From 7e2e5c3c581521203d65bd224aec089de00a1f61 Mon Sep 17 00:00:00 2001 From: Daniel Neilson Date: Wed, 11 Apr 2018 16:50:04 +0000 Subject: [PATCH] [DSE] Regenerate tests with update_test_checks.py (NFC) Summary: In preparation for a future commit, this regenerates the test checks for test/Transforms/DeadStoreElimination/simple.ll test/Transforms/DeadStoreElimination/memintrinsics.ll llvm-svn: 329824 --- .../DeadStoreElimination/memintrinsics.ll | 13 +- .../test/Transforms/DeadStoreElimination/simple.ll | 470 +++++++++++++-------- 2 files changed, 308 insertions(+), 175 deletions(-) diff --git a/llvm/test/Transforms/DeadStoreElimination/memintrinsics.ll b/llvm/test/Transforms/DeadStoreElimination/memintrinsics.ll index aa9f1a1..5c6af10 100644 --- a/llvm/test/Transforms/DeadStoreElimination/memintrinsics.ll +++ b/llvm/test/Transforms/DeadStoreElimination/memintrinsics.ll @@ -1,3 +1,4 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py ; RUN: opt -S -dse < %s | FileCheck %s declare void @llvm.memcpy.p0i8.p0i8.i8(i8* nocapture, i8* nocapture, i8, i1) nounwind @@ -6,42 +7,42 @@ declare void @llvm.memset.p0i8.i8(i8* nocapture, i8, i8, i1) nounwind define void @test1() { ; CHECK-LABEL: @test1( +; CHECK-NEXT: ret void +; %A = alloca i8 %B = alloca i8 store i8 0, i8* %A ;; Written to by memcpy -; CHECK-NOT: store call void @llvm.memcpy.p0i8.p0i8.i8(i8* %A, i8* %B, i8 -1, i1 false) ret void -; CHECK: ret void } define void @test2() { ; CHECK-LABEL: @test2( +; CHECK-NEXT: ret void +; %A = alloca i8 %B = alloca i8 store i8 0, i8* %A ;; Written to by memmove -; CHECK-NOT: store call void @llvm.memmove.p0i8.p0i8.i8(i8* %A, i8* %B, i8 -1, i1 false) ret void -; CHECK: ret void } define void @test3() { ; CHECK-LABEL: @test3( +; CHECK-NEXT: ret void +; %A = alloca i8 %B = alloca i8 store i8 0, i8* %A ;; Written to by memset -; CHECK-NOT: store call void @llvm.memset.p0i8.i8(i8* %A, i8 0, i8 -1, i1 false) ret void -; CHECK: ret void } diff --git a/llvm/test/Transforms/DeadStoreElimination/simple.ll b/llvm/test/Transforms/DeadStoreElimination/simple.ll index 6046211..c842190 100644 --- a/llvm/test/Transforms/DeadStoreElimination/simple.ll +++ b/llvm/test/Transforms/DeadStoreElimination/simple.ll @@ -1,3 +1,4 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py ; RUN: opt < %s -basicaa -dse -S | FileCheck %s ; RUN: opt < %s -aa-pipeline=basic-aa -passes=dse -S | FileCheck %s target datalayout = "E-p:64:64:64-a0:0:8-f32:32:32-f64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-v64:64:64-v128:128:128" @@ -7,23 +8,27 @@ declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) n declare void @llvm.init.trampoline(i8*, i8*, i8*) define void @test1(i32* %Q, i32* %P) { - %DEAD = load i32, i32* %Q - store i32 %DEAD, i32* %P - store i32 0, i32* %P - ret void ; CHECK-LABEL: @test1( -; CHECK-NEXT: store i32 0, i32* %P -; CHECK-NEXT: ret void +; CHECK-NEXT: store i32 0, i32* [[P:%.*]] +; CHECK-NEXT: ret void +; + %DEAD = load i32, i32* %Q + store i32 %DEAD, i32* %P + store i32 0, i32* %P + ret void } ; PR8576 - Should delete store of 10 even though p/q are may aliases. define void @test2(i32 *%p, i32 *%q) { +; CHECK-LABEL: @test2( +; CHECK-NEXT: store i32 20, i32* [[Q:%.*]], align 4 +; CHECK-NEXT: store i32 30, i32* [[P:%.