[llvm] 009d56d - [test][msa] Add more sse,avx intrinsics tests
Vitaly Buka via llvm-commits
llvm-commits at lists.llvm.org
Sat May 7 20:16:30 PDT 2022
Author: Vitaly Buka
Date: 2022-05-07T20:16:22-07:00
New Revision: 009d56da5c4ea3666c4753ce7564c8c20d7e0255
URL: https://github.com/llvm/llvm-project/commit/009d56da5c4ea3666c4753ce7564c8c20d7e0255
DIFF: https://github.com/llvm/llvm-project/commit/009d56da5c4ea3666c4753ce7564c8c20d7e0255.diff
LOG: [test][msa] Add more sse,avx intrinsics tests
Added:
llvm/test/Instrumentation/MemorySanitizer/avx-intrinsics-x86.ll
llvm/test/Instrumentation/MemorySanitizer/avx2-intrinsics-x86.ll
llvm/test/Instrumentation/MemorySanitizer/sse-intrinsics-x86.ll
llvm/test/Instrumentation/MemorySanitizer/sse2-intrinsics-x86.ll
llvm/test/Instrumentation/MemorySanitizer/sse41-intrinsics-x86.ll
Modified:
Removed:
################################################################################
diff --git a/llvm/test/Instrumentation/MemorySanitizer/avx-intrinsics-x86.ll b/llvm/test/Instrumentation/MemorySanitizer/avx-intrinsics-x86.ll
new file mode 100644
index 0000000000000..70f4d6f237a41
--- /dev/null
+++ b/llvm/test/Instrumentation/MemorySanitizer/avx-intrinsics-x86.ll
@@ -0,0 +1,1818 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt %s -S -passes=msan 2>&1 | FileCheck %s
+
+target datalayout = "e-m:o-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+define <4 x double> @test_x86_avx_addsub_pd_256(<4 x double> %a0, <4 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_addsub_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.addsub.pd.256(<4 x double> [[A0:%.*]], <4 x double> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i64> [[_MSPROP]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx.addsub.pd.256(<4 x double> %a0, <4 x double> %a1) ; <<4 x double>> [#uses=1]
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.addsub.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+
+define <8 x float> @test_x86_avx_addsub_ps_256(<8 x float> %a0, <8 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_addsub_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.addsub.ps.256(<8 x float> [[A0:%.*]], <8 x float> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> [[_MSPROP]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.addsub.ps.256(<8 x float> %a0, <8 x float> %a1) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.addsub.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+
+define <4 x double> @test_x86_avx_blendv_pd_256(<4 x double> %a0, <4 x double> %a1, <4 x double> %a2) #0 {
+; CHECK-LABEL: @test_x86_avx_blendv_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 64) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[_MSPROP1:%.*]] = or <4 x i64> [[_MSPROP]], [[TMP3]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.blendv.pd.256(<4 x double> [[A0:%.*]], <4 x double> [[A1:%.*]], <4 x double> [[A2:%.*]])
+; CHECK-NEXT: store <4 x i64> [[_MSPROP1]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx.blendv.pd.256(<4 x double> %a0, <4 x double> %a1, <4 x double> %a2) ; <<4 x double>> [#uses=1]
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.blendv.pd.256(<4 x double>, <4 x double>, <4 x double>) nounwind readnone
+
+
+define <8 x float> @test_x86_avx_blendv_ps_256(<8 x float> %a0, <8 x float> %a1, <8 x float> %a2) #0 {
+; CHECK-LABEL: @test_x86_avx_blendv_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 64) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[_MSPROP1:%.*]] = or <8 x i32> [[_MSPROP]], [[TMP3]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.blendv.ps.256(<8 x float> [[A0:%.*]], <8 x float> [[A1:%.*]], <8 x float> [[A2:%.*]])
+; CHECK-NEXT: store <8 x i32> [[_MSPROP1]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.blendv.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float> %a2) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.blendv.ps.256(<8 x float>, <8 x float>, <8 x float>) nounwind readnone
+
+
+define <4 x double> @test_x86_avx_cmp_pd_256(<4 x double> %a0, <4 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_cmp_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0:![0-9]+]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8:[0-9]+]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i64> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.cmp.pd.256(<4 x double> [[A0:%.*]], <4 x double> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <4 x i64> zeroinitializer, <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx.cmp.pd.256(<4 x double> %a0, <4 x double> %a1, i8 7) ; <<4 x double>> [#uses=1]
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.cmp.pd.256(<4 x double>, <4 x double>, i8) nounwind readnone
+
+
+define <8 x float> @test_x86_avx_cmp_ps_256(<8 x float> %a0, <8 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_cmp_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0:%.*]], <8 x float> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a1, i8 7) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+
+define <8 x float> @test_x86_avx_cmp_ps_256_pseudo_op(<8 x float> %a0, <8 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_cmp_ps_256_pseudo_op(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[A2:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0:%.*]], <8 x float> [[A1:%.*]], i8 0)
+; CHECK-NEXT: [[TMP9:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP9]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP10:%.*]], label [[TMP11:%.*]], !prof [[PROF0]]
+; CHECK: 10:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 11:
+; CHECK-NEXT: [[A3:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A2]], i8 1)
+; CHECK-NEXT: [[TMP12:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i256 [[TMP12]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP13:%.*]], label [[TMP14:%.*]], !prof [[PROF0]]
+; CHECK: 13:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 14:
+; CHECK-NEXT: [[A4:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A3]], i8 2)
+; CHECK-NEXT: [[TMP15:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP4:%.*]] = icmp ne i256 [[TMP15]], 0
+; CHECK-NEXT: br i1 [[_MSCMP4]], label [[TMP16:%.*]], label [[TMP17:%.*]], !prof [[PROF0]]
+; CHECK: 16:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 17:
+; CHECK-NEXT: [[A5:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A4]], i8 3)
+; CHECK-NEXT: [[TMP18:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP5:%.*]] = icmp ne i256 [[TMP18]], 0
+; CHECK-NEXT: br i1 [[_MSCMP5]], label [[TMP19:%.*]], label [[TMP20:%.*]], !prof [[PROF0]]
+; CHECK: 19:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 20:
+; CHECK-NEXT: [[A6:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A5]], i8 4)
+; CHECK-NEXT: [[TMP21:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP6:%.*]] = icmp ne i256 [[TMP21]], 0
+; CHECK-NEXT: br i1 [[_MSCMP6]], label [[TMP22:%.*]], label [[TMP23:%.*]], !prof [[PROF0]]
+; CHECK: 22:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 23:
+; CHECK-NEXT: [[A7:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A6]], i8 5)
+; CHECK-NEXT: [[TMP24:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP7:%.*]] = icmp ne i256 [[TMP24]], 0
+; CHECK-NEXT: br i1 [[_MSCMP7]], label [[TMP25:%.*]], label [[TMP26:%.*]], !prof [[PROF0]]
+; CHECK: 25:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 26:
+; CHECK-NEXT: [[A8:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A7]], i8 6)
+; CHECK-NEXT: [[TMP27:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP8:%.*]] = icmp ne i256 [[TMP27]], 0
+; CHECK-NEXT: br i1 [[_MSCMP8]], label [[TMP28:%.*]], label [[TMP29:%.*]], !prof [[PROF0]]
+; CHECK: 28:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 29:
+; CHECK-NEXT: [[A9:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A8]], i8 7)
+; CHECK-NEXT: [[TMP30:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP9:%.*]] = icmp ne i256 [[TMP30]], 0
+; CHECK-NEXT: br i1 [[_MSCMP9]], label [[TMP31:%.*]], label [[TMP32:%.*]], !prof [[PROF0]]
+; CHECK: 31:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 32:
+; CHECK-NEXT: [[A10:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A9]], i8 8)
+; CHECK-NEXT: [[TMP33:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP10:%.*]] = icmp ne i256 [[TMP33]], 0
+; CHECK-NEXT: br i1 [[_MSCMP10]], label [[TMP34:%.*]], label [[TMP35:%.*]], !prof [[PROF0]]
+; CHECK: 34:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 35:
+; CHECK-NEXT: [[A11:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A10]], i8 9)
+; CHECK-NEXT: [[TMP36:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP11:%.*]] = icmp ne i256 [[TMP36]], 0
+; CHECK-NEXT: br i1 [[_MSCMP11]], label [[TMP37:%.*]], label [[TMP38:%.*]], !prof [[PROF0]]
+; CHECK: 37:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 38:
+; CHECK-NEXT: [[A12:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A11]], i8 10)
+; CHECK-NEXT: [[TMP39:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP12:%.*]] = icmp ne i256 [[TMP39]], 0
+; CHECK-NEXT: br i1 [[_MSCMP12]], label [[TMP40:%.*]], label [[TMP41:%.*]], !prof [[PROF0]]
+; CHECK: 40:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 41:
+; CHECK-NEXT: [[A13:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A12]], i8 11)
+; CHECK-NEXT: [[TMP42:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP13:%.*]] = icmp ne i256 [[TMP42]], 0
+; CHECK-NEXT: br i1 [[_MSCMP13]], label [[TMP43:%.*]], label [[TMP44:%.*]], !prof [[PROF0]]
+; CHECK: 43:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 44:
+; CHECK-NEXT: [[A14:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A13]], i8 12)
+; CHECK-NEXT: [[TMP45:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP14:%.*]] = icmp ne i256 [[TMP45]], 0
+; CHECK-NEXT: br i1 [[_MSCMP14]], label [[TMP46:%.*]], label [[TMP47:%.*]], !prof [[PROF0]]
+; CHECK: 46:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 47:
+; CHECK-NEXT: [[A15:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A14]], i8 13)
+; CHECK-NEXT: [[TMP48:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP15:%.*]] = icmp ne i256 [[TMP48]], 0
+; CHECK-NEXT: br i1 [[_MSCMP15]], label [[TMP49:%.*]], label [[TMP50:%.*]], !prof [[PROF0]]
+; CHECK: 49:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 50:
+; CHECK-NEXT: [[A16:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A15]], i8 14)
+; CHECK-NEXT: [[TMP51:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP16:%.*]] = icmp ne i256 [[TMP51]], 0
+; CHECK-NEXT: br i1 [[_MSCMP16]], label [[TMP52:%.*]], label [[TMP53:%.*]], !prof [[PROF0]]
+; CHECK: 52:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 53:
+; CHECK-NEXT: [[A17:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A16]], i8 15)
+; CHECK-NEXT: [[TMP54:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP17:%.*]] = icmp ne i256 [[TMP54]], 0
+; CHECK-NEXT: br i1 [[_MSCMP17]], label [[TMP55:%.*]], label [[TMP56:%.*]], !prof [[PROF0]]
+; CHECK: 55:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 56:
+; CHECK-NEXT: [[A18:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A17]], i8 16)
+; CHECK-NEXT: [[TMP57:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP18:%.*]] = icmp ne i256 [[TMP57]], 0
+; CHECK-NEXT: br i1 [[_MSCMP18]], label [[TMP58:%.*]], label [[TMP59:%.*]], !prof [[PROF0]]
+; CHECK: 58:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 59:
+; CHECK-NEXT: [[A19:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A18]], i8 17)
+; CHECK-NEXT: [[TMP60:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP19:%.*]] = icmp ne i256 [[TMP60]], 0
+; CHECK-NEXT: br i1 [[_MSCMP19]], label [[TMP61:%.*]], label [[TMP62:%.*]], !prof [[PROF0]]
+; CHECK: 61:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 62:
+; CHECK-NEXT: [[A20:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A19]], i8 18)
+; CHECK-NEXT: [[TMP63:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP20:%.*]] = icmp ne i256 [[TMP63]], 0
+; CHECK-NEXT: br i1 [[_MSCMP20]], label [[TMP64:%.*]], label [[TMP65:%.*]], !prof [[PROF0]]
+; CHECK: 64:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 65:
+; CHECK-NEXT: [[A21:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A20]], i8 19)
+; CHECK-NEXT: [[TMP66:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP21:%.*]] = icmp ne i256 [[TMP66]], 0
+; CHECK-NEXT: br i1 [[_MSCMP21]], label [[TMP67:%.*]], label [[TMP68:%.*]], !prof [[PROF0]]
+; CHECK: 67:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 68:
+; CHECK-NEXT: [[A22:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A21]], i8 20)
+; CHECK-NEXT: [[TMP69:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP22:%.*]] = icmp ne i256 [[TMP69]], 0
+; CHECK-NEXT: br i1 [[_MSCMP22]], label [[TMP70:%.*]], label [[TMP71:%.*]], !prof [[PROF0]]
+; CHECK: 70:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 71:
+; CHECK-NEXT: [[A23:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A22]], i8 21)
+; CHECK-NEXT: [[TMP72:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP23:%.*]] = icmp ne i256 [[TMP72]], 0
+; CHECK-NEXT: br i1 [[_MSCMP23]], label [[TMP73:%.*]], label [[TMP74:%.*]], !prof [[PROF0]]
+; CHECK: 73:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 74:
+; CHECK-NEXT: [[A24:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A23]], i8 22)
+; CHECK-NEXT: [[TMP75:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP24:%.*]] = icmp ne i256 [[TMP75]], 0
+; CHECK-NEXT: br i1 [[_MSCMP24]], label [[TMP76:%.*]], label [[TMP77:%.*]], !prof [[PROF0]]
+; CHECK: 76:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 77:
+; CHECK-NEXT: [[A25:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A24]], i8 23)
+; CHECK-NEXT: [[TMP78:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP25:%.*]] = icmp ne i256 [[TMP78]], 0
+; CHECK-NEXT: br i1 [[_MSCMP25]], label [[TMP79:%.*]], label [[TMP80:%.*]], !prof [[PROF0]]
+; CHECK: 79:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 80:
+; CHECK-NEXT: [[A26:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A25]], i8 24)
+; CHECK-NEXT: [[TMP81:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP26:%.*]] = icmp ne i256 [[TMP81]], 0
+; CHECK-NEXT: br i1 [[_MSCMP26]], label [[TMP82:%.*]], label [[TMP83:%.*]], !prof [[PROF0]]
+; CHECK: 82:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 83:
+; CHECK-NEXT: [[A27:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A26]], i8 25)
+; CHECK-NEXT: [[TMP84:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP27:%.*]] = icmp ne i256 [[TMP84]], 0
+; CHECK-NEXT: br i1 [[_MSCMP27]], label [[TMP85:%.*]], label [[TMP86:%.*]], !prof [[PROF0]]
+; CHECK: 85:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 86:
+; CHECK-NEXT: [[A28:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A27]], i8 26)
+; CHECK-NEXT: [[TMP87:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP28:%.*]] = icmp ne i256 [[TMP87]], 0
+; CHECK-NEXT: br i1 [[_MSCMP28]], label [[TMP88:%.*]], label [[TMP89:%.*]], !prof [[PROF0]]
+; CHECK: 88:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 89:
+; CHECK-NEXT: [[A29:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A28]], i8 27)
+; CHECK-NEXT: [[TMP90:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP29:%.*]] = icmp ne i256 [[TMP90]], 0
+; CHECK-NEXT: br i1 [[_MSCMP29]], label [[TMP91:%.*]], label [[TMP92:%.*]], !prof [[PROF0]]
+; CHECK: 91:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 92:
+; CHECK-NEXT: [[A30:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A29]], i8 28)
+; CHECK-NEXT: [[TMP93:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP30:%.*]] = icmp ne i256 [[TMP93]], 0
+; CHECK-NEXT: br i1 [[_MSCMP30]], label [[TMP94:%.*]], label [[TMP95:%.*]], !prof [[PROF0]]
+; CHECK: 94:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 95:
+; CHECK-NEXT: [[A31:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A30]], i8 29)
+; CHECK-NEXT: [[TMP96:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP31:%.*]] = icmp ne i256 [[TMP96]], 0
+; CHECK-NEXT: br i1 [[_MSCMP31]], label [[TMP97:%.*]], label [[TMP98:%.*]], !prof [[PROF0]]
+; CHECK: 97:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 98:
+; CHECK-NEXT: [[A32:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A31]], i8 30)
+; CHECK-NEXT: [[TMP99:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP32:%.*]] = icmp ne i256 [[TMP99]], 0
+; CHECK-NEXT: br i1 [[_MSCMP32]], label [[TMP100:%.*]], label [[TMP101:%.*]], !prof [[PROF0]]
+; CHECK: 100:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 101:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> [[A0]], <8 x float> [[A32]], i8 31)
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %a2 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a1, i8 0) ; <<8 x float>> [#uses=1]
+ %a3 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a2, i8 1) ; <<8 x float>> [#uses=1]
+ %a4 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a3, i8 2) ; <<8 x float>> [#uses=1]
+ %a5 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a4, i8 3) ; <<8 x float>> [#uses=1]
+ %a6 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a5, i8 4) ; <<8 x float>> [#uses=1]
