Optimize some 64-bit multiplication by constants into two lea's or one lea + shl since imulq is slow (latency 5). e.g.
x * 40
=>
shlq $3, %rdi
leaq (%rdi,%rdi,4), %rax
This has the added benefit of allowing more multiply to be folded into addressing mode. e.g.
a * 24 + b
=>
leaq (%rdi,%rdi,2), %rax
leaq (%rsi,%rax,8), %rax
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@67917 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/lib/CodeGen/SelectionDAG/DAGCombiner.cpp b/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
index 041c500..78d5d40 100644
--- a/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
+++ b/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
@@ -93,14 +93,14 @@
}
SDValue CombineTo(SDNode *N, const SDValue *To, unsigned NumTo,
- bool AddTo = true);
+ bool AddTo = true);
SDValue CombineTo(SDNode *N, SDValue Res, bool AddTo = true) {
return CombineTo(N, &Res, 1, AddTo);
}
SDValue CombineTo(SDNode *N, SDValue Res0, SDValue Res1,
- bool AddTo = true) {
+ bool AddTo = true) {
SDValue To[] = { Res0, Res1 };
return CombineTo(N, To, 2, AddTo);
}
@@ -293,19 +293,19 @@
}
SDValue TargetLowering::DAGCombinerInfo::
-CombineTo(SDNode *N, const std::vector<SDValue> &To) {
- return ((DAGCombiner*)DC)->CombineTo(N, &To[0], To.size());
+CombineTo(SDNode *N, const std::vector<SDValue> &To, bool AddTo) {
+ return ((DAGCombiner*)DC)->CombineTo(N, &To[0], To.size(), AddTo);
}
SDValue TargetLowering::DAGCombinerInfo::
-CombineTo(SDNode *N, SDValue Res) {
- return ((DAGCombiner*)DC)->CombineTo(N, Res);
+CombineTo(SDNode *N, SDValue Res, bool AddTo) {
+ return ((DAGCombiner*)DC)->CombineTo(N, Res, AddTo);
}
SDValue TargetLowering::DAGCombinerInfo::
-CombineTo(SDNode *N, SDValue Res0, SDValue Res1) {
- return ((DAGCombiner*)DC)->CombineTo(N, Res0, Res1);
+CombineTo(SDNode *N, SDValue Res0, SDValue Res1, bool AddTo) {
+ return ((DAGCombiner*)DC)->CombineTo(N, Res0, Res1, AddTo);
}
void TargetLowering::DAGCombinerInfo::