*]], align 4 +; CHECK-NEXT: ret void +; store i32 10, i32* %p, align 4 store i32 20, i32* %q, align 4 store i32 30, i32* %p, align 4 ret void -; CHECK-LABEL: @test2( -; CHECK-NEXT: store i32 20 } @@ -32,7 +37,12 @@ define void @test2(i32 *%p, i32 *%q) { define i32 @test3(i32* %g_addr) nounwind { ; CHECK-LABEL: @test3( -; CHECK: load i32, i32* %g_addr +; CHECK-NEXT: [[G_VALUE:%.*]] = load i32, i32* [[G_ADDR:%.*]], align 4 +; CHECK-NEXT: store i32 -1, i32* @g, align 4 +; CHECK-NEXT: store i32 [[G_VALUE]], i32* [[G_ADDR]], align 4 +; CHECK-NEXT: [[TMP3:%.*]] = load i32, i32* @g, align 4 +; CHECK-NEXT: ret i32 [[TMP3]] +; %g_value = load i32, i32* %g_addr, align 4 store i32 -1, i32* @g, align 4 store i32 %g_value, i32* %g_addr, align 4 @@ -42,87 +52,105 @@ define i32 @test3(i32* %g_addr) nounwind { define void @test4(i32* %Q) { - %a = load i32, i32* %Q - store volatile i32 %a, i32* %Q - ret void ; CHECK-LABEL: @test4( -; CHECK-NEXT: load i32 -; CHECK-NEXT: store volatile -; CHECK-NEXT: ret void +; CHECK-NEXT: [[A:%.*]] = load i32, i32* [[Q:%.*]] +; CHECK-NEXT: store volatile i32 [[A]], i32* [[Q]] +; CHECK-NEXT: ret void +; + %a = load i32, i32* %Q + store volatile i32 %a, i32* %Q + ret void } define void @test5(i32* %Q) { - %a = load volatile i32, i32* %Q - store i32 %a, i32* %Q - ret void ; CHECK-LABEL: @test5( -; CHECK-NEXT: load volatile -; CHECK-NEXT: ret void +; CHECK-NEXT: [[A:%.*]] = load volatile i32, i32* [[Q:%.*]] +; CHECK-NEXT: ret void +; + %a = load volatile i32, i32* %Q + store i32 %a, i32* %Q + ret void } ; Should delete store of 10 even though memset is a may-store to P (P and Q may ; alias). define void @test6(i32 *%p, i8 *%q) { +; CHECK-LABEL: @test6( +; CHECK-NEXT: call void @llvm.memset.p0i8.i64(i8* [[Q:%.*]], i8 42, i64 900, i1 false) +; CHECK-NEXT: store i32 30, i32* [[P:%.*]], align 4 +; CHECK-NEXT: ret void +; store i32 10, i32* %p, align 4 ;; dead. call void @llvm.memset.p0i8.i64(i8* %q, i8 42, i64 900, i1 false) store i32 30, i32* %p, align 4 ret void -; CHECK-LABEL: @test6( -; CHECK-NEXT: call void @llvm.memset } ; Should delete store of 10 even though memcpy is a may-store to P (P and Q may ; alias). define void @test7(i32 *%p, i8 *%q, i8* noalias %r) { +; CHECK-LABEL: @test7( +; CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[Q:%.*]], i8* [[R:%.*]], i64 900, i1 false) +; CHECK-NEXT: store i32 30, i32* [[P:%.*]], align 4 +; CHECK-NEXT: ret void +; store i32 10, i32* %p, align 4 ;; dead. call void @llvm.memcpy.p0i8.p0i8.i64(i8* %q, i8* %r, i64 900, i1 false) store i32 30, i32* %p, align 4 ret void -; CHECK-LABEL: @test7( -; CHECK-NEXT: call void @llvm.memcpy } ; Do not delete stores that are only partially killed. define i32 @test8() { - %V = alloca i32 - store i32 1234567, i32* %V - %V2 = bitcast i32* %V to i8* - store i8 0, i8* %V2 - %X = load i32, i32* %V - ret i32 %X - ; CHECK-LABEL: @test8( -; CHECK: store i32 1234567 +; CHECK-NEXT: [[V:%.*]] = alloca i32 +; CHECK-NEXT: store i32 1234567, i32* [[V]] +; CHECK-NEXT: [[X:%.