RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2025-01-31 20:51:52 +01:00
Code Issues Projects Releases Wiki Activity

Fix a logic bug in inline expansion of memcpy / memset with an overlapping

Browse Source 
load / store pair. It's not legal to use a wider load than the size of
the remaining bytes if it's the first pair of load / store.

llvm-svn: 170018
This commit is contained in:
Evan Cheng 2012-12-12 20:43:23 +00:00
parent e47df808ad
commit e2dde0e860
2 changed files with 16 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
lib/CodeGen/SelectionDAG/SelectionDAG.cpp
View File

@ -3469,9 +3469,6 @@ static bool FindOptimalMemOpLowering(std::vector<EVT> &MemOps,
unsigned NumMemOps = 0; unsigned NumMemOps = 0;
while (Size != 0) { while (Size != 0) {
if (++NumMemOps > Limit)
return false;
unsigned VTSize = VT.getSizeInBits() / 8; unsigned VTSize = VT.getSizeInBits() / 8;
while (VTSize > Size) { while (VTSize > Size) {
// For now, only use non-vector load / store's for the left-over pieces. // For now, only use non-vector load / store's for the left-over pieces.
@ -3507,7 +3504,8 @@ static bool FindOptimalMemOpLowering(std::vector<EVT> &MemOps,
// FIXME: Only does this for 64-bit or more since we don't have proper // FIXME: Only does this for 64-bit or more since we don't have proper
// cost model for unaligned load / store. // cost model for unaligned load / store.
bool Fast; bool Fast;
if (AllowOverlap && VTSize >= 8 && NewVTSize < Size && if (NumMemOps && AllowOverlap &&
VTSize >= 8 && NewVTSize < Size &&
TLI.allowsUnalignedMemoryAccesses(VT, &Fast) && Fast) TLI.allowsUnalignedMemoryAccesses(VT, &Fast) && Fast)
VTSize = Size; VTSize = Size;
else { else {
@ -3516,6 +3514,9 @@ static bool FindOptimalMemOpLowering(std::vector<EVT> &MemOps,
} }
} }
if (++NumMemOps > Limit)
return false;
MemOps.push_back(VT); MemOps.push_back(VT);
Size -= VTSize; Size -= VTSize;
} }

11
test/CodeGen/Mips/2012-12-12-ExpandMemcpy.ll Normal file
View File

@ -0,0 +1,11 @@
; RUN: llc -march=mips64el -mcpu=mips64r2 < %s
@.str = private unnamed_addr constant [7 x i8] c"hello\0A\00", align 1
define void @t(i8* %ptr) {
entry:
tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %ptr, i8* getelementptr inbounds ([7 x i8]* @.str, i64 0, i64 0), i64 7, i32 1, i1 false)
ret void
}
declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind