[llvm] [llvm-objdump][ELF] Add Section size check. (#86612) (PR #124470)
via llvm-commits
llvm-commits at lists.llvm.org
Tue Jan 28 04:07:38 PST 2025
https://github.com/cabbaken updated https://github.com/llvm/llvm-project/pull/124470
>From c9d4137a065d401b373b44cb4e139124936c17ee Mon Sep 17 00:00:00 2001
From: cabbaken <cabbaken at outlook.com>
Date: Sun, 26 Jan 2025 17:37:46 +0800
Subject: [PATCH 1/2] [llvm-objdump][ELF] Add Section size check. (#86612)
This change make the check of the section size to
avoid crashing of llvm-objdump when processing
misformated elf file.
Signed-off-by: cabbaken <cabbaken at outlook.com>
---
llvm/tools/llvm-objdump/ELFDump.cpp | 12 ++++++++++++
1 file changed, 12 insertions(+)
diff --git a/llvm/tools/llvm-objdump/ELFDump.cpp b/llvm/tools/llvm-objdump/ELFDump.cpp
index e9e5b059f1786e..83699f8267575c 100644
--- a/llvm/tools/llvm-objdump/ELFDump.cpp
+++ b/llvm/tools/llvm-objdump/ELFDump.cpp
@@ -221,6 +221,13 @@ template <class ELFT> void ELFDumper<ELFT>::printDynamicSection() {
std::string TagFmt = " %-" + std::to_string(MaxLen) + "s ";
outs() << "\nDynamic Section:\n";
+ auto DynamicSectionOrErr = Elf.getSection(ELF::SHT_DYNAMIC);
+ if (!DynamicSectionOrErr) {
+ reportWarning(toString(DynamicSectionOrErr.takeError()), Obj.getFileName());
+ return;
+ }
+ const auto StringTableSize = (*DynamicSectionOrErr)->sh_size;
+
for (const typename ELFT::Dyn &Dyn : DynamicEntries) {
if (Dyn.d_tag == ELF::DT_NULL)
continue;
@@ -235,6 +242,11 @@ template <class ELFT> void ELFDumper<ELFT>::printDynamicSection() {
Expected<StringRef> StrTabOrErr = getDynamicStrTab(Elf);
if (StrTabOrErr) {
const char *Data = StrTabOrErr->data();
+ if (Dyn.getVal() > StringTableSize) {
+ reportWarning("Invalid string table offset for section .dynstr",
+ Obj.getFileName());
+ continue;
+ }
outs() << format(TagFmt.c_str(), Str.c_str()) << Data + Dyn.getVal()
<< "\n";
continue;
>From 7d720ebbd1d6e8c1d8f545ed3a710e4d0099c909 Mon Sep 17 00:00:00 2001
From: Florian Hahn <flo at fhahn.com>
Date: Sun, 26 Jan 2025 13:46:00 +0000
Subject: [PATCH 2/2] [LV] Add more tests for narrowing interleave groups for
AArch64.
Add additional tests for
https://github.com/llvm/llvm-project/pull/106441.
