Fix one x86_64 abi issue and the test to actually look for the right thing,
which is: { <4 x float>, <4 x float> } should continue to go through memory.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@134946 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/lib/CodeGen/TargetInfo.cpp b/lib/CodeGen/TargetInfo.cpp
index 0c070a1..7c850af 100644
--- a/lib/CodeGen/TargetInfo.cpp
+++ b/lib/CodeGen/TargetInfo.cpp
@@ -1228,6 +1228,13 @@
const RecordDecl *RD = RT->getDecl();
+ // The only case a 256-bit wide vector could be used is when the struct
+ // contains a single 256-bit element. Since Lo and Hi logic isn't extended
+ // to work for sizes wider than 128, early check and fallback to memory.
+ RecordDecl::field_iterator FirstElt = RD->field_begin();
+ if (Size > 128 && getContext().getTypeSize(FirstElt->getType()) != 256)
+ return;
+
// Assume variable sized types are passed in memory.
if (RD->hasFlexibleArrayMember())
return;
@@ -1263,7 +1270,7 @@
// Classify the fields one at a time, merging the results.
unsigned idx = 0;
- for (RecordDecl::field_iterator i = RD->field_begin(), e = RD->field_end();
+ for (RecordDecl::field_iterator i = FirstElt, e = RD->field_end();
i != e; ++i, ++idx) {
uint64_t Offset = OffsetBase + Layout.getFieldOffset(idx);
bool BitField = i->isBitField();