Author: Florian Hahn Date: 2021-06-11T14:06:35+01:00 New Revision: 31ea973802050b5bacf4d5261e000b6875108c49
URL: https://github.com/llvm/llvm-project/commit/31ea973802050b5bacf4d5261e000b6875108c49 DIFF: https://github.com/llvm/llvm-project/commit/31ea973802050b5bacf4d5261e000b6875108c49.diff LOG: [SLP] Add some tests that require memory runtime checks. Added: llvm/test/Transforms/SLPVectorizer/AArch64/memory-runtime-checks.ll llvm/test/Transforms/SLPVectorizer/X86/memory-runtime-checks.ll Modified: Removed: ################################################################################ diff --git a/llvm/test/Transforms/SLPVectorizer/AArch64/memory-runtime-checks.ll b/llvm/test/Transforms/SLPVectorizer/AArch64/memory-runtime-checks.ll new file mode 100644 index 0000000000000..3ecb0f20cbbd4 --- /dev/null +++ b/llvm/test/Transforms/SLPVectorizer/AArch64/memory-runtime-checks.ll @@ -0,0 +1,354 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py +; RUN: opt -scoped-noalias-aa -slp-vectorizer -mtriple=arm64-apple-darwin -enable-new-pm=false -S %s | FileCheck %s + +define void @needs_versioning_not_profitable(i32* %dst, i32* %src) { +; CHECK-LABEL: @needs_versioning_not_profitable( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[SRC_0:%.*]] = load i32, i32* [[SRC:%.*]], align 4 +; CHECK-NEXT: [[R_0:%.*]] = ashr i32 [[SRC_0]], 16 +; CHECK-NEXT: store i32 [[R_0]], i32* [[DST:%.*]], align 4 +; CHECK-NEXT: [[SRC_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[SRC]], i64 1 +; CHECK-NEXT: [[SRC_1:%.*]] = load i32, i32* [[SRC_GEP_1]], align 4 +; CHECK-NEXT: [[R_1:%.*]] = ashr i32 [[SRC_1]], 16 +; CHECK-NEXT: [[DST_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[DST]], i64 1 +; CHECK-NEXT: store i32 [[R_1]], i32* [[DST_GEP_1]], align 4 +; CHECK-NEXT: ret void +; +entry: + %src.0 = load i32, i32* %src, align 4 + %r.0 = ashr i32 %src.0, 16 + store i32 %r.0, i32* %dst, align 4 + %src.gep.1 = getelementptr inbounds i32, i32* %src, i64 1 + %src.1 = load i32, i32* %src.gep.1, align 4 + %r.1 = ashr i32 %src.1, 16 + %dst.gep.1 = getelementptr inbounds i32, i32* %dst, i64 1 + store i32 %r.1, i32* %dst.gep.1, align 4 + ret void +} + +define void @needs_versioning_profitable(i32* %dst, i32* %src) { +; CHECK-LABEL: @needs_versioning_profitable( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[SRC_0:%.*]] = load i32, i32* [[SRC:%.*]], align 4 +; CHECK-NEXT: [[R_0:%.*]] = ashr i32 [[SRC_0]], 16 +; CHECK-NEXT: store i32 [[R_0]], i32* [[DST:%.*]], align 4 +; CHECK-NEXT: [[SRC_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[SRC]], i64 1 +; CHECK-NEXT: [[SRC_1:%.*]] = load i32, i32* [[SRC_GEP_1]], align 4 +; CHECK-NEXT: [[R_1:%.*]] = ashr i32 [[SRC_1]], 16 +; CHECK-NEXT: [[DST_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[DST]], i64 1 +; CHECK-NEXT: store i32 [[R_1]], i32* [[DST_GEP_1]], align 4 +; CHECK-NEXT: [[SRC_GEP_2:%.*]] = getelementptr inbounds i32, i32* [[SRC]], i64 2 +; CHECK-NEXT: [[SRC_2:%.*]] = load i32, i32* [[SRC_GEP_2]], align 4 +; CHECK-NEXT: [[R_2:%.*]] = ashr i32 [[SRC_2]], 16 +; CHECK-NEXT: [[DST_GEP_2:%.*]] = getelementptr inbounds i32, i32* [[DST]], i64 2 +; CHECK-NEXT: store i32 [[R_2]], i32* [[DST_GEP_2]], align 4 +; CHECK-NEXT: [[SRC_GEP_3:%.