---
...sform-narrow-interleave-to-widen-memory.ll | 205 ++++++++++++++++++
1 file changed, 205 insertions(+)
create mode 100644 llvm/test/Transforms/LoopVectorize/AArch64/transform-narrow-interleave-to-widen-memory.ll
diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/transform-narrow-interleave-to-widen-memory.ll b/llvm/test/Transforms/LoopVectorize/AArch64/transform-narrow-interleave-to-widen-memory.ll
new file mode 100644
index 00000000000000..3fca274a3bb12b
--- /dev/null
+++ b/llvm/test/Transforms/LoopVectorize/AArch64/transform-narrow-interleave-to-widen-memory.ll
@@ -0,0 +1,205 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals none --version 5
+; RUN: opt -p loop-vectorize -S %s | FileCheck --check-prefixes=CHECK %s
+
+target datalayout = "e-m:o-p270:32:32-p271:32:32-p272:64:64-i64:64-i128:128-n32:64-S128-Fn32"
+target triple = "arm64-apple-macosx15.0.0"
+
+define void @test_complex_add_float(ptr %res, ptr noalias %A, ptr noalias %B, i64 %N) {
+; CHECK-LABEL: define void @test_complex_add_float(
+; CHECK-SAME: ptr [[RES:%.*]], ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], i64 [[N:%.*]]) {
+; CHECK-NEXT: [[ENTRY:.*]]:
+; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[N]], 8
+; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
+; CHECK: [[VECTOR_PH]]:
+; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N]], 8
+; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[N]], [[N_MOD_VF]]
+; CHECK-NEXT: br label %[[VECTOR_BODY:.*]]
+; CHECK: [[VECTOR_BODY]]:
+; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
+; CHECK-NEXT: [[IV:%.*]] = add i64 [[INDEX]], 0
+; CHECK-NEXT: [[TMP1:%.*]] = add i64 [[INDEX]], 4
+; CHECK-NEXT: [[GEP_A_0:%.*]] = getelementptr inbounds nuw { float, float }, ptr [[A]], i64 [[IV]]
+; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds nuw { float, float }, ptr [[A]], i64 [[TMP1]]
+; CHECK-NEXT: [[GEP_B_0:%.*]] = getelementptr inbounds nuw { float, float }, ptr [[B]], i64 [[IV]]
+; CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds nuw { float, float }, ptr [[B]], i64 [[TMP1]]
+; CHECK-NEXT: [[WIDE_VEC:%.*]] = load <8 x float>, ptr [[GEP_A_0]], align 4
+; CHECK-NEXT: [[STRIDED_VEC:%.*]] = shufflevector <8 x float> [[WIDE_VEC]], <8 x float> poison, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
+; CHECK-NEXT: [[STRIDED_VEC1:%.*]] = shufflevector <8 x float> [[WIDE_VEC]], <8 x float> poison, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
+; CHECK-NEXT: [[WIDE_VEC2:%.*]] = load <8 x float>, ptr [[TMP3]], align 4
+; CHECK-NEXT: [[STRIDED_VEC3:%.*]] = shufflevector <8 x float> [[WIDE_VEC2]], <8 x float> poison, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
+; CHECK-NEXT: [[STRIDED_VEC4:%.*]] = shufflevector <8 x float> [[WIDE_VEC2]], <8 x float> poison, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
+; CHECK-NEXT: [[WIDE_VEC5:%.*]] = load <8 x float>, ptr [[GEP_B_0]], align 4
