608, avail 8368128 (c:\bin\sh.exe 1959) 0x2EB5000 = sbrk (0x800) (total 0x5800) (c:\bin\sh.exe 1959) line from passwd file Administrator::500:513:::/bin/sh (c:\bin\sh.exe 1959) line from passwd file ahuger::1004:513:Alfred Huger://c/usr/home/ahuger:/bin/sh (c:\bin\sh.exe 1959) line from passwd file artwong::2009:513:art wong::/bin/sh (c:\bin\sh.exe 1959) line from passwd file Blas::1003:513:Blas Zappa://linux/blas:/bin/sh (c:\bin\sh.exe 1959) line from passwd file Guest::501:514:::/bin/sh (c:\bin\sh.exe 1959) line from passwd file IUSR_INFIDEL::1001:513:Internet Guest Account::/bin/sh (c:\bin\sh.exe 1959) line from passwd file newsham::2013:513:tim newsham://c/home/newsham:/bin/sh (c:\bin\sh.exe 1959) _sbrk (0) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB5800, end 0x36B5800, size 8388608, avail 8366080 (c:\bin\sh.exe 1959) 0x2EB5800 = sbrk (0x0) (total 0x5800) (c:\bin\sh.exe 1959) _sbrk (2048) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB5800, end 0x36B5800, size 8388608, avail 8366080 (c:\bin\sh.exe 1959) 0x2EB5800 = sbrk (0x800) (total 0x6000) (c:\bin\sh.exe 1959) _sbrk (0) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB6000, end 0x36B6000, size 8388608, avail 8364032 (c:\bin\sh.exe 1959) 0x2EB6000 = sbrk (0x0) (total 0x6000) (c:\bin\sh.exe 1959) _sbrk (2048) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB6000, end 0x36B6000, size 8388608, avail 8364032 (c:\bin\sh.exe 1959) 0x2EB6000 = sbrk (0x800) (total 0x6800) (c:\bin\sh.exe 1959) _sbrk (0) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB6800, end 0x36B6800, size 8388608, avail 8361984 (c:\bin\sh.exe 1959) 0x2EB6800 = sbrk (0x0) (total 0x6800) (c:\bin\sh.exe 1959) _sbrk (2048) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB6800, end 0x36B6800, size 8388608, avail 8361984 (c:\bin\sh.exe 1959) 0x2EB6800 = sbrk (0x800) (total 0x7000) (c:\bin\sh.exe 1959) conv_to_posix_path (c:\ballista, no-keep-rel) (c:\bin\sh.exe 1959) c:\ballista = normalize_win32_path (c:\ballista) (c:\bin\sh.exe 1959) /ballista = conv_to_posix_path (c:\ballista) (c:\bin\sh.exe 1959) 0x2EB3604 (/ballista) = getcwd_inner (0x2EB3604, 259, posix) (c:\bin\sh.exe 1959) stat (., 0x244F130) (c:\bin\sh.exe 1959) conv_to_win32_path (.) (c:\bin\sh.exe 1959) conv_to_posix_path (c:\ballista, no-keep-rel) (c:\bin\sh.exe 1959) c:\ballista = normalize_win32_path (c:\ballista) (c:\bin\sh.exe 1959) /ballista = conv_to_posix_path (c:\ballista) (c:\bin\sh.exe 1959) 0x244E358 (/ballista) = getcwd_inner (0x244E358, 260, posix) (c:\bin\sh.exe 1959) /ballista = normalize_posix_path (.) (c:\bin\sh.exe 1959) . = conv_to_win32_path (.) (c:\bin\sh.exe 1959) 0 = symlink_follow (.) -> . (c:\bin\sh.exe 1959) conv_to_win32_path (.) (c:\bin\sh.exe 1959) conv_to_posix_path (c:\ballista, no-keep-rel) (c:\bin\sh.exe 1959) c:\ballista = normalize_win32_path (c:\ballista) (c:\bin\sh.exe 1959) /ballista = conv_to_posix_path (c:\ballista) (c:\bin\sh.exe 1959) 0x244E920 (/ballista) = getcwd_inner (0x244E920, 260, posix) (c:\bin\sh.exe 1959) /ballista = normalize_posix_path (.) (c:\bin\sh.exe 1959) . = conv_to_win32_path (.) (c:\bin\sh.exe 1959) 16 = GetFileAttributesA (.) (c:\bin\sh.exe 1959) 0 = stat (., 0x244F130) (c:\bin\sh.exe 1959) stat (/bin/sh.exe, 0x244F0BC) (c:\bin\sh.exe 1959) conv_to_win32_path (/bin) (c:\bin\sh.exe 1959) /bin = normalize_posix_path (/bin) (c:\bin\sh.exe 1959) c:\bin = conv_to_win32_path (/bin) (c:\bin\sh.exe 1959) conv_to_win32_path (/bin) (c:\bin\sh.exe 1959) /bin = normalize_posix_path (/bin) (c:\bin\sh.exe 1959) c:\bin = conv_to_win32_path (/bin) (c:\bin\sh.exe 1959) conv_to_win32_path (/bin/sh.exe) (c:\bin\sh.exe 1959) /bin/sh.exe = normalize_posix_path (/bin/sh.exe) (c:\bin\sh.exe 1959) c:\bin\sh.exe = conv_to_win32_path (/bin/sh.exe) (c:\bin\sh.exe 1959) conv_to_win32_path (/bin) (c:\bin\sh.exe 1959) /bin = normalize_posix_path (/bin) (c:\bin\sh.exe 1959) c:\bin = conv_to_win32_path (/bin) (c:\bin\sh.exe 1959) 0 = symlink_follow (/bin/sh.exe) -> /bin/sh.exe (c:\bin\sh.exe 1959) conv_to_win32_path (/bin/sh.exe) (c:\bin\sh.exe 1959) /bin/sh.exe = normalize_posix_path (/bin/sh.exe) (c:\bin\sh.exe 1959) c:\bin\sh.exe = conv_to_win32_path (/bin/sh.exe) (c:\bin\sh.exe 1959) 32 = GetFileAttributesA (c:\bin\sh.exe) (c:\bin\sh.exe 1959) open (c:\bin\sh.exe, 0x10000) (c:\bin\sh.exe 1959) fhandler_disk_file::open (c:\bin\sh.exe, 0x10000) (c:\bin\sh.exe 1959) conv_to_win32_path (c:\bin\sh.exe) (c:\bin\sh.exe 1959) conv_to_win32_path: c:\bin\sh.exe already win32 (c:\bin\sh.exe 1959) 0 = symlink_follow (c:\bin\sh.exe) -> c:\bin\sh.exe (c:\bin\sh.exe 1959) fhandler_base::open (c:\bin\sh.exe, 0x10000) (c:\bin\sh.exe 1959) conv_to_win32_path (c:\bin\sh.exe) (c:\bin\sh.exe 1959) conv_to_win32_path: c:\bin\sh.exe already win32 (c:\bin\sh.exe 1959) 52 = CreateFileA (c:\bin\sh.exe, 0x80000000, 0x7, 0x244EB84, 0x3, 0x80, 0) (c:\bin\sh.exe 1959) set handle to 52 (c:\bin\sh.exe 1959) 0x2EB0170 = fhandler_base::open (c:\bin\sh.exe, 0x10000) (c:\bin\sh.exe 1959) 0x2EB0170 = fhandler_disk_file::open (c:\bin\sh.exe, 0x10000) (c:\bin\sh.exe 1959) 3 = open (c:\bin\sh.exe, 0x10000) (c:\bin\sh.exe 1959) 1 = GetFileInformationByHandle (c:\bin\sh.exe, 52) (c:\bin\sh.exe 1959) conv_to_win32_path (c:\bin\sh.exe) (c:\bin\sh.exe 1959) conv_to_win32_path: c:\bin\sh.exe already win32 (c:\bin\sh.exe 1959) 0 = fstat (, 0x244F0BC) st_atime=33FCAF55 st_size=260096, st_mode=0x8180, st_ino=345968682, sizeof=64 (c:\bin\sh.exe 1959) 0 = fstat (3, 244F0BC) (c:\bin\sh.exe 1959) close (3) (c:\bin\sh.exe 1959) fhandler_base::close (handle 52) (c:\bin\sh.exe 1959) set handle to -99 (c:\bin\sh.exe 1959) 0 = close (3) (c:\bin\sh.exe 1959) Running delqueue (c:\bin\sh.exe 1959) 0 = stat (/bin/sh.exe, 0x244F0BC) (c:\bin\sh.exe 1959) stat (/gnuwin32/b18/H-i386-cygwin32/bin/sh.exe, 0x244F0BC) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32) (c:\bin\sh.exe 1959) /gnuwin32 = normalize_posix_path (/gnuwin32) (c:\bin\sh.exe 1959) c:\gnuwin32 = conv_to_win32_path (/gnuwin32) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18) (c:\bin\sh.exe 1959) /gnuwin32/b18 = normalize_posix_path (/gnuwin32/b18) (c:\bin\sh.exe 1959) c:\gnuwin32\b18 = conv_to_win32_path (/gnuwin32/b18) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32) (c:\bin\sh.exe 1959) /gnuwin32/b18/H-i386-cygwin32 = normalize_posix_path (/gnuwin32/b18/H-i386-cygwin32) (c:\bin\sh.exe 1959) c:\gnuwin32\b18\H-i386-cygwin32 = conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\bin\sh.exe 1959) /gnuwin32/b18/H-i386-cygwin32/bin = normalize_posix_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\bin\sh.exe 1959) c:\gnuwin32\b18\H-i386-cygwin32\bin = conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32) (c:\bin\sh.exe 1959) /gnuwin32 = normalize_posix_path (/gnuwin32) (c:\bin\sh.exe 1959) c:\gnuwin32 = conv_to_win32_path (/gnuwin32) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18) (c:\bin\sh.exe 1959) /gnuwin32/b18 = normalize_posix_path (/gnuwin32/b18) (c:\bin\sh.exe 1959) c:\gnuwin32\b18 = conv_to_win32_path (/gnuwin32/b18) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32) (c:\bin\sh.exe 1959) /gnuwin32/b18/H-i386-cygwin32 = normalize_posix_path (/gnuwin32/b18/H-i386-cygwin32) (c:\bin\sh.exe 1959) c:\gnuwin32\b18\H-i386-cygwin32 = conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\bin\sh.exe 1959) /gnuwin32/b18/H-i386-cygwin32/bin = normalize_posix_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\bin\sh.exe 1959) c:\gnuwin32\b18\H-i386-cygwin32\bin = conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin/sh.exe) (c:\bin\sh.exe 1959) /gnuwin32/b18/H-i386-cygwin32/bin/sh.exe = normalize_posix_path (/gnuwin32/b18/H-i386-cygwin32/bin/sh.exe) (c:\bin\sh.exe 1959) c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe = conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin/sh.exe) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32) (c:\bin\sh.exe 1959) /gnuwin32 = normalize_posix_path (/gnuwin32) (c:\bin\sh.exe 1959) c:\gnuwin32 = conv_to_win32_path (/gnuwin32) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18) (c:\bin\sh.exe 1959) /gnuwin32/b18 = normalize_posix_path (/gnuwin32/b18) (c:\bin\sh.exe 1959) c:\gnuwin32\b18 = conv_to_win32_path (/gnuwin32/b18) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32) (c:\bin\sh.exe 1959) /gnuwin32/b18/H-i386-cygwin32 = normalize_posix_path (/gnuwin32/b18/H-i386-cygwin32) (c:\bin\sh.exe 1959) c:\gnuwin32\b18\H-i386-cygwin32 = conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\bin\sh.exe 1959) /gnuwin32/b18/H-i386-cygwin32/bin = normalize_posix_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\bin\sh.exe 1959) c:\gnuwin32\b18\H-i386-cygwin32\bin = conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\bin\sh.exe 1959) 0 = symlink_follow (/gnuwin32/b18/H-i386-cygwin32/bin/sh.exe) -> /gnuwin32/b18/H-i386-cygwin32/bin/sh.exe (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin/sh.exe) (c:\bin\sh.exe 1959) /gnuwin32/b18/H-i386-cygwin32/bin/sh.exe = normalize_posix_path (/gnuwin32/b18/H-i386-cygwin32/bin/sh.exe) (c:\bin\sh.exe 1959) c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe = conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin/sh.exe) (c:\bin\sh.exe 1959) 32 = GetFileAttributesA (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe) (c:\bin\sh.exe 1959) open (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe, 0x10000) (c:\bin\sh.exe 1959) fhandler_disk_file::open (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe, 0x10000) (c:\bin\sh.exe 1959) conv_to_win32_path (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe) (c:\bin\sh.exe 1959) conv_to_win32_path: c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe already win32 (c:\bin\sh.exe 1959) 0 = symlink_follow (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe) -> c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe (c:\bin\sh.exe 1959) fhandler_base::open (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe, 0x10000) (c:\bin\sh.exe 1959) conv_to_win32_path (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe) (c:\bin\sh.exe 1959) conv_to_win32_path: c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe already win32 (c:\bin\sh.exe 1959) 80 = CreateFileA (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe, 0x80000000, 0x7, 0x244EB84, 0x3, 0x80, 0) (c:\bin\sh.exe 1959) set handle to 80 (c:\bin\sh.exe 1959) 0x2EB0170 = fhandler_base::open (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe, 0x10000) (c:\bin\sh.exe 1959) 0x2EB0170 = fhandler_disk_file::open (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe, 0x10000) (c:\bin\sh.exe 1959) 3 = open (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe, 0x10000) (c:\bin\sh.exe 1959) 1 = GetFileInformationByHandle (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe, 80) (c:\bin\sh.exe 1959) conv_to_win32_path (c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe) (c:\bin\sh.exe 1959) conv_to_win32_path: c:\gnuwin32\b18\H-i386-cygwin32\bin\sh.exe already win32 (c:\bin\sh.exe 1959) 0 = fstat (, 0x244F0BC) st_atime=33F3C50C st_size=260096, st_mode=0x81ED, st_ino=791209137, sizeof=64 (c:\bin\sh.exe 1959) 0 = fstat (3, 244F0BC) (c:\bin\sh.exe 1959) close (3) (c:\bin\sh.exe 1959) fhandler_base::close (handle 80) (c:\bin\sh.exe 1959) set handle to -99 (c:\bin\sh.exe 1959) 0 = close (3) (c:\bin\sh.exe 1959) Running delqueue (c:\bin\sh.exe 1959) 0 = stat (/gnuwin32/b18/H-i386-cygwin32/bin/sh.exe, 0x244F0BC) (c:\bin\sh.exe 1959) _sbrk (0) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB7000, end 0x36B7000, size 8388608, avail 8359936 (c:\bin\sh.exe 1959) 0x2EB7000 = sbrk (0x0) (total 0x7000) (c:\bin\sh.exe 1959) _sbrk (2048) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB7000, end 0x36B7000, size 8388608, avail 8359936 (c:\bin\sh.exe 1959) 0x2EB7000 = sbrk (0x800) (total 0x7800) (c:\bin\sh.exe 1959) _sbrk (0) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB7800, end 0x36B7800, size 8388608, avail 8357888 (c:\bin\sh.exe 1959) 0x2EB7800 = sbrk (0x0) (total 0x7800) (c:\bin\sh.exe 1959) _sbrk (4096) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB7800, end 0x36B7800, size 8388608, avail 8357888 (c:\bin\sh.exe 1959) 0x2EB7800 = sbrk (0x1000) (total 0x8800) (c:\bin\sh.exe 1959) open (/dev/tty, 0x0) (c:\bin\sh.exe 1959) set handle to -1 (c:\bin\sh.exe 1959) set handle to 23 (c:\bin\sh.exe 1959) fhandler_console: opened conin$ = 23, conout$ = 27 (c:\bin\sh.exe 1959) 3 = open (/dev/tty, 0x0) (c:\bin\sh.exe 1959) 0 = tcgetattr (console) (43C324) (win32 ENABLE_ 7) (t->lflag 6) (t->iflag 1) (c:\bin\sh.exe 1959) 0 = tcgetattr (3, 43C324); (c:\bin\sh.exe 1959) close (3) (c:\bin\sh.exe 1959) set handle to -1 (c:\bin\sh.exe 1959) 0 = close (3) (c:\bin\sh.exe 1959) Running delqueue (c:\bin\sh.exe 1959) _sbrk (0) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB8800, end 0x36B8800, size 8388608, avail 8353792 (c:\bin\sh.exe 1959) 0x2EB8800 = sbrk (0x0) (total 0x8800) (c:\bin\sh.exe 1959) _sbrk (2048) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB8800, end 0x36B8800, size 8388608, avail 8353792 (c:\bin\sh.exe 1959) 0x2EB8800 = sbrk (0x800) (total 0x9000) (c:\bin\sh.exe 1959) _sbrk (0) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB9000, end 0x36B9000, size 8388608, avail 8351744 (c:\bin\sh.exe 1959) 0x2EB9000 = sbrk (0x0) (total 0x9000) (c:\bin\sh.exe 1959) _sbrk (4096) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EB9000, end 0x36B9000, size 8388608, avail 8351744 (c:\bin\sh.exe 1959) 0x2EB9000 = sbrk (0x1000) (total 0xA000) (c:\bin\sh.exe 1959) lock_pinfo_for update : locking handle 44, timeout -1 (c:\bin\sh.exe 1959) unlock_pinfo : handle 44 (c:\bin\sh.exe 1959) -1 = waitpid (-1, 0x244F220, 1) (c:\bin\sh.exe 1959) _sbrk (0) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EBA000, end 0x36BA000, size 8388608, avail 8347648 (c:\bin\sh.exe 1959) 0x2EBA000 = sbrk (0x0) (total 0xA000) (c:\bin\sh.exe 1959) _sbrk (2048) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EBA000, end 0x36BA000, size 8388608, avail 8347648 (c:\bin\sh.exe 1959) 0x2EBA000 = sbrk (0x800) (total 0xA800) (c:\bin\sh.exe 1959) _sbrk (0) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EBA800, end 0x36BA800, size 8388608, avail 8345600 (c:\bin\sh.exe 1959) 0x2EBA800 = sbrk (0x0) (total 0xA800) (c:\bin\sh.exe 1959) _sbrk (4096) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EBA800, end 0x36BA800, size 8388608, avail 8345600 (c:\bin\sh.exe 1959) 0x2EBA800 = sbrk (0x1000) (total 0xB800) (c:\bin\sh.exe 1959) _sbrk (0) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EBB800, end 0x36BB800, size 8388608, avail 8341504 (c:\bin\sh.exe 1959) 0x2EBB800 = sbrk (0x0) (total 0xB800) (c:\bin\sh.exe 1959) _sbrk (2048) (c:\bin\sh.exe 1959) _sbrk: base 0x2EB0000, ptr 0x2EBB800, end 0x36BB800, size 8388608, avail 8341504 (c:\bin\sh.exe 1959) 0x2EBB800 = sbrk (0x800) (total 0xC000) (c:\bin\sh.exe 1959) _execve (/winnt/system32/arp, /winnt/system32/arp, 2EB6604) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32/arp) (c:\bin\sh.exe 1959) /winnt/system32/arp = normalize_posix_path (/winnt/system32/arp) (c:\bin\sh.exe 1959) c:\winnt\system32\arp = conv_to_win32_path (/winnt/system32/arp) (c:\bin\sh.exe 1959) /src/current/cdk/winsup/path.cc:1123 seterrno: 2 (ERROR_FILE_NOT_FOUND) -> 2 (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) 0 = symlink_follow (/winnt/system32/arp) -> /winnt/system32/arp (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32/arp) (c:\bin\sh.exe 1959) /winnt/system32/arp = normalize_posix_path (/winnt/system32/arp) (c:\bin\sh.exe 1959) c:\winnt\system32\arp = conv_to_win32_path (/winnt/system32/arp) (c:\bin\sh.exe 1959) can't find /winnt/system32/arp, trying with .exe appended (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) /winnt/system32/arp.exe = normalize_posix_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) c:\winnt\system32\arp.exe = conv_to_win32_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) 0 = symlink_follow (/winnt/system32/arp.exe) -> /winnt/system32/arp.exe (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) /winnt/system32/arp.exe = normalize_posix_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) c:\winnt\system32\arp.exe = conv_to_win32_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) spawn_guts (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) spawn_guts () argv[0] in `/winnt/system32/arp' (c:\bin\sh.exe 1959) spawn_guts () argv[0] out `/winnt/system32/arp' (c:\bin\sh.exe 1959) spawn_guts () argv[1] in `-a' (c:\bin\sh.exe 1959) spawn_guts () argv[1] out `-a' (c:\bin\sh.exe 1959) in init_from_exec (c:\bin\sh.exe 1959) spawn_guts: translating PATH=/bin:/gnuwin32/b18/H-i386-cygwin32/bin:/gnuwin32/b18/tcl/bin:/winnt/system32:/winnt:/perl5/bin:/usr/local/bin:/usr/X11/bin:. (c:\bin\sh.exe 1959) conv_to_win32_path (/bin) (c:\bin\sh.exe 1959) /bin = normalize_posix_path (/bin) (c:\bin\sh.exe 1959) c:\bin = conv_to_win32_path (/bin) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\bin\sh.exe 1959) /gnuwin32/b18/H-i386-cygwin32/bin = normalize_posix_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\bin\sh.exe 1959) c:\gnuwin32\b18\H-i386-cygwin32\bin = conv_to_win32_path (/gnuwin32/b18/H-i386-cygwin32/bin) (c:\bin\sh.exe 1959) conv_to_win32_path (/gnuwin32/b18/tcl/bin) (c:\bin\sh.exe 1959) /gnuwin32/b18/tcl/bin = normalize_posix_path (/gnuwin32/b18/tcl/bin) (c:\bin\sh.exe 1959) c:\gnuwin32\b18\tcl\bin = conv_to_win32_path (/gnuwin32/b18/tcl/bin) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/perl5/bin) (c:\bin\sh.exe 1959) /perl5/bin = normalize_posix_path (/perl5/bin) (c:\bin\sh.exe 1959) c:\perl5\bin = conv_to_win32_path (/perl5/bin) (c:\bin\sh.exe 1959) conv_to_win32_path (/usr/local/bin) (c:\bin\sh.exe 1959) /usr/local/bin = normalize_posix_path (/usr/local/bin) (c:\bin\sh.exe 1959) c:\usr\local\bin = conv_to_win32_path (/usr/local/bin) (c:\bin\sh.exe 1959) conv_to_win32_path (/usr/X11/bin) (c:\bin\sh.exe 1959) /usr/X11/bin = normalize_posix_path (/usr/X11/bin) (c:\bin\sh.exe 1959) c:\usr\X11\bin = conv_to_win32_path (/usr/X11/bin) (c:\bin\sh.exe 1959) conv_to_win32_path (.) (c:\bin\sh.exe 1959) conv_to_posix_path (c:\ballista, no-keep-rel) (c:\bin\sh.exe 1959) c:\ballista = normalize_win32_path (c:\ballista) (c:\bin\sh.exe 1959) /ballista = conv_to_posix_path (c:\ballista) (c:\bin\sh.exe 1959) 0x244E88C (/ballista) = getcwd_inner (0x244E88C, 260, posix) (c:\bin\sh.exe 1959) /ballista = normalize_posix_path (.) (c:\bin\sh.exe 1959) . = conv_to_win32_path (.) (c:\bin\sh.exe 1959) spawn_guts: setting PATH=c:\bin;c:\gnuwin32\b18\H-i386-cygwin32\bin;c:\gnuwin32\b18\tcl\bin;c:\winnt\system32;c:\winnt;c:\perl5\bin;c:\usr\local\bin;c:\usr\X11\bin;. (c:\bin\sh.exe 1959) spawn_guts: env count=35 bytes=922 (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) /winnt/system32/arp.exe = normalize_posix_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) c:\winnt\system32\arp.exe = conv_to_win32_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) 0 = symlink_follow (/winnt/system32/arp.exe) -> /winnt/system32/arp.exe (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) /winnt/system32/arp.exe = normalize_posix_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) c:\winnt\system32\arp.exe = conv_to_win32_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) /winnt = normalize_posix_path (/winnt) (c:\bin\sh.exe 1959) c:\winnt = conv_to_win32_path (/winnt) (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) /winnt/system32 = normalize_posix_path (/winnt/system32) (c:\bin\sh.exe 1959) c:\winnt\system32 = conv_to_win32_path (/winnt/system32) (c:\bin\sh.exe 1959) 0 = symlink_follow (/winnt/system32/arp.exe) -> /winnt/system32/arp.exe (c:\bin\sh.exe 1959) conv_to_win32_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) /winnt/system32/arp.exe = normalize_posix_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) c:\winnt\system32\arp.exe = conv_to_win32_path (/winnt/system32/arp.exe) (c:\bin\sh.exe 1959) spawn_guts (c:\winnt\system32\arp.exe, /winnt/system32/arp -a ) (c:\bin\sh.exe 1959) create_shared_fd_table: pid = 1959, no of fds = 64, map_name = cygwin32.fd_map.17.1959 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: buf = 0, buflen = 0 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing open fd 0:/dev/conin (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: len was 102 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing open fd 1:out (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: len was 30 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing open fd 2:out (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: len was 30 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 3 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 4 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 5 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 6 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 7 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 8 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 9 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 10 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 11 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 12 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 13 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 14 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 15 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 16 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 17 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 18 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 19 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 20 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 21 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 22 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 23 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 24 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 25 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 26 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 27 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 28 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 29 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 30 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 31 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 32 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 33 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 34 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 35 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 36 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 37 