[llvm-commits] [Fwd: [llvm] r82689 - in /llvm/trunk: lib/Transforms/IPO/IPConstantPropagation.cpp lib/Transforms/Scalar/SCCP.cpp test/Transforms/IPConstantProp/2009-09-24-byval-ptr.ll]

[Török Edwin]

On 2009-09-24 21:59, Tanya M. Lattner wrote:
>
>>> I agree that this would be nice to fix for 2.6, but it may be too
>>> late.  It depends on the fate of PR5004, which is still pending.
>>
>> Ok, if you do pull this patch into 2.6 please pull 82700 too (it fixes
>> the code pessimization I mentioned, and its just a few lines again).
>
> I'll do this tonight when I get home from work.
>
> However, I'd like to say that normally its way too late for these
> sorts of patches to be merged in and it introduces risk that I
> personally think we should avoid at this stage in the release process.
> Normally, the answer would be no. If it introduces any sort of
> problem, it will be reverted and not fixed.
>

Sounds good.

Thanks,
--Edwin