[llvm] [msan][NFCI] Add tests for Arm NEON add reduction to scalar (PR #125271)
Thurston Dang via llvm-commits
llvm-commits at lists.llvm.org
Fri Jan 31 11:16:08 PST 2025
https://github.com/thurstond created https://github.com/llvm/llvm-project/pull/125271
Forked from llvm/test/CodeGen/AArch64/arm64-vaddv.ll
Incorrectly handled by handleUnknownInstruction:
- llvm.aarch64.neon.faddv
- llvm.aarch64.neon.saddv
- llvm.aarch64.neon.uaddv
>From 97b147803021b6dd8da1256568d4a36d51a8502e Mon Sep 17 00:00:00 2001
From: Thurston Dang <thurston at google.com>
Date: Fri, 31 Jan 2025 19:07:07 +0000
Subject: [PATCH 1/2] [msan][NFCI] Add tests for Arm NEON add reduction to
scalar
Forked from llvm/test/CodeGen/AArch64/arm64-vaddv.ll
Incorrectly handled by handleUnknownInstruction:
- llvm.aarch64.neon.faddv
- llvm.aarch64.neon.saddv
- llvm.aarch64.neon.uaddv
---
.../MemorySanitizer/AArch64/arm64-vaddv.ll | 899 ++++++++++++++++++
1 file changed, 899 insertions(+)
create mode 100644 llvm/test/Instrumentation/MemorySanitizer/AArch64/arm64-vaddv.ll
diff --git a/llvm/test/Instrumentation/MemorySanitizer/AArch64/arm64-vaddv.ll b/llvm/test/Instrumentation/MemorySanitizer/AArch64/arm64-vaddv.ll
new file mode 100644
index 00000000000000..d8abb65894c7f2
--- /dev/null
+++ b/llvm/test/Instrumentation/MemorySanitizer/AArch64/arm64-vaddv.ll
@@ -0,0 +1,899 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
+; RUN: opt < %s -passes=msan -S | FileCheck %s
+;
+; Forked from llvm/test/CodeGen/AArch64/arm64-vaddv.ll
+;
+; Incorrectly handled by handleUnknownInstruction:
+; - llvm.aarch64.neon.faddv
+; - llvm.aarch64.neon.saddv
+; - llvm.aarch64.neon.uaddv
+
+target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128"
+target triple = "aarch64--linux-android9001"
+
+define signext i8 @test_vaddv_s8(<8 x i8> %a1) #0 {
+; CHECK-LABEL: define signext i8 @test_vaddv_s8(
+; CHECK-SAME: <8 x i8> [[A1:%.*]]) #[[ATTR0:[0-9]+]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <8 x i8>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <8 x i8> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1:![0-9]+]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3:[0-9]+]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.saddv.i32.v8i8(<8 x i8> [[A1]])
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i32 [[VADDV_I]] to i8
+; CHECK-NEXT: store i8 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i8 [[TMP4]]
+;
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.saddv.i32.v8i8(<8 x i8> %a1)
+ %0 = trunc i32 %vaddv.i to i8
+ ret i8 %0
+}
+
+define <8 x i8> @test_vaddv_s8_used_by_laneop(<8 x i8> %a1, <8 x i8> %a2) #0 {
+; CHECK-LABEL: define <8 x i8> @test_vaddv_s8_used_by_laneop(
+; CHECK-SAME: <8 x i8> [[A1:%.*]], <8 x i8> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <8 x i8>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 8) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i8>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <8 x i8> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @llvm.aarch64.neon.saddv.i32.v8i8(<8 x i8> [[A2]])
+; CHECK-NEXT: [[TMP6:%.*]] = trunc i32 [[TMP5]] to i8
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <8 x i8> [[TMP1]], i8 0, i32 3
+; CHECK-NEXT: [[TMP7:%.*]] = insertelement <8 x i8> [[A1]], i8 [[TMP6]], i32 3
+; CHECK-NEXT: store <8 x i8> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <8 x i8> [[TMP7]]
+;
+entry:
+ %0 = tail call i32 @llvm.aarch64.neon.saddv.i32.v8i8(<8 x i8> %a2)
+ %1 = trunc i32 %0 to i8
+ %2 = insertelement <8 x i8> %a1, i8 %1, i32 3
+ ret <8 x i8> %2
+}
+
+define signext i16 @test_vaddv_s16(<4 x i16> %a1) #0 {
+; CHECK-LABEL: define signext i16 @test_vaddv_s16(
+; CHECK-SAME: <4 x i16> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <4 x i16>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <4 x i16> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.saddv.i32.v4i16(<4 x i16> [[A1]])
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i32 [[VADDV_I]] to i16
+; CHECK-NEXT: store i16 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i16 [[TMP4]]
+;
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.saddv.i32.v4i16(<4 x i16> %a1)