*]] = getelementptr inbounds i32, i32* [[SRC]], i64 3 +; CHECK-NEXT: [[SRC_3:%.*]] = load i32, i32* [[SRC_GEP_3]], align 4 +; CHECK-NEXT: [[R_3:%.*]] = ashr i32 [[SRC_3]], 16 +; CHECK-NEXT: [[DST_GEP_3:%.*]] = getelementptr inbounds i32, i32* [[DST]], i64 3 +; CHECK-NEXT: store i32 [[R_3]], i32* [[DST_GEP_3]], align 4 +; CHECK-NEXT: ret void +; +entry: + %src.0 = load i32, i32* %src, align 4 + %r.0 = ashr i32 %src.0, 16 + store i32 %r.0, i32* %dst, align 4 + %src.gep.1 = getelementptr inbounds i32, i32* %src, i64 1 + %src.1 = load i32, i32* %src.gep.1, align 4 + %r.1 = ashr i32 %src.1, 16 + %dst.gep.1 = getelementptr inbounds i32, i32* %dst, i64 1 + store i32 %r.1, i32* %dst.gep.1, align 4 + %src.gep.2 = getelementptr inbounds i32, i32* %src, i64 2 + %src.2 = load i32, i32* %src.gep.2, align 4 + %r.2 = ashr i32 %src.2, 16 + %dst.gep.2 = getelementptr inbounds i32, i32* %dst, i64 2 + store i32 %r.2, i32* %dst.gep.2, align 4 + %src.gep.3 = getelementptr inbounds i32, i32* %src, i64 3 + %src.3 = load i32, i32* %src.gep.3, align 4 + %r.3 = ashr i32 %src.3, 16 + %dst.gep.3 = getelementptr inbounds i32, i32* %dst, i64 3 + store i32 %r.3, i32* %dst.gep.3, align 4 + + ret void +} + + +define void @no_version(i32* nocapture %dst, i32* nocapture readonly %src) { +; CHECK-LABEL: @no_version( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[SRC_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[SRC:%.*]], i64 1 +; CHECK-NEXT: [[TMP0:%.*]] = bitcast i32* [[SRC]] to <2 x i32>* +; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i32>, <2 x i32>* [[TMP0]], align 4 +; CHECK-NEXT: [[TMP2:%.*]] = ashr <2 x i32> [[TMP1]], <i32 16, i32 16> +; CHECK-NEXT: [[DST_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[DST:%.*]], i64 1 +; CHECK-NEXT: [[TMP3:%.*]] = bitcast i32* [[DST]] to <2 x i32>* +; CHECK-NEXT: store <2 x i32> [[TMP2]], <2 x i32>* [[TMP3]], align 4 +; CHECK-NEXT: ret void +; +entry: + %src.0 = load i32, i32* %src, align 4 + %src.gep.1 = getelementptr inbounds i32, i32* %src, i64 1 + %src.1 = load i32, i32* %src.gep.1, align 4 + %r.0 = ashr i32 %src.0, 16 + %r.1 = ashr i32 %src.1, 16 + %dst.gep.1 = getelementptr inbounds i32, i32* %dst, i64 1 + store i32 %r.0, i32* %dst, align 4 + store i32 %r.1, i32* %dst.gep.1, align 4 + ret void +} + +define void @version_multiple(i32* nocapture %out_block, i32* nocapture readonly %counter) { +; CHECK-LABEL: @version_multiple( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[TMP0:%.*]] = load i32, i32* [[COUNTER:%.*]], align 4 +; CHECK-NEXT: [[TMP1:%.*]] = load i32, i32* [[OUT_BLOCK:%.*]], align 4 +; CHECK-NEXT: [[XOR:%.*]] = xor i32 [[TMP1]], [[TMP0]] +; CHECK-NEXT: store i32 [[XOR]], i32* [[OUT_BLOCK]], align 4 +; CHECK-NEXT: [[ARRAYIDX_1:%.*]] = getelementptr inbounds i32, i32* [[COUNTER]], i64 1 +; CHECK-NEXT: [[TMP2:%.*]] = load i32, i32* [[ARRAYIDX_1]], align 4 +; CHECK-NEXT: [[ARRAYIDX2_1:%.*]] = getelementptr inbounds i32, i32* [[OUT_BLOCK]], i64 1 +; CHECK-NEXT: [[TMP3:%.*]] = load i32, i32* [[ARRAYIDX2_1]], align 4 +; CHECK-NEXT: [[XOR_1:%.*]] = xor i32 [[TMP3]], [[TMP2]] +; CHECK-NEXT: store i32 [[XOR_1]], i32* [[ARRAYIDX2_1]], align 4 +; CHECK-NEXT: [[ARRAYIDX_2:%.*]] = getelementptr inbounds i32, i32* [[COUNTER]], i64 2 +; CHECK-NEXT: [[TMP4:%.