From 1f5b3589c8d0a603267ac4bfd8983e177f27fc1e Mon Sep 17 00:00:00 2001
From: Simon Pilgrim <llvm-dev@redking.me.uk>
Date: Sat, 28 Oct 2017 22:10:40 +0000
Subject: [PATCH] [SelectionDAG] Add support for INSERT_SUBVECTOR to
 computeKnownBits

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316847 91177308-0d34-0410-b5e6-96231b3b80d8
---
 lib/CodeGen/SelectionDAG/SelectionDAG.cpp | 34 +++++++++++++++++++++++
 test/CodeGen/X86/vector-trunc.ll          | 20 ++++++-------
 2 files changed, 42 insertions(+), 12 deletions(-)

diff --git a/lib/CodeGen/SelectionDAG/SelectionDAG.cpp b/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
index 7f47853b3fc..377f84aa0d0 100644
--- a/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
+++ b/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
@@ -2207,6 +2207,40 @@ void SelectionDAG::computeKnownBits(SDValue Op, KnownBits &Known,
     }
     break;
   }
+  case ISD::INSERT_SUBVECTOR: {
+    // If we know the element index, demand any elements from the subvector and
+    // the remainder from the src its inserted into, otherwise demand them all.
+    SDValue Src = Op.getOperand(0);
+    SDValue Sub = Op.getOperand(1);
+    ConstantSDNode *SubIdx = dyn_cast<ConstantSDNode>(Op.getOperand(2));
+    unsigned NumSubElts = Sub.getValueType().getVectorNumElements();
+    if (SubIdx && SubIdx->getAPIntValue().ule(NumElts - NumSubElts)) {
+      Known.One.setAllBits();
+      Known.Zero.setAllBits();
+      uint64_t Idx = SubIdx->getZExtValue();
+      APInt DemandedSubElts = DemandedElts.extractBits(NumSubElts, Idx);
+      if (!!DemandedSubElts) {
+        computeKnownBits(Sub, Known, DemandedSubElts, Depth + 1);
+        if (Known.isUnknown())
+          break; // early-out.
+      }
+      APInt SubMask = APInt::getBitsSet(NumElts, Idx, Idx + NumSubElts);
+      APInt DemandedSrcElts = DemandedElts & ~SubMask;
+      if (!!DemandedSrcElts) {
+        computeKnownBits(Src, Known2, DemandedSrcElts, Depth + 1);
+        Known.One &= Known2.One;
+        Known.Zero &= Known2.Zero;
+      }
+    } else {
+      computeKnownBits(Sub, Known, Depth + 1);
+      if (Known.isUnknown())
+        break; // early-out.
+      computeKnownBits(Src, Known2, Depth + 1);
+      Known.One &= Known2.One;
+      Known.Zero &= Known2.Zero;
+    }
+    break;
+  }
   case ISD::EXTRACT_SUBVECTOR: {
     // If we know the element index, just demand that subvector elements,
     // otherwise demand them all.
diff --git a/test/CodeGen/X86/vector-trunc.ll b/test/CodeGen/X86/vector-trunc.ll
index 2a6d9453e37..91d09831622 100644
--- a/test/CodeGen/X86/vector-trunc.ll
+++ b/test/CodeGen/X86/vector-trunc.ll
@@ -1227,9 +1227,8 @@ define void @trunc16i16_16i8_lshr(<16 x i16> %a) {
 ; AVX1:       # BB#0: # %entry
 ; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
 ; AVX1-NEXT:    vpsrlw $8, %xmm1, %xmm1
-; AVX1-NEXT:    vpackuswb %xmm0, %xmm1, %xmm1
-; AVX1-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[1,3,5,7,9,11,13,15,u,u,u,u,u,u,u,u]
-; AVX1-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX1-NEXT:    vpsrlw $8, %xmm0, %xmm0
+; AVX1-NEXT:    vpackuswb %xmm1, %xmm0, %xmm0
 ; AVX1-NEXT:    vmovdqu %xmm0, (%rax)
 ; AVX1-NEXT:    vzeroupper
 ; AVX1-NEXT:    retq
@@ -1953,15 +1952,12 @@ define void @PR34773(i16* %a0, i8* %a1) {
 ; AVX1-NEXT:    vmovdqu 32(%rdi), %ymm1
 ; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
 ; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
-; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm3
-; AVX1-NEXT:    vpsrlw $8, %xmm3, %xmm3
-; AVX1-NEXT:    vmovdqa {{.*#+}} xmm4 = <1,3,5,7,9,11,13,15,u,u,u,u,u,u,u,u>
-; AVX1-NEXT:    vpshufb %xmm4, %xmm0, %xmm0
-; AVX1-NEXT:    vpackuswb %xmm0, %xmm2, %xmm2
-; AVX1-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
-; AVX1-NEXT:    vpshufb %xmm4, %xmm1, %xmm1
-; AVX1-NEXT:    vpackuswb %xmm0, %xmm3, %xmm2
-; AVX1-NEXT:    vpunpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
+; AVX1-NEXT:    vpsrlw $8, %xmm0, %xmm0
+; AVX1-NEXT:    vpackuswb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
+; AVX1-NEXT:    vpsrlw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
 ; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
 ; AVX1-NEXT:    vmovups %ymm0, (%rsi)
 ; AVX1-NEXT:    vzeroupper
-- 
2.50.1