+ %0 = trunc i32 %vaddv.i to i16
+ ret i16 %0
+}
+
+define <4 x i16> @test_vaddv_s16_used_by_laneop(<4 x i16> %a1, <4 x i16> %a2) #0 {
+; CHECK-LABEL: define <4 x i16> @test_vaddv_s16_used_by_laneop(
+; CHECK-SAME: <4 x i16> [[A1:%.*]], <4 x i16> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <4 x i16>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 8) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i16>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i16> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @llvm.aarch64.neon.saddv.i32.v4i16(<4 x i16> [[A2]])
+; CHECK-NEXT: [[TMP6:%.*]] = trunc i32 [[TMP5]] to i16
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <4 x i16> [[TMP1]], i16 0, i32 3
+; CHECK-NEXT: [[TMP7:%.*]] = insertelement <4 x i16> [[A1]], i16 [[TMP6]], i32 3
+; CHECK-NEXT: store <4 x i16> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <4 x i16> [[TMP7]]
+;
+entry:
+ %0 = tail call i32 @llvm.aarch64.neon.saddv.i32.v4i16(<4 x i16> %a2)
+ %1 = trunc i32 %0 to i16
+ %2 = insertelement <4 x i16> %a1, i16 %1, i32 3
+ ret <4 x i16> %2
+}
+
+define i32 @test_vaddv_s32(<2 x i32> %a1) #0 {
+; CHECK-LABEL: define i32 @test_vaddv_s32(
+; CHECK-SAME: <2 x i32> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <2 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <2 x i32> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.saddv.i32.v2i32(<2 x i32> [[A1]])
+; CHECK-NEXT: store i32 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i32 [[VADDV_I]]
+;
+; 2 x i32 is not supported by the ISA, thus, this is a special case
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.saddv.i32.v2i32(<2 x i32> %a1)
+ ret i32 %vaddv.i
+}
+
+define <2 x i32> @test_vaddv_s32_used_by_laneop(<2 x i32> %a1, <2 x i32> %a2) #0 {
+; CHECK-LABEL: define <2 x i32> @test_vaddv_s32_used_by_laneop(
+; CHECK-SAME: <2 x i32> [[A1:%.*]], <2 x i32> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <2 x i32>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 8) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <2 x i32> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @llvm.aarch64.neon.saddv.i32.v2i32(<2 x i32> [[A2]])
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <2 x i32> [[TMP1]], i32 0, i32 1
+; CHECK-NEXT: [[TMP6:%.*]] = insertelement <2 x i32> [[A1]], i32 [[TMP5]], i32 1
+; CHECK-NEXT: store <2 x i32> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <2 x i32> [[TMP6]]
+;
+entry:
+ %0 = tail call i32 @llvm.aarch64.neon.saddv.i32.v2i32(<2 x i32> %a2)
+ %1 = insertelement <2 x i32> %a1, i32 %0, i32 1
+ ret <2 x i32> %1
+}
+
+define i64 @test_vaddv_s64(<2 x i64> %a1) #0 {
+; CHECK-LABEL: define i64 @test_vaddv_s64(
+; CHECK-SAME: <2 x i64> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <2 x i64>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <2 x i64> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i64 @llvm.aarch64.neon.saddv.i64.v2i64(<2 x i64> [[A1]])
+; CHECK-NEXT: store i64 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i64 [[VADDV_I]]
+;
+entry:
+ %vaddv.i = tail call i64 @llvm.aarch64.neon.saddv.i64.v2i64(<2 x i64> %a1)
+ ret i64 %vaddv.i
+}
+
+define <2 x i64> @test_vaddv_s64_used_by_laneop(<2 x i64> %a1, <2 x i64> %a2) #0 {
+; CHECK-LABEL: define <2 x i64> @test_vaddv_s64_used_by_laneop(
+; CHECK-SAME: <2 x i64> [[A1:%.*]], <2 x i64> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <2 x i64>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 16) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <2 x i64> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i64 @llvm.aarch64.neon.saddv.i64.v2i64(<2 x i64> [[A2]])
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <2 x i64> [[TMP1]], i64 0, i64 1
+; CHECK-NEXT: [[TMP6:%.*]] = insertelement <2 x i64> [[A1]], i64 [[TMP5]], i64 1
+; CHECK-NEXT: store <2 x i64> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <2 x i64> [[TMP6]]
+;
+entry:
+ %0 = tail call i64 @llvm.aarch64.neon.saddv.i64.v2i64(<2 x i64> %a2)
+ %1 = insertelement <2 x i64> %a1, i64 %0, i64 1
+ ret <2 x i64> %1
+}
+
+define zeroext i8 @test_vaddv_u8(<8 x i8> %a1) #0 {
+; CHECK-LABEL: define zeroext i8 @test_vaddv_u8(
