[DAGCombiner][x86] add transform/hook to decompose integer multiply into shift/add

This is an alternative to D37896. I don't see a way to decompose multiplies 
generically without a target hook to tell us when it's profitable. 

ARM and AArch64 may be able to remove some duplicate code that overlaps with 
this transform.

As a first step, we're only getting the most clear wins on the vector examples
requested in PR34474:
https://bugs.llvm.org/show_bug.cgi?id=34474

As noted in the code comment, it's likely that the x86 constraints are tighter
than necessary, but it may not always be a win to replace a pmullw/pmulld.

Differential Revision: https://reviews.llvm.org/D52195

llvm-svn: 342554

2 files changed, 19 insertions, 0 deletions

diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index ce1dd6bfc5f..ff35748a750 100644
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -4722,6 +4722,23 @@ bool X86TargetLowering::convertSelectOfConstantsToMath(EVT VT) const {
   return true;
 }
 
+bool X86TargetLowering::decomposeMulByConstant(EVT VT, SDValue C) const {
+  // TODO: We handle scalars using custom code, but generic combining could make
+  // that unnecessary.
+  APInt MulC;
+  if (!ISD::isConstantSplatVector(C.getNode(), MulC))
+    return false;
+
+  // If vector multiply is legal, assume that's faster than shl + add/sub.
+  // TODO: Multiply is a complex op with higher latency and lower througput in
+  //       most implementations, so this check could be loosened based on type
+  //       and/or a CPU attribute.
+  if (isOperationLegal(ISD::MUL, VT))
+    return false;
+
+  return (MulC + 1).isPowerOf2() || (MulC - 1).isPowerOf2();
+}
+
 bool X86TargetLowering::isExtractSubvectorCheap(EVT ResVT, EVT SrcVT,
                                                 unsigned Index) const {
   if (!isOperationLegalOrCustom(ISD::EXTRACT_SUBVECTOR, ResVT))
diff --git a/llvm/lib/Target/X86/X86ISelLowering.h b/llvm/lib/Target/X86/X86ISelLowering.h
index caed0219962..9a3726e566e 100644
--- a/llvm/lib/Target/X86/X86ISelLowering.h
+++ b/llvm/lib/Target/X86/X86ISelLowering.h
@@ -1034,6 +1034,8 @@ namespace llvm {
 
     bool convertSelectOfConstantsToMath(EVT VT) const override;
 
+    bool decomposeMulByConstant(EVT VT, SDValue C) const override;
+
     /// Return true if EXTRACT_SUBVECTOR is cheap for this result type
     /// with this index.
     bool isExtractSubvectorCheap(EVT ResVT, EVT SrcVT,