From f01b9aa89e8cd5d3cca0e13835302f69c1f879d2 Mon Sep 17 00:00:00 2001 From: David Green Date: Tue, 5 Nov 2019 11:54:22 +0000 Subject: [PATCH] [MachineScheduler] Enable AA in PostRA Machine scheduler This adds AA to Post-RA Machine Scheduling, allowing the pass more freedom when handling memory operations. My understanding is that this was just never done, not that it is inherently incorrect to do so. The older PostRA List scheduler already makes use of AA, it's just that the MI PostRA Scheduler was never taught to use it. Differential Revision: https://reviews.llvm.org/D69814 --- llvm/lib/CodeGen/MachineScheduler.cpp | 2 ++ .../test/CodeGen/AArch64/merge-store-dependency.ll | 6 ++-- llvm/test/CodeGen/PowerPC/extract-and-store.ll | 24 ++++++++-------- llvm/test/CodeGen/PowerPC/f128-aggregates.ll | 4 +-- .../CodeGen/PowerPC/vec_conv_fp32_to_i64_elts.ll | 32 +++++++++++----------- .../CodeGen/PowerPC/vec_conv_i16_to_fp64_elts.ll | 8 +++--- .../CodeGen/PowerPC/vec_conv_i64_to_fp32_elts.ll | 24 ++++++++-------- 7 files changed, 51 insertions(+), 49 deletions(-) diff --git a/llvm/lib/CodeGen/MachineScheduler.cpp b/llvm/lib/CodeGen/MachineScheduler.cpp index caebb95..df88a87 100644 --- a/llvm/lib/CodeGen/MachineScheduler.cpp +++ b/llvm/lib/CodeGen/MachineScheduler.cpp @@ -238,6 +238,7 @@ void PostMachineScheduler::getAnalysisUsage(AnalysisUsage &AU) const { AU.setPreservesCFG(); AU.addRequired(); AU.addRequired(); + AU.addRequired(); AU.addRequired(); MachineFunctionPass::getAnalysisUsage(AU); } @@ -412,6 +413,7 @@ bool PostMachineScheduler::runOnMachineFunction(MachineFunction &mf) { MF = &mf; MLI = &getAnalysis(); PassConfig = &getAnalysis(); + AA = &getAnalysis().getAAResults(); if (VerifyScheduling) MF->verify(this, "Before post machine scheduling."); diff --git a/llvm/test/CodeGen/AArch64/merge-store-dependency.ll b/llvm/test/CodeGen/AArch64/merge-store-dependency.ll index 4c561f3..5613db1 100644 --- a/llvm/test/CodeGen/AArch64/merge-store-dependency.ll +++ b/llvm/test/CodeGen/AArch64/merge-store-dependency.ll @@ -17,15 +17,15 @@ define void @test(%struct1* %fde, i32 %fd, void (i32, i32, i8*)* %func, i8* %arg ; A53-NEXT: movi v0.2d, #0000000000000000 ; A53-NEXT: mov x8, x0 ; A53-NEXT: mov x19, x8 -; A53-NEXT: mov w9, #256 ; A53-NEXT: mov w0, w1 -; A53-NEXT: str q0, [x8] +; A53-NEXT: mov w9, #256 ; A53-NEXT: str q0, [x19, #16]! -; A53-NEXT: strh w9, [x8, #24] ; A53-NEXT: str w1, [x19] ; A53-NEXT: mov w1, #4 ; A53-NEXT: stp x2, x3, [x8, #32] ; A53-NEXT: mov x2, x8 +; A53-NEXT: str q0, [x8] +; A53-NEXT: strh w9, [x8, #24] ; A53-NEXT: str wzr, [x8, #20] ; A53-NEXT: bl fcntl ; A53-NEXT: adrp x9, gv0 diff --git a/llvm/test/CodeGen/PowerPC/extract-and-store.ll b/llvm/test/CodeGen/PowerPC/extract-and-store.ll index 9a5bacd..2731ffd 100644 --- a/llvm/test/CodeGen/PowerPC/extract-and-store.ll +++ b/llvm/test/CodeGen/PowerPC/extract-and-store.ll @@ -484,8 +484,8 @@ define dso_local void @test_consecutive_i32(<4 x i32> %a, i32* nocapture %b) loc ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxsldwi vs0, vs34, vs34, 2 ; CHECK-NEXT: li r3, 4 -; CHECK-NEXT: stfiwx f0, 0, r5 ; CHECK-NEXT: stxsiwx vs34, r5, r3 +; CHECK-NEXT: stfiwx f0, 0, r5 ; CHECK-NEXT: blr ; ; CHECK-BE-LABEL: test_consecutive_i32: @@ -501,8 +501,8 @@ define dso_local void @test_consecutive_i32(<4 x i32> %a, i32* nocapture %b) loc ; CHECK-P9: # %bb.0: # %entry ; CHECK-P9-NEXT: xxsldwi vs0, vs34, vs34, 2 ; CHECK-P9-NEXT: li r3, 4 -; CHECK-P9-NEXT: stfiwx f0, 0, r5 ; CHECK-P9-NEXT: stxsiwx vs34, r5, r3 +; CHECK-P9-NEXT: stfiwx f0, 0, r5 ; CHECK-P9-NEXT: blr ; ; CHECK-P9-BE-LABEL: test_consecutive_i32: @@ -590,8 +590,8 @@ define dso_local void @test_stores_exceed_vec_size(<4 x i32> %a, i32* nocapture ; CHECK-BE-NEXT: li r4, 20 ; CHECK-BE-NEXT: stxsiwx vs34, r5, r3 ; CHECK-BE-NEXT: xxsldwi vs0, vs34, vs0, 2 -; CHECK-BE-NEXT: stxvw4x vs0, 0, r5 ; CHECK-BE-NEXT: stfiwx f1, r5, r4 +; CHECK-BE-NEXT: stxvw4x vs0, 0, r5 ; CHECK-BE-NEXT: blr ; ; CHECK-P9-LABEL: test_stores_exceed_vec_size: @@ -599,13 +599,13 @@ define dso_local void @test_stores_exceed_vec_size(<4 x i32> %a, i32* nocapture ; CHECK-P9-NEXT: addis r3, r2, .LCPI16_0@toc@ha ; CHECK-P9-NEXT: addi r3, r3, .LCPI16_0@toc@l ; CHECK-P9-NEXT: lxvx vs35, 0, r3 -; CHECK-P9-NEXT: li r3, 16 -; CHECK-P9-NEXT: vperm v3, v2, v2, v3 ; CHECK-P9-NEXT: xxsldwi vs0, vs34, vs34, 1 -; CHECK-P9-NEXT: stxv vs35, 0(r5) +; CHECK-P9-NEXT: li r3, 16 ; CHECK-P9-NEXT: stfiwx f0, r5, r3 ; CHECK-P9-NEXT: li r3, 20 ; CHECK-P9-NEXT: stxsiwx vs34, r5, r3 +; CHECK-P9-NEXT: vperm v3, v2, v2, v3 +; CHECK-P9-NEXT: stxv vs35, 0(r5) ; CHECK-P9-NEXT: blr ; ; CHECK-P9-BE-LABEL: test_stores_exceed_vec_size: @@ -613,10 +613,10 @@ define dso_local void @test_stores_exceed_vec_size(<4 x i32> %a, i32* nocapture ; CHECK-P9-BE-NEXT: xxspltw vs0, vs34, 0 ; CHECK-P9-BE-NEXT: xxsldwi vs0, vs34, vs0, 2 ; CHECK-P9-BE-NEXT: li r3, 16 -; CHECK-P9-BE-NEXT: stxv vs0, 0(r5) -; CHECK-P9-BE-NEXT: xxsldwi vs0, vs34, vs34, 1 ; CHECK-P9-BE-NEXT: stxsiwx vs34, r5, r3 ; CHECK-P9-BE-NEXT: li r3, 20 +; CHECK-P9-BE-NEXT: stxv vs0, 0(r5) +; CHECK-P9-BE-NEXT: xxsldwi vs0, vs34, vs34, 1 ; CHECK-P9-BE-NEXT: stfiwx f0, r5, r3 ; CHECK-P9-BE-NEXT: blr entry: @@ -930,8 +930,8 @@ define void @test_elements_from_two_vec(<4 x i32> %a, <4 x i32> %b, i32* nocaptu ; CHECK-BE: # %bb.0: # %entry ; CHECK-BE-NEXT: xxsldwi vs0, vs34, vs34, 3 ; CHECK-BE-NEXT: li r3, 4 -; CHECK-BE-NEXT: stfiwx f0, r7, r3 ; CHECK-BE-NEXT: stxsiwx vs35, 0, r7 +; CHECK-BE-NEXT: stfiwx f0, r7, r3 ; CHECK-BE-NEXT: blr ; ; CHECK-P9-LABEL: test_elements_from_two_vec: @@ -977,19 +977,19 @@ define dso_local void @test_elements_from_three_vec(<4 x float> %a, <4 x float> ; CHECK-BE-NEXT: xxsldwi vs1, vs35, vs35, 1 ; CHECK-BE-NEXT: li r3, 4 ; CHECK-BE-NEXT: li r4, 8 +; CHECK-BE-NEXT: stxsiwx vs36, r9, r4 ; CHECK-BE-NEXT: stfiwx f1, r9, r3 ; CHECK-BE-NEXT: stfiwx f0, 0, r9 -; CHECK-BE-NEXT: stxsiwx vs36, r9, r4 ; CHECK-BE-NEXT: blr ; ; CHECK-P9-LABEL: test_elements_from_three_vec: ; CHECK-P9: # %bb.0: # %entry ; CHECK-P9-NEXT: xxsldwi vs0, vs34, vs34, 3 ; CHECK-P9-NEXT: li r3, 4 -; CHECK-P9-NEXT: stfiwx f0, 0, r9 -; CHECK-P9-NEXT: xxsldwi vs0, vs36, vs36, 1 ; CHECK-P9-NEXT: stxsiwx vs35, r9, r3 ; CHECK-P9-NEXT: li r3, 8 +; CHECK-P9-NEXT: stfiwx f0, 0, r9 +; CHECK-P9-NEXT: xxsldwi vs0, vs36, vs36, 1 ; CHECK-P9-NEXT: stfiwx f0, r9, r3 ; CHECK-P9-NEXT: blr ; diff --git a/llvm/test/CodeGen/PowerPC/f128-aggregates.ll b/llvm/test/CodeGen/PowerPC/f128-aggregates.ll