+; CHECK-SAME: <8 x i8> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <8 x i8>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <8 x i8> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v8i8(<8 x i8> [[A1]])
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i32 [[VADDV_I]] to i8
+; CHECK-NEXT: store i8 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i8 [[TMP4]]
+;
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.uaddv.i32.v8i8(<8 x i8> %a1)
+ %0 = trunc i32 %vaddv.i to i8
+ ret i8 %0
+}
+
+define <8 x i8> @test_vaddv_u8_used_by_laneop(<8 x i8> %a1, <8 x i8> %a2) #0 {
+; CHECK-LABEL: define <8 x i8> @test_vaddv_u8_used_by_laneop(
+; CHECK-SAME: <8 x i8> [[A1:%.*]], <8 x i8> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <8 x i8>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 8) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i8>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <8 x i8> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v8i8(<8 x i8> [[A2]])
+; CHECK-NEXT: [[TMP6:%.*]] = trunc i32 [[TMP5]] to i8
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <8 x i8> [[TMP1]], i8 0, i32 3
+; CHECK-NEXT: [[TMP7:%.*]] = insertelement <8 x i8> [[A1]], i8 [[TMP6]], i32 3
+; CHECK-NEXT: store <8 x i8> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <8 x i8> [[TMP7]]
+;
+entry:
+ %0 = tail call i32 @llvm.aarch64.neon.uaddv.i32.v8i8(<8 x i8> %a2)
+ %1 = trunc i32 %0 to i8
+ %2 = insertelement <8 x i8> %a1, i8 %1, i32 3
+ ret <8 x i8> %2
+}
+
+define i32 @test_vaddv_u8_masked(<8 x i8> %a1) #0 {
+; CHECK-LABEL: define i32 @test_vaddv_u8_masked(
+; CHECK-SAME: <8 x i8> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <8 x i8>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <8 x i8> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v8i8(<8 x i8> [[A1]])
+; CHECK-NEXT: [[TMP4:%.*]] = and i32 [[VADDV_I]], 0
+; CHECK-NEXT: [[TMP5:%.*]] = or i32 0, [[TMP4]]
+; CHECK-NEXT: [[TMP6:%.*]] = or i32 [[TMP5]], 0
+; CHECK-NEXT: [[TMP7:%.*]] = and i32 [[VADDV_I]], 511
+; CHECK-NEXT: store i32 [[TMP6]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i32 [[TMP7]]
+;
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.uaddv.i32.v8i8(<8 x i8> %a1)
+ %0 = and i32 %vaddv.i, 511 ; 0x1ff
+ ret i32 %0
+}
+
+define zeroext i16 @test_vaddv_u16(<4 x i16> %a1) #0 {
+; CHECK-LABEL: define zeroext i16 @test_vaddv_u16(
+; CHECK-SAME: <4 x i16> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <4 x i16>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <4 x i16> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v4i16(<4 x i16> [[A1]])
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i32 [[VADDV_I]] to i16
+; CHECK-NEXT: store i16 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i16 [[TMP4]]
+;
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.uaddv.i32.v4i16(<4 x i16> %a1)
+ %0 = trunc i32 %vaddv.i to i16
+ ret i16 %0
+}
+
+define <4 x i16> @test_vaddv_u16_used_by_laneop(<4 x i16> %a1, <4 x i16> %a2) #0 {
+; CHECK-LABEL: define <4 x i16> @test_vaddv_u16_used_by_laneop(
+; CHECK-SAME: <4 x i16> [[A1:%.*]], <4 x i16> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <4 x i16>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 8) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i16>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i16> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v4i16(<4 x i16> [[A2]])
+; CHECK-NEXT: [[TMP6:%.*]] = trunc i32 [[TMP5]] to i16
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <4 x i16> [[TMP1]], i16 0, i32 3
+; CHECK-NEXT: [[TMP7:%.*]] = insertelement <4 x i16> [[A1]], i16 [[TMP6]], i32 3
+; CHECK-NEXT: store <4 x i16> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <4 x i16> [[TMP7]]
+;
+entry:
+ %0 = tail call i32 @llvm.aarch64.neon.uaddv.i32.v4i16(<4 x i16> %a2)
+ %1 = trunc i32 %0 to i16
+ %2 = insertelement <4 x i16> %a1, i16 %1, i32 3
+ ret <4 x i16> %2
+}
+
+define i32 @test_vaddv_u16_masked(<4 x i16> %a1) #0 {
+; CHECK-LABEL: define i32 @test_vaddv_u16_masked(