*]] = load i32, i32* [[ARRAYIDX_2]], align 4 +; CHECK-NEXT: [[ARRAYIDX2_2:%.*]] = getelementptr inbounds i32, i32* [[OUT_BLOCK]], i64 2 +; CHECK-NEXT: [[TMP5:%.*]] = load i32, i32* [[ARRAYIDX2_2]], align 4 +; CHECK-NEXT: [[XOR_2:%.*]] = xor i32 [[TMP5]], [[TMP4]] +; CHECK-NEXT: store i32 [[XOR_2]], i32* [[ARRAYIDX2_2]], align 4 +; CHECK-NEXT: [[ARRAYIDX_3:%.*]] = getelementptr inbounds i32, i32* [[COUNTER]], i64 3 +; CHECK-NEXT: [[TMP6:%.*]] = load i32, i32* [[ARRAYIDX_3]], align 4 +; CHECK-NEXT: [[ARRAYIDX2_3:%.*]] = getelementptr inbounds i32, i32* [[OUT_BLOCK]], i64 3 +; CHECK-NEXT: [[TMP7:%.*]] = load i32, i32* [[ARRAYIDX2_3]], align 4 +; CHECK-NEXT: [[XOR_3:%.*]] = xor i32 [[TMP7]], [[TMP6]] +; CHECK-NEXT: store i32 [[XOR_3]], i32* [[ARRAYIDX2_3]], align 4 +; CHECK-NEXT: ret void +; +entry: + %0 = load i32, i32* %counter, align 4 + %1 = load i32, i32* %out_block, align 4 + %xor = xor i32 %1, %0 + store i32 %xor, i32* %out_block, align 4 + %arrayidx.1 = getelementptr inbounds i32, i32* %counter, i64 1 + %2 = load i32, i32* %arrayidx.1, align 4 + %arrayidx2.1 = getelementptr inbounds i32, i32* %out_block, i64 1 + %3 = load i32, i32* %arrayidx2.1, align 4 + %xor.1 = xor i32 %3, %2 + store i32 %xor.1, i32* %arrayidx2.1, align 4 + %arrayidx.2 = getelementptr inbounds i32, i32* %counter, i64 2 + %4 = load i32, i32* %arrayidx.2, align 4 + %arrayidx2.2 = getelementptr inbounds i32, i32* %out_block, i64 2 + %5 = load i32, i32* %arrayidx2.2, align 4 + %xor.2 = xor i32 %5, %4 + store i32 %xor.2, i32* %arrayidx2.2, align 4 + %arrayidx.3 = getelementptr inbounds i32, i32* %counter, i64 3 + %6 = load i32, i32* %arrayidx.3, align 4 + %arrayidx2.3 = getelementptr inbounds i32, i32* %out_block, i64 3 + %7 = load i32, i32* %arrayidx2.3, align 4 + %xor.3 = xor i32 %7, %6 + store i32 %xor.3, i32* %arrayidx2.3, align 4 + ret void +} + +define i32 @use_outside_version_bb(i32* %dst, i32* %src, i1 %c.1) { +; CHECK-LABEL: @use_outside_version_bb( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[SRC_0:%.*]] = load i32, i32* [[SRC:%.*]], align 4 +; CHECK-NEXT: [[R_0:%.*]] = ashr i32 [[SRC_0]], 16 +; CHECK-NEXT: store i32 [[R_0]], i32* [[DST:%.*]], align 4 +; CHECK-NEXT: [[SRC_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[SRC]], i64 1 +; CHECK-NEXT: [[SRC_1:%.*]] = load i32, i32* [[SRC_GEP_1]], align 4 +; CHECK-NEXT: [[R_1:%.*]] = ashr i32 [[SRC_1]], 16 +; CHECK-NEXT: [[DST_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[DST]], i64 1 +; CHECK-NEXT: store i32 [[R_1]], i32* [[DST_GEP_1]], align 4 +; CHECK-NEXT: br label [[EXIT:%.*]] +; CHECK: exit: +; CHECK-NEXT: ret i32 [[R_0]] +; +entry: + %src.0 = load i32, i32* %src, align 4 + %r.0 = ashr i32 %src.0, 16 + store i32 %r.0, i32* %dst, align 4 + %src.gep.1 = getelementptr inbounds i32, i32* %src, i64 1 + %src.1 = load i32, i32* %src.gep.1, align 4 + %r.1 = ashr i32 %src.1, 16 + %dst.gep.1 = getelementptr inbounds i32, i32* %dst, i64 1 + store i32 %r.1, i32* %dst.gep.1, align 4 + br label %exit + +exit: + ret i32 %r.0 +} + +define i32 @value_used_in_return(i32* %dst, i32* %src, i32 %x) { +; CHECK-LABEL: @value_used_in_return( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[SRC_0:%.*]] = load i32, i32* [[SRC:%.*]], align 4 +; CHECK-NEXT: [[R_0:%.*]] = ashr i32 [[SRC_0]], 16 +; CHECK-NEXT: store i32 [[R_0]], i32* [[DST:%.