[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.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@250324 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/PowerPC/PPCISelDAGToDAG.cpp b/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
index 4fd57be8c9db2417f2c2ecd94583fa47f885c555..e876a186f866de675b4c48ccf9b8f14dc9aea08d 100644 (file)
--- a/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
+++ b/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
@@ -2799,6 +2799,8 @@ SDNode *PPCDAGToDAGISel::Select(SDNode *N) {
         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 };

diff --git a/test/CodeGen/PowerPC/pr25157.ll b/test/CodeGen/PowerPC/pr25157.ll
new file mode 100644 (file)
index 0000000..7137d67
--- /dev/null
+++ b/test/CodeGen/PowerPC/pr25157.ll
@@ -0,0 +1,58 @@
+; RUN: llc -mcpu=pwr8 -mtriple=powerpc64le-unknown-linux-gnu < %s | FileCheck %s
+
+; Verify correct generation of an lxsspx rather than an invalid optimization
+; to lxvdsx.  Bugpoint-reduced test from Eric Schweitz.
+
+%struct.BSS38.51.4488.9911.14348.16813.20264.24701.28152.31603.35054.39491.44914.45407.46393.46886.47872.49351.49844.50830.51323.52309.53295.53788.54281.55267.55760.59211.61625 = type <{ [28 x i8] }>
+%struct_main1_2_.491.4928.10351.14788.17253.20704.25141.28592.32043.35494.39931.45354.45847.46833.47326.48312.49791.50284.51270.51763.52749.53735.54228.54721.55707.56200.59651.61626 = type <{ [64 x i8] }>
+
+@.BSS38 = external global %struct.BSS38.51.4488.9911.14348.16813.20264.24701.28152.31603.35054.39491.44914.45407.46393.46886.47872.49351.49844.50830.51323.52309.53295.53788.54281.55267.55760.59211.61625, align 32
+@_main1_2_ = external global %struct_main1_2_.491.4928.10351.14788.17253.20704.25141.28592.32043.35494.39931.45354.45847.46833.47326.48312.49791.50284.51270.51763.52749.53735.54228.54721.55707.56200.59651.61626, section ".comm", align 16
+
+define void @aercalc_() {
+L.entry:
+  br i1 undef, label %L.LB38_2426, label %L.LB38_2911
+
+L.LB38_2911:
+  br i1 undef, label %L.LB38_2140, label %L.LB38_2640
+
+L.LB38_2640:
+  unreachable
+
+L.LB38_2426:
+  br i1 undef, label %L.LB38_2438, label %L.LB38_2920
+
+L.LB38_2920:
+  br i1 undef, label %L.LB38_2438, label %L.LB38_2921
+
+L.LB38_2921:
+  br label %L.LB38_2140
+
+L.LB38_2140:
+  ret void
+
+L.LB38_2438:
+  br i1 undef, label %L.LB38_2451, label %L.LB38_2935
+
+L.LB38_2935:
+  br i1 undef, label %L.LB38_2451, label %L.LB38_2936
+
+L.LB38_2936:
+  unreachable
+
+L.LB38_2451:
+  br i1 undef, label %L.LB38_2452, label %L.LB38_2937
+
+L.LB38_2937:
+  unreachable
+
+L.LB38_2452:
+  %0 = load float, float* bitcast (i8* getelementptr inbounds (%struct.BSS38.51.4488.9911.14348.16813.20264.24701.28152.31603.35054.39491.44914.45407.46393.46886.47872.49351.49844.50830.51323.52309.53295.53788.54281.55267.55760.59211.61625, %struct.BSS38.51.4488.9911.14348.16813.20264.24701.28152.31603.35054.39491.44914.45407.46393.46886.47872.49351.49844.50830.51323.52309.53295.53788.54281.55267.55760.59211.61625* @.BSS38, i64 0, i32 0, i64 16) to float*), align 16
+  %1 = fpext float %0 to double
+  %2 = insertelement <2 x double> undef, double %1, i32 1
+  store <2 x double> %2, <2 x double>* bitcast (i8* getelementptr inbounds (%struct_main1_2_.491.4928.10351.14788.17253.20704.25141.28592.32043.35494.39931.45354.45847.46833.47326.48312.49791.50284.51270.51763.52749.53735.54228.54721.55707.56200.59651.61626, %struct_main1_2_.491.4928.10351.14788.17253.20704.25141.28592.32043.35494.39931.45354.45847.46833.47326.48312.49791.50284.51270.51763.52749.53735.54228.54721.55707.56200.59651.61626* @_main1_2_, i64 0, i32 0, i64 32) to <2 x double>*), align 16
+  unreachable
+}
+
+; CHECK-LABEL: @aercalc_
+; CHECK: lxsspx