[llvm-bugs] [Bug 34359] [X86][AVX512] 128-bit shuffles of 256-bit vectors prefer AVX2 instructions instead of AVX512 preventing combining masks into the instruction

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=34359

Simon Pilgrim <llvm-dev at redking.me.uk> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|CONFIRMED                   |RESOLVED

--- Comment #4 from Simon Pilgrim <llvm-dev at redking.me.uk> ---
Resolving - we've found that avoiding constant kmasks for 128/256 blends tend
to be the better solution. And the rest of the shuffle combining is optimal
now.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20190712/eca81456/attachment.html>