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 38 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 39 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 40 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 41 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 42 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 43 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 44 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 45 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 46 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 47 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 48 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 49 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 50 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 51 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 52 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 53 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 54 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 55 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 56 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 57 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 58 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 59 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 60 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 61 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 62 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 63 (c:\bin\sh.exe 1959) create_shared_fd_table: area = 56B0000, u->self->hmap.vec = 2EB0004, size = 246 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: buf = 56B0004, buflen = 246 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing open fd 0:/dev/conin (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: len was 102 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing open fd 1:out (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: len was 30 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing open fd 2:out (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: len was 30 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 3 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 4 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 5 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 6 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 7 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 8 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 9 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 10 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 11 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 12 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 13 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 14 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 15 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 16 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 17 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 18 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 19 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 20 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 21 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 22 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 23 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 24 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 25 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 26 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 27 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 28 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 29 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 30 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 31 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 32 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 33 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 34 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 35 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 36 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 37 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 38 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 39 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 40 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 41 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 42 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 43 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 44 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 45 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 46 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 47 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 48 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 49 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 50 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 51 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 52 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 53 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 54 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 55 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 56 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 57 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 58 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 59 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 60 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 61 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 62 (c:\bin\sh.exe 1959) hinfo_vec::linearize_fd_array: linearizing closed fd 63 (c:\bin\sh.exe 1959) 164 = spawn_guts (/winnt/system32/arp.exe, /winnt/system32/arp -a ) (c:\bin\sh.exe 1959) spawn_guts: parent = A000D90 (c:\bin\sh.exe 1959) spawn_guts: parent's handle = 136 (c:\bin\sh.exe 1959) lock_pinfo_for update : locking handle 44, timeout -1 (c:\bin\sh.exe 1959) spawn_guts: Dup hP 1 (c:\bin\sh.exe 1959) spawn_guts: Dup hT 1 (c:\bin\sh.exe 1959) unlock_pinfo : handle 44 (c:\bin\sh.exe 1959) spawn_guts: ExitProcess (REPARENTING) (c:\a.exe 1958) wait_for_single: 0 = WFMO (104, -1) (c:\a.exe 1958) lock_pinfo_for update : locking handle 44, timeout -1 (c:\a.exe 1958) wait_found: result = 100000 (c:\a.exe 1958) unlock_pinfo : handle 44 (c:\a.exe 1958) wait_found: reparenting (c:\a.exe 1958) lock_pinfo_for update : locking handle 44, timeout -1 (c:\a.exe 1958) unlock_pinfo : handle 44 (c:\a.exe 1958) wait_for_single: child pid 1959 (c:\a.exe 1958) wait_for_single: 0 = WFMO (160, -1) (c:\a.exe 1958) lock_pinfo_for update : locking handle 44, timeout -1 (c:\a.exe 1958) wait_found: result = 0 (c:\a.exe 1958) unlock_pinfo : handle 44 (c:\a.exe 1958) wait_found: 1959 = waitpid (-1, 0x240EF44, 0) (0x0) (c:\a.exe 1958) In cygwin_except_handler sig = 7 (c:\a.exe 1958) _exit (2) (c:\a.exe 1958) Terminating. (c:\a.exe 1958) 0 = kill_worker (1030, 20); (c:\a.exe 1958) Running delqueue (c:\a.exe 1958) lock_pinfo_for update : locking handle 44, timeout -1 (c:\a.exe 1958) unlock_pinfo : handle 44 (c:\a.exe 1958) destroy_pinfo_lock : handle 44 - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From acherman@inf.ethz.ch Fri Aug 22 07:05:00 1997 From: acherman@inf.ethz.ch (Erwin Achermann) Date: Fri, 22 Aug 1997 07:05:00 -0000 Subject: Vt100 ? Message-ID: <33FD92FC.DEB59A24@inf.ethz.ch> Hi, I am happily running Sergey's bash (GNU bash, version 2.01.0(4)-release (i386-pc-cygwin32)), But i would be even more happy if my $TERM would be set to vt100 or better. 'less' seems not to cope with the currently 'dumb' Terminal. Are there plans to give the bash vt100 capabilities per default? Where is (if at all) a vt100-emaulation for NT? Thanks for any help Erwin -- |Erwin Achermann phone: +41-1-632 74 40 | |Institute of Scientific Computing fax: +41-1-632 11 72 | |ETH Zurich, IFW C29.2 mailto:acherman@inf.ethz.ch | |8092 Zurich, Switzerland http://www.inf.ethz.ch/personal/acherman | - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From hgeleff@spidernet.com.cy Fri Aug 22 07:16:00 1997 From: hgeleff@spidernet.com.cy (Henrik Geleff) Date: Fri, 22 Aug 1997 07:16:00 -0000 Subject: winnt/95 authentication Message-ID: <199708221342.QAA04644@mail0.spidernet.net> Try starting by entering URL http://www.microsoft.com/isapi/goprem.asp?TARGET=/msdn/library/ and then navigate 'SDK Documentation'->'Windows Base Services'->Security. If you have a MSDN library CD handy, you can access the same information, just faster. Regards and good luck Henrik Geleff - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From rob@cygnus.com Fri Aug 22 10:57:00 1997 From: rob@cygnus.com (Rob Savoye) Date: Fri, 22 Aug 1997 10:57:00 -0000 Subject: expect/dejagnu for cygwin32? References: <199708182042.NAA08246.cygnus.gnu-win32@big.aa.net> Message-ID: <33FDD301.4B2E1628@cygnus.com> We're working on it now, and hopefully NT support will be in the next release of DejaGnu. (but probably not until sometime this fall. Expect is now working adequately, but needs more work. Tcl already runs. As DejaGnu is just a large Tcl/Expect script, once expect is ported, it'll "just work". - rob - - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mchase@ix.netcom.com Fri Aug 22 10:57:00 1997 From: mchase@ix.netcom.com (Michael A. Chase) Date: Fri, 22 Aug 1997 10:57:00 -0000 Subject: setting HOME Message-ID: <199708221446.JAA06251@dfw-ix1.ix.netcom.com> I've set $HOME in the environment of each of my NT accounts to /home/username and it seems to work for BASH. I don't use emacs. Mac :}) -----Original Message----- From: kendall shaw To: gnu-win32@cygnus.com Date: Thursday, August 21, 1997 22:28 Subject: setting HOME >1. Emacs wants HOME to have DOS filename conventions. If I set HOME to >c:/, then when I run bash with PS1='\w ' I get no prompt at root, and >"./usr" in /usr. Do you know a solution to this? > >2. I have multiple accounts on my nt box. I which I could do something >like HOME=c:/users/%username%, but bash doesn't know what %username% is, >and cmd bash just runs cmd. > >Any suggestions? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ashaikh@eecs.umich.edu Fri Aug 22 11:12:00 1997 From: ashaikh@eecs.umich.edu (Anees A. Shaikh) Date: Fri, 22 Aug 1997 11:12:00 -0000 Subject: Anybody tried compiling xfig 3.2? Message-ID: Hi, I'm trying to get xfig running on my Win95 box under b18. Using precompiled X11 libs from Arlindo and cygwin.dll from Sergey (not the latest) I was able to compile with the following mods: - replace calls to random/srandom with rand/srand - force it to use dirent.h structs rather than dir.h -- there is no dir.h I still got warnings because xfig unfortunately uses a macro called O_TEXT that is also in sys/fcntl.h (if I remember correctly) but I didn't think this would cause a problem. I also compiled libjpeg and libXpm pretty much out of the box (necessary for xfig). I did get some "incompatible pointer type" warnings for bcopy calls though that I ignored. My assumption that getting it to compile would result in a working binary didn't hold. It seems to start up okay but then immediately crashes with an exception. I haven't really given enough detail for someone to help debug what I've done but I'm hoping someone has been successful in getting a working xfig and can tell me what I've overlooked perhaps. Thanks very much. Anees ashaikh@eecs.umich.edu - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dgold@brown.edu Fri Aug 22 12:21:00 1997 From: dgold@brown.edu (Dan Goldwater) Date: Fri, 22 Aug 1997 12:21:00 -0000 Subject: problem building a dll that uses other dll's Message-ID: <199708221537.LAA15984@nut.brown.edu> Hi! I am trying to build a DLL which has function calls which must be resolved into existing win32 DLL's. I am using b18 on winNT 4.0. i have MSVC 5.0 available. Following the instructions for building a relocatable dll, i get an output but each of the three ld steps reports all the undefined references for the functions which are in the existing win32 dll's. i had to add the --noinhibit-exec option to ld so that it wouldn't delete the output dll due to the errors. if i try to use the result dll, it says it is an invalid image... when i build this project on solaris, the linker doesn't complain about the undefined references, and they are resolved to appropriate existing .so files at runtime. is there some option i need to give ld so it knows i want to resolve the references at runtime to the existing dll's? or am i clueless? thanks! --------------------------------------------------------- Dan Goldwater dgold@brown.edu http://fatmac.ee.cornell.edu/~goldwada --------------------------------------------------------- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Anthony@computronix.com Fri Aug 22 12:21:00 1997 From: Anthony@computronix.com (Anthony Tuininga) Date: Fri, 22 Aug 1997 12:21:00 -0000 Subject: Link Failure with 3rd party Windows DLLs Message-ID: <408173704E9ED011A9DB00009296CA0010D003@alpha.computronix.com> I looked at the mailing list archive and noticed that one other person got this result and had queried the list but I haven't seen a response to date. Perhaps this information will make it easier to track down the problem and suggest a solution -- I am a programmer by trade and so do not necessarily have a problem with making the fix myself. Note that I am not a compiler/linker programmer, however. :-) I compiled and linked successfully a C program (one of Oracle's OCI demo programs) against the OCIW32.LIB (Oracle Call Interface library) which contains pointers to the OCIW32.DLL and I got the following result when I tried to run the program: The procedure entry point could not be found in the dynamic link library OCIW32.DLL. Using nm and the link I noticed that the call points are null in the library OCIW32.LIB which seems the root cause of the problem (or is this normal?). Trying nm on the DLL itself simply says "no symbols". Any pointers at all are appreciated! I really like the GCC/G++ compilers and have used them extensively over the past five years -- this port to Windows NT will really make life a lot easier for me! Anthony - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Anthony@computronix.com Fri Aug 22 12:26:00 1997 From: Anthony@computronix.com (Anthony Tuininga) Date: Fri, 22 Aug 1997 12:26:00 -0000 Subject: Link Failure with 3rd party Windows DLLs Message-ID: <408173704E9ED011A9DB00009296CA0010D004@alpha.computronix.com> Oops! I forgot to mention that this is beta 18 under Windows NT 4.0 SP2. Sorry to waste bandwith. > -----Original Message----- > From: Anthony Tuininga > Sent: Friday, August 22, 1997 1:24 PM > To: 'gnu-win32@cygnus.com' > Subject: Link Failure with 3rd party Windows DLLs > > I looked at the mailing list archive and noticed that one other person > got this result and had queried the list but I haven't seen a response > to date. Perhaps this information will make it easier to track down > the problem and suggest a solution -- I am a programmer by trade and > so do not necessarily have a problem with making the fix myself. Note > that I am not a compiler/linker programmer, however. :-) > > I compiled and linked successfully a C program (one of Oracle's OCI > demo programs) against the OCIW32.LIB (Oracle Call Interface library) > which contains pointers to the OCIW32.DLL and I got the following > result when I tried to run the program: > > The procedure entry point could not be found in the dynamic link > library OCIW32.DLL. > > Using nm and the link I noticed that the call points are null in the > library OCIW32.LIB which seems the root cause of the problem (or is > this normal?). Trying nm on the DLL itself simply says "no symbols". > > Any pointers at all are appreciated! I really like the GCC/G++ > compilers and have used them extensively over the past five years -- > this port to Windows NT will really make life a lot easier for me! > > Anthony > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From suicide@freakz.org Sun Aug 24 17:03:00 1997 From: suicide@freakz.org (justin.) Date: Sun, 24 Aug 1997 17:03:00 -0000 Subject: beta 19. Message-ID: <3400CC4E.FC1B5FC2@freakz.org> Does anyone know when the expected release of beta 19 will be? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From suicide@freakz.org Sun Aug 24 19:22:00 1997 From: suicide@freakz.org (justin.) Date: Sun, 24 Aug 1997 19:22:00 -0000 Subject: .h files. Message-ID: <3400ECC7.2105D09A@freakz.org> Would any of these (linux) files every be included in a distrubition. netinet/in_systm.h netinet/tcp.h: etc... these are used in lots of internet based programs and it would provide a great deal of functionality to gnu-win32 if they were ported somehow. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Fuehrer@seabase.com Sun Aug 24 19:22:00 1997 From: Fuehrer@seabase.com (Gary Fuehrer) Date: Sun, 24 Aug 1997 19:22:00 -0000 Subject: Bug with Fix: mmap(MAP_PRIVATE) region handling in fork() Message-ID: <6CC63E2E4FC1D011A2A700609716117A266F11@seawolf> After a fork, a child does not see the same contents in its mmap regions as the parent does when the region was created with mmap(,,,MAP_PRIVATE,,) and the parent has written to that regjon prior to the fork. This bug is a result of the fork code failing to make the parent copy into the child process its version of mmap regions that have the FILE_MAP_COPY bit set. The following changes to three files (winsup.h, fork.cc, mmap.cc) will fix this bug. ================ winsup.h ================ Eliminate these declarations: -------------------------------- /* For mmaps across fork() */ int recreate_mmaps_after_fork(void *); void set_child_mmap_ptr(pinfo *); -------------------------------- ================ fork.cc ================ Rename the static "copy" function to the non-static "copy_memory_to_forkee" function: -------------------------------- int /* NOTE: not "static" anymore */ copy_memory_to_forkee (HANDLE child, void *low, void *high, int) : rc = copy_memory_to_forkee (pi.hProcess, (char *)proc_data_start, (char *)proc_data_end, 0); : rc = copy_memory_to_forkee (pi.hProcess, (char *)proc_bss_start, (char *)proc_bss_end, 1); : rc = copy_memory_to_forkee (child->hProcess, u->base, u->ptr, 2); : rc = copy_memory_to_forkee (child->hProcess, &x, u->initial_sp, 3); -------------------------------- Add these function declarations just before the "cygwin_fork_helper1" function: -------------------------------- int copy_mmap_records_to_forkee(pinfo *child); int recreate_mmaps_after_fork(void *); int copy_copyonwrite_mmaps_to_forkee(HANDLE hChild); -------------------------------- Replace the line "set_child_mmap_ptr (child);" with: -------------------------------- if (copy_mmap_records_to_forkee(child)) { small_printf ("fork_helper: copy of mmap_records failed\n"); set_errno (EAGAIN); syscall_printf ("-1 = fork(), copy of mmap_records failed\n"); TerminateProcess (child->hProcess, 1); u->forkee = 0; goto cleanup; } -------------------------------- Insert this block of code before the comment "* Now we have started...": -------------------------------- if(copy_copyonwrite_mmaps_to_forkee(child->hProcess)) { small_printf("fork_helper: copy_copyonwrite_mmaps_to_forkee failed\n"); set_errno (EAGAIN); syscall_printf ("-1 = fork(), copy of copyonwrite memory failed\n"); TerminateProcess (child->hProcess, 1); goto cleanup; } -------------------------------- Move this block of code so that it comes before the line "/* Tell our parent we've started." -------------------------------- if(recreate_mmaps_after_fork(u->self->mmap_ptr)) { small_printf("fork child: recreate_mmaps_after_fork_failed\n"); ExitProcess (1); } -------------------------------- ================ mmap.cc ================ Replace the functions "recreate_mmaps_after_fork", and "set_child_mmap_ptr" with these three functions that follow: -------------------------------- /* * Called by the parent process durring a fork. At this point, the child has * not yet run. When it does run and initialize, we want it to re-establish * the same mmap regions that we have. Because our heap, where our * mmap_records are stored, will not have been copied to the child at the time * it initializes, we can't simply pass it the "mmapped_areas" pointer. * Instead, we must copy the contents of our "mmapped_areas" into an array of * mmap_records where the child process can read them. The "child->mmap_ptr" * is set to a handle for the memory where the the child will find the array. * After we return, we will suspend and the child will initialize and call * "recreate_mmaps_after_fork" below. */ int copy_mmap_records_to_forkee(pinfo *child) { /* * Count the number of mmap_record entries are in the "mmapped_areas" map. */ unsigned count = 0; map< int, list * >::iterator it; if (mmapped_areas != NULL) { for( it = mmapped_areas->begin(); it != mmapped_areas->end(); ++it) { list *l = (*it).second; if(l != 0) { list::iterator li; for( li = l->begin(); li != l->end(); ++li) count++; } } } /* * Return with "child->mmap_ptr == NULL"; there are no mmap regions. */ if (count == 0) { child->mmap_ptr = NULL; debug_printf("copy_mmap_records_to_forkee: succeeded\n"); return 0; } /* * Create a file mapping that will be made available to the child process * and will hold an array of all the mmap_records in "mmapped_areas". */ HANDLE h = CreateFileMapping((HANDLE)-1, NULL, PAGE_READWRITE, 0, (count+1)*sizeof(mmap_record), NULL); if (h == NULL) { syscall_printf("copy_mmap_records_to_forkee: CreateFileMapping failed\n"); return -1; } mmap_record* pRec = (mmap_record*)MapViewOfFile(h, FILE_MAP_WRITE, 0, 0, 0); if (pRec == NULL) { CloseHandle(h); syscall_printf("copy_mmap_records_to_forkee: MapViewOfFile failed\n"); return -1; } /* * Now copy all the mmap_records into the mapped memory array. * The count of elements in this array are in the first record * along with some other useful info. */ *pRec = mmap_record(0, 0, (DWORD)mmapped_areas, count, pRec); if (mmapped_areas != NULL) { for( it = mmapped_areas->begin(); it != mmapped_areas->end(); ++it) { list *l = (*it).second; if(l != 0) { list::iterator li; for( li = l->begin(); li != l->end(); ++li) *(++pRec) = *li; } } } /* * Duplicate the handle for use by the child process. The child process * will use the duplicate handle in recreate_mmaps_after_fork. * We also unmap our view and close our handle, since we're done with it. */ UnmapViewOfFile(pRec - count); DuplicateHandle(GetCurrentProcess(), h, child->hProcess, (LPHANDLE)&child->mmap_ptr, FILE_MAP_READ, FALSE, DUPLICATE_CLOSE_SOURCE); debug_printf("copy_mmap_records_to_forkee: succeeded\n"); return 0; } /* * Call to re-create all the file mappings in a fork()'ed child. Called from * "cygwin_fork_helper1" by