This is the mail archive of the cygwin@cygwin.com mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: make error: cannot exec gcc (after many compiles) : Win95 : Cygwin of 9/19/2001


Charles Hixson schrieb am 2001-09-20, 13:44:

>Summary:
>dllwrap: installation problem, cannot exec `gcc': No such file
>or directory
>dllwrap: gcc exited with status 255
>make: *** [dummy] Error 1

After configure and compiling, I cannot say more as before, sorry.
It works for me.  I have done:

./configure 2>&1 | tee log.configure
make 2>&1 | tee log.make
make install 2>&1 | tee log.install

Everything compiles and all is installed in place now.
Works as before:)

$ ruby --version
ruby 1.6.5 (2001-09-19) [i686-cygwin]

>Details:
>With the new installation of CygWin, I was attempting to compile
>   the new release of Ruby.  The compiles worked fine, the error
>occured during the attempt at a creation of a shared library.  I
>don't really understand make, but it looks relatively
>straightforward.
>
>The out.txt file has been edited to include the error message.
>It didn't appear in a simply redirected output file.  Strangely
>it didn't appear in a redirected stderr file either:
>make > out.txt 2>err.txt
>err.txt ends up blank, and out.txt doesn't include the error
>message, but it appears on the screen if I don't do a redirection.

I do redirection like shown above, that is o.k.

>P.S.:  Is there a bugzilla or something?  This seems a bit long
>to include in a post to a mailing list.

No, just post it. Better if you send it as attachment, so the output
gets not mangeled that much and is better readable then.

>Out.txt:
>gcc -g -O2 -I. -I. -I/usr/local/include -c array.c
[...]
>gcc -g -O2 -I. -I. -I/usr/local/include -c main.c

Up to here iyt is identical.

>gcc -g -O2  main.o dmyext.o librubys.a -lcrypt   -o miniruby.exe

gcc -g -O2 -s main.o dmyext.o librubys.a -lcrypt   -o miniruby.exe

Now here is a difference. '-s' is an additional flag here at my build.
'-s' comes with my standard LDFLAGS, that is also the main difference
in the Makefile. NOTE, for gnu make, GNUmkefile is the first Makefile.

>echo EXPORTS > ruby.def
>nm --extern-only --defined-only librubys.a | \
>   ./miniruby.exe -ne 'puts $1 if / [CDT] _(.*)$/' >> ruby.def
>windres --include-dir . --include-dir . --include-dir ./win32 
>dummy.rc dummy.res.o
>dllwrap --target=cygwin --as=as --dlltool-name=dlltool 
>--driver-name=gcc --export-all -s --output-exp=ruby.exp 
>--dllname=ruby.exe --output-lib=libcygwin-ruby16.a 
>--add-stdcall-alias --def=ruby.def array.o bignum.o class.o 
>compar.o dir.o dln.o enum.o error.o eval.o file.o gc.o hash.o 
>inits.o io.o marshal.o math.o numeric.o object.o pack.o parse.o 
>process.o prec.o random.o range.o re.o regex.o ruby.o signal.o 
>sprintf.o st.o string.o struct.o time.o util.o variable.o 
>version.o flock.o strftime.o  dmyext.o dummy.res.o -lcrypt   -o 
>dummy
>
>dllwrap: installation problem, cannot exec `gcc': No such file 
>or directory
>dllwrap: gcc exited with status 255
>make: *** [dummy] Error 1

And I get no error here. All builds fine.

>LDFLAGS =  $(CFLAGS)
LDFLAGS =  $(CFLAGS) -s

>a:  fd           N/A    N/A
>c:  hd  FAT32   9756Mb  86% CP    UN
>d:  hd  FAT32   1003Mb   4% CP    UN           WORK

No NT, no NTFS.
Win95 is pretty old nowadays...

One thing is interesting. As i built ruby-1.6.4 (IMO) it was build a
ruby.dll by default which it is not now, only a static ruby binary was
built.

Gerrit


-- 
=^..^=

--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Bug reporting:         http://cygwin.com/bugs.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]