*]] = load i32, i32* [[V]] +; CHECK-NEXT: ret i32 [[X]] +; + %V = alloca i32 + store i32 1234567, i32* %V + %V2 = bitcast i32* %V to i8* + store i8 0, i8* %V2 + %X = load i32, i32* %V + ret i32 %X + } ; Test for byval handling. %struct.x = type { i32, i32, i32, i32 } define void @test9(%struct.x* byval %a) nounwind { - %tmp2 = getelementptr %struct.x, %struct.x* %a, i32 0, i32 0 - store i32 1, i32* %tmp2, align 4 - ret void ; CHECK-LABEL: @test9( -; CHECK-NEXT: ret void +; CHECK-NEXT: ret void +; + %tmp2 = getelementptr %struct.x, %struct.x* %a, i32 0, i32 0 + store i32 1, i32* %tmp2, align 4 + ret void } ; Test for inalloca handling. define void @test9_2(%struct.x* inalloca %a) nounwind { - %tmp2 = getelementptr %struct.x, %struct.x* %a, i32 0, i32 0 - store i32 1, i32* %tmp2, align 4 - ret void ; CHECK-LABEL: @test9_2( -; CHECK-NEXT: ret void +; CHECK-NEXT: ret void +; + %tmp2 = getelementptr %struct.x, %struct.x* %a, i32 0, i32 0 + store i32 1, i32* %tmp2, align 4 + ret void } ; va_arg has fuzzy dependence, the store shouldn't be zapped. define double @test10(i8* %X) { - %X_addr = alloca i8* - store i8* %X, i8** %X_addr - %tmp.0 = va_arg i8** %X_addr, double - ret double %tmp.0 ; CHECK-LABEL: @test10( -; CHECK: store +; CHECK-NEXT: [[X_ADDR:%.*]] = alloca i8* +; CHECK-NEXT: store i8* [[X:%.*]], i8** [[X_ADDR]] +; CHECK-NEXT: [[TMP_0:%.*]] = va_arg i8** [[X_ADDR]], double +; CHECK-NEXT: ret double [[TMP_0]] +; + %X_addr = alloca i8* + store i8* %X, i8** %X_addr + %tmp.0 = va_arg i8** %X_addr, double + ret double %tmp.0 } @@ -130,50 +158,64 @@ define double @test10(i8* %X) { declare void @test11f() define void @test11() { ; CHECK-LABEL: @test11( - %storage = alloca [10 x i8], align 16 ; <[10 x i8]*> [#uses=1] -; CHECK-NOT: alloca - %cast = getelementptr [10 x i8], [10 x i8]* %storage, i32 0, i32 0 ; [#uses=1] - call void @llvm.init.trampoline( i8* %cast, i8* bitcast (void ()* @test11f to i8*), i8* null ) ; [#uses=1] -; CHECK-NOT: trampoline - ret void -; CHECK: ret void +; CHECK-NEXT: ret void +; + %storage = alloca [10 x i8], align 16 ; <[10 x i8]*> [#uses=1] + %cast = getelementptr [10 x i8], [10 x i8]* %storage, i32 0, i32 0 ; [#uses=1] + call void @llvm.init.trampoline( i8* %cast, i8* bitcast (void ()* @test11f to i8*), i8* null ) ; [#uses=1] + ret void } ; PR2599 - load -> store to same address. define void @test12({ i32, i32 }* %x) nounwind { - %tmp4 = getelementptr { i32, i32 }, { i32, i32 }* %x, i32 0, i32 0 - %tmp5 = load i32, i32* %tmp4, align 4 - %tmp7 = getelementptr { i32, i32 }, { i32, i32 }* %x, i32 0, i32 1 - %tmp8 = load i32, i32* %tmp7, align 4 - %tmp17 = sub i32 0, %tmp8 - store i32 %tmp5, i32* %tmp4, align 4 - store i32 %tmp17, i32* %tmp7, align 4 - ret void ; CHECK-LABEL: @test12( -; CHECK-NOT: tmp5 -; CHECK: ret void +; CHECK-NEXT: [[TMP7:%.*]] = getelementptr { i32, i32 }, { i32, i32 }* [[X:%.*]], i32 0, i32 1 +; CHECK-NEXT: [[TMP8:%.*]] = load i32, i32* [[TMP7]], align 4 +; CHECK-NEXT: [[TMP17:%.*]] = sub i32 0, [[TMP8]] +; CHECK-NEXT: store i32 [[TMP17]], i32* [[TMP7]], align 4 +; CHECK-NEXT: ret void +; + %tmp4 = getelementptr { i32, i32 }, { i32, i32 }* %x, i32 0, i32 0 + %tmp5 = load i32, i32* %tmp4, align 4 + %tmp7 = getelementptr { i32, i32 }, { i32, i32 }* %x, i32 0, i32 1 + %tmp8 = load i32, i32* %tmp7, align 4 + %tmp17 = sub i32 0, %tmp8 + store i32 %tmp5, i32* %tmp4, align 4 + store i32 %tmp17, i32* %tmp7, align 4 + ret void } ; %P doesn't escape, the DEAD instructions should be removed. declare void @test13f() define i32* @test13() { - %p = tail call i8* @malloc(i32 4) - %P = bitcast i8* %p to i32* - %DEAD = load i32, i32* %P - %DEAD2 = add i32 %DEAD, 1 - store i32 %DEAD2, i32* %P - call void @test13f( ) - store i32 0, i32* %P - ret i32* %P -; CHECK: @test13() -; CHECK-NEXT: malloc -; CHECK-NEXT: bitcast -; CHECK-NEXT: call void +; CHECK-LABEL: @test13( +; CHECK-NEXT: [[PTR:%.