diff options
Diffstat (limited to 'test/Transforms')
46 files changed, 1038 insertions, 181 deletions
diff --git a/test/Transforms/DeadArgElim/dbginfo.ll b/test/Transforms/DeadArgElim/dbginfo.ll index 59eb458..351925b 100644 --- a/test/Transforms/DeadArgElim/dbginfo.ll +++ b/test/Transforms/DeadArgElim/dbginfo.ll @@ -36,11 +36,11 @@ entry: !llvm.dbg.cu = !{!0} -!0 = metadata !{i32 786449, i32 0, i32 4, metadata !"test.cc", metadata !"/home/samsonov/tmp/clang-di", metadata !"clang version 3.2 (trunk 165305)", i1 true, i1 false, metadata !"", i32 0, metadata !1, metadata !1, metadata !3, metadata !1} ; [ DW_TAG_compile_unit ] [/home/samsonov/tmp/clang-di/test.cc] [DW_LANG_C_plus_plus] +!0 = metadata !{i32 786449, i32 0, i32 4, metadata !6, metadata !"clang version 3.2 (trunk 165305)", i1 false, metadata !"", i32 0, metadata !1, metadata !1, metadata !3, metadata !1, metadata !""} ; [ DW_TAG_compile_unit ] [/home/samsonov/tmp/clang-di/test.cc] [DW_LANG_C_plus_plus] !1 = metadata !{i32 0} !3 = metadata !{metadata !5, metadata !8, metadata !9} !5 = metadata !{i32 786478, i32 0, metadata !6, metadata !"run", metadata !"run", metadata !"", metadata !6, i32 8, metadata !7, i1 false, i1 true, i32 0, i32 0, null, i32 256, i1 false, void ()* @_Z3runv, null, null, metadata !1, i32 8} ; [ DW_TAG_subprogram ] [line 8] [def] [run] -!6 = metadata !{i32 786473, metadata !"test.cc", metadata !"/home/samsonov/tmp/clang-di", null} ; [ DW_TAG_file_type ] +!6 = metadata !{i32 786473, metadata !20} ; [ DW_TAG_file_type ] !7 = metadata !{i32 786453, i32 0, metadata !"", i32 0, i32 0, i64 0, i64 0, i64 0, i32 0, null, metadata !1, i32 0, i32 0} ; [ DW_TAG_subroutine_type ] [line 0, size 0, align 0, offset 0] [from ] !8 = metadata !{i32 786478, i32 0, metadata !6, metadata !"dead_vararg", metadata !"dead_vararg", metadata !"", metadata !6, i32 5, metadata !7, i1 true, i1 true, i32 0, i32 0, null, i32 256, i1 false, void (...)* @_ZN12_GLOBAL__N_111dead_varargEz, null, null, metadata !1, i32 5} ; [ DW_TAG_subprogram ] [line 5] [local] [def] [dead_vararg] @@ -60,3 +60,4 @@ entry: !17 = metadata !{i32 5, i32 25, metadata !18, null} !18 = metadata !{i32 786443, metadata !8, i32 5, i32 23, metadata !6, i32 1} ; [ DW_TAG_lexical_block ] [/home/samsonov/tmp/clang-di/test.cc] !19 = metadata !{i32 5, i32 30, metadata !18, null} +!20 = metadata !{metadata !"test.cc", metadata !"/home/samsonov/tmp/clang-di"} diff --git a/test/Transforms/EarlyCSE/floatingpoint.ll b/test/Transforms/EarlyCSE/floatingpoint.ll new file mode 100644 index 0000000..2abecd7 --- /dev/null +++ b/test/Transforms/EarlyCSE/floatingpoint.ll @@ -0,0 +1,14 @@ +; RUN: opt < %s -S -early-cse | FileCheck %s + +; Ensure we don't simplify away additions vectors of +0.0's (same as scalars). +define <4 x float> @fV( <4 x float> %a) { + ; CHECK: %b = fadd <4 x float> %a, zeroinitializer + %b = fadd <4 x float> %a, <float 0.0,float 0.0,float 0.0,float 0.0> + ret <4 x float> %b +} + +define <4 x float> @fW( <4 x float> %a) { + ; CHECK: ret <4 x float> %a + %b = fadd <4 x float> %a, <float -0.0,float -0.0,float -0.0,float -0.0> + ret <4 x float> %b +} diff --git a/test/Transforms/GCOVProfiling/lit.local.cfg b/test/Transforms/GCOVProfiling/lit.local.cfg new file mode 100644 index 0000000..19eebc0 --- /dev/null +++ b/test/Transforms/GCOVProfiling/lit.local.cfg @@ -0,0 +1 @@ +config.suffixes = ['.ll', '.c', '.cpp'] diff --git a/test/Transforms/GCOVProfiling/version.ll b/test/Transforms/GCOVProfiling/version.ll new file mode 100644 index 0000000..efd633a --- /dev/null +++ b/test/Transforms/GCOVProfiling/version.ll @@ -0,0 +1,27 @@ +; RUN: echo '!9 = metadata !{metadata !"%s", metadata !0}' > %t1 +; RUN: cat %s %t1 > %t2 +; RUN: opt -insert-gcov-profiling -disable-output < %t2 +; RUN: head -c12 %S/version.gcno | grep '^oncg\*204MVLL$' +; RUN: rm %S/version.gcno +; RUN: not opt -insert-gcov-profiling -default-gcov-version=asdfasdf -disable-output < %t2 +; RUN: opt -insert-gcov-profiling -default-gcov-version=407* -disable-output < %t2 +; RUN: head -c12 %S/version.gcno | grep '^oncg\*704MVLL$' +; RUN: rm %S/version.gcno + +define void @test() { + ret void, !dbg !8 +} + +!llvm.gcov = !{!9} +!llvm.dbg.cu = !{!0} + +!0 = metadata !{metadata !"./version", metadata !1} +!1 = metadata !{i32 786449, i32 0, i32 4, metadata !2, metadata !"clang version 3.3 (trunk 176994)", i1 false, metadata !"", i32 0, metadata !3, metadata !3, metadata !4, metadata !3, metadata !""} ; [ DW_TAG_compile_unit ] [./version] [DW_LANG_C_plus_plus] +!2 = metadata !{i32 786473, metadata !"version", metadata !"/usr/local/google/home/nlewycky"} ; [ DW_TAG_file_type ] +!3 = metadata !{i32 0} +!4 = metadata !{metadata !5} +!5 = metadata !{i32 786478, i32 0, metadata !6, metadata !"test", metadata !"test", metadata !"", metadata !6, i32 1, metadata !7, i1 false, i1 true, i32 0, i32 0, null, i32 256, i1 false, void ()* @test, null, null, metadata !3, i32 1} ; [ DW_TAG_subprogram ] [line 1] [def] [test] +!6 = metadata !{i32 786473, metadata !"<stdin>", metadata !"."} ; [ DW_TAG_file_type ] +!7 = metadata !{i32 786453, i32 0, metadata !"", i32 0, i32 0, i64 0, i64 0, i64 0, i32 0, null, metadata !3, i32 0, i32 0} ; [ DW_TAG_subroutine_type ] [line 0, size 0, align 0, offset 0] [from ] +!8 = metadata !{i32 1, i32 0, metadata !5, null} +;; !9 is added through the echo line at the top. diff --git a/test/Transforms/GVN/nonescaping-malloc.ll b/test/Transforms/GVN/nonescaping-malloc.ll index afcb7fe..c2eeed5 100644 --- a/test/Transforms/GVN/nonescaping-malloc.ll +++ b/test/Transforms/GVN/nonescaping-malloc.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt < %s -basicaa -gvn -stats -disable-output 2>&1 | grep "Number of loads deleted" ; rdar://7363102 diff --git a/test/Transforms/GlobalOpt/2009-03-05-dbg.ll b/test/Transforms/GlobalOpt/2009-03-05-dbg.ll index 0f3efa0..e71aed9 100644 --- a/test/Transforms/GlobalOpt/2009-03-05-dbg.ll +++ b/test/Transforms/GlobalOpt/2009-03-05-dbg.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt < %s -globalopt -stats -disable-output 2>&1 | grep "1 globalopt - Number of global vars shrunk to booleans" @Stop = internal global i32 0 ; <i32*> [#uses=3] diff --git a/test/Transforms/GlobalOpt/integer-bool.ll b/test/Transforms/GlobalOpt/integer-bool.ll index cf025ec..5185806 100644 --- a/test/Transforms/GlobalOpt/integer-bool.ll +++ b/test/Transforms/GlobalOpt/integer-bool.ll @@ -2,9 +2,9 @@ ;; check that global opt turns integers that only hold 0 or 1 into bools. @G = internal addrspace(1) global i32 0 -; CHECK @G.b -; CHECK addrspace(1) -; CHECK global i1 0 +; CHECK: @G +; CHECK: addrspace(1) +; CHECK: global i1 false define void @set1() { store i32 0, i32 addrspace(1)* @G @@ -19,7 +19,7 @@ define void @set2() { } define i1 @get() { -; CHECK @get +; CHECK: @get %A = load i32 addrspace(1) * @G %C = icmp slt i32 %A, 2 ret i1 %C diff --git a/test/Transforms/IndVarSimplify/phi-uses-value-multiple-times.ll b/test/Transforms/IndVarSimplify/phi-uses-value-multiple-times.ll index 52c9e5c..dc36b99 100644 --- a/test/Transforms/IndVarSimplify/phi-uses-value-multiple-times.ll +++ b/test/Transforms/IndVarSimplify/phi-uses-value-multiple-times.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt < %s -indvars -disable-output -stats -info-output-file - | FileCheck %s ; Check that IndVarSimplify is not creating unnecessary canonical IVs ; that will never be used. diff --git a/test/Transforms/Inline/delete-call.ll b/test/Transforms/Inline/delete-call.ll index 0afd2ee..97c52af 100644 --- a/test/Transforms/Inline/delete-call.ll +++ b/test/Transforms/Inline/delete-call.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt -S -inline -stats < %s 2>&1 | FileCheck %s ; CHECK: Number of functions inlined diff --git a/test/Transforms/Inline/devirtualize.ll b/test/Transforms/Inline/devirtualize.ll index 18bbf7a..d46154e 100644 --- a/test/Transforms/Inline/devirtualize.ll +++ b/test/Transforms/Inline/devirtualize.ll @@ -1,4 +1,4 @@ -; RUN: opt -S -basicaa -inline -scalarrepl -instcombine -simplifycfg -instcombine -gvn -globaldce < %s | FileCheck %s +; RUN: opt -S -Os < %s | FileCheck %s target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64" target triple = "x86_64-apple-darwin10.0.0" diff --git a/test/Transforms/InstCombine/2013-03-05-Combine-BitcastTy-Into-Alloca.ll b/test/Transforms/InstCombine/2013-03-05-Combine-BitcastTy-Into-Alloca.ll new file mode 100644 index 0000000..b20c3a0 --- /dev/null +++ b/test/Transforms/InstCombine/2013-03-05-Combine-BitcastTy-Into-Alloca.ll @@ -0,0 +1,45 @@ +; RUN: opt < %s -instcombine -S | FileCheck %s +target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128" +target triple = "x86_64-apple-macosx10.9.0" + +%struct._my_struct = type <{ [12 x i8], [4 x i8] }> + +@initval = common global %struct._my_struct zeroinitializer, align 1 + +; InstCombine will try to change the %struct._my_struct alloca into an +; allocation of an i96 because of the bitcast to create %2. That's not valid, +; as the other 32 bits of the structure still feed into the return value +define { i64, i64 } @function(i32 %x, i32 %y, i32 %z) nounwind { +; CHECK: @function +; CHECK-NEXT: entry: +; CHECK-NEXT: %retval = alloca %struct._my_struct, align 8 +; CHECK-NOT: bitcast i96* %retval to %struct._my_struct* +entry: + %retval = alloca %struct._my_struct, align 8 + %k.sroa.0.0.copyload = load i96* bitcast (%struct._my_struct* @initval to i96*), align 1 + %k.sroa.1.12.copyload = load i32* bitcast ([4 x i8]* getelementptr inbounds (%struct._my_struct* @initval, i64 0, i32 1) to i32*), align 1 + %0 = zext i32 %x to i96 + %bf.value = shl nuw nsw i96 %0, 6 + %bf.clear = and i96 %k.sroa.0.0.copyload, -288230376151711744 + %1 = zext i32 %y to i96 + %bf.value2 = shl nuw nsw i96 %1, 32 + %bf.shl3 = and i96 %bf.value2, 288230371856744448 + %bf.value.masked = and i96 %bf.value, 4294967232 + %2 = zext i32 %z to i96 + %bf.value8 = and i96 %2, 63 + %bf.clear4 = or i96 %bf.shl3, %bf.value.masked + %bf.set5 = or i96 %bf.clear4, %bf.value8 + %bf.set10 = or i96 %bf.set5, %bf.clear + %retval.0.cast7 = bitcast %struct._my_struct* %retval to i96* + store i96 %bf.set10, i96* %retval.0.cast7, align 8 + %retval.12.idx8 = getelementptr inbounds %struct._my_struct* %retval, i64 0, i32 1 + %retval.12.cast9 = bitcast [4 x i8]* %retval.12.idx8 to i32* + store i32 %k.sroa.1.12.copyload, i32* %retval.12.cast9, align 4 + %trunc = trunc i96 %bf.set10 to i64 + %.fca.0.insert = insertvalue { i64, i64 } undef, i64 %trunc, 0 + %retval.8.idx12 = getelementptr inbounds %struct._my_struct* %retval, i64 0, i32 0, i64 8 + %retval.8.cast13 = bitcast i8* %retval.8.idx12 to i64* + %retval.8.load14 = load i64* %retval.8.cast13, align 8 + %.fca.1.insert = insertvalue { i64, i64 } %.fca.0.insert, i64 %retval.8.load14, 1 + ret { i64, i64 } %.fca.1.insert +} diff --git a/test/Transforms/InstCombine/debuginfo.ll b/test/Transforms/InstCombine/debuginfo.ll index f6892fc..e7fe71d 100644 --- a/test/Transforms/InstCombine/debuginfo.ll +++ b/test/Transforms/InstCombine/debuginfo.ll @@ -28,22 +28,21 @@ entry: ret i8* %call, !dbg !21 } -!llvm.dbg.lv.foobar = !