+ %a7 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a6, i8 5) ; <<8 x float>> [#uses=1]
+ %a8 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a7, i8 6) ; <<8 x float>> [#uses=1]
+ %a9 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a8, i8 7) ; <<8 x float>> [#uses=1]
+ %a10 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a9, i8 8) ; <<8 x float>> [#uses=1]
+ %a11 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a10, i8 9) ; <<8 x float>> [#uses=1]
+ %a12 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a11, i8 10) ; <<8 x float>> [#uses=1]
+ %a13 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a12, i8 11) ; <<8 x float>> [#uses=1]
+ %a14 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a13, i8 12) ; <<8 x float>> [#uses=1]
+ %a15 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a14, i8 13) ; <<8 x float>> [#uses=1]
+ %a16 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a15, i8 14) ; <<8 x float>> [#uses=1]
+ %a17 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a16, i8 15) ; <<8 x float>> [#uses=1]
+ %a18 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a17, i8 16) ; <<8 x float>> [#uses=1]
+ %a19 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a18, i8 17) ; <<8 x float>> [#uses=1]
+ %a20 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a19, i8 18) ; <<8 x float>> [#uses=1]
+ %a21 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a20, i8 19) ; <<8 x float>> [#uses=1]
+ %a22 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a21, i8 20) ; <<8 x float>> [#uses=1]
+ %a23 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a22, i8 21) ; <<8 x float>> [#uses=1]
+ %a24 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a23, i8 22) ; <<8 x float>> [#uses=1]
+ %a25 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a24, i8 23) ; <<8 x float>> [#uses=1]
+ %a26 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a25, i8 24) ; <<8 x float>> [#uses=1]
+ %a27 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a26, i8 25) ; <<8 x float>> [#uses=1]
+ %a28 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a27, i8 26) ; <<8 x float>> [#uses=1]
+ %a29 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a28, i8 27) ; <<8 x float>> [#uses=1]
+ %a30 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a29, i8 28) ; <<8 x float>> [#uses=1]
+ %a31 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a30, i8 29) ; <<8 x float>> [#uses=1]
+ %a32 = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a31, i8 30) ; <<8 x float>> [#uses=1]
+ %res = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a32, i8 31) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float>, <8 x float>, i8) nounwind readnone
+
+
+define <4 x float> @test_x86_avx_cvt_pd2_ps_256(<4 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_cvt_pd2_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.avx.cvt.pd2.ps.256(<4 x double> [[A0:%.*]])
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.avx.cvt.pd2.ps.256(<4 x double> %a0) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx.cvt.pd2.ps.256(<4 x double>) nounwind readnone
+
+
+define <4 x i32> @test_x86_avx_cvt_pd2dq_256(<4 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_cvt_pd2dq_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.avx.cvt.pd2dq.256(<4 x double> [[A0:%.*]])
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.avx.cvt.pd2dq.256(<4 x double> %a0) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.avx.cvt.pd2dq.256(<4 x double>) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx_cvt_ps2dq_256(<8 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_cvt_ps2dq_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx.cvt.ps2dq.256(<8 x float> [[A0:%.*]])
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx.cvt.ps2dq.256(<8 x float> %a0) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx.cvt.ps2dq.256(<8 x float>) nounwind readnone
+
+
+define <4 x i32> @test_x86_avx_cvtt_pd2dq_256(<4 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_cvtt_pd2dq_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.avx.cvtt.pd2dq.256(<4 x double> [[A0:%.*]])
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.avx.cvtt.pd2dq.256(<4 x double> %a0) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.avx.cvtt.pd2dq.256(<4 x double>) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx_cvtt_ps2dq_256(<8 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_cvtt_ps2dq_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx.cvtt.ps2dq.256(<8 x float> [[A0:%.*]])
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx.cvtt.ps2dq.256(<8 x float> %a0) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx.cvtt.ps2dq.256(<8 x float>) nounwind readnone
+
+
+define <8 x float> @test_x86_avx_dp_ps_256(<8 x float> %a0, <8 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_dp_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.dp.ps.256(<8 x float> [[A0:%.*]], <8 x float> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.dp.ps.256(<8 x float> %a0, <8 x float> %a1, i8 7) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.dp.ps.256(<8 x float>, <8 x float>, i8) nounwind readnone
+
+
+define <4 x double> @test_x86_avx_hadd_pd_256(<4 x double> %a0, <4 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_hadd_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.hadd.pd.256(<4 x double> [[A0:%.*]], <4 x double> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i64> [[_MSPROP]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx.hadd.pd.256(<4 x double> %a0, <4 x double> %a1) ; <<4 x double>> [#uses=1]
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.hadd.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+
+define <8 x float> @test_x86_avx_hadd_ps_256(<8 x float> %a0, <8 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_hadd_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.hadd.ps.256(<8 x float> [[A0:%.*]], <8 x float> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> [[_MSPROP]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.hadd.ps.256(<8 x float> %a0, <8 x float> %a1) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.hadd.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+
+define <4 x double> @test_x86_avx_hsub_pd_256(<4 x double> %a0, <4 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_hsub_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.hsub.pd.256(<4 x double> [[A0:%.*]], <4 x double> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i64> [[_MSPROP]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx.hsub.pd.256(<4 x double> %a0, <4 x double> %a1) ; <<4 x double>> [#uses=1]
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.hsub.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+
+define <8 x float> @test_x86_avx_hsub_ps_256(<8 x float> %a0, <8 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_hsub_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.hsub.ps.256(<8 x float> [[A0:%.*]], <8 x float> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> [[_MSPROP]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.hsub.ps.256(<8 x float> %a0, <8 x float> %a1) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.hsub.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+
+define <32 x i8> @test_x86_avx_ldu_dq_256(i8* %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_ldu_dq_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = ptrtoint i8* [[A0:%.*]] to i64
+; CHECK-NEXT: [[TMP3:%.*]] = xor i64 [[TMP2]], 87960930222080
+; CHECK-NEXT: [[TMP4:%.*]] = inttoptr i64 [[TMP3]] to <32 x i8>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <32 x i8>, <32 x i8>* [[TMP4]], align 1
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP5:%.*]], label [[TMP6:%.*]], !prof [[PROF0]]
+; CHECK: 5:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 6:
+; CHECK-NEXT: [[RES:%.*]] = call <32 x i8> @llvm.x86.avx.ldu.dq.256(i8* [[A0]])
+; CHECK-NEXT: store <32 x i8> [[_MSLD]], <32 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <32 x i8>*), align 8
+; CHECK-NEXT: ret <32 x i8> [[RES]]
+;
+ %res = call <32 x i8> @llvm.x86.avx.ldu.dq.256(i8* %a0) ; <<32 x i8>> [#uses=1]
+ ret <32 x i8> %res
+}
+declare <32 x i8> @llvm.x86.avx.ldu.dq.256(i8*) nounwind readonly
+
+
+define <2 x double> @test_x86_avx_maskload_pd(i8* %a0, <2 x i64> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx_maskload_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.avx.maskload.pd(i8* [[A0:%.*]], <2 x i64> [[MASK:%.*]])
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.avx.maskload.pd(i8* %a0, <2 x i64> %mask) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.avx.maskload.pd(i8*, <2 x i64>) nounwind readonly
+
+
+define <4 x double> @test_x86_avx_maskload_pd_256(i8* %a0, <4 x i64> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx_maskload_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i64> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.maskload.pd.256(i8* [[A0:%.*]], <4 x i64> [[MASK:%.*]])
+; CHECK-NEXT: store <4 x i64> zeroinitializer, <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx.maskload.pd.256(i8* %a0, <4 x i64> %mask) ; <<4 x double>> [#uses=1]
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.maskload.pd.256(i8*, <4 x i64>) nounwind readonly
+
+
+define <4 x float> @test_x86_avx_maskload_ps(i8* %a0, <4 x i32> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx_maskload_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.avx.maskload.ps(i8* [[A0:%.*]], <4 x i32> [[MASK:%.*]])
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.avx.maskload.ps(i8* %a0, <4 x i32> %mask) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx.maskload.ps(i8*, <4 x i32>) nounwind readonly
+
+
+define <8 x float> @test_x86_avx_maskload_ps_256(i8* %a0, <8 x i32> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx_maskload_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.maskload.ps.256(i8* [[A0:%.*]], <8 x i32> [[MASK:%.*]])
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.maskload.ps.256(i8* %a0, <8 x i32> %mask) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.maskload.ps.256(i8*, <8 x i32>) nounwind readonly
+
+
+define void @test_x86_avx_maskstore_pd(i8* %a0, <2 x i64> %mask, <2 x double> %a2) #0 {
+; CHECK-LABEL: @test_x86_avx_maskstore_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[TMP9:%.*]] = bitcast <2 x i64> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP9]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP10:%.*]], label [[TMP11:%.*]], !prof [[PROF0]]
+; CHECK: 10:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 11:
+; CHECK-NEXT: call void @llvm.x86.avx.maskstore.pd(i8* [[A0:%.*]], <2 x i64> [[MASK:%.*]], <2 x double> [[A2:%.*]])
+; CHECK-NEXT: ret void
+;
+ call void @llvm.x86.avx.maskstore.pd(i8* %a0, <2 x i64> %mask, <2 x double> %a2)
+ ret void
+}
+declare void @llvm.x86.avx.maskstore.pd(i8*, <2 x i64>, <2 x double>) nounwind
+
+
+define void @test_x86_avx_maskstore_pd_256(i8* %a0, <4 x i64> %mask, <4 x double> %a2) #0 {
+; CHECK-LABEL: @test_x86_avx_maskstore_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i64> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[TMP9:%.*]] = bitcast <4 x i64> [[TMP3]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP9]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP10:%.*]], label [[TMP11:%.*]], !prof [[PROF0]]
+; CHECK: 10:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 11:
+; CHECK-NEXT: call void @llvm.x86.avx.maskstore.pd.256(i8* [[A0:%.*]], <4 x i64> [[MASK:%.*]], <4 x double> [[A2:%.*]])
+; CHECK-NEXT: ret void
+;
+ call void @llvm.x86.avx.maskstore.pd.256(i8* %a0, <4 x i64> %mask, <4 x double> %a2)
+ ret void
+}
+declare void @llvm.x86.avx.maskstore.pd.256(i8*, <4 x i64>, <4 x double>) nounwind
+
+
+define void @test_x86_avx_maskstore_ps(i8* %a0, <4 x i32> %mask, <4 x float> %a2) #0 {
+; CHECK-LABEL: @test_x86_avx_maskstore_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[TMP9:%.*]] = bitcast <4 x i32> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP9]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP10:%.*]], label [[TMP11:%.*]], !prof [[PROF0]]
+; CHECK: 10:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 11:
+; CHECK-NEXT: call void @llvm.x86.avx.maskstore.ps(i8* [[A0:%.*]], <4 x i32> [[MASK:%.*]], <4 x float> [[A2:%.*]])
+; CHECK-NEXT: ret void
+;
+ call void @llvm.x86.avx.maskstore.ps(i8* %a0, <4 x i32> %mask, <4 x float> %a2)
+ ret void
+}
+declare void @llvm.x86.avx.maskstore.ps(i8*, <4 x i32>, <4 x float>) nounwind
+
+
+define void @test_x86_avx_maskstore_ps_256(i8* %a0, <8 x i32> %mask, <8 x float> %a2) #0 {
+; CHECK-LABEL: @test_x86_avx_maskstore_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[TMP9:%.*]] = bitcast <8 x i32> [[TMP3]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP9]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP10:%.*]], label [[TMP11:%.*]], !prof [[PROF0]]
+; CHECK: 10:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 11:
+; CHECK-NEXT: call void @llvm.x86.avx.maskstore.ps.256(i8* [[A0:%.*]], <8 x i32> [[MASK:%.*]], <8 x float> [[A2:%.*]])
+; CHECK-NEXT: ret void
+;
+ call void @llvm.x86.avx.maskstore.ps.256(i8* %a0, <8 x i32> %mask, <8 x float> %a2)
+ ret void
+}
+declare void @llvm.x86.avx.maskstore.ps.256(i8*, <8 x i32>, <8 x float>) nounwind
+
+
+define <4 x double> @test_x86_avx_max_pd_256(<4 x double> %a0, <4 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_max_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.max.pd.256(<4 x double> [[A0:%.*]], <4 x double> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i64> [[_MSPROP]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx.max.pd.256(<4 x double> %a0, <4 x double> %a1) ; <<4 x double>> [#uses=1]
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.max.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+
+define <8 x float> @test_x86_avx_max_ps_256(<8 x float> %a0, <8 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_max_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.max.ps.256(<8 x float> [[A0:%.*]], <8 x float> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> [[_MSPROP]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.max.ps.256(<8 x float> %a0, <8 x float> %a1) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.max.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+
+define <4 x double> @test_x86_avx_min_pd_256(<4 x double> %a0, <4 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_min_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.min.pd.256(<4 x double> [[A0:%.*]], <4 x double> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i64> [[_MSPROP]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx.min.pd.256(<4 x double> %a0, <4 x double> %a1) ; <<4 x double>> [#uses=1]
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.min.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+
+define <8 x float> @test_x86_avx_min_ps_256(<8 x float> %a0, <8 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_min_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.min.ps.256(<8 x float> [[A0:%.*]], <8 x float> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> [[_MSPROP]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.min.ps.256(<8 x float> %a0, <8 x float> %a1) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.min.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+
+define i32 @test_x86_avx_movmsk_pd_256(<4 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_movmsk_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.movmsk.pd.256(<4 x double> [[A0:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.movmsk.pd.256(<4 x double> %a0) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.movmsk.pd.256(<4 x double>) nounwind readnone
+
+
+define i32 @test_x86_avx_movmsk_ps_256(<8 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_movmsk_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.movmsk.ps.256(<8 x float> [[A0:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.movmsk.ps.256(<8 x float> %a0) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.movmsk.ps.256(<8 x float>) nounwind readnone
+
+
+define i32 @test_x86_avx_ptestc_256(<4 x i64> %a0, <4 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_ptestc_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i64> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.ptestc.256(<4 x i64> [[A0:%.*]], <4 x i64> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.ptestc.256(<4 x i64> %a0, <4 x i64> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.ptestc.256(<4 x i64>, <4 x i64>) nounwind readnone
+
+
+define i32 @test_x86_avx_ptestnzc_256(<4 x i64> %a0, <4 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_ptestnzc_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i64> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.ptestnzc.256(<4 x i64> [[A0:%.*]], <4 x i64> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.ptestnzc.256(<4 x i64> %a0, <4 x i64> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.ptestnzc.256(<4 x i64>, <4 x i64>) nounwind readnone