+; CHECK-SAME: <4 x i16> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <4 x i16>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <4 x i16> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v4i16(<4 x i16> [[A1]])
+; CHECK-NEXT: [[TMP4:%.*]] = and i32 [[VADDV_I]], 0
+; CHECK-NEXT: [[TMP5:%.*]] = or i32 0, [[TMP4]]
+; CHECK-NEXT: [[TMP6:%.*]] = or i32 [[TMP5]], 0
+; CHECK-NEXT: [[TMP7:%.*]] = and i32 [[VADDV_I]], 3276799
+; CHECK-NEXT: store i32 [[TMP6]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i32 [[TMP7]]
+;
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.uaddv.i32.v4i16(<4 x i16> %a1)
+ %0 = and i32 %vaddv.i, 3276799 ; 0x31ffff
+ ret i32 %0
+}
+
+define i32 @test_vaddv_u32(<2 x i32> %a1) #0 {
+; CHECK-LABEL: define i32 @test_vaddv_u32(
+; CHECK-SAME: <2 x i32> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <2 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <2 x i32> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v2i32(<2 x i32> [[A1]])
+; CHECK-NEXT: store i32 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i32 [[VADDV_I]]
+;
+; 2 x i32 is not supported by the ISA, thus, this is a special case
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.uaddv.i32.v2i32(<2 x i32> %a1)
+ ret i32 %vaddv.i
+}
+
+define <2 x i32> @test_vaddv_u32_used_by_laneop(<2 x i32> %a1, <2 x i32> %a2) #0 {
+; CHECK-LABEL: define <2 x i32> @test_vaddv_u32_used_by_laneop(
+; CHECK-SAME: <2 x i32> [[A1:%.*]], <2 x i32> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <2 x i32>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 8) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <2 x i32> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v2i32(<2 x i32> [[A2]])
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <2 x i32> [[TMP1]], i32 0, i32 1
+; CHECK-NEXT: [[TMP6:%.*]] = insertelement <2 x i32> [[A1]], i32 [[TMP5]], i32 1
+; CHECK-NEXT: store <2 x i32> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <2 x i32> [[TMP6]]
+;
+entry:
+ %0 = tail call i32 @llvm.aarch64.neon.uaddv.i32.v2i32(<2 x i32> %a2)
+ %1 = insertelement <2 x i32> %a1, i32 %0, i32 1
+ ret <2 x i32> %1
+}
+
+define float @test_vaddv_f32(<2 x float> %a1) #0 {
+; CHECK-LABEL: define float @test_vaddv_f32(
+; CHECK-SAME: <2 x float> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <2 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <2 x i32> [[TMP0]] to i64
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call float @llvm.aarch64.neon.faddv.f32.v2f32(<2 x float> [[A1]])
+; CHECK-NEXT: store i32 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret float [[VADDV_I]]
+;
+entry:
+ %vaddv.i = tail call float @llvm.aarch64.neon.faddv.f32.v2f32(<2 x float> %a1)
+ ret float %vaddv.i
+}
+
+define float @test_vaddv_v4f32(<4 x float> %a1) #0 {
+; CHECK-LABEL: define float @test_vaddv_v4f32(
+; CHECK-SAME: <4 x float> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <4 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <4 x i32> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call float @llvm.aarch64.neon.faddv.f32.v4f32(<4 x float> [[A1]])
+; CHECK-NEXT: store i32 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret float [[VADDV_I]]
+;
+entry:
+ %vaddv.i = tail call float @llvm.aarch64.neon.faddv.f32.v4f32(<4 x float> %a1)
+ ret float %vaddv.i
+}
+
+define double @test_vaddv_f64(<2 x double> %a1) #0 {
+; CHECK-LABEL: define double @test_vaddv_f64(
+; CHECK-SAME: <2 x double> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <2 x i64>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <2 x i64> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call double @llvm.aarch64.neon.faddv.f64.v2f64(<2 x double> [[A1]])
+; CHECK-NEXT: store i64 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret double [[VADDV_I]]
+;
+entry:
+ %vaddv.i = tail call double @llvm.aarch64.neon.faddv.f64.v2f64(<2 x double> %a1)
+ ret double %vaddv.i
+}
+
+define i64 @test_vaddv_u64(<2 x i64> %a1) #0 {
+; CHECK-LABEL: define i64 @test_vaddv_u64(
+; CHECK-SAME: <2 x i64> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <2 x i64>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <2 x i64> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i64 @llvm.aarch64.neon.uaddv.i64.v2i64(<2 x i64> [[A1]])