index 8c21b85..6e782c2 100644 --- a/llvm/test/CodeGen/PowerPC/f128-aggregates.ll +++ b/llvm/test/CodeGen/PowerPC/f128-aggregates.ll @@ -353,10 +353,10 @@ define fp128 @sum_float128(i32 signext %count, ...) { ; CHECK-NEXT: addi r3, r1, 40 ; CHECK-NEXT: lxvx v3, 0, r3 ; CHECK-NEXT: xsaddqp v2, v3, v2 -; CHECK-NEXT: addi [[REG2:r[0-9]+]], r1, 72 -; CHECK-NEXT: std [[REG2]], -8(r1) ; CHECK-NEXT: lxv v3, 16(r3) ; CHECK-NEXT: xsaddqp v2, v2, v3 +; CHECK-NEXT: addi [[REG2:r[0-9]+]], r1, 72 +; CHECK-NEXT: std [[REG2]], -8(r1) ; CHECK-NEXT: blr entry: %ap = alloca i8*, align 8 diff --git a/llvm/test/CodeGen/PowerPC/vec_conv_fp32_to_i64_elts.ll b/llvm/test/CodeGen/PowerPC/vec_conv_fp32_to_i64_elts.ll index 6b945d4..cf4a6d6 100644 --- a/llvm/test/CodeGen/PowerPC/vec_conv_fp32_to_i64_elts.ll +++ b/llvm/test/CodeGen/PowerPC/vec_conv_fp32_to_i64_elts.ll @@ -328,27 +328,27 @@ define void @test16elt(<16 x i64>* noalias nocapture sret %agg.result, <16 x flo ; CHECK-P9-NEXT: lxv vs2, 48(r4) ; CHECK-P9-NEXT: xxswapd vs8, vs2 ; CHECK-P9-NEXT: xscvspdpn f8, vs8 -; CHECK-P9-NEXT: xvcvdpuxds vs0, vs0 -; CHECK-P9-NEXT: stxv vs5, 32(r3) -; CHECK-P9-NEXT: xvcvdpuxds vs6, vs6 +; CHECK-P9-NEXT: xvcvdpuxds vs4, vs4 ; CHECK-P9-NEXT: xscvspdpn f3, vs3 ; CHECK-P9-NEXT: xxmrghd vs3, vs7, vs3 ; CHECK-P9-NEXT: xxsldwi vs7, vs2, vs2, 3 +; CHECK-P9-NEXT: xvcvdpuxds vs0, vs0 +; CHECK-P9-NEXT: xvcvdpuxds vs6, vs6 +; CHECK-P9-NEXT: stxv vs6, 64(r3) ; CHECK-P9-NEXT: xscvspdpn f7, vs7 ; CHECK-P9-NEXT: xxmrghd vs7, vs8, vs7 ; CHECK-P9-NEXT: xscvspdpn f8, vs2 ; CHECK-P9-NEXT: xxsldwi vs2, vs2, vs2, 1 -; CHECK-P9-NEXT: stxv vs6, 64(r3) -; CHECK-P9-NEXT: xvcvdpuxds vs4, vs4 +; CHECK-P9-NEXT: xscvspdpn f2, vs2 +; CHECK-P9-NEXT: xxmrghd vs2, vs8, vs2 ; CHECK-P9-NEXT: xvcvdpuxds vs3, vs3 ; CHECK-P9-NEXT: xvcvdpuxds vs7, vs7 -; CHECK-P9-NEXT: xscvspdpn f2, vs2 ; CHECK-P9-NEXT: stxv vs3, 80(r3) -; CHECK-P9-NEXT: xxmrghd vs2, vs8, vs2 ; CHECK-P9-NEXT: xvcvdpuxds vs2, vs2 -; CHECK-P9-NEXT: stxv vs2, 112(r3) ; CHECK-P9-NEXT: stxv vs7, 96(r3) +; CHECK-P9-NEXT: stxv vs2, 112(r3) ; CHECK-P9-NEXT: stxv vs4, 48(r3) +; CHECK-P9-NEXT: stxv vs5, 