+; CHECK-NEXT: [[STRIDED_VEC6:%.*]] = shufflevector <8 x float> [[WIDE_VEC5]], <8 x float> poison, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
+; CHECK-NEXT: [[STRIDED_VEC7:%.*]] = shufflevector <8 x float> [[WIDE_VEC5]], <8 x float> poison, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
+; CHECK-NEXT: [[WIDE_VEC8:%.*]] = load <8 x float>, ptr [[TMP5]], align 4
+; CHECK-NEXT: [[STRIDED_VEC9:%.*]] = shufflevector <8 x float> [[WIDE_VEC8]], <8 x float> poison, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
+; CHECK-NEXT: [[STRIDED_VEC10:%.*]] = shufflevector <8 x float> [[WIDE_VEC8]], <8 x float> poison, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
+; CHECK-NEXT: [[TMP6:%.*]] = fadd <4 x float> [[STRIDED_VEC]], [[STRIDED_VEC6]]
+; CHECK-NEXT: [[TMP7:%.*]] = fadd <4 x float> [[STRIDED_VEC3]], [[STRIDED_VEC9]]
+; CHECK-NEXT: [[TMP8:%.*]] = fadd <4 x float> [[STRIDED_VEC1]], [[STRIDED_VEC7]]
+; CHECK-NEXT: [[TMP9:%.*]] = fadd <4 x float> [[STRIDED_VEC4]], [[STRIDED_VEC10]]
+; CHECK-NEXT: [[TMP10:%.*]] = getelementptr inbounds nuw { float, float }, ptr [[RES]], i64 [[IV]]
+; CHECK-NEXT: [[TMP11:%.*]] = getelementptr inbounds nuw { float, float }, ptr [[RES]], i64 [[TMP1]]
+; CHECK-NEXT: [[TMP12:%.*]] = shufflevector <4 x float> [[TMP6]], <4 x float> [[TMP8]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+; CHECK-NEXT: [[INTERLEAVED_VEC:%.*]] = shufflevector <8 x float> [[TMP12]], <8 x float> poison, <8 x i32> <i32 0, i32 4, i32 1, i32 5, i32 2, i32 6, i32 3, i32 7>
+; CHECK-NEXT: store <8 x float> [[INTERLEAVED_VEC]], ptr [[TMP10]], align 4
+; CHECK-NEXT: [[TMP13:%.*]] = shufflevector <4 x float> [[TMP7]], <4 x float> [[TMP9]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+; CHECK-NEXT: [[INTERLEAVED_VEC11:%.*]] = shufflevector <8 x float> [[TMP13]], <8 x float> poison, <8 x i32> <i32 0, i32 4, i32 1, i32 5, i32 2, i32 6, i32 3, i32 7>
+; CHECK-NEXT: store <8 x float> [[INTERLEAVED_VEC11]], ptr [[TMP11]], align 4
+; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 8
+; CHECK-NEXT: [[TMP14:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
+; CHECK-NEXT: br i1 [[TMP14]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
+; CHECK: [[MIDDLE_BLOCK]]:
+; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[N]], [[N_VEC]]
+; CHECK-NEXT: br i1 [[CMP_N]], label %[[EXIT:.*]], label %[[SCALAR_PH]]
+; CHECK: [[SCALAR_PH]]:
+; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ]
+; CHECK-NEXT: br label %[[LOOP:.*]]
+; CHECK: [[LOOP]]:
+; CHECK-NEXT: [[IV1:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ]
+; CHECK-NEXT: [[GEP_A_2:%.*]] = getelementptr inbounds nuw { float, float }, ptr [[A]], i64 [[IV1]]
+; CHECK-NEXT: [[GEP_B_2:%.*]] = getelementptr inbounds nuw { float, float }, ptr [[B]], i64 [[IV1]]
+; CHECK-NEXT: [[L_A_0:%.*]] = load float, ptr [[GEP_A_2]], align 4
+; CHECK-NEXT: [[GEP_A_1:%.*]] = getelementptr inbounds nuw i8, ptr [[GEP_A_2]], i64 4
+; CHECK-NEXT: [[L_A_1:%.*]] = load float, ptr [[GEP_A_1]], align 4
+; CHECK-NEXT: [[L_B_0:%.*]] = load float, ptr [[GEP_B_2]], align 4
+; CHECK-NEXT: [[ADD_0:%.*]] = fadd float [[L_A_0]], [[L_B_0]]
+; CHECK-NEXT: [[GEP_B_1:%.*]] = getelementptr inbounds nuw i8, ptr [[GEP_B_2]], i64 4
+; CHECK-NEXT: [[L_B_1:%.*]] = load float, ptr [[GEP_B_1]], align 4
+; CHECK-NEXT: [[ADD_1:%.*]] = fadd float [[L_A_1]], [[L_B_1]]
+; CHECK-NEXT: [[GEP_RES_0:%.*]] = getelementptr inbounds nuw { float, float }, ptr [[RES]], i64 [[IV1]]
+; CHECK-NEXT: store float [[ADD_0]], ptr [[GEP_RES_0]], align 4
+; CHECK-NEXT: [[GEP_RES_1:%.*]] = getelementptr inbounds nuw i8, ptr [[GEP_RES_0]], i64 4
+; CHECK-NEXT: store float [[ADD_1]], ptr [[GEP_RES_1]], align 4
+; CHECK-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV1]], 1
+; CHECK-NEXT: [[EC:%.*]] = icmp eq i64 [[IV_NEXT]], [[N]]
+; CHECK-NEXT: br i1 [[EC]], label %[[EXIT]], label %[[LOOP]], !llvm.loop [[LOOP3:![0-9]+]]
+; CHECK: [[EXIT]]:
+; CHECK-NEXT: ret void
+;
+entry:
+ br label %loop
+
+loop:
+ %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
+ %gep.A.0 = getelementptr inbounds nuw { float, float }, ptr %A, i64 %iv
+ %gep.B.0 = getelementptr inbounds nuw { float, float }, ptr %B, i64 %iv
+ %l.A.0 = load float, ptr %gep.A.0, align 4
+ %gep.A.1 = getelementptr inbounds nuw i8, ptr %gep.A.0, i64 4
+ %l.A.1 = load float, ptr %gep.A.1, align 4
+ %l.B.0 = load float, ptr %gep.B.0, align 4
+ %add.0 = fadd float %l.A.0, %l.B.0
+ %gep.B.1 = getelementptr inbounds nuw i8, ptr %gep.B.0, i64 4
+ %l.B.1 = load float, ptr %gep.B.1, align 4
+ %add.1 = fadd float %l.A.1, %l.B.1
+ %gep.res.0 = getelementptr inbounds nuw { float, float }, ptr %res, i64 %iv
+ store float %add.0, ptr %gep.res.0, align 4
+ %gep.res.1 = getelementptr inbounds nuw i8, ptr %gep.res.0, i64 4
+ store float %add.1, ptr %gep.res.1, align 4
+ %iv.next = add nuw nsw i64 %iv, 1
+ %ec = icmp eq i64 %iv.next, %N
+ br i1 %ec, label %exit, label %loop
+
+exit:
+ ret void
+}
+
+define void @test_complex_add_double(ptr %res, ptr noalias %A, ptr noalias %B, i64 %N) {
+; CHECK-LABEL: define void @test_complex_add_double(
+; CHECK-SAME: ptr [[RES:%.*]], ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], i64 [[N:%.*]]) {
+; CHECK-NEXT: [[ENTRY:.*]]:
+; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[N]], 4
+; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
+; CHECK: [[VECTOR_PH]]:
+; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N]], 4
+; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[N]], [[N_MOD_VF]]
+; CHECK-NEXT: br label %[[VECTOR_BODY:.*]]
+; CHECK: [[VECTOR_BODY]]:
+; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
+; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0
+; CHECK-NEXT: [[TMP1:%.*]] = add i64 [[INDEX]], 2
+; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds nuw { double, double }, ptr [[A]], i64 [[TMP0]]
+; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds nuw { double, double }, ptr [[A]], i64 [[TMP1]]
+; CHECK-NEXT: [[TMP4:%.*]] = getelementptr inbounds nuw { double, double }, ptr [[B]], i64 [[TMP0]]
+; CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds nuw { double, double }, ptr [[B]], i64 [[TMP1]]
+; CHECK-NEXT: [[WIDE_VEC:%.*]] = load <4 x double>, ptr [[TMP2]], align 4
+; CHECK-NEXT: [[STRIDED_VEC:%.*]] = shufflevector <4 x double> [[WIDE_VEC]], <4 x double> poison, <2 x i32> <i32 0, i32 2>
+; CHECK-NEXT: [[STRIDED_VEC1:%.*]] = shufflevector <4 x double> [[WIDE_VEC]], <4 x double> poison, <2 x i32> <i32 1, i32 3>
+; CHECK-NEXT: [[WIDE_VEC2:%.*]] = load <4 x double>, ptr [[TMP3]], align 4
+; CHECK-NEXT: [[STRIDED_VEC3:%.*]] = shufflevector <4 x double> [[WIDE_VEC2]], <4 x double> poison, <2 x i32> <i32 0, i32 2>
+; CHECK-NEXT: [[STRIDED_VEC4:%.*]] = shufflevector <4 x double> [[WIDE_VEC2]], <4 x double> poison, <2 x i32> <i32 1, i32 3>
+; CHECK-NEXT: [[WIDE_VEC5:%.*]] = load <4 x double>, ptr [[TMP4]], align 4
+; CHECK-NEXT: [[STRIDED_VEC6:%.*]] = shufflevector <4 x double> [[WIDE_VEC5]], <4 x double> poison, <2 x i32> <i32 0, i32 2>
+; CHECK-NEXT: [[STRIDED_VEC7:%.*]] = shufflevector <4 x double> [[WIDE_VEC5]], <4 x double> poison, <2 x i32> <i32 1, i32 3>
+; CHECK-NEXT: [[WIDE_VEC8:%.*]] = load <4 x double>, ptr [[TMP5]], align 4
+; CHECK-NEXT: [[STRIDED_VEC9:%.*]] = shufflevector <4 x double> [[WIDE_VEC8]], <4 x double> poison, <2 x i32> <i32 0, i32 2>
+; CHECK-NEXT: [[STRIDED_VEC10:%.*]] = shufflevector <4 x double> [[WIDE_VEC8]], <4 x double> poison, <2 x i32> <i32 1, i32 3>
+; CHECK-NEXT: [[TMP6:%.*]] = fadd <2 x double> [[STRIDED_VEC]], [[STRIDED_VEC6]]
+; CHECK-NEXT: [[TMP7:%.*]] = fadd <2 x double> [[STRIDED_VEC3]], [[STRIDED_VEC9]]
+; CHECK-NEXT: [[TMP8:%.*]] = fadd <2 x double> [[STRIDED_VEC1]], [[STRIDED_VEC7]]
+; CHECK-NEXT: [[TMP9:%.*]] = fadd <2 x double> [[STRIDED_VEC4]], [[STRIDED_VEC10]]
+; CHECK-NEXT: [[TMP10:%.*]] = getelementptr inbounds nuw { double, double }, ptr [[RES]], i64 [[TMP0]]
+; CHECK-NEXT: [[TMP11:%.*]] = getelementptr inbounds nuw { double, double }, ptr [[RES]], i64 [[TMP1]]
+; CHECK-NEXT: [[TMP12:%.*]] = shufflevector <2 x double> [[TMP6]], <2 x double> [[TMP8]], <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+; CHECK-NEXT: [[INTERLEAVED_VEC:%.*]] = shufflevector <4 x double> [[TMP12]], <4 x double> poison, <4 x i32> <i32 0, i32 2, i32 1, i32 3>
+; CHECK-NEXT: store <4 x double> [[INTERLEAVED_VEC]], ptr [[TMP10]], align 4
+; CHECK-NEXT: [[TMP13:%.*]] = shufflevector <2 x double> [[TMP7]], <2 x double> [[TMP9]], <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+; CHECK-NEXT: [[INTERLEAVED_VEC11:%.*]] = shufflevector <4 x double> [[TMP13]], <4 x double> poison, <4 x i32> <i32 0, i32 2, i32 1, i32 3>
+; CHECK-NEXT: store <4 x double> [[INTERLEAVED_VEC11]], ptr [[TMP11]], align 4
+; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