*]] = tail call i8* @malloc(i32 4) +; CHECK-NEXT: [[P:%.*]] = bitcast i8* [[PTR]] to i32* +; CHECK-NEXT: call void @test13f() +; CHECK-NEXT: store i32 0, i32* [[P]] +; CHECK-NEXT: ret i32* [[P]] +; + %ptr = tail call i8* @malloc(i32 4) + %P = bitcast i8* %ptr to i32* + %DEAD = load i32, i32* %P + %DEAD2 = add i32 %DEAD, 1 + store i32 %DEAD2, i32* %P + call void @test13f( ) + store i32 0, i32* %P + ret i32* %P } define i32 addrspace(1)* @test13_addrspacecast() { +; CHECK-LABEL: @test13_addrspacecast( +; CHECK-NEXT: [[P:%.*]] = tail call i8* @malloc(i32 4) +; CHECK-NEXT: [[P_BC:%.*]] = bitcast i8* [[P]] to i32* +; CHECK-NEXT: [[P:%.*]] = addrspacecast i32* [[P_BC]] to i32 addrspace(1)* +; CHECK-NEXT: call void @test13f() +; CHECK-NEXT: store i32 0, i32 addrspace(1)* [[P]] +; CHECK-NEXT: ret i32 addrspace(1)* [[P]] +; %p = tail call i8* @malloc(i32 4) %p.bc = bitcast i8* %p to i32* %P = addrspacecast i32* %p.bc to i32 addrspace(1)* @@ -183,11 +225,6 @@ define i32 addrspace(1)* @test13_addrspacecast() { call void @test13f( ) store i32 0, i32 addrspace(1)* %P ret i32 addrspace(1)* %P -; CHECK: @test13_addrspacecast() -; CHECK-NEXT: malloc -; CHECK-NEXT: bitcast -; CHECK-NEXT: addrspacecast -; CHECK-NEXT: call void } declare noalias i8* @malloc(i32) @@ -195,13 +232,14 @@ declare noalias i8* @calloc(i32, i32) define void @test14(i32* %Q) { - %P = alloca i32 - %DEAD = load i32, i32* %Q - store i32 %DEAD, i32* %P - ret void - ; CHECK-LABEL: @test14( -; CHECK-NEXT: ret void +; CHECK-NEXT: ret void +; + %P = alloca i32 + %DEAD = load i32, i32* %Q + store i32 %DEAD, i32* %P + ret void + } @@ -209,43 +247,47 @@ define void @test14(i32* %Q) { ;; Fully dead overwrite of memcpy. define void @test15(i8* %P, i8* %Q) nounwind ssp { +; CHECK-LABEL: @test15( +; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[P:%.*]], i8* [[Q:%.*]], i64 12, i1 false) +; CHECK-NEXT: ret void +; tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) ret void -; CHECK-LABEL: @test15( -; CHECK-NEXT: call void @llvm.memcpy -; CHECK-NEXT: ret } ;; Full overwrite of smaller memcpy. define void @test16(i8* %P, i8* %Q) nounwind ssp { +; CHECK-LABEL: @test16( +; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[P:%.*]], i8* [[Q:%.*]], i64 12, i1 false) +; CHECK-NEXT: ret void +; tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 8, i1 false) tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) ret void -; CHECK-LABEL: @test16( -; CHECK-NEXT: call void @llvm.memcpy -; CHECK-NEXT: ret } ;; Overwrite of memset by memcpy. define void @test17(i8* %P, i8* noalias %Q) nounwind ssp { +; CHECK-LABEL: @test17( +; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[P:%.*]], i8* [[Q:%.*]], i64 12, i1 false) +; CHECK-NEXT: ret void +; tail call void @llvm.memset.p0i8.i64(i8* %P, i8 42, i64 8, i1 false) tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) ret void -; CHECK-LABEL: @test17( -; CHECK-NEXT: call void @llvm.memcpy -; CHECK-NEXT: ret } ; Should not delete the volatile memset. define void @test17v(i8* %P, i8* %Q) nounwind ssp { +; CHECK-LABEL: @test17v( +; CHECK-NEXT: tail call void @llvm.memset.p0i8.i64(i8* [[P:%.*]], i8 42, i64 8, i1 true) +; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[P]], i8* [[Q:%.