[llvm-testresults] buildbot failure in llvm on clang-x86_64-linux-selfhost-rel

[Evan Cheng]

I've reverted it for now. The patch is to canonicalize brcond + br codegen by not optimizing away an unconditional branch to a fallthrough block. That indicates a bug downstream.

Evan

On Sep 23, 2010, at 11:32 AM, Jakob Stoklund Olesen wrote:

> 
> On Sep 23, 2010, at 11:25 AM, Evan Cheng wrote:
> 
>> I can't see how r114634 can break this. I'll revert it for now.
> 
> Thanks.
> 
> There's a whole bunch of bots failing since last night. They are all stage2 failures or lnt test failures, so it looks like we are currently miscompiling code.
> 
> This one is also blaming that patch: http://google1.osuosl.org:8011/builders/clang-x86_64-linux-fnt/builds/3264
> 
> Possibly this one: http://google1.osuosl.org:8011/builders/dragonegg-i386-linux/builds/1601
> 
> 
> 
>> 
>> Evan
>> 
>> On Sep 23, 2010, at 11:14 AM, Jakob Stoklund Olesen wrote:
>> 
>>> 
>>> On Sep 23, 2010, at 2:30 AM, buildbot at google1.osuosl.org wrote:
>>> 
>>>> The Buildbot has detected a new failure of clang-x86_64-linux-selfhost-rel on llvm.
>>>> Full details are available at:
>>>> http://google1.osuosl.org:8011/builders/clang-x86_64-linux-selfhost-rel/builds/2569
>>>> 
>>>> Buildbot URL: http://google1.osuosl.org:8011/
>>>> 
>>>> Buildslave for this Build: osu7
>>>> 
>>>> Build Reason: 
>>>> Build Source Stamp: 114634
>>>> Blamelist: evancheng
>>>> 
>>>> BUILD FAILED: failed compile.llvm.stage2
>>> 
>>> Evan,
>>> 
>>> This bot has not recovered yet. It looks like you broke something last night.
>>> 
>>> Please take a look.
>>> 
>>> /jakob
>>> 
>> 
>