+
+
+define i32 @test_x86_avx_ptestz_256(<4 x i64> %a0, <4 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_ptestz_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i64> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.ptestz.256(<4 x i64> [[A0:%.*]], <4 x i64> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.ptestz.256(<4 x i64> %a0, <4 x i64> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.ptestz.256(<4 x i64>, <4 x i64>) nounwind readnone
+
+
+define <8 x float> @test_x86_avx_rcp_ps_256(<8 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_rcp_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.rcp.ps.256(<8 x float> [[A0:%.*]])
+; CHECK-NEXT: store <8 x i32> [[TMP1]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.rcp.ps.256(<8 x float> %a0) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.rcp.ps.256(<8 x float>) nounwind readnone
+
+
+define <4 x double> @test_x86_avx_round_pd_256(<4 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_round_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.round.pd.256(<4 x double> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <4 x i64> zeroinitializer, <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx.round.pd.256(<4 x double> %a0, i32 7) ; <<4 x double>> [#uses=1]
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.round.pd.256(<4 x double>, i32) nounwind readnone
+
+
+define <8 x float> @test_x86_avx_round_ps_256(<8 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_round_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.round.ps.256(<8 x float> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.round.ps.256(<8 x float> %a0, i32 7) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.round.ps.256(<8 x float>, i32) nounwind readnone
+
+
+define <8 x float> @test_x86_avx_rsqrt_ps_256(<8 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_rsqrt_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.rsqrt.ps.256(<8 x float> [[A0:%.*]])
+; CHECK-NEXT: store <8 x i32> [[TMP1]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.rsqrt.ps.256(<8 x float> %a0) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.rsqrt.ps.256(<8 x float>) nounwind readnone
+
+define <2 x double> @test_x86_avx_vpermilvar_pd(<2 x double> %a0, <2 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vpermilvar_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.avx.vpermilvar.pd(<2 x double> [[A0:%.*]], <2 x i64> [[A1:%.*]])
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.avx.vpermilvar.pd(<2 x double> %a0, <2 x i64> %a1) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.avx.vpermilvar.pd(<2 x double>, <2 x i64>) nounwind readnone
+
+
+define <4 x double> @test_x86_avx_vpermilvar_pd_256(<4 x double> %a0, <4 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vpermilvar_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i64> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double> [[A0:%.*]], <4 x i64> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i64> zeroinitializer, <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double> %a0, <4 x i64> %a1) ; <<4 x double>> [#uses=1]
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double>, <4 x i64>) nounwind readnone
+
+define <4 x double> @test_x86_avx_vpermilvar_pd_256_2(<4 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx_vpermilvar_pd_256_2(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double> [[A0:%.*]], <4 x i64> <i64 2, i64 0, i64 0, i64 2>)
+; CHECK-NEXT: store <4 x i64> zeroinitializer, <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double> %a0, <4 x i64> <i64 2, i64 0, i64 0, i64 2>) ; <<4 x double>> [#uses=1]
+ ret <4 x double> %res
+}
+
+define <4 x float> @test_x86_avx_vpermilvar_ps(<4 x float> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vpermilvar_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float> [[A0:%.*]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float> %a0, <4 x i32> %a1) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+define <4 x float> @test_x86_avx_vpermilvar_ps_load(<4 x float> %a0, <4 x i32>* %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vpermilvar_ps_load(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[A2:%.*]] = load <4 x i32>, <4 x i32>* [[A1:%.*]], align 16
+; CHECK-NEXT: [[TMP5:%.*]] = ptrtoint <4 x i32>* [[A1]] to i64
+; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
+; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to <4 x i32>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <4 x i32>, <4 x i32>* [[TMP7]], align 16
+; CHECK-NEXT: [[TMP8:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP8]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP9:%.*]], label [[TMP10:%.*]], !prof [[PROF0]]
+; CHECK: 9:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 10:
+; CHECK-NEXT: [[TMP11:%.*]] = bitcast <4 x i32> [[_MSLD]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP11]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP12:%.*]], label [[TMP13:%.*]], !prof [[PROF0]]
+; CHECK: 12:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 13:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float> [[A0:%.*]], <4 x i32> [[A2]])
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %a2 = load <4 x i32>, <4 x i32>* %a1
+ %res = call <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float> %a0, <4 x i32> %a2) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float>, <4 x i32>) nounwind readnone
+
+
+define <8 x float> @test_x86_avx_vpermilvar_ps_256(<8 x float> %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vpermilvar_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.vpermilvar.ps.256(<8 x float> [[A0:%.*]], <8 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx.vpermilvar.ps.256(<8 x float> %a0, <8 x i32> %a1) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.vpermilvar.ps.256(<8 x float>, <8 x i32>) nounwind readnone
+
+
+define i32 @test_x86_avx_vtestc_pd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vtestc_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.vtestc.pd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.vtestc.pd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestc.pd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_avx_vtestc_pd_256(<4 x double> %a0, <4 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vtestc_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i64> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.vtestc.pd.256(<4 x double> [[A0:%.*]], <4 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.vtestc.pd.256(<4 x double> %a0, <4 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestc.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+
+define i32 @test_x86_avx_vtestc_ps(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vtestc_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.vtestc.ps(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.vtestc.ps(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestc.ps(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_avx_vtestc_ps_256(<8 x float> %a0, <8 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vtestc_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.vtestc.ps.256(<8 x float> [[A0:%.*]], <8 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.vtestc.ps.256(<8 x float> %a0, <8 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestc.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+
+define i32 @test_x86_avx_vtestnzc_pd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vtestnzc_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.vtestnzc.pd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.vtestnzc.pd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestnzc.pd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_avx_vtestnzc_pd_256(<4 x double> %a0, <4 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vtestnzc_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i64> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.vtestnzc.pd.256(<4 x double> [[A0:%.*]], <4 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.vtestnzc.pd.256(<4 x double> %a0, <4 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestnzc.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+
+define i32 @test_x86_avx_vtestnzc_ps(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vtestnzc_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.vtestnzc.ps(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.vtestnzc.ps(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestnzc.ps(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_avx_vtestnzc_ps_256(<8 x float> %a0, <8 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vtestnzc_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.vtestnzc.ps.256(<8 x float> [[A0:%.*]], <8 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.vtestnzc.ps.256(<8 x float> %a0, <8 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestnzc.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+
+define i32 @test_x86_avx_vtestz_pd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vtestz_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.vtestz.pd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.vtestz.pd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestz.pd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_avx_vtestz_pd_256(<4 x double> %a0, <4 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vtestz_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i64> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.vtestz.pd.256(<4 x double> [[A0:%.*]], <4 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.vtestz.pd.256(<4 x double> %a0, <4 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestz.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+
+define i32 @test_x86_avx_vtestz_ps(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vtestz_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.vtestz.ps(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.vtestz.ps(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestz.ps(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_avx_vtestz_ps_256(<8 x float> %a0, <8 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx_vtestz_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx.vtestz.ps.256(<8 x float> [[A0:%.*]], <8 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx.vtestz.ps.256(<8 x float> %a0, <8 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestz.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+
+define void @test_x86_avx_vzeroall() #0 {
+; CHECK-LABEL: @test_x86_avx_vzeroall(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: call void @llvm.x86.avx.vzeroall()
+; CHECK-NEXT: ret void
+;
+ call void @llvm.x86.avx.vzeroall()
+ ret void
+}
+declare void @llvm.x86.avx.vzeroall() nounwind
+
+
+define void @test_x86_avx_vzeroupper() #0 {
+; CHECK-LABEL: @test_x86_avx_vzeroupper(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: call void @llvm.x86.avx.vzeroupper()
+; CHECK-NEXT: ret void
+;
+ call void @llvm.x86.avx.vzeroupper()
+ ret void
+}
+declare void @llvm.x86.avx.vzeroupper() nounwind
+
+define void @movnt_dq(i8* %p, <2 x i64> %a1) nounwind #0 {
+; CHECK-LABEL: @movnt_dq(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <2 x i64> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[A2:%.*]] = add <2 x i64> [[A1:%.*]], <i64 1, i64 1>
+; CHECK-NEXT: [[_MSPROP1:%.*]] = shufflevector <2 x i64> [[_MSPROP]], <2 x i64> <i64 -1, i64 -1>, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+; CHECK-NEXT: [[A3:%.*]] = shufflevector <2 x i64> [[A2]], <2 x i64> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+; CHECK-NEXT: [[CAST:%.*]] = bitcast i8* [[P:%.*]] to <4 x i64>*
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = ptrtoint <4 x i64>* [[CAST]] to i64
+; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
+; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to <4 x i64>*
+; CHECK-NEXT: store <4 x i64> [[_MSPROP1]], <4 x i64>* [[TMP7]], align 32
+; CHECK-NEXT: store <4 x i64> [[A3]], <4 x i64>* [[CAST]], align 32, !nontemporal !1
+; CHECK-NEXT: ret void
+;
+ %a2 = add <2 x i64> %a1, <i64 1, i64 1>
+ %a3 = shufflevector <2 x i64> %a2, <2 x i64> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+ tail call void @llvm.x86.avx.movnt.dq.256(i8* %p, <4 x i64> %a3) nounwind
+ ret void
+}
+declare void @llvm.x86.avx.movnt.dq.256(i8*, <4 x i64>) nounwind
+
+define void @movnt_ps(i8* %p, <8 x float> %a) nounwind #0 {
+; CHECK-LABEL: @movnt_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[CAST:%.*]] = bitcast i8* [[P:%.*]] to <8 x float>*
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = ptrtoint <8 x float>* [[CAST]] to i64
+; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
+; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to <8 x i32>*
+; CHECK-NEXT: store <8 x i32> [[TMP2]], <8 x i32>* [[TMP7]], align 32
+; CHECK-NEXT: store <8 x float> [[A:%.*]], <8 x float>* [[CAST]], align 32, !nontemporal !1
+; CHECK-NEXT: ret void
+;
+ tail call void @llvm.x86.avx.movnt.ps.256(i8* %p, <8 x float> %a) nounwind
+ ret void
+}
+declare void @llvm.x86.avx.movnt.ps.256(i8*, <8 x float>) nounwind
+
+define void @movnt_pd(i8* %p, <4 x double> %a1) nounwind #0 {
+ ; add operation forces the execution domain.
+; CHECK-LABEL: @movnt_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i64> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[A2:%.*]] = fadd <4 x double> [[A1:%.*]], zeroinitializer
+; CHECK-NEXT: [[CAST:%.*]] = bitcast i8* [[P:%.*]] to <4 x double>*
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR8]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = ptrtoint <4 x double>* [[CAST]] to i64
+; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
+; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to <4 x i64>*
+; CHECK-NEXT: store <4 x i64> [[_MSPROP]], <4 x i64>* [[TMP7]], align 32
+; CHECK-NEXT: store <4 x double> [[A2]], <4 x double>* [[CAST]], align 32, !nontemporal !1
+; CHECK-NEXT: ret void
+;
+ %a2 = fadd <4 x double> %a1, <double 0x0, double 0x0, double 0x0, double 0x0>
+ tail call void @llvm.x86.avx.movnt.pd.256(i8* %p, <4 x double> %a2) nounwind
+ ret void
+}
+declare void @llvm.x86.avx.movnt.pd.256(i8*, <4 x double>) nounwind
+
+
+define <2 x i64> @test_x86_pclmulqdq(<2 x i64> %a0, <2 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_pclmulqdq(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <2 x i64> [[TMP1]], <2 x i64> poison, <2 x i32> zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = shufflevector <2 x i64> [[TMP2]], <2 x i64> poison, <2 x i32> zeroinitializer
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <2 x i64> [[TMP3]], [[TMP4]]
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.pclmulqdq(<2 x i64> [[A0:%.*]], <2 x i64> [[A1:%.*]], i8 0)
+; CHECK-NEXT: store <2 x i64> [[_MSPROP]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.pclmulqdq(<2 x i64> %a0, <2 x i64> %a1, i8 0) ; <<2 x i64>> [#uses=1]
+ ret <2 x i64> %res
+}
+declare <2 x i64> @llvm.x86.pclmulqdq(<2 x i64>, <2 x i64>, i8) nounwind readnone
+
+attributes #0 = { sanitize_memory }
diff --git a/llvm/test/Instrumentation/MemorySanitizer/avx2-intrinsics-x86.ll b/llvm/test/Instrumentation/MemorySanitizer/avx2-intrinsics-x86.ll
new file mode 100644
index 0000000000000..7c896c6ff0088
--- /dev/null
+++ b/llvm/test/Instrumentation/MemorySanitizer/avx2-intrinsics-x86.ll
@@ -0,0 +1,2427 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt %s -S -passes=msan 2>&1 | FileCheck %s
+
+target datalayout = "e-m:o-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+define <16 x i16> @test_x86_avx2_packssdw(<8 x i32> %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_packssdw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <8 x i32> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <8 x i1> [[TMP3]] to <8 x i32>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <8 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <8 x i1> [[TMP5]] to <8 x i32>
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <16 x i16> @llvm.x86.avx2.packssdw(<8 x i32> [[TMP4]], <8 x i32> [[TMP6]])
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.packssdw(<8 x i32> [[A0:%.*]], <8 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[_MSPROP_VECTOR_PACK]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.packssdw(<8 x i32> %a0, <8 x i32> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.packssdw(<8 x i32>, <8 x i32>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_packssdw_fold() #0 {
+; CHECK-LABEL: @test_x86_avx2_packssdw_fold(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <16 x i16> @llvm.x86.avx2.packssdw(<8 x i32> zeroinitializer, <8 x i32> zeroinitializer)
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.packssdw(<8 x i32> zeroinitializer, <8 x i32> <i32 255, i32 32767, i32 65535, i32 -1, i32 -32767, i32 -65535, i32 0, i32 -256>)
+; CHECK-NEXT: store <16 x i16> [[_MSPROP_VECTOR_PACK]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.packssdw(<8 x i32> zeroinitializer, <8 x i32> <i32 255, i32 32767, i32 65535, i32 -1, i32 -32767, i32 -65535, i32 0, i32 -256>)