32(r3) ; CHECK-P9-NEXT: stxv vs0, 16(r3) ; CHECK-P9-NEXT: stxv vs1, 0(r3) ; CHECK-P9-NEXT: blr @@ -738,27 +738,27 @@ define void @test16elt_signed(<16 x i64>* noalias nocapture sret %agg.result, <1 ; CHECK-P9-NEXT: lxv vs2, 48(r4) ; CHECK-P9-NEXT: xxswapd vs8, vs2 ; CHECK-P9-NEXT: xscvspdpn f8, vs8 -; CHECK-P9-NEXT: xvcvdpuxds vs0, vs0 -; CHECK-P9-NEXT: stxv vs5, 32(r3) -; CHECK-P9-NEXT: xvcvdpuxds vs6, vs6 +; CHECK-P9-NEXT: xvcvdpuxds vs4, vs4 ; CHECK-P9-NEXT: xscvspdpn f3, vs3 ; CHECK-P9-NEXT: xxmrghd vs3, vs7, vs3 ; CHECK-P9-NEXT: xxsldwi vs7, vs2, vs2, 3 +; CHECK-P9-NEXT: xvcvdpuxds vs0, vs0 +; CHECK-P9-NEXT: xvcvdpuxds vs6, vs6 +; CHECK-P9-NEXT: stxv vs6, 64(r3) ; CHECK-P9-NEXT: xscvspdpn f7, vs7 ; CHECK-P9-NEXT: xxmrghd vs7, vs8, vs7 ; CHECK-P9-NEXT: xscvspdpn f8, vs2 ; CHECK-P9-NEXT: xxsldwi vs2, vs2, vs2, 1 -; CHECK-P9-NEXT: stxv vs6, 64(r3) -; CHECK-P9-NEXT: xvcvdpuxds vs4, vs4 +; CHECK-P9-NEXT: xscvspdpn f2, vs2 +; CHECK-P9-NEXT: xxmrghd vs2, vs8, vs2 ; CHECK-P9-NEXT: xvcvdpuxds vs3, vs3 ; CHECK-P9-NEXT: xvcvdpuxds vs7, vs7 -; CHECK-P9-NEXT: xscvspdpn f2, vs2 ; CHECK-P9-NEXT: stxv vs3, 80(r3) -; CHECK-P9-NEXT: xxmrghd vs2, vs8, vs2 ; CHECK-P9-NEXT: xvcvdpuxds vs2, vs2 -; CHECK-P9-NEXT: stxv vs2, 112(r3) ; CHECK-P9-NEXT: stxv vs7, 96(r3) +; CHECK-P9-NEXT: stxv vs2, 112(r3) ; CHECK-P9-NEXT: stxv vs4, 48(r3) +; CHECK-P9-NEXT: stxv vs5, 32(r3) ; CHECK-P9-NEXT: stxv vs0, 16(r3) ; CHECK-P9-NEXT: stxv vs1, 0(r3) ; CHECK-P9-NEXT: blr diff --git a/llvm/test/CodeGen/PowerPC/vec_conv_i16_to_fp64_elts.ll b/llvm/test/CodeGen/PowerPC/vec_conv_i16_to_fp64_elts.ll index f03d069..883cf7e5 100644 --- a/llvm/test/CodeGen/PowerPC/vec_conv_i16_to_fp64_elts.ll +++ b/llvm/test/CodeGen/PowerPC/vec_conv_i16_to_fp64_elts.ll @@ -712,24 +712,24 @@ define void @test16elt_signed(<16 x double>* noalias nocapture sret %agg.result, ; CHECK-P9-NEXT: vextsh2d v2, v2 ; CHECK-P9-NEXT: xvcvsxddp vs3, v2 ; CHECK-P9-NEXT: vperm v2, v4, v4, v3 +; CHECK-P9-NEXT: stxv vs2, 32(r3) ; CHECK-P9-NEXT: vextsh2d v2, v2 ; CHECK-P9-NEXT: stxv vs3, 48(r3) +; CHECK-P9-NEXT: stxv vs1, 16(r3) ; CHECK-P9-NEXT: xvcvsxddp vs4, v2 ; CHECK-P9-NEXT: vperm v2, v4, v4, v5 ; CHECK-P9-NEXT: vextsh2d v2, v2 ; CHECK-P9-NEXT: xvcvsxddp vs5, v2 ; CHECK-P9-NEXT: vperm v2, v4, v4, v0 ; CHECK-P9-NEXT: stxv vs4, 64(r3) -; CHECK-P9-NEXT: stxv vs5, 80(r3) ; CHECK-P9-NEXT: vextsh2d v2, v2 ; CHECK-P9-NEXT: xvcvsxddp vs6, v2 ; CHECK-P9-NEXT: vperm v2, v4, v4, v1 -; CHECK-P9-NEXT: vextsh2d v2, v2 +; CHECK-P9-NEXT: stxv vs5, 80(r3) ; CHECK-P9-NEXT: stxv vs6, 96(r3) +; CHECK-P9-NEXT: vextsh2d v2, v2 ; CHECK-P9-NEXT: xvcvsxddp vs7, v2 ; CHECK-P9-NEXT: stxv vs7, 112(r3) -; CHECK-P9-NEXT: stxv vs2, 32(r3) -; CHECK-P9-NEXT: stxv vs1, 16(r3) ; CHECK-P9-NEXT: stxv vs0, 0(r3) ; CHECK-P9-NEXT: blr ; diff --git a/llvm/test/CodeGen/PowerPC/vec_conv_i64_to_fp32_elts.ll b/llvm/test/CodeGen/PowerPC/vec_conv_i64_to_fp32_elts.ll index 08e6f70..66e85e9 100644 --- a/llvm/test/CodeGen/PowerPC/vec_conv_i64_to_fp32_elts.ll +++ b/llvm/test/CodeGen/PowerPC/vec_conv_i64_to_fp32_elts.ll @@ -239,18 +239,18 @@ define void @test16elt(<16 x float>* noalias nocapture sret %agg.result, <16 x i ; CHECK-P9-NEXT: xxsldwi v6, vs0, vs0, 3 ; CHECK-P9-NEXT: xvcvuxdsp vs0, v0 ; CHECK-P9-NEXT: lxv v5, 64(r4) +; CHECK-P9-NEXT: stxv v1, 0(r3) ; CHECK-P9-NEXT: xxsldwi v0, vs0, vs0, 3 ; CHECK-P9-NEXT: xvcvuxdsp vs0, v5 ; CHECK-P9-NEXT: lxv v4, 80(r4) ; CHECK-P9-NEXT: vpkudum v0, v0, v6 +; CHECK-P9-NEXT: stxv v0, 16(r3) ; CHECK-P9-NEXT: xxsldwi v5, vs0, vs0, 3 -; CHECK-P9-NEXT: lxv v3, 96(r4) ; CHECK-P9-NEXT: xvcvuxdsp vs0, v4 +; CHECK-P9-NEXT: lxv v3, 96(r4) ; CHECK-P9-NEXT: xxsldwi v4, vs0, vs0, 3 ; CHECK-P9-NEXT: xvcvuxdsp vs0, v3 ; CHECK-P9-NEXT: lxv v2, 112(r4) -; CHECK-P9-NEXT: stxv v0, 16(r3) -; CHECK-P9-NEXT: stxv v1, 0(r3) ; CHECK-P9-NEXT: vpkudum v4, v4, v5 ; CHECK-P9-NEXT: stxv v4, 32(r3) ; CHECK-P9-NEXT: xxsldwi v3, vs0, vs0, 3 @@ -275,18 +275,18 @@ define void @test16elt(<16 x float>* noalias nocapture sret %agg.result, <16 x i ; CHECK-BE-NEXT: xxsldwi v6, vs0, vs0, 3 ; CHECK-BE-NEXT: xvcvuxdsp vs0, v0 ; CHECK-BE-NEXT: lxv v5, 80(r4) +; CHECK-BE-NEXT: stxv v1, 0(r3) ; CHECK-BE-NEXT: xxsldwi v0, vs0, vs0, 3 ; CHECK-BE-NEXT: xvcvuxdsp vs0, v5 ; CHECK-BE-NEXT: lxv v4, 64(r4) ; CHECK-BE-NEXT: vpkudum v0, v0, v6 +; CHECK-BE-NEXT: stxv v0, 16(r3) ; CHECK-BE-NEXT: xxsldwi v5, vs0, vs0, 3 -; CHECK-BE-NEXT: lxv v3, 112(r4) ; CHECK-BE-NEXT: xvcvuxdsp vs0, v4 +; CHECK-BE-NEXT: lxv v3, 112(r4) ; CHECK-BE-NEXT: xxsldwi v4, vs0, vs0, 3 ; CHECK-BE-NEXT: xvcvuxdsp vs0, v3 ; CHECK-BE-NEXT: lxv v2, 96(r4) -; CHECK-BE-NEXT: stxv v0, 16(r3) -; CHECK-BE-NEXT: stxv v1, 0(r3) ; CHECK-BE-NEXT: vpkudum v4, v4, v5 ; CHECK-BE-NEXT: stxv v4, 32(r3) ; CHECK-BE-NEXT: xxsldwi v3, vs0, vs0, 3 @@ -532,18 +532,18 @@ define void @test16elt_signed(<16 x float>* noalias nocapture sret %agg.result, ; CHECK-P9-NEXT: xxsldwi v6, vs0, vs0, 3 ; CHECK-P9-NEXT: xvcvsxdsp vs0, v0 ; CHECK-P9-NEXT: lxv v5, 64(r4) +; CHECK-P9-NEXT: stxv v1, 0(r3) ; CHECK-P9-NEXT: xxsldwi v0, vs0, vs0, 3 ; CHECK-P9-NEXT: xvcvsxdsp vs0, v5 ; CHECK-P9-NEXT: lxv v4, 80(r4) ; CHECK-P9-NEXT: vpkudum v0, v0, v6 +; CHECK-P9-NEXT: stxv v0, 16(r3) ; CHECK-P9-NEXT: xxsldwi v5, vs0, vs0, 3 -; CHECK-P9-NEXT: lxv v3, 96(r4) ; CHECK-P9-NEXT: xvcvsxdsp vs0, v4 +; CHECK-P9-NEXT: lxv v3, 96(r4) ; CHECK-P9-NEXT: xxsldwi v4, vs0, vs0, 3 ; CHECK-P9-NEXT: xvcvsxdsp vs0, v3 ; CHECK-P9-NEXT: lxv v2, 112(r4) -; CHECK-P9-NEXT: stxv v0, 16(r3) -; CHECK-P9-NEXT: stxv v1, 0(r3) ; CHECK-P9-NEXT: vpkudum v4, v4, v5 ; CHECK-P9-NEXT: stxv v4, 32(r3) ; CHECK-P9-NEXT: xxsldwi v3, vs0, vs0, 3 @@ -568,18 +568,18 @@ define void @test16elt_signed(<16 x float>* noalias nocapture sret %agg.result, ; CHECK-BE-NEXT: xxsldwi v6, vs0, vs0, 3 ; CHECK-BE-NEXT: xvcvsxdsp vs0, v0 ; CHECK-BE-NEXT: lxv v5, 80(r4) +; CHECK-BE-NEXT: stxv v1, 0(r3) ; CHECK-BE-NEXT: xxsldwi v0, vs0, vs0, 3 ; CHECK-BE-NEXT: xvcvsxdsp vs0, v5 ; CHECK-BE-NEXT: lxv v4, 64(r4) ; CHECK-BE-NEXT: vpkudum v0, v0, v6 +; CHECK-BE-NEXT: stxv v0, 16(r3) ; CHECK-BE-NEXT: xxsldwi v5, vs0, vs0, 3 -; CHECK-BE-NEXT: lxv v3, 112(r4) ; CHECK-BE-NEXT: xvcvsxdsp vs0, v4 +; CHECK-BE-NEXT: lxv v3, 112(r4) ; CHECK-BE-NEXT: xxsldwi v4, vs0, vs0, 3 ; CHECK-BE-NEXT: xvcvsxdsp vs0, v3 ; CHECK-BE-NEXT: lxv v2, 96(r4) -; CHECK-BE-NEXT: stxv v0, 16(r3) -; CHECK-BE-NEXT: stxv v1, 0(r3) ; CHECK-BE-NEXT: vpkudum v4, v4, v5 ; CHECK-BE-NEXT: stxv v4, 32(r3) ; CHECK-BE-NEXT: xxsldwi v3, vs0, vs0, 3 -- 2.7.4