the child durring initialization. At this point, * the parent has duplicated its .data and .bss, but not its stack or heap. * We are passed a handle to memory via "u->self->mmap_ptr" where an array of * mmap_records can be found. This was put there by our parent in its call * to "copy_mmap_records_to_forkee". The first record in the array contains * information like the count of records. All the HANDLE's in those records * are valid for us (we inherited them), but none of the mapped areas * are in our address space. We need to iterate through the array and do the * MapViewOfFileEx calls. Initially, the "mmapped_areas" pointer is assumed to * be NULL, since this process was just created. Before returning, both * "mmapped_areas" and "u->self->mmap_ptr" are set to point to the mmap data * structure in the heap. This pointer is also passed to us in the first * mmap_record. Remember, that pointer refers to heap memory which the parent * process hasn't yet copied to us. After we return, we will suspend while we * wait for our parent to copy our heap (so that "mmaped_areas" is valid). * All of our mmap regions, except the ones that the parent created with * FILE_MAP_COPY, will refer to the same physical memory that the cooresponding * region in our parent refers to. In the case of FILE_MAP_COPY regions, we * will see the contents of the original file at the time the parent mmapped it. * Any changes made by the parent to these regions will not be seen by us. So, * the parent will also need to call "copy_copyonwrite_mmaps_to_forkee" below so * that we are in sync with the parent immediately after the fork. */ int recreate_mmaps_after_fork(void *param) { if (param != NULL) { HANDLE h = (HANDLE)param; mmap_record* pRec = (mmap_record*)MapViewOfFile(h, FILE_MAP_READ, 0, 0, sizeof(mmap_record)); if (pRec == NULL) { small_printf("-1 = recreate_mmaps_after_fork(): MapViewOfFile failed " "with GetLastError = %x\n", GetLastError()); return -1; } /* * Copy the information that was put in the first record * (see set_child_mmap_ptr). */ void* pRecDesired = pRec->get_address(); unsigned count = pRec->get_size(); mmapped_areas = (map< int, list * > *)pRec->get_offset(); /* * We need to have the array of mmap_records at the address "pRecDesired". * Otherwise, this array may be overlapping one of the mapped regions * in the parent that we are trying to duplicate. It is unlikely that * we are so lucky as to have "pRec" mapped correctly the first time, so * unmap this view and re-map it using MapViewOfFileEx. */ UnmapViewOfFile(pRec); if (count > 0) { pRec = (mmap_record*)MapViewOfFileEx(h, FILE_MAP_READ, 0, 0, 0, pRecDesired); if (pRec == NULL) { small_printf("-1 = recreate_mmaps_after_fork(): MapViewOfFileEx failed " "with GetLastError = %x\n", GetLastError()); return -1; } if (pRec != pRecDesired) { small_printf("-1 = recreate_mmaps_after_fork(): MapViewOfFileEx wrong address\n"); return -1; } /* * Copy the "pRec" mmap_record array into the * "mmaped_areas" map of mmap_record lists. */ while (count-- > 0) { ++pRec; /* Now re-create the MapViewOfFileEx call */ void* base = MapViewOfFileEx(pRec->get_handle(), pRec->get_access(), 0, pRec->get_offset(), pRec->get_size(), pRec->get_address()); if(base != pRec->get_address()) { small_printf("recreate_mmaps_after_fork: base address %x\ fails to match requested address %x\n", base, pRec->get_address()); return -1; } debug_printf("recreate_mmaps_after_fork: h = %x, access = %x, offset = %d, \ size = %d, address = %x\n", pRec->get_handle(), pRec->get_access(), pRec->get_offset(), pRec->get_size(), pRec->get_address()); } UnmapViewOfFile(pRecDesired); } CloseHandle(h); } /* Now set our mmap record in case the child forks. */ u->self->mmap_ptr = mmapped_areas; debug_printf("recreate_mmaps_after_fork: succeeded\n"); return 0; } /* * Called by the parent process durring a fork. At this point, * the address space of the child process has been established; * that is, the child has already called "recreate_mmaps_after_fork". * Any mmap region created with "FILE_MAP_COPY" will need to be * copied to the cooresponding memory addresses in the child process, * or else the parent and child may not see the same contents at those * addresses. They will be different for those FILE_MAP_COPY regions * that the parent has written to (unless the parent wrote what was * there originally). */ extern int copy_memory_to_forkee (HANDLE child, void *low, void *high, int); int copy_copyonwrite_mmaps_to_forkee(HANDLE hChild) { /* * Iterate through the array if mmap_records and copy those entries * that have FILE_MAP_COPY set. */ map< int, list * >::iterator it; if (mmapped_areas != NULL) { for( it = mmapped_areas->begin(); it != mmapped_areas->end(); ++it) { list *l = (*it).second; if(l != 0) { list::iterator li; for( li = l->begin(); li != l->end(); ++li) { mmap_record& rec = *li; char* start = (char*)rec.get_address() + rec.get_offset(); if ((rec.get_access() & FILE_MAP_COPY) && !copy_memory_to_forkee(hChild, start, start + rec.get_size(), 0)) { syscall_printf("copy_copyonwrite_mmaps_to_forkee: copy_memory_to_forkee failed\n"); return -1; } debug_printf("copy_copyonwrite_mmaps_to_forkee: h = %x, " "access = %x, offset = %d, size = %d, address = %x\n", rec.get_handle(), rec.get_access(), rec.get_offset(), rec.get_size(), rec.get_address()); } } } } debug_printf("copy_copyonwrite_mmaps_to_forkee: succeeded\n"); return 0; } -------------------------------- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From colin@bird.fu.is.saga-u.ac.jp Sun Aug 24 19:22:00 1997 From: colin@bird.fu.is.saga-u.ac.jp (Colin Peters) Date: Sun, 24 Aug 1997 19:22:00 -0000 Subject: Resource Compiler Message-ID: <01BCB146.DD5B3360@gbird0> Lucas Madar[SMTP:madar@lightlink.com] wrote: > Are there any freeware resource compilers/builders out there? I've tried >rcl, which gives me tons and tons of errors no matter how I try running the >preprocessor.. There is a resource compiler included with LCC if I remember correctly (slap me if I'm wrong) and there's also a resource compiler included with rsxnt (a package to build NT programs with DJGPP). I don't know if either of these alternatives will work much better though. You can get to them both from my homepage (see .sig, follow the GCC link, and then go to the Extras page). You might also try posting a short example of your .rc script and the typical error output. I have managed to get rcl to compile some reasonable scripts although there are still some severe bugs and holes in functionality (the version I have won't handle accelerators and seems to have problems with dialog box definitions sometimes which I have been able to work around by including extra dummy dialog box definitions in the resource file). Anyway, good luck, Colin. -- Colin Peters - Saga Univ. Dept. of Information Science -- colin@bird.fu.is.saga-u.ac.jp - finger for PGP public key -- http://www.fu.is.saga-u.ac.jp/~colin/index.html -- http://www.geocities.com/Tokyo/Towers/6162/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sun Aug 24 20:48:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sun, 24 Aug 1997 20:48:00 -0000 Subject: make cacking! Message-ID: <01BCB12B.82B21430@sos> Sandeep Neema wrote: > hi all, > excuse my ignorance but the make that comes with b18 distribution > always cacks on all my makefiles with error: > > Makefile:: *** missing seperator. Stop. > > I am lost. please help!! Gnu make requires rules lines to begin with a char. Your editor replaces tabs with the spaces. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sun Aug 24 20:52:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sun, 24 Aug 1997 20:52:00 -0000 Subject: Vt100 ? Message-ID: <01BCB12B.F53F55D0@sos> Erwin Achermann wrote: > Hi, > > I am happily running Sergey's bash (GNU bash, version 2.01.0(4)-release > (i386-pc-cygwin32)), But i would be even more happy if my $TERM would be > set to vt100 or better. 'less' seems not to cope with the currently > 'dumb' Terminal. Are there plans to give the bash vt100 capabilities per > default? Cygwin.dll emulates linux console terminal, so your TERM variable shall be set to "linux". -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Bardley09@aol.com Sun Aug 24 20:52:00 1997 From: Bardley09@aol.com (Bardley09@aol.com) Date: Sun, 24 Aug 1997 20:52:00 -0000 Subject: Resource Compiler Message-ID: <970824085049_671842487@emout01.mail.aol.com> I've used rc.exe, included in Microsoft's SDK (probably the hugest file, I don't remember). Anyway, I used it with the /r flag and then ran res2coff.exe on the .res file it produced. It linked dandily. I suppose it isn't exactly freeware, and probably binds you in some unpleasant way to service to Bill Gates in the future. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From root@jacob.remcomp.fr Sun Aug 24 20:52:00 1997 From: root@jacob.remcomp.fr (root) Date: Sun, 24 Aug 1997 20:52:00 -0000 Subject: Resource Compiler References: <3.0.2.32.19970820192621.00839240@light.lightlink.com> Message-ID: Try lcc at http://www.remcomp.com/lcc-win32 the resource compiler is called lrc, the resource editor weditres. Please report any bugs to me. -- Jacob Navia Logiciels/Informatique 41 rue Maurice Ravel Tel 01 48.23.51.44 93430 Villetaneuse Fax 01 48.23.95.39 France - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jamesl@albany.net Sun Aug 24 21:07:00 1997 From: jamesl@albany.net (Jim Lefavour) Date: Sun, 24 Aug 1997 21:07:00 -0000 Subject: Native GUI pe executables Message-ID: <33FE1CB4.830B67F5@albany.net> Hi! Please email a copy of the answer directly to me, if you don't mind - I wish to keep the load down, so I am not subscribed... According to the faq, adding -Wl --subsystem,windows to the gcc command line will result in a type '2' application (win32_GUI) however, a type 3 (Win32_CUI) app is still created. What is the correct way to fix this problem? Thanks! Jim Lefavour - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jeffdbREMOVETHIS@netzone.com Sun Aug 24 21:07:00 1997 From: jeffdbREMOVETHIS@netzone.com (Mikey) Date: Sun, 24 Aug 1997 21:07:00 -0000 Subject: Vt100 ? References: <33FD92FC.DEB59A24@inf.ethz.ch> Message-ID: <3401f68e.80333473@smtp.netzone.com> just put B18 termcap in /etc and set TERM to linux in ~/.bashrc On Fri, 22 Aug 1997 15:24:12 +0200, you wrote: >Hi, > >I am happily running Sergey's bash (GNU bash, version 2.01.0(4)-release >(i386-pc-cygwin32)), But i would be even more happy if my $TERM would be >set to vt100 or better. 'less' seems not to cope with the currently >'dumb' Terminal. Are there plans to give the bash vt100 capabilities per >default? >Where is (if at all) a vt100-emaulation for NT? > >Thanks for any help > >Erwin >-- >|Erwin Achermann phone: +41-1-632 74 40 | >|Institute of Scientific Computing fax: +41-1-632 11 72 | >|ETH Zurich, IFW C29.2 mailto:acherman@inf.ethz.ch | >|8092 Zurich, Switzerland http://www.inf.ethz.ch/personal/acherman | >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > (jeffdbREMOVETHIS@netzone.com) delete REMOVETHIS from the above to reply Mikey - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sun Aug 24 21:07:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sun, 24 Aug 1997 21:07:00 -0000 Subject: system() bug Message-ID: <01BCB12E.2CF9C670@sos> Tim Newsham wrote: > I've noticed occasional problems with system() before. Well at > the moment the problem seems to be happening every time so I figured > I'd take the occasion to get some debugging info. The test > program I used is: > > main() > { > system("/winnt/system32/arp -a"); > printf("done!\n"); > } > > At the moment, every time I run this program I get a crash. Did you try cygwin.dll with my patches? Interface: 201.1.1.2 on Interface 2 Internet Address Physical Address Type 201.1.1.1 00-40-05-14-18-64 dynamic done! -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From dahms@ifk20.mach.uni-karlsruhe.de Sun Aug 24 21:07:00 1997 From: dahms@ifk20.mach.uni-karlsruhe.de (dahms@ifk20.mach.uni-karlsruhe.de) Date: Sun, 24 Aug 1997 21:07:00 -0000 Subject: NFS Server Message-ID: <009B92FA.32696940.21873@ifk20.mach.uni-karlsruhe.de> Hi Murat, you wrote: : Is there a free NFS server for NT ? Is it possible to port Unix NFS : servers to NT with gnu-32. ? I do it the other way using SAMBA on unix. (Only gotcha it stopped after applying NT4.0 SP3! I had to set a registry entry to re-enable plaintext passwords, as explained in MS Knowledge Base article Q166730.) Note the traditional NFS uses UDP, not TCP. IIRC UDP isn't supported yet. Bye, Heribert (dahms@ifk20.mach.uni-karlsruhe.de) - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ebritten@uci.edu Sun Aug 24 21:07:00 1997 From: ebritten@uci.edu (Eric Britten) Date: Sun, 24 Aug 1997 21:07:00 -0000 Subject: Link Failure with 3rd party Windows DLLs References: <408173704E9ED011A9DB00009296CA0010D003@alpha.computronix.com> Message-ID: <34010311.5453AC1C@uci.edu> Anthony Tuininga wrote: > I looked at the mailing list archive and noticed that one other person > > got this result and had queried the list but I haven't seen a response > > to date. Perhaps this information will make it easier to track down > the > problem and suggest a solution -- I am a programmer by trade and so do > > not necessarily have a problem with making the fix myself. Note that I > > am not a compiler/linker programmer, however. :-) > > I compiled and linked successfully a C program (one of Oracle's OCI > demo > programs) against the OCIW32.LIB (Oracle Call Interface library) which > > contains pointers to the OCIW32.DLL and I got the following result > when > I tried to run the program: > You may have to make the import library yourself. First you will need a list of all the functions in the DLL.If they are stdcall functions make sure they have the @n on them. Place all this in a DEF file. for example(I have no what functions OCIW32.DLL exports) EXPORTS ociInitialize@4 ociUninitilize@0 This may generate the def file. ociw32.lib is the LIB you have now not the one you will be creating to use with GCC. echo EXPORTS > OCIW32.def nm ociw32.lib | egrep '^........ [T] _' | sed 's/[^_]*_//' >> OCIW32.def Use the following command line to generate the import library. dlltool --output-lib OCIW32.a --dllname OCIW32.dll --def OCIW32.def -k -k is to remove @n in the exported names. It will remove static and runtime linking errors. I have had this problem too and the above solution worked on setupapi.dll. In this case the import library that you would use to link with your program is ociw32.a. OCIW32.LIB must be a MSVC++ LIB. Hope this helps Eric Britten > The procedure entry point could not be found in the dynamic link > library OCIW32.DLL. > > Using nm and the link I noticed that the call points are null in the > library OCIW32.LIB which seems the root cause of the problem (or is > this > normal?). Trying nm on the DLL itself simply says "no symbols". > > Any pointers at all are appreciated! I really like the GCC/G++ > compilers > and have used them extensively over the past five years -- this port > to > Windows NT will really make life a lot easier for me! > > Anthony - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fjh@cs.mu.OZ.AU Sun Aug 24 23:36:00 1997 From: fjh@cs.mu.OZ.AU (Fergus Henderson) Date: Sun, 24 Aug 1997 23:36:00 -0000 Subject: DLL Problem References: <199708212048.WAA07599@bombadill.forwiss.uni-passau.de> Message-ID: <199708250613.QAA27153@murlibobo.cs.mu.OZ.AU> Roland Geier wrote: > [someone wrote:] > > I'm trying to port a tcl module from unix to win95 as a dll to load within > > wish. ... > > ext+0x15):libcmain.cc: undefined reference to `GetCommandLineA@0' > > c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t > > > > ext+0x1e):libcmain.cc: undefined reference to `WinMain@16' > > > > This seems to be a reentrancy problem, since a simple call like : > > fprintf(stderr, ...) > > ...for me this seems to be a forgotten '-lkernel32' at link-time :). The > undefined ref to 'WinMain' can be avoided with: > > echo 'int main() { return 0; }' > fixup.c; > echo 'asm (".section .idata$3\n" ".long 0,0,0,0,0,0,0,0");' >> fixup.c; > gcc -c fixup.c > # link fixup.o with your dll... Defining main() will supress the link error, but it will not actually fix the problem. The startup file (crt0.o, I think) containing a definition of `_impure_ptr' will be linked in, but _impure_ptr will not be initialized. This means that any reference to global variables such as `stdout' or `stderr' from within your DLL will be broken, and so your DLL will probably crash. To avoid the problem, you need to initialize your DLL's _impure_ptr from the main program. See < http://www.cs.mu.oz.au/~fjh/gnu-win32 > for an example of how to do this. -- Fergus Henderson | "I have always known that the pursuit WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From colin@bird.fu.is.saga-u.ac.jp Sun Aug 24 23:36:00 1997 From: colin@bird.fu.is.saga-u.ac.jp (Colin Peters) Date: Sun, 24 Aug 1997 23:36:00 -0000 Subject: CYGNUS-WIN32 and Oracle OCI7.2 Library Message-ID: <01BCB14B.9198B920@gbird0> Minh Le Cong[SMTP:Minh.Le_Cong@spacebel.be] wrote: >As you may know, Personal Oracle 7.2 for NT include an API called OCI (Oracle Call Interface). >This is used to access the Oracle database. >This is a C API . It consists of some .h files , a DLL (ociw32.dll) and OCIw32.lib > >The ociw32.lib is intended for MSVC users. > >So my question is that how can I link my program with OCI library if I use CYGNUS-WIN32 ? > how gcc and ld deal with .LIB libary files ( not .a library file)? In your case I think it might be easier to build a .a file using dlltool and a .def file. Some time ago Valery Fine (Valeri.Faine@cern.ch) posted a program called dumpexts, based on pedump, that creates a .def file from a .dll. Run this on ociw32.dll and then run dlltool on the resulting .def file to make a libociw32.a. Then you should be able to build your programs using -lociw32 on your link line. This all assumes that the library file contains only import records, not extra statically linked functions. From the looks of it, it seems like a good bet in this case. I can send you the source file dumpexts.c if you wish. Colin. -- Colin Peters - Saga Univ. Dept. of Information Science -- colin@bird.fu.is.saga-u.ac.jp - finger for PGP public key -- http://www.fu.is.saga-u.ac.jp/~colin/index.html -- http://www.geocities.com/Tokyo/Towers/6162/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From spong@usa.net Sun Aug 24 23:36:00 1997 From: spong@usa.net (Matt Spong) Date: Sun, 24 Aug 1997 23:36:00 -0000 Subject: expect/dejagnu for cygwin32? References: <199708182042.NAA08246@big.aa.net> Message-ID: <33FE3BEA.2781E494@usa.net> Wayne Stewart wrote: > > Have tcl, expect, and/or dejagnu been ported to cygwin32? > > Thanks, > Wayne Stewart > wstewart@aa.net > - > For help on using this list (especially unsubscribing), send a message to > "gnu-win32-request@cygnus.com" with one line of text: "help". Yeah... cygwin32 comes with a version of Tcl (haven't tried it though). -------------------------------------------- Matt Spong spong@usa.net - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From root@jacob.remcomp.fr Sun Aug 24 23:36:00 1997 From: root@jacob.remcomp.fr (root) Date: Sun, 24 Aug 1997 23:36:00 -0000 Subject: make cacking! References: <33FD61E2.77D0341E@cc.usu.edu> Message-ID: > > hi all, > excuse my ignorance but the make that comes with b18 distribution > always cacks on all my makefiles with error: > > Makefile:: *** missing seperator. Stop. > > I am lost. please help!! > - Isn't obvious man? You are missing the separator! More than 10 years ago I discovered UNIX, that strange OS of sorts. There I learned that make would stop telling me: > Makefile:: *** missing separator. Stop when I replaced inadvertenltly a TAB char with SPACES!!!!! Since TAB and SPACE are SO *easy* to distinguish in your computer screen or printout that was a HELL! 2 years ago someone posted the same question to this list, and I answered in similar terms: MAKE SHOULDN'T MAKE A DIFFERENCE BETWEEN TABS AND SPACES To no avail as it seems. I have always wandered why this bug doesn't go away... The reasons are many. People at cygnus are overworked, and this is a minor problem, and once you know it, you do not do it again, so this bug will never bite them, it harms only beginners... Maybe there are some crazy makefiles that rely in this 'feature', and people think that hypothetical makefiles would break down... bugs of this style tend to go forever: look at the Ctrl-Z bug of Microsoft that is still hunting us well over 10 years since DOS 2.0 that fixed it. Maybe we should start a contest of the longuest living bug around. I think this make *feature* would be a good candidate. -- Jacob Navia Logiciels/Informatique 41 rue Maurice Ravel Tel 01 48.23.51.44 93430 Villetaneuse Fax 01 48.23.95.39 France - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From colin@bird.fu.is.saga-u.ac.jp Sun Aug 24 23:36:00 1997 From: colin@bird.fu.is.saga-u.ac.jp (Colin Peters) Date: Sun, 24 Aug 1997 23:36:00 -0000 Subject: Include Path Error Message-ID: <01BCB14C.9D7D59C0@gbird0> Dan Rearick[SMTP:dan.rearick@cdmspa.com] wrote: [snip] >Entered: > gcc hello.c >Results: > hello.c:1: No include path in which to find stdio.h > GCC.EXE: Internal compiler error: program cpp got fatal signal 33 [snip] >GCC_EXEC_PREFIX = C:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\ >TCL_LIBRARY = C:/gnuwin32/b18/tcl/lib/tcl7.6 >GDBTK_LIBRARY = C:/gnuwin32/b18/share/gdbtcl >PATH=C:\WINDOWS;C:\WINDOWS\COMMAND;C:\MSSQL\BIN;C:\UTL;C:\PERL\BIN;C:\OPTIALL;C: >\KIPP\BIN;C:\PROGRAMF\MTI\OCRFF_16;C:\OCRFF\OCR\WORDSCAN.30;C:\OCRFF\KIPP\BIN;C: >\GNUWIN32\B18\H-I386-CYGWIN32\BIN;C:\GNUWIN32\B18\TCL\BIN;C:\GNUWIN32\B18\H-I386 >-CYGWIN32\LIB\GCC-LIB\I386-CYGWIN32\CYGNUS-2.7.2-970404;C:\GNUWIN32\B18\H-I386-C >YGWIN32\I386-CYGWIN32\INCLUDE [snip] You shouldn't need C:\GNUWIN32\B18\H-I386-CYGWIN32\LIB\GCC-LIB\I386-CYGWIN32\ CYGNUS-2.7.2-970404 in your PATH (unless you are running cpp directly... possibly an issue with rcl). It *might* be necessary to add it to LIBRARY_PATH. Neither do you need C:\GNUWIN32\B18\H-I386-CYGWIN32\I386-CYGWIN32\INCLUDE in your PATH, but you might try adding it to your C_INCLUDE_PATH and/or CPLUS_INCLUDE_PATH instead. On the other hand, GCC generally searches these paths automagically (and before environment variable or command line specified paths, which can be annoying if you want to exclude them) if everything is set up normally. Try running gcc with the -v option and looking at the least of searched include paths. Colin. -- Colin Peters - Saga