+ ret <16 x i16> %res
+}
+
+
+define <32 x i8> @test_x86_avx2_packsswb(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_packsswb(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <16 x i16> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <16 x i1> [[TMP3]] to <16 x i16>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <16 x i16> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <16 x i1> [[TMP5]] to <16 x i16>
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <32 x i8> @llvm.x86.avx2.packsswb(<16 x i16> [[TMP4]], <16 x i16> [[TMP6]])
+; CHECK-NEXT: [[RES:%.*]] = call <32 x i8> @llvm.x86.avx2.packsswb(<16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <32 x i8> [[_MSPROP_VECTOR_PACK]], <32 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <32 x i8>*), align 8
+; CHECK-NEXT: ret <32 x i8> [[RES]]
+;
+ %res = call <32 x i8> @llvm.x86.avx2.packsswb(<16 x i16> %a0, <16 x i16> %a1) ; <<32 x i8>> [#uses=1]
+ ret <32 x i8> %res
+}
+declare <32 x i8> @llvm.x86.avx2.packsswb(<16 x i16>, <16 x i16>) nounwind readnone
+
+
+define <32 x i8> @test_x86_avx2_packsswb_fold() #0 {
+; CHECK-LABEL: @test_x86_avx2_packsswb_fold(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <32 x i8> @llvm.x86.avx2.packsswb(<16 x i16> zeroinitializer, <16 x i16> zeroinitializer)
+; CHECK-NEXT: [[RES:%.*]] = call <32 x i8> @llvm.x86.avx2.packsswb(<16 x i16> <i16 0, i16 255, i16 256, i16 -1, i16 -1, i16 -255, i16 -256, i16 -32678, i16 0, i16 255, i16 256, i16 -1, i16 -1, i16 -255, i16 -256, i16 -32678>, <16 x i16> zeroinitializer)
+; CHECK-NEXT: store <32 x i8> [[_MSPROP_VECTOR_PACK]], <32 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <32 x i8>*), align 8
+; CHECK-NEXT: ret <32 x i8> [[RES]]
+;
+ %res = call <32 x i8> @llvm.x86.avx2.packsswb(<16 x i16> <i16 0, i16 255, i16 256, i16 65535, i16 -1, i16 -255, i16 -256, i16 -32678, i16 0, i16 255, i16 256, i16 65535, i16 -1, i16 -255, i16 -256, i16 -32678>, <16 x i16> zeroinitializer)
+ ret <32 x i8> %res
+}
+
+
+define <32 x i8> @test_x86_avx2_packuswb(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_packuswb(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <16 x i16> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <16 x i1> [[TMP3]] to <16 x i16>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <16 x i16> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <16 x i1> [[TMP5]] to <16 x i16>
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <32 x i8> @llvm.x86.avx2.packsswb(<16 x i16> [[TMP4]], <16 x i16> [[TMP6]])
+; CHECK-NEXT: [[RES:%.*]] = call <32 x i8> @llvm.x86.avx2.packuswb(<16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <32 x i8> [[_MSPROP_VECTOR_PACK]], <32 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <32 x i8>*), align 8
+; CHECK-NEXT: ret <32 x i8> [[RES]]
+;
+ %res = call <32 x i8> @llvm.x86.avx2.packuswb(<16 x i16> %a0, <16 x i16> %a1) ; <<32 x i8>> [#uses=1]
+ ret <32 x i8> %res
+}
+declare <32 x i8> @llvm.x86.avx2.packuswb(<16 x i16>, <16 x i16>) nounwind readnone
+
+
+define <32 x i8> @test_x86_avx2_packuswb_fold() #0 {
+; CHECK-LABEL: @test_x86_avx2_packuswb_fold(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <32 x i8> @llvm.x86.avx2.packsswb(<16 x i16> zeroinitializer, <16 x i16> zeroinitializer)
+; CHECK-NEXT: [[RES:%.*]] = call <32 x i8> @llvm.x86.avx2.packuswb(<16 x i16> <i16 0, i16 255, i16 256, i16 -1, i16 -1, i16 -255, i16 -256, i16 -32678, i16 0, i16 255, i16 256, i16 -1, i16 -1, i16 -255, i16 -256, i16 -32678>, <16 x i16> zeroinitializer)
+; CHECK-NEXT: store <32 x i8> [[_MSPROP_VECTOR_PACK]], <32 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <32 x i8>*), align 8
+; CHECK-NEXT: ret <32 x i8> [[RES]]
+;
+ %res = call <32 x i8> @llvm.x86.avx2.packuswb(<16 x i16> <i16 0, i16 255, i16 256, i16 65535, i16 -1, i16 -255, i16 -256, i16 -32678, i16 0, i16 255, i16 256, i16 65535, i16 -1, i16 -255, i16 -256, i16 -32678>, <16 x i16> zeroinitializer)
+ ret <32 x i8> %res
+}
+
+
+define <32 x i8> @test_x86_avx2_pavg_b(<32 x i8> %a0, <32 x i8> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_pavg_b(
+; CHECK-NEXT: [[TMP1:%.*]] = load <32 x i8>, <32 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <32 x i8>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <32 x i8>, <32 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <32 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <32 x i8> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <32 x i8> @llvm.x86.avx2.pavg.b(<32 x i8> [[A0:%.*]], <32 x i8> [[A1:%.*]])
+; CHECK-NEXT: store <32 x i8> [[_MSPROP]], <32 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <32 x i8>*), align 8
+; CHECK-NEXT: ret <32 x i8> [[RES]]
+;
+ %res = call <32 x i8> @llvm.x86.avx2.pavg.b(<32 x i8> %a0, <32 x i8> %a1) ; <<32 x i8>> [#uses=1]
+ ret <32 x i8> %res
+}
+declare <32 x i8> @llvm.x86.avx2.pavg.b(<32 x i8>, <32 x i8>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_pavg_w(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_pavg_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <16 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.pavg.w(<16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[_MSPROP]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.pavg.w(<16 x i16> %a0, <16 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.pavg.w(<16 x i16>, <16 x i16>) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_pmadd_wd(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_pmadd_wd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <16 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = bitcast <16 x i16> [[TMP3]] to <8 x i32>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <8 x i32> [[TMP4]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <8 x i1> [[TMP5]] to <8 x i32>
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.pmadd.wd(<16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> [[TMP6]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.pmadd.wd(<16 x i16> %a0, <16 x i16> %a1) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.pmadd.wd(<16 x i16>, <16 x i16>) nounwind readnone
+
+
+define i32 @test_x86_avx2_pmovmskb(<32 x i8> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx2_pmovmskb(
+; CHECK-NEXT: [[TMP1:%.*]] = load <32 x i8>, <32 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <32 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <32 x i8> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0:![0-9]+]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6:[0-9]+]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.avx2.pmovmskb(<32 x i8> [[A0:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.avx2.pmovmskb(<32 x i8> %a0) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx2.pmovmskb(<32 x i8>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_pmulh_w(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_pmulh_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <16 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.pmulh.w(<16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[_MSPROP]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.pmulh.w(<16 x i16> %a0, <16 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.pmulh.w(<16 x i16>, <16 x i16>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_pmulhu_w(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_pmulhu_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <16 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.pmulhu.w(<16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[_MSPROP]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.pmulhu.w(<16 x i16> %a0, <16 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.pmulhu.w(<16 x i16>, <16 x i16>) nounwind readnone
+
+
+define <4 x i64> @test_x86_avx2_psad_bw(<32 x i8> %a0, <32 x i8> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psad_bw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <32 x i8>, <32 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <32 x i8>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <32 x i8>, <32 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <32 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <32 x i8> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = bitcast <32 x i8> [[TMP3]] to <4 x i64>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <4 x i64> [[TMP4]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <4 x i1> [[TMP5]] to <4 x i64>
+; CHECK-NEXT: [[TMP7:%.*]] = lshr <4 x i64> [[TMP6]], <i64 48, i64 48, i64 48, i64 48>
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i64> @llvm.x86.avx2.psad.bw(<32 x i8> [[A0:%.*]], <32 x i8> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i64> [[TMP7]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x i64> [[RES]]
+;
+ %res = call <4 x i64> @llvm.x86.avx2.psad.bw(<32 x i8> %a0, <32 x i8> %a1) ; <<4 x i64>> [#uses=1]
+ ret <4 x i64> %res
+}
+declare <4 x i64> @llvm.x86.avx2.psad.bw(<32 x i8>, <32 x i8>) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_psll_d(<8 x i32> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psll_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i256
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i256 [[TMP6]] to <8 x i32>
+; CHECK-NEXT: [[TMP8:%.*]] = call <8 x i32> @llvm.x86.avx2.psll.d(<8 x i32> [[TMP1]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <8 x i32> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.psll.d(<8 x i32> [[A0:%.*]], <4 x i32> [[A1]])
+; CHECK-NEXT: store <8 x i32> [[TMP9]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.psll.d(<8 x i32> %a0, <4 x i32> %a1) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.psll.d(<8 x i32>, <4 x i32>) nounwind readnone
+
+
+define <4 x i64> @test_x86_avx2_psll_q(<4 x i64> %a0, <2 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psll_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i256
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i256 [[TMP6]] to <4 x i64>
+; CHECK-NEXT: [[TMP8:%.*]] = call <4 x i64> @llvm.x86.avx2.psll.q(<4 x i64> [[TMP1]], <2 x i64> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <4 x i64> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i64> @llvm.x86.avx2.psll.q(<4 x i64> [[A0:%.*]], <2 x i64> [[A1]])
+; CHECK-NEXT: store <4 x i64> [[TMP9]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x i64> [[RES]]
+;
+ %res = call <4 x i64> @llvm.x86.avx2.psll.q(<4 x i64> %a0, <2 x i64> %a1) ; <<4 x i64>> [#uses=1]
+ ret <4 x i64> %res
+}
+declare <4 x i64> @llvm.x86.avx2.psll.q(<4 x i64>, <2 x i64>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_psll_w(<16 x i16> %a0, <8 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psll_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i16> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i256
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i256 [[TMP6]] to <16 x i16>
+; CHECK-NEXT: [[TMP8:%.*]] = call <16 x i16> @llvm.x86.avx2.psll.w(<16 x i16> [[TMP1]], <8 x i16> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <16 x i16> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.psll.w(<16 x i16> [[A0:%.*]], <8 x i16> [[A1]])
+; CHECK-NEXT: store <16 x i16> [[TMP9]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.psll.w(<16 x i16> %a0, <8 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.psll.w(<16 x i16>, <8 x i16>) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_pslli_d(<8 x i32> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx2_pslli_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <8 x i32> @llvm.x86.avx2.pslli.d(<8 x i32> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <8 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.pslli.d(<8 x i32> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <8 x i32> [[TMP3]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.pslli.d(<8 x i32> %a0, i32 7) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.pslli.d(<8 x i32>, i32) nounwind readnone
+
+
+define <4 x i64> @test_x86_avx2_pslli_q(<4 x i64> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx2_pslli_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <4 x i64> @llvm.x86.avx2.pslli.q(<4 x i64> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i64> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i64> @llvm.x86.avx2.pslli.q(<4 x i64> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <4 x i64> [[TMP3]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x i64> [[RES]]
+;
+ %res = call <4 x i64> @llvm.x86.avx2.pslli.q(<4 x i64> %a0, i32 7) ; <<4 x i64>> [#uses=1]
+ ret <4 x i64> %res
+}
+declare <4 x i64> @llvm.x86.avx2.pslli.q(<4 x i64>, i32) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_pslli_w(<16 x i16> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx2_pslli_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <16 x i16> @llvm.x86.avx2.pslli.w(<16 x i16> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <16 x i16> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.pslli.w(<16 x i16> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <16 x i16> [[TMP3]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.pslli.w(<16 x i16> %a0, i32 7) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.pslli.w(<16 x i16>, i32) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_psra_d(<8 x i32> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psra_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i256
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i256 [[TMP6]] to <8 x i32>
+; CHECK-NEXT: [[TMP8:%.*]] = call <8 x i32> @llvm.x86.avx2.psra.d(<8 x i32> [[TMP1]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <8 x i32> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.psra.d(<8 x i32> [[A0:%.*]], <4 x i32> [[A1]])
+; CHECK-NEXT: store <8 x i32> [[TMP9]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.psra.d(<8 x i32> %a0, <4 x i32> %a1) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.psra.d(<8 x i32>, <4 x i32>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_psra_w(<16 x i16> %a0, <8 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psra_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i16> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i256
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i256 [[TMP6]] to <16 x i16>
+; CHECK-NEXT: [[TMP8:%.*]] = call <16 x i16> @llvm.x86.avx2.psra.w(<16 x i16> [[TMP1]], <8 x i16> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <16 x i16> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.psra.w(<16 x i16> [[A0:%.*]], <8 x i16> [[A1]])
+; CHECK-NEXT: store <16 x i16> [[TMP9]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.psra.w(<16 x i16> %a0, <8 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.psra.w(<16 x i16>, <8 x i16>) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_psrai_d(<8 x i32> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrai_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <8 x i32> @llvm.x86.avx2.psrai.d(<8 x i32> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <8 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.psrai.d(<8 x i32> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <8 x i32> [[TMP3]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.psrai.d(<8 x i32> %a0, i32 7) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.psrai.d(<8 x i32>, i32) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_psrai_w(<16 x i16> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrai_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <16 x i16> @llvm.x86.avx2.psrai.w(<16 x i16> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <16 x i16> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.psrai.w(<16 x i16> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <16 x i16> [[TMP3]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.psrai.w(<16 x i16> %a0, i32 7) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.psrai.w(<16 x i16>, i32) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_psrl_d(<8 x i32> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrl_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i256
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i256 [[TMP6]] to <8 x i32>
+; CHECK-NEXT: [[TMP8:%.*]] = call <8 x i32> @llvm.x86.avx2.psrl.d(<8 x i32> [[TMP1]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <8 x i32> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.psrl.d(<8 x i32> [[A0:%.*]], <4 x i32> [[A1]])
+; CHECK-NEXT: store <8 x i32> [[TMP9]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.psrl.d(<8 x i32> %a0, <4 x i32> %a1) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.psrl.d(<8 x i32>, <4 x i32>) nounwind readnone
+
+
+define <4 x i64> @test_x86_avx2_psrl_q(<4 x i64> %a0, <2 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrl_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i256
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i256 [[TMP6]] to <4 x i64>
