[PATCH] D140603: Resolve a long-standing FIXME in memcpyopt.

[Owen Anderson via Phabricator via llvm-commits]

resistor created this revision.
Herald added a subscriber: hiraditya.
Herald added a project: All.
resistor requested review of this revision.
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.

Inspecting the downstream use of the cpyAlign, it is clear that
`performCallSlotOptzn` is expecting it to represent the alignment
of the copy destination, not the minimum of the src and dest
alignments. This patch renames the parameter to make this more
obvious.

I believe this change is NFC, because the downstream code has
alignment checks such that it all works out in the end. I have not
been able to construct a test case that actually triggers a change
in output.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D140603

Files:
  llvm/lib/Transforms/Scalar/MemCpyOptimizer.cpp
  llvm/test/Transforms/MemCpyOpt/callslot.ll


Index: llvm/test/Transforms/MemCpyOpt/callslot.ll
===================================================================
--- llvm/test/Transforms/MemCpyOpt/callslot.ll
+++ llvm/test/Transforms/MemCpyOpt/callslot.ll
@@ -211,6 +211,19 @@
   ret void
 }
 
+define void @alignment(ptr align 1 %dest) {
+; CHECK-LABEL: @alignment(ptr align 1 %dest) {
+; CHECK-NEXT:    [[SRC:%.*]] = alloca [16 x i8], align 4
+; CHECK-NEXT:    call void @llvm.memset.p0.i64(ptr align 4 [[SRC]], i8 0, i64 16, i1 false)
+; CHECK-NEXT:    call void @llvm.memset.p0.i64(ptr align 1 %dest, i8 0, i64 16, i1 false)
+; CHECK-NEXT:    ret void
+
+  %src = alloca [16 x i8], align 4
+  call void @llvm.memset.p0.i64(ptr align 4 %src, i8 0, i64 16, i1 false)
+  call void @llvm.memcpy.p0.p0.i64(ptr align 1 %dest, ptr align 4 %src, i64 16, i1 false)
+  ret void
+}
+
 declare void @may_throw()
 declare void @accept_ptr(ptr)
 declare void @llvm.memcpy.p0.p0.i64(ptr, ptr, i64, i1)
Index: llvm/lib/Transforms/Scalar/MemCpyOptimizer.cpp
===================================================================
--- llvm/lib/Transforms/Scalar/MemCpyOptimizer.cpp
+++ llvm/lib/Transforms/Scalar/MemCpyOptimizer.cpp
@@ -873,7 +873,7 @@
 bool MemCpyOptPass::performCallSlotOptzn(Instruction *cpyLoad,
                                          Instruction *cpyStore, Value *cpyDest,
                                          Value *cpySrc, TypeSize cpySize,
-                                         Align cpyAlign, BatchAAResults &BAA,
+                                         Align cpyDestAlign, BatchAAResults &BAA,
                                          std::function<CallInst *()> GetC) {
   // The general transformation to keep in mind is
   //
@@ -978,7 +978,7 @@
 
   // Check that dest points to memory that is at least as aligned as src.
   Align srcAlign = srcAlloca->getAlign();
-  bool isDestSufficientlyAligned = srcAlign <= cpyAlign;
+  bool isDestSufficientlyAligned = srcAlign <= cpyDestAlign;
   // If dest is not aligned enough and we can't increase its alignment then
   // bail out.
   if (!isDestSufficientlyAligned && !isa<AllocaInst>(cpyDest)) {
@@ -1503,11 +1503,9 @@
         if (auto *C = dyn_cast<CallInst>(MI)) {
           // FIXME: Can we pass in either of dest/src alignment here instead
           // of conservatively taking the minimum?
-          Align Alignment = std::min(M->getDestAlign().valueOrOne(),
-                                     M->getSourceAlign().valueOrOne());
           if (performCallSlotOptzn(M, M, M->getDest(), M->getSource(),
                                    TypeSize::getFixed(CopySize->getZExtValue()),
-                                   Alignment, BAA,
+                                   M->getDestAlign().valueOrOne(), BAA,
                                    [C]() -> CallInst * { return C; })) {
             LLVM_DEBUG(dbgs() << "Performed call slot optimization:\n"
                               << "    call: " << *C << "\n"


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D140603.485029.patch
Type: text/x-patch
Size: 2959 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20221223/913127d9/attachment.bin>