Univ. Dept. of Information Science -- colin@bird.fu.is.saga-u.ac.jp - finger for PGP public key -- http://www.fu.is.saga-u.ac.jp/~colin/index.html -- http://www.geocities.com/Tokyo/Towers/6162/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From acherman@inf.ethz.ch Sun Aug 24 23:36:00 1997 From: acherman@inf.ethz.ch (Erwin Achermann) Date: Sun, 24 Aug 1997 23:36:00 -0000 Subject: bash: non English keyboard support References: <33FB560D.638425E1@informatik.uni-hamburg.de> Message-ID: <340127B8.13DB1A82@inf.ethz.ch> Dejan wrote: > > Hello, > > I have tested your bash for windows, and it works really fine,but > my keyboard is a german one, an some special characters like > the backslash the pipe symbol and other ones are usually typed with > the right ALT key pressed down, but this doesn't work with your bash. > > Have you any idea how to solve this problem?? > -- > D.D. Hamburg - Germany > email: 3djokic@informatik.uni-hamburg.de > or: dejan@stollmann.de > - > For help on using this list (especially unsubscribing), send a message > to > "gnu-win32-request@cygnus.com" with one line of text: "help". Check out Sergys page at http://miracle.geol.msu.ru/sos/ That bash has several bugs fixed, but it doesnt understand crlf as eol only lf accepted. Cheers Erwin -- |Erwin Achermann phone: +41-1-632 74 40 | |Institute of Scientific Computing fax: +41-1-632 11 72 | |ETH Zurich, IFW C29.2 mailto:acherman@inf.ethz.ch | |8092 Zurich, Switzerland http://www.inf.ethz.ch/personal/acherman | - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From hinds@inreach.com Mon Aug 25 00:28:00 1997 From: hinds@inreach.com (Alexander Hinds) Date: Mon, 25 Aug 1997 00:28:00 -0000 Subject: Has anybody succesfully compiled nuweb0.90? Message-ID: <3.0.3.32.19970825002808.02d35b50@inreach.com> I've recently become interested in literate programming and tried compiling nuweb 0.90; it actually compiles ok, but on running dies a horrible death. Does anybody have a working version? Thanks in advance. ---------------------------------------------------------------------------- --------- 22:03:06:[~/nuweb]$ ./nuweb nuweb.w ./nuweb: you'll need to rerun nuweb after running latex (c:\home\ahinds\nuweb\nuweb.exe 1180) In cygwin_except_handler (c:\home\ahinds\nuweb\nuweb.exe 1180) Exception trapped! (c:\home\ahinds\nuweb\nuweb.exe 1180) exception C0000005 at 1002A38C (c:\home\ahinds\nuweb\nuweb.exe 1180) exception: ax 0 bx 1 cx 2EB3804 dx 0 (c:\home\ahinds\nuweb\nuweb.exe 1180) exception: si 403370 di 100476C8 bp 240EB4 C sp 240EB4C (c:\home\ahinds\nuweb\nuweb.exe 1180) exception is: STATUS_ACCESS_VIOLATION (c:\home\ahinds\nuweb\nuweb.exe 1180) Stack trace: (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 0: sp = 0x240E980, pc = 0x1000CEC2 (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 1: sp = 0x240E99C, pc = 0x77F94512 (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 2: sp = 0x240E9C0, pc = 0x77F88EEB (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 3: sp = 0x240EA4C, pc = 0x77F76266 (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 4: sp = 0x240EB4C, pc = 0x1002B5C0 (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 5: sp = 0x240EB68, pc = 0x1002B610 (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 6: sp = 0x240EB7C, pc = 0x4033D9 (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 7: sp = 0x240ED98, pc = 0x4033CD (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 8: sp = 0x240EFAC, pc = 0x4013C7 (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 9: sp = 0x240F104, pc = 0x1000C102 (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 10: sp = 0x240FF94, pc = 0x1000C113 (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 11: sp = 0x240FFA0, pc = 0x405D21 (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 12: sp = 0x240FFB0, pc = 0x40103B (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 13: sp = 0x240FFC0, pc = 0x77F1B304 (c:\home\ahinds\nuweb\nuweb.exe 1180) frame 14: sp = 0x240FFF0, pc = 0x0 (c:\home\ahinds\nuweb\nuweb.exe 1180) End of stack trace - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From walter.jenny@gmsg.ch Mon Aug 25 03:49:00 1997 From: walter.jenny@gmsg.ch (walter.jenny) Date: Mon, 25 Aug 1997 03:49:00 -0000 Subject: Startup of daemon process Message-ID: <9708251048.AA17926@nxser> Hi whenn running a process as a daemon with fork() and setsid() the startup code in crt0 is not executed properly. A number of calls (getpwent(), gethostbyname()) fail after becoming a daemon since some internal data structures are not initialized. Will this be fixed in b19? Thanks Walter Jenny - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From hubert.fauque@wanadoo.fr Mon Aug 25 06:36:00 1997 From: hubert.fauque@wanadoo.fr (Hubert FAUQUE) Date: Mon, 25 Aug 1997 06:36:00 -0000 Subject: problem with mounting / as binary or not Message-ID: <34038957.76387559@smtp.wanadoo.fr> I have installed Sergey's cygwin.dll and bash and I am having a problem: bash didn't find .bashrc at startup, so as I have seen on a previous message I have mounted c: as / with text=binary and it works for bash; but now is the problem: make doesn't find any include files; if there is include file in the Makefile, it gives the message file^M: no such file or directory it doesn't find the file because it adds a ^M at the end of the name; Has anybody found a solution? thanks Hubert - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From rminnich@sarnoff.com Mon Aug 25 06:51:00 1997 From: rminnich@sarnoff.com (Ron G. Minnich) Date: Mon, 25 Aug 1997 06:51:00 -0000 Subject: NFS Server References: <009B92FA.32696940.21873@ifk20.mach.uni-karlsruhe.de> Message-ID: try the linux nfs server, user mode. I almost got it going once just for fun. Ron Minnich |Java: an operating-system-independent, rminnich@sarnoff.com |architecture-independent programming language (609)-734-3120 |for Windows/95 and Windows/NT on the Pentium ftp://ftp.sarnoff.com/pub/mnfs/www/docs/cluster.html - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From madar@lightlink.com Mon Aug 25 06:51:00 1997 From: madar@lightlink.com (Lucas Madar) Date: Mon, 25 Aug 1997 06:51:00 -0000 Subject: Broken libraries? Message-ID: <3.0.2.32.19970825065017.00846760@light.lightlink.com> This may be a stupid question.. but here. gcc -o test test.c test.o -luser32 -lwsock32 -lgdi32 -lshell32 -Wl,--subsystem,windows test.c: In function `listenfor': test.c:133: warning: assignment makes pointer from integer without a cast //d/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/bin/ld.exe: warning: cannot find entry symbol _WinMainCRTStartup; defaulting to 00401000 C:\WINDOWS\TEMP\cc0010001.o(.text+0x5b2):test.c: undefined reference to `Shell_N otifyIconA' GCC.EXE: Internal compiler error: program ld got fatal signal 1 Shell_NotifyIconA is obviously a symbol in libshell32.a gcc -o ddetest dde.c -luser32 dde.c: In function `main': dde.c:86: warning: assignment makes pointer from integer without a cast dde.c:87: warning: assignment makes pointer from integer without a cast dde.c:90: warning: assignment makes pointer from integer without a cast C:\WINDOWS\TEMP\cc0010001.o(.text+0x218):dde.c: undefined reference to `DdeClientTransaction' GCC.EXE: Internal compiler error: program ld got fatal signal 1 DdeClientTransaction is obviously a symbol in libuser32.a Why are these not working at all? ----- Lucas Madar / darkrot IRC Administrator, hebron.in.us.dal.net email: madar@lightlink.com / darkrot@dal.net whois: LM1691 Warning labels thwart the process of natural selection. If at first you don't succeed, well that's pretty much it for skydiving. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jeffdbREMOVETHIS@netzone.com Mon Aug 25 06:51:00 1997 From: jeffdbREMOVETHIS@netzone.com (Mikey) Date: Mon, 25 Aug 1997 06:51:00 -0000 Subject: termcap References: <34012F71.8D0CF6A8@inf.ethz.ch> Message-ID: <34014c0d.102224177@smtp.netzone.com> Sorry, I believe it's hardcoded in most unix apps so unless you want to rebuild from source, and modify EVERY ref, you're probably better off either making an /etc directory, or making a link ln -s /YourInstallPath/H-i386-cygwin32/etc/ /etc/ and then place all of your admin files in one or the other. for termcap, you could put it as ~/.termcap, most termcap implementations look there first. but passwd, group, and the rest of the admin files need to be in /etc On Mon, 25 Aug 1997 09:08:33 +0200, you wrote: >Mikey wrote: >> >> just put B18 termcap in /etc >> and set TERM to linux in >> ~/.bashrc >> >Thank you, that helped indeed. Say, where is the /etc Path coded is it >hardwired, i'd rather leafe termcap where it is and not introdoce a new >/etc directory. > >Thank you > >Erwin >-- >|Erwin Achermann phone: +41-1-632 74 40 | >|Institute of Scientific Computing fax: +41-1-632 11 72 | >|ETH Zurich, IFW C29.2 mailto:acherman@inf.ethz.ch | >|8092 Zurich, Switzerland http://www.inf.ethz.ch/personal/acherman | > (jeffdbREMOVETHIS@netzone.com) delete REMOVETHIS from the above to reply Mikey - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From demircio@boun.edu.tr Mon Aug 25 06:51:00 1997 From: demircio@boun.edu.tr (Murat Demircioglu) Date: Mon, 25 Aug 1997 06:51:00 -0000 Subject: NFS Server References: <199708241719.MAA32270@cirrusnet.com> Message-ID: Hi, Thanx for your help.. I got sossnt. It also has the source code and it works great... And one more question: If i port linux user level nfs server to NT with cygnus tools may it be sold? I'm writing a commercial product for newspapers and i want to bundle an NFS server with it.. Murat Demircioglu - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kcantrel@dg-rtp.dg.com Mon Aug 25 07:28:00 1997 From: kcantrel@dg-rtp.dg.com (Ken Cantrell) Date: Mon, 25 Aug 1997 07:28:00 -0000 Subject: Unix spooler Message-ID: <9708251316.AA02771@dg-rtp.dg.com> You should have a lpr.exe in \WINNT\SYSTEM32 (assuming you named directories in the standard way) which supports TCP/IP printing. It probably isn't what you're wanting, but it might be closer than the suggestions mentioned so far. ******************************************************************** Ken Cantrell DG Performance Modeling x5879 kcantrel@dg.rtp.com *** Don't forget to smile...that's why God gave us Fangs *** -----Original Message----- [chunks have been removed from the following ... ] From: Rob Hegt Subject: Re: Re[2]: Unix spooler >> Date: Mon, 18 Aug 1997 21:16:13 -0700 (PDT) >> From: Alex >> On Mon, 18 Aug 1997 arifi_koseoglu@karmaint.com wrote: >> > How'bout if the printer is not on te Parallel port but on the network ? >> Perhaps use the UNC pathname for it? > >This may all be true, but I guess that what was initially intended was an >lpr/lpd suite for win32. So, what is the status on that? > >Rob Hegt - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From koi@ssa.crane.navy.mil Mon Aug 25 07:28:00 1997 From: koi@ssa.crane.navy.mil (mark koi) Date: Mon, 25 Aug 1997 07:28:00 -0000 Subject: Out of environment space Message-ID: <34018643.156B02B4@ssa.crane.navy.mil> When adding new environment variables for gnu-win32 on win95, I was receiving OUT OF ENVIRONMENT SPACE. I added the following to my config.sys file shell=c:\windows\command.com /e:512 /p And it worked fine, but damn my system is slow. I imagine I am running a command interpreter on top of the WIN95 kernel(?). Is there a better way. I would hope so, because I could learn a second language with the time it takes to load apps. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From zeil@cs.odu.edu Mon Aug 25 16:21:00 1997 From: zeil@cs.odu.edu (Steven J. Zeil) Date: Mon, 25 Aug 1997 16:21:00 -0000 Subject: Anybody tried compiling xfig 3.2? References: Message-ID: On Fri, 22 Aug 1997, Anees A. Shaikh wrote: > Hi, > > I'm trying to get xfig running on my Win95 box under b18. Using > precompiled X11 libs from Arlindo and cygwin.dll from Sergey (not the > latest) I was able to compile with the following mods: > > - replace calls to random/srandom with rand/srand > - force it to use dirent.h structs rather than dir.h -- there is no > dir.h > I compiled xfig 3.1.4 using the same process. > I also compiled libjpeg Yes, I used version 6a > and libXpm pretty much out of the box > (necessary for xfig). I didn't need that. Is it optional, or a change between xfig 3.1.4 and 3.2? > My assumption that getting it to compile would result in a working > binary didn't hold. It seems to start up okay but then immediately > crashes with an exception. > I had a few of those that I traced to having set up my X server stuff incorrectly. Make sure that you can run the various binaries that come with Arlindo's X libs. SJZ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mskow@earthling.net Mon Aug 25 16:21:00 1997 From: mskow@earthling.net (Michael Skowronski) Date: Mon, 25 Aug 1997 16:21:00 -0000 Subject: cc1plus.exe error Message-ID: Hi, i get the following error: (cc1plus.exe 1002) In cygwin_except_handler when i try to compile certain c++ sources. it doesn;t happen with all of them. anyone have an idea? Mike - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Fuehrer@seabase.com Mon Aug 25 16:21:00 1997 From: Fuehrer@seabase.com (Gary Fuehrer) Date: Mon, 25 Aug 1997 16:21:00 -0000 Subject: SML-NJ for CygWin32 Message-ID: <6CC63E2E4FC1D011A2A700609716117A266F12@seawolf> > Sergey Okhapkin wrote: > > ---------- > From: Sergey Okhapkin[SMTP:sos@prospect.com.ru] > Sent: Sunday, August 24, 1997 12:09 AM > To: 'gnu-win32@cygnus.com'; Gary Fuehrer > Cc: 'Geoffrey Noer' > Subject: RE: SML-NJ for CygWin32 > > Gary Fuehrer wrote: > > > > 2. The signal handling in CygWin32 doesn't let me do the > following: > > > > a. Get and Set the "eip" register (or any register) of the > > > > excepting thread. > > > > b. Ascertain the kind of floating point exception that > occurred. > > > > The need is for "siginfo_t" and "sigcontext" parameters that can > be > > > > optionally received and modified by signal handlers. > > I have a patch for exception.cc in winsup to add this functionality > and > > fix a related bug. > > I hope you made modifications in call_handler() and wrap assembly > code? > Current sources contains all you need to implement this > functionality: > (Body of call_handler() here...) > The saved integer registers are accessible from signal handler. Save > float > context just near pusha/pushf and modify sigarg/sigfunc/oldmask > offsets. > I turns out that technique won't quite work. The exceptions that I need to handle start in "__cygwin_exception_handler" such as "integer divide by zero". But the current source for "call_handler" is only good for dispatching events starting in "ctrl_c_handler". For Beta 18, someone mistakenly thought that having "__cygwin_exception_handler" call "ctrl_c_handler" for its implementation would be a good idea. While it was a good idea to have a common signal dispatcher, the current "call_handler" can't dispatch both kinds of events. The reason is that "call_handler" assumes the current thread is not the excepting thread; one of the first things "call_handler" does is it suspends the excepting thread. But for any event starting in "__cygwin_exception_handler", the current thread is the excepting thread. Consequently, "call_handler" never dispatches and never returns for events like "integer divide by zero". So, yes, I made modifications to call_handler() and the wrap assembly code to get things to work. But it's not worth sharing for these reasons: 1. My patch is an awful kludge on top of a mistake. 2. I didn't take the time yet to provide the second and third arguments to signal handlers for events that start in "ctrl_c_handler". 3. The second and third arguments that I'm passing to the signal handlers do not conform to POSIX.4 which it should and which wouldn't be hard to do. I'd be glad to make exceptions.cc a correct, POSIX.4 compliant signal dispatcher. But to do it, I need the correct versions of exceptions.cc and signal.cc to start from. The Beta 18 starting point that I have now is missing the new beta 19 stuff and the Sergey Okhapkin patches. How can we coordinate this effort? > > > > 3. There is no SIGVTALRM signal. > > SML-NJ doesn't seem to need this in order to function. For now, I'm > > providing a dummy #define. > > I don't know how to implement virtual timer on win32. > According to Geoffrey Noer, setitimer() and getitimer() are implemented in beta 19. If I understand it correctly, that is the virtual timer and therefore SIGVTALRM will be available. > > > > 4. "L_ctermid" is not defined (but "L_cuserid" is). > > I don't remember why SML-NJ needs this. For now, I'm providing a > dummy > > #define. It's not critical. > > L_ctermid define should be added to stdio.h, and ctermid() call should > be > added to winsup sources for POSIX conformance. > That's the way I see it, too. Even a stub for ctermid() would be helpful for beta 19. > > > > 5. The following definitions are not in the Posix sockets > include > > > > files (only the Windows socket stuff) > > The missing defines already added to headers. > Which ones? Beta 19? Or one of your patches? > -- > Sergey Okhapkin, http://www.lexa.ru/sos > Moscow, Russia > Looking for a job > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From racke@gundel.han.de Mon Aug 25 16:21:00 1997 From: racke@gundel.han.de (Stefan Hornburg) Date: Mon, 25 Aug 1997 16:21:00 -0000 Subject: make cacking! References: <33FD61E2.77D0341E@cc.usu.edu> Message-ID: Sandeep Neema writes: > > hi all, > excuse my ignorance but the make that comes with b18 distribution > always cacks on all my makefiles with error: > > Makefile:: *** missing seperator. Stop. > > I am lost. please help!! GNU Make expects the following syntax: : \t \t .... Other makes (like Watcom's wmake (really broken at all)) are satisfied with whitespace .... Ciao Racke -- This site is powered by Linux 2.0.29 and his friends GNU CC 2.7.2, C Library 5.4.23, C++ Library 2.7.2.1, Binutils 2.8.1.0.1. Surf to Racke's Webnode (URL: http://www.han.de/~racke )! - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mich0101@tc.umn.edu Mon Aug 25 16:21:00 1997 From: mich0101@tc.umn.edu (Ares) Date: Mon, 25 Aug 1997 16:21:00 -0000 Subject: Compiling with templates Message-ID: I'm in need of using some templates in C++ code, to implement a linked list, and am having problems with ld linking it. It seems to find that there is no such functions from my Templated class. The object files for the linked list come from several different files. I was wondering what switches to use on the command line to solve my problem. Thanks Jason _/_/_/_/_/ _/_/_/_/ _/ _/ Jason D. Michaelson _/ _/ _/ _// _// Jason.D.Michaelson-1@umn.edu _/ _/ _/ _/_/ _/_/ jason13@geocities.com _/ _/ _/ _/ _/ _/ _/ mich0101@itlabs.umn.edu _/ _/ _/ _/ _/ _/ _/ WebPage: http://www.geocities.com _/ _/ _/ _/ _/ _/_/ _/ /Siliww.tc.umn.edu/~mich0101 _/_/_/ _/_/_/_/ _/ _/ _/ http://www.umn.edu/~mich0101 Don't you wish you were using Pegasus Mail for Windoze. If you were, you'd be able to click on the URL's. :-). Get it at http://www.pegasus.usa.com/ . Damn, another one of those URL's. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Scott.H.Chang@nab02.usace.army.mil Mon Aug 25 16:21:00 1997 From: Scott.H.Chang@nab02.usace.army.mil (Chang NAB02, Scott H) Date: Mon, 25 Aug 1997 16:21:00 -0000 Subject: gs5.03 Message-ID: Hello, Please tell me what specific files in gs503w32 and gs503ini are for Windows NT 4.0 to read PostScript. Thank you in advance, Scott Chang - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From wa_perkins@pnl.gov Mon Aug 25 16:21:00 1997 From: wa_perkins@pnl.gov (William A. Perkins) Date: Mon, 25 Aug 1997 16:21:00 -0000 Subject: Off topic filesystem question References: <199708220021.RAA13778@emptiness.arceneaux.com> Message-ID: <199708251507.IAA25054@yama.phoebus> >>>>> "Joe" == Joseph Arceneaux writes: Joe> This is not really a cygwin question, so I must apologize in Joe> advance. Joe> I have a Unix file system which I'm trying to render onto a CD. Joe> The CD burner only runs off a Windows system (or potentially, Joe> NT). The Question is how to preserve the many softlinks in Joe> the file system so they appear properly on the CD. Joe> Any suggestions would be appreciated. Joe> Joe Joe, I have done this successfully by first making a ISO9660 image using mkisofs on the UNIX system (in my case Solaris and SunOS) first, then copying the image to the PC and burning the CD there. Here are some helpful links to check out: * Linux CD-Writing < http://www.shop.de/~winni/linux/CD-Writing/ > discusses the use of mkisofs. * Golden Hawk Technology < http://www.goldenhawk.com/ > has some free software which will burn a CD using an image, if your software won't do it. * mkisofs source can be found in any of the Linux archives one place would be < ftp://tsx-11.mit.edu/pub/linux/packages/mkisofs/mkisofs-1.11.tar.gz > I have successfully compiled mkisofs under cygwin32-b18 (on Win95) by defining NON_UNIXFS, but it won't make a good image. isovfy and isoinfo work under cygwin32 though. I didn't put much effort into finding the problem. Has anyone else had any luck? Hope this helps. Bill -- Bill Perkins Battelle Pacific Northwest Laboratories Environmental Technologies Division, Hydrology Group P.O. Box 999 MSIN K9-36 Richland, Washington, USA 99352 voice: (509) 372-6131 fax: (509) 372-6089 email: wa_perkins@pnl.gov - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jpump@researchmag.com Mon Aug 25 18:19:00 1997 From: jpump@researchmag.com (Jason Pump) Date: Mon, 25 Aug 1997 18:19:00 -0000 Subject: Porting Unix code to NT Message-ID: <3.0.1.32.19970825180319.0093d4a0@sf.looksmart.com.au> Hello, I'm looking at porting a large series of interdependent fast cgi programs for Apache web server to NT. They use such wonderful features as sockets, mmap'ing, some arrays that are defined with non constant expressions, gdbm, templates, and a host of other goodies. I've gotten about halfway through an NT port and decided to take a look at gnu-win32 after running across some unresolvable architecture conflicts. Has anyone out there done serious, stable, code with gnu-win32, or would I be better off continuing down my conversion path? - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From startnet@nation.org Mon Aug 25 18:19:00 1997 From: startnet@nation.org (Derek Reynolds) Date: Mon, 25 Aug 1997 18:19:00 -0000 Subject: Cygwin Ports Archieve here Message-ID: <1.5.4.32.19970826011407.0069c134@nation.org> Hi there. I have finnaly posted all of the programs i have ported via cygwin. I have also made an archive of other *nix ports to win32 