{!0, !7, !9} -!llvm.dbg.sp = !{!1} +!llvm.dbg.cu = !{!3} -!0 = metadata !{i32 590081, metadata !1, metadata !"__dest", metadata !2, i32 16777294, metadata !6, i32 0} ; [ DW_TAG_arg_variable ] -!1 = metadata !{i32 589870, i32 0, metadata !2, metadata !"foobar", metadata !"foobar", metadata !"", metadata !2, i32 79, metadata !4, i1 true, i1 true, i32 0, i32 0, i32 0, i32 256, i1 true, i8* (i8*, i32, i64)* @foobar} ; [ DW_TAG_subprogram ] -!2 = metadata !{i32 589865, metadata !"string.h", metadata !"Game", metadata !3} ; [ DW_TAG_file_type ] -!3 = metadata !{i32 589841, i32 0, i32 12, metadata !"bits.c", metadata !"Game", metadata !"clang version 3.0 (trunk 127710)", i1 true, i1 true, metadata !"", i32 0} ; [ DW_TAG_compile_unit ] -!4 = metadata !{i32 589845, metadata !2, metadata !"", metadata !2, i32 0, i64 0, i64 0, i32 0, i32 0, i32 0, metadata !5, i32 0, i32 0} ; [ DW_TAG_subroutine_type ] +!0 = metadata !{i32 786689, metadata !1, metadata !"__dest", metadata !2, i32 16777294, metadata !6, i32 0, null} ; [ DW_TAG_arg_variable ] +!1 = metadata !{i32 786478, i32 0, metadata !2, metadata !"foobar", metadata !"foobar", metadata !"", metadata !2, i32 79, metadata !4, i1 true, i1 true, i32 0, i32 0, i32 0, i32 256, i1 true, i8* (i8*, i32, i64)* @foobar, null, null, metadata !25, i32 79} ; [ DW_TAG_subprogram ] +!2 = metadata !{i32 786473, metadata !27} ; [ DW_TAG_file_type ] +!3 = metadata !{i32 786449, i32 0, i32 12, metadata !26, metadata !"clang version 3.0 (trunk 127710)", i1 true, metadata !"", i32 0, null, null, metadata !24, null, null} ; [ DW_TAG_compile_unit ] +!4 = metadata !{i32 786453, metadata !2, metadata !"", metadata !2, i32 0, i64 0, i64 0, i32 0, i32 0, i32 0, metadata !5, i32 0, i32 0} ; [ DW_TAG_subroutine_type ] !5 = metadata !{metadata !6} -!6 = metadata !{i32 589839, metadata !3, metadata !"", null, i32 0, i64 64, i64 64, i64 0, i32 0, null} ; [ DW_TAG_pointer_type ] -!7 = metadata !{i32 590081, metadata !1, metadata !"__val", metadata !2, i32 33554510, metadata !8, i32 0} ; [ DW_TAG_arg_variable ] -!8 = metadata !{i32 589860, metadata !3, metadata !"int", null, i32 0, i64 32, i64 32, i64 0, i32 0, i32 5} ; [ DW_TAG_base_type ] -!9 = metadata !{i32 590081, metadata !1, metadata !"__len", metadata !2, i32 50331726, metadata !10, i32 0} ; [ DW_TAG_arg_variable ] +!6 = metadata !{i32 786447, metadata !3, metadata !"", null, i32 0, i64 64, i64 64, i64 0, i32 0, null} ; [ DW_TAG_pointer_type ] +!7 = metadata !{i32 786689, metadata !1, metadata !"__val", metadata !2, i32 33554510, metadata !8, i32 0, null} ; [ DW_TAG_arg_variable ] +!8 = metadata !{i32 786468, metadata !3, metadata !"int", null, i32 0, i64 32, i64 32, i64 0, i32 0, i32 5} ; [ DW_TAG_base_type ] +!9 = metadata !{i32 786689, metadata !1, metadata !"__len", metadata !2, i32 50331726, metadata !10, i32 0, null} ; [ DW_TAG_arg_variable ] !10 = metadata !{i32 589846, metadata !3, metadata !"size_t", metadata !2, i32 80, i64 0, i64 0, i64 0, i32 0, metadata !11} ; [ DW_TAG_typedef ] !11 = metadata !{i32 589846, metadata !3, metadata !"__darwin_size_t", metadata !2, i32 90, i64 0, i64 0, i64 0, i32 0, metadata !12} ; [ DW_TAG_typedef ] -!12 = metadata !{i32 589860, metadata !3, metadata !"long unsigned int", null, i32 0, i64 64, i64 64, i64 0, i32 0, i32 7} ; [ DW_TAG_base_type ] +!12 = metadata !{i32 786468, metadata !3, metadata !"long unsigned int", null, i32 0, i64 64, i64 64, i64 0, i32 0, i32 7} ; [ DW_TAG_base_type ] !13 = metadata !{metadata !"any pointer", metadata !14} !14 = metadata !{metadata !"omnipotent char", metadata !15} !15 = metadata !{metadata !"Simple C/C++ TBAA", null} @@ -53,5 +52,10 @@ entry: !19 = metadata !{metadata !"long", metadata !14} !20 = metadata !{i32 78, i32 54, metadata !1, null} !21 = metadata !{i32 80, i32 3, metadata !22, null} -!22 = metadata !{i32 589835, metadata !23, i32 80, i32 3, metadata !2, i32 7} ; [ DW_TAG_lexical_block ] -!23 = metadata !{i32 589835, metadata !1, i32 79, i32 1, metadata !2, i32 6} ; [ DW_TAG_lexical_block ] +!22 = metadata !{i32 786443, metadata !23, i32 80, i32 3, metadata !2, i32 7} ; [ DW_TAG_lexical_block ] +!23 = metadata !{i32 786443, metadata !1, i32 79, i32 1, metadata !2, i32 6} ; [ DW_TAG_lexical_block ] +!24 = metadata !{metadata !1} +!25 = metadata !{metadata !0, metadata !7, metadata !9} +!26 = metadata !{i32 786473, metadata !28} ; [ DW_TAG_file_type ] +!27 = metadata !{metadata !"string.h", metadata !"Game"} +!28 = metadata !{metadata !"bits.c", metadata !"Game"} diff --git a/test/Transforms/InstCombine/fast-math.ll b/test/Transforms/InstCombine/fast-math.ll index c97bd28..47f1ec4 100644 --- a/test/Transforms/InstCombine/fast-math.ll +++ b/test/Transforms/InstCombine/fast-math.ll @@ -172,6 +172,17 @@ define double @fmul_distribute3(double %f1) { ; CHECK: fmul fast double %t2, 0x10000000000000 } +; ((X*C1) + C2) * C3 => (X * (C1*C3)) + (C2*C3) (i.e. distribution) +define float @fmul_distribute4(float %f1) { + %t1 = fmul float %f1, 6.0e+3 + %t2 = fsub float 2.0e+3, %t1 + %t3 = fmul fast float %t2, 5.0e+3 + ret float %t3 +; CHECK: @fmul_distribute4 +; CHECK: %1 = fmul fast float %f1, 3.000000e+07 +; CHECK: %t3 = fsub fast float 1.000000e+07, %1 +} + ; C1/X * C2 => (C1*C2) / X define float @fmul2(float %f1) { %t1 = fdiv float 2.0e+3, %f1 @@ -339,3 +350,108 @@ define float @fdiv9(float %x) { ; CHECK: @fdiv9 ; CHECK: fmul fast float %x, 5.000000e+00 } + +; ========================================================================= +; +; Testing-cases about factorization +; +; ========================================================================= +; x*z + y*z => (x+y) * z +define float @fact_mul1(float %x, float %y, float %z) { + %t1 = fmul fast float %x, %z + %t2 = fmul fast float %y, %z + %t3 = fadd fast float %t1, %t2 + ret float %t3 +; CHECK: @fact_mul1 +; CHECK: fmul fast float %1, %z +} + +; z*x + y*z => (x+y) * z +define float @fact_mul2(float %x, float %y, float %z) { + %t1 = fmul fast float %z, %x + %t2 = fmul fast float %y, %z + %t3 = fsub fast float %t1, %t2 + ret float %t3 +; CHECK: @fact_mul2 +; CHECK: fmul fast float %1, %z +} + +; z*x - z*y => (x-y) * z +define float @fact_mul3(float %x, float %y, float %z) { + %t2 = fmul fast float %z, %y + %t1 = fmul fast float %z, %x + %t3 = fsub fast float %t1, %t2 + ret float %t3 +; CHECK: @fact_mul3 +; CHECK: fmul fast float %1, %z +} + +; x*z - z*y => (x-y) * z +define float @fact_mul4(float %x, float %y, float %z) { + %t1 = fmul fast float %x, %z + %t2 = fmul fast float %z, %y + %t3 = fsub fast float %t1, %t2 + ret float %t3 +; CHECK: @fact_mul4 +; CHECK: fmul fast float %1, %z +} + +; x/y + x/z, no xform +define float @fact_div1(float %x, float %y, float %z) { + %t1 = fdiv fast float %x, %y + %t2 = fdiv fast float %x, %z + %t3 = fadd fast float %t1, %t2 + ret float %t3 +; CHECK: fact_div1 +; CHECK: fadd fast float %t1, %t2 +} + +; x/y + z/x; no xform +define float @fact_div2(float %x, float %y, float %z) { + %t1 = fdiv fast float %x, %y + %t2 = fdiv fast float %z, %x + %t3 = fadd fast float %t1, %t2 + ret float %t3 +; CHECK: fact_div2 +; CHECK: fadd fast float %t1, %t2 +} + +; y/x + z/x => (y+z)/x +define float @fact_div3(float %x, float %y, float %z) { + %t1 = fdiv fast float %y, %x + %t2 = fdiv fast float %z, %x + %t3 = fadd fast float %t1, %t2 + ret float %t3 +; CHECK: fact_div3 +; CHECK: fdiv fast float %1, %x +} + +; y/x - z/x => (y-z)/x +define float @fact_div4(float %x, float %y, float %z) { + %t1 = fdiv fast float %y, %x + %t2 = fdiv fast float %z, %x + %t3 = fsub fast float %t1, %t2 + ret float %t3 +; CHECK: fact_div4 +; CHECK: fdiv fast float %1, %x +} + +; y/x - z/x => (y-z)/x is disabled if y-z is denormal. +define float @fact_div5(float %x) { + %t1 = fdiv fast float 0x3810000000000000, %x + %t2 = fdiv fast float 0x3800000000000000, %x + %t3 = fadd fast float %t1, %t2 + ret float %t3 +; CHECK: fact_div5 +; CHECK: fdiv fast float 0x3818000000000000, %x +} + +; y/x - z/x => (y-z)/x is disabled if y-z is denormal. +define float @fact_div6(float %x) { + %t1 = fdiv fast float 0x3810000000000000, %x + %t2 = fdiv fast float 0x3800000000000000, %x + %t3 = fsub fast float %t1, %t2 + ret float %t3 +; CHECK: fact_div6 +; CHECK: %t3 = fsub fast float %t1, %t2 +} diff --git a/test/Transforms/InstCombine/objsize.ll b/test/Transforms/InstCombine/objsize.ll index 31a3cb4..0ead9d1 100644 --- a/test/Transforms/InstCombine/objsize.ll +++ b/test/Transforms/InstCombine/objsize.ll @@ -256,3 +256,131 @@ xpto: return: ret i32 7 } + +declare noalias i8* @valloc(i32) nounwind + +; CHECK: @test14 +; CHECK: ret i32 6 +define i32 @test14(i32 %a) nounwind { + switch i32 %a, label %sw.default [ + i32 