From:	Richard Dawe <rich AT phekda DOT freeserve DOT co DOT uk>
Newsgroups:	comp.os.msdos.djgpp
Subject:	Re: Updated GCC 3.0.4 -> 3.1, now it crashes
Date:	Tue, 11 Jun 2002 22:08:18 +0100
Lines:	93
Message-ID:	<3D0666C2.7B056D86@phekda.freeserve.co.uk>
References:	<3D06DD32 DOT 5090109 AT removethis DOT edu DOT stadia DOT fi>
NNTP-Posting-Host:	modem-117.elmiron.dialup.pol.co.uk
Mime-Version:	1.0
X-Trace:	news5.svr.pol.co.uk 1023830648 25100 62.136.92.117 (11 Jun 2002 21:24:08 GMT)
NNTP-Posting-Date:	11 Jun 2002 21:24:08 GMT
X-Complaints-To:	abuse AT theplanet DOT net
X-Mailer:	Mozilla 4.77 [en] (X11; U; Linux 2.2.19 i586)
X-Accept-Language:	de,fr
To:	djgpp AT delorie DOT com
DJ-Gateway:	from newsgroup comp.os.msdos.djgpp
Reply-To:	djgpp AT delorie DOT com

Hello.

Antti Koskipaa wrote:
> 
> Ok. I just downloaded gcc31b.zip and unzipped it at c:\djgpp with the -d
> option, as usual. I had a previously fully installed and fresh gcc 3.0.4
> with binutils 2.11, djdev203 etc.

Are you running binutils 2.11.2? Try 'as --version' to see what version you
have. If you're not running 2.11.2, then I suggest you upgrade. It's possible
there are some compatibility problems between gcc 3.1 and binutils 2.11.
 
> Now when I try to compile something, it waits for about 10 seconds
> apparently doing nothing, and then prints this:
> 
> c:\source>gcc -v hello.c -o hello.exe
> Reading specs from c:/djgpp/lib/gcc-lib/djgpp/3.1/specs
> Configured with: ../configure i586-pc-msdosdjgpp --prefix=/dev/env/DJDIR
> --disable-nls
> Thread model: single
> gcc version 3.1
> Abort!

This message below is a traceback. Could you try running:

    symify c:/djgpp/bin/gcc.exe

right after it crashes and post the result here please? Kiitos!

> Exiting due to signal SIGABRT
> Raised at eip=00016aee
> eax=00a8d594 ebx=00000120 ecx=00000000 edx=00000000 esi=00000004 edi=ffb534f0
> ebp=00a8d640 esp=00a8d590 program=C:\DJGPP\BIN\GCC.EXE
> cs: sel=00a7  base=83519000  limit=ffb56fff
> ds: sel=00af  base=83519000  limit=ffb56fff
> es: sel=00af  base=83519000  limit=ffb56fff
> fs: sel=0087  base=0000d9f0  limit=0000ffff
> gs: sel=00bf  base=00000000  limit=0010ffff
> ss: sel=00af  base=83519000  limit=ffb56fff
> App stack: [00a8e000..00a0e000]  Exceptn stack: [0002c430..0002a4f0]
> 
> Call frame traceback EIPs:
>    0x00016a14
>    0x00016aee
>    0x00015e8b
>    0x0001231c
>    0x0000d4af
>    0x0000df03
>    0x0000dcb1
>    0x0000df03
>    0x0000dcb1
>    0x0000c543
>    0x0000df03
>    0x0000dcb1
>    0x0000df03
>    0x0000dcb1
>    0x0000df03
>    0x0000dcb1
>    0x0000df03
>    0x0000dcb1
>    0x0000ba48
>    0x0000f711
>    0x00014248
> 
> I verified that I had no old files that the readme mentioned, like
> lib/specs, bin/cc1.exe etc.
> 
> What's even more weird, compiling to assembly works, ie. the '-S' -option,
> as well as with optimizations. the 'compile only' option, '-c' crashes too.
> It seems to hang at the assembly phase. Running 'as' by hand afterwards
> seems to work.

The assembler 'as' is part of binutils. An upgrade to the latest DJGPP port of
binutils may help.

> Oh yes, my system: athlon, 512MB, win98, go32-v2 reports about 330MB of free
> virtual memory. I unzipped without LFN and also tried setting LFN=N.

I have a similar configuration - Athlon, 640MB, Win98 SE. go32-v2 gives me
these statistics:

    DPMI memory available: 475675 Kb
    DPMI swap space available: 173717 Kb

I've been using gcc 3.1 with binutils 2.11.2 for a few weeks and it seems to
work fine for me.

I hope that helps somehow.

Regards,

-- 
Richard Dawe [ http://www.phekda.freeserve.co.uk/richdawe/ ]

- Raw text -