*]], i64 12, i1 false) +; CHECK-NEXT: ret void +; tail call void @llvm.memset.p0i8.i64(i8* %P, i8 42, i64 8, i1 true) tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) ret void -; CHECK-LABEL: @test17v( -; CHECK-NEXT: call void @llvm.memset -; CHECK-NEXT: call void @llvm.memcpy -; CHECK-NEXT: ret } ; PR8728 @@ -256,13 +298,14 @@ define void @test17v(i8* %P, i8* %Q) nounwind ssp { ; NB! See PR11763 - currently LLVM allows memcpy's source and destination to be ; equal (but not inequal and overlapping). define void @test18(i8* %P, i8* %Q, i8* %R) nounwind ssp { +; CHECK-LABEL: @test18( +; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[P:%.*]], i8* [[Q:%.*]], i64 12, i1 false) +; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[P]], i8* [[R:%.*]], i64 12, i1 false) +; CHECK-NEXT: ret void +; tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %R, i64 12, i1 false) ret void -; CHECK-LABEL: @test18( -; CHECK-NEXT: call void @llvm.memcpy -; CHECK-NEXT: call void @llvm.memcpy -; CHECK-NEXT: ret } @@ -270,49 +313,62 @@ define void @test18(i8* %P, i8* %Q, i8* %R) nounwind ssp { declare void @test19f({i32}* byval align 4 %P) define void @test19({i32} * nocapture byval align 4 %arg5) nounwind ssp { +; CHECK-LABEL: @test19( +; CHECK-NEXT: bb: +; CHECK-NEXT: [[TMP7:%.*]] = getelementptr inbounds { i32 }, { i32 }* [[ARG5:%.*]], i32 0, i32 0 +; CHECK-NEXT: store i32 912, i32* [[TMP7]] +; CHECK-NEXT: call void @test19f({ i32 }* byval align 4 [[ARG5]]) +; CHECK-NEXT: ret void +; bb: %tmp7 = getelementptr inbounds {i32}, {i32}* %arg5, i32 0, i32 0 store i32 912, i32* %tmp7 call void @test19f({i32}* byval align 4 %arg5) ret void -; CHECK-LABEL: @test19( -; CHECK: store i32 912 -; CHECK: call void @test19f } define void @test20() { +; CHECK-LABEL: @test20( +; CHECK-NEXT: ret void +; %m = call i8* @malloc(i32 24) store i8 0, i8* %m ret void } -; CHECK-LABEL: @test20( -; CHECK-NEXT: ret void -; CHECK-LABEL: @test21( define void @test21() { +; CHECK-LABEL: @test21( +; CHECK-NEXT: ret void +; %m = call i8* @calloc(i32 9, i32 7) store i8 0, i8* %m -; CHECK-NEXT: ret void ret void } -; CHECK-LABEL: @test22( define void @test22(i1 %i, i32 %k, i32 %m) nounwind { +; CHECK-LABEL: @test22( +; CHECK-NEXT: ret void +; %k.addr = alloca i32 %m.addr = alloca i32 %k.addr.m.addr = select i1 %i, i32* %k.addr, i32* %m.addr store i32 0, i32* %k.addr.m.addr, align 4 -; CHECK-NEXT: ret void ret void } ; PR13547 -; CHECK-LABEL: @test23( -; CHECK: store i8 97 -; CHECK: store i8 0 declare noalias i8* @strdup(i8* nocapture) nounwind define noalias i8* @test23() nounwind uwtable ssp { +; CHECK-LABEL: @test23( +; CHECK-NEXT: [[X:%.*]] = alloca [2 x i8], align 1 +; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds [2 x i8], [2 x i8]* [[X]], i64 0, i64 0 +; CHECK-NEXT: store i8 97, i8* [[ARRAYIDX]], align 1 +; CHECK-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds [2 x i8], [2 x i8]* [[X]], i64 0, i64 1 +; CHECK-NEXT: store i8 0, i8* [[ARRAYIDX1]], align 1 +; CHECK-NEXT: [[CALL:%.*]] = call i8* @strdup(i8* [[ARRAYIDX]]) #1 +; CHECK-NEXT: ret i8* [[CALL]] +; %x = alloca [2 x i8], align 1 %arrayidx = getelementptr inbounds [2 x i8], [2 x i8]* %x, i64 0, i64 0 store i8 97, i8* %arrayidx, align 1 @@ -323,13 +379,14 @@ define noalias i8* @test23() nounwind uwtable ssp { } ; Make sure same sized store to later element is deleted -; CHECK-LABEL: @test24( -; CHECK-NOT: store i32 0 -; CHECK-NOT: store i32 0 -; CHECK: store i32 %b -; CHECK: store i32 %c -; CHECK: ret void define void @test24([2 x i32]* %a, i32 %b, i32 %c) nounwind { +; CHECK-LABEL: @test24( +; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds [2 x i32], [2 x i32]* [[A:%.