1, label %sw.bb + i32 2, label %sw.bb1 + ] + +sw.bb: + %call = tail call noalias i8* @malloc(i32 6) nounwind + br label %sw.epilog + +sw.bb1: + %call2 = tail call noalias i8* @calloc(i32 3, i32 2) nounwind + br label %sw.epilog + +sw.default: + %call3 = tail call noalias i8* @valloc(i32 6) nounwind + br label %sw.epilog + +sw.epilog: + %b.0 = phi i8* [ %call3, %sw.default ], [ %call2, %sw.bb1 ], [ %call, %sw.bb ] + %1 = tail call i32 @llvm.objectsize.i32(i8* %b.0, i1 false) + ret i32 %1 +} + +; CHECK: @test15 +; CHECK: llvm.objectsize +define i32 @test15(i32 %a) nounwind { + switch i32 %a, label %sw.default [ + i32 1, label %sw.bb + i32 2, label %sw.bb1 + ] + +sw.bb: + %call = tail call noalias i8* @malloc(i32 3) nounwind + br label %sw.epilog + +sw.bb1: + %call2 = tail call noalias i8* @calloc(i32 2, i32 1) nounwind + br label %sw.epilog + +sw.default: + %call3 = tail call noalias i8* @valloc(i32 3) nounwind + br label %sw.epilog + +sw.epilog: + %b.0 = phi i8* [ %call3, %sw.default ], [ %call2, %sw.bb1 ], [ %call, %sw.bb ] + %1 = tail call i32 @llvm.objectsize.i32(i8* %b.0, i1 false) + ret i32 %1 +} + +; CHECK: @test16 +; CHECK: llvm.objectsize +define i32 @test16(i8* %a, i32 %n) nounwind { + %b = alloca [5 x i8], align 1 + %c = alloca [5 x i8], align 1 + switch i32 %n, label %sw.default [ + i32 1, label %sw.bb + i32 2, label %sw.bb1 + ] + +sw.bb: + %bp = bitcast [5 x i8]* %b to i8* + br label %sw.epilog + +sw.bb1: + %cp = bitcast [5 x i8]* %c to i8* + br label %sw.epilog + +sw.default: + br label %sw.epilog + +sw.epilog: + %phi = phi i8* [ %a, %sw.default ], [ %cp, %sw.bb1 ], [ %bp, %sw.bb ] + %sz = call i32 @llvm.objectsize.i32(i8* %phi, i1 false) + ret i32 %sz +} + +; CHECK: @test17 +; CHECK: ret i32 5 +define i32 @test17(i32 %n) nounwind { + %b = alloca [5 x i8], align 1 + %c = alloca [5 x i8], align 1 + %bp = bitcast [5 x i8]* %b to i8* + switch i32 %n, label %sw.default [ + i32 1, label %sw.bb + i32 2, label %sw.bb1 + ] + +sw.bb: + br label %sw.epilog + +sw.bb1: + %cp = bitcast [5 x i8]* %c to i8* + br label %sw.epilog + +sw.default: + br label %sw.epilog + +sw.epilog: + %phi = phi i8* [ %bp, %sw.default ], [ %cp, %sw.bb1 ], [ %bp, %sw.bb ] + %sz = call i32 @llvm.objectsize.i32(i8* %phi, i1 false) + ret i32 %sz +} + +@globalalias = alias internal [60 x i8]* @a + +; CHECK: @test18 +; CHECK-NEXT: ret i32 60 +define i32 @test18() { + %bc = bitcast [60 x i8]* @globalalias to i8* + %1 = call i32 @llvm.objectsize.i32(i8* %bc, i1 false) + ret i32 %1 +} + +@globalalias2 = alias weak [60 x i8]* @a + +; CHECK: @test19 +; CHECK: llvm.objectsize +define i32 @test19() { + %bc = bitcast [60 x i8]* @globalalias2 to i8* + %1 = call i32 @llvm.objectsize.i32(i8* %bc, i1 false) + ret i32 %1 +} diff --git a/test/Transforms/InstSimplify/compare.ll b/test/Transforms/InstSimplify/compare.ll index 0ecfb1f..b764c76 100644 --- a/test/Transforms/InstSimplify/compare.ll +++ b/test/Transforms/InstSimplify/compare.ll @@ -682,3 +682,14 @@ define zeroext i1 @external_compare(i32* noalias %x) { ; CHECK: external_compare ; CHECK: ret i1 %cmp } + +define i1 @alloca_gep(i64 %a, i64 %b) { +; CHECK: @alloca_gep +; We can prove this GEP is non-null because it is inbounds and the pointer +; is non-null. + %strs = alloca [1000 x [1001 x i8]], align 16 + %x = getelementptr inbounds [1000 x [1001 x i8]]* %strs, i64 0, i64 %a, i64 %b + %cmp = icmp eq i8* %x, null + ret i1 %cmp +; CHECK-NEXT: ret i1 false +} diff --git a/test/Transforms/LICM/hoist-invariant-load.ll b/test/Transforms/LICM/hoist-invariant-load.ll index f9fc551..1ba94d6 100644 --- a/test/Transforms/LICM/hoist-invariant-load.ll +++ b/test/Transforms/LICM/hoist-invariant-load.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt < %s -licm -stats -S 2>&1 | grep "1 licm" @"\01L_OBJC_METH_VAR_NAME_" = internal global [4 x i8] c"foo\00", section "__TEXT,__objc_methname,cstring_literals", align 1 diff --git a/test/Transforms/LoopUnroll/runtime-loop3.ll b/test/Transforms/LoopUnroll/runtime-loop3.ll index 55cf223..aa928cc 100644 --- a/test/Transforms/LoopUnroll/runtime-loop3.ll +++ b/test/Transforms/LoopUnroll/runtime-loop3.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt < %s -disable-output -stats -loop-unroll -unroll-runtime -unroll-threshold=400 -info-output-file - | FileCheck %s --check-prefix=STATS ; Test that nested loops can be unrolled. We need to increase threshold to do it diff --git a/test/Transforms/LoopUnswitch/2008-11-03-Invariant.ll b/test/Transforms/LoopUnswitch/2008-11-03-Invariant.ll index 9d73d31..31dba79 100644 --- a/test/Transforms/LoopUnswitch/2008-11-03-Invariant.ll +++ b/test/Transforms/LoopUnswitch/2008-11-03-Invariant.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt < %s -loop-unswitch -stats -disable-output 2>&1 | grep "1 loop-unswitch - Number of branches unswitched" | count 1 ; PR 3170 define i32 @a(i32 %x, i32 %y) nounwind { diff --git a/test/Transforms/LoopUnswitch/2011-11-18-SimpleSwitch.ll b/test/Transforms/LoopUnswitch/2011-11-18-SimpleSwitch.ll index bde52da..a8608b87 100644 --- a/test/Transforms/LoopUnswitch/2011-11-18-SimpleSwitch.ll +++ b/test/Transforms/LoopUnswitch/2011-11-18-SimpleSwitch.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt -loop-unswitch -disable-output -stats -info-output-file - < %s | FileCheck --check-prefix=STATS %s ; RUN: opt -S -loop-unswitch -verify-loop-info -verify-dom-info < %s | FileCheck %s diff --git a/test/Transforms/LoopUnswitch/2011-11-18-TwoSwitches-Threshold.ll b/test/Transforms/LoopUnswitch/2011-11-18-TwoSwitches-Threshold.ll index c3bf596..686cedb 100644 --- a/test/Transforms/LoopUnswitch/2011-11-18-TwoSwitches-Threshold.ll +++ b/test/Transforms/LoopUnswitch/2011-11-18-TwoSwitches-Threshold.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt -loop-unswitch -loop-unswitch-threshold 13 -disable-output -stats -info-output-file - < %s | FileCheck --check-prefix=STATS %s ; RUN: opt -S -loop-unswitch -loop-unswitch-threshold 13 -verify-loop-info -verify-dom-info < %s | FileCheck %s diff --git a/test/Transforms/LoopUnswitch/2011-11-18-TwoSwitches.ll b/test/Transforms/LoopUnswitch/2011-11-18-TwoSwitches.ll index 9530333..3ba9fc2 100644 --- a/test/Transforms/LoopUnswitch/2011-11-18-TwoSwitches.ll +++ b/test/Transforms/LoopUnswitch/2011-11-18-TwoSwitches.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt -loop-unswitch -loop-unswitch-threshold 1000 -disable-output -stats -info-output-file - < %s | FileCheck --check-prefix=STATS %s ; RUN: opt -S -loop-unswitch -loop-unswitch-threshold 1000 -verify-loop-info -verify-dom-info < %s | FileCheck %s diff --git a/test/Transforms/LoopUnswitch/infinite-loop.ll b/test/Transforms/LoopUnswitch/infinite-loop.ll index f3fba64..8261e38 100644 --- a/test/Transforms/LoopUnswitch/infinite-loop.ll +++ b/test/Transforms/LoopUnswitch/infinite-loop.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt -loop-unswitch -disable-output -stats -info-output-file - < %s | FileCheck --check-prefix=STATS %s ; RUN: opt -loop-unswitch -simplifycfg -S < %s | FileCheck %s ; PR5373 diff --git a/test/Transforms/LoopVectorize/X86/avx1.ll b/test/Transforms/LoopVectorize/X86/avx1.ll index a85c6fe..6c0366e 100644 --- a/test/Transforms/LoopVectorize/X86/avx1.ll +++ b/test/Transforms/LoopVectorize/X86/avx1.ll @@ -27,7 +27,7 @@ define i32 @read_mod_write_single_ptr(float* nocapture %a, i32 %n) nounwind uwta ;CHECK: @read_mod_i64 -;CHECK: load <4 x i64> +;CHECK: load <2 x i64> ;CHECK: ret i32 define i32 @read_mod_i64(i64* nocapture %a, i32 %n) nounwind uwtable ssp { %1 = icmp sgt i32 %n, 0 @@ -37,7 +37,7 @@ define i32 @read_mod_i64(i64* nocapture %a, i32 %n) nounwind uwtable ssp { %indvars.iv = phi i64 [ %indvars.iv.next, %.lr.ph ], [ 0, %0 ] %2 = getelementptr inbounds i64* %a, i64 %indvars.iv %3 = load i64* %2, align 4 - %4 = mul i64 %3, 3 + %4 = add i64 %3, 3 store i64 %4, i64* %2, align 4 %indvars.iv.next = add i64 %indvars.iv, 1 %lftr.wideiv = trunc i64 %indvars.iv.next to i32 diff --git a/test/Transforms/LoopVectorize/X86/reduction-crash.ll b/test/Transforms/LoopVectorize/X86/reduction-crash.ll new file mode 100644 index 0000000..f580846 --- /dev/null +++ b/test/Transforms/LoopVectorize/X86/reduction-crash.ll @@ -0,0 +1,35 @@ +; RUN: opt -S -loop-vectorize -mcpu=prescott < %s | FileCheck %s + +target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32-n8:16:32-S128" +target triple = "i386-apple-darwin" + +; PR15344 +define void @test1(float* nocapture %arg, i32 %arg1) nounwind { +; CHECK: @test1 +; CHECK: preheader +; CHECK: insertelement <2 x double> zeroinitializer, double %tmp, i32 0 +; CHECK: vector.memcheck + +bb: + br label %bb2 + +bb2: ; preds = %bb + %tmp = load double* null, align 8 + br i1 undef, label %bb3, label %bb12 + +bb3: ; preds = %bb3, %bb2 + %tmp4 = phi double [ %tmp9, %bb3 ], [ %tmp, %bb2 ] + %tmp5 = phi i32 [ %tmp8, %bb3 ], [ 0, %bb2 ] + %tmp6 = getelementptr inbounds [16 x double]* undef, i32 0, i32 %tmp5 + %tmp7 = load double* %tmp6, align 4 + %tmp8 = add nsw i32 %tmp5, 1 + %tmp9 = fadd fast double %tmp4, undef + %tmp10 = getelementptr inbounds float* %arg, i32 %tmp5 + store float undef, float* %tmp10, align 4 + %tmp11 = icmp eq i32 %tmp8, %arg1 + br i1 %tmp11, label %bb12, label %bb3 + +bb12: ; preds = %bb3, %bb2 + %tmp13 = phi double [ %tmp, %bb2 ], [ %tmp9, %bb3 ] + ret void +} diff --git a/test/Transforms/LoopVectorize/X86/vector-scalar-select-cost.ll b/test/Transforms/LoopVectorize/X86/vector-scalar-select-cost.ll new file mode 100644 index 0000000..3b3a787 --- /dev/null +++ b/test/Transforms/LoopVectorize/X86/vector-scalar-select-cost.ll @@ -0,0 +1,66 @@ +; RUN: opt < %s -loop-vectorize -mcpu=core2 -debug-only=loop-vectorize 2>&1 -S | FileCheck %s +; REQUIRES: asserts +; Make sure we use the right select kind when querying select costs. + +target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128" +target triple = "x86_64-apple-macosx10.8.0" + +@a = common global [2048 x i32] zeroinitializer, align 16 +@b = common global [2048 x i32] zeroinitializer, align 16 +@c = common global [2048 x i32] zeroinitializer, align 16 + +; CHECK: Checking a loop in "scalarselect" +define void @scalarselect(i1 %cond) { + br label %1 + +; <label>:1 + %indvars.iv = phi i64 [ 0, %0 ], [ %indvars.iv.next, %1 ] + %2 = getelementptr inbounds [2048 x i32]* @b, i64 0, i64 %indvars.iv + %3 = load i32* %2, align 4 + %4 = getelementptr inbounds [2048 x i32]* @c, i64 0, i64 %indvars.iv + %5 = load i32* %4, align 4 + %6 = add nsw i32 %5, %3 + %7 = getelementptr inbounds [2048 x i32]* @a, i64 0, i64 %indvars.iv + +; A scalar select has a cost of 1 on core2 +; CHECK: cost of 1 for VF 2 {{.