[llvm-testresults] Red bots at night, buildczar's delight?

[Bill Wendling]

Yeah. It made it different. I think I know why, and put another fix in.

Stab-in-the-dark...

-bw

On Nov 17, 2010, at 12:01 PM, dalej wrote:

> Thanks, but this didn't fix the problem.
> 
> On Nov 16, 2010, at 9:34 PM, Bill Wendling wrote:
> 
>> On Nov 16, 2010, at 9:05 PM, Dale Johannesen wrote:
>> 
>>> llvm-arm-linux developed new failures between 119281-119315 (build 4631).   That's an unusually wide range, and several people touched the ARM BE.  Bob looks most likely to me, with Bill second.  This is typical:
>>> 
>>> Unknown addressing sub-mode!
>>> UNREACHABLE executed at ARMCodeEmitter.cpp:1164!
>>> 0  lli       0x00dfa6c8
>>> 1  lli       0x00dfb050
>>> 2  libc.so.6 0x4021cc10 __default_sa_restorer_v2 + 0
>>> 
>> This is probably fall-out from my sub-mode conversion stuff. It was trying to deduce the submode from some random machine operand, which no longer *is* the sub-mode. I committed r119461 to try to fix this...
>> 
>> I don't know why this didn't fail on Darwin...
>> 
>> -bw
>> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-testresults/attachments/20101117/f56d086e/attachment.html>