*]], i64 0, i64 0 +; CHECK-NEXT: store i32 [[B:%.*]], i32* [[TMP1]], align 4 +; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds [2 x i32], [2 x i32]* [[A]], i64 0, i64 1 +; CHECK-NEXT: store i32 [[C:%.*]], i32* [[TMP2]], align 4 +; CHECK-NEXT: ret void +; %1 = getelementptr inbounds [2 x i32], [2 x i32]* %a, i64 0, i64 0 store i32 0, i32* %1, align 4 %2 = getelementptr inbounds [2 x i32], [2 x i32]* %a, i64 0, i64 1 @@ -342,11 +399,15 @@ define void @test24([2 x i32]* %a, i32 %b, i32 %c) nounwind { } ; Check another case like PR13547 where strdup is not like malloc. -; CHECK-LABEL: @test25( -; CHECK: load i8 -; CHECK: store i8 0 -; CHECK: store i8 %tmp define i8* @test25(i8* %p) nounwind { +; CHECK-LABEL: @test25( +; CHECK-NEXT: [[P_4:%.*]] = getelementptr i8, i8* [[P:%.*]], i64 4 +; CHECK-NEXT: [[TMP:%.*]] = load i8, i8* [[P_4]], align 1 +; CHECK-NEXT: store i8 0, i8* [[P_4]], align 1 +; CHECK-NEXT: [[Q:%.*]] = call i8* @strdup(i8* [[P]]) #4 +; CHECK-NEXT: store i8 [[TMP]], i8* [[P_4]], align 1 +; CHECK-NEXT: ret i8* [[Q]] +; %p.4 = getelementptr i8, i8* %p, i64 4 %tmp = load i8, i8* %p.4, align 1 store i8 0, i8* %p.4, align 1 @@ -356,10 +417,17 @@ define i8* @test25(i8* %p) nounwind { } ; Remove redundant store if loaded value is in another block. -; CHECK-LABEL: @test26( -; CHECK-NOT: store -; CHECK: ret define i32 @test26(i1 %c, i32* %p) { +; CHECK-LABEL: @test26( +; CHECK-NEXT: entry: +; CHECK-NEXT: br i1 [[C:%.*]], label [[BB1:%.*]], label [[BB2:%.*]] +; CHECK: bb1: +; CHECK-NEXT: br label [[BB3:%.*]] +; CHECK: bb2: +; CHECK-NEXT: br label [[BB3]] +; CHECK: bb3: +; CHECK-NEXT: ret i32 0 +; entry: %v = load i32, i32* %p, align 4 br i1 %c, label %bb1, label %bb2 @@ -373,10 +441,17 @@ bb3: } ; Remove redundant store if loaded value is in another block. -; CHECK-LABEL: @test27( -; CHECK-NOT: store -; CHECK: ret define i32 @test27(i1 %c, i32* %p) { +; CHECK-LABEL: @test27( +; CHECK-NEXT: entry: +; CHECK-NEXT: br i1 [[C:%.*]], label [[BB1:%.*]], label [[BB2:%.*]] +; CHECK: bb1: +; CHECK-NEXT: br label [[BB3:%.*]] +; CHECK: bb2: +; CHECK-NEXT: br label [[BB3]] +; CHECK: bb3: +; CHECK-NEXT: ret i32 0 +; entry: %v = load i32, i32* %p, align 4 br i1 %c, label %bb1, label %bb2 @@ -390,10 +465,20 @@ bb3: } ; Don't remove redundant store because of may-aliased store. -; CHECK-LABEL: @test28( -; CHECK: bb3: -; CHECK-NEXT: store i32 %v define i32 @test28(i1 %c, i32* %p, i32* %p2, i32 %i) { +; CHECK-LABEL: @test28( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[V:%.*]] = load i32, i32* [[P:%.*]], align 4 +; CHECK-NEXT: store i32 [[I:%.*]], i32* [[P2:%.*]], align 4 +; CHECK-NEXT: br i1 [[C:%.*]], label [[BB1:%.*]], label [[BB2:%.*]] +; CHECK: bb1: +; CHECK-NEXT: br label [[BB3:%.*]] +; CHECK: bb2: +; CHECK-NEXT: br label [[BB3]] +; CHECK: bb3: +; CHECK-NEXT: store i32 [[V]], i32* [[P]], align 4 +; CHECK-NEXT: ret i32 0 +; entry: %v = load i32, i32* %p, align 4 @@ -410,10 +495,20 @@ bb3: } ; Don't remove redundant store because of may-aliased store. -; CHECK-LABEL: @test29( -; CHECK: bb3: -; CHECK-NEXT: store i32 %v define i32 @test29(i1 %c, i32* %p, i32* %p2, i32 %i) { +; CHECK-LABEL: @test29( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[V:%.