*}} select i1 %cond, i32 %6, i32 0 + + %sel = select i1 %cond, i32 %6, i32 zeroinitializer + store i32 %sel, i32* %7, align 4 + %indvars.iv.next = add i64 %indvars.iv, 1 + %lftr.wideiv = trunc i64 %indvars.iv.next to i32 + %exitcond = icmp eq i32 %lftr.wideiv, 256 + br i1 %exitcond, label %8, label %1 + +; <label>:8 + ret void +} + +; CHECK: Checking a loop in "vectorselect" +define void @vectorselect(i1 %cond) { + br label %1 + +; <label>:1 + %indvars.iv = phi i64 [ 0, %0 ], [ %indvars.iv.next, %1 ] + %2 = getelementptr inbounds [2048 x i32]* @b, i64 0, i64 %indvars.iv + %3 = load i32* %2, align 4 + %4 = getelementptr inbounds [2048 x i32]* @c, i64 0, i64 %indvars.iv + %5 = load i32* %4, align 4 + %6 = add nsw i32 %5, %3 + %7 = getelementptr inbounds [2048 x i32]* @a, i64 0, i64 %indvars.iv + %8 = icmp ult i64 %indvars.iv, 8 + +; A vector select has a cost of 4 on core2 +; CHECK: cost of 4 for VF 2 {{.*}} select i1 %8, i32 %6, i32 0 + + %sel = select i1 %8, i32 %6, i32 zeroinitializer + store i32 %sel, i32* %7, align 4 + %indvars.iv.next = add i64 %indvars.iv, 1 + %lftr.wideiv = trunc i64 %indvars.iv.next to i32 + %exitcond = icmp eq i32 %lftr.wideiv, 256 + br i1 %exitcond, label %9, label %1 + +; <label>:9 + ret void +} + diff --git a/test/Transforms/LoopVectorize/dbg.value.ll b/test/Transforms/LoopVectorize/dbg.value.ll new file mode 100644 index 0000000..a2ea951 --- /dev/null +++ b/test/Transforms/LoopVectorize/dbg.value.ll @@ -0,0 +1,70 @@ +; RUN: opt < %s -S -loop-vectorize -force-vector-unroll=1 -force-vector-width=4 -dce -instcombine | FileCheck %s +; Make sure we vectorize with debugging turned on. + +target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128" +target triple = "x86_64-apple-macosx10.8.0" + +@A = global [1024 x i32] zeroinitializer, align 16 +@B = global [1024 x i32] zeroinitializer, align 16 +@C = global [1024 x i32] zeroinitializer, align 16 + +; CHECK: @test +define i32 @test() #0 { +entry: + tail call void @llvm.dbg.value(metadata !1, i64 0, metadata !9), !dbg !18 + br label %for.body, !dbg !18 + +for.body: + ;CHECK: load <4 x i32> + %indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ] + %arrayidx = getelementptr inbounds [1024 x i32]* @B, i64 0, i64 %indvars.iv, !dbg !19 + %0 = load i32* %arrayidx, align 4, !dbg !19, !tbaa !21 + %arrayidx2 = getelementptr inbounds [1024 x i32]* @C, i64 0, i64 %indvars.iv, !dbg !19 + %1 = load i32* %arrayidx2, align 4, !dbg !19, !tbaa !21 + %add = add nsw i32 %1, %0, !dbg !19 + %arrayidx4 = getelementptr inbounds [1024 x i32]* @A, i64 0, i64 %indvars.iv, !dbg !19 + store i32 %add, i32* %arrayidx4, align 4, !dbg !19, !tbaa !21 + %indvars.iv.next = add i64 %indvars.iv, 1, !dbg !18 + tail call void @llvm.dbg.value(metadata !{null}, i64 0, metadata !9), !dbg !18 + %lftr.wideiv = trunc i64 %indvars.iv.next to i32, !dbg !18 + %exitcond = icmp ne i32 %lftr.wideiv, 1024, !dbg !18 + br i1 %exitcond, label %for.body, label %for.end, !dbg !18 + +for.end: + ret i32 0, !dbg !24 +} + +declare void @llvm.dbg.declare(metadata, metadata) #1 + +declare void @llvm.dbg.value(metadata, i64, metadata) #1 + +attributes #0 = { nounwind ssp uwtable "fp-contract-model"="standard" "no-frame-pointer-elim" "no-frame-pointer-elim-non-leaf" "realign-stack" "relocation-model"="pic" "ssp-buffers-size"="8" } +attributes #1 = { nounwind readnone } + +!llvm.dbg.cu = !{!0} + +!0 = metadata !{i32 786449, i32 0, i32 4, metadata !"test", metadata !"/path/to/somewhere", metadata !"clang", i1 true, i1 true, metadata !"", i32 0, metadata !1, metadata !1, metadata !2, metadata !11, metadata !""} +!1 = metadata !{i32 0} +!2 = metadata !{metadata !3} +!3 = metadata !{i32 786478, i32 0, metadata !4, metadata !"test", metadata !"test", metadata !"test", metadata !4, i32 5, metadata !5, i1 false, i1 true, i32 0, i32 0, null, i32 256, i1 true, i32 ()* @test, null, null, metadata !8, i32 5} +!4 = metadata !{i32 786473, metadata !"test", metadata !"/path/to/somewhere", null} +!5 = metadata !{i32 786453, i32 0, metadata !"", i32 0, i32 0, i64 0, i64 0, i64 0, i32 0, null, metadata !6, i32 0, i32 0} +!6 = metadata !{metadata !7} +!7 = metadata !{i32 786468, null, metadata !"int", null, i32 0, i64 32, i64 32, i64 0, i32 0, i32 5} +!8 = metadata !{metadata !9} +!9 = metadata !{i32 786688, metadata !10, metadata !"i", metadata !4, i32 6, metadata !7, i32 0, i32 0} +!10 = metadata !{i32 786443, metadata !3, i32 6, i32 0, metadata !4, i32 0} +!11 = metadata !{metadata !12, metadata !16, metadata !17} +!12 = metadata !{i32 786484, i32 0, null, metadata !"A", metadata !"A", metadata !"", metadata !4, i32 1, metadata !13, i32 0, i32 1, [1024 x i32]* @A, null} +!13 = metadata !{i32 786433, null, metadata !"", null, i32 0, i64 32768, i64 32, i32 0, i32 0, metadata !7, metadata !14, i32 0, i32 0} +!14 = metadata !{metadata !15} +!15 = metadata !{i32 786465, i64 0, i64 1024} +!16 = metadata !{i32 786484, i32 0, null, metadata !"B", metadata !"B", metadata !"", metadata !4, i32 2, metadata !13, i32 0, i32 1, [1024 x i32]* @B, null} +!17 = metadata !{i32 786484, i32 0, null, metadata !"C", metadata !"C", metadata !"", metadata !4, i32 3, metadata !13, i32 0, i32 1, [1024 x i32]* @C, null} +!18 = metadata !{i32 6, i32 0, metadata !10, null} +!19 = metadata !{i32 7, i32 0, metadata !20, null} +!20 = metadata !{i32 786443, metadata !10, i32 6, i32 0, metadata !4, i32 1} +!21 = metadata !{metadata !"int", metadata !22} +!22 = metadata !{metadata !"omnipotent char", metadata !23} +!23 = metadata !{metadata !"Simple C/C++ TBAA"} +!24 = metadata !{i32 9, i32 0, metadata !3, null} diff --git a/test/Transforms/LoopVectorize/global_alias.ll b/test/Transforms/LoopVectorize/global_alias.ll index 24e698b..121da8b 100644 --- a/test/Transforms/LoopVectorize/global_alias.ll +++ b/test/Transforms/LoopVectorize/global_alias.ll @@ -24,7 +24,7 @@ target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f3 ; } ; CHECK: define i32 @noAlias01 ; CHECK: add nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @noAlias01(i32 %a) nounwind { entry: @@ -72,7 +72,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias02 ; CHECK: add nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @noAlias02(i32 %a) { entry: @@ -121,7 +121,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias03 ; CHECK: add nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @noAlias03(i32 %a) { entry: @@ -170,7 +170,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias04 ; CHECK-NOT: add nsw <4 x i32> -; CHECK ret +; CHECK: ret ; ; TODO: This test vectorizes (with run-time check) on real targets with -O3) ; Check why it's not being vectorized even when forcing vectorization @@ -224,7 +224,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias05 ; CHECK: add nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @noAlias05(i32 %a) #0 { entry: @@ -280,7 +280,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias06 ; CHECK: add nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @noAlias06(i32 %a) #0 { entry: @@ -337,7 +337,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias07 ; CHECK: sub nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @noAlias07(i32 %a) #0 { entry: @@ -389,7 +389,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias08 ; CHECK: sub nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @noAlias08(i32 %a) #0 { entry: @@ -441,7 +441,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias09 ; CHECK: sub nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @noAlias09(i32 %a) #0 { entry: @@ -493,7 +493,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias10 ; CHECK-NOT: sub nsw <4 x i32> -; CHECK ret +; CHECK: ret ; ; TODO: This test vectorizes (with run-time check) on real targets with -O3) ; Check why it's not being vectorized even when forcing vectorization @@ -553,7 +553,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias11 ; CHECK: sub nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @noAlias11(i32 %a) #0 { entry: @@ -613,7 +613,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias12 ; CHECK: sub nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @noAlias12(i32 %a) #0 { entry: @@ -674,7 +674,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias13 ; CHECK: add nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @noAlias13(i32 %a) #0 { entry: @@ -723,7 +723,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @noAlias14 ; CHECK: sub nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @noAlias14(i32 %a) #0 { entry: @@ -779,7 +779,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @mayAlias01 ; CHECK-NOT: add nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @mayAlias01(i32 %a) nounwind { entry: @@ -829,7 +829,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @mayAlias02 ; CHECK-NOT: add nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @mayAlias02(i32 %a) nounwind { entry: @@ -879,7 +879,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @mayAlias03 ; CHECK-NOT: add nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @mayAlias03(i32 %a) nounwind { entry: @@ -936,7 +936,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @mustAlias01 ; CHECK-NOT: add nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @mustAlias01(i32 %a) nounwind { entry: @@ -986,7 +986,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @mustAlias02 ; CHECK-NOT: add nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @mustAlias02(i32 %a) nounwind { entry: @@ -1035,7 +1035,7 @@ for.end: ; preds = %for.cond ; } ; CHECK: define i32 @mustAlias03 ; CHECK-NOT: add nsw <4 x i32> -; CHECK ret +; CHECK: ret define i32 @mustAlias03(i32 %a) nounwind { entry: diff --git a/test/Transforms/LoopVectorize/intrinsic.ll b/test/Transforms/LoopVectorize/intrinsic.ll index 7d5a5d7..e79d78d 100644 --- a/test/Transforms/LoopVectorize/intrinsic.ll +++ b/test/Transforms/LoopVectorize/intrinsic.ll @@ -902,6 +902,30 @@ for.end: ; preds = %for.body, %entry ret void } +; CHECK: fabs_libm +; CHECK: call <4 x float> @llvm.fabs.v4f32 +; CHECK: ret void +define void @fabs_libm(float* nocapture %x) nounwind { +entry: + br label %for.body + +for.body: ; preds = %entry, %for.body + %indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ] + %arrayidx = getelementptr inbounds float* %x, i64 %indvars.iv + %0 = load float* %arrayidx, align 4 + %call = tail call float @fabsf(float %0) nounwind readnone + store float %call, float* %arrayidx, align 4 + %indvars.iv.next = add i64 %indvars.iv, 1 + %lftr.wideiv = trunc i64 %indvars.iv.next to i32 + %exitcond = icmp eq i32 %lftr.wideiv, 1024 + br i1 %exitcond, label %for.end, label %for.body + +for.end: ; preds = %for.body + ret void +} + +declare float @fabsf(float) nounwind readnone + declare double @llvm.pow.f64(double, double) nounwind readnone !0 = metadata !{metadata !"float", metadata !1} diff --git a/test/Transforms/LoopVectorize/phi-hang.ll b/test/Transforms/LoopVectorize/phi-hang.ll new file mode 100644 index 0000000..b80d459 --- /dev/null +++ b/test/Transforms/LoopVectorize/phi-hang.ll @@ -0,0 +1,29 @@ +; RUN: opt -S -loop-vectorize < %s + +; PR15384 +define void @test1(i32 %arg) { +bb: + br label %bb1 + +bb1: ; preds = %bb5, %bb + %tmp = phi i32 [ 1, %bb ], [ %tmp7, %bb5 ] + %tmp2 = phi i32 [ %arg, %bb ], [ %tmp9, %bb5 ] + br i1 true, label %bb5, label %bb3 + +bb3: ; preds = %bb1 + br label %bb4 + +bb4: ; preds = %bb3 + br label %bb5 + +bb5: ; preds = %bb4, %bb1 + %tmp6 = phi i32 [ 0, %bb4 ], [ %tmp, %bb1 ] + %tmp7 = phi i32 [ 0, %bb4 ], [ %tmp6, %bb1 ] + %tmp8 = phi i32 [ 0, %bb4 ], [ %tmp, %bb1 ] + %tmp9 = add nsw i32 %tmp2, 1 + %tmp10 = icmp eq i32 %tmp9, 0 + br i1 %tmp10, label %bb11, label %bb1 + +bb11: ; preds = %bb5 + ret void +} diff --git a/test/Transforms/LoopVectorize/vectorize-once.ll b/test/Transforms/LoopVectorize/vectorize-once.ll new file mode 100644 index 0000000..ac16948 --- /dev/null +++ b/test/Transforms/LoopVectorize/vectorize-once.ll @@ -0,0 +1,75 @@ +; RUN: opt < %s -loop-vectorize -force-vector-unroll=1 -force-vector-width=4 -dce -instcombine -S -simplifycfg | FileCheck %s + +target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128" +target triple = "x86_64-apple-macosx10.8.0" + +; +; We want to make sure that we are vectorizeing the scalar loop only once +; even if the pass manager runs the vectorizer multiple times due to inlining. + + +; This test checks that we add metadata to vectorized loops +; CHECK: _Z4foo1Pii +; CHECK: <4 x i32> +; CHECK: llvm.vectorizer.already_vectorized +; CHECK: ret + +; This test comes from the loop: +; +;int foo (int *A, int n) { +; return std::accumulate(A, A + n, 0); +;} +define i32 @_Z4foo1Pii(i32* %A, i32 %n) #0 { +entry: + %idx.ext = sext i32 %n to i64 + %add.ptr = getelementptr inbounds i32* %A, i64 %idx.ext + %cmp3.i = icmp eq i32 %n, 0 + br i1 %cmp3.i, label %_ZSt10accumulateIPiiET0_T_S2_S1_.exit, label %for.body.i + +for.body.i: ; preds = %entry, %for.body.i + %__init.addr.05.i = phi i32 [ %add.i, %for.body.i ], [ 0, %entry ] + %__first.addr.04.i = phi i32* [ %incdec.ptr.i, %for.body.i ], [ %A, %entry ] + %0 = load i32* %__first.addr.04.i, align 4, !tbaa !0 + %add.i = add nsw i32 %0, %__init.addr.05.i + %incdec.ptr.i = getelementptr inbounds i32* %__first.addr.04.i, i64 1 + %cmp.i = icmp eq i32* %incdec.ptr.i, %add.ptr + br i1 %cmp.i, label %_ZSt10accumulateIPiiET0_T_S2_S1_.exit, label %for.body.i + +_ZSt10accumulateIPiiET0_T_S2_S1_.exit: ; preds = %for.body.i, %entry + %__init.addr.0.lcssa.i = phi i32 [ 0, %entry ], [ %add.i, %for.body.i ] + ret i32 %__init.addr.0.lcssa.i +} + +; This test checks that we don't vectorize loops that are marked with the "already vectorized" metadata. +; CHECK: _Z4foo2Pii +; CHECK-NOT: <4 x i32> +; CHECK: llvm.vectorizer.already_vectorized +; CHECK: ret +define i32 @_Z4foo2Pii(i32* %A, i32 %n) #0 { +entry: + %idx.ext = sext i32 %n to i64 + %add.ptr = getelementptr inbounds i32* %A, i64 %idx.ext + %cmp3.i = icmp eq i32 %n, 0 + br i1 %cmp3.i, label %_ZSt10accumulateIPiiET0_T_S2_S1_.exit, label %for.body.i + +for.body.i: ; preds = %entry, %for.body.i + %__init.addr.05.i = phi i32 [ %add.i, %for.body.i ], [ 0, %entry ] + %__first.addr.04.i = phi i32* [ %incdec.ptr.i, %for.body.i ], [ %A, %entry ] + %0 = load i32* %__first.addr.04.i, align 4, !tbaa !0 + %add.i = add nsw i32 %0, %__init.addr.05.i + %incdec.ptr.i = getelementptr inbounds i32* %__first.addr.04.i, i64 1 + %cmp.i = icmp eq i32* %incdec.ptr.i, %add.ptr + br i1 %cmp.i, label %_ZSt10accumulateIPiiET0_T_S2_S1_.exit, label %for.body.i, !llvm.vectorizer.already_vectorized !3 + +_ZSt10accumulateIPiiET0_T_S2_S1_.exit: ; preds = %for.body.i, %entry + %__init.addr.0.lcssa.i = phi i32 [ 0, %entry ], [ %add.i, %for.body.i ] + ret i32 %__init.addr.0.lcssa.i +} + +attributes #0 = { nounwind readonly ssp uwtable "fp-contract-model"="standard" "no-frame-pointer-elim" "no-frame-pointer-elim-non-leaf" "realign-stack" "relocation-model"="pic" "ssp-buffers-size"="8" } + +!0 = metadata !{metadata !"int", metadata !1} +!1 = metadata !{metadata !"omnipotent char", metadata !2} +!2 = metadata !{metadata !"Simple C/C++ TBAA"} +!3 = metadata !{} + diff --git a/test/Transforms/Mem2Reg/ConvertDebugInfo.ll b/test/Transforms/Mem2Reg/ConvertDebugInfo.ll index 2f1ccb4..8abd851 100644 --- a/test/Transforms/Mem2Reg/ConvertDebugInfo.ll +++ b/test/Transforms/Mem2Reg/ConvertDebugInfo.ll @@ -8,8 +8,10 @@ entry: %0 = alloca double ; <double*> [#uses=2] %"alloca point" = bitcast i32 0 to i32 ; <i32> [#uses=0] call void @llvm.dbg.declare(metadata !{i32* %i_addr}, metadata !0), !dbg !8 -; CHECK: call void @llvm.dbg.value(metadata !{i32 %i}, i64 0, metadata !0) -; CHECK: call void @llvm.dbg.value(metadata !{double %j}, i64 0, metadata !9) +; CHECK: call void @llvm.dbg.value(metadata !{i32 %i}, i64 0, metadata ![[IVAR:[0-9]*]]) +; CHECK: call void @llvm.dbg.value(metadata !{double %j}, i64 0, metadata ![[JVAR:[0-9]*]]) +; CHECK: ![[IVAR]] = {{.*}} ; [ DW_TAG_arg_variable ] [i] +; CHECK: ![[JVAR]] = {{.*}} ; [ DW_TAG_arg_variable ] [j] store i32 %i, i32* %i_addr call void @llvm.dbg.declare(metadata !{double* %j_addr}, metadata !9), !dbg !8 store double %j, double* %j_addr @@ -30,16 +32,18 @@ return: ; preds = %entry declare void @llvm.dbg.declare(metadata, metadata) nounwind readnone -!0 = metadata !{i32 524545, metadata !1, metadata !"i", metadata !2, i32 2, metadata !7} ; [ DW_TAG_arg_variable ] -!1 = metadata !{i32 524334, i32 0, metadata !2, metadata !"testfunc", metadata !"testfunc", metadata !"testfunc", metadata !2, i32 2, metadata !4, i1 false, i1 true, i32 0, i32 0, null, i1 false, i1 false} ; [ DW_TAG_subprogram ] -!2 = metadata !{i32 524329, metadata !"testfunc.c", metadata !"/tmp", metadata !3} ; [ DW_TAG_file_type ] -!3 = metadata !{i32 524305, i32 0, i32 1, metadata !"testfunc.c", metadata !"/tmp", metadata !"4.2.1 (Based on Apple Inc. build 5658) (LLVM build)", i1 true, i1 false, metadata !"", i32 0} ; [ DW_TAG_compile_unit ] -!4 = metadata !{i32 524309, metadata !2, metadata !"", metadata !2, i32 0, i64 0, i64 0, i64 0, i32 0, null, metadata !5, i32 0, null} ; [ DW_TAG_subroutine_type ] +!llvm.dbg.cu = !{!3} + +!0 = metadata !{i32 786689, metadata !1, metadata !"i", metadata !2, i32 2, metadata !7, i32 0, null} ; [ DW_TAG_arg_variable ] +!1 = metadata !{i32 786478, i32 0, metadata !2, metadata !"testfunc", metadata !"testfunc", metadata !"testfunc", metadata !2, i32 2, metadata !4, i1 false, i1 true, i32 0, i32 0, null, i32 0, i1 false, double (i32, double)* @testfunc, null, null, null, i32 2} ; [ DW_TAG_subprogram ] +!2 = metadata !{i32 786473, metadata !12} ; [ DW_TAG_file_type ] +!3 = metadata !{i32 786449, i32 0, i32 1, metadata !"testfunc.c", metadata !"/tmp", metadata !"4.2.1 (Based on Apple Inc. build 5658) (LLVM build)", i1 true, i1 false, metadata !"", i32 0} ; [ DW_TAG_compile_unit ] +!4 = metadata !{i32 786453, metadata !2, metadata !"", metadata !2, i32 0, i64 0, i64 0, i64 0, i32 0, null, metadata !5, i32 0, null} ; [ DW_TAG_subroutine_type ] !5 = metadata !{metadata !6, metadata !7, metadata !6} -!6 = metadata !{i32 524324, metadata !2, metadata !"double", metadata !2, i32 0, i64 64, i64 64, i64 0, i32 0, i32 4} ; [ DW_TAG_base_type ] -!7 = metadata !{i32 524324, metadata !2, metadata !"int", metadata !2, i32 0, i64 32, i64 32, i64 0, i32 0, i32 5} ; [ DW_TAG_base_type ] +!6 = metadata !{i32 786468, metadata !2, metadata !"double", metadata !2, i32 0, i64 64, i64 64, i64 0, i32 0, i32 4} ; [ DW_TAG_base_type ] +!7 = metadata !{i32 786468, metadata !2, metadata !"int", metadata !2, i32 0, i64 32, i64 32, i64 0, i32 0, i32 5} ; [ DW_TAG_base_type ] !8 = metadata !{i32 2, i32 0, metadata !1, null} -!9 = metadata !{i32 524545, metadata !1, metadata !"j", metadata !2, i32 2, metadata !6} ; [ DW_TAG_arg_variable ] +!9 = metadata !{i32 786689, metadata !1, metadata !"j", metadata !2, i32 2, metadata !6, i32 0, null} ; [ DW_TAG_arg_variable ] !10 = metadata !{i32 3, i32 0, metadata !11, null} -!11 = metadata !{i32 524299, metadata !1, i32 2, i32 0} ; [ DW_TAG_lexical_block ] - +!11 = metadata !