commit 048cc97fb1900dc2143753f851c742ae30ece4bb
author Bill Schmidt <wschmidt@linux.vnet.ibm.com> Wed Oct 14 20:45:00 2015 +0000
committer Bill Schmidt <wschmidt@linux.vnet.ibm.com> Wed Oct 14 20:45:00 2015 +0000
tree 1f8ba5f417a3468204acf4446376c6ed7560a437
parent 147bb89d520c0868d0a857e37c714de73558c3cc

[PowerPC] Fix invalid lxvdsx optimization (PR25157)

PR25157 identifies a bug where a load plus a vector shuffle is
incorrectly converted into an LXVDSX instruction.  That optimization
is only valid if the load is of a doubleword, and in the noted case,
it was not.  This corrects that problem.

Joint patch with Eric Schweitz, who provided the bugpoint-reduced test
case.

llvm-svn: 250324

llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp

diff --git a/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp b/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
index 4fd57be..e876a18 100644
--- a/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
+++ b/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
@@ -2799,6 +2799,8 @@
         SDValue Base, Offset;
 
         if (LD->isUnindexed() &&
+            (LD->getMemoryVT() == MVT::f64 ||
+             LD->getMemoryVT() == MVT::i64) &&
             SelectAddrIdxOnly(LD->getBasePtr(), Base, Offset)) {
           SDValue Chain = LD->getChain();
           SDValue Ops[] = { Base, Offset, Chain };