Re: Perl instabilities

["Gerrit P. Haase" <freeweb at nyckelpiga dot de>]

Gerrit schrieb:

> Hallo Greg,

>> Breakpoint 2, 0x0040375c in ?? ()
>> (gdb) bt
>> #0  0x0040375c in ?? ()
>> #1  0x0040103c in ?? ()
>> #2  0xbff8b6be in KERNEL32!EnumTimeFormatsA () from
>> /cy/c/WINDOWS/SYSTEM/KERNEL32.DLL
>> #3  0xbff8b570 in KERNEL32!FT_Thunk () from
>> /cy/c/WINDOWS/SYSTEM/KERNEL32.DLL
>> #4  0xbff89f33 in KERNEL32!BeginUpdateResourceA () from
>> /cy/c/WINDOWS/SYSTEM/KERNEL32.DLL
>> Cannot access memory at address 0x856dbfec
>> (gdb) nexti

> Hmmm, I'll build a version with Cygwin malloc instead of Perl malloc,
> maybe a bug in malloc()?

Finally it happens perl using system malloc() instead perl malloc():
http://anfaenger.de/cygwin/perl/perl-5.8.0-umymalloc.tar.bz2

This version is compiled with -Uusemymalloc which means that the system
malloc is used instead of the perl malloc, it is also compiled with debug
infos, so if you also want to debug perl then the sourcetre is also
needed (it is the same source as always):
http://anfaenger.de/cygwin/perl/perl-5.8.0-umymalloc-srctree.tar.bz2
this package unpacks in /src/cygwinperl/perl-5.8.0 if unpacked from the
Cygwin root where it should be for the debugger to find it.

So could you and everyone else with Win98 please try if it makes a
difference if we use the Cygwin malloc() instead of Perl's own malloc()?

Buildscripts and stuff available on request only.


Gerrit
-- 
=^..^=


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/