{i32 786443, metadata !1, i32 2, i32 0} ; [ DW_TAG_lexical_block ] +!12 = metadata !{metadata !"testfunc.c", metadata !"/tmp"} diff --git a/test/Transforms/Mem2Reg/ConvertDebugInfo2.ll b/test/Transforms/Mem2Reg/ConvertDebugInfo2.ll index 4cb621f..f601532 100644 --- a/test/Transforms/Mem2Reg/ConvertDebugInfo2.ll +++ b/test/Transforms/Mem2Reg/ConvertDebugInfo2.ll @@ -30,23 +30,24 @@ return: ; preds = %entry ret void, !dbg !19 } -!0 = metadata !{i32 524545, metadata !1, metadata !"a", metadata !2, i32 8, metadata !6} ; [ DW_TAG_arg_variable ] -!1 = metadata !{i32 524334, i32 0, metadata !2, metadata !"baz", metadata !"baz", metadata !"baz", metadata !2, i32 8, metadata !4, i1 false, i1 true, i32 0, i32 0, null, i1 false, i1 false} ; [ DW_TAG_subprogram ] -!2 = metadata !{i32 524329, metadata !"bar.c", metadata !"/tmp/", metadata !3} ; [ DW_TAG_file_type ] -!3 = metadata !{i32 524305, i32 0, i32 1, metadata !"bar.c", metadata !"/tmp/", metadata !"4.2.1 (Based on Apple Inc. build 5658) (LLVM build)", i1 true, i1 false, metadata !"", i32 0} ; [ DW_TAG_compile_unit ] -!4 = metadata !{i32 524309, metadata !2, metadata !"", metadata !2, i32 0, i64 0, i64 0, i64 0, i32 0, null, metadata !5, i32 0, null} ; [ DW_TAG_subroutine_type ] +!0 = metadata !{i32 786689, metadata !1, metadata !"a", metadata !2, i32 8, metadata !6, i32 0, null} ; [ DW_TAG_arg_variable ] +!1 = metadata !{i32 786478, i32 0, metadata !2, metadata !"baz", metadata !"baz", metadata !"baz", metadata !2, i32 8, metadata !4, i1 false, i1 true, i32 0, i32 0, null, i32 0, i1 false, void (i32)* @baz, null, null, null, i32 8} ; [ DW_TAG_subprogram ] +!2 = metadata !{i32 786473, metadata !20} ; [ DW_TAG_file_type ] +!3 = metadata !{i32 786449, i32 0, i32 1, metadata !"bar.c", metadata !"/tmp/", metadata !"4.2.1 (Based on Apple Inc. build 5658) (LLVM build)", i1 true, i1 false, metadata !"", i32 0} ; [ DW_TAG_compile_unit ] +!4 = metadata !{i32 786453, metadata !2, metadata !"", metadata !2, i32 0, i64 0, i64 0, i64 0, i32 0, null, metadata !5, i32 0, null} ; [ DW_TAG_subroutine_type ] !5 = metadata !{null, metadata !6} -!6 = metadata !{i32 524324, metadata !2, metadata !"int", metadata !2, i32 0, i64 32, i64 32, i64 0, i32 0, i32 5} ; [ DW_TAG_base_type ] +!6 = metadata !{i32 786468, metadata !2, metadata !"int", metadata !2, i32 0, i64 32, i64 32, i64 0, i32 0, i32 5} ; [ DW_TAG_base_type ] !7 = metadata !{i32 8, i32 0, metadata !1, null} !8 = metadata !{i32 9, i32 0, metadata !1, null} -!9 = metadata !{i32 524545, metadata !10, metadata !"x", metadata !2, i32 4, metadata !6} ; [ DW_TAG_arg_variable ] -!10 = metadata !{i32 524334, i32 0, metadata !2, metadata !"bar", metadata !"bar", metadata !"bar", metadata !2, i32 4, metadata !11, i1 true, i1 true, i32 0, i32 0, null, i1 false, i1 false} ; [ DW_TAG_subprogram ] -!11 = metadata !{i32 524309, metadata !2, metadata !"", metadata !2, i32 0, i64 0, i64 0, i64 0, i32 0, null, metadata !12, i32 0, null} ; [ DW_TAG_subroutine_type ] +!9 = metadata !{i32 786689, metadata !10, metadata !"x", metadata !2, i32 4, metadata !6, i32 0, null} ; [ DW_TAG_arg_variable ] +!10 = metadata !{i32 786478, i32 0, metadata !2, metadata !"bar", metadata !"bar", metadata !"bar", metadata !2, i32 4, metadata !11, i1 true, i1 true, i32 0, i32 0, null, i1 false, i1 false, null, null, null, null, i32 4} ; [ DW_TAG_subprogram ] +!11 = metadata !{i32 786453, metadata !2, metadata !"", metadata !2, i32 0, i64 0, i64 0, i64 0, i32 0, null, metadata !12, i32 0, null} ; [ DW_TAG_subroutine_type ] !12 = metadata !{null, metadata !6, metadata !13, metadata !14} -!13 = metadata !{i32 524324, metadata !2, metadata !"long int", metadata !2, i32 0, i64 64, i64 64, i64 0, i32 0, i32 5} ; [ DW_TAG_base_type ] -!14 = metadata !{i32 524303, metadata !2, metadata !"", metadata !2, i32 0, i64 64, i64 64, i64 0, i32 0, null} ; [ DW_TAG_pointer_type ] +!13 = metadata !{i32 786468, metadata !2, metadata !"long int", metadata !2, i32 0, i64 64, i64 64, i64 0, i32 0, i32 5} ; [ DW_TAG_base_type ] +!14 = metadata !{i32 786447, metadata !2, metadata !"", metadata !2, i32 0, i64 64, i64 64, i64 0, i32 0, null} ; [ DW_TAG_pointer_type ] !15 = metadata !{i32 4, i32 0, metadata !10, metadata !8} -!16 = metadata !{i32 524545, metadata !10, metadata !"y", metadata !2, i32 4, metadata !13} ; [ DW_TAG_arg_variable ] -!17 = metadata !{i32 524545, metadata !10, metadata !"z", metadata !2, i32 4, metadata !14} ; [ DW_TAG_arg_variable ] +!16 = metadata !{i32 786689, metadata !10, metadata !"y", metadata !2, i32 4, metadata !13, i32 0, null} ; [ DW_TAG_arg_variable ] +!17 = metadata !{i32 786689, metadata !10, metadata !"z", metadata !2, i32 4, metadata !14, i32 0, null} ; [ DW_TAG_arg_variable ] !18 = metadata !{i32 5, i32 0, metadata !10, metadata !8} !19 = metadata !{i32 10, i32 0, metadata !1, null} +!20 = metadata !{metadata !"bar.c", metadata !"/tmp/"} diff --git a/test/Transforms/MergeFunc/phi-speculation1.ll b/test/Transforms/MergeFunc/phi-speculation1.ll index fd0baff..548e510 100644 --- a/test/Transforms/MergeFunc/phi-speculation1.ll +++ b/test/Transforms/MergeFunc/phi-speculation1.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt < %s -mergefunc -stats -disable-output 2>&1 | not grep "functions merged" define i32 @foo1(i32 %x) { diff --git a/test/Transforms/MergeFunc/phi-speculation2.ll b/test/Transforms/MergeFunc/phi-speculation2.ll index eec8b5c..d42a465 100644 --- a/test/Transforms/MergeFunc/phi-speculation2.ll +++ b/test/Transforms/MergeFunc/phi-speculation2.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt < %s -mergefunc -stats -disable-output 2>&1 | grep "functions merged" define i32 @foo1(i32 %x) { diff --git a/test/Transforms/MergeFunc/vector.ll b/test/Transforms/MergeFunc/vector.ll index 4af079f..dba5fa3 100644 --- a/test/Transforms/MergeFunc/vector.ll +++ b/test/Transforms/MergeFunc/vector.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt -mergefunc -stats -disable-output < %s 2>&1 | grep "functions merged" ; This test is checks whether we can merge diff --git a/test/Transforms/MergeFunc/vectors-and-arrays.ll b/test/Transforms/MergeFunc/vectors-and-arrays.ll index dc64a08..2274722 100644 --- a/test/Transforms/MergeFunc/vectors-and-arrays.ll +++ b/test/Transforms/MergeFunc/vectors-and-arrays.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt -mergefunc < %s -disable-output -stats | not grep merged ; This used to crash with an assert. diff --git a/test/Transforms/ObjCARC/dont-infinite-loop-during-block-escape-analysis.ll b/test/Transforms/ObjCARC/dont-infinite-loop-during-block-escape-analysis.ll deleted file mode 100644 index bdee2be..0000000 --- a/test/Transforms/ObjCARC/dont-infinite-loop-during-block-escape-analysis.ll +++ /dev/null @@ -1,87 +0,0 @@ -; RUN: opt -S -objc-arc < %s -; bugzilla://14551 -; rdar://12851911 - -; Make sure that we do not hang clang during escape analysis. - -target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128" -target triple = "x86_64-darwin" - -%struct.__block_descriptor = type { i64, i64 } -%struct.__block_byref_foo = type { i8*, %struct.__block_byref_foo*, i32, i32, i32 } - -@_NSConcreteGlobalBlock = external global i8* -@.str = private unnamed_addr constant [6 x i8] c"v8@?0\00", align 1 -@__block_descriptor_tmp = internal constant { i64, i64, i8*, i8* } { i64 0, i64 32, i8* getelementptr inbounds ([6 x i8]* @.str, i32 0, i32 0), i8* null } -@__block_literal_global = internal constant { i8**, i32, i32, i8*, %struct.__block_descriptor* } { i8** @_NSConcreteGlobalBlock, i32 1342177280, i32 0, i8* bitcast (void (i8*)* @__hang_clang_block_invoke to i8*), %struct.__block_descriptor* bitcast ({ i64, i64, i8*, i8* }* @__block_descriptor_tmp to %struct.__block_descriptor*) }, align 8 - -define void @hang_clang() uwtable optsize ssp { -entry: - %foo = alloca %struct.__block_byref_foo, align 8 - %byref.isa = getelementptr inbounds %struct.__block_byref_foo* %foo, i64 0, i32 0 - store i8* null, i8** %byref.isa, align 8 - %byref.forwarding = getelementptr inbounds %struct.__block_byref_foo* %foo, i64 0, i32 1 - store %struct.__block_byref_foo* %foo, %struct.__block_byref_foo** %byref.forwarding, align 8 - %byref.flags = getelementptr inbounds %struct.__block_byref_foo* %foo, i64 0, i32 2 - store i32 536870912, i32* %byref.flags, align 8 - %byref.size = getelementptr inbounds %struct.__block_byref_foo* %foo, i64 0, i32 3 - store i32 32, i32* %byref.size, align 4 - %foo1 = getelementptr inbounds %struct.__block_byref_foo* %foo, i64 0, i32 4 - store i32 0, i32* %foo1, align 8, !tbaa !4 - br label %for.body - -for.body: ; preds = %for.inc.for.body_crit_edge, %entry - %0 = phi i1 [ true, %entry ], [ %phitmp, %for.inc.for.body_crit_edge ] - %i.06 = phi i32 [ 1, %entry ], [ %phitmp8, %for.inc.for.body_crit_edge ] - %block.05 = phi void (...)* [ null, %entry ], [ %block.1, %for.inc.for.body_crit_edge ] - br i1 %0, label %for.inc, label %if.then - -if.then: ; preds = %for.body - %1 = call i8* @objc_retainBlock(i8* bitcast ({ i8**, i32, i32, i8*, %struct.__block_descriptor* }* @__block_literal_global to i8*)) nounwind, !clang.arc.copy_on_escape !7 - %2 = bitcast i8* %1 to void (...)* - %3 = bitcast void (...)* %block.05 to i8* - call void @objc_release(i8* %3) nounwind, !clang.imprecise_release !7 - br label %for.inc - -for.inc: ; preds = %for.body, %if.then - %block.1 = phi void (...)* [ %2, %if.then ], [ %block.05, %for.body ] - %exitcond = icmp eq i32 %i.06, 10 - br i1 %exitcond, label %for.end, label %for.inc.for.body_crit_edge - -for.inc.for.body_crit_edge: ; preds = %for.inc - %.pre = load %struct.__block_byref_foo** %byref.forwarding, align 8 - %foo2.phi.trans.insert = getelementptr inbounds %struct.__block_byref_foo* %.pre, i64 0, i32 4 - %.pre7 = load i32* %foo2.phi.trans.insert, align 4, !tbaa !4 - %phitmp = icmp eq i32 %.pre7, 0 - %phitmp8 = add i32 %i.06, 1 - br label %for.body - -for.end: ; preds = %for.inc - %4 = bitcast %struct.