+; CHECK-NEXT: [[TMP8:%.*]] = call <4 x i64> @llvm.x86.avx2.psrl.q(<4 x i64> [[TMP1]], <2 x i64> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <4 x i64> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i64> @llvm.x86.avx2.psrl.q(<4 x i64> [[A0:%.*]], <2 x i64> [[A1]])
+; CHECK-NEXT: store <4 x i64> [[TMP9]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x i64> [[RES]]
+;
+ %res = call <4 x i64> @llvm.x86.avx2.psrl.q(<4 x i64> %a0, <2 x i64> %a1) ; <<4 x i64>> [#uses=1]
+ ret <4 x i64> %res
+}
+declare <4 x i64> @llvm.x86.avx2.psrl.q(<4 x i64>, <2 x i64>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_psrl_w(<16 x i16> %a0, <8 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrl_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i16> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i256
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i256 [[TMP6]] to <16 x i16>
+; CHECK-NEXT: [[TMP8:%.*]] = call <16 x i16> @llvm.x86.avx2.psrl.w(<16 x i16> [[TMP1]], <8 x i16> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <16 x i16> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.psrl.w(<16 x i16> [[A0:%.*]], <8 x i16> [[A1]])
+; CHECK-NEXT: store <16 x i16> [[TMP9]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.psrl.w(<16 x i16> %a0, <8 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.psrl.w(<16 x i16>, <8 x i16>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_psrl_w_load(<16 x i16> %a0, <8 x i16>* %p) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrl_w_load(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to i64*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[A1:%.*]] = load <8 x i16>, <8 x i16>* [[P:%.*]], align 16
+; CHECK-NEXT: [[TMP5:%.*]] = ptrtoint <8 x i16>* [[P]] to i64
+; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
+; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to <8 x i16>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <8 x i16>, <8 x i16>* [[TMP7]], align 16
+; CHECK-NEXT: [[TMP8:%.*]] = bitcast <8 x i16> [[_MSLD]] to i128
+; CHECK-NEXT: [[TMP9:%.*]] = trunc i128 [[TMP8]] to i64
+; CHECK-NEXT: [[TMP10:%.*]] = icmp ne i64 [[TMP9]], 0
+; CHECK-NEXT: [[TMP11:%.*]] = sext i1 [[TMP10]] to i256
+; CHECK-NEXT: [[TMP12:%.*]] = bitcast i256 [[TMP11]] to <16 x i16>
+; CHECK-NEXT: [[TMP13:%.*]] = call <16 x i16> @llvm.x86.avx2.psrl.w(<16 x i16> [[TMP2]], <8 x i16> [[A1]])
+; CHECK-NEXT: [[TMP14:%.*]] = or <16 x i16> [[TMP13]], [[TMP12]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.psrl.w(<16 x i16> [[A0:%.*]], <8 x i16> [[A1]])
+; CHECK-NEXT: store <16 x i16> [[TMP14]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %a1 = load <8 x i16>, <8 x i16>* %p
+ %res = call <16 x i16> @llvm.x86.avx2.psrl.w(<16 x i16> %a0, <8 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+
+
+define <8 x i32> @test_x86_avx2_psrli_d(<8 x i32> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrli_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <8 x i32> @llvm.x86.avx2.psrli.d(<8 x i32> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <8 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.psrli.d(<8 x i32> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <8 x i32> [[TMP3]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.psrli.d(<8 x i32> %a0, i32 7) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.psrli.d(<8 x i32>, i32) nounwind readnone
+
+
+define <4 x i64> @test_x86_avx2_psrli_q(<4 x i64> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrli_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <4 x i64> @llvm.x86.avx2.psrli.q(<4 x i64> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i64> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i64> @llvm.x86.avx2.psrli.q(<4 x i64> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <4 x i64> [[TMP3]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x i64> [[RES]]
+;
+ %res = call <4 x i64> @llvm.x86.avx2.psrli.q(<4 x i64> %a0, i32 7) ; <<4 x i64>> [#uses=1]
+ ret <4 x i64> %res
+}
+declare <4 x i64> @llvm.x86.avx2.psrli.q(<4 x i64>, i32) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_psrli_w(<16 x i16> %a0) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrli_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <16 x i16> @llvm.x86.avx2.psrli.w(<16 x i16> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <16 x i16> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.psrli.w(<16 x i16> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <16 x i16> [[TMP3]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.psrli.w(<16 x i16> %a0, i32 7) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.psrli.w(<16 x i16>, i32) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_phadd_d(<8 x i32> %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_phadd_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.phadd.d(<8 x i32> [[A0:%.*]], <8 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> [[_MSPROP]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.phadd.d(<8 x i32> %a0, <8 x i32> %a1) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.phadd.d(<8 x i32>, <8 x i32>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_phadd_sw(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_phadd_sw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <16 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.phadd.sw(<16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[_MSPROP]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.phadd.sw(<16 x i16> %a0, <16 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.phadd.sw(<16 x i16>, <16 x i16>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_phadd_w(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_phadd_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <16 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.phadd.w(<16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[_MSPROP]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.phadd.w(<16 x i16> %a0, <16 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.phadd.w(<16 x i16>, <16 x i16>) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_phsub_d(<8 x i32> %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_phsub_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.phsub.d(<8 x i32> [[A0:%.*]], <8 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> [[_MSPROP]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.phsub.d(<8 x i32> %a0, <8 x i32> %a1) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.phsub.d(<8 x i32>, <8 x i32>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_phsub_sw(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_phsub_sw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <16 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.phsub.sw(<16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[_MSPROP]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.phsub.sw(<16 x i16> %a0, <16 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.phsub.sw(<16 x i16>, <16 x i16>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_phsub_w(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_phsub_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <16 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.phsub.w(<16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[_MSPROP]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.phsub.w(<16 x i16> %a0, <16 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.phsub.w(<16 x i16>, <16 x i16>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_pmadd_ub_sw(<32 x i8> %a0, <32 x i8> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_pmadd_ub_sw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <32 x i8>, <32 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <32 x i8>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <32 x i8>, <32 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <32 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <32 x i8> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = bitcast <32 x i8> [[TMP3]] to <16 x i16>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <16 x i16> [[TMP4]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <16 x i1> [[TMP5]] to <16 x i16>
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.pmadd.ub.sw(<32 x i8> [[A0:%.*]], <32 x i8> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[TMP6]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.pmadd.ub.sw(<32 x i8> %a0, <32 x i8> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.pmadd.ub.sw(<32 x i8>, <32 x i8>) nounwind readnone
+
+define <16 x i16> @test_x86_avx2_pmadd_ub_sw_load_op0(<32 x i8>* %ptr, <32 x i8> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_pmadd_ub_sw_load_op0(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <32 x i8>, <32 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <32 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[A0:%.*]] = load <32 x i8>, <32 x i8>* [[PTR:%.*]], align 32
+; CHECK-NEXT: [[TMP5:%.*]] = ptrtoint <32 x i8>* [[PTR]] to i64
+; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
+; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to <32 x i8>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <32 x i8>, <32 x i8>* [[TMP7]], align 32
+; CHECK-NEXT: [[TMP8:%.*]] = or <32 x i8> [[_MSLD]], [[TMP2]]
+; CHECK-NEXT: [[TMP9:%.*]] = bitcast <32 x i8> [[TMP8]] to <16 x i16>
+; CHECK-NEXT: [[TMP10:%.*]] = icmp ne <16 x i16> [[TMP9]], zeroinitializer
+; CHECK-NEXT: [[TMP11:%.*]] = sext <16 x i1> [[TMP10]] to <16 x i16>
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.pmadd.ub.sw(<32 x i8> [[A0]], <32 x i8> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[TMP11]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %a0 = load <32 x i8>, <32 x i8>* %ptr
+ %res = call <16 x i16> @llvm.x86.avx2.pmadd.ub.sw(<32 x i8> %a0, <32 x i8> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+
+define <16 x i16> @test_x86_avx2_pmul_hr_sw(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_pmul_hr_sw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <16 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.pmul.hr.sw(<16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[_MSPROP]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.pmul.hr.sw(<16 x i16> %a0, <16 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.pmul.hr.sw(<16 x i16>, <16 x i16>) nounwind readnone
+
+
+define <32 x i8> @test_x86_avx2_pshuf_b(<32 x i8> %a0, <32 x i8> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_pshuf_b(
+; CHECK-NEXT: [[TMP1:%.*]] = load <32 x i8>, <32 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <32 x i8>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <32 x i8>, <32 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <32 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <32 x i8> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <32 x i8> @llvm.x86.avx2.pshuf.b(<32 x i8> [[A0:%.*]], <32 x i8> [[A1:%.*]])
+; CHECK-NEXT: store <32 x i8> [[_MSPROP]], <32 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <32 x i8>*), align 8
+; CHECK-NEXT: ret <32 x i8> [[RES]]
+;
+ %res = call <32 x i8> @llvm.x86.avx2.pshuf.b(<32 x i8> %a0, <32 x i8> %a1) ; <<16 x i8>> [#uses=1]
+ ret <32 x i8> %res
+}
+declare <32 x i8> @llvm.x86.avx2.pshuf.b(<32 x i8>, <32 x i8>) nounwind readnone
+
+
+define <32 x i8> @test_x86_avx2_psign_b(<32 x i8> %a0, <32 x i8> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psign_b(
+; CHECK-NEXT: [[TMP1:%.*]] = load <32 x i8>, <32 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <32 x i8>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <32 x i8>, <32 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <32 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <32 x i8> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <32 x i8> @llvm.x86.avx2.psign.b(<32 x i8> [[A0:%.*]], <32 x i8> [[A1:%.*]])
+; CHECK-NEXT: store <32 x i8> [[_MSPROP]], <32 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <32 x i8>*), align 8
+; CHECK-NEXT: ret <32 x i8> [[RES]]
+;
+ %res = call <32 x i8> @llvm.x86.avx2.psign.b(<32 x i8> %a0, <32 x i8> %a1) ; <<32 x i8>> [#uses=1]
+ ret <32 x i8> %res
+}
+declare <32 x i8> @llvm.x86.avx2.psign.b(<32 x i8>, <32 x i8>) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_psign_d(<8 x i32> %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psign_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.psign.d(<8 x i32> [[A0:%.*]], <8 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> [[_MSPROP]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.psign.d(<8 x i32> %a0, <8 x i32> %a1) ; <<4 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.psign.d(<8 x i32>, <8 x i32>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_psign_w(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psign_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <16 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.psign.w(<16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[_MSPROP]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.psign.w(<16 x i16> %a0, <16 x i16> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.psign.w(<16 x i16>, <16 x i16>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_mpsadbw(<32 x i8> %a0, <32 x i8> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_mpsadbw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <32 x i8>, <32 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <32 x i8>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <32 x i8>, <32 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <32 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <32 x i8> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <32 x i8> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.mpsadbw(<32 x i8> [[A0:%.*]], <32 x i8> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <16 x i16> zeroinitializer, <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.mpsadbw(<32 x i8> %a0, <32 x i8> %a1, i8 7) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.mpsadbw(<32 x i8>, <32 x i8>, i8) nounwind readnone
+
+define <16 x i16> @test_x86_avx2_mpsadbw_load_op0(<32 x i8>* %ptr, <32 x i8> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_mpsadbw_load_op0(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <32 x i8>, <32 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <32 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[A0:%.*]] = load <32 x i8>, <32 x i8>* [[PTR:%.*]], align 32
+; CHECK-NEXT: [[TMP5:%.*]] = ptrtoint <32 x i8>* [[PTR]] to i64
+; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
+; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to <32 x i8>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <32 x i8>, <32 x i8>* [[TMP7]], align 32
+; CHECK-NEXT: [[TMP8:%.*]] = bitcast <32 x i8> [[_MSLD]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP8]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP9:%.*]], label [[TMP10:%.*]], !prof [[PROF0]]
+; CHECK: 9:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 10:
+; CHECK-NEXT: [[TMP11:%.*]] = bitcast <32 x i8> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP11]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP12:%.*]], label [[TMP13:%.*]], !prof [[PROF0]]
+; CHECK: 12:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 13:
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.mpsadbw(<32 x i8> [[A0]], <32 x i8> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <16 x i16> zeroinitializer, <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %a0 = load <32 x i8>, <32 x i8>* %ptr
+ %res = call <16 x i16> @llvm.x86.avx2.mpsadbw(<32 x i8> %a0, <32 x i8> %a1, i8 7) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+
+define <16 x i16> @test_x86_avx2_packusdw(<8 x i32> %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_packusdw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <8 x i32> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <8 x i1> [[TMP3]] to <8 x i32>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <8 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <8 x i1> [[TMP5]] to <8 x i32>
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <16 x i16> @llvm.x86.avx2.packssdw(<8 x i32> [[TMP4]], <8 x i32> [[TMP6]])
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.packusdw(<8 x i32> [[A0:%.*]], <8 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i16> [[_MSPROP_VECTOR_PACK]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.packusdw(<8 x i32> %a0, <8 x i32> %a1) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.packusdw(<8 x i32>, <8 x i32>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_packusdw_fold() #0 {
+; CHECK-LABEL: @test_x86_avx2_packusdw_fold(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <16 x i16> @llvm.x86.avx2.packssdw(<8 x i32> zeroinitializer, <8 x i32> zeroinitializer)
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i16> @llvm.x86.avx2.packusdw(<8 x i32> zeroinitializer, <8 x i32> <i32 255, i32 32767, i32 65535, i32 -1, i32 -32767, i32 -65535, i32 0, i32 -256>)
+; CHECK-NEXT: store <16 x i16> [[_MSPROP_VECTOR_PACK]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[RES]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.packusdw(<8 x i32> zeroinitializer, <8 x i32> <i32 255, i32 32767, i32 65535, i32 -1, i32 -32767, i32 -65535, i32 0, i32 -256>)
