[X86][SSE] Added ComputeNumSignBitsForTargetNode support for (V)PSRAI

Part 2 of 3.

Differential Revision: https://reviews.llvm.org/D31347

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@298780 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/X86/X86ISelLowering.cpp b/lib/Target/X86/X86ISelLowering.cpp
index 47f20a2b45634957550182d77f16e27a755b0a3e..547250375aa242bbace811e47a8eba3687bc2969 100644 (file)
--- a/lib/Target/X86/X86ISelLowering.cpp
+++ b/lib/Target/X86/X86ISelLowering.cpp
@@ -26661,6 +26661,15 @@ unsigned X86TargetLowering::ComputeNumSignBitsForTargetNode(
     return Tmp;
   }
 
+  case X86ISD::VSRAI: {
+    SDValue Src = Op.getOperand(0);
+    unsigned Tmp = DAG.ComputeNumSignBits(Src, Depth + 1);
+    unsigned VTBits = Op.getValueType().getScalarSizeInBits();
+    APInt ShiftVal = cast<ConstantSDNode>(Op.getOperand(1))->getAPIntValue();
+    ShiftVal += Tmp;
+    return ShiftVal.uge(VTBits) ? VTBits : ShiftVal.getZExtValue();
+  }
+
   case X86ISD::PCMPGT:
   case X86ISD::PCMPEQ:
   case X86ISD::CMPP:

diff --git a/test/CodeGen/X86/combine-and.ll b/test/CodeGen/X86/combine-and.ll
index 00e5f8f847aaec89d1f37bb0496f8696016d5d7d..352705b48d9db8b1bd34ec199d71405a5f9b1dd9 100644 (file)
--- a/test/CodeGen/X86/combine-and.ll
+++ b/test/CodeGen/X86/combine-and.ll
@@ -254,7 +254,7 @@ define <8 x i16> @ashr_mask1_v8i16(<8 x i16> %a0) {
 ; CHECK-LABEL: ashr_mask1_v8i16:
 ; CHECK:       # BB#0:
 ; CHECK-NEXT:    psraw $15, %xmm0
-; CHECK-NEXT:    pand {{.*}}(%rip), %xmm0
+; CHECK-NEXT:    psrlw $15, %xmm0
 ; CHECK-NEXT:    retq
   %1 = ashr <8 x i16> %a0, <i16 15, i16 15, i16 15, i16 15, i16 15, i16 15, i16 15, i16 15>
   %2 = and <8 x i16> %1, <i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1>
@@ -265,7 +265,7 @@ define <4 x i32> @ashr_mask7_v4i32(<4 x i32> %a0) {
 ; CHECK-LABEL: ashr_mask7_v4i32:
 ; CHECK:       # BB#0:
 ; CHECK-NEXT:    psrad $31, %xmm0
-; CHECK-NEXT:    pand {{.*}}(%rip), %xmm0
+; CHECK-NEXT:    psrld $29, %xmm0
 ; CHECK-NEXT:    retq
   %1 = ashr <4 x i32> %a0, <i32 31, i32 31, i32 31, i32 31>
   %2 = and <4 x i32> %1, <i32 7, i32 7, i32 7, i32 7>