*]] = load i32, i32* [[P:%.*]], align 4 +; CHECK-NEXT: br i1 [[C:%.*]], label [[BB1:%.*]], label [[BB2:%.*]] +; CHECK: bb1: +; CHECK-NEXT: br label [[BB3:%.*]] +; CHECK: bb2: +; CHECK-NEXT: store i32 [[I:%.*]], i32* [[P2:%.*]], align 4 +; CHECK-NEXT: br label [[BB3]] +; CHECK: bb3: +; CHECK-NEXT: store i32 [[V]], i32* [[P]], align 4 +; CHECK-NEXT: ret i32 0 +; entry: %v = load i32, i32* %p, align 4 br i1 %c, label %bb1, label %bb2 @@ -431,10 +526,20 @@ bb3: declare void @unknown_func() ; Don't remove redundant store because of unknown call. -; CHECK-LABEL: @test30( -; CHECK: bb3: -; CHECK-NEXT: store i32 %v define i32 @test30(i1 %c, i32* %p, i32 %i) { +; CHECK-LABEL: @test30( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[V:%.*]] = load i32, i32* [[P:%.*]], align 4 +; CHECK-NEXT: br i1 [[C:%.*]], label [[BB1:%.*]], label [[BB2:%.*]] +; CHECK: bb1: +; CHECK-NEXT: br label [[BB3:%.*]] +; CHECK: bb2: +; CHECK-NEXT: call void @unknown_func() +; CHECK-NEXT: br label [[BB3]] +; CHECK: bb3: +; CHECK-NEXT: store i32 [[V]], i32* [[P]], align 4 +; CHECK-NEXT: ret i32 0 +; entry: %v = load i32, i32* %p, align 4 br i1 %c, label %bb1, label %bb2 @@ -450,10 +555,15 @@ bb3: } ; Remove redundant store if loaded value is in another block inside a loop. -; CHECK-LABEL: @test31( -; CHECK-NOT: store -; CHECK: ret define i32 @test31(i1 %c, i32* %p, i32 %i) { +; CHECK-LABEL: @test31( +; CHECK-NEXT: entry: +; CHECK-NEXT: br label [[BB1:%.*]] +; CHECK: bb1: +; CHECK-NEXT: br i1 undef, label [[BB1]], label [[BB2:%.*]] +; CHECK: bb2: +; CHECK-NEXT: ret i32 0 +; entry: %v = load i32, i32* %p, align 4 br label %bb1 @@ -465,11 +575,18 @@ bb2: } ; Don't remove redundant store in a loop with a may-alias store. -; CHECK-LABEL: @test32( -; CHECK: bb1: -; CHECK-NEXT: store i32 %v -; CHECK-NEXT: call void @unknown_func define i32 @test32(i1 %c, i32* %p, i32 %i) { +; CHECK-LABEL: @test32( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[V:%.*]] = load i32, i32* [[P:%.*]], align 4 +; CHECK-NEXT: br label [[BB1:%.*]] +; CHECK: bb1: +; CHECK-NEXT: store i32 [[V]], i32* [[P]], align 4 +; CHECK-NEXT: call void @unknown_func() +; CHECK-NEXT: br i1 undef, label [[BB1]], label [[BB2:%.*]] +; CHECK: bb2: +; CHECK-NEXT: ret i32 0 +; entry: %v = load i32, i32* %p, align 4 br label %bb1 @@ -483,10 +600,18 @@ bb2: } ; Remove redundant store, which is in the lame loop as the load. -; CHECK-LABEL: @test33( -; CHECK-NOT: store -; CHECK: ret define i32 @test33(i1 %c, i32* %p, i32 %i) { +; CHECK-LABEL: @test33( +; CHECK-NEXT: entry: +; CHECK-NEXT: br label [[BB1:%.*]] +; CHECK: bb1: +; CHECK-NEXT: br label [[BB2:%.*]] +; CHECK: bb2: +; CHECK-NEXT: call void @unknown_func() +; CHECK-NEXT: br i1 undef, label [[BB1]], label [[BB3:%.*]] +; CHECK: bb3: +; CHECK-NEXT: ret i32 0 +; entry: br label %bb1 bb1: @@ -502,11 +627,13 @@ bb3: } ; Don't remove redundant store: unknown_func could unwind -; CHECK-LABEL: @test34( -; CHECK: store i32 1 -; CHECK: store i32 0 -; CHECK: ret define void @test34(i32* noalias %p) { +; CHECK-LABEL: @test34( +; CHECK-NEXT: store i32 1, i32* [[P:%.