+; CHECK-NEXT: store i64 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i64 [[VADDV_I]]
+;
+entry:
+ %vaddv.i = tail call i64 @llvm.aarch64.neon.uaddv.i64.v2i64(<2 x i64> %a1)
+ ret i64 %vaddv.i
+}
+
+define <2 x i64> @test_vaddv_u64_used_by_laneop(<2 x i64> %a1, <2 x i64> %a2) #0 {
+; CHECK-LABEL: define <2 x i64> @test_vaddv_u64_used_by_laneop(
+; CHECK-SAME: <2 x i64> [[A1:%.*]], <2 x i64> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <2 x i64>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 16) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <2 x i64> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i64 @llvm.aarch64.neon.uaddv.i64.v2i64(<2 x i64> [[A2]])
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <2 x i64> [[TMP1]], i64 0, i64 1
+; CHECK-NEXT: [[TMP6:%.*]] = insertelement <2 x i64> [[A1]], i64 [[TMP5]], i64 1
+; CHECK-NEXT: store <2 x i64> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <2 x i64> [[TMP6]]
+;
+entry:
+ %0 = tail call i64 @llvm.aarch64.neon.uaddv.i64.v2i64(<2 x i64> %a2)
+ %1 = insertelement <2 x i64> %a1, i64 %0, i64 1
+ ret <2 x i64> %1
+}
+
+define <1 x i64> @test_vaddv_u64_to_vec(<2 x i64> %a1) #0 {
+; CHECK-LABEL: define <1 x i64> @test_vaddv_u64_to_vec(
+; CHECK-SAME: <2 x i64> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <2 x i64>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <2 x i64> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i64 @llvm.aarch64.neon.uaddv.i64.v2i64(<2 x i64> [[A1]])
+; CHECK-NEXT: [[VEC:%.*]] = insertelement <1 x i64> undef, i64 [[VADDV_I]], i32 0
+; CHECK-NEXT: store <1 x i64> zeroinitializer, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <1 x i64> [[VEC]]
+;
+entry:
+ %vaddv.i = tail call i64 @llvm.aarch64.neon.uaddv.i64.v2i64(<2 x i64> %a1)
+ %vec = insertelement <1 x i64> undef, i64 %vaddv.i, i32 0
+ ret <1 x i64> %vec
+}
+
+define signext i8 @test_vaddvq_s8(<16 x i8> %a1) #0 {
+; CHECK-LABEL: define signext i8 @test_vaddvq_s8(
+; CHECK-SAME: <16 x i8> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <16 x i8>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <16 x i8> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.saddv.i32.v16i8(<16 x i8> [[A1]])
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i32 [[VADDV_I]] to i8
+; CHECK-NEXT: store i8 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i8 [[TMP4]]
+;
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.saddv.i32.v16i8(<16 x i8> %a1)
+ %0 = trunc i32 %vaddv.i to i8
+ ret i8 %0
+}
+
+define <16 x i8> @test_vaddvq_s8_used_by_laneop(<16 x i8> %a1, <16 x i8> %a2) #0 {
+; CHECK-LABEL: define <16 x i8> @test_vaddvq_s8_used_by_laneop(
+; CHECK-SAME: <16 x i8> [[A1:%.*]], <16 x i8> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <16 x i8>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 16) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i8>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <16 x i8> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @llvm.aarch64.neon.saddv.i32.v16i8(<16 x i8> [[A2]])
+; CHECK-NEXT: [[TMP6:%.*]] = trunc i32 [[TMP5]] to i8
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <16 x i8> [[TMP1]], i8 0, i32 3
+; CHECK-NEXT: [[TMP7:%.*]] = insertelement <16 x i8> [[A1]], i8 [[TMP6]], i32 3
+; CHECK-NEXT: store <16 x i8> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <16 x i8> [[TMP7]]
+;
+entry:
+ %0 = tail call i32 @llvm.aarch64.neon.saddv.i32.v16i8(<16 x i8> %a2)
+ %1 = trunc i32 %0 to i8
+ %2 = insertelement <16 x i8> %a1, i8 %1, i32 3
+ ret <16 x i8> %2
+}
+
+define signext i16 @test_vaddvq_s16(<8 x i16> %a1) #0 {
+; CHECK-LABEL: define signext i16 @test_vaddvq_s16(
+; CHECK-SAME: <8 x i16> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <8 x i16>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <8 x i16> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.saddv.i32.v8i16(<8 x i16> [[A1]])
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i32 [[VADDV_I]] to i16
+; CHECK-NEXT: store i16 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i16 [[TMP4]]
+;
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.saddv.i32.v8i16(<8 x i16> %a1)