+ ret <16 x i16> %res
+}
+
+
+define <32 x i8> @test_x86_avx2_pblendvb(<32 x i8> %a0, <32 x i8> %a1, <32 x i8> %a2) #0 {
+; CHECK-LABEL: @test_x86_avx2_pblendvb(
+; CHECK-NEXT: [[TMP1:%.*]] = load <32 x i8>, <32 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <32 x i8>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <32 x i8>, <32 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <32 x i8>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <32 x i8>, <32 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 64) to <32 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <32 x i8> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[_MSPROP1:%.*]] = or <32 x i8> [[_MSPROP]], [[TMP3]]
+; CHECK-NEXT: [[RES:%.*]] = call <32 x i8> @llvm.x86.avx2.pblendvb(<32 x i8> [[A0:%.*]], <32 x i8> [[A1:%.*]], <32 x i8> [[A2:%.*]])
+; CHECK-NEXT: store <32 x i8> [[_MSPROP1]], <32 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <32 x i8>*), align 8
+; CHECK-NEXT: ret <32 x i8> [[RES]]
+;
+ %res = call <32 x i8> @llvm.x86.avx2.pblendvb(<32 x i8> %a0, <32 x i8> %a1, <32 x i8> %a2) ; <<32 x i8>> [#uses=1]
+ ret <32 x i8> %res
+}
+declare <32 x i8> @llvm.x86.avx2.pblendvb(<32 x i8>, <32 x i8>, <32 x i8>) nounwind readnone
+
+
+define <16 x i16> @test_x86_avx2_pblendw(<16 x i16> %a0, <16 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_pblendw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i16>, <16 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i16>, <16 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = shufflevector <16 x i16> [[TMP1]], <16 x i16> [[TMP2]], <16 x i32> <i32 16, i32 17, i32 18, i32 3, i32 4, i32 5, i32 6, i32 7, i32 24, i32 25, i32 26, i32 11, i32 12, i32 13, i32 14, i32 15>
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <16 x i16> [[A0:%.*]], <16 x i16> [[A1:%.*]], <16 x i32> <i32 16, i32 17, i32 18, i32 3, i32 4, i32 5, i32 6, i32 7, i32 24, i32 25, i32 26, i32 11, i32 12, i32 13, i32 14, i32 15>
+; CHECK-NEXT: store <16 x i16> [[_MSPROP]], <16 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i16>*), align 8
+; CHECK-NEXT: ret <16 x i16> [[TMP3]]
+;
+ %res = call <16 x i16> @llvm.x86.avx2.pblendw(<16 x i16> %a0, <16 x i16> %a1, i8 7) ; <<16 x i16>> [#uses=1]
+ ret <16 x i16> %res
+}
+declare <16 x i16> @llvm.x86.avx2.pblendw(<16 x i16>, <16 x i16>, i8) nounwind readnone
+
+
+define <4 x i32> @test_x86_avx2_pblendd_128(<4 x i32> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_pblendd_128(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = shufflevector <4 x i32> [[TMP1]], <4 x i32> [[TMP2]], <4 x i32> <i32 4, i32 5, i32 6, i32 3>
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <4 x i32> [[A0:%.*]], <4 x i32> [[A1:%.*]], <4 x i32> <i32 4, i32 5, i32 6, i32 3>
+; CHECK-NEXT: store <4 x i32> [[_MSPROP]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[TMP3]]
+;
+ %res = call <4 x i32> @llvm.x86.avx2.pblendd.128(<4 x i32> %a0, <4 x i32> %a1, i8 7) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.avx2.pblendd.128(<4 x i32>, <4 x i32>, i8) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_pblendd_256(<8 x i32> %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_pblendd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = shufflevector <8 x i32> [[TMP1]], <8 x i32> [[TMP2]], <8 x i32> <i32 8, i32 9, i32 10, i32 3, i32 4, i32 5, i32 6, i32 7>
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <8 x i32> [[A0:%.*]], <8 x i32> [[A1:%.*]], <8 x i32> <i32 8, i32 9, i32 10, i32 3, i32 4, i32 5, i32 6, i32 7>
+; CHECK-NEXT: store <8 x i32> [[_MSPROP]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[TMP3]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.pblendd.256(<8 x i32> %a0, <8 x i32> %a1, i8 7) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.pblendd.256(<8 x i32>, <8 x i32>, i8) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_permd(<8 x i32> %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_permd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.permd(<8 x i32> [[A0:%.*]], <8 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> [[_MSPROP]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.permd(<8 x i32> %a0, <8 x i32> %a1) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.permd(<8 x i32>, <8 x i32>) nounwind readonly
+
+
+define <8 x float> @test_x86_avx2_permps(<8 x float> %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_permps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx2.permps(<8 x float> [[A0:%.*]], <8 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx2.permps(<8 x float> %a0, <8 x i32> %a1) ; <<8 x float>> [#uses=1]
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx2.permps(<8 x float>, <8 x i32>) nounwind readonly
+
+
+define <2 x i64> @test_x86_avx2_maskload_q(i8* %a0, <2 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_maskload_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.avx2.maskload.q(i8* [[A0:%.*]], <2 x i64> [[A1:%.*]])
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.avx2.maskload.q(i8* %a0, <2 x i64> %a1) ; <<2 x i64>> [#uses=1]
+ ret <2 x i64> %res
+}
+declare <2 x i64> @llvm.x86.avx2.maskload.q(i8*, <2 x i64>) nounwind readonly
+
+
+define <4 x i64> @test_x86_avx2_maskload_q_256(i8* %a0, <4 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_maskload_q_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i64> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i64> @llvm.x86.avx2.maskload.q.256(i8* [[A0:%.*]], <4 x i64> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i64> zeroinitializer, <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x i64> [[RES]]
+;
+ %res = call <4 x i64> @llvm.x86.avx2.maskload.q.256(i8* %a0, <4 x i64> %a1) ; <<4 x i64>> [#uses=1]
+ ret <4 x i64> %res
+}
+declare <4 x i64> @llvm.x86.avx2.maskload.q.256(i8*, <4 x i64>) nounwind readonly
+
+
+define <4 x i32> @test_x86_avx2_maskload_d(i8* %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_maskload_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.avx2.maskload.d(i8* [[A0:%.*]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.avx2.maskload.d(i8* %a0, <4 x i32> %a1) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.avx2.maskload.d(i8*, <4 x i32>) nounwind readonly
+
+
+define <8 x i32> @test_x86_avx2_maskload_d_256(i8* %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_maskload_d_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.maskload.d.256(i8* [[A0:%.*]], <8 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.maskload.d.256(i8* %a0, <8 x i32> %a1) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.maskload.d.256(i8*, <8 x i32>) nounwind readonly
+
+
+define void @test_x86_avx2_maskstore_q(i8* %a0, <2 x i64> %a1, <2 x i64> %a2) #0 {
+; CHECK-LABEL: @test_x86_avx2_maskstore_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[TMP9:%.*]] = bitcast <2 x i64> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP9]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP10:%.*]], label [[TMP11:%.*]], !prof [[PROF0]]
+; CHECK: 10:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 11:
+; CHECK-NEXT: call void @llvm.x86.avx2.maskstore.q(i8* [[A0:%.*]], <2 x i64> [[A1:%.*]], <2 x i64> [[A2:%.*]])
+; CHECK-NEXT: ret void
+;
+ call void @llvm.x86.avx2.maskstore.q(i8* %a0, <2 x i64> %a1, <2 x i64> %a2)
+ ret void
+}
+declare void @llvm.x86.avx2.maskstore.q(i8*, <2 x i64>, <2 x i64>) nounwind
+
+
+define void @test_x86_avx2_maskstore_q_256(i8* %a0, <4 x i64> %a1, <4 x i64> %a2) #0 {
+; CHECK-LABEL: @test_x86_avx2_maskstore_q_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i64> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[TMP9:%.*]] = bitcast <4 x i64> [[TMP3]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP9]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP10:%.*]], label [[TMP11:%.*]], !prof [[PROF0]]
+; CHECK: 10:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 11:
+; CHECK-NEXT: call void @llvm.x86.avx2.maskstore.q.256(i8* [[A0:%.*]], <4 x i64> [[A1:%.*]], <4 x i64> [[A2:%.*]])
+; CHECK-NEXT: ret void
+;
+ call void @llvm.x86.avx2.maskstore.q.256(i8* %a0, <4 x i64> %a1, <4 x i64> %a2)
+ ret void
+}
+declare void @llvm.x86.avx2.maskstore.q.256(i8*, <4 x i64>, <4 x i64>) nounwind
+
+
+define void @test_x86_avx2_maskstore_d(i8* %a0, <4 x i32> %a1, <4 x i32> %a2) #0 {
+; CHECK-LABEL: @test_x86_avx2_maskstore_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[TMP9:%.*]] = bitcast <4 x i32> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP9]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP10:%.*]], label [[TMP11:%.*]], !prof [[PROF0]]
+; CHECK: 10:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 11:
+; CHECK-NEXT: call void @llvm.x86.avx2.maskstore.d(i8* [[A0:%.*]], <4 x i32> [[A1:%.*]], <4 x i32> [[A2:%.*]])
+; CHECK-NEXT: ret void
+;
+ call void @llvm.x86.avx2.maskstore.d(i8* %a0, <4 x i32> %a1, <4 x i32> %a2)
+ ret void
+}
+declare void @llvm.x86.avx2.maskstore.d(i8*, <4 x i32>, <4 x i32>) nounwind
+
+
+define void @test_x86_avx2_maskstore_d_256(i8* %a0, <8 x i32> %a1, <8 x i32> %a2) #0 {
+; CHECK-LABEL: @test_x86_avx2_maskstore_d_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[TMP9:%.*]] = bitcast <8 x i32> [[TMP3]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP9]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP10:%.*]], label [[TMP11:%.*]], !prof [[PROF0]]
+; CHECK: 10:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 11:
+; CHECK-NEXT: call void @llvm.x86.avx2.maskstore.d.256(i8* [[A0:%.*]], <8 x i32> [[A1:%.*]], <8 x i32> [[A2:%.*]])
+; CHECK-NEXT: ret void
+;
+ call void @llvm.x86.avx2.maskstore.d.256(i8* %a0, <8 x i32> %a1, <8 x i32> %a2)
+ ret void
+}
+declare void @llvm.x86.avx2.maskstore.d.256(i8*, <8 x i32>, <8 x i32>) nounwind
+
+
+define <4 x i32> @test_x86_avx2_psllv_d(<4 x i32> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psllv_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <4 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <4 x i1> [[TMP3]] to <4 x i32>
+; CHECK-NEXT: [[TMP5:%.*]] = call <4 x i32> @llvm.x86.avx2.psllv.d(<4 x i32> [[TMP1]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: [[TMP6:%.*]] = or <4 x i32> [[TMP5]], [[TMP4]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.avx2.psllv.d(<4 x i32> [[A0:%.*]], <4 x i32> [[A1]])
+; CHECK-NEXT: store <4 x i32> [[TMP6]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.avx2.psllv.d(<4 x i32> %a0, <4 x i32> %a1) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+
+define <4 x i32> @test_x86_avx2_psllv_d_const() #0 {
+; CHECK-LABEL: @test_x86_avx2_psllv_d_const(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = call <4 x i32> @llvm.x86.avx2.psllv.d(<4 x i32> zeroinitializer, <4 x i32> <i32 1, i32 0, i32 33, i32 -1>)
+; CHECK-NEXT: [[TMP2:%.*]] = or <4 x i32> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[RES0:%.*]] = call <4 x i32> @llvm.x86.avx2.psllv.d(<4 x i32> <i32 2, i32 9, i32 0, i32 -1>, <4 x i32> <i32 1, i32 0, i32 33, i32 -1>)
+; CHECK-NEXT: [[TMP3:%.*]] = call <4 x i32> @llvm.x86.avx2.psllv.d(<4 x i32> zeroinitializer, <4 x i32> <i32 1, i32 1, i32 1, i32 -1>)
+; CHECK-NEXT: [[TMP4:%.*]] = or <4 x i32> [[TMP3]], zeroinitializer
+; CHECK-NEXT: [[RES1:%.*]] = call <4 x i32> @llvm.x86.avx2.psllv.d(<4 x i32> <i32 1, i32 1, i32 1, i32 -1>, <4 x i32> <i32 1, i32 1, i32 1, i32 -1>)
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i32> [[TMP2]], [[TMP4]]
+; CHECK-NEXT: [[RES2:%.*]] = add <4 x i32> [[RES0]], [[RES1]]
+; CHECK-NEXT: store <4 x i32> [[_MSPROP]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES2]]
+;
+ %res0 = call <4 x i32> @llvm.x86.avx2.psllv.d(<4 x i32> <i32 2, i32 9, i32 0, i32 -1>, <4 x i32> <i32 1, i32 0, i32 33, i32 -1>)
+ %res1 = call <4 x i32> @llvm.x86.avx2.psllv.d(<4 x i32> <i32 1, i32 1, i32 1, i32 -1>, <4 x i32> <i32 1, i32 1, i32 1, i32 -1>)
+ %res2 = add <4 x i32> %res0, %res1
+ ret <4 x i32> %res2
+}
+declare <4 x i32> @llvm.x86.avx2.psllv.d(<4 x i32>, <4 x i32>) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_psllv_d_256(<8 x i32> %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psllv_d_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <8 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <8 x i1> [[TMP3]] to <8 x i32>
+; CHECK-NEXT: [[TMP5:%.*]] = call <8 x i32> @llvm.x86.avx2.psllv.d.256(<8 x i32> [[TMP1]], <8 x i32> [[A1:%.*]])
+; CHECK-NEXT: [[TMP6:%.*]] = or <8 x i32> [[TMP5]], [[TMP4]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.psllv.d.256(<8 x i32> [[A0:%.*]], <8 x i32> [[A1]])
+; CHECK-NEXT: store <8 x i32> [[TMP6]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.psllv.d.256(<8 x i32> %a0, <8 x i32> %a1) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+
+define <8 x i32> @test_x86_avx2_psllv_d_256_const() #0 {
+; CHECK-LABEL: @test_x86_avx2_psllv_d_256_const(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = call <8 x i32> @llvm.x86.avx2.psllv.d.256(<8 x i32> zeroinitializer, <8 x i32> <i32 1, i32 0, i32 33, i32 -1, i32 2, i32 0, i32 34, i32 -2>)
+; CHECK-NEXT: [[TMP2:%.*]] = or <8 x i32> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[RES0:%.*]] = call <8 x i32> @llvm.x86.avx2.psllv.d.256(<8 x i32> <i32 2, i32 9, i32 0, i32 -1, i32 3, i32 7, i32 -1, i32 0>, <8 x i32> <i32 1, i32 0, i32 33, i32 -1, i32 2, i32 0, i32 34, i32 -2>)
+; CHECK-NEXT: [[TMP3:%.*]] = call <8 x i32> @llvm.x86.avx2.psllv.d.256(<8 x i32> zeroinitializer, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 -1>)
+; CHECK-NEXT: [[TMP4:%.*]] = or <8 x i32> [[TMP3]], zeroinitializer
+; CHECK-NEXT: [[RES1:%.*]] = call <8 x i32> @llvm.x86.avx2.psllv.d.256(<8 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 -1>, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 -1>)
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i32> [[TMP2]], [[TMP4]]
+; CHECK-NEXT: [[RES2:%.*]] = add <8 x i32> [[RES0]], [[RES1]]
+; CHECK-NEXT: store <8 x i32> [[_MSPROP]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES2]]
+;
+ %res0 = call <8 x i32> @llvm.x86.avx2.psllv.d.256(<8 x i32> <i32 2, i32 9, i32 0, i32 -1, i32 3, i32 7, i32 -1, i32 0>, <8 x i32> <i32 1, i32 0, i32 33, i32 -1,i32 2, i32 0, i32 34, i32 -2>)
+ %res1 = call <8 x i32> @llvm.x86.avx2.psllv.d.256(<8 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 -1>, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 -1>)
+ %res2 = add <8 x i32> %res0, %res1
+ ret <8 x i32> %res2
+}
+declare <8 x i32> @llvm.x86.avx2.psllv.d.256(<8 x i32>, <8 x i32>) nounwind readnone
+
+
+define <2 x i64> @test_x86_avx2_psllv_q(<2 x i64> %a0, <2 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psllv_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <2 x i64> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <2 x i1> [[TMP3]] to <2 x i64>
+; CHECK-NEXT: [[TMP5:%.*]] = call <2 x i64> @llvm.x86.avx2.psllv.q(<2 x i64> [[TMP1]], <2 x i64> [[A1:%.*]])
+; CHECK-NEXT: [[TMP6:%.*]] = or <2 x i64> [[TMP5]], [[TMP4]]
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.avx2.psllv.q(<2 x i64> [[A0:%.*]], <2 x i64> [[A1]])
+; CHECK-NEXT: store <2 x i64> [[TMP6]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.avx2.psllv.q(<2 x i64> %a0, <2 x i64> %a1) ; <<2 x i64>> [#uses=1]
+ ret <2 x i64> %res
+}
+define <2 x i64> @test_x86_avx2_psllv_q_const() #0 {
+; CHECK-LABEL: @test_x86_avx2_psllv_q_const(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = call <2 x i64> @llvm.x86.avx2.psllv.q(<2 x i64> zeroinitializer, <2 x i64> <i64 1, i64 -1>)
+; CHECK-NEXT: [[TMP2:%.*]] = or <2 x i64> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.avx2.psllv.q(<2 x i64> <i64 4, i64 -1>, <2 x i64> <i64 1, i64 -1>)
+; CHECK-NEXT: store <2 x i64> [[TMP2]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.avx2.psllv.q(<2 x i64> <i64 4, i64 -1>, <2 x i64> <i64 1, i64 -1>)
+ ret <2 x i64> %res
+}
+declare <2 x i64> @llvm.x86.avx2.psllv.q(<2 x i64>, <2 x i64>) nounwind readnone
+
+
+define <4 x i64> @test_x86_avx2_psllv_q_256(<4 x i64> %a0, <4 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psllv_q_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <4 x i64> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <4 x i1> [[TMP3]] to <4 x i64>
+; CHECK-NEXT: [[TMP5:%.*]] = call <4 x i64> @llvm.x86.avx2.psllv.q.256(<4 x i64> [[TMP1]], <4 x i64> [[A1:%.*]])
+; CHECK-NEXT: [[TMP6:%.*]] = or <4 x i64> [[TMP5]], [[TMP4]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i64> @llvm.x86.avx2.psllv.q.256(<4 x i64> [[A0:%.*]], <4 x i64> [[A1]])
+; CHECK-NEXT: store <4 x i64> [[TMP6]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x i64> [[RES]]
+;
+ %res = call <4 x i64> @llvm.x86.avx2.psllv.q.256(<4 x i64> %a0, <4 x i64> %a1) ; <<4 x i64>> [#uses=1]
+ ret <4 x i64> %res
+}
+
+define <4 x i64> @test_x86_avx2_psllv_q_256_const() #0 {
+; CHECK-LABEL: @test_x86_avx2_psllv_q_256_const(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = call <4 x i64> @llvm.x86.avx2.psllv.q.256(<4 x i64> zeroinitializer, <4 x i64> <i64 1, i64 1, i64 1, i64 -1>)