*]], align 4 +; CHECK-NEXT: [[SRC_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[SRC]], i64 1 +; CHECK-NEXT: [[SRC_1:%.*]] = load i32, i32* [[SRC_GEP_1]], align 4 +; CHECK-NEXT: [[R_1:%.*]] = ashr i32 [[SRC_1]], 16 +; CHECK-NEXT: [[DST_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[DST]], i64 1 +; CHECK-NEXT: store i32 [[R_1]], i32* [[DST_GEP_1]], align 4 +; CHECK-NEXT: [[ADD:%.*]] = add i32 [[X:%.*]], 20 +; CHECK-NEXT: ret i32 [[ADD]] +; +entry: + %src.0 = load i32, i32* %src, align 4 + %r.0 = ashr i32 %src.0, 16 + store i32 %r.0, i32* %dst, align 4 + %src.gep.1 = getelementptr inbounds i32, i32* %src, i64 1 + %src.1 = load i32, i32* %src.gep.1, align 4 + %r.1 = ashr i32 %src.1, 16 + %dst.gep.1 = getelementptr inbounds i32, i32* %dst, i64 1 + store i32 %r.1, i32* %dst.gep.1, align 4 + %add = add i32 %x, 20 + ret i32 %add +} +define i32 @needs_versioning2_cond_br(i32* %dst, i32* %src, i1 %c.1) { +; CHECK-LABEL: @needs_versioning2_cond_br( +; CHECK-NEXT: entry: +; CHECK-NEXT: br i1 [[C_1:%.*]], label [[THEN:%.*]], label [[ELSE:%.*]] +; CHECK: then: +; CHECK-NEXT: [[SRC_0:%.*]] = load i32, i32* [[SRC:%.*]], align 4 +; CHECK-NEXT: [[R_0:%.*]] = ashr i32 [[SRC_0]], 16 +; CHECK-NEXT: store i32 [[R_0]], i32* [[DST:%.*]], align 4 +; CHECK-NEXT: [[SRC_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[SRC]], i64 1 +; CHECK-NEXT: [[SRC_1:%.*]] = load i32, i32* [[SRC_GEP_1]], align 4 +; CHECK-NEXT: [[R_1:%.*]] = ashr i32 [[SRC_1]], 16 +; CHECK-NEXT: [[DST_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[DST]], i64 1 +; CHECK-NEXT: store i32 [[R_1]], i32* [[DST_GEP_1]], align 4 +; CHECK-NEXT: ret i32 10 +; CHECK: else: +; CHECK-NEXT: ret i32 0 +; +entry: + br i1 %c.1, label %then, label %else + +then: + %src.0 = load i32, i32* %src, align 4 + %r.0 = ashr i32 %src.0, 16 + store i32 %r.0, i32* %dst, align 4 + %src.gep.1 = getelementptr inbounds i32, i32* %src, i64 1 + %src.1 = load i32, i32* %src.gep.1, align 4 + %r.1 = ashr i32 %src.1, 16 + %dst.gep.1 = getelementptr inbounds i32, i32* %dst, i64 1 + store i32 %r.1, i32* %dst.gep.1, align 4 + ret i32 10 + + +else: + ret i32 0 +} + +define void @pointer_defined_in_bb(i32* %dst, i32** %src.p) { +; CHECK-LABEL: @pointer_defined_in_bb( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[SRC:%.*]] = load i32*, i32** [[SRC_P:%.*]], align 8 +; CHECK-NEXT: [[SRC_0:%.*]] = load i32, i32* [[SRC]], align 4 +; CHECK-NEXT: [[R_0:%.*]] = ashr i32 [[SRC_0]], 16 +; CHECK-NEXT: store i32 [[R_0]], i32* [[DST:%.*]], align 4 +; CHECK-NEXT: [[SRC_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[SRC]], i64 1 +; CHECK-NEXT: [[SRC_1:%.*]] = load i32, i32* [[SRC_GEP_1]], align 4 +; CHECK-NEXT: [[R_1:%.*]] = ashr i32 [[SRC_1]], 16 +; CHECK-NEXT: [[DST_GEP_1:%.*]] = getelementptr inbounds i32, i32* [[DST]], i64 1 +; CHECK-NEXT: store i32 [[R_1]], i32* [[DST_GEP_1]], align 4 +; CHECK-NEXT: ret void +; +entry: + %src = load i32*, i32** %src.p + %src.0 = load i32, i32* %src, align 4 + %r.0 = ashr i32 %src.0, 16 + store i32 %r.0, i32* %dst, align 4 + %src.gep.1 = getelementptr inbounds i32, i32* %src, i64 1 + %src.1 = load i32, i32* %src.gep.1, align 4 + %r.1 = ashr i32 %src.1, 16 + %dst.gep.1 = getelementptr inbounds i32, i32* %dst, i64 1 + store i32 %r.1, i32* %dst.gep.1, align 4 + ret void +} + +define void @clobber_same_underlying_object(i32* %this) { +; CHECK-LABEL: @clobber_same_underlying_object( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[P_3:%.