+ %0 = trunc i32 %vaddv.i to i16
+ ret i16 %0
+}
+
+define <8 x i16> @test_vaddvq_s16_used_by_laneop(<8 x i16> %a1, <8 x i16> %a2) #0 {
+; CHECK-LABEL: define <8 x i16> @test_vaddvq_s16_used_by_laneop(
+; CHECK-SAME: <8 x i16> [[A1:%.*]], <8 x i16> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <8 x i16>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 16) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <8 x i16> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @llvm.aarch64.neon.saddv.i32.v8i16(<8 x i16> [[A2]])
+; CHECK-NEXT: [[TMP6:%.*]] = trunc i32 [[TMP5]] to i16
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <8 x i16> [[TMP1]], i16 0, i32 3
+; CHECK-NEXT: [[TMP7:%.*]] = insertelement <8 x i16> [[A1]], i16 [[TMP6]], i32 3
+; CHECK-NEXT: store <8 x i16> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <8 x i16> [[TMP7]]
+;
+entry:
+ %0 = tail call i32 @llvm.aarch64.neon.saddv.i32.v8i16(<8 x i16> %a2)
+ %1 = trunc i32 %0 to i16
+ %2 = insertelement <8 x i16> %a1, i16 %1, i32 3
+ ret <8 x i16> %2
+}
+
+define i32 @test_vaddvq_s32(<4 x i32> %a1) #0 {
+; CHECK-LABEL: define i32 @test_vaddvq_s32(
+; CHECK-SAME: <4 x i32> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <4 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <4 x i32> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.saddv.i32.v4i32(<4 x i32> [[A1]])
+; CHECK-NEXT: store i32 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i32 [[VADDV_I]]
+;
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.saddv.i32.v4i32(<4 x i32> %a1)
+ ret i32 %vaddv.i
+}
+
+define <4 x i32> @test_vaddvq_s32_used_by_laneop(<4 x i32> %a1, <4 x i32> %a2) #0 {
+; CHECK-LABEL: define <4 x i32> @test_vaddvq_s32_used_by_laneop(
+; CHECK-SAME: <4 x i32> [[A1:%.*]], <4 x i32> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <4 x i32>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 16) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i32> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @llvm.aarch64.neon.saddv.i32.v4i32(<4 x i32> [[A2]])
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <4 x i32> [[TMP1]], i32 0, i32 3
+; CHECK-NEXT: [[TMP6:%.*]] = insertelement <4 x i32> [[A1]], i32 [[TMP5]], i32 3
+; CHECK-NEXT: store <4 x i32> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <4 x i32> [[TMP6]]
+;
+entry:
+ %0 = tail call i32 @llvm.aarch64.neon.saddv.i32.v4i32(<4 x i32> %a2)
+ %1 = insertelement <4 x i32> %a1, i32 %0, i32 3
+ ret <4 x i32> %1
+}
+
+define zeroext i8 @test_vaddvq_u8(<16 x i8> %a1) #0 {
+; CHECK-LABEL: define zeroext i8 @test_vaddvq_u8(
+; CHECK-SAME: <16 x i8> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <16 x i8>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <16 x i8> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v16i8(<16 x i8> [[A1]])
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i32 [[VADDV_I]] to i8
+; CHECK-NEXT: store i8 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i8 [[TMP4]]
+;
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.uaddv.i32.v16i8(<16 x i8> %a1)
+ %0 = trunc i32 %vaddv.i to i8
+ ret i8 %0
+}
+
+define <16 x i8> @test_vaddvq_u8_used_by_laneop(<16 x i8> %a1, <16 x i8> %a2) #0 {
+; CHECK-LABEL: define <16 x i8> @test_vaddvq_u8_used_by_laneop(
+; CHECK-SAME: <16 x i8> [[A1:%.*]], <16 x i8> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <16 x i8>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 16) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i8>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <16 x i8> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v16i8(<16 x i8> [[A2]])
+; CHECK-NEXT: [[TMP6:%.*]] = trunc i32 [[TMP5]] to i8
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <16 x i8> [[TMP1]], i8 0, i32 3
+; CHECK-NEXT: [[TMP7:%.*]] = insertelement <16 x i8> [[A1]], i8 [[TMP6]], i32 3
+; CHECK-NEXT: store <16 x i8> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <16 x i8> [[TMP7]]
+;
+entry:
+ %0 = tail call i32 @llvm.aarch64.neon.uaddv.i32.v16i8(<16 x i8> %a2)
+ %1 = trunc i32 %0 to i8
+ %2 = insertelement <16 x i8> %a1, i8 %1, i32 3
+ ret <16 x i8> %2
+}
+
+define zeroext i16 @test_vaddvq_u16(<8 x i16> %a1) #0 {