+; CHECK-NEXT: [[TMP2:%.*]] = or <4 x i64> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i64> @llvm.x86.avx2.psllv.q.256(<4 x i64> <i64 4, i64 4, i64 4, i64 -1>, <4 x i64> <i64 1, i64 1, i64 1, i64 -1>)
+; CHECK-NEXT: store <4 x i64> [[TMP2]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x i64> [[RES]]
+;
+ %res = call <4 x i64> @llvm.x86.avx2.psllv.q.256(<4 x i64> <i64 4, i64 4, i64 4, i64 -1>, <4 x i64> <i64 1, i64 1, i64 1, i64 -1>)
+ ret <4 x i64> %res
+}
+declare <4 x i64> @llvm.x86.avx2.psllv.q.256(<4 x i64>, <4 x i64>) nounwind readnone
+
+
+define <4 x i32> @test_x86_avx2_psrlv_d(<4 x i32> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrlv_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <4 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <4 x i1> [[TMP3]] to <4 x i32>
+; CHECK-NEXT: [[TMP5:%.*]] = call <4 x i32> @llvm.x86.avx2.psrlv.d(<4 x i32> [[TMP1]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: [[TMP6:%.*]] = or <4 x i32> [[TMP5]], [[TMP4]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.avx2.psrlv.d(<4 x i32> [[A0:%.*]], <4 x i32> [[A1]])
+; CHECK-NEXT: store <4 x i32> [[TMP6]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.avx2.psrlv.d(<4 x i32> %a0, <4 x i32> %a1) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+
+define <4 x i32> @test_x86_avx2_psrlv_d_const() #0 {
+; CHECK-LABEL: @test_x86_avx2_psrlv_d_const(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = call <4 x i32> @llvm.x86.avx2.psrlv.d(<4 x i32> zeroinitializer, <4 x i32> <i32 1, i32 0, i32 33, i32 -1>)
+; CHECK-NEXT: [[TMP2:%.*]] = or <4 x i32> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[RES0:%.*]] = call <4 x i32> @llvm.x86.avx2.psrlv.d(<4 x i32> <i32 2, i32 9, i32 0, i32 -1>, <4 x i32> <i32 1, i32 0, i32 33, i32 -1>)
+; CHECK-NEXT: [[TMP3:%.*]] = call <4 x i32> @llvm.x86.avx2.psrlv.d(<4 x i32> zeroinitializer, <4 x i32> <i32 1, i32 1, i32 1, i32 -1>)
+; CHECK-NEXT: [[TMP4:%.*]] = or <4 x i32> [[TMP3]], zeroinitializer
+; CHECK-NEXT: [[RES1:%.*]] = call <4 x i32> @llvm.x86.avx2.psrlv.d(<4 x i32> <i32 4, i32 4, i32 4, i32 -1>, <4 x i32> <i32 1, i32 1, i32 1, i32 -1>)
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i32> [[TMP2]], [[TMP4]]
+; CHECK-NEXT: [[RES2:%.*]] = add <4 x i32> [[RES0]], [[RES1]]
+; CHECK-NEXT: store <4 x i32> [[_MSPROP]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES2]]
+;
+ %res0 = call <4 x i32> @llvm.x86.avx2.psrlv.d(<4 x i32> <i32 2, i32 9, i32 0, i32 -1>, <4 x i32> <i32 1, i32 0, i32 33, i32 -1>)
+ %res1 = call <4 x i32> @llvm.x86.avx2.psrlv.d(<4 x i32> <i32 4, i32 4, i32 4, i32 -1>, <4 x i32> <i32 1, i32 1, i32 1, i32 -1>)
+ %res2 = add <4 x i32> %res0, %res1
+ ret <4 x i32> %res2
+}
+declare <4 x i32> @llvm.x86.avx2.psrlv.d(<4 x i32>, <4 x i32>) nounwind readnone
+
+
+define <8 x i32> @test_x86_avx2_psrlv_d_256(<8 x i32> %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrlv_d_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <8 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <8 x i1> [[TMP3]] to <8 x i32>
+; CHECK-NEXT: [[TMP5:%.*]] = call <8 x i32> @llvm.x86.avx2.psrlv.d.256(<8 x i32> [[TMP1]], <8 x i32> [[A1:%.*]])
+; CHECK-NEXT: [[TMP6:%.*]] = or <8 x i32> [[TMP5]], [[TMP4]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.psrlv.d.256(<8 x i32> [[A0:%.*]], <8 x i32> [[A1]])
+; CHECK-NEXT: store <8 x i32> [[TMP6]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.psrlv.d.256(<8 x i32> %a0, <8 x i32> %a1) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+
+define <8 x i32> @test_x86_avx2_psrlv_d_256_const() #0 {
+; CHECK-LABEL: @test_x86_avx2_psrlv_d_256_const(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = call <8 x i32> @llvm.x86.avx2.psrlv.d.256(<8 x i32> zeroinitializer, <8 x i32> <i32 1, i32 0, i32 33, i32 -1, i32 2, i32 0, i32 34, i32 -2>)
+; CHECK-NEXT: [[TMP2:%.*]] = or <8 x i32> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[RES0:%.*]] = call <8 x i32> @llvm.x86.avx2.psrlv.d.256(<8 x i32> <i32 2, i32 9, i32 0, i32 -1, i32 3, i32 7, i32 -1, i32 0>, <8 x i32> <i32 1, i32 0, i32 33, i32 -1, i32 2, i32 0, i32 34, i32 -2>)
+; CHECK-NEXT: [[TMP3:%.*]] = call <8 x i32> @llvm.x86.avx2.psrlv.d.256(<8 x i32> zeroinitializer, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 -1>)
+; CHECK-NEXT: [[TMP4:%.*]] = or <8 x i32> [[TMP3]], zeroinitializer
+; CHECK-NEXT: [[RES1:%.*]] = call <8 x i32> @llvm.x86.avx2.psrlv.d.256(<8 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 -1>, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 -1>)
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i32> [[TMP2]], [[TMP4]]
+; CHECK-NEXT: [[RES2:%.*]] = add <8 x i32> [[RES0]], [[RES1]]
+; CHECK-NEXT: store <8 x i32> [[_MSPROP]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES2]]
+;
+ %res0 = call <8 x i32> @llvm.x86.avx2.psrlv.d.256(<8 x i32> <i32 2, i32 9, i32 0, i32 -1, i32 3, i32 7, i32 -1, i32 0>, <8 x i32> <i32 1, i32 0, i32 33, i32 -1,i32 2, i32 0, i32 34, i32 -2>)
+ %res1 = call <8 x i32> @llvm.x86.avx2.psrlv.d.256(<8 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 -1>, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 -1>)
+ %res2 = add <8 x i32> %res0, %res1
+ ret <8 x i32> %res2
+}
+declare <8 x i32> @llvm.x86.avx2.psrlv.d.256(<8 x i32>, <8 x i32>) nounwind readnone
+
+
+define <2 x i64> @test_x86_avx2_psrlv_q(<2 x i64> %a0, <2 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrlv_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <2 x i64> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <2 x i1> [[TMP3]] to <2 x i64>
+; CHECK-NEXT: [[TMP5:%.*]] = call <2 x i64> @llvm.x86.avx2.psrlv.q(<2 x i64> [[TMP1]], <2 x i64> [[A1:%.*]])
+; CHECK-NEXT: [[TMP6:%.*]] = or <2 x i64> [[TMP5]], [[TMP4]]
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.avx2.psrlv.q(<2 x i64> [[A0:%.*]], <2 x i64> [[A1]])
+; CHECK-NEXT: store <2 x i64> [[TMP6]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.avx2.psrlv.q(<2 x i64> %a0, <2 x i64> %a1) ; <<2 x i64>> [#uses=1]
+ ret <2 x i64> %res
+}
+
+define <2 x i64> @test_x86_avx2_psrlv_q_const() #0 {
+; CHECK-LABEL: @test_x86_avx2_psrlv_q_const(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = call <2 x i64> @llvm.x86.avx2.psrlv.q(<2 x i64> zeroinitializer, <2 x i64> <i64 1, i64 -1>)
+; CHECK-NEXT: [[TMP2:%.*]] = or <2 x i64> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.avx2.psrlv.q(<2 x i64> <i64 4, i64 4>, <2 x i64> <i64 1, i64 -1>)
+; CHECK-NEXT: store <2 x i64> [[TMP2]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.avx2.psrlv.q(<2 x i64> <i64 4, i64 4>, <2 x i64> <i64 1, i64 -1>)
+ ret <2 x i64> %res
+}
+declare <2 x i64> @llvm.x86.avx2.psrlv.q(<2 x i64>, <2 x i64>) nounwind readnone
+
+
+define <4 x i64> @test_x86_avx2_psrlv_q_256(<4 x i64> %a0, <4 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrlv_q_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <4 x i64> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <4 x i1> [[TMP3]] to <4 x i64>
+; CHECK-NEXT: [[TMP5:%.*]] = call <4 x i64> @llvm.x86.avx2.psrlv.q.256(<4 x i64> [[TMP1]], <4 x i64> [[A1:%.*]])
+; CHECK-NEXT: [[TMP6:%.*]] = or <4 x i64> [[TMP5]], [[TMP4]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i64> @llvm.x86.avx2.psrlv.q.256(<4 x i64> [[A0:%.*]], <4 x i64> [[A1]])
+; CHECK-NEXT: store <4 x i64> [[TMP6]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x i64> [[RES]]
+;
+ %res = call <4 x i64> @llvm.x86.avx2.psrlv.q.256(<4 x i64> %a0, <4 x i64> %a1) ; <<4 x i64>> [#uses=1]
+ ret <4 x i64> %res
+}
+
+
+define <4 x i64> @test_x86_avx2_psrlv_q_256_const() #0 {
+; CHECK-LABEL: @test_x86_avx2_psrlv_q_256_const(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = call <4 x i64> @llvm.x86.avx2.psrlv.q.256(<4 x i64> zeroinitializer, <4 x i64> <i64 1, i64 1, i64 1, i64 -1>)
+; CHECK-NEXT: [[TMP2:%.*]] = or <4 x i64> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i64> @llvm.x86.avx2.psrlv.q.256(<4 x i64> <i64 4, i64 4, i64 4, i64 4>, <4 x i64> <i64 1, i64 1, i64 1, i64 -1>)
+; CHECK-NEXT: store <4 x i64> [[TMP2]], <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x i64> [[RES]]
+;
+ %res = call <4 x i64> @llvm.x86.avx2.psrlv.q.256(<4 x i64> <i64 4, i64 4, i64 4, i64 4>, <4 x i64> <i64 1, i64 1, i64 1, i64 -1>)
+ ret <4 x i64> %res
+}
+declare <4 x i64> @llvm.x86.avx2.psrlv.q.256(<4 x i64>, <4 x i64>) nounwind readnone
+
+
+define <4 x i32> @test_x86_avx2_psrav_d(<4 x i32> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrav_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <4 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <4 x i1> [[TMP3]] to <4 x i32>
+; CHECK-NEXT: [[TMP5:%.*]] = call <4 x i32> @llvm.x86.avx2.psrav.d(<4 x i32> [[TMP1]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: [[TMP6:%.*]] = or <4 x i32> [[TMP5]], [[TMP4]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.avx2.psrav.d(<4 x i32> [[A0:%.*]], <4 x i32> [[A1]])
+; CHECK-NEXT: store <4 x i32> [[TMP6]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.avx2.psrav.d(<4 x i32> %a0, <4 x i32> %a1) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+
+define <4 x i32> @test_x86_avx2_psrav_d_const() #0 {
+; CHECK-LABEL: @test_x86_avx2_psrav_d_const(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = call <4 x i32> @llvm.x86.avx2.psrav.d(<4 x i32> zeroinitializer, <4 x i32> <i32 1, i32 18, i32 35, i32 52>)
+; CHECK-NEXT: [[TMP2:%.*]] = or <4 x i32> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.avx2.psrav.d(<4 x i32> <i32 2, i32 9, i32 -12, i32 23>, <4 x i32> <i32 1, i32 18, i32 35, i32 52>)
+; CHECK-NEXT: store <4 x i32> [[TMP2]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.avx2.psrav.d(<4 x i32> <i32 2, i32 9, i32 -12, i32 23>, <4 x i32> <i32 1, i32 18, i32 35, i32 52>)
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.avx2.psrav.d(<4 x i32>, <4 x i32>) nounwind readnone
+
+define <8 x i32> @test_x86_avx2_psrav_d_256(<8 x i32> %a0, <8 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_avx2_psrav_d_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <8 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <8 x i1> [[TMP3]] to <8 x i32>
+; CHECK-NEXT: [[TMP5:%.*]] = call <8 x i32> @llvm.x86.avx2.psrav.d.256(<8 x i32> [[TMP1]], <8 x i32> [[A1:%.*]])
+; CHECK-NEXT: [[TMP6:%.*]] = or <8 x i32> [[TMP5]], [[TMP4]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.psrav.d.256(<8 x i32> [[A0:%.*]], <8 x i32> [[A1]])
+; CHECK-NEXT: store <8 x i32> [[TMP6]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.psrav.d.256(<8 x i32> %a0, <8 x i32> %a1) ; <<8 x i32>> [#uses=1]
+ ret <8 x i32> %res
+}
+
+define <8 x i32> @test_x86_avx2_psrav_d_256_const() #0 {
+; CHECK-LABEL: @test_x86_avx2_psrav_d_256_const(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP1:%.*]] = call <8 x i32> @llvm.x86.avx2.psrav.d.256(<8 x i32> zeroinitializer, <8 x i32> <i32 1, i32 18, i32 35, i32 52, i32 69, i32 15, i32 32, i32 49>)
+; CHECK-NEXT: [[TMP2:%.*]] = or <8 x i32> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.psrav.d.256(<8 x i32> <i32 2, i32 9, i32 -12, i32 23, i32 -26, i32 37, i32 -40, i32 51>, <8 x i32> <i32 1, i32 18, i32 35, i32 52, i32 69, i32 15, i32 32, i32 49>)
+; CHECK-NEXT: store <8 x i32> [[TMP2]], <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.psrav.d.256(<8 x i32> <i32 2, i32 9, i32 -12, i32 23, i32 -26, i32 37, i32 -40, i32 51>, <8 x i32> <i32 1, i32 18, i32 35, i32 52, i32 69, i32 15, i32 32, i32 49>)
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.psrav.d.256(<8 x i32>, <8 x i32>) nounwind readnone
+
+define <2 x double> @test_x86_avx2_gather_d_pd(<2 x double> %a0, i8* %a1, <4 x i32> %idx, <2 x double> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_d_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <4 x i32> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <2 x i64> [[TMP4]] to i128
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i128 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.avx2.gather.d.pd(<2 x double> [[A0:%.*]], i8* [[A1:%.*]], <4 x i32> [[IDX:%.*]], <2 x double> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.avx2.gather.d.pd(<2 x double> %a0,
+ i8* %a1, <4 x i32> %idx, <2 x double> %mask, i8 2) ;
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.avx2.gather.d.pd(<2 x double>, i8*,
+ <4 x i32>, <2 x double>, i8) nounwind readonly
+
+define <4 x double> @test_x86_avx2_gather_d_pd_256(<4 x double> %a0, i8* %a1, <4 x i32> %idx, <4 x double> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_d_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 56) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <4 x i32> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <4 x i64> [[TMP4]] to i256
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i256 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx2.gather.d.pd.256(<4 x double> [[A0:%.*]], i8* [[A1:%.*]], <4 x i32> [[IDX:%.*]], <4 x double> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <4 x i64> zeroinitializer, <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx2.gather.d.pd.256(<4 x double> %a0,
+ i8* %a1, <4 x i32> %idx, <4 x double> %mask, i8 2) ;
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx2.gather.d.pd.256(<4 x double>, i8*,
+ <4 x i32>, <4 x double>, i8) nounwind readonly
+
+define <2 x double> @test_x86_avx2_gather_q_pd(<2 x double> %a0, i8* %a1, <2 x i64> %idx, <2 x double> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_q_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <2 x i64> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <2 x i64> [[TMP4]] to i128
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i128 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.avx2.gather.q.pd(<2 x double> [[A0:%.*]], i8* [[A1:%.*]], <2 x i64> [[IDX:%.*]], <2 x double> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.avx2.gather.q.pd(<2 x double> %a0,
+ i8* %a1, <2 x i64> %idx, <2 x double> %mask, i8 2) ;
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.avx2.gather.q.pd(<2 x double>, i8*,
+ <2 x i64>, <2 x double>, i8) nounwind readonly
+
+define <4 x double> @test_x86_avx2_gather_q_pd_256(<4 x double> %a0, i8* %a1, <4 x i64> %idx, <4 x double> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_q_pd_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 72) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <4 x i64> [[TMP3]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <4 x i64> [[TMP4]] to i256
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i256 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx2.gather.q.pd.256(<4 x double> [[A0:%.*]], i8* [[A1:%.*]], <4 x i64> [[IDX:%.*]], <4 x double> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <4 x i64> zeroinitializer, <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x double> [[RES]]
+;
+ %res = call <4 x double> @llvm.x86.avx2.gather.q.pd.256(<4 x double> %a0,
+ i8* %a1, <4 x i64> %idx, <4 x double> %mask, i8 2) ;
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx2.gather.q.pd.256(<4 x double>, i8*,
+ <4 x i64>, <4 x double>, i8) nounwind readonly
+
+define <4 x float> @test_x86_avx2_gather_d_ps(<4 x float> %a0, i8* %a1, <4 x i32> %idx, <4 x float> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_d_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <4 x i32> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <4 x i32> [[TMP4]] to i128
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i128 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.avx2.gather.d.ps(<4 x float> [[A0:%.*]], i8* [[A1:%.*]], <4 x i32> [[IDX:%.*]], <4 x float> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.avx2.gather.d.ps(<4 x float> %a0,
+ i8* %a1, <4 x i32> %idx, <4 x float> %mask, i8 2) ;
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx2.gather.d.ps(<4 x float>, i8*,
+ <4 x i32>, <4 x float>, i8) nounwind readonly
+
+define <8 x float> @test_x86_avx2_gather_d_ps_256(<8 x float> %a0, i8* %a1, <8 x i32> %idx, <8 x float> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_d_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 72) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <8 x i32> [[TMP3]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <8 x i32> [[TMP4]] to i256
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i256 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx2.gather.d.ps.256(<8 x float> [[A0:%.*]], i8* [[A1:%.*]], <8 x i32> [[IDX:%.*]], <8 x float> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %res = call <8 x float> @llvm.x86.avx2.gather.d.ps.256(<8 x float> %a0,
+ i8* %a1, <8 x i32> %idx, <8 x float> %mask, i8 2) ;
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx2.gather.d.ps.256(<8 x float>, i8*,
+ <8 x i32>, <8 x float>, i8) nounwind readonly
+
+define <4 x float> @test_x86_avx2_gather_q_ps(<4 x float> %a0, i8* %a1, <2 x i64> %idx, <4 x float> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_q_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <2 x i64> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <4 x i32> [[TMP4]] to i128
