diff options
| author | David Zarzycki <dave@znu.io> | 2019-09-23 05:00:59 +0000 |
|---|---|---|
| committer | David Zarzycki <dave@znu.io> | 2019-09-23 05:00:59 +0000 |
| commit | a7a515cb7738d77282fc52fc41c97fd914515aae (patch) | |
| tree | 16178cc76dc9d41ba4238bb03cd594461b4451be /llvm/lib/Target | |
| parent | d8ac51ab8fde3dbf7d55e5940584019330991dd7 (diff) | |
| download | bcm5719-llvm-a7a515cb7738d77282fc52fc41c97fd914515aae.tar.gz bcm5719-llvm-a7a515cb7738d77282fc52fc41c97fd914515aae.zip | |
Prefer AVX512 memcpy when applicable
When AVX512 is available and the preferred vector width is 512-bits or
more, we should prefer AVX512 for memcpy().
https://bugs.llvm.org/show_bug.cgi?id=43240
https://reviews.llvm.org/D67874
llvm-svn: 372540
Diffstat (limited to 'llvm/lib/Target')
| -rw-r--r-- | llvm/lib/Target/X86/X86ISelLowering.cpp | 5 |
1 files changed, 5 insertions, 0 deletions
diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp index 2e40d8805f3..7dbd0c1e45d 100644 --- a/llvm/lib/Target/X86/X86ISelLowering.cpp +++ b/llvm/lib/Target/X86/X86ISelLowering.cpp @@ -2149,6 +2149,11 @@ EVT X86TargetLowering::getOptimalMemOpType( if (Size >= 16 && (!Subtarget.isUnalignedMem16Slow() || ((DstAlign == 0 || DstAlign >= 16) && (SrcAlign == 0 || SrcAlign >= 16)))) { + // FIXME: Check if unaligned 64-byte accesses are slow. + if (Size >= 64 && Subtarget.hasAVX512() && + (Subtarget.getPreferVectorWidth() >= 512)) { + return Subtarget.hasBWI() ? MVT::v64i8 : MVT::v16i32; + } // FIXME: Check if unaligned 32-byte accesses are slow. if (Size >= 32 && Subtarget.hasAVX() && (Subtarget.getPreferVectorWidth() >= 256)) { |
