[PATCH] D60801: [X86] Disable argument copy elision for arguments passed via pointers

[Craig Topper via Phabricator via llvm-commits]

craig.topper created this revision.
craig.topper added a reviewer: rnk.
Herald added a subscriber: hiraditya.
Herald added a project: LLVM.

If you pass two 1024 bit vectors in IR with AVX2 on Windows 64. Both vectors will be split in four 256 bit pieces. The four pieces of the first argument will be passed indirectly using 4 gprs. The second argument will get passed via pointers in memory.

The PartOffsets stored for the second argument are all in terms of its original 1024 bit size. So the PartOffsets for each piece are 32 bytes apart. So if we consider it for copy elision we'll only load an 8 byte pointer, but we'll move the address 32 bytes. The stack object size we create for the first part is probably wrong too.

This issue was encountered by ISPC. I'm working on getting a reduce test case, but wanted to go ahead and get feedback on the fix.


https://reviews.llvm.org/D60801

Files:
  llvm/lib/Target/X86/X86ISelLowering.cpp


Index: llvm/lib/Target/X86/X86ISelLowering.cpp
===================================================================
--- llvm/lib/Target/X86/X86ISelLowering.cpp
+++ llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -3011,7 +3011,9 @@
   }
 
   // This is an argument in memory. We might be able to perform copy elision.
-  if (Flags.isCopyElisionCandidate()) {
+  // Unless its passed by pointer.
+  if (Flags.isCopyElisionCandidate() &&
+      VA.getLocInfo() != CCValAssign::Indirect && !ExtendedInMem) {
     EVT ArgVT = Ins[i].ArgVT;
     SDValue PartAddr;
     if (Ins[i].PartOffset == 0) {


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D60801.195474.patch
Type: text/x-patch
Size: 587 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190416/5f47d9f1/attachment.bin>