*]] = getelementptr inbounds i32, i32* [[THIS:%.*]], i32 3 +; CHECK-NEXT: store i32 10, i32* [[P_3]], align 8 +; CHECK-NEXT: tail call void @clobber() +; CHECK-NEXT: [[P_4:%.*]] = getelementptr inbounds i32, i32* [[THIS]], i32 4 +; CHECK-NEXT: [[L2:%.*]] = load i32, i32* [[P_4]], align 8 +; CHECK-NEXT: store i32 20, i32* [[P_4]], align 8 +; CHECK-NEXT: ret void +; +entry: + %p.3 = getelementptr inbounds i32, i32* %this, i32 3 + store i32 10, i32* %p.3, align 8 + tail call void @clobber() + %p.4 = getelementptr inbounds i32, i32* %this, i32 4 + %l2 = load i32, i32* %p.4, align 8 + store i32 20, i32* %p.4, align 8 + ret void +} + +declare void @clobber() + +define void @slp_not_beneficial(i32* %A, i32* %B) { +; CHECK-LABEL: @slp_not_beneficial( +; CHECK-NEXT: bb: +; CHECK-NEXT: [[TMP:%.*]] = getelementptr inbounds i32, i32* [[A:%.*]], i32 4 +; CHECK-NEXT: store i32 0, i32* [[TMP]], align 8 +; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i32, i32* [[A]], i32 5 +; CHECK-NEXT: [[TMP4:%.*]] = getelementptr inbounds i32, i32* [[B:%.*]], i32 4 +; CHECK-NEXT: [[TMP5:%.*]] = load i32, i32* [[TMP4]], align 8 +; CHECK-NEXT: store i32 [[TMP5]], i32* [[TMP3]], align 8 +; CHECK-NEXT: ret void +; +bb: + %tmp = getelementptr inbounds i32, i32* %A, i32 4 + store i32 0, i32* %tmp, align 8 + %tmp3 = getelementptr inbounds i32, i32* %A, i32 5 + %tmp4 = getelementptr inbounds i32, i32* %B, i32 4 + %tmp5 = load i32, i32* %tmp4, align 8 + store i32 %tmp5, i32* %tmp3, align 8 + ret void +} + +define void @widget(double* %ptr, double* %ptr.2) { +; CHECK-LABEL: @widget( +; CHECK-NEXT: bb1: +; CHECK-NEXT: [[TMP3:%.*]] = load double, double* null, align 8 +; CHECK-NEXT: [[TMP4:%.*]] = fmul double undef, [[TMP3]] +; CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds double, double* [[PTR:%.*]], i32 0 +; CHECK-NEXT: [[TMP6:%.*]] = load double, double* [[TMP5]], align 8 +; CHECK-NEXT: [[TMP7:%.*]] = fadd double [[TMP6]], [[TMP4]] +; CHECK-NEXT: store double [[TMP7]], double* [[TMP5]], align 8 +; CHECK-NEXT: [[TMP8:%.*]] = getelementptr inbounds double, double* [[PTR_2:%.*]], i64 0 +; CHECK-NEXT: [[TMP9:%.*]] = load double, double* [[TMP8]], align 8 +; CHECK-NEXT: [[TMP10:%.*]] = fmul double undef, [[TMP9]] +; CHECK-NEXT: [[TMP11:%.*]] = getelementptr inbounds double, double* [[PTR]], i32 1 +; CHECK-NEXT: [[TMP12:%.*]] = load double, double* [[TMP11]], align 8 +; CHECK-NEXT: [[TMP13:%.*]] = fadd double [[TMP12]], [[TMP10]] +; CHECK-NEXT: store double [[TMP13]], double* [[TMP11]], align 8 +; CHECK-NEXT: br label [[BB15:%.*]] +; CHECK: bb15: +; CHECK-NEXT: br label [[BB15]] +; +bb1: ; preds = %bb + %tmp3 = load double, double* null, align 8 + %tmp4 = fmul double undef, %tmp3 + %tmp5 = getelementptr inbounds double, double* %ptr, i32 0 + %tmp6 = load double, double* %tmp5, align 8 + %tmp7 = fadd double %tmp6, %tmp4 + store double %tmp7, double* %tmp5, align 8 + %tmp8 = getelementptr inbounds double, double* %ptr.2, i64 0 + %tmp9 = load double, double* %tmp8, align 8 + %tmp10 = fmul double undef, %tmp9 + %tmp11 = getelementptr inbounds double, double* %ptr, i32 1 + %tmp12 = load double, double* %tmp11, align 8 + %tmp13 = fadd double %tmp12, %tmp10 + store double %tmp13, double* %tmp11, align 8 + br label %bb15 + +bb15: ; preds = %bb15, %bb14 + br label %bb15 +} diff --git a/llvm/test/Transforms/SLPVectorizer/X86/memory-runtime-checks.ll b/llvm/test/Transforms/SLPVectorizer/X86/memory-runtime-checks.ll new file mode 100644 index 0000000000000..cf76d0a98da29 --- /dev/null +++ b/llvm/test/Transforms/SLPVectorizer/X86/memory-runtime-checks.ll @@ -0,0 +1,188 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py +; RUN: opt -scoped-noalias-aa -slp-vectorizer -mtriple=x86_64-apple-darwin -enable-new-pm=false -S %s | FileCheck %s + +define void @version_multiple(i32* nocapture %out_block, i32* nocapture readonly %counter) { +; CHECK-LABEL: @version_multiple( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[TMP0:%.*]] = load i32, i32* [[COUNTER:%.*]], align 4 +; CHECK-NEXT: [[TMP1:%.*]] = load i32, i32* [[OUT_BLOCK:%.*]], align 4 +; CHECK-NEXT: [[XOR:%.*]] = xor i32 [[TMP1]], [[TMP0]] +; CHECK-NEXT: store i32 [[XOR]], i32* [[OUT_BLOCK]], align 4 +; CHECK-NEXT: [[ARRAYIDX_1:%.*]] = getelementptr inbounds i32, i32* [[COUNTER]], i64 1 +; CHECK-NEXT: [[TMP2:%.*]] = load i32, i32* [[ARRAYIDX_1]], align 4 +; CHECK-NEXT: [[ARRAYIDX2_1:%.*]] = getelementptr inbounds i32, i32* [[OUT_BLOCK]], i64 1 +; CHECK-NEXT: [[TMP3:%.*]] = load i32, i32* [[ARRAYIDX2_1]], align 4 +; CHECK-NEXT: [[XOR_1:%.*]] = xor i32 [[TMP3]], [[TMP2]] +; CHECK-NEXT: store i32 [[XOR_1]], i32* [[ARRAYIDX2_1]], align 4 +; CHECK-NEXT: [[ARRAYIDX_2:%.*]] = getelementptr inbounds i32, i32* [[COUNTER]], i64 2 +; CHECK-NEXT: [[TMP4:%.*]] = load i32, i32* [[ARRAYIDX_2]], align 4 +; CHECK-NEXT: [[ARRAYIDX2_2:%.*]] = getelementptr inbounds i32, i32* [[OUT_BLOCK]], i64 2 +; CHECK-NEXT: [[TMP5:%.*]] = load i32, i32* [[ARRAYIDX2_2]], align 4 +; CHECK-NEXT: [[XOR_2:%.*]] = xor i32 [[TMP5]], [[TMP4]] +; CHECK-NEXT: store i32 [[XOR_2]], i32* [[ARRAYIDX2_2]], align 4 +; CHECK-NEXT: [[ARRAYIDX_3:%.*]] = getelementptr inbounds i32, i32* [[COUNTER]], i64 3 +; CHECK-NEXT: [[TMP6:%.*]] = load i32, i32* [[ARRAYIDX_3]], align 4 +; CHECK-NEXT: [[ARRAYIDX2_3:%.*]] = getelementptr inbounds i32, i32* [[OUT_BLOCK]], i64 3 +; CHECK-NEXT: [[TMP7:%.*]] = load i32, i32* [[ARRAYIDX2_3]], align 4 +; CHECK-NEXT: [[XOR_3:%.*]] = xor i32 [[TMP7]], [[TMP6]] +; CHECK-NEXT: store i32 [[XOR_3]], i32* [[ARRAYIDX2_3]], align 4 +; CHECK-NEXT: ret void +; +entry: + %0 = load i32, i32* %counter, align 4 + %1 = load i32, i32* %out_block, align 4 + %xor = xor i32 %1, %0 + store i32 %xor, i32* %out_block, align 4 + %arrayidx.1 = getelementptr inbounds i32, i32* %counter, i64 1 + %2 = load i32, i32* %arrayidx.1, align 4 + %arrayidx2.1 = getelementptr inbounds i32, i32* %out_block, i64 1 + %3 = load i32, i32* %arrayidx2.1, align 4 + %xor.1 = xor i32 %3, %2 + store i32 %xor.1, i32* %arrayidx2.1, align 4 + %arrayidx.2 = getelementptr inbounds i32, i32* %counter, i64 2 + %4 = load i32, i32* %arrayidx.2, align 4 + %arrayidx2.2 = getelementptr inbounds i32, i32* %out_block, i64 2 + %5 = load i32, i32* %arrayidx2.2, align 4 + %xor.2 = xor i32 %5, %4 + store i32 %xor.2, i32* %arrayidx2.2, align 4 + %arrayidx.3 = getelementptr inbounds i32, i32* %counter, i64 3 + %6 = load i32, i32* %arrayidx.3, align 4 + %arrayidx2.3 = getelementptr inbounds i32, i32* %out_block, i64 3 + %7 = load i32, i32* %arrayidx2.3, align 4 + %xor.3 = xor i32 %7, %6 + store i32 %xor.3, i32* %arrayidx2.3, align 4 + ret void +} + +declare void @use(<8 x float>) +define void @delete_pointer_bound(float* %a, float* %b, i1 %c) #0 { +; CHECK-LABEL: @delete_pointer_bound( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[TMP0:%.*]] = insertelement <2 x float*> poison, float* [[B:%.*]], i32 0 +; CHECK-NEXT: [[TMP1:%.*]] = insertelement <2 x float*> [[TMP0]], float* [[B]], i32 1 +; CHECK-NEXT: [[TMP2:%.*]] = getelementptr float, <2 x float*> [[TMP1]], <2 x i64> <i64 10, i64 14> +; CHECK-NEXT: br i1 [[C:%.*]], label [[THEN:%.*]], label [[ELSE:%.*]] +; CHECK: else: +; CHECK-NEXT: [[TMP3:%.*]] = call <2 x float> @llvm.masked.gather.v2f32.v2p0f32(<2 x float*> [[TMP2]], i32 4, <2 x i1> <i1 true, i1 true>, <2 x float> undef) +; CHECK-NEXT: [[SHUFFLE:%.*]] = shufflevector <2 x float> [[TMP3]], <2 x float> poison, <4 x i32> <i32 0, i32 0, i32 1, i32 1> +; CHECK-NEXT: [[TMP4:%.*]] = shufflevector <4 x float> [[SHUFFLE]], <4 x float> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 undef, i32 undef, i32 3, i32 undef, i32 undef> +; CHECK-NEXT: [[I71:%.*]] = shufflevector <8 x float> undef, <8 x float> [[TMP4]], <8 x i32> <i32 0, i32 1, i32 8, i32 9, i32 10, i32 5, i32 6, i32 13> +; CHECK-NEXT: call void @use(<8 x float> [[I71]]) +; CHECK-NEXT: ret void +; CHECK: then: +; CHECK-NEXT: [[A_8:%.*]] = getelementptr inbounds float, float* [[A:%.*]], i64 8 +; CHECK-NEXT: store float 0.000000e+00, float* [[A_8]], align 4 +; CHECK-NEXT: [[TMP5:%.*]] = extractelement <2 x float*> [[TMP2]], i32 1 +; CHECK-NEXT: [[L6:%.*]] = load float, float* [[TMP5]], align 4 +; CHECK-NEXT: [[A_5:%.*]] = getelementptr inbounds float, float* [[A]], i64 5 +; CHECK-NEXT: store float [[L6]], float* [[A_5]], align 4 +; CHECK-NEXT: [[A_6:%.*]] = getelementptr inbounds float, float* [[A]], i64 6 +; CHECK-NEXT: store float 0.000000e+00, float* [[A_6]], align 4 +; CHECK-NEXT: [[A_7:%.*]] = getelementptr inbounds float, float* [[A]], i64 7 +; CHECK-NEXT: store float 0.000000e+00, float* [[A_7]], align 4 +; CHECK-NEXT: ret void +; +entry: + %b.10 = getelementptr inbounds float, float* %b, i64 10 + %b.14 = getelementptr inbounds float, float* %b, i64 14 + br i1 %c, label %then, label %else + +else: + %l0 = load float, float* %b.10, align 4 + %l1 = load float, float* %b.14, align 4 + %i2 = insertelement <8 x float> undef, float %l0, i32 2 + %i3 = insertelement <8 x float> %i2, float %l0, i32 3 + %i4 = insertelement <8 x float> %i3, float %l1, i32 4 + %i7 = insertelement <8 x float> %i4, float %l1, i32 7 + call void @use(<8 x float> %i7) + ret void + +then: + %a.8 = getelementptr inbounds float, float* %a, i64 8 + store float 0.0, float* %a.8, align 4 + %l6 = load float, float* %b.14, align 4 + %a.5 = getelementptr inbounds float, float* %a, i64 5 + store float %l6, float* %a.5, align 4 + %a.6 = getelementptr inbounds float, float* %a, i64 6 + store float 0.0, float* %a.6, align 4 + %a.7 = getelementptr inbounds float, float* %a, i64 7 + store float 0.0, float* %a.7, align 4 + ret void +} + +%struct.zot = type { i16, i16, i16, i32, float, float, float, %struct.quux*, %struct.zot*, %struct.wombat*, %struct.wombat.0 } +%struct.quux = type { i16, %struct.quux*, %struct.quux* } +%struct.wombat = type { i32, i16, i8, i8, %struct.eggs* } +%struct.eggs = type { float, i8, %struct.ham } +%struct.ham = type { [2 x double], [8 x i8] } +%struct.wombat.0 = type { %struct.bar } +%struct.bar = type { [3 x double], [3 x double], double, double, i16, [3 x double]*, i32, [3 x double] } + +define double @preserve_loop_info(%struct.zot* %arg) { +; CHECK-LABEL: @preserve_loop_info( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[TMP:%.*]] = alloca [3 x double], align 16 +; CHECK-NEXT: br label [[OUTER_HEADER:%.*]] +; CHECK: outer.header: +; CHECK-NEXT: br label [[INNER:%.*]] +; CHECK: inner: +; CHECK-NEXT: br i1 undef, label [[OUTER_LATCH:%.*]], label [[INNER]] +; CHECK: outer.latch: +; CHECK-NEXT: br i1 undef, label [[BB:%.*]], label [[OUTER_HEADER]] +; CHECK: bb: +; CHECK-NEXT: [[TMP5:%.*]] = load [3 x double]*, [3 x double]** undef, align 8 +; CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds [3 x double], [3 x double]* [[TMP]], i64 0, i64 0 +; CHECK-NEXT: [[TMP7:%.*]] = getelementptr inbounds [3 x double], [3 x double]* [[TMP]], i64 0, i64 1 +; CHECK-NEXT: br label [[LOOP_3HEADER:%.*]] +; CHECK: loop.3header: +; CHECK-NEXT: br i1 undef, label [[LOOP_3LATCH:%.*]], label [[BB9:%.*]] +; CHECK: bb9: +; CHECK-NEXT: [[TMP10:%.*]] = getelementptr inbounds [3 x double], [3 x double]* [[TMP5]], i64 undef, i64 1 +; CHECK-NEXT: store double undef, double* [[TMP6]], align 16 +; CHECK-NEXT: [[TMP12:%.*]] = load double, double* [[TMP10]], align 8 +; CHECK-NEXT: store double [[TMP12]], double* [[TMP7]], align 8 +; CHECK-NEXT: br label [[LOOP_3LATCH]] +; CHECK: loop.3latch: +; CHECK-NEXT: br i1 undef, label [[BB14:%.*]], label [[LOOP_3HEADER]] +; CHECK: bb14: +; CHECK-NEXT: [[TMP15:%.*]] = call double undef(double* [[TMP6]], %struct.zot* [[ARG:%.*]]) +; CHECK-NEXT: ret double undef +; +entry: + %tmp = alloca [3 x double], align 16 + br label %outer.header + +outer.header: ; preds = %bb3, %bb + br label %inner + +inner: + br i1 undef, label %outer.latch, label %inner + +outer.latch: ; preds = %bb16 + br i1 undef, label %bb, label %outer.header + +bb: ; preds = %bb3 + %tmp5 = load [3 x double]*, [3 x double]** undef, align 8 + %tmp6 = getelementptr inbounds [3 x double], [3 x double]* %tmp, i64 0, i64 0 + %tmp7 = getelementptr inbounds [3 x double], [3 x double]* %tmp, i64 0, i64 1 + br label %loop.3header + +loop.3header: ; preds = %bb13, %bb4 + br i1 undef, label %loop.3latch, label %bb9 + +bb9: ; preds = %bb8 + %tmp10 = getelementptr inbounds [3 x double], [3 x double]* %tmp5, i64 undef, i64 1 + store double undef, double* %tmp6, align 16 + %tmp12 = load double, double* %tmp10, align 8 + store double %tmp12, double* %tmp7, align 8 + br label %loop.3latch + +loop.3latch: ; preds = %bb11, %bb8 + br i1 undef, label %bb14, label %loop.3header + +bb14: ; preds = %bb13 + %tmp15 = call double undef(double* %tmp6, %struct.zot* %arg) + ret double undef +} + +attributes #0 = { "target-features"="+avx2" } _______________________________________________ llvm-branch-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