+; CHECK-NEXT: [[TMP14:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
+; CHECK-NEXT: br i1 [[TMP14]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
+; CHECK: [[MIDDLE_BLOCK]]:
+; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[N]], [[N_VEC]]
+; CHECK-NEXT: br i1 [[CMP_N]], label %[[EXIT:.*]], label %[[SCALAR_PH]]
+; CHECK: [[SCALAR_PH]]:
+; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ]
+; CHECK-NEXT: br label %[[LOOP:.*]]
+; CHECK: [[LOOP]]:
+; CHECK-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ]
+; CHECK-NEXT: [[GEP_A_0:%.*]] = getelementptr inbounds nuw { double, double }, ptr [[A]], i64 [[IV]]
+; CHECK-NEXT: [[GEP_B_0:%.*]] = getelementptr inbounds nuw { double, double }, ptr [[B]], i64 [[IV]]
+; CHECK-NEXT: [[L_A_0:%.*]] = load double, ptr [[GEP_A_0]], align 4
+; CHECK-NEXT: [[GEP_A_1:%.*]] = getelementptr inbounds nuw i8, ptr [[GEP_A_0]], i64 8
+; CHECK-NEXT: [[L_A_1:%.*]] = load double, ptr [[GEP_A_1]], align 4
+; CHECK-NEXT: [[L_B_0:%.*]] = load double, ptr [[GEP_B_0]], align 4
+; CHECK-NEXT: [[ADD_0:%.*]] = fadd double [[L_A_0]], [[L_B_0]]
+; CHECK-NEXT: [[GEP_B_1:%.*]] = getelementptr inbounds nuw i8, ptr [[GEP_B_0]], i64 8
+; CHECK-NEXT: [[L_B_1:%.*]] = load double, ptr [[GEP_B_1]], align 4
+; CHECK-NEXT: [[ADD_1:%.*]] = fadd double [[L_A_1]], [[L_B_1]]
+; CHECK-NEXT: [[GEP_RES_0:%.*]] = getelementptr inbounds nuw { double, double }, ptr [[RES]], i64 [[IV]]
+; CHECK-NEXT: store double [[ADD_0]], ptr [[GEP_RES_0]], align 4
+; CHECK-NEXT: [[GEP_RES_1:%.*]] = getelementptr inbounds nuw i8, ptr [[GEP_RES_0]], i64 8
+; CHECK-NEXT: store double [[ADD_1]], ptr [[GEP_RES_1]], align 4
+; CHECK-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
+; CHECK-NEXT: [[EC:%.*]] = icmp eq i64 [[IV_NEXT]], [[N]]
+; CHECK-NEXT: br i1 [[EC]], label %[[EXIT]], label %[[LOOP]], !llvm.loop [[LOOP5:![0-9]+]]
+; CHECK: [[EXIT]]:
+; CHECK-NEXT: ret void
+;
+entry:
+ br label %loop
+
+loop:
+ %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
+ %gep.A.0 = getelementptr inbounds nuw { double, double }, ptr %A, i64 %iv
+ %gep.B.0 = getelementptr inbounds nuw { double, double }, ptr %B, i64 %iv
+ %l.A.0 = load double, ptr %gep.A.0, align 4
+ %gep.A.1 = getelementptr inbounds nuw i8, ptr %gep.A.0, i64 8
+ %l.A.1 = load double, ptr %gep.A.1, align 4
+ %l.B.0 = load double, ptr %gep.B.0, align 4
+ %add.0 = fadd double %l.A.0, %l.B.0
+ %gep.B.1 = getelementptr inbounds nuw i8, ptr %gep.B.0, i64 8
+ %l.B.1 = load double, ptr %gep.B.1, align 4
+ %add.1 = fadd double %l.A.1, %l.B.1
+ %gep.res.0 = getelementptr inbounds nuw { double, double }, ptr %res, i64 %iv
+ store double %add.0, ptr %gep.res.0, align 4
+ %gep.res.1 = getelementptr inbounds nuw i8, ptr %gep.res.0, i64 8
+ store double %add.1, ptr %gep.res.1, align 4
+ %iv.next = add nuw nsw i64 %iv, 1
+ %ec = icmp eq i64 %iv.next, %N
+ br i1 %ec, label %exit, label %loop
+
+exit:
+ ret void
+}
More information about the llvm-commits
mailing list