+; CHECK-LABEL: define zeroext i16 @test_vaddvq_u16(
+; CHECK-SAME: <8 x i16> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <8 x i16>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <8 x i16> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v8i16(<8 x i16> [[A1]])
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i32 [[VADDV_I]] to i16
+; CHECK-NEXT: store i16 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i16 [[TMP4]]
+;
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.uaddv.i32.v8i16(<8 x i16> %a1)
+ %0 = trunc i32 %vaddv.i to i16
+ ret i16 %0
+}
+
+define <8 x i16> @test_vaddvq_u16_used_by_laneop(<8 x i16> %a1, <8 x i16> %a2) #0 {
+; CHECK-LABEL: define <8 x i16> @test_vaddvq_u16_used_by_laneop(
+; CHECK-SAME: <8 x i16> [[A1:%.*]], <8 x i16> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <8 x i16>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 16) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <8 x i16> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v8i16(<8 x i16> [[A2]])
+; CHECK-NEXT: [[TMP6:%.*]] = trunc i32 [[TMP5]] to i16
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <8 x i16> [[TMP1]], i16 0, i32 3
+; CHECK-NEXT: [[TMP7:%.*]] = insertelement <8 x i16> [[A1]], i16 [[TMP6]], i32 3
+; CHECK-NEXT: store <8 x i16> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <8 x i16> [[TMP7]]
+;
+entry:
+ %0 = tail call i32 @llvm.aarch64.neon.uaddv.i32.v8i16(<8 x i16> %a2)
+ %1 = trunc i32 %0 to i16
+ %2 = insertelement <8 x i16> %a1, i16 %1, i32 3
+ ret <8 x i16> %2
+}
+
+define i32 @test_vaddvq_u32(<4 x i32> %a1) #0 {
+; CHECK-LABEL: define i32 @test_vaddvq_u32(
+; CHECK-SAME: <4 x i32> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <4 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <4 x i32> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v4i32(<4 x i32> [[A1]])
+; CHECK-NEXT: store i32 0, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret i32 [[VADDV_I]]
+;
+entry:
+ %vaddv.i = tail call i32 @llvm.aarch64.neon.uaddv.i32.v4i32(<4 x i32> %a1)
+ ret i32 %vaddv.i
+}
+
+define <4 x i32> @test_vaddvq_u32_used_by_laneop(<4 x i32> %a1, <4 x i32> %a2) #0 {
+; CHECK-LABEL: define <4 x i32> @test_vaddvq_u32_used_by_laneop(
+; CHECK-SAME: <4 x i32> [[A1:%.*]], <4 x i32> [[A2:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = load <4 x i32>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 16) to ptr), align 8
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i32> [[TMP0]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @llvm.aarch64.neon.uaddv.i32.v4i32(<4 x i32> [[A2]])
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <4 x i32> [[TMP1]], i32 0, i32 3
+; CHECK-NEXT: [[TMP6:%.*]] = insertelement <4 x i32> [[A1]], i32 [[TMP5]], i32 3
+; CHECK-NEXT: store <4 x i32> [[_MSPROP]], ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: ret <4 x i32> [[TMP6]]
+;
+entry:
+ %0 = tail call i32 @llvm.aarch64.neon.uaddv.i32.v4i32(<4 x i32> %a2)
+ %1 = insertelement <4 x i32> %a1, i32 %0, i32 3
+ ret <4 x i32> %1
+}
+
+declare i32 @llvm.aarch64.neon.uaddv.i32.v4i32(<4 x i32>)
+
+declare i32 @llvm.aarch64.neon.uaddv.i32.v8i16(<8 x i16>)
+
+declare i32 @llvm.aarch64.neon.uaddv.i32.v16i8(<16 x i8>)
+
+declare i32 @llvm.aarch64.neon.saddv.i32.v4i32(<4 x i32>)
+
+declare i32 @llvm.aarch64.neon.saddv.i32.v8i16(<8 x i16>)
+
+declare i32 @llvm.aarch64.neon.saddv.i32.v16i8(<16 x i8>)
+
+declare i64 @llvm.aarch64.neon.uaddv.i64.v2i64(<2 x i64>)
+
+declare i32 @llvm.aarch64.neon.uaddv.i32.v2i32(<2 x i32>)
+
+declare i32 @llvm.aarch64.neon.uaddv.i32.v4i16(<4 x i16>)
+
+declare i32 @llvm.aarch64.neon.uaddv.i32.v8i8(<8 x i8>)
+
+declare i32 @llvm.aarch64.neon.saddv.i32.v2i32(<2 x i32>)
+
+declare i64 @llvm.aarch64.neon.saddv.i64.v2i64(<2 x i64>)
+
+declare i32 @llvm.aarch64.neon.saddv.i32.v4i16(<4 x i16>)
+
+declare i32 @llvm.aarch64.neon.saddv.i32.v8i8(<8 x i8>)
+
+declare float @llvm.aarch64.neon.faddv.f32.v2f32(<2 x float> %a1)
+declare float @llvm.aarch64.neon.faddv.f32.v4f32(<4 x float> %a1)
+declare double @llvm.aarch64.neon.faddv.f64.v2f64(<2 x double> %a1)
+
+attributes #0 = { sanitize_memory }
+;.