environment. Please check it out: http://www.h0ar.org/myn If there is anything i am missing please let me know. If you have any ported programs that you think should be on here, please email me startnet@nation.org with the location to where it can be downloaded (its an archive). Thanks Derek Reynolds - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jeffdbREMOVETHIS@netzone.com Mon Aug 25 23:11:00 1997 From: jeffdbREMOVETHIS@netzone.com (Mikey) Date: Mon, 25 Aug 1997 23:11:00 -0000 Subject: problem building a dll that uses other dll's References: <199708251740.NAA14070@nut.brown.edu> <199708251852.OAA02984@nut.brown.edu> Message-ID: <34023e4c.26952970@smtp.netzone.com> These are from the SDK_4_ming_cygwin32 lib2def does the obvious, the Makefile should show you how to use it. If you don't have the .lib file, you can use impdef.exe to generate the .def file, you may need to edit it to eliminate c++ junk, (vc++ has different name mangling conventions then g++) just redirect the output of impdef to a file, then run dlltool on it as shown in the enclosed Makefile. On Mon, 25 Aug 1997 14:54:39 -500, you wrote: >> The dll is not the stub library, the stub library >> for foo.dll is libfoo.a > >ok, the thing is that i don't have the stub library for foo.dll. foo.dll >is from a commercial product. does this mean i'm stuck? if i had the .a >files i would have been all set long ago :) > >is there some way to generate the .a file for foo.dll? > >> BTW all of this stuff is in the ld man page, and also >> in gcc.info, you might try reading them. > >really i already knew all the basics of ld, the stuff i'm having the >problem with is specific to gnuwin32. Then why are you asking about -l "small L"? One of the basics of ld, is that you don't link directly with coff or aout shared libraries, you have to make a stub library first. That's the main reason why linux when to ELF. > >thanks. >--------------------------------------------------------- >Dan Goldwater dgold@brown.edu > http://fatmac.ee.cornell.edu/~goldwada >--------------------------------------------------------- > > > (jeffdbREMOVETHIS@netzone.com) delete REMOVETHIS from the above to reply Mikey -------------- next part -------------- A non-text attachment was scrubbed... Name: impdef.c Type: text/x-c Size: 10780 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: lib2def Type: text/x-shellscript Size: 519 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Makefile Type: text/x-makefile Size: 2961 bytes Desc: not available URL: From Fuehrer@seabase.com Mon Aug 25 23:11:00 1997 From: Fuehrer@seabase.com (Gary Fuehrer) Date: Mon, 25 Aug 1997 23:11:00 -0000 Subject: SML-NJ for CygWin32 Message-ID: <6CC63E2E4FC1D011A2A700609716117A266F13@seawolf> Brad, Shortly, I will e-mail you a couple of tar.gz files with the stuff you need plus instructions. I will mail this to anyone who wants it. Just e-mail your request to me directly (fuehrer@seabase.com) so we don't clutter up the gnu-win32 mailing list. Please specify whether or not you want me to send a pre-built cygwin.dll that includes my patches to winsup, as it is 544 kBytes uncompressed. The SML-NJ team has accepted these changes and are incorporating them now. I'm working through this mailing list to get the CygWin32 changes incorporated into beta 19. My goal is to get SML-NJ 109.31 and CygWin32 beta 19 working together without patches. We'll see how it goes. Since I haven't pointed it out explicitly anywhere else, I just inform everyone now: SML-NJ can't work under CygWin32 on top of Win95. Only NT. It's because the SML-NJ runtime for unix variants (of which CygWin32 is one) depends on a functioning mmap(). And mmap() can't be implemented by CygWin32 (or anyone else) for Win95 (see winsup/mmap.cc for a special comment about Win95). Thanks for your interest, Gary PS. Denis, are you still interested? > ---------- > From: Brad Porter[SMTP:Brad.Porter@MCI.Com] > Sent: Sunday, August 24, 1997 1:22 AM > To: Gary Fuehrer > Subject: Re: SML-NJ for CygWin32 > > Do you have diff files or something I can use to get 109.30 to run > under > CygWin32 as well? > > Thanks. > > Gary Fuehrer wrote: > > > I have now successfully gotten SML-NJ 109.30 to build and run under > > CygWin32. To accomplish this, changes had to be made to both > cygwin.dll > > (the winsup project) as well as to the SML-NJ runtime source code. > The > > changes to SML-NJ are just to add the necessary settings for the > > CygWin32 platform. I am now attempting to get the SML-NJ team to > > incorporate those changes. Because they are just additional > settings to > > support another platform, I anticipate the SML-NJ team will gladly > > accept them. > > ... > > > -- > Brad Porter > 719-535-4135 > v622-4135 > > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jeffdbREMOVETHIS@netzone.com Mon Aug 25 23:11:00 1997 From: jeffdbREMOVETHIS@netzone.com (Mikey) Date: Mon, 25 Aug 1997 23:11:00 -0000 Subject: problem building a dll that uses other dll's References: <199708251740.NAA14070@nut.brown.edu> Message-ID: <3402d0f2.136250896@smtp.netzone.com> The dll is not the stub library, the stub library for foo.dll is libfoo.a to use ld.exe directly, you would need to do ld --subsystem [ windows/console ] --dll -e _dll_entry@12 -o prog.exe prog.o -L/search/dir -lfoo the -l automatically adds the lib prefix, and the .a suffix. to the stub library name -L is used to add a search path for ld. The order of files on the link line IS important, since ld is a one pass linker, references must be resolved by object files farther allong in the link line. BTW all of this stuff is in the ld man page, and also in gcc.info, you might try reading them. On Mon, 25 Aug 1997 13:42:04 -500, you wrote: >> You must tell ld (or link.exe) which libraries to >> link with. > >hmm.. what option are you using to do this? i tried "-l foo.dll" but it >doesn't seem to work: > >gnuwin32/H-i386-cygwin32/i386-cygwin32/bin/ld.exe: cannot open >-lhttpdlw.dll: No such file or directory > >maybe i just need a different syntax? but i think it is expecting a .a >library. do you know how to do this? > >> easeist way is to let gcc do it. >> >> to link do >> >> gcc -Wl,--dll,-e,_dll_entry@12 -o prog.exe prog.o >> >> the libraries will automatically be linked in. > >i tried this, but without specifying the libraries (since i couldn't >figure out the syntax for that), and it did the same thing as before - >"undefined reference ...". > >thanks! > >> >I am trying to build a DLL which has function calls which must be >> >resolved into existing win32 DLL's. >> > >> >I am using b18 on winNT 4.0. i have MSVC 5.0 available. >> > >> >Following the instructions for building a relocatable dll, i get an >> >output but each of the three ld steps reports all the undefined >> >references for the functions which are in the existing win32 dll's. i had >> >to add the --noinhibit-exec option to ld so that it wouldn't delete the >> >output dll due to the errors. if i try to use the result dll, it says it >> >is an invalid image... > >--------------------------------------------------------- >Dan Goldwater dgold@brown.edu > http://fatmac.ee.cornell.edu/~goldwada >--------------------------------------------------------- > > > (jeffdbREMOVETHIS@netzone.com) delete REMOVETHIS from the above to reply Mikey - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From speck@maighels.informatik.uni-tuebingen.de Mon Aug 25 23:11:00 1997 From: speck@maighels.informatik.uni-tuebingen.de (Andreas Speck) Date: Mon, 25 Aug 1997 23:11:00 -0000 Subject: POSIX IPC in GNU WIN32 C++-Compiler Message-ID: <199708251843.UAA08062@maighels.wsi> Dear CYGNUS Team Currently I'm porting software (including IPC mechanisms like semaphores and shared memory) from UNIX to WindowsNT. The compiler I use is the CYGNUS GNU WIN32 C++-Compiler. - Does the CYGNUS GNU WIN32 C++-Compiler support POSIX (or any other UNIX, e.g. System V) semaphores (or other synchronizing mechanism) and shared memory? - Are there any other concepts used in the CYGNUS GNU WIN32 C++-Compiler for exchanging large amount of data among processes? - Is there any sample code ilustrating the use of IPC mechanisms? Thank you! Regards Andreas Speck email: speck@informatik.uni-tuebingen.de http: http://www-sr.informatik.uni-tuebingen.de/~speck ------------------------------------------------------------------------------ University of Tuebingen Wilhelm-Schickard-Department for Computer Science Research Group Symbolic Computation Sand 13 D-72076 Tuebingen Phone: +49(7071) 29-78961 Fax: +49(7071) 67540 ------------------------------------------------------------------------------ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bwomack@ptc.com Mon Aug 25 23:11:00 1997 From: bwomack@ptc.com (Brian David Womack) Date: Mon, 25 Aug 1997 23:11:00 -0000 Subject: make: vfork (heap_chunk_size too small) Error Message-ID: <3401F817.914521FD@ptc.com> Several people have reported this fairly serious error in either make or gcc before. It occurs while making dependencies for a file with a LOT of includes. Could Cygnus please correct this problem in Beta 19? gcc -M -w -D_M_IX86=400 -D_MSC_VER=1020 -U__STDC__ -IC:/Progra~1/MSDEV/include -IC:/Progra~1/MSDEV/mfc/include -DDSG_DEVELOPMENT -D_DEBUG -DCRTAPI1=_cdecl -DCRTAPI2=_cdecl -D_X86_=1 -DWINVER=0x0400 -DWIN32 -D_WIN32 -D_MT -DRUNNING_NT=1 -DDONT_AA_POINTS -Di486_nt -Ic:/users/bwomack/nt_dbg/dsg/util/ui++/resources -Id:/users/bwomack/trunk/third_party/uslnew/src/incl -IS:/Progra~1/Galaxy.cxx/include -Id:/users/bwomack/trunk/render/include -Id:/users/bwomack/trunk/render/include -Id:/users/bwomack/trunk/dsg/modeler/include -Ic:/users/bwomack/nt_dbg/dsg/modeler/ui/resources -Ic:/users/bwomack/nt_dbg/dsg/modeler/ui/resources/messages -IS:/Progra~1/ProE/proe/toolkit/includes -Ic:/users/bwomack/nt_dbg/dsg/modeler/include -Id:/users/bwomack/trunk/dsg/util/include -IS:/Progra~1/ProE/prodevelop/includes -IS:/Progra~1/ProE/prodevelop/prodev_appls/includes -IS:/Progra~1/ProE/prodevelop/ptc_includes -IS:/Progra~1/ProE/protoolkit/includes -Id:/users/bwomack/trunk/dsg/ptc/proimage //d/users/bwomack/trunk/dsg/modeler/mh/crv_edit_pts_eh.cxx > tmp.d The heap has been split, CYGWIN can't fork this process. Increase the heap_chunk_size in the registry and try again. make: vfork: Not enough space --------------------------------------------------------------- Brian Womack Parametric Technology Corp 801-588-4611/4650 Voice/FAX bwomack@ptc.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From grospell@limeil.cea.fr Mon Aug 25 23:11:00 1997 From: grospell@limeil.cea.fr (G Grospellier) Date: Mon, 25 Aug 1997 23:11:00 -0000 Subject: Macro SNDMSG definition missing in Functions.h [GNUWIN32-B18 Win95] Message-ID: <199708260611.IAA03265@fermi.limeil.s> I am using GnuWin32-B18 under W95. When I use the functions CommDlg_*, defined in 'Windows32/Functions.h', the symbol SNDMSG is undefined at link time. I think it is because SNDMSG is a macro but you did not defined it. In BC5.0 or MSVC4.0, it is defined like that: #ifdef __cplusplus #define SNDMSG ::SendMessage #else #define SNDMSG SendMessage #endif I am sorry if is bug has already been reported but I did not seen it anywhere. Gilles. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mphalpin@iag.net Tue Aug 26 02:52:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Tue, 26 Aug 1997 02:52:00 -0000 Subject: Here are 3 socket problems I've found in GNUWIN32 Message-ID: <199708260207.WAA20559@iag.net> I have the following problems with socket functions: (1) SO_USELOOPBACK (error message: SO_USELOOPBACK - not supported) opt = 1; if (setsockopt(fd, SOL_SOCKET, SO_USELOOPBACK, (char *)&opt, sizeof(opt)) < 0) report_error("setsockopt(SO_USELOOPBACK) %s:%s", cptr); (2) SNDBUF and RCVBUF (not even defined) if (setsockopt(fd, SOL_SOCKET, SO_SNDBUF, (char *)&opt, sizeof(opt)) < 0) report_error("setsockopt(SO_SNDBUF) %s:%s", cptr); (3) select() doesn't allow for enough fds select only allows for around 60 or so FDs, while some Unix apps use 4095! (4) /proc/net/tcp isn't output. This is used by Identd. If anyone wants to add this, I'll send them all the info. It shouldn't be too hard. This isn't a problem, but it is something I think should be added. - SJ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Tue Aug 26 02:52:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Tue, 26 Aug 1997 02:52:00 -0000 Subject: SML-NJ for CygWin32 Message-ID: <01BCB20F.682E9C70@gater.krystalbank.msk.ru> Gary Fuehrer wrote: > I turns out that technique won't quite work. The exceptions that I need > to handle start in "__cygwin_exception_handler" such as "integer divide > by zero". But the current source for "call_handler" is only good for > dispatching events starting in "ctrl_c_handler". For Beta 18, someone > mistakenly thought that having "__cygwin_exception_handler" call > "ctrl_c_handler" for its implementation would be a good idea. While it > was a good idea to have a common signal dispatcher, the current > "call_handler" can't dispatch both kinds of events. The reason is that > "call_handler" assumes the current thread is not the excepting thread; It's already fixed, but not a good way - __cygwin_exception_handler just raises a corresponding signal now and returns. The failed context is lost when call_handler is called. > I'd be glad to make exceptions.cc a correct, POSIX.4 compliant signal > dispatcher. But to do it, I need the correct versions of exceptions.cc > and signal.cc to start from. The Beta 18 starting point that I have now > is missing the new beta 19 stuff and the Sergey Okhapkin patches. How > can we coordinate this effort? You'll need the entire current winsup sources - there was too many changes after B18. Ask Geoffrey. > According to Geoffrey Noer, setitimer() and getitimer() are implemented > in beta 19. If I understand it correctly, that is the virtual timer and > therefore SIGVTALRM will be available. Only the real, but not a virtual timer is implemented. > > L_ctermid define should be added to stdio.h, and ctermid() call should > > be > > added to winsup sources for POSIX conformance. > > > That's the way I see it, too. Even a stub for ctermid() would be > helpful for beta 19. Geoffrey, please add the following lines to termios.cc (and to cygwin.din, of course): extern "C" char * ctermid (char *s) { static char buf[L_ctermid]; if (u->self->ctty != -1) { sprintf (buf, "/dev/tty%d", u->self->ctty); } else buf[0] = '\0'; if (s) strcpy (s, buf); return buf; } And #define L_ctermid 16 to stdio.h. > > The missing defines already added to headers. > > > Which ones? Beta 19? Or one of your patches? Both B19 sources and my patches. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Steven.Pemberton@cwi.nl Tue Aug 26 02:52:00 1997 From: Steven.Pemberton@cwi.nl (Steven Pemberton) Date: Tue, 26 Aug 1997 02:52:00 -0000 Subject: b18: cygwin_except_handler failure on deref NULL Message-ID: I have a program that generates limits.h and float.h. It is also an extremely good compiler tester. You can find it at: http://www.cwi.nl/~steven/enquire.html An earlier version is part of the gcc distribution and apparently was used to generate the float.h include file of b18. Compiling and running it with b18 causes a run-time failure (and bash hangs) with the message: (ENQUIRE.EXE 1011) In cygwin_except_handler I don't know what the 1011 refers to, but if it's a line number, that is not where it failed. It failed at the point where the program looks to see if dereferencing NULL causes a trap (which is done using setjmp, and dereferencing NULL). Steven Pemberton, CWI, Amsterdam; Steven.Pemberton@cwi.nl - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Daviid.Smethurst@pancredit.com Tue Aug 26 08:08:00 1997 From: Daviid.Smethurst@pancredit.com (David Smethurst) Date: Tue, 26 Aug 1997 08:08:00 -0000 Subject: The patch ? Message-ID: Hi, I've seen alot of talk about a patch that is available and I was wondering where I could get this from . Thanks Dave --- D.Smethurst Developer - PanCredit Systems Ltd. #include - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mmccourt@iona.com Tue Aug 26 11:26:00 1997 From: mmccourt@iona.com (Maurice McCourt) Date: Tue, 26 Aug 1997 11:26:00 -0000 Subject: _main: arg/environment allocation error. Message-ID: <199708261623.RAA14479@dublin.iona.ie> Hello, I keep getting a message: _main: arg/environment allocation error. when trying to run 'vi' or using scripts which use pipes extensively. I'm running beta 17.1 on NT 4.0. Any solutions suggestions appreciated. thanks, Maurice. --------------------------------------------------------------------------- Maurice McCourt----------| email: mmccourt@iona.com --| ---- voice xtn:2283 IONA Technologies -------| ftp: ftp.iona.com -------| tel: +353-1-6625255 The IONA Building -------| WWW: http://www.iona.com/ | fax: +353-1-6625244 8-10 Lower Pembroke St. -| In the USA call: 1-800 orbix4u, 1-800 6724948 Dublin 2 ----------------------------------------------------------------- Ireland ------------------------------------------------------------------ Orbix wins prestigious 'best ORB middleware' Well Connected Award from Network Computing Magazine http://techweb.cmp.com/nc/808/808wellcon22.html --------------------------------------------------------------------------- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Tue Aug 26 11:26:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Tue, 26 Aug 1997 11:26:00 -0000 Subject: tar from/to floppy disk Message-ID: <01BCB091.A98EB620@sos> Michael Reinelt wrote: > As I said, it only should. With NT4 Server, the floppy starts up, and > after a few seconds, I get the following: > > tar: Cannot open /dev/fd0: Not super-user > Take cygwin.dll from my web page. It works with tar floppies on NT (single-volume archives only). -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Bardley09@aol.com Tue Aug 26 11:26:00 1997 From: Bardley09@aol.com (Bardley09@aol.com) Date: Tue, 26 Aug 1997 11:26:00 -0000 Subject: Winsock.h compilation error Message-ID: <970824090739_1092303419@emout13.mail.aol.com> I can't figure this out. I'm trying to compile a program which includes the winsock.h header file. Here are the errors I get, followed by the corresponding areas in winsock.h. At first I thought it was the IN thing, which I was able to locate (I think) and the file it's in says it's a "pseudo keyword" #defined to nothing. You may be thinking, "gee, this guy's helpless." You are wise. C:\mingw32\win32\include/winsock.h:843: syntax error before `SOCKET' C:\mingw32\win32\include/winsock.h:855: syntax error before `SOCKET' C:\mingw32\win32\include/winsock.h:868: syntax error before `PVOID' BOOL PASCAL FAR TransmitFile ( IN SOCKET hSocket, [843] IN HANDLE hFile, IN DWORD nNumberOfBytesToWrite, IN DWORD nNumberOfBytesPerSend, IN LPOVERLAPPED lpOverlapped, IN LPTRANSMIT_FILE_BUFFERS lpTransmitBuffers, IN DWORD dwReserved ); BOOL PASCAL FAR AcceptEx ( IN SOCKET sListenSocket, [855] IN SOCKET sAcceptSocket, IN PVOID lpOutputBuffer, IN DWORD dwReceiveDataLength, IN DWORD dwLocalAddressLength, IN DWORD dwRemoteAddressLength, OUT LPDWORD lpdwBytesReceived, IN LPOVERLAPPED lpOverlapped ); VOID PASCAL FAR GetAcceptExSockaddrs ( IN PVOID lpOutputBuffer, [868] IN DWORD dwReceiveDataLength, IN DWORD dwLocalAddressLength, IN DWORD dwRemoteAddressLength, OUT struct sockaddr **LocalSockaddr, OUT LPINT LocalSockaddrLength, OUT struct sockaddr **RemoteSockaddr, OUT LPINT RemoteSockaddrLength ); - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jman@lx.net Tue Aug 26 11:26:00 1997 From: jman@lx.net (jman) Date: Tue, 26 Aug 1997 11:26:00 -0000 Subject: tcl swig1.1 Message-ID: <3.0.3.32.19970826112639.006b67dc@lx.net> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Howdy, Has anyone worked with libtcl76.a that comes with B18? Has anyone used swig. I got it to compile but anything I try to compile comes up with undefined refrence's to tcl_???? so on. I am not that good that I can figure it all out so I was wondering if someone else has. I'd love to be able to use swig with gnuwin32. Thanks, -----BEGIN PGP SIGNATURE----- Version: PGP for Personal Privacy 5.0 Charset: noconv iQA/AwUBNAMDvg6ne3t4b32aEQKLoQCdHzQgdZbCfnRikgfid69mg2kWZ9kAoLfD LmQIUbuLZwyZWqaLRe3FUK/3 =UZoH -----END PGP SIGNATURE----- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jaalto@tre.tele.nokia.fi Tue Aug 26 11:26:00 1997 From: jaalto@tre.tele.nokia.fi (Jari Aalto) Date: Tue, 26 Aug 1997 11:26:00 -0000 Subject: Your text files could be be in .html Message-ID: <199708241702.NAA01692@helene.tele.nokia.fi> Hi, I just Browsed the great http://www.cygnus.com/misc/gnu-win32/ and looked into the ftp directories where you have some text files. I'm and Emacs elisp programmer and I have developed a simple text file format which I call 'Technical text' (TF). I Just though to drop you a note if you're interested in keeping your text documentation in that format. ftp://cs.uta.fi/pub/ssjaaa/tiny-docs.html#378 The benefits: o You write text file as usual, Only indent it a bit differently (TAB set to 4) o There is special Emacs minor mode to help you: format paragraphs, create TOC automatically etc. o You can feed the file to t2html.pls Perl4 script that turns the apge into identical .html page. All links are automatically made clickable, link verification feature is coming real soon In short: With few (2 minute) work, you can convert any .txt file into .html. And you don't need html editor for that. As an example, see my web pages that are created from plain text files. ftp://cs.uta.fi/pub/ssjaaa/tiny-tools.tar.gz -- Whole kit ftp://cs.uta.fi/pub/ssjaaa/t2html.pls -- Perl script "What's Technical format's layout like?". Well, you just look at in this mail message. Nothing special, just indentation. Cheers! jari - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sbiswas@novell.com Tue Aug 26 11:26:00 1997 From: sbiswas@novell.com (Soumen Biswas) Date: Tue, 26 Aug 1997 11:26:00 -0000 Subject: fork - Is there a problem: Message-ID: Hi , I am new to gnu-win32 . I have the following prob: while forking ( ver - b18 .) /* Example Code*/ int main() { int x; if( !fork() ) printf("Child\n" ); else printf("Parent\n"); /* If wait is commented out the program shows errors*/ wait( &x ) ; return x ; } The the above code works fine BUT IF wait() is commented out the program becomes unpredictable ( sometimes it works ... sometimes shows EXCEPTION TRAPPED ... I think it depends on which process finishes first - chid || parent . Is the above behaviour a bug or is this the way fork() is supposed to work in cygwin Is there a workaround ????? thanx Soumen sbiswas@novell.