Use movups to lower memcpy and memset even if it's not fast (like corei7).
The theory is it's still faster than a pair of movq / a quad of movl. This
will probably hurt older chips like P4 but should run faster on current
and future Intel processors. rdar://8817010
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122955 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/lib/Target/X86/X86ISelLowering.cpp b/lib/Target/X86/X86ISelLowering.cpp
index f871b5a..ddec78b 100644
--- a/lib/Target/X86/X86ISelLowering.cpp
+++ b/lib/Target/X86/X86ISelLowering.cpp
@@ -1063,12 +1063,8 @@
// linux. This is because the stack realignment code can't handle certain
// cases like PR2962. This should be removed when PR2962 is fixed.
const Function *F = MF.getFunction();
- if (NonScalarIntSafe &&
- !F->hasFnAttr(Attribute::NoImplicitFloat)) {
+ if (NonScalarIntSafe && !F->hasFnAttr(Attribute::NoImplicitFloat)) {
if (Size >= 16 &&
- (Subtarget->isUnalignedMemAccessFast() ||
- ((DstAlign == 0 || DstAlign >= 16) &&
- (SrcAlign == 0 || SrcAlign >= 16))) &&
Subtarget->getStackAlignment() >= 16) {
if (Subtarget->hasSSE2())
return MVT::v4i32;