__block_byref_foo* %foo to i8* - call void @_Block_object_dispose(i8* %4, i32 8) - %5 = bitcast void (...)* %block.1 to i8* - call void @objc_release(i8* %5) nounwind, !clang.imprecise_release !7 - ret void -} - -define internal void @__hang_clang_block_invoke(i8* nocapture %.block_descriptor) nounwind uwtable readnone optsize ssp { -entry: - ret void -} - -declare i8* @objc_retainBlock(i8*) - -declare void @objc_release(i8*) nonlazybind - -declare void @_Block_object_dispose(i8*, i32) - -!llvm.module.flags = !{!0, !1, !2, !3} - -!0 = metadata !{i32 1, metadata !"Objective-C Version", i32 2} -!1 = metadata !{i32 1, metadata !"Objective-C Image Info Version", i32 0} -!2 = metadata !{i32 1, metadata !"Objective-C Image Info Section", metadata !"__DATA, __objc_imageinfo, regular, no_dead_strip"} -!3 = metadata !{i32 4, metadata !"Objective-C Garbage Collection", i32 0} -!4 = metadata !{metadata !"int", metadata !5} -!5 = metadata !{metadata !"omnipotent char", metadata !6} -!6 = metadata !{metadata !"Simple C/C++ TBAA"} -!7 = metadata !{} diff --git a/test/Transforms/ObjCARC/retain-block-escape-analysis.ll b/test/Transforms/ObjCARC/retain-block-escape-analysis.ll new file mode 100644 index 0000000..2c1ddce --- /dev/null +++ b/test/Transforms/ObjCARC/retain-block-escape-analysis.ll @@ -0,0 +1,127 @@ +; RUN: opt -S -objc-arc < %s | FileCheck %s + +declare i8* @objc_retain(i8*) nonlazybind +declare void @objc_release(i8*) nonlazybind +declare i8* @objc_retainBlock(i8*) + +;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; +; Use by an instruction which copies the value is an escape if the ; +; result is an escape. The current instructions with this property are: ; +; ; +; 1. BitCast. ; +; 2. GEP. ; +; 3. PhiNode. ; +; 4. SelectInst. ; +; ; +; Make sure that such instructions do not confuse the optimizer into removing ; +; an objc_retainBlock that is needed. ; +; ; +; rdar://13273675. (With extra test cases to handle bitcast, phi, and select. ; +;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; + +define void @bitcasttest(i8* %storage, void (...)* %block) { +; CHECK: define void @bitcasttest +entry: + %t1 = bitcast void (...)* %block to i8* +; CHECK-NOT: tail call i8* @objc_retain + %t2 = tail call i8* @objc_retain(i8* %t1) +; CHECK: tail call i8* @objc_retainBlock + %t3 = tail call i8* @objc_retainBlock(i8* %t1), !clang.arc.copy_on_escape !0 + %t4 = bitcast i8* %storage to void (...)** + %t5 = bitcast i8* %t3 to void (...)* + store void (...)* %t5, void (...)** %t4, align 8 +; CHECK-NOT: call void @objc_release + call void @objc_release(i8* %t1) + ret void +} + +define void @geptest(void (...)** %storage_array, void (...)* %block) { +; CHECK: define void @geptest +entry: + %t1 = bitcast void (...)* %block to i8* +; CHECK-NOT: tail call i8* @objc_retain + %t2 = tail call i8* @objc_retain(i8* %t1) +; CHECK: tail call i8* @objc_retainBlock + %t3 = tail call i8* @objc_retainBlock(i8* %t1), !clang.arc.copy_on_escape !0 + %t4 = bitcast i8* %t3 to void (...)* + + %storage = getelementptr inbounds void (...)** %storage_array, i64 0 + + store void (...)* %t4, void (...)** %storage, align 8 +; CHECK-NOT: call void @objc_release + call void @objc_release(i8* %t1) + ret void +} + +define void @selecttest(void (...)** %store1, void (...)** %store2, + void (...)* %block) { +; CHECK: define void @selecttest +entry: + %t1 = bitcast void (...)* %block to i8* +; CHECK-NOT: tail call i8* @objc_retain + %t2 = tail call i8* @objc_retain(i8* %t1) +; CHECK: tail call i8* @objc_retainBlock + %t3 = tail call i8* @objc_retainBlock(i8* %t1), !clang.arc.copy_on_escape !0 + %t4 = bitcast i8* %t3 to void (...)* + %store = select i1 undef, void (...)** %store1, void (...)** %store2 + store void (...)* %t4, void (...)** %store, align 8 +; CHECK-NOT: call void @objc_release + call void @objc_release(i8* %t1) + ret void +} + +define void @phinodetest(void (...)** %storage1, + void (...)** %storage2, + void (...)* %block) { +; CHECK: define void @phinodetest +entry: + %t1 = bitcast void (...)* %block to i8* +; CHECK-NOT: tail call i8* @objc_retain + %t2 = tail call i8* @objc_retain(i8* %t1) +; CHECK: tail call i8* @objc_retainBlock + %t3 = tail call i8* @objc_retainBlock(i8* %t1), !clang.arc.copy_on_escape !0 + %t4 = bitcast i8* %t3 to void (...)* + br i1 undef, label %store1_set, label %store2_set + +store1_set: + br label %end + +store2_set: + br label %end + +end: + %storage = phi void (...)** [ %storage1, %store1_set ], [ %storage2, %store2_set] + store void (...)* %t4, void (...)** %storage, align 8 +; CHECK-NOT: call void @objc_release + call void @objc_release(i8* %t1) + ret void +} + +;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; +; This test makes sure that we do not hang clang when visiting a use ; +; cycle caused by phi nodes during objc-arc analysis. *NOTE* This ; +; test case looks a little convoluted since it was produced by ; +; bugpoint. ; +; ; +; bugzilla://14551 ; +; rdar://12851911 ; +;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; + +define void @phinode_use_cycle(i8* %block) uwtable optsize ssp { +; CHECK: define void @phinode_use_cycle(i8* %block) +entry: + br label %for.body + +for.body: ; preds = %if.then, %for.body, %entry + %block.05 = phi void (...)* [ null, %entry ], [ %1, %if.then ], [ %block.05, %for.body ] + br i1 undef, label %for.body, label %if.then + +if.then: ; preds = %for.body + %0 = call i8* @objc_retainBlock(i8* %block), !clang.arc.copy_on_escape !0 + %1 = bitcast i8* %0 to void (...)* + %2 = bitcast void (...)* %block.05 to i8* + call void @objc_release(i8* %2) nounwind, !clang.imprecise_release !0 + br label %for.body +} + +!0 = metadata !{} diff --git a/test/Transforms/SROA/basictest.ll b/test/Transforms/SROA/basictest.ll index efc01ac..30dd217 100644 --- a/test/Transforms/SROA/basictest.ll +++ b/test/Transforms/SROA/basictest.ll @@ -500,14 +500,27 @@ entry: define i64 @test9() { ; Ensure we can handle loads off the end of an alloca even when wrapped in -; weird bit casts and types. The result is undef, but this shouldn't crash -; anything. +; weird bit casts and types. This is valid IR due to the alignment and masking +; off the bits past the end of the alloca. +; ; CHECK: @test9 ; CHECK-NOT: alloca -; CHECK: ret i64 undef +; CHECK: %[[b2:.*]] = zext i8 26 to i64 +; CHECK-NEXT: %[[s2:.*]] = shl i64 %[[b2]], 16 +; CHECK-NEXT: %[[m2:.*]] = and i64 undef, -16711681 +; CHECK-NEXT: %[[i2:.*]] = or i64 %[[m2]], %[[s2]] +; CHECK-NEXT: %[[b1:.*]] = zext i8 0 to i64 +; CHECK-NEXT: %[[s1:.*]] = shl i64 %[[b1]], 8 +; CHECK-NEXT: %[[m1:.*]] = and i64 %[[i2]], -65281 +; CHECK-NEXT: %[[i1:.*]] = or i64 %[[m1]], %[[s1]] +; CHECK-NEXT: %[[b0:.*]] = zext i8 0 to i64 +; CHECK-NEXT: %[[m0:.*]] = and i64 %[[i1]], -256 +; CHECK-NEXT: %[[i0:.*]] = or i64 %[[m0]], %[[b0]] +; CHECK-NEXT: %[[result:.*]] = and i64 %[[i0]], 16777215 +; CHECK-NEXT: ret i64 %[[result]] entry: - %a = alloca { [3 x i8] } + %a = alloca { [3 x i8] }, align 8 %gep1 = getelementptr inbounds { [3 x i8] }* %a, i32 0, i32 0, i32 0 store i8 0, i8* %gep1, align 1 %gep2 = getelementptr inbounds { [3 x i8] }* %a, i32 0, i32 0, i32 1 @@ -516,7 +529,8 @@ entry: store i8 26, i8* %gep3, align 1 %cast = bitcast { [3 x i8] }* %a to { i64 }* %elt = getelementptr inbounds { i64 }* %cast, i32 0, i32 0 - %result = load i64* %elt + %load = load i64* %elt + %result = and i64 %load, 16777215 ret i64 %result } @@ -617,11 +631,12 @@ define i32 @test13() { ; Ensure we don't crash and handle undefined loads that straddle the end of the ; allocation. ; CHECK: @test13 -; CHECK: %[[ret:.*]] = zext i16 undef to i32 -; CHECK: ret i32 %[[ret]] +; CHECK: %[[value:.*]] = zext i8 0 to i16 +; CHECK-NEXT: %[[ret:.*]] = zext i16 %[[value]] to i32 +; CHECK-NEXT: ret i32 %[[ret]] entry: - %a = alloca [3 x i8] + %a = alloca [3 x i8], align 2 %b0ptr = getelementptr [3 x i8]* %a, i64 0, i32 0 store i8 0, i8* %b0ptr %b1ptr = getelementptr [3 x i8]* %a, i64 0, i32 1 @@ -1160,20 +1175,25 @@ define void @PR14548(i1 %x) { entry: %a = alloca <{ i1 }>, align 8 %b = alloca <{ i1 }>, align 8 -; Nothing of interest is simplified here. -; CHECK: alloca -; CHECK: alloca +; CHECK: %[[a:.*]] = alloca i8, align 8 %b.i1 = bitcast <{ i1 }>* %b to i1* store i1 %x, i1* %b.i1, align 8 %b.i8 = bitcast <{ i1 }>* %b to i8* %foo = load i8* %b.i8, align 1 +; CHECK-NEXT: {{.*}} = zext i1 %x to i8 +; CHECK-NEXT: %[[ext:.*]] = zext i1 %x to i8 +; CHECK-NEXT: store i8 %[[ext]], i8* %[[a]], align 8 +; CHECK-NEXT: {{.*}} = load i8* %[[a]], align 8 %a.i8 = bitcast <{ i1 }>* %a to i8* call void @llvm.memcpy.p0i8.p0i8.i32(i8* %a.i8, i8* %b.i8, i32 1, i32 1, i1 false) nounwind %bar = load i8* %a.i8, align 1 %a.i1 = getelementptr inbounds <{ i1 }>* %a, i32 0, i32 0 %baz = load i1* %a.i1, align 1 +; CHECK-NEXT: %[[a_cast:.*]] = bitcast i8* %[[a]] to i1* +; CHECK-NEXT: {{.*}} = load i1* %[[a_cast]], align 8 + ret void } diff --git a/test/Transforms/SROA/phi-and-select.ll b/test/Transforms/SROA/phi-and-select.ll index 921016a..b993180 100644 --- a/test/Transforms/SROA/phi-and-select.ll +++ b/test/Transforms/SROA/phi-and-select.ll @@ -396,9 +396,10 @@ define i64 @PR14132(i1 %flag) { ; Here we form a PHI-node by promoting the pointer alloca first, and then in ; order to promote the other two allocas, we speculate the load of the ; now-phi-node-pointer. In doing so we end up loading a 64-bit value from an i8 -; alloca, which is completely bogus. However, we were asserting on trying to -; rewrite it. Now it is replaced with undef. Eventually we may replace it with -; unrechable and even the CFG will go away here. +; alloca. While this is a bit dubious, we were asserting on trying to +; rewrite it. The trick is that the code using the value may carefully take +; steps to only use the not-undef bits, and so we need to at least loosely +; support this.. entry: %a = alloca i64 %b = alloca i8 @@ -414,13 +415,14 @@ entry: if.then: store i8* %b, i8** %ptr.cast br label %if.end +; CHECK-NOT: store +; CHECK: %[[ext:.