diff options
| author | Evan Cheng <evan.cheng@apple.com> | 2011-08-31 02:05:24 +0000 |
|---|---|---|
| committer | Evan Cheng <evan.cheng@apple.com> | 2011-08-31 02:05:24 +0000 |
| commit | cb1e5bae4c483bdffcde1acc51c4070d1b59c38c (patch) | |
| tree | 9c99f643ffcfb41f2eeaae54c23f6ea052942007 /llvm/lib/Target/X86 | |
| parent | 0399ec26c0c9c7fdda5b903bda4c99f974920aee (diff) | |
| download | bcm5719-llvm-cb1e5bae4c483bdffcde1acc51c4070d1b59c38c.tar.gz bcm5719-llvm-cb1e5bae4c483bdffcde1acc51c4070d1b59c38c.zip | |
Fix (movhps load) lowering / pattern to match more cases. rdar://10050549
llvm-svn: 138848
Diffstat (limited to 'llvm/lib/Target/X86')
| -rw-r--r-- | llvm/lib/Target/X86/X86ISelLowering.cpp | 7 | ||||
| -rw-r--r-- | llvm/lib/Target/X86/X86InstrSSE.td | 3 |
2 files changed, 7 insertions, 3 deletions
diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp index effe06e2a9c..7fa0dba9a01 100644 --- a/llvm/lib/Target/X86/X86ISelLowering.cpp +++ b/llvm/lib/Target/X86/X86ISelLowering.cpp @@ -6236,8 +6236,11 @@ SDValue getMOVLowToHigh(SDValue &Op, DebugLoc &dl, SelectionDAG &DAG, if (HasSSE2 && VT == MVT::v2f64) return getTargetShuffleNode(X86ISD::MOVLHPD, dl, VT, V1, V2, DAG); - // v4f32 or v4i32 - return getTargetShuffleNode(X86ISD::MOVLHPS, dl, VT, V1, V2, DAG); + // v4f32 or v4i32: canonizalized to v4f32 (which is legal for SSE1) + return DAG.getNode(ISD::BITCAST, dl, VT, + getTargetShuffleNode(X86ISD::MOVLHPS, dl, MVT::v4f32, + DAG.getNode(ISD::BITCAST, dl, MVT::v4f32, V1), + DAG.getNode(ISD::BITCAST, dl, MVT::v4f32, V2), DAG)); } static diff --git a/llvm/lib/Target/X86/X86InstrSSE.td b/llvm/lib/Target/X86/X86InstrSSE.td index e8a38f5ed83..595010d8055 100644 --- a/llvm/lib/Target/X86/X86InstrSSE.td +++ b/llvm/lib/Target/X86/X86InstrSSE.td @@ -691,11 +691,12 @@ let Predicates = [HasSSE1] in { // MOVHPS patterns def : Pat<(movlhps VR128:$src1, (bc_v4i32 (v2i64 (X86vzload addr:$src2)))), (MOVHPSrm (v4i32 VR128:$src1), addr:$src2)>; + def : Pat<(X86Movlhps VR128:$src1, (bc_v4f32 (v2f64 (scalar_to_vector (loadf64 addr:$src2))))), (MOVHPSrm VR128:$src1, addr:$src2)>; def : Pat<(X86Movlhps VR128:$src1, - (bc_v4i32 (v2i64 (X86vzload addr:$src2)))), + (bc_v4f32 (v2i64 (X86vzload addr:$src2)))), (MOVHPSrm VR128:$src1, addr:$src2)>; // MOVLHPS patterns |
