6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2025-07-01 12:24:24 +00:00
Code Issues Projects Releases Wiki Activity

Add a fixme message about canonicalizing floating-point vector types.

Browse Source 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@78897 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Bob Wilson
2009-08-13 06:01:30 +00:00
parent bab812b4b0
commit bfcbb507c2
1 changed files with 2 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
lib/Target/ARM/ARMISelLowering.cpp
View File

@ -2442,6 +2442,8 @@ static SDValue LowerVECTOR_SHUFFLE(SDValue Op, SelectionDAG &DAG) {
// DAG nodes, instead of keeping them as shuffles and matching them again // DAG nodes, instead of keeping them as shuffles and matching them again
// during code selection. This is more efficient and avoids the possibility // during code selection. This is more efficient and avoids the possibility
// of inconsistencies between legalization and selection. // of inconsistencies between legalization and selection.
// FIXME: floating-point vectors should be canonicalized to integer vectors
// of the same time so that they get CSEd properly.
if (isVREVMask(SVN, 64)) if (isVREVMask(SVN, 64))
return DAG.getNode(ARMISD::VREV64, dl, VT, SVN->getOperand(0)); return DAG.getNode(ARMISD::VREV64, dl, VT, SVN->getOperand(0));
if (isVREVMask(SVN, 32)) if (isVREVMask(SVN, 32))