[llvm-bugs] [Bug 34370] [X86][AVX512] suboptimal shuffle sequence instead of one vpermps instruction

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=34370

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

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

--- Comment #1 from Simon Pilgrim <llvm-dev at redking.me.uk> ---
Current codegen combine the 2 shuffles AND the select into a single binary
shuffle:

    vmovaps (%rdi), %ymm2
    vmovaps {{.*#+}} ymm1 = [7,6,3,11,7,6,14,15]
    vpermi2ps %ymm0, %ymm2, %ymm1
    vmovaps %ymm1, %ymm0
    retq

Added a test case at rL323106

-- 
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/20180122/0aea2714/attachment-0001.html>