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bartlee@hotmail.com Tue Aug 26 11:26:00 1997 From: bartlee@hotmail.com (Bartlee Anderson) Date: Tue, 26 Aug 1997 11:26:00 -0000 Subject: Looking for java implementation of X Windows Message-ID: <19970824073906.12588.qmail@hotmail.com> Has anyone seen a Java only implementation of X. I saw something that I thought I saved, but can't find. May have been on another list. I know Sergey would also be interested in this. ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mh@mike.franken.de Tue Aug 26 11:26:00 1997 From: mh@mike.franken.de (Michael Hirmke) Date: Tue, 26 Aug 1997 11:26:00 -0000 Subject: bash: non English keyboard support References: <33FB560D.638425E1@informatik.uni-hamburg.de> Message-ID: <6cTSwIW4pfB@mike.franken.de> Hi Dejan, > Hello, > > I have tested your bash for windows, and it works really fine,but > my keyboard is a german one, an some special characters like > the backslash the pipe symbol and other ones are usually typed with > the right ALT key pressed down, but this doesn't work with your bash. > > Have you any idea how to solve this problem?? Get Sergey's patched cygwin.dll off his web site: http://miracle.geol.msu.ru/sos/ He has fixed this bug and many others. Bye. Michael. -- Michael Hirmke | Telefon +49 (911) 557999 Georg-Strobel-Strasse 81 | FAX +49 (911) 557664 90489 Nuernberg | E-Mail mh@mike.franken.de | WWW http://minimike.franken.de/ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Wed Aug 27 07:06:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: SML-NJ for CygWin32 Message-ID: <01BCB077.8501DD60@sos> Gary Fuehrer wrote: > > > 2. The signal handling in CygWin32 doesn't let me do the following: > > > a. Get and Set the "eip" register (or any register) of the > > > excepting thread. > > > b. Ascertain the kind of floating point exception that occurred. > > > The need is for "siginfo_t" and "sigcontext" parameters that can be > > > optionally received and modified by signal handlers. > I have a patch for exception.cc in winsup to add this functionality and > fix a related bug. I hope you made modifications in call_handler() and wrap assembly code? Current sources contains all you need to implement this functionality: res = SuspendThread (OurhThread); debug_printf ("suspend said %d error %d\n", res, GetLastError()); orig.ContextFlags = CONTEXT_CONTROL | CONTEXT_INTEGER; res = GetThreadContext (OurhThread, &orig); unsigned int *sp = (unsigned int *) orig.Esp; *(--sp) = orig.Eip; /* rethere = orig.Eip;*/ orig.Eip = (unsigned) &&wrap; *(--sp) = oldmask; *(--sp) = n; *(--sp) = (DWORD) pfunc; orig.Esp = (DWORD) sp; SetThreadContext (OurhThread, &orig); /* Restart the thread */ ResumeThread (OurhThread); /* It looks like a bug in Windows NT, but several waiting threads are not always released on SetEvent() without this sleep... */ Sleep(1); PulseEvent (signal_arrived); WSACancelBlockingCall(); return; /* This code is run in the standard thread space */ /* Commented out instructions are emulated already */ wrap: asm ("orl $0x0,(%esp)\n" /* probe out enough room to play */ "orl $0x0,-0x800(%esp)\n" "orl $0x0,-0x1000(%esp)\n" // "pushl _rethere\n" // "pushl _oldmask\n" // "pushl _sigarg\n" // "pushl _sigfunc\n"); "pusha\n" "pushf\n" "pushl %ebp\n" "movl %esp,%ebp\n" "pushl 0xb*4(%ebp)\n" // sigarg "movl 0xa*4(%ebp),%eax\n" // sigfunc "call *%eax\n" "pushl 0xc*4(%ebp)\n" // oldmask "call _set_process_mask\n" "leave\n" "popf\n" "popa\n" "popl _rethere\n" "popl _rethere\n" "popl _rethere\n" "ret"); The saved integer registers are accessible from signal handler. Save float context just near pusha/pushf and modify sigarg/sigfunc/oldmask offsets. > > > 3. There is no SIGVTALRM signal. > SML-NJ doesn't seem to need this in order to function. For now, I'm > providing a dummy #define. I don't know how to implement virtual timer on win32. > > > 4. "L_ctermid" is not defined (but "L_cuserid" is). > I don't remember why SML-NJ needs this. For now, I'm providing a dummy > #define. It's not critical. L_ctermid define should be added to stdio.h, and ctermid() call should be added to winsup sources for POSIX conformance. > > > 5. The following definitions are not in the Posix sockets include > > > files (only the Windows socket stuff) The missing defines already added to headers. -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Fuehrer@seabase.com Wed Aug 27 07:06:00 1997 From: Fuehrer@seabase.com (Gary Fuehrer) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: Bug with Fix: mmap(MAP_PRIVATE) region handling in fork() Message-ID: <6CC63E2E4FC1D011A2A700609716117A266F16@seawolf> > Download cdk-src.tar.gz from /pub/gnu-win32/gnu-win32-b18/ at > ftp.cygnus.com. After expanding it out, you will have a "cdk" > subdirectory. The "winsup" directory is in there. > > Attached are my copies of the files (except for "winsup.h" which was > trivial). My e-mail about the changes came out crappy because of all > the gratuitous line wrapping. So, maybe it would be easier to diff my > version with the original. > > > > > Gary > > > ---------- > From: Martin Schrape[SMTP:schrape@atomika.com] > Sent: Tuesday, August 26, 1997 1:56 PM > To: Gary Fuehrer > Subject: RE: Bug with Fix: mmap(MAP_PRIVATE) region handling in > fork() > > > > -----Original Message----- > From: Gary Fuehrer [SMTP:Fuehrer@seabase.com] > Sent: Friday, August 22, 1997 10:42 PM > To: 'gnu-win32@cygnus.com' > Subject: Bug with Fix: mmap(MAP_PRIVATE) region handling in > fork() > > > The following changes to three files (winsup.h, fork.cc, mmap.cc) will > fix this bug. > > ================ > winsup.h > ================ > > [schrape] Gary, > sorry but where did you get the winsup files from. I can't find them > on the cygnus ftp server. > --martin > -------------- next part -------------- A non-text attachment was scrubbed... Name: fork.cc Type: text/x-c Size: 15836 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: mmap.cc Type: text/x-c++ Size: 17516 bytes Desc: not available URL: From fjh@cs.mu.OZ.AU Wed Aug 27 07:06:00 1997 From: fjh@cs.mu.OZ.AU (Fergus Henderson) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: DLL problem References: <199708190947.LAA20111@hugo.int-evry.fr> Message-ID: <199708240910.TAA18108@murlibobo.cs.mu.OZ.AU> Stephane Mariel, you wrote: > I'm trying to port a unix network tool (at least the client) to Win95/NT , > since there is a DLL to create, I tested the example provided on > www.cygnus.com. > > It works fine, i got my foo.dll, main.exe, but moving to my real software > when I run ld for the first time I got > > c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libccrt0.o)(.t > ext+0x8a):libccrt0.cc: undefined reference to `GetModuleHandleA@4' > c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t > ext+0xc):libcmain.cc: undefined reference to `GetModuleHandleA@4' > c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t > ext+0x15):libcmain.cc: undefined reference to `GetCommandLineA@0' > c:/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib/libcygwin.a(libcmain.o)(.t > ext+0x1e):libcmain.cc: undefined reference to `WinMain@16' This occurs because there is no definition for `_impure_ptr' in your DLL. See < http://www.cs.mu.oz.au/~fjh/gnu-win32/how-to-build-dlls.html > for a brief explanation about `_impure_ptr'. -- Fergus Henderson | "I have always known that the pursuit WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ebritten@uci.edu Wed Aug 27 07:06:00 1997 From: ebritten@uci.edu (Eric Britten) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: Missing virtual functions Message-ID: <33FF524B.7B256D82@uci.edu> Yesterday I found that if one declares virtual functions in a class like: class A { A(); virtual void B(); }; and the definitions for the functions(inline or external) are not found by the linker then two types of linking errors result. 1. Undefined reference to a vtable in the constructors of the class. 2. Undefined reference to any other inline functions of the class that are used by other code. It took me a while to figure out that if all virtual functions are defined, then these errors go away. Is this GNU specific? It would help if the linker complained of undefined references to the virtual function without a definition, rather than the missing vtable. Of course this would help those unfamiliar with this error, I already know the cause. Eric Britten - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From weiku@utkux.utcc.utk.edu Wed Aug 27 07:06:00 1997 From: weiku@utkux.utcc.utk.edu (Wei Ku) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: VT-100 Message-ID: <199708241955.PAA29690@solid.ssd.ornl.gov> I was told to set TERM=linux and every thing works fine. Please notice that you need to copy or to link /gnuwin32/b18/H-i386-cygwin32/ to /etc so that the termcap can be found. Sincerely, Wei Ku -----Original Message----- From: Erwin Achermann To: gnu-win32@cygnus.com Date: Sunday, August 24, 1997 3:21 AM Subject: VT-100 >Howdy all, > >Finally i'am as far as it is really cute to work on my NT4sp3 emacs, >bash and less are all working hand in hand, handle CR LF correctly. >NICE JOB you all did! > >But still there is a little wich i have, My $TERM is set tu "dumb" >Is there a VT100 (or even better emulation arround somewhere? > >thanks > >Erwin >-- >|Erwin Achermann phone: +41-1-632 74 40 | >|Institute of Scientific Computing fax: +41-1-632 11 72 | >|ETH Zurich, IFW C29.2 mailto:acherman@inf.ethz.ch | >|8092 Zurich, Switzerland http://www.inf.ethz.ch/personal/acherman | >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From weiku@utkux.utcc.utk.edu Wed Aug 27 07:06:00 1997 From: weiku@utkux.utcc.utk.edu (Wei Ku) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: Include Path Error Message-ID: <199708241959.PAA37900@solid.ssd.ornl.gov> Try to put these two line in the file where the env. variables are set: CPLUS_INCLUDE_PATH=d:\Cygnus\include\g++ C_INCLUDE_PATH=d:\Cygnus\include -----Original Message----- From: Dan Rearick To: gnu-win32@cygnus.com Date: Sunday, August 24, 1997 6:55 AM Subject: Include Path Error >Problem: I am attempting to compile my first GNUWin32 program, the infamous HELLO.C, to verify the installation of GNUWin32 under Windows 95 and I am receiving the following error: > >Entered: > gcc hello.c >Results: > hello.c:1: No include path in which to find stdio.h > GCC.EXE: Internal compiler error: program cpp got fatal signal 33 > >Below is a listing of my environment parameters obtained with the Win95/DOS SET command: > >/* Begin parameters list */ > >TMP=C:\WINDOWS\TEMP >TEMP=C:\WINDOWS\TEMP >PROMPT=$p$g >winbootdir=C:\WINDOWS >COMSPEC=C:\WINDOWS\COMMAND.COM >NESTORPATH=C:\OCRFF\NESTOR30 >GCC_EXEC_PREFIX = C:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\ >TCL_LIBRARY = C:/gnuwin32/b18/tcl/lib/tcl7.6 >GDBTK_LIBRARY = C:/gnuwin32/b18/share/gdbtcl >PATH=C:\WINDOWS;C:\WINDOWS\COMMAND;C:\MSSQL\BIN;C:\UTL;C:\PERL\BIN;C:\OPTIA LL;C: >\KIPP\BIN;C:\PROGRAMF\MTI\OCRFF_16;C:\OCRFF\OCR\WORDSCAN.30;C:\OCRFF\KIPP\B IN;C: >\GNUWIN32\B18\H-I386-CYGWIN32\BIN;C:\GNUWIN32\B18\TCL\BIN;C:\GNUWIN32\B18\H -I386 >-CYGWIN32\LIB\GCC-LIB\I386-CYGWIN32\CYGNUS-2.7.2-970404;C:\GNUWIN32\B18\H-I 386-C >YGWIN32\I386-CYGWIN32\INCLUDE >windir=C:\WINDOWS >SNDSCAPE=C:\WINDOWS >BLASTER=A220 I5 D1 T1 >CMDLINE=gcc hello.c > >/* End parameters list */ > >Would someone please enlighten me as to what is missing. The software was installed from a full version of GNUWin32 (CDK.EXE, size 14MB) and the installation completed error free. Thanks. > >DBR... > >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From urcr@rz.uni-karlsruhe.de Wed Aug 27 07:06:00 1997 From: urcr@rz.uni-karlsruhe.de (urcr@rz.uni-karlsruhe.de) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: dll questions Message-ID: <199708271406.HAA17893@cygnus.com> hi. i managed to build a relocable dll with gnuwin32. a) is there a dll entry handler somewhere, that handles everything needed ? maybe my handler is a bit short ("{}"). b) how can i use the gnuwin32 craeted dll with borland or microsoft c compilers ? i don't think that they can work with libDLL.a like gcc, but i don't know. c) cygwin32.dll : gpl or lgpl ? i didn't follow the copyright discussions. thanks for your help. andreas - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fine@mail.cern.ch Wed Aug 27 07:06:00 1997 From: fine@mail.cern.ch (Valery Fine) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: Unix spooler References: <420CC47239E@fs-nw-ps.oce.nl> Message-ID: <199708230857.KAA112928@sp058.cern.ch> On 19 Aug 97 at 11:16, Rob Hegt wrote: > This may all be true, but I guess that what was initially intended > was an lpr/lpd suite for win32. So, what is the status on that? It is supplied with Windows. Goto "Control Panel" -> Networking" and setup Microsoft TCP/IP printing staff. If this is not enough one may support this from UNIX side via SAMBA daemon as well. Hope this helps, Valery Dr. Valeri Faine (Valery Fine) ------------ ------------- Phone: +41 22 767 4921 CERN FAX : +41 22 767 7155 CH-1211 Geneva, 23 mailto:fine@mail.cern.ch Switzerland http://nicewww.cern.ch/~fine - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bilbo@ameritech.net Wed Aug 27 07:06:00 1997 From: bilbo@ameritech.net (Dan Maynard) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: winnt/95 authentication References: <3.0.3.32.19970821192623.0069e5d8@pop.mindspring.com> Message-ID: > >Is there any way of authenticating a password against windows nt server > >domain security in gnu-win32? snippet from Sergey's patch to login.c: ============================ HANDLE hToken; char *username; char *pp; <=- password if (LogonUser(username, NULL, pp, LOGON32_LOGON_INTERACTIVE, LOGON32_PROVIDER_DEFAULT, &hToken)) ============================ returns: nonzero success zero failed - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mryan@aet.net Wed Aug 27 07:06:00 1997 From: mryan@aet.net (Mark Ryan) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: gnu compiler Message-ID: <199708270217.WAA14826@candy.micro-net.com> Cygnus: >>Thank you in advance for your help. >> >>I am running a DX4100 16MB RAM 850MB hard drive, Windows 95. >> >>I have installed the GNU compiler, following the directions in the FAQ, >>i.e., unzipping all.tgz file, setting gcc_exec_prefix environment variable, >>and adding \cygnus\h-i386-c\bin to the path statement. >> >>The faq indicates I should set the environment variable to >>c:\cygnus\h-i386-cygwin32\bin\, however it needs to be set to >>c:\cygnus\h-i386-c\bin as there is an 8 character limit to folder names. >> >>Is this a problem as long as the environment variable, path and directories >>all match?? >> >>I am getting errors that "no include path is found for iostream.h..." >> >>The errors change depending on whether or not I use CAPS or change the path >>to include more of the cygnus directories. >> >>I have talked with Macmillan technical support on a number of occasions. >>Can someone help with this problem?? All I want to do is learn C++, I >>don't want to get hung up installing a compiler. >> >>Please help. >>I can be reached any of the following ways: >>email: mryan@aet.net (home) >> mryan@tsgpro.com (work) >>Phone: 1-704-482-8673 (home 6:30p - 10:00p) >> 1-800-249-7269 (work 8:00a - 5:00p) >> >>Thanks again. >>Mark Ryan - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From startnet@nation.org Wed Aug 27 07:06:00 1997 From: startnet@nation.org (Derek Reynolds) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: Cygwin Archive and my Ports Message-ID: <1.5.4.32.19970823172624.0066dca4@nation.org> Hi there. I have finnaly posted all of the programs i have ported via cygwin. I have also made an archive of other *nix ports to win32 environment. Please check it out: http://www.h0ar.org/myn If there is anything i am missing please let me know. If you have any ported programs that you think should be on here, please email me startnet@nation.org with the location to where it can be downloaded (its an archive). Thanks Derek Reynolds - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fjh@cs.mu.OZ.AU Wed Aug 27 07:06:00 1997 From: fjh@cs.mu.OZ.AU (Fergus Henderson) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: how to build a dll that uses other dll's? References: <33FBCC2D.98CBE385@brown.edu> Message-ID: <199708240913.TAA18566@murlibobo.cs.mu.OZ.AU> Dan Goldwater, you wrote: > I am trying to build a DLL which has function calls which must be > resolved into existing win32 DLL's. > > Following the instructions for building a relocatable dll, i get an > output but each of the three ld steps reports all the undefined > references for the functions which are in the existing win32 dll's. i > had to add the --noinhibit-exec option to ld so that it wouldn't delete > the output dll due to the errors. You need to build import libraries for the DLLs you want to link your DLL with, and then you need to link against those import libraries when creating your DLL. > when i build this project on solaris, the linker doesn't complain about > the undefined references, and they are resolved to appropriate existing > .so files at runtime. Yeah, ELF is much nicer than gnu-win32 DLLs... -- Fergus Henderson | "I have always known that the pursuit WWW: < http://www.cs.mu.oz.au/~fjh > | of excellence is a lethal habit" PGP: finger fjh@128.250.37.3 | -- the last words of T. S. Garp. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bartlee@hotmail.com Wed Aug 27 07:06:00 1997 From: bartlee@hotmail.com (Bartlee Anderson) Date: Wed, 27 Aug 1997 07:06:00 -0000 Subject: df.exe gives me an error message... Message-ID: <19970824073056.3679.qmail@hotmail.com> Another reply to something like this was that the ExA is the NT version. You need to rebuild the executable so that it is linked to the ..A (sans Ex) as I seem to recall. And to steal somebodies one liner... If memory serves me right... it would be the first time. ;) >From gnu-win32-owner@cygnus.com Sat Aug 23 20:58:07 1997 >Received: from cygnus.com (cygnus.com [205.180.230.20]) > by yooper.switch.rockwell.com (8.8.5/8.8.4) with ESMTP > id WAA03641 for ; Sat, 23 Aug 1997 22:57:52 -0500 (CDT) >Received: (from daemon@localhost) > by cygnus.com (8.8.5/8.8.5) id MAA05021 > for gnu-win32-outgoing; Mon, 18 Aug 1997 12:08:23 -0700 (PDT) >Received: from prometheus.hol.gr (root@prometheus.hol.gr [194.30.220.120]) > by cygnus.com (8.8.5/8.8.5) with ESMTP id MAA05013 > for ; Mon, 18 Aug 1997 12:08:11 -0700 (PDT) >Received: from panos ([194.30.193.9]) > by prometheus.hol.gr (8.8.5/11.4.0) with SMTP id WAA14581; > Mon, 18 Aug 1997 22:05:04 -0200 (GMT) >Posted-Date: Mon, 18 Aug 1997 22:05:04 -0200 (GMT) >Received: by panos with Microsoft Mail > id < 01BCAC23.D7422380@panos >; Mon, 18 Aug 1997 22:12:39 +-300 >Message-ID: < 01BCAC23.D7422380@panos > >From: Panos Platon Tsapralis >To: "'Herry'" >Cc: "'Cygnus GNU Programming Environment'" >Subject: df.exe gives me an error message... >Date: Mon, 18 Aug 1997 22:12:10 +-300 >MIME-Version: 1.0 >Content-Type: text/plain; charset="us-ascii" >Content-Transfer-Encoding: 8bit >X-MIME-Autoconverted: from quoted-printable to 8bit by cygnus.com id MAA05015 >Sender: owner-gnu-win32@cygnus.com >Precedence: bulk > > I tried to use the df.exe file that you sent to me, attached to your last message, and it gives me the following message in the command line of BASH : > >bash.exe: /GNUWIN32/B18/H-I386-CYGWIN32/BIN/df.exe: (null) > >while, at the same time a dialog box appears on-screen, with a caption "Error Starting Program", a "OK" button and a message that says : > >"The DF.EXE file is linked to missing export KERNEL32.DLL:GetDiskFreeSpaceExA. > >What am I missing ? > >+----------------------------------------------+ >| Panos Platon Tsapralis,Software Engineer, | >| M.I.S.Applications Manager - TeleStet, | >+----------------------------------------------+ >| E-mail Address = pan@hol.gr, | >| Home Address = 3 Aishylou Str., 15234, | >| Halandri, Athens, | >| Attica, GREECE, | >| Home Phone No. = (00301)68.58.298, | >| Bysiness Address = 60 Kifissias Ave., 15125, | >| Amaroussion, Athens, | >| Attica, GREECE, | >| Office Phone No. = (00301)61.86.013, | >| Mobile Phone No. = (003093)23.90.13. | >+----------------------------------------------+ >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ckirgios@ajboggs.com Wed Aug 27 08:46:00 1997 From: ckirgios@ajboggs.com (Chris Kirgios) Date: Wed, 27 Aug 1997 08:46:00 -0000 Subject: Linking with ORACLE Pro*C/C++ Libraries on NT Message-ID: <01BCB2DE.CA94E250@stones.ajboggs.com> I have a g++ compiled library that contains embeded SQL. It was precompiled with Oracle's Pro*C/C++ compiler on Windows NT. In order to link that library with my executable I need to link with Oracle's sqllib80.lib (or sqllib80.dll, which is NOT a .a library). When I run gdb I get a bunch of DLL relocation errors. Question: Does anyone know how to 'transform' the Oracle DLL (by the way, Oracle provides copies of it only for Vicual C++ and Borland C++ compilers) into a gnu-win32 library? Has anyone else come across the same or similar problem? I suspect that this might be a more general problem for situations where source code is not available but DLLs (created with non-cygwin32 compilers) are. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From keresztg@podolin.piar.hu Wed Aug 27 10:53:00 1997 From: keresztg@podolin.piar.hu (Keresztfalvi Gábor 97b) Date: Wed, 27 Aug 1997 10:53:00 -0000 Subject: Problem with relocatable DLL in b18 Message-ID: Hi, I'm new in GnuWin32 programming and I want to make a relocatable DLL according to the example residing at www.cygnus.com. I compiled it under Beta17 and it worked fine, but if I compile it under Beta18, it doesn't work. The 'start main.exe' command sais: "The program is in an invalid format, and cannot be run. It may be damaged." And I got an "Error Starting Program" dialog box with a message: "Unable to run f:\tmp\main.exe" and an "OK" button. What is wrong? Me or Beta18? Please help me! Thanks in advance, Keresztg P.s.: Here is the shell script: #! /bin/sh # Example Script to compile and link a relocatable DLL # Files that make up the DLL = foo.c foo2.c init.cc fixup.c. # (init.cc and fixup.c are housekeeping routines needed for the DLL. The actual # library routines are in foo.c and foo2.c) # ***Fill in your path to libcygwin.a here (with no trailing slash)*** LIBPATH=/f/gnuwin32/b18/H-i386-cygwin32/i386-cygwin32/lib # Compile source files: gcc -c foo.c gcc -c foo2.c gcc -c init.cc gcc -c fixup.c # Make .def file: echo EXPORTS > fooB.def nm foo.o foo2.o init.o fixup.o | grep '^........ [T] _' | sed 's/[^_]*_//' >> fooB.def # Link DLL. ld --base-file fooB.base --dll -o fooB.dll foo.o foo2.o init.o fixup.o \ $LIBPATH/libcygwin.a -e _dll_entry@12 dlltool --as=as --dllname fooB.dll --def fooB.def --base-file fooB.base --output-exp fooB.exp ld --base-file fooB.base fooB.exp --dll -o fooB.dll foo.o foo2.o init.o fixup.o \ $LIBPATH/libcygwin.a -e _dll_entry@12 dlltool --as=as --dllname fooB.dll --def fooB.def --base-file fooB.base --output-exp fooB.exp ld fooB.exp --dll -o fooB.dll foo.o foo2.o init.o fixup.o \ $LIBPATH/libcygwin.a -e _dll_entry@12 # Build the fooB.a lib to link to: dlltool --as=as --dllname fooB.dll --def fooB.def --output-lib fooB.a # Linking with main gcc main.c fooB.a -o main.exe +-----------------------------------------------------------------------------+ | Keresztfalvi Gabor Budapesti Piarista Gimnazium | | Web Page: http://www.piar.hu/~keresztg/ | | E-Mail: keresztg@podolin.piar.hu :-) | +---------- This message was transmitted on 100% recycled electrons ----------+ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mphalpin@iag.net Wed Aug 27 14:16:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Wed, 27 Aug 1997 14:16:00 -0000 Subject: Vim? Message-ID: <199708272106.RAA19188@iag.net> Where can I get vim? Thanks, - SJ - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sshteingold@cctrading.com Wed Aug 27 14:16:00 1997 From: sshteingold@cctrading.com (sshteingold@cctrading.com) Date: Wed, 27 Aug 1997 14:16:00 -0000 Subject: RCS for WinNT Message-ID: <9708278727.AA872716059@inet.stknhlg.com> I looked at this mailing list's archive and there were only 3 messages about RCS - 2 about compiling it, and one requesting binaries. None answered, so I repeat the question: Where can I get the working RCS binaries for winnt? I have the latest cygwin32 (rcs is not included), but neither who nor whoami work: C:\home\sds>who am i who: /var/run/utmp: No such file or directory C:\home\sds>whoami whoami: cannot find username for UID 500 (I have an environment var USER set, but apparently it's not enough). Also, I use binary (aka UNIX) file format (no CRLF, just LF). Will RCS work with it? (someone claimed that RCS 5.6 converts everything to text - aka DOS - file format) If possible, please send me a mail copy of your kind reply. Thanks. Sam - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Fuehrer@seabase.com Wed Aug 27 14:16:00 1997 From: Fuehrer@seabase.com (Gary Fuehrer) Date: Wed, 27 Aug 1997 14:16:00 -0000 Subject: Bug with Fix: mmap(MAP_PRIVATE) region handling in fork() Message-ID: <6CC63E2E4FC1D011A2A700609716117A266F18@seawolf> I can use WinZip 6.2 and gzip successfully on the non-split cdk-scr.tar.gz. I'd bet the split one works fine too. Once you get a successful download of cdk-src.tar.gz, I suggest you not use WinZip to expand it. WinZip does at least two annoying things without telling the user: it converts LF to CRLF when it thinks the file is text and it ignores symbolic links. It sounds like I can now add a third annoying thing to my list: WinZip doesn't tell the user when it has quit early with a CRC error. Gary > ---------- > From: Martin Schrape[SMTP:schrape@atomika.com] > Sent: Wednesday, August 27, 1997 11:13 AM > To: Gary Fuehrer > Subject: RE:RE: Bug with Fix: mmap(MAP_PRIVATE) region handling > in fork() > > From: Gary Fuehrer [SMTP:Fuehrer@seabase.com] > Sent: Wednesday, August 27, 1997 12:44 AM > To: 'Martin Schrape' > Cc: 'gnu-win32@cygnus.com' > Subject: RE: Bug with Fix: mmap(MAP_PRIVATE) region handling in > fork() > > > Download cdk-src.tar.gz from /pub/gnu-win32/gnu-win32-b18/ at > > ftp.cygnus.com. After expanding it out, you will have a "cdk" > > subdirectory. The "winsup" directory is in there. > > Did you download the splitted version? > Having used Winzip with the cdk-src.tar.gz 26MB archive will not > include the winsup directory and gzip gives an crc error. Seems to me > that Winzip just ignores the error. > > --martin > > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From DGREGORY@email.usps.gov Wed Aug 27 14:23:00 1997 From: DGREGORY@email.usps.gov (DGREGORY@email.usps.gov) Date: Wed, 27 Aug 1997 14:23:00 -0000 Subject: C++ link problem on Windows 95 - 17.1 & 18 Message-ID: <0033000004089002000004*@MHS> I am having problems compiling C++ programs with versions 17.1 and 18 on Windows 95. I have tried this on two systems with beta 18 the same results, and on one system I tried it with both 18 and 17.1. I am trying to use beta 18 to learn C++, so please excuse any terminology problems. Is anyone using beta 18 to compile C++ on Windows 95? The results: C:\>gcc hello.cc C:\tmp/cc0010001.o(.text+0x1e):hello.cc: undefined reference to `cout' C:\tmp/cc0010001.o(.text+0x23):hello.cc: undefined reference to `ostream::operator<<(char const *)' I suspect that this is a linking issue, because if I execute: gcc -c hello.cc I get no error messages, and an object file is created (hello.o). The program: #include void main () { cout << "Hello world!\n"; } Additional information... I have tried playing with all of the environment settings including LIBRARY_PATH and nothing seems to work. I am able to compile a C hello.c program using printf without any problems. When I try to copy libg++.a to play around, I need to put the file name in quotes, or the copy fails, so I think there is a possibilitty that this may be a contributor to the problem. I have tried linking the hello.o module with ld, but I can not seem to successfully use the -l parameter to specify libg++.a Any help will be greatly appreciated. ... Darwin PS - I have checked the FAQ and the mailing list archive which showed a few others having this problem, but no solution posted. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jsmat@roman.net Wed Aug 27 14:39:00 1997 From: jsmat@roman.net (Jeff Smathers) Date: Wed, 27 Aug 1997 14:39:00 -0000 Subject: Installation Problems with GNU-WIN32 Message-ID: <199708272117.RAA03402@web2.america.net> ????I am new to GNU (sorry about that). I am planning to take Jesse Liberty's class on C++ at ZD University. I run Win95 on a P-166 Intel box with 24Meg RAM. I am also using MS IE4.0.???? I have followed the install directions to the best of my ability including set the path setting and the environment variables.???? I have installed version 18 (or is it 1.8?).???? I get this message: "GCC.EXE: installation problem, cannot exec 'cpp'. No such file or directory. I know that that is supposed to mean I have not set the GCC_EXEC_PREFIX variable correctly. However, when I run SET I can see this (among all the other variables) "GCC_EXEC_PREFIX=D:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\. So, I am stumped. I have searched various places for help but cannot get beyond the basic directions.???? Can someone tell me where to look (or exactly what is wrong)? TIA.....Jeff Smathers ???? ???? From Ian@kiwiplan.co.nz Wed Aug 27 17:28:00 1997 From: Ian@kiwiplan.co.nz (Ian Collins) Date: Wed, 27 Aug 1997 17:28:00 -0000 Subject: fcntl call on win95 Message-ID: <07551A47EAD6D01186430060B025C6F40A6C99@ntserver.kiwiplan2.co.nz> I have a routine that write locks a file. The essential code is, #include auto struct flock lck; auto int flc; auto int filedesc; filedesc = open(filename, O_RDWR|O_CREAT, 0666); lck.l_type = F_WRLCK; lck.l_start = 0; lck.l_whence = 0; lck.l_len = 0; flc = fcntl(fildesc, F_SETLK, &lck); The fcntl provides a write lock on Windows NT fine (on both FAT and NTFS) filesystems, but fails (flc == -1) on Windows 95. Can anyone help me on this? Many Regards, Ian Collins - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From colin@bird.fu.is.saga-u.ac.jp Wed Aug 27 19:57:00 1997 From: colin@bird.fu.is.saga-u.ac.jp (Colin Peters) Date: Wed, 27 Aug 1997 19:57:00 -0000 Subject: Making cygnus-win32 libraries from DLLs Message-ID: <01BCB3A8.5EED0300@gbird0> Chris Kirgios[SMTP:ckirgios@ajboggs.com] wrote: >I have a similar problem with Minh Le Cong to whom you replied in the cygnus mailing list. I have ported all my C++ code from Solaris to Windows NT with the exception of a library that uses Oracle Pro*C/C++ (embedded SQL). Modules that link with this library must link with an Oracle dll. Question: How do I link my g++ linked code with that library/dll? > >I am an NT 'ignorant' so I appreciate all the help and input I can get. I have heard a few people ask this question, so I dug around a bit in my files and found out that the program best suited for this is not really dumpexts (that works with object files) but one posted by Ismael Jurado way back in April called impdef. I've included the source code and will put up the source and exe on my web site shortly. This program compiles and works on my system with no modifications (since it includes the missing info from the winnt.h header than dumpexts does not). Anyway, here's how it works. Assuming you have a .dll and the appropriate .h files (you don't actually need the .lib at all) run impdef on the .dll: impdef foo.dll >foo.def Then run foo.def through dlltool to get a libfoo.a file: dlltool --dllname foo.dll --def foo.def --output-lib libfoo.a -k NOTE: Don't let the --dllname argument fool you. Dlltool does NOT look for any file called foo.dll and definitely does NOT extract the exports and create the right library for them. You need impdef. The dllname is just the name that is put into the library to mark what file should be loaded to resolve the imports (don't put a path name in here, just the basic file name). Compile your code including the .h files and linking with the libfoo.a library. Your code, upon execution, will search for foo.dll and resolve the references to the functions within it (if all goes well). That's it, good luck. Colin. PS. Apologies to earlier inquirers, including Minh Le Cong, who were told that dumpexts didn't do the job I thought it would. It doesn't, but impdef does. -- Colin Peters - Saga Univ. Dept. of Information Science -- colin@bird.fu.is.saga-u.ac.jp - finger for PGP public key -- http://www.fu.is.saga-u.ac.jp/~colin/index.html -- http://www.geocities.com/Tokyo/Towers/6162/ -------------- next part -------------- A non-text attachment was scrubbed... Name: impdef.cpp Type: text/x-c Size: 10895 bytes Desc: not available URL: From Ian@kiwiplan.co.nz Wed Aug 27 23:18:00 1997 From: Ian@kiwiplan.co.nz (Ian Collins) Date: Wed, 27 Aug 1997 23:18:00 -0000 Subject: Signal Handling Message-ID: <07551A47EAD6D01186430060B025C6F40A6C9B@modem1.kiwiplan2.co.nz> My signal handling is not functioning as expected under gnu-win32. I notice that if in a read, when a signal (e.g, a SIGALRM) triggers, the signal handler is run, and then control is passed back to the read, which just carries on. It is expected that after a signal has been triggored and the handler run, then the read should return. I have tried this using both, struct sigaction vec; vec.sa_handler = myroutinename; if (sigemptyset(*vec.sa_mask) == -1) return -1; vec.sa_flags = SIG_INTERRUPT; if (sigaction(SIGALRM, &vec, NULL) == -1) return -1; alarm(10); and, if (signal(SIGALRM, myroutinename) == SIG_ERR) return -1; alarm(10); Many Regards, Ian Collins - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Ian@kiwiplan.co.nz Wed Aug 27 23:19:00 1997 From: Ian@kiwiplan.co.nz (Ian Collins) Date: Wed, 27 Aug 1997 23:19:00 -0000 Subject: FW: fcntl call on win95 Message-ID: <07551A47EAD6D01186430060B025C6F40A6C9C@modem1.kiwiplan2.co.nz> Many Regards, Ian Collins > -----Original Message----- > From: Ian Collins > Sent: Thursday, August 28, 1997 12:26 PM > To: 'Gnu win32 mailer' > Subject: fcntl call on win95 > > I have a routine that write locks a file. The essential code is, > > #include > auto struct flock lck; > auto int flc; > auto int filedesc; > > filedesc = open(filename, O_RDWR|O_CREAT, 0666); > lck.l_type = F_WRLCK; > lck.l_start = 0; > lck.l_whence = 0; > lck.l_len = 0; > flc = fcntl(fildesc, F_SETLK, &lck); > > The fcntl provides a write lock on Windows NT fine (on both FAT and > NTFS) filesystems, but fails (flc == -1) on Windows 95. > Can anyone help me on this? > > Many Regards, > Ian Collins > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From hubert.fauque@wanadoo.fr Thu Aug 28 07:18:00 1997 From: hubert.fauque@wanadoo.fr (Hubert FAUQUE) Date: Thu, 28 Aug 1997 07:18:00 -0000 Subject: problem with mounting / as binary or not Message-ID: <3405e496.230818880@smtp.wanadoo.fr> I have installed Sergey's cygwin.dll and bash and I am having a problem: bash didn't find .bashrc at startup, so as I have seen on a previous message I have mounted c: as / with text=binary and it works for bash; but now is the problem: make doesn't find any include files; if there is include file in the Makefile, it gives the message file^M: no such file or directory it doesn't find the file because it adds a ^M at the end of the name; Has anybody found a solution? thanks Hubert - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Gilles.Bourque@nrc.ca Thu Aug 28 07:18:00 1997 From: Gilles.Bourque@nrc.ca (Bourque, Gilles) Date: Thu, 28 Aug 1997 07:18:00 -0000 Subject: GLUT32 + gnuwin = Examples? Message-ID: Hi, Does anybody has working examples with makefile that use GLUT32 and/or opengl. Thanks in advance, Gilles Bourque Gilles.Bourque@nrc.ca - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From schrape@atomika.com Thu Aug 28 07:18:00 1997 From: schrape@atomika.com (Martin Schrape) Date: Thu, 28 Aug 1997 07:18:00 -0000 Subject: cdk-src.tar.gz installation failed Message-ID: <01BCB3B4.09942440@MS> I have a problem unzipping the b18 cdk-src.tar.gz archive. I have already downloaded it twice. The first was 27,561,754 bytes the 2nd 27,666,697. The first one can be unzipped with Winzip 6.0 and 6.2 but I get only about 26Mbyte. A lot of files are missing like the winsup directory. gzip failed on this archive with an crc error. On the 2nd archive Winzip and gzip failed (invalid compressed data--format violated). Before starting another download try and paying another $3/MB to my provider I would like to know if someone knows this problem and a workaround. --martin --------------------------------------------------------------- Martin Schrape schrape@atomika.com Senior Software Engineer Tel: +49 89 315 891-34 Atomika Instruments GmbH Fax: +49 89 315 59 21 Bruckmannring 40 85764 Oberschleissheim/Munich, Germany --------------------------------------------------------------- - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bond@cs.chalmers.se Thu Aug 28 08:30:00 1997 From: bond@cs.chalmers.se (Magnus Bondesson) Date: Thu, 28 Aug 1997 08:30:00 -0000 Subject: DirectX Message-ID: <199708281427.QAA18702@muppet67.cs.chalmers.se> I'm trying to use parts of DirectX using the material supplied at http://dragon.acadiau.ca/~025668b/develop/cygnus/ However I miss a macro MAKE_HRESULT and perhaps two constants E_FAIL and CLR_INVALID. They are not supplied in the DirectX SDK. I suspect that they are defined in some other Microsoft product. MAKE_HRESULT is needed to interpret the error code given when I try to create a surface. I would be thankful for any help. Magnus Bondesson Dep of Computing Science Chalmers University of Technology S-412 96 Gothenburg SWEDEN bond@cs.chalmers.se - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Xavier.Fornari@sophia.inria.fr Thu Aug 28 09:55:00 1997 From: Xavier.Fornari@sophia.inria.fr (Xavier.Fornari@sophia.inria.fr) Date: Thu, 28 Aug 1997 09:55:00 -0000 Subject: Cannot generate relocatable object Message-ID: <199708281502.RAA17182@filtre.inria.fr> I generated an object file using the -r option of the linker, but when I link this object to have the final executable I got: "bad reloc address 0xb in section .text" My aim is to create a library .a. This has already be done for Sun4OS, Solaris, DecAlpha, Linux, Aix, HP, and I try to port our software on the NT platform. The global scheme is: ld -r -o common.o common1.o common2.o ... ar ru libcommon.a common.o extra1.o extra2.o And then: gcc -o myapp app.c libcommon.a Direct e-mail answers prefered since I only subscribed to gnu-win32-announce-request. Thanks. -- Xavier Fornari ================================================== http://www.inria.fr/meije/personnel/Xavier.Fornari CMA/MEIJE Tel: 33 (0)4 93 65 79 65 INRIA Fax: 33 (0)4 93 65 79 98 2004 Route des Lucioles BP 93 06902 Sophia-Antipolis Cedex - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Gilles.Bourque@nrc.ca Thu Aug 28 10:47:00 1997 From: Gilles.Bourque@nrc.ca (Bourque, Gilles) Date: Thu, 28 Aug 1997 10:47:00 -0000 Subject: FW: GLUT32 + gnuwin = Examples? Message-ID: ---------- From: Bourque, Gilles To: Cygnus ML Subject: GLUT32 + gnuwin = Examples? Date: Thursday, 28 August, 1997 10:00AM Hi, Does anybody has working examples with makefile that use GLUT32 and/or opengl. Thanks in advance, Gilles Bourque Gilles.Bourque@nrc.ca - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sof@dcs.gla.ac.uk Fri Aug 29 03:25:00 1997 From: sof@dcs.gla.ac.uk (Sigbjorn Finne) Date: Fri, 29 Aug 1997 03:25:00 -0000 Subject: SML-NJ for CygWin32 References: <6CC63E2E4FC1D011A2A700609716117A266F0A@seawolf> Message-ID: <19970828191744.FMU39440@savage> > > > > > > This is clearly incorrect behavior on the part of CygWin's gcc. > > Dollar > > > signs are not valid leading characters of C (or C++) identifiers. > > > > Until this is fixed in the preprocessor, the work-around for building > SML-NJ is to add a space separator between the "$" and the constant. I > will NOT submit this change to the SML-NJ team. They would likely > reject it anyhow (I would). Hi, AFAIK, it is ANSI that forbids $ in ids. I'd expect this restriction to be enabled via -ansi (as per gcc docs.) For whatever reason, the gcc2 snap included in b18 doesn't pass -$ to cpp with -ansi.. foo$ gcc -v Reading specs from c:\usr\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32\cygnus-2.7.2-970404\specs gcc version cygnus-2.7.2-970404 foo$ echo 'subl $FOO, %eax' | gcc -E -Wp,-P -DFOO=4 - subl $FOO, %eax foo$ echo 'subl $FOO, %eax' | gcc -E -Wp,-P -DFOO=4 -ansi - subl $FOO, %eax Suggested workaround: use -Wp,-$ foo$ echo 'subl $FOO, %eax' | gcc -E -Wp,-P -DFOO=4 -Wp,-$ - subl $4 , %eax HTH, --Sigbjorn - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ROHITVERMA@Inf.COM Fri Aug 29 03:25:00 1997 From: ROHITVERMA@Inf.COM (ROHITVERMA) Date: Fri, 29 Aug 1997 03:25:00 -0000 Subject: No Subject Message-ID: <001091EC.3391@smtp_gw.inf.com> Hi, I am a novice in the area of system programming who has been asked to replicate the functionality of fork on NT within a week. I have understood the steps taken during the fork in UNIX.However,on going throuh fork.cc and other files,I get confused in the actual flow of the process. Can you please help me by listing a)the files which contain the complete implementation of fork b)the flow of the process(a description which is more detailed than that provided in the FAQ is needed.) Thanks & regards, Rohit Verma - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From bischoff@ise.com Fri Aug 29 04:08:00 1997 From: bischoff@ise.com (Andreas Bischoff) Date: Fri, 29 Aug 1997 04:08:00 -0000 Subject: gcc optimization for Pentium Pro ? Message-ID: <199708290602.XAA17789@ise20.ise.com> Hi all, is there a compiler option to optimize the code for a Pentium Pro processor ? Regards, Andreas - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From etlmllg@etlxdmx.ericsson.se Fri Aug 29 04:08:00 1997 From: etlmllg@etlxdmx.ericsson.se (Mike Legg) Date: Fri, 29 Aug 1997 04:08:00 -0000 Subject: Templates and Linking Message-ID: <3406AD50.BA7598A4@etlxdmx.ericsson.se> Hi, Does anyone know how to get around this error which is reported during linking when using the STL: Table.o(.data+0xc):Table.cc: multiple definition of `__default_alloc_template::free_list' Main.o(.data+0x0):Main.cc: first defined here The exact same code compiles fine on gcc2.7.2 with djgpp. Is there any reason why template intansiation is different? Attached is a trivial example. regards, Mike. -------------- next part -------------- A non-text attachment was scrubbed... Name: Main.cc Type: text/x-c Size: 187 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Table.cc Type: text/x-c Size: 365 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Table.h Type: text/x-c++ Size: 306 bytes Desc: not available URL: From Don.Sharp@dddandr.octacon.co.uk Fri Aug 29 07:10:00 1997 From: Don.Sharp@dddandr.octacon.co.uk (Don Sharp) Date: Fri, 29 Aug 1997 07:10:00 -0000 Subject: Non-receipt of traffic Message-ID: <3405EF5C.3CF9@dddandr.octacon.co.uk> Hi There, I haven't received anything from this list since the early hours of Monday morning. Is there anybody out there? I have had regular traffic on other mail groups so I know my mail is OK. Hoping to receive some GNu news soon! Don Sharp - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Gilles.Bourque@nrc.ca Fri Aug 29 07:10:00 1997 From: Gilles.Bourque@nrc.ca (Bourque, Gilles) Date: Fri, 29 Aug 1997 07:10:00 -0000 Subject: Makefile creation tutorial? Where? Message-ID: Hi, The title says it all. Gilles Bourque Gilles.Bourque@nrc.ca - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Keetnet@wilmington.net Sat Aug 30 17:49:00 1997 From: Keetnet@wilmington.net (Keet / Foxbird) Date: Sat, 30 Aug 1997 17:49:00 -0000 Subject: REPEAT: Date / Time Stuff.. Message-ID: <3.0.32.19970830204746.0079c770@wilmington.net> Had to repeat the send of this message.. evidentally it never got through (from what I can tell) -------- I tried compiling some 'out-of-the-box' examples from a learn C++ guide and I get a really nice error that looks like this: C:\WINDOWS\Desktop\fb5.56\test\c++tutor\source>c++ -g -O -o date date.cpp C:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32\cygnus-2.7.2-97040 4\.. /../../../i386-cygwin32\lib/libcygwin.a(libcmain.o)(.text+0x1e):libcmain.cc: und efined reference to `WinMain@16' C__~1.EXE: Internal compiler error: program ld got fatal signal 1 Am I missing a library, an environment setting or something along those lines? I've compiled other date stuff before, but this is the first one that doesn't work. If you need the source, just ask, I've got it available. - Greg Neujahr - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From kmahan@novell.com Sat Aug 30 17:49:00 1997 From: kmahan@novell.com (Kurt Mahan) Date: Sat, 30 Aug 1997 17:49:00 -0000 Subject: Looking for java implementation of X Windows References: <19970824073906.12588.qmail@hotmail.com> Message-ID: <3408b47c0.10bc@bolt.usg.provo.novell.com> > Has anyone seen a Java only implementation of X. > I saw something that I thought I saved, but can't find. > > May have been on another list. Attached is what you might be referring to (from the jolt list).. Kurt /* * This version of Kurt Mahan is currently being evaluated. Words he speaks * are those of him only and not those of Novell or anybody else. * * Novell Java Technologies R&D Group * * Kurt Mahan * kmahan@novell.com */ ------------------------------ ATTACH --------------------------- >From jolt-list-request@redhat.com Wed Aug 13 14:05 MDT 1997 Return-path: From: "Eugene O'Neil" To: Subject: XTC, an X Window library implemented in pure Java Date: Wed, 13 Aug 1997 15:39:23 -0400 X-URL: http://www.redhat.com I am writing a Java windowing toolkit named XTC, the X Tool Collection. Instead of using native methods that depend on Motif or even xlib, I have instead opted to implement the underlying X Window Protocol in pure Java. My library is small, powerful, object-oriented, multi-threaded, reasonably efficient, and public domain. If you find this interesting, download my working demo from the following web page: http://www.cs.umb.edu/~eugene/X.html -Eugene -- To unsubscribe: mail -s unsubscribe jolt-list-request@redhat.com < /dev/null - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pbrant@pcbsys.demon.co.uk Sat Aug 30 17:49:00 1997 From: pbrant@pcbsys.demon.co.uk (Paul C. Brant) Date: Sat, 30 Aug 1997 17:49:00 -0000 Subject: Looking for java implementation of X Windows Message-ID: <872984445.133531.0@pcbsys.demon.co.uk> Hi, The implementation I know of comes through part of the NTRIGUE Java client from Insignia ( www.insignia.com ) . The system provides extensions to the citrix winframe gubbins and allows non microsoft type machines to run windows apps. They call the JAVA thing keoke ! Hope this helps ! Branty -----Original Message----- From: Bartlee Anderson To: gnu-win32@cygnus.com Date: Sunday, August 31, 1997 1:11 AM Subject: Looking for java implementation of X Windows > >Has anyone seen a Java only implementation of X. >I saw something that I thought I saved, but can't find. > >May have been on another list. > >I know Sergey would also be interested in this. > >______________________________________________________ >Get Your Private, Free Email at http://www.hotmail.com >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From penanen@xray.harvard.edu Sat Aug 30 17:49:00 1997 From: penanen@xray.harvard.edu (Konstantin Penanen) Date: Sat, 30 Aug 1997 17:49:00 -0000 Subject: Sergey's login.w95 Message-ID: <199708282145.RAA13248@xray.harvard.edu> Windows95-- Sergey's login.w95 in the remote-package, that gets renamed to login, should say exec /bin/sh -login instead of exec /bin/sh --login otherwise sh complains about bad option --login. Sergey -- will you include ftpd in your remote package? Thanks for the other utilities. I tried to build inetutils1.3a, but it had too many problems. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From weiku@utkux.utcc.utk.edu Sun Aug 31 13:43:00 1997 From: weiku@utkux.utcc.utk.edu (Wei Ku) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: Compiling with templates Message-ID: <199708311623.MAA27071@solid.ssd.ornl.gov> Did you put the decleration of the templated classes or functions in the header files ? -----Original Message----- From: Ares To: gnu-win32@cygnus.com Date: Saturday, August 30, 1997 2:52 PM Subject: Compiling with templates >I'm in need of using some templates in C++ code, to implement a linked >list, and am having problems with ld linking it. It seems to find that >there is no such functions from my Templated class. The object files for >the linked list come from several different files. I was wondering what >switches to use on the command line to solve my problem. > >Thanks >Jason > > > _/_/_/_/_/ _/_/_/_/ _/ _/ Jason D. Michaelson > _/ _/ _/ _// _// Jason.D.Michaelson-1@umn.edu > _/ _/ _/ _/_/ _/_/ jason13@geocities.com > _/ _/ _/ _/ _/ _/ _/ mich0101@itlabs.umn.edu > _/ _/ _/ _/ _/ _/ _/ WebPage: http://www.geocities.com > _/ _/ _/ _/ _/ _/_/ _/ /Siliww.tc.umn.edu/~mich0101 >_/_/_/ _/_/_/_/ _/ _/ _/ http://www.umn.edu/~mich0101 > >Don't you wish you were using Pegasus Mail for Windoze. If you were, >you'd be able to click on the URL's. :-). Get it at > http://www.pegasus.usa.com/ . Damn, another one of those URL's. > >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From fabundiz@correo_rcb.gfnorte.com.mx Sun Aug 31 13:43:00 1997 From: fabundiz@correo_rcb.gfnorte.com.mx (Felix Abundiz Charles) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: No Subject Message-ID: Alguien sabe como hacer un trigger o un evento que se dispare cuando por ejemplo un usuario copie cierto archivo al servidor Felix Abundiz Charles mailto:fabundiz@gfnorte.com.mx http:\\www.gfnorte.com.mx - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From room_cygnus@uncnsrd.mt-kisco.ny.us Sun Aug 31 13:43:00 1997 From: room_cygnus@uncnsrd.mt-kisco.ny.us (IGnatius T Foobar) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: Has anyone attempted to port Apache? Message-ID: <338259@uncnsrd.mt-kisco.ny.us> Has anyone attempted to port the Apache webserver to the GNU-Win32 environment? One would assume that it would be a _much_ nicer option than Micro$oft's IIS. - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From jim@jimpick.com Sun Aug 31 13:43:00 1997 From: jim@jimpick.com (Jim Pick) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: Looking for java implementation of X Windows References: <19970824073906.12588.qmail@hotmail.com> Message-ID: <199708312043.NAA06577@fleming.jimpick.com> > > Has anyone seen a Java only implementation of X. > I saw something that I thought I saved, but can't find. > > May have been on another list. > > I know Sergey would also be interested in this. Is this what you were interested in? I tried it, couldn't get it to work. I guess I wasn't very persistent. :-) Cheers, - Jim > From: "Eugene O'Neil" > To: > Subject: XTC, an X Window library implemented in pure Java > Date: Wed, 13 Aug 1997 15:39:23 -0400 > > I am writing a Java windowing toolkit named XTC, the X Tool > Collection. Instead of using native methods that depend on Motif or > even xlib, I have instead opted to implement the underlying X Window > Protocol in pure Java. > > My library is small, powerful, object-oriented, multi-threaded, > reasonably efficient, and public domain. If you find this interesting, > download my working demo from the following web page: > > http://www.cs.umb.edu/~eugene/X.html > > -Eugene Cheers, - Jim -- -----BEGIN PGP MESSAGE----- Version: 2.6.3a iQCVAwUBNAnXT+Qz770qyIfJAQHC6QP/ZF95CDVEdMRD/S3CpvQvNyGLBOk/m149 4oeGZOCnAMUHQ2IhKnvCUMt4ZzYydEJJd0n1erfcPOhMHGYSLvlzZIRXqp6CSVqE xB5hx94Et49wzjO7kXt72vqqUcgqd476pwzCJvgw2teYSTarXti8Y1a2f8gmt298 f74DLiTuMy8= =slbY -----END PGP MESSAGE----- From dworrells@coppermountain.com Sun Aug 31 13:43:00 1997 From: dworrells@coppermountain.com (Dan Worrells) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: cscope on NT Message-ID: <3.0.32.19970829173621.009532d0@coppermountain.com> Hi, I've just moved to an NT4.0 shop and sorely miss cscope. I don't see any specific reference to cscope on your site. Do you have any knowledge of how this ports to NT? I suspect that it's not a simple recompile since screen control is involved. Thanks for your help. ____________________________________________________________________ Dan Worrells dworrells@coppermountain.com Copper Mountain Networks, Inc. 619.453.8799 x178 6650 Lusk Blvd. B-103 619.453.9244 (FAX) San Diego, CA 92121-2744 http://www.coppermountain.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sun Aug 31 13:43:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: Troubles with ML software on cygnus? Message-ID: <01BCB53F.98108D80@sos> Hi! What happened with this mailing list for a last two weeks? Almost nobody writes here? Looks like a troubles with ML software on Cygnus - here is the header of the message written Aug 25, but received Aug 30... Received: from cygnus.com (cygnus.COM [205.180.230.20]) by buggy.prospect.com.ru (8.7.5/8.7.3) with ESMTP id JAA01459 for ; Sat, 30 Aug 1997 09:45:24 +0400 Received: (from daemon@localhost) by cygnus.com (8.8.5/8.8.5) id KAA12059 for gnu-win32-outgoing; Mon, 25 Aug 1997 10:56:22 -0700 (PDT) Received: from seawolf.seabase.com (SEAWOLF.seabase.com [38.254.223.4]) by cygnus.com (8.8.5/8.8.5) with ESMTP id KAA12041; Mon, 25 Aug 1997 10:56:07 -0700 (PDT) -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From yung@pobox.com Sun Aug 31 13:43:00 1997 From: yung@pobox.com (Yung Hsien) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: Problems with shell script (bash) Message-ID: <199708281923.MAA17886@rx7.connectinc.com> Hello, I'm having a problem with bash, tr, or echo... I'm not sure who the culprit is here. In any case, here's a short script: >#!/bin/sh > >OPTIONS="A B C" > >FOUND="FALSE" >FIND=`echo a | tr '[a-z]' '[A-Z]'` > >echo $FIND > >for opt in $OPTIONS ; do > if [ $opt = $FIND ] ; then > FOUND="TRUE" > fi >done > >echo $FOUND > What it should echo at the end is TRUE since 'A' is in "A B C". But it echos FALSE. Note, that this works on my UNIX box. Also, if I say FIND=`echo A | tr '[a-z'] '[A-Z]'` that fails also. Thanks, Yung Hsien yung@connectinc.com - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From nealk@net66.com Sun Aug 31 13:43:00 1997 From: nealk@net66.com (Neal Krawetz) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: Read from keyboard? Message-ID: <199708311543.KAA01411@winslow.net66.com> Hello, I have a text-based application that has been successfully ported to OS/2, SunOS, Linux, IRIX, and HP-UX. (The only difference is the non-blocking read from a keyboard function.) I'd like to port it to Win95. I have tried GCC and an older version of Cygnus-GCC, but both seem to be unable to perform a non-blocking read from a text-based application. I'm basically looking for "int getch(stdin)" where it returns the character in the read buffer, or -1 if no character is available. Under GCC for OS/2, the function _read_kbd(0,0,0) does exactly what I want. Under Unix, we use curses to set stdin to non-blocking. Is a possible to make a text-based application with non-blocking read from stdin using the current Cygnus GCC compiler? If so, how? (I wrote about 6 months ago with the same question and was told that it currently cannot be done, but may be possible in the future.) -Neal - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From ellson@lucent.com Sun Aug 31 13:43:00 1997 From: ellson@lucent.com (John Ellson) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: UWIN Message-ID: <340815EF.C2127B34@lucent.com> Hi, I was surpised not to see a reference from the gnu-win32 home page to the UWIN effort by David Korn at AT&T Research. There seems to be a lot of overlap in the two projects. Ref: http://www.research.att.com/sw/tools/uwin/ John Ellson Lucent Technologies (no longer a part of AT&T) - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From hgeleff@spidernet.com.cy Sun Aug 31 13:43:00 1997 From: hgeleff@spidernet.com.cy (Henrik Geleff) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: RCS for WinNT Message-ID: <199708311345.QAA27151@mail0.spidernet.net> I think it is posible to compile RCS with the cygwin tools, I have tried but not succeeded -- I think I did not try hard enough. I know two sources of working binaries: http://www.componentsoftware.com/csrcs/ and http://www.mks.com/solution/si/ Don't tell me that both are comercial, I know, but they work and US $49 for the Component Source solution is not alot !!! Regards with the hope that it helps you Henrik Geleff -----Original Message--- From: sshteingold@cctrading.com To: gnu-win32@cygnus.com Date: 31 August 1997 15:38 Subject: RCS for WinNT > > I looked at this mailing list's archive and there were only 3 messages > about RCS - 2 about compiling it, and one requesting binaries. None > answered, so I repeat the question: > Where can I get the working RCS binaries for winnt? > I have the latest cygwin32 (rcs is not included), but neither who nor > whoami work: > > C:\home\sds>who am i > who: /var/run/utmp: No such file or directory > > C:\home\sds>whoami > whoami: cannot find username for UID 500 > > (I have an environment var USER set, but apparently it's not enough). > Also, I use binary (aka UNIX) file format (no CRLF, just LF). Will RCS > work with it? (someone claimed that RCS 5.6 converts everything to > text - aka DOS - file format) > > If possible, please send me a mail copy of your kind reply. > Thanks. > Sam > > >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From sos@prospect.com.ru Sun Aug 31 13:43:00 1997 From: sos@prospect.com.ru (Sergey Okhapkin) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: Off topic filesystem question Message-ID: <01BCB53E.2928C9B0@sos> William A. Perkins wrote: > I have successfully compiled mkisofs under cygwin32-b18 (on Win95) by > defining NON_UNIXFS, but it won't make a good image. isovfy and > isoinfo work under cygwin32 though. I didn't put much effort into > finding the problem. Has anyone else had any luck? I didn't try to compile mkisofs with cygwin, but it seems to me the troubles are related with garbage filling on file seeks. Would you like to try the same binaries on NT? -- Sergey Okhapkin, http://www.lexa.ru/sos Moscow, Russia Looking for a job - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From pete@lilith.demon.co.uk Sun Aug 31 13:43:00 1997 From: pete@lilith.demon.co.uk (Pete John) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: Flex/Byacc Message-ID: <872930487.0510474.0@lilith.demon.co.uk> ????Hi.???? I'm using the Flex/Byacc utilities distributed with the Cygwin32 b18 package (and they are excellent utillities too), but I'm concerned about the commercial implications for redistributing my software because of the GPL. I have read that software developed with Flex is freely redistributable, but I'm not sure about either Flex or Byacc.. Please can you help? Many thanks Pete John. From weiku@utkux.utcc.utk.edu Sun Aug 31 13:43:00 1997 From: weiku@utkux.utcc.utk.edu (Wei Ku) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: Out of environment space Message-ID: <199708311628.MAA27100@solid.ssd.ornl.gov> This can be done by setting the environment variable memory in the prooperty dialog box. ( Try to click the small icon at left top corner of the window). Wei Ku -----Original Message----- From: mark koi To: gnu-win32@cygnus.com Date: Saturday, August 30, 1997 11:50 AM Subject: Out of environment space >When adding new environment variables for gnu-win32 on win95, I was >receiving OUT OF ENVIRONMENT SPACE. I added the following >to my config.sys file > >shell=c:\windows\command.com /e:512 /p > >And it worked fine, but damn my system is slow. I imagine I am running >a command >interpreter on top of the WIN95 kernel(?). Is there a better way. I >would hope >so, because I could learn a second language with the time it takes to >load apps. > >- >For help on using this list (especially unsubscribing), send a message to >"gnu-win32-request@cygnus.com" with one line of text: "help". > - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From rsvagera@atpibm6000.tuwien.acdot Sun Aug 31 13:43:00 1997 From: rsvagera@atpibm6000.tuwien.acdot (rsvagera@atpibm6000.tuwien.acdot) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: getch() without curses ? Message-ID: Is it necessary to use curses library to have a function compatible to getch() of DOS compilers or would it be possible to change the line buffered input stream temporarily to unbuffered (and the terminal mode as well, if necessary), read one character and switch back to line buffered mode? Is it done this way by DOS C-compilers? Thank you in advance, Robert - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From koi@ssa.crane.navy.mil Sun Aug 31 13:43:00 1997 From: koi@ssa.crane.navy.mil (Mark Koi) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: Undefined Reference?? Message-ID: I have installed both gnu-win32 and mingw32 but both packages give me the same error. It must be an option or something. My brain is fried. gcc -o myprg.exe myprg.c Error: c:\windows\temp\cc9350811.o : undefined reference to 'SetTimer@16' GCC.EXE: Internal compiler error: program ld got fatal signal 1 koi@ssa.crane.navy.mil program: ------------------------------------------------------- #include #include VOID CALLBACK handler(HWND hwnd, UINT umsg, UINT id, DWORD dwTime) { fprintf(stdout,"Callback invoked dwTime=%d\n",dwTime); } int main(int argc, char **argv) { UINT cc; cc=SetTimer(NULL, 0, 500, (TIMERPROC) handler); if (cc == 0) { fprintf(stderr,"SetTimer failed\n"); exit(-1); } fprintf(stdout,"Created timer succesfull\n"); Sleep(10000); fprintf(stdout,"After sleep \n"); } - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Keetnet@wilmington.net Sun Aug 31 13:43:00 1997 From: Keetnet@wilmington.net (Keet / Foxbird) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: Date / Time Stuff.. Message-ID: <3.0.32.19970828205419.007a5e10@wilmington.net> I tried compiling some 'out-of-the-box' examples from a learn C++ guide and I get a really nice error that looks like this: C:\WINDOWS\Desktop\fb5.56\test\c++tutor\source>c++ -g -O -o date date.cpp C:\gnuwin32\b18\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32\cygnus-2.7.2-97040 4\.. /../../../i386-cygwin32\lib/libcygwin.a(libcmain.o)(.text+0x1e):libcmain.cc: und efined reference to `WinMain@16' C__~1.EXE: Internal compiler error: program ld got fatal signal 1 Am I missing a library, an environment setting or something along those lines? I've compiled other date stuff before, but this is the first one that doesn't work. If you need the source, just ask, I've got it available. - Greg Neujahr - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From Graham.Dumpleton@nms.otc.com.au Sun Aug 31 13:43:00 1997 From: Graham.Dumpleton@nms.otc.com.au (Graham Dumpleton) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: sys/socket #define problems Message-ID: <199708300607.QAA26304@baldric.pad.otc.com.au> In sys/socket.h, there exists a whole bunch of #define's like: #define bind cygwin32_bind ... The intent is to ensure that the Win32 versions of these functions are not used. Can some other way be found of doing this, or for C++ at least, can they be made into C++ inline functions. The problem is that some of the names are common enough that one might use them as member function names in C++ classes. To an extent this isn't a problem, however, it is if the order of #include's is such that you include sys/socket.h after the header file which used one of these names, eg. bind. After the point where sys/socket.h is included, your use of the name gets changed, with the cygwin32_ prefix being added. A number of things can happen. If your version of the function had the same prototype, but, was a member of a class, the compiler will quite silently compile the file as okay, however, in reality, it has called the wrong function, not calling your function in the class, but the system bind. A second possibility is that your function had a different prototype in which case it will not find the function and the compiler will given an error. A similar case would be where the name was used as a member variable name, the compiler simply wouldn't find it. -- Graham Dumpleton (grahamd@nms.otc.com.au) - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From mphalpin@iag.net Sun Aug 31 13:43:00 1997 From: mphalpin@iag.net (Sonic Junior) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: Request for Beta 19 Message-ID: <199708301913.PAA06131@iag.net> In Beta 19, could someone please fix this problem: Applications can only have like 64 file descriptors? select() Is it possible to make that number registry selectable? For example, I have one app that wants between 256 and 4095. Registry selectable (like the mounts) would be a very good thing. - SJ P.S. If someone will do this, please let me know! - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From 07535020@mercury.pc-express.com Sun Aug 31 13:43:00 1997 From: 07535020@mercury.pc-express.com (07535020@mercury.pc-express.com) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: UCE - ** SEX! Has Met Its Match! ** Message-ID: Dear gnu-win32@cygnus.com, **** ** Money, Sex, Technology & Home Based Business ** The 4 Most Powerful Forces On Earth Have Merged To Create The Most Explosive Opportunity Of This CENTURY! PLUS: FREE ADULT ENTERTAINMENT WEB SITE!! Become a part of the biggest income opportunity to hit the Internet, while remaining TOTALLY ANONYMOUS. To receive additional information, Send e-mail to: cp1387@usa-now.com or Click Here and type, "send info" in the subject box. ***BY REQUESTING THIS INFO, YOU ARE CONFIRMING THAT YOU ARE AT LEAST 18 YEARS OF AGE*** ********************************************************* ********************************************************* To Be Removed send email to: remove@usa-now.com or Remove Me ********************************************************* ********************************************************* - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help". From root@jacob.remcomp.fr Sun Aug 31 13:43:00 1997 From: root@jacob.remcomp.fr (root) Date: Sun, 31 Aug 1997 13:43:00 -0000 Subject: enquire Message-ID: > I have a program that generates limits.h and float.h. It is also an > extremely good compiler tester. You can find it at: > > http://www.cwi.nl/~steven/enquire.html > > An earlier version is part of the gcc distribution and apparently was > used to generate the float.h include file of b18. > > Compiling and running it with b18 causes a run-time failure (and bash > hangs) with the message: > > (ENQUIRE.EXE 1011) In cygwin_except_handler > > I don't know what the 1011 refers to, but if it's a line number, that > is not where it failed. > > It failed at the point where the program looks to see if dereferencing > NULL causes a trap (which is done using setjmp, and dereferencing NULL). > What a coincidence! I downloaded enquire 3 days ago to test my compiler system (lcc-win32) and I was dismayed that it crashed... in the same place that Cygnus!!!! Meanwhile I corrected this problem. My system wasn't calling 'raise' when a trap occurred, so the workings of 'signal' were impossible. I corrected this, and enquire runs OK in my system. I wish good luck to the friends at cygnus... :-) -- Jacob Navia Logiciels/Informatique 41 rue Maurice Ravel Tel 01 48.23.51.44 93430 Villetaneuse Fax 01 48.23.95.39 France - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request@cygnus.com" with one line of text: "help".