[PATCH] AVX intrinsics were broken on CPU with AVX-512 instruction set

[Adam Nemet]

On Jul 21, 2014, at 1:33 PM, Robert Khasanov <rob.khasanov at gmail.com> wrote:

>>> ! In D4605#5, @anemet wrote:
>> Hi Robert,
>> 
>> Is this a correctness or performance fix?
>> 
>> Thanks,
>> Adam
> 
> Hi Adam,
> 
> This is correctness fix.

OK, thanks.  Can you please explain what the bug is and what the fix does in more details.  I would expect AddedComplexity in performance fixes (i.e. even the instruction we’re trying to avoid selecting should be correct to generate just less efficient).

Also the diff is on phabricator is missing the context for some reason...

Adam

> Thanks, 
> Robert
> 
> http://reviews.llvm.org/D4605
> 
>