*]] = zext i8 1 to i64 if.end: %tmp = load i64** %ptr %result = load i64* %tmp -; CHECK-NOT: store ; CHECK-NOT: load -; CHECK: %[[result:.*]] = phi i64 [ undef, %if.then ], [ 0, %entry ] +; CHECK: %[[result:.*]] = phi i64 [ %[[ext]], %if.then ], [ 0, %entry ] ret i64 %result ; CHECK-NEXT: ret i64 %[[result]] diff --git a/test/Transforms/ScalarRepl/debuginfo-preserved.ll b/test/Transforms/ScalarRepl/debuginfo-preserved.ll index c149134..d1dd01d 100644 --- a/test/Transforms/ScalarRepl/debuginfo-preserved.ll +++ b/test/Transforms/ScalarRepl/debuginfo-preserved.ll @@ -40,22 +40,23 @@ entry: declare void @llvm.dbg.declare(metadata, metadata) nounwind readnone !llvm.dbg.cu = !{!0} -!llvm.dbg.sp = !{!1} -!0 = metadata !{i32 589841, i32 0, i32 12, metadata !"/d/j/debug-test.c", metadata !"/Volumes/Data/b", metadata !"clang version 3.0 (trunk 131941)", i1 true, i1 false, metadata !"", i32 0} ; [ DW_TAG_compile_unit ] -!1 = metadata !{i32 589870, i32 0, metadata !2, metadata !"f", metadata !"f", metadata !"", metadata !2, i32 1, metadata !3, i1 false, i1 true, i32 0, i32 0, i32 0, i32 256, i1 false, i32 (i32, i32)* @f, null, null} ; [ DW_TAG_subprogram ] -!2 = metadata !{i32 589865, metadata !"/d/j/debug-test.c", metadata !"/Volumes/Data/b", metadata !0} ; [ DW_TAG_file_type ] -!3 = metadata !{i32 589845, metadata !2, metadata !"", metadata !2, i32 0, i64 0, i64 0, i32 0, i32 0, i32 0, metadata !4, i32 0, i32 0} ; [ DW_TAG_subroutine_type ] +!0 = metadata !{i32 786449, i32 0, i32 12, metadata !2, metadata !"clang version 3.0 (trunk 131941)", i1 false, metadata !"", i32 0, null, null, metadata !17, null, null} ; [ DW_TAG_compile_unit ] +!1 = metadata !{i32 786478, i32 0, metadata !2, metadata !"f", metadata !"f", metadata !"", metadata !2, i32 1, metadata !3, i1 false, i1 true, i32 0, i32 0, i32 0, i32 256, i1 false, i32 (i32, i32)* @f, null, null, null, i32 1} ; [ DW_TAG_subprogram ] +!2 = metadata !{i32 786473, metadata !18} ; [ DW_TAG_file_type ] +!3 = metadata !{i32 786453, metadata !2, metadata !"", metadata !2, i32 0, i64 0, i64 0, i32 0, i32 0, i32 0, metadata !4, i32 0, i32 0} ; [ DW_TAG_subroutine_type ] !4 = metadata !{metadata !5} -!5 = metadata !{i32 589860, metadata !0, metadata !"int", null, i32 0, i64 32, i64 32, i64 0, i32 0, i32 5} ; [ DW_TAG_base_type ] -!6 = metadata !{i32 590081, metadata !1, metadata !"a", metadata !2, i32 16777217, metadata !5, i32 0} ; [ DW_TAG_arg_variable ] +!5 = metadata !{i32 786468, metadata !0, metadata !"int", null, i32 0, i64 32, i64 32, i64 0, i32 0, i32 5} ; [ DW_TAG_base_type ] +!6 = metadata !{i32 786689, metadata !1, metadata !"a", metadata !2, i32 16777217, metadata !5, i32 0, null} ; [ DW_TAG_arg_variable ] !7 = metadata !{i32 1, i32 11, metadata !1, null} -!8 = metadata !{i32 590081, metadata !1, metadata !"b", metadata !2, i32 33554433, metadata !5, i32 0} ; [ DW_TAG_arg_variable ] +!8 = metadata !{i32 786689, metadata !1, metadata !"b", metadata !2, i32 33554433, metadata !5, i32 0, null} ; [ DW_TAG_arg_variable ] !9 = metadata !{i32 1, i32 18, metadata !1, null} -!10 = metadata !{i32 590080, metadata !11, metadata !"c", metadata !2, i32 2, metadata !5, i32 0} ; [ DW_TAG_auto_variable ] -!11 = metadata !{i32 589835, metadata !1, i32 1, i32 21, metadata !2, i32 0} ; [ DW_TAG_lexical_block ] +!10 = metadata !{i32 786688, metadata !11, metadata !"c", metadata !2, i32 2, metadata !5, i32 0, null} ; [ DW_TAG_auto_variable ] +!11 = metadata !{i32 786443, metadata !1, i32 1, i32 21, metadata !2, i32 0} ; [ DW_TAG_lexical_block ] !12 = metadata !{i32 2, i32 9, metadata !11, null} !13 = metadata !{i32 2, i32 14, metadata !11, null} !14 = metadata !{i32 3, i32 5, metadata !11, null} !15 = metadata !{i32 4, i32 5, metadata !11, null} !16 = metadata !{i32 5, i32 5, metadata !11, null} +!17 = metadata !{metadata !1} +!18 = metadata !{metadata !"/d/j/debug-test.c", metadata !"/Volumes/Data/b"} diff --git a/test/Transforms/SimplifyCFG/trivial-throw.ll b/test/Transforms/SimplifyCFG/trivial-throw.ll new file mode 100644 index 0000000..ca2b569 --- /dev/null +++ b/test/Transforms/SimplifyCFG/trivial-throw.ll @@ -0,0 +1,77 @@ +; RUN: opt -simplifycfg -S < %s | FileCheck %s +; <rdar://problem/13360379> + +@_ZTVN10__cxxabiv117__class_type_infoE = external global i8* +@_ZTS13TestException = linkonce_odr constant [16 x i8] c"13TestException\00" +@_ZTI13TestException = linkonce_odr unnamed_addr constant { i8*, i8* } { i8* bitcast (i8** getelementptr inbounds (i8** @_ZTVN10__cxxabiv117__class_type_infoE, i64 2) to i8*), i8* getelementptr inbounds ([16 x i8]* @_ZTS13TestException, i32 0, i32 0) } + +define void @throw(i32 %n) #0 { +entry: + %exception = call i8* @__cxa_allocate_exception(i64 1) #4 + call void @__cxa_throw(i8* %exception, i8* bitcast ({ i8*, i8* }* @_ZTI13TestException to i8*), i8* null) #2 + unreachable +} + +define void @func() #0 { +entry: +; CHECK: func() +; CHECK: invoke void @throw +; CHECK-NOT: call void @throw + invoke void @throw(i32 42) #0 + to label %exit unwind label %lpad + +lpad: + %tmp0 = landingpad { i8*, i32 } personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) + cleanup + resume { i8*, i32 } %tmp0 + +exit: + invoke void @abort() #2 + to label %invoke.cont unwind label %lpad1 + +invoke.cont: + unreachable + +lpad1: + %tmp1 = landingpad { i8*, i32 } personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) + catch i8* bitcast ({ i8*, i8* }* @_ZTI13TestException to i8*) + %tmp2 = extractvalue { i8*, i32 } %tmp1, 1 + %tmp3 = call i32 @llvm.eh.typeid.for(i8* bitcast ({ i8*, i8* }* @_ZTI13TestException to i8*)) #4 + %matches = icmp eq i32 %tmp2, %tmp3 + br i1 %matches, label %catch, label %eh.resume + +catch: + ret void + +eh.resume: + resume { i8*, i32 } %tmp1 +} + +define linkonce_odr hidden void @__clang_call_terminate(i8*) #1 { + %2 = call i8* @__cxa_begin_catch(i8* %0) #4 + call void @_ZSt9terminatev() #5 + unreachable +} + +declare void @abort() #2 + +declare i32 @llvm.eh.typeid.for(i8*) #3 + +declare void @__cxa_end_catch() + +declare i8* @__cxa_allocate_exception(i64) + +declare i32 @__gxx_personality_v0(...) + +declare void @__cxa_throw(i8*, i8*, i8*) + +declare i8* @__cxa_begin_catch(i8*) + +declare void @_ZSt9terminatev() + +attributes #0 = { ssp uwtable } +attributes #1 = { noinline noreturn nounwind } +attributes #2 = { noreturn } +attributes #3 = { nounwind readnone } +attributes #4 = { nounwind } +attributes #5 = { noreturn nounwind } diff --git a/test/Transforms/SimplifyCFG/volatile-phioper.ll b/test/Transforms/SimplifyCFG/volatile-phioper.ll new file mode 100644 index 0000000..1648988 --- /dev/null +++ b/test/Transforms/SimplifyCFG/volatile-phioper.ll @@ -0,0 +1,48 @@ +; RUN: opt < %s -simplifycfg -S | FileCheck %s +; +; rdar:13349374 +; +; SimplifyCFG should not eliminate blocks with volatile stores. +; Essentially, volatile needs to be backdoor that tells the optimizer +; it can no longer use language standard as an excuse. The compiler +; needs to expose the volatile access to the platform. +; +; CHECK: @test +; CHECK: entry: +; CHECK: @Trace +; CHECK: while.body: +; CHECK: store volatile +; CHECK: end: +target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128" + +define void @test(i8** nocapture %PeiServices) #0 { +entry: + %call = tail call i32 (...)* @Trace() #2 + %tobool = icmp eq i32 %call, 0 + br i1 %tobool, label %while.body, label %if.then + +if.then: ; preds = %entry + %call1 = tail call i32 (...)* @Trace() #2 + br label %while.body + +while.body: ; preds = %entry, %if.then, %while.body + %Addr.017 = phi i8* [ %incdec.ptr, %while.body ], [ null, %if.then ], [ null, %entry ] + %x.016 = phi i8 [ %inc, %while.body ], [ 0, %if.then ], [ 0, %entry ] + %inc = add i8 %x.016, 1 + %incdec.ptr = getelementptr inbounds i8* %Addr.017, i64 1 + store volatile i8 %x.016, i8* %Addr.017, align 1 + %0 = ptrtoint i8* %incdec.ptr to i64 + %1 = trunc i64 %0 to i32 + %cmp = icmp ult i32 %1, 4096 + br i1 %cmp, label %while.body, label %end + +end: + ret void +} +declare i32 @Trace(...) #1 + +attributes #0 = { nounwind ssp uwtable "fp-contract-model"="standard" "no-frame-pointer-elim" "no-frame-pointer-elim-non-leaf" "realign-stack" "relocation-model"="pic" "ssp-buffers-size"="8" } +attributes #1 = { "fp-contract-model"="standard" "no-frame-pointer-elim" "no-frame-pointer-elim-non-leaf" "realign-stack" "relocation-model"="pic" "ssp-buffers-size"="8" } +attributes #2 = { nounwind } + +!0 = metadata !{i32 1039} diff --git a/test/Transforms/TailCallElim/ackermann.ll b/test/Transforms/TailCallElim/ackermann.ll index 5b5dbcc..83d98b8 100644 --- a/test/Transforms/TailCallElim/ackermann.ll +++ b/test/Transforms/TailCallElim/ackermann.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; This function contains two tail calls, which should be eliminated ; RUN: opt < %s -tailcallelim -stats -disable-output 2>&1 | grep "2 tailcallelim" diff --git a/test/Transforms/TailCallElim/dup_tail.ll b/test/Transforms/TailCallElim/dup_tail.ll index 3b87ed3..f5b87f2 100644 --- a/test/Transforms/TailCallElim/dup_tail.ll +++ b/test/Transforms/TailCallElim/dup_tail.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; Duplicate the return into if.end to enable TCE. ; RUN: opt -tailcallelim -stats -disable-output < %s 2>&1 | FileCheck %s diff --git a/test/Transforms/TailDup/2008-06-11-AvoidDupLoopHeader.ll b/test/Transforms/TailDup/2008-06-11-AvoidDupLoopHeader.ll index 7853d7b..2921860 100644 --- a/test/Transforms/TailDup/2008-06-11-AvoidDupLoopHeader.ll +++ b/test/Transforms/TailDup/2008-06-11-AvoidDupLoopHeader.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: opt < %s -tailduplicate -taildup-threshold=3 -stats -disable-output 2>&1 | not grep tailduplicate ; XFAIL: * |