[blarson@blars.org: Bug#323126: slate: ftbfs [sparc] Bus error]

[Nick Forde]

Hi Brian,

The alignment problem I hit in March was UltraSPARC specific and
hence I wouldn't expect it to appear on a G5. The quick fix to
keep the Debian builds happy for now would be to switch to -O0
when uname -s = "SunOS". Slate runs extremely slow with this
option, even on high end boxes, but at least it doesn't crash.

Nick.

Brian Rice wrote:
> Hi, Nick!
> 
> I've been trying GCC 4.0 again on OS X Tiger/G5, and without that  
> option it seems to work fine now. I'll check again and see what is  
> actually breaking if I can.
> 
> I've also seen reports that GCC 4 can break assumptions about  argument 
> evaluation order, but I've not been able to reproduce a bug  under it 
> yet - when I do, I'll be looking for this kind of bug, and  it's 
> probably a good idea if others do, too.
> 
> On Aug 16, 2005, at 7:17 AM, Nick Forde wrote:
> 
>> Helmut,
>>
>> I believe this is the GCC UltraSPARC alignment problem I reported
>> back in March. It can be avoided by switching to -O0 or using the
>> compiler directive workaround in:
>>
>> http://lists.tunes.org/archives/slate/2005-March/001205.html
>>
>> I've just tested the -mno-faster-structs option in GCC 4.0 and
>> it still fails as it did with v3.2.3.
>>
>> Regards,
>>
>> Nick.
>>
>> Brian Rice wrote:
>>
>>> That's not really correct. There was an incomplete and therefore   
>>> broken VM patch to the platform sources that I accidentally  pushed  
>>> onto the main repository which I then rolled back. It's  NOT in 
>>> 0.3.5  (never was) and the latest slate-current tarball  doesn't have 
>>> it,  either (now). I only mentioned it to you online  because you 
>>> reported  a problem at about the same time.
>>> I'll repeat for clarity: The problem I noticed only happened  after  
>>> the 0.3.5 release, and was only present in the repository  for 
>>> several  hours. Any problem you're seeing for 0.3.5 is unknown  and 
>>> you will  have to debug it yourself with GDB or whatever.
>>> On Aug 16, 2005, at 3:44 AM, Helmut Grohne wrote:
>>>
>>>> Hi,
>>>>
>>>> this bug was recently reported against slate. This build failure on
>>>> sparc is a known bug as far as i know.
>>>>
>>>> Brian, you said that you think you solved this problem. Is this   
>>>> included
>>>> in the 0.3.5 tarball?
>>>>
>>>> Greetings
>>>>
>>>> Helmut Grohne
>>>>
>>>> ----- Forwarded message from Blars Blarson <blarson at blars.org> -----
>>>>
>>>> X-Sieve: CMU Sieve 2.2
>>>> Subject: Bug#323126: slate: ftbfs [sparc] Bus error
>>>> Reply-To: Blars Blarson <blarson at blars.org>, 323126 at bugs.debian.org
>>>> Resent-From: Blars Blarson <blarson at blars.org>
>>>> Resent-To: debian-bugs-dist at lists.debian.org
>>>> Resent-Cc: Helmut Grohne <helmut at subdivi.de>
>>>> Resent-Date: Sun, 14 Aug 2005 21:48:04 UTC
>>>> Resent-Message-ID: <handler.323126.B.11240553976142 at bugs.debian.org>
>>>> X-Debian-PR-Message: report 323126
>>>> X-Debian-PR-Package: slate
>>>> X-Debian-PR-Keywords:
>>>> From: Blars Blarson <blarson at blars.org>
>>>> To: Debian Bug Tracking System <submit at bugs.debian.org>
>>>> X-Reportbug-Version: 3.15
>>>> X-Spam-Checker-Version: SpamAssassin 2.60-bugs.debian.org_2005_01_02
>>>>     (1.212-2003-09-23-exp) on spohr.debian.org
>>>> X-Spam-Level:
>>>> X-Spam-Status: No, hits=-8.0 required=4.0 tests=BAYES_00,HAS_PACKAGE
>>>>     autolearn=no version=2.60-bugs.debian.org_2005_01_02
>>>> Resent-Sender: Debian BTS <debbugs at bugs.debian.org>
>>>> X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000,  version=0.94.13
>>>>
>>>> Package: slate
>>>> Version: 0.3.4.3-2
>>>> Severity: important
>>>> Justification: fails to build from source
>>>>
>>>> slate failed to build on a sparc buildd, duplicated on my sparc   
>>>> pbuilder.
>>>>
>>>> chmod 644 /build/buildd/slate-0.3.4.3/src/plugins/ socket/../../../ 
>>>> socket.a
>>>> libtool: install: warning: remember to run `libtool --finish / 
>>>> build/ buildd/slate-0.3.4.3/src/plugins/socket'
>>>> rm unix/socket.la
>>>> make[2]: Leaving directory `/build/buildd/slate-0.3.4.3/src/ 
>>>> plugins/ socket'
>>>> make[1]: Leaving directory `/build/buildd/slate-0.3.4.3'
>>>> gcc -o debian/endianess debian/endianess.c
>>>> cd /build/buildd/slate-0.3.4.3 && echo "repl reset. Image   
>>>> saveNamed: 'slate.image'. quit." | ./vm "`debian/endianess`.image"
>>>> /bin/sh: line 1: 15728 Done                    echo "repl reset.   
>>>> Image saveNamed: 'slate.image'. quit."
>>>>      15729 Bus error               | ./vm "`debian/endianess`.image"
>>>> make: *** [build-stamp] Error 138
>>>>
>>>>
>>>> ----- End forwarded message -----
> 
> 
> -- 
> -Brian
> 
>