[InstCombine] reverse 'trunc X to <N x i1>' canonicalization

icmp ne (and X, 1), 0 --> trunc X to N x i1

Ideally, we'd do the same for scalars, but there will likely be 
regressions unless we add more trunc folds as we're doing here 
for vectors.

The motivating vector case is from PR37549:
https://bugs.llvm.org/show_bug.cgi?id=37549

define <4 x float> @bitwise_select(<4 x float> %x, <4 x float> %y, <4 x float> %z, <4 x float> %w) {
  %c = fcmp ole <4 x float> %x, %y
  %s = sext <4 x i1> %c to <4 x i32>
  %s1 = shufflevector <4 x i32> %s, <4 x i32> undef, <4 x i32> <i32 0, i32 0, i32 1, i32 1>
  %s2 = shufflevector <4 x i32> %s, <4 x i32> undef, <4 x i32> <i32 2, i32 2, i32 3, i32 3>
  %cond = or <4 x i32> %s1, %s2
  %condtr = trunc <4 x i32> %cond to <4 x i1>
  %r = select <4 x i1> %condtr, <4 x float> %z, <4 x float> %w
  ret <4 x float> %r
}

Here's a sampling of the vector codegen for that case using 
mask+icmp (current behavior) vs. trunc (with this patch):

AVX before:

vcmpleps	%xmm1, %xmm0, %xmm0
vpermilps	$80, %xmm0, %xmm1 ## xmm1 = xmm0[0,0,1,1]
vpermilps	$250, %xmm0, %xmm0 ## xmm0 = xmm0[2,2,3,3]
vorps	%xmm0, %xmm1, %xmm0
vandps	LCPI0_0(%rip), %xmm0, %xmm0
vxorps	%xmm1, %xmm1, %xmm1
vpcmpeqd	%xmm1, %xmm0, %xmm0
vblendvps	%xmm0, %xmm3, %xmm2, %xmm0

AVX after:

vcmpleps	%xmm1, %xmm0, %xmm0
vpermilps	$80, %xmm0, %xmm1 ## xmm1 = xmm0[0,0,1,1]
vpermilps	$250, %xmm0, %xmm0 ## xmm0 = xmm0[2,2,3,3]
vorps	%xmm0, %xmm1, %xmm0
vblendvps	%xmm0, %xmm2, %xmm3, %xmm0

AVX512f before:

vcmpleps	%xmm1, %xmm0, %xmm0
vpermilps	$80, %xmm0, %xmm1 ## xmm1 = xmm0[0,0,1,1]
vpermilps	$250, %xmm0, %xmm0 ## xmm0 = xmm0[2,2,3,3]
vorps	%xmm0, %xmm1, %xmm0
vpbroadcastd	LCPI0_0(%rip), %xmm1 ## xmm1 = [1,1,1,1]
vptestnmd	%zmm1, %zmm0, %k1
vblendmps	%zmm3, %zmm2, %zmm0 {%k1}

AVX512f after:

vcmpleps	%xmm1, %xmm0, %xmm0
vpermilps	$80, %xmm0, %xmm1 ## xmm1 = xmm0[0,0,1,1]
vpermilps	$250, %xmm0, %xmm0 ## xmm0 = xmm0[2,2,3,3]
vorps	%xmm0, %xmm1, %xmm0
vpslld	$31, %xmm0, %xmm0
vptestmd	%zmm0, %zmm0, %k1
vblendmps	%zmm2, %zmm3, %zmm0 {%k1}

AArch64 before:

fcmge	v0.4s, v1.4s, v0.4s
zip1	v1.4s, v0.4s, v0.4s
zip2	v0.4s, v0.4s, v0.4s
orr	v0.16b, v1.16b, v0.16b
movi	v1.4s, #1
and	v0.16b, v0.16b, v1.16b
cmeq	v0.4s, v0.4s, #0
bsl	v0.16b, v3.16b, v2.16b

AArch64 after:

fcmge	v0.4s, v1.4s, v0.4s
zip1	v1.4s, v0.4s, v0.4s
zip2	v0.4s, v0.4s, v0.4s
orr	v0.16b, v1.16b, v0.16b
bsl	v0.16b, v2.16b, v3.16b

PowerPC-le before:

xvcmpgesp 34, 35, 34
vspltisw 0, 1
vmrglw 3, 2, 2
vmrghw 2, 2, 2
xxlor 0, 35, 34
xxlxor 35, 35, 35
xxland 34, 0, 32
vcmpequw 2, 2, 3
xxsel 34, 36, 37, 34

PowerPC-le after:

xvcmpgesp 34, 35, 34
vmrglw 3, 2, 2
vmrghw 2, 2, 2
xxlor 0, 35, 34
xxsel 34, 37, 36, 0

Differential Revision: https://reviews.llvm.org/D52747

llvm-svn: 344082

1 files changed, 27 insertions, 4 deletions

diff --git a/llvm/lib/Transforms/InstCombine/InstCombineCasts.cpp b/llvm/lib/Transforms/InstCombine/InstCombineCasts.cpp
index fd59c3a7c0c..74f1e695ff6 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombineCasts.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineCasts.cpp
@@ -706,12 +706,35 @@ Instruction *InstCombiner::visitTrunc(TruncInst &CI) {
   if (SimplifyDemandedInstructionBits(CI))
     return &CI;
 
-  // Canonicalize trunc x to i1 -> (icmp ne (and x, 1), 0), likewise for vector.
   if (DestTy->getScalarSizeInBits() == 1) {
-    Constant *One = ConstantInt::get(SrcTy, 1);
-    Src = Builder.CreateAnd(Src, One);
     Value *Zero = Constant::getNullValue(Src->getType());
-    return new ICmpInst(ICmpInst::ICMP_NE, Src, Zero);
+    if (DestTy->isIntegerTy()) {
+      // Canonicalize trunc x to i1 -> icmp ne (and x, 1), 0 (scalar only).
+      // TODO: We canonicalize to more instructions here because we are probably
+      // lacking equivalent analysis for trunc relative to icmp. There may also
+      // be codegen concerns. If those trunc limitations were removed, we could
+      // remove this transform.
+      Value *And = Builder.CreateAnd(Src, ConstantInt::get(SrcTy, 1));
+      return new ICmpInst(ICmpInst::ICMP_NE, And, Zero);
+    }
+
+    // For vectors, we do not canonicalize all truncs to icmp, so optimize
+    // patterns that would be covered within visitICmpInst.
+    Value *X;
+    const APInt *C;
+    if (match(Src, m_OneUse(m_LShr(m_Value(X), m_APInt(C))))) {
+      // trunc (lshr X, C) to i1 --> icmp ne (and X, C'), 0
+      APInt MaskC = APInt(SrcTy->getScalarSizeInBits(), 1).shl(*C);
+      Value *And = Builder.CreateAnd(X, ConstantInt::get(SrcTy, MaskC));
+      return new ICmpInst(ICmpInst::ICMP_NE, And, Zero);
+    }
+    if (match(Src, m_OneUse(m_c_Or(m_LShr(m_Value(X), m_APInt(C)),
+                                   m_Deferred(X))))) {
+      // trunc (or (lshr X, C), X) to i1 --> icmp ne (and X, C'), 0
+      APInt MaskC = APInt(SrcTy->getScalarSizeInBits(), 1).shl(*C) | 1;
+      Value *And = Builder.CreateAnd(X, ConstantInt::get(SrcTy, MaskC));
+      return new ICmpInst(ICmpInst::ICMP_NE, And, Zero);
+    }
   }
 
   // FIXME: Maybe combine the next two transforms to handle the no cast case