RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2025-02-10 14:36:12 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

[X86] Remove unnecessary code from getMaskNode.

Browse Source 
We no longer need to extend mask scalars before bitcasting them to vXi1. This was only needed for the truncate intrinsics. And was really a bug in our lowering of them.

llvm-svn: 350991
This commit is contained in:
Craig Topper 2019-01-12 06:13:44 +00:00
parent 5785c01760
commit 1fbb0d9ea4
1 changed files with 1 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
lib/Target/X86/X86ISelLowering.cpp
View File

@ -21332,11 +21332,7 @@ static SDValue getMaskNode(SDValue Mask, MVT MaskVT,
if (X86::isZeroNode(Mask))
return DAG.getConstant(0, dl, MaskVT);
if (MaskVT.bitsGT(Mask.getSimpleValueType())) {
// Mask should be extended
Mask = DAG.getNode(ISD::ANY_EXTEND, dl,
MVT::getIntegerVT(MaskVT.getSizeInBits()), Mask);
}
assert(MaskVT.bitsLE(Mask.getSimpleValueType()) && "Unexpected mask size!");
if (Mask.getSimpleValueType() == MVT::i64 && Subtarget.is32Bit()) {
assert(MaskVT == MVT::v64i1 && "Expected v64i1 mask!");