X86: Open up some opportunities for constant folding by postponing shift lowering.

Fixes PR15141.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@174327 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/X86/X86ISelLowering.cpp b/lib/Target/X86/X86ISelLowering.cpp
index 5b3f06689f24de7dd4fe88aa584425a9a987251c..496202380af8d07a018c762f123f76458713d0c0 100644 (file)
--- a/lib/Target/X86/X86ISelLowering.cpp
+++ b/lib/Target/X86/X86ISelLowering.cpp
@@ -11583,8 +11583,7 @@ SDValue X86TargetLowering::LowerShift(SDValue Op, SelectionDAG &DAG) const {
 
   // Lower SHL with variable shift amount.
   if (VT == MVT::v4i32 && Op->getOpcode() == ISD::SHL) {
-    Op = DAG.getNode(X86ISD::VSHLI, dl, VT, Op.getOperand(1),
-                     DAG.getConstant(23, MVT::i32));
+    Op = DAG.getNode(ISD::SHL, dl, VT, Amt, DAG.getConstant(23, VT));
 
     Op = DAG.getNode(ISD::ADD, dl, VT, Op, DAG.getConstant(0x3f800000U, VT));
     Op = DAG.getNode(ISD::BITCAST, dl, MVT::v4f32, Op);
@@ -11595,8 +11594,7 @@ SDValue X86TargetLowering::LowerShift(SDValue Op, SelectionDAG &DAG) const {
     assert(Subtarget->hasSSE2() && "Need SSE2 for pslli/pcmpeq.");
 
     // a = a << 5;
-    Op = DAG.getNode(X86ISD::VSHLI, dl, MVT::v8i16, Op.getOperand(1),
-                     DAG.getConstant(5, MVT::i32));
+    Op = DAG.getNode(ISD::SHL, dl, VT, Amt, DAG.getConstant(5, VT));
     Op = DAG.getNode(ISD::BITCAST, dl, VT, Op);
 
     // Turn 'a' into a mask suitable for VSELECT

diff --git a/test/CodeGen/X86/avx-shift.ll b/test/CodeGen/X86/avx-shift.ll
index 681747b844a03491b9e1bbb83b1aa8091522c642..b0bff454c18684be0c91f0c6e1c1a70853760bf0 100644 (file)
--- a/test/CodeGen/X86/avx-shift.ll
+++ b/test/CodeGen/X86/avx-shift.ll
@@ -112,6 +112,16 @@ define <8 x i32> @vshift08(<8 x i32> %a) nounwind {
   ret <8 x i32> %bitop
 }
 
+; PR15141
+; CHECK: _vshift13:
+; CHECK-NOT: vpsll
+; CHECK: vcvttps2dq
+; CHECK-NEXT: vpmulld
+define <4 x i32> @vshift13(<4 x i32> %in) {
+  %T = shl <4 x i32> %in, <i32 0, i32 1, i32 2, i32 4>
+  ret <4 x i32> %T
+}
+
 ;;; Uses shifts for sign extension
 ; CHECK: _sext_v16i16
 ; CHECK: vpsllw