+; CHECK: [[PROF1]] = !{!"branch_weights", i32 1, i32 1048575}
+;.
>From d62f2ff3f203e3aad46aa832ffc1af16c8cafa3d Mon Sep 17 00:00:00 2001
From: Thurston Dang <thurston at google.com>
Date: Fri, 31 Jan 2025 19:14:06 +0000
Subject: [PATCH 2/2] Replace undef with parameter
---
.../MemorySanitizer/AArch64/arm64-vaddv.ll | 18 ++++++++++--------
1 file changed, 10 insertions(+), 8 deletions(-)
diff --git a/llvm/test/Instrumentation/MemorySanitizer/AArch64/arm64-vaddv.ll b/llvm/test/Instrumentation/MemorySanitizer/AArch64/arm64-vaddv.ll
index d8abb65894c7f2..f0e607db9d281e 100644
--- a/llvm/test/Instrumentation/MemorySanitizer/AArch64/arm64-vaddv.ll
+++ b/llvm/test/Instrumentation/MemorySanitizer/AArch64/arm64-vaddv.ll
@@ -533,27 +533,29 @@ entry:
ret <2 x i64> %1
}
-define <1 x i64> @test_vaddv_u64_to_vec(<2 x i64> %a1) #0 {
+define <1 x i64> @test_vaddv_u64_to_vec(<2 x i64> %a1, <1 x i64> %param1) #0 {
; CHECK-LABEL: define <1 x i64> @test_vaddv_u64_to_vec(
-; CHECK-SAME: <2 x i64> [[A1:%.*]]) #[[ATTR0]] {
+; CHECK-SAME: <2 x i64> [[A1:%.*]], <1 x i64> [[PARAM1:%.*]]) #[[ATTR0]] {
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = load <2 x i64>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <1 x i64>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 16) to ptr), align 8
; CHECK-NEXT: call void @llvm.donothing()
; CHECK-NEXT: [[TMP1:%.*]] = bitcast <2 x i64> [[TMP0]] to i128
; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP1]], 0
-; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF1]]
-; CHECK: 2:
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF1]]
+; CHECK: 3:
; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR3]]
; CHECK-NEXT: unreachable
-; CHECK: 3:
+; CHECK: 4:
; CHECK-NEXT: [[VADDV_I:%.*]] = tail call i64 @llvm.aarch64.neon.uaddv.i64.v2i64(<2 x i64> [[A1]])
-; CHECK-NEXT: [[VEC:%.*]] = insertelement <1 x i64> undef, i64 [[VADDV_I]], i32 0
-; CHECK-NEXT: store <1 x i64> zeroinitializer, ptr @__msan_retval_tls, align 8
+; CHECK-NEXT: [[_MSPROP:%.*]] = insertelement <1 x i64> [[TMP2]], i64 0, i32 0
+; CHECK-NEXT: [[VEC:%.*]] = insertelement <1 x i64> [[PARAM1]], i64 [[VADDV_I]], i32 0
+; CHECK-NEXT: store <1 x i64> [[_MSPROP]], ptr @__msan_retval_tls, align 8
; CHECK-NEXT: ret <1 x i64> [[VEC]]
;
entry:
%vaddv.i = tail call i64 @llvm.aarch64.neon.uaddv.i64.v2i64(<2 x i64> %a1)
- %vec = insertelement <1 x i64> undef, i64 %vaddv.i, i32 0
+ %vec = insertelement <1 x i64> %param1, i64 %vaddv.i, i32 0
ret <1 x i64> %vec
}
More information about the llvm-commits
mailing list