Re: 5.8.1 and gcc-3.2.3

Robert Allerstorfer <roal@anet.at> writes:
>Hi,
>
>I have been able to use gcc 3.2.3 (from the  MinGW distribution
>3.0.0-1) to build perl-5.8.0 patched by ActiveState (starting with AP
>804), but *not* with the original perl-5.8.0 from CPAN. 

perl5.8 pre-dates gcc-3.* and its new fussyness^Wcorrectnes.
Also Mingw's gcc is a pre-release and is a little noisy 
(e.g. about signed/unsigned mixtures.)

>Therefore, I
>assume that the CPAN tarballs are still not good quality stuff (at
>least to be used with MinGW), including the 5.8.1 release candidates.

Please let us know of problems with 5.8.1 release candidates.
I am running a perl5.8.1 built with Mingw on my WinXP I am using for
Tk testing. (But mine is from perforce not a snapshot...)

>I don't know what ActiveState changed that the build succeeds, so we
>probably have to wait until they release their first 5.8.1 based AP
>source. And, again, gcc 3.2.3 seems to work fine, it's not its fault.
>
>best,
>rob.
>
>--
>On Fri, 5 Sep 2003, 15:13 GMT+08 (09:13 local time) greg wrote:
>
>> Has anyone built perl-5.8.1 with gcc-3.2.3.
>> I am having problems with all the Release Candidates, on Win98.
>> Things like target ambiguities between B.xs and B.c, etc
>> And Encode seems to be trying to execute some illegal
>> instrtuction. I get a Windows error message, which terminates
>> the build process.
>
>> Going back to 5.8.0, I get:
>> C:\MSYS\1.0\MINGW\BIN\..\lib\gcc-lib\mingw32\3.2.3\..\..\..\..\mingw32\bin\l
>> d.exe: cannot find .\mini\win32.o
>
>> I was able to build 5.8.0 last year with the patch for post
>> gcc-2.95 builds. Perhaps rather than dig that out again, I will
>> try gcc-2.95 with 5.8.1.
>
>> --
>> Greg Matheson
>> Taiwan
>
>_______________________________________________
>Perl-Win32-Porters mailing list
>Perl-Win32-Porters@listserv.ActiveState.com
>To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs

• Re: 5.8.1 and gcc-3.2.3 by Nick Ing-Simmons
• Re: 5.8.1 and gcc-3.2.3 (& Win98) by Greg Matheson