*]] +; CHECK-NEXT: call void @unknown_func() +; CHECK-NEXT: store i32 0, i32* [[P]] +; CHECK-NEXT: ret void +; store i32 1, i32* %p call void @unknown_func() store i32 0, i32* %p @@ -514,11 +641,12 @@ define void @test34(i32* noalias %p) { } ; Remove redundant store even with an unwinding function in the same block -; CHECK-LABEL: @test35( -; CHECK: call void @unknown_func -; CHECK-NEXT: store i32 0 -; CHECK-NEXT: ret void define void @test35(i32* noalias %p) { +; CHECK-LABEL: @test35( +; CHECK-NEXT: call void @unknown_func() +; CHECK-NEXT: store i32 0, i32* [[P:%.*]] +; CHECK-NEXT: ret void +; call void @unknown_func() store i32 1, i32* %p store i32 0, i32* %p @@ -528,9 +656,10 @@ define void @test35(i32* noalias %p) { ; We cannot optimize away the first memmove since %P could overlap with %Q. define void @test36(i8* %P, i8* %Q) { ; CHECK-LABEL: @test36( -; CHECK-NEXT: tail call void @llvm.memmove.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) -; CHECK-NEXT: tail call void @llvm.memmove.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) -; CHECK-NEXT: ret +; CHECK-NEXT: tail call void @llvm.memmove.p0i8.p0i8.i64(i8* [[P:%.*]], i8* [[Q:%.*]], i64 12, i1 false) +; CHECK-NEXT: tail call void @llvm.memmove.p0i8.p0i8.i64(i8* [[P]], i8* [[Q]], i64 12, i1 false) +; CHECK-NEXT: ret void +; tail call void @llvm.memmove.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) tail call void @llvm.memmove.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) @@ -539,9 +668,10 @@ define void @test36(i8* %P, i8* %Q) { define void @test37(i8* %P, i8* %Q, i8* %R) { ; CHECK-LABEL: @test37( -; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) -; CHECK-NEXT: tail call void @llvm.memmove.p0i8.p0i8.i64(i8* %P, i8* %R, i64 12, i1 false) -; CHECK-NEXT: ret +; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[P:%.*]], i8* [[Q:%.*]], i64 12, i1 false) +; CHECK-NEXT: tail call void @llvm.memmove.p0i8.p0i8.i64(i8* [[P]], i8* [[R:%.*]], i64 12, i1 false) +; CHECK-NEXT: ret void +; tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) tail call void @llvm.memmove.p0i8.p0i8.i64(i8* %P, i8* %R, i64 12, i1 false) @@ -551,9 +681,10 @@ define void @test37(i8* %P, i8* %Q, i8* %R) { ; Same caveat about memcpy as in @test18 applies here. define void @test38(i8* %P, i8* %Q, i8* %R) { ; CHECK-LABEL: @test38( -; CHECK-NEXT: tail call void @llvm.memmove.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) -; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %R, i64 12, i1 false) -; CHECK-NEXT: ret +; CHECK-NEXT: tail call void @llvm.memmove.p0i8.p0i8.i64(i8* [[P:%.*]], i8* [[Q:%.*]], i64 12, i1 false) +; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[P]], i8* [[R:%.*]], i64 12, i1 false) +; CHECK-NEXT: ret void +; tail call void @llvm.memmove.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %R, i64 12, i1 false) @@ -562,9 +693,10 @@ define void @test38(i8* %P, i8* %Q, i8* %R) { define void @test39(i8* %P, i8* %Q, i8* %R) { ; CHECK-LABEL: @test39( -; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) -; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %R, i64 8, i1 false) -; CHECK-NEXT: ret +; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[P:%.*]], i8* [[Q:%.*]], i64 12, i1 false) +; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[P]], i8* [[R:%.*]], i64 8, i1 false) +; CHECK-NEXT: ret void +; tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %Q, i64 12, i1 false) tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %P, i8* %R, i64 8, i1 false) -- 2.7.4