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i128 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.avx2.gather.q.ps(<4 x float> [[A0:%.*]], i8* [[A1:%.*]], <2 x i64> [[IDX:%.*]], <4 x float> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.avx2.gather.q.ps(<4 x float> %a0,
+ i8* %a1, <2 x i64> %idx, <4 x float> %mask, i8 2) ;
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx2.gather.q.ps(<4 x float>, i8*,
+ <2 x i64>, <4 x float>, i8) nounwind readonly
+
+define <4 x float> @test_x86_avx2_gather_q_ps_256(<4 x float> %a0, i8* %a1, <4 x i64> %idx, <4 x float> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_q_ps_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 56) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <4 x i64> [[TMP3]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <4 x i32> [[TMP4]] to i128
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i128 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.avx2.gather.q.ps.256(<4 x float> [[A0:%.*]], i8* [[A1:%.*]], <4 x i64> [[IDX:%.*]], <4 x float> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.avx2.gather.q.ps.256(<4 x float> %a0,
+ i8* %a1, <4 x i64> %idx, <4 x float> %mask, i8 2) ;
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx2.gather.q.ps.256(<4 x float>, i8*,
+ <4 x i64>, <4 x float>, i8) nounwind readonly
+
+define <2 x i64> @test_x86_avx2_gather_d_q(<2 x i64> %a0, i8* %a1, <4 x i32> %idx, <2 x i64> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_d_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <4 x i32> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <2 x i64> [[TMP4]] to i128
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i128 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.avx2.gather.d.q(<2 x i64> [[A0:%.*]], i8* [[A1:%.*]], <4 x i32> [[IDX:%.*]], <2 x i64> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.avx2.gather.d.q(<2 x i64> %a0,
+ i8* %a1, <4 x i32> %idx, <2 x i64> %mask, i8 2) ;
+ ret <2 x i64> %res
+}
+declare <2 x i64> @llvm.x86.avx2.gather.d.q(<2 x i64>, i8*,
+ <4 x i32>, <2 x i64>, i8) nounwind readonly
+
+define <4 x i64> @test_x86_avx2_gather_d_q_256(<4 x i64> %a0, i8* %a1, <4 x i32> %idx, <4 x i64> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_d_q_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 56) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <4 x i32> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <4 x i64> [[TMP4]] to i256
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i256 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i64> @llvm.x86.avx2.gather.d.q.256(<4 x i64> [[A0:%.*]], i8* [[A1:%.*]], <4 x i32> [[IDX:%.*]], <4 x i64> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <4 x i64> zeroinitializer, <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x i64> [[RES]]
+;
+ %res = call <4 x i64> @llvm.x86.avx2.gather.d.q.256(<4 x i64> %a0,
+ i8* %a1, <4 x i32> %idx, <4 x i64> %mask, i8 2) ;
+ ret <4 x i64> %res
+}
+declare <4 x i64> @llvm.x86.avx2.gather.d.q.256(<4 x i64>, i8*,
+ <4 x i32>, <4 x i64>, i8) nounwind readonly
+
+define <2 x i64> @test_x86_avx2_gather_q_q(<2 x i64> %a0, i8* %a1, <2 x i64> %idx, <2 x i64> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_q_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <2 x i64> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <2 x i64> [[TMP4]] to i128
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i128 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.avx2.gather.q.q(<2 x i64> [[A0:%.*]], i8* [[A1:%.*]], <2 x i64> [[IDX:%.*]], <2 x i64> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.avx2.gather.q.q(<2 x i64> %a0,
+ i8* %a1, <2 x i64> %idx, <2 x i64> %mask, i8 2) ;
+ ret <2 x i64> %res
+}
+declare <2 x i64> @llvm.x86.avx2.gather.q.q(<2 x i64>, i8*,
+ <2 x i64>, <2 x i64>, i8) nounwind readonly
+
+define <4 x i64> @test_x86_avx2_gather_q_q_256(<4 x i64> %a0, i8* %a1, <4 x i64> %idx, <4 x i64> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_q_q_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, <4 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 72) to <4 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i64> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <4 x i64> [[TMP3]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <4 x i64> [[TMP4]] to i256
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i256 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i64> @llvm.x86.avx2.gather.q.q.256(<4 x i64> [[A0:%.*]], i8* [[A1:%.*]], <4 x i64> [[IDX:%.*]], <4 x i64> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <4 x i64> zeroinitializer, <4 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i64>*), align 8
+; CHECK-NEXT: ret <4 x i64> [[RES]]
+;
+ %res = call <4 x i64> @llvm.x86.avx2.gather.q.q.256(<4 x i64> %a0,
+ i8* %a1, <4 x i64> %idx, <4 x i64> %mask, i8 2) ;
+ ret <4 x i64> %res
+}
+declare <4 x i64> @llvm.x86.avx2.gather.q.q.256(<4 x i64>, i8*,
+ <4 x i64>, <4 x i64>, i8) nounwind readonly
+
+define <4 x i32> @test_x86_avx2_gather_d_d(<4 x i32> %a0, i8* %a1, <4 x i32> %idx, <4 x i32> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_d_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <4 x i32> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <4 x i32> [[TMP4]] to i128
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i128 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.avx2.gather.d.d(<4 x i32> [[A0:%.*]], i8* [[A1:%.*]], <4 x i32> [[IDX:%.*]], <4 x i32> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.avx2.gather.d.d(<4 x i32> %a0,
+ i8* %a1, <4 x i32> %idx, <4 x i32> %mask, i8 2) ;
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.avx2.gather.d.d(<4 x i32>, i8*,
+ <4 x i32>, <4 x i32>, i8) nounwind readonly
+
+define <8 x i32> @test_x86_avx2_gather_d_d_256(<8 x i32> %a0, i8* %a1, <8 x i32> %idx, <8 x i32> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_d_d_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 72) to <8 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <8 x i32> [[TMP1]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <8 x i32> [[TMP3]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <8 x i32> [[TMP4]] to i256
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i256 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i32> @llvm.x86.avx2.gather.d.d.256(<8 x i32> [[A0:%.*]], i8* [[A1:%.*]], <8 x i32> [[IDX:%.*]], <8 x i32> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x i32> [[RES]]
+;
+ %res = call <8 x i32> @llvm.x86.avx2.gather.d.d.256(<8 x i32> %a0,
+ i8* %a1, <8 x i32> %idx, <8 x i32> %mask, i8 2) ;
+ ret <8 x i32> %res
+}
+declare <8 x i32> @llvm.x86.avx2.gather.d.d.256(<8 x i32>, i8*,
+ <8 x i32>, <8 x i32>, i8) nounwind readonly
+
+define <4 x i32> @test_x86_avx2_gather_q_d(<4 x i32> %a0, i8* %a1, <2 x i64> %idx, <4 x i32> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_q_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <2 x i64> [[TMP3]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <4 x i32> [[TMP4]] to i128
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i128 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.avx2.gather.q.d(<4 x i32> [[A0:%.*]], i8* [[A1:%.*]], <2 x i64> [[IDX:%.*]], <4 x i32> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.avx2.gather.q.d(<4 x i32> %a0,
+ i8* %a1, <2 x i64> %idx, <4 x i32> %mask, i8 2) ;
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.avx2.gather.q.d(<4 x i32>, i8*,
+ <2 x i64>, <4 x i32>, i8) nounwind readonly
+
+define <4 x i32> @test_x86_avx2_gather_q_d_256(<4 x i32> %a0, i8* %a1, <4 x i64> %idx, <4 x i32> %mask) #0 {
+; CHECK-LABEL: @test_x86_avx2_gather_q_d_256(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i64>, <4 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <4 x i64>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 56) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <4 x i64> [[TMP3]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <4 x i32> [[TMP4]] to i128
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i128 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.avx2.gather.q.d.256(<4 x i32> [[A0:%.*]], i8* [[A1:%.*]], <4 x i64> [[IDX:%.*]], <4 x i32> [[MASK:%.*]], i8 2)
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.avx2.gather.q.d.256(<4 x i32> %a0,
+ i8* %a1, <4 x i64> %idx, <4 x i32> %mask, i8 2) ;
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.avx2.gather.q.d.256(<4 x i32>, i8*,
+ <4 x i64>, <4 x i32>, i8) nounwind readonly
+
+define <8 x float> @test_gather_mask(<8 x float> %a0, float* %a, <8 x i32> %idx, <8 x float> %mask, float* nocapture %out) #0 {
+; CHECK-LABEL: @test_gather_mask(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to i64*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i32>, <8 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <8 x i32>, <8 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 72) to <8 x i32>*), align 8
+; CHECK-NEXT: [[TMP5:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 104) to i64*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[A_I8:%.*]] = bitcast float* [[A:%.*]] to i8*
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i32> [[TMP2]] to i256
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i256 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP9:%.*]], label [[TMP10:%.*]], !prof [[PROF0]]
+; CHECK: 9:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 10:
+; CHECK-NEXT: [[TMP11:%.*]] = bitcast <8 x i32> [[TMP3]] to i256
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i256 [[TMP11]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP12:%.*]], label [[TMP13:%.*]], !prof [[PROF0]]
+; CHECK: 12:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 13:
+; CHECK-NEXT: [[TMP14:%.*]] = bitcast <8 x i32> [[TMP4]] to i256
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i256 [[TMP14]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP15:%.*]], label [[TMP16:%.*]], !prof [[PROF0]]
+; CHECK: 15:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 16:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx2.gather.d.ps.256(<8 x float> [[A0:%.*]], i8* [[A_I8]], <8 x i32> [[IDX:%.*]], <8 x float> [[MASK:%.*]], i8 4)
+; CHECK-NEXT: [[OUT_PTR:%.*]] = bitcast float* [[OUT:%.*]] to <8 x float>*
+; CHECK-NEXT: [[_MSCMP4:%.*]] = icmp ne i64 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP4]], label [[TMP17:%.*]], label [[TMP18:%.*]], !prof [[PROF0]]
+; CHECK: 17:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 18:
+; CHECK-NEXT: [[TMP19:%.*]] = ptrtoint <8 x float>* [[OUT_PTR]] to i64
+; CHECK-NEXT: [[TMP20:%.*]] = xor i64 [[TMP19]], 87960930222080
+; CHECK-NEXT: [[TMP21:%.*]] = inttoptr i64 [[TMP20]] to <8 x i32>*
+; CHECK-NEXT: store <8 x i32> [[TMP4]], <8 x i32>* [[TMP21]], align 4
+; CHECK-NEXT: store <8 x float> [[MASK]], <8 x float>* [[OUT_PTR]], align 4
+; CHECK-NEXT: store <8 x i32> zeroinitializer, <8 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i32>*), align 8
+; CHECK-NEXT: ret <8 x float> [[RES]]
+;
+ %a_i8 = bitcast float* %a to i8*
+ %res = call <8 x float> @llvm.x86.avx2.gather.d.ps.256(<8 x float> %a0,
+ i8* %a_i8, <8 x i32> %idx, <8 x float> %mask, i8 4) ;
+
+ %out_ptr = bitcast float * %out to <8 x float> *
+ store <8 x float> %mask, <8 x float> * %out_ptr, align 4
+
+ ret <8 x float> %res
+}
+
+define <2 x i64> @test_mask_demanded_bits(<2 x i64> %a0, i8* %a1, <2 x i64> %idx, <2 x i1> %mask) #0 {
+; CHECK-LABEL: @test_mask_demanded_bits(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i1>, <2 x i1>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 40) to <2 x i1>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP4:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 24) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = sext <2 x i1> [[TMP1]] to <2 x i64>
+; CHECK-NEXT: [[MASK1:%.*]] = sext <2 x i1> [[MASK:%.*]] to <2 x i64>
+; CHECK-NEXT: [[TMP5:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP5]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP6:%.*]], label [[TMP7:%.*]], !prof [[PROF0]]
+; CHECK: 6:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 7:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[TMP10:%.*]] = bitcast <2 x i64> [[TMP4]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP10]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP11:%.*]], label [[TMP12:%.*]], !prof [[PROF0]]
+; CHECK: 11:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 12:
+; CHECK-NEXT: [[TMP13:%.*]] = bitcast <2 x i64> [[_MSPROP]] to i128
+; CHECK-NEXT: [[_MSCMP3:%.*]] = icmp ne i128 [[TMP13]], 0
+; CHECK-NEXT: br i1 [[_MSCMP3]], label [[TMP14:%.*]], label [[TMP15:%.*]], !prof [[PROF0]]
+; CHECK: 14:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 15:
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.avx2.gather.q.q(<2 x i64> [[A0:%.*]], i8* [[A1:%.*]], <2 x i64> [[IDX:%.*]], <2 x i64> [[MASK1]], i8 2)
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %mask1 = sext <2 x i1> %mask to <2 x i64>
+ %res = call <2 x i64> @llvm.x86.avx2.gather.q.q(<2 x i64> %a0,
+ i8* %a1, <2 x i64> %idx, <2 x i64> %mask1, i8 2) ;
+ ret <2 x i64> %res
+}
+
+attributes #0 = { sanitize_memory }
diff --git a/llvm/test/Instrumentation/MemorySanitizer/sse-intrinsics-x86.ll b/llvm/test/Instrumentation/MemorySanitizer/sse-intrinsics-x86.ll
new file mode 100644
index 0000000000000..c3b98628649e7
--- /dev/null
+++ b/llvm/test/Instrumentation/MemorySanitizer/sse-intrinsics-x86.ll
@@ -0,0 +1,521 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt %s -S -passes=msan 2>&1 | FileCheck %s
+
+target datalayout = "e-m:o-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+define <4 x float> @test_x86_sse_cmp_ps(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_cmp_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = icmp ne <4 x i32> [[TMP3]], zeroinitializer
+; CHECK-NEXT: [[TMP5:%.*]] = sext <4 x i1> [[TMP4]] to <4 x i32>
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse.cmp.ps(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <4 x i32> [[TMP5]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse.cmp.ps(<4 x float> %a0, <4 x float> %a1, i8 7) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse.cmp.ps(<4 x float>, <4 x float>, i8) nounwind readnone
+
+
+define <4 x float> @test_x86_sse_cmp_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_cmp_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i128
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i128 [[TMP6]] to <4 x i32>
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse.cmp.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <4 x i32> [[TMP7]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse.cmp.ss(<4 x float> %a0, <4 x float> %a1, i8 7) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse.cmp.ss(<4 x float>, <4 x float>, i8) nounwind readnone
+
+
+define i32 @test_x86_sse_comieq_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_comieq_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.comieq.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.comieq.ss(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.comieq.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_comige_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_comige_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.comige.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.comige.ss(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.comige.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_comigt_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_comigt_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.comigt.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.comigt.ss(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.comigt.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_comile_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_comile_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.comile.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.comile.ss(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.comile.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_comilt_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_comilt_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.comilt.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.comilt.ss(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.comilt.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_comineq_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_comineq_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.comineq.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.comineq.ss(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.comineq.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_cvtss2si(<4 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse_cvtss2si(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = extractelement <4 x i32> [[TMP1]], i32 0
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i32 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0:![0-9]+]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR5:[0-9]+]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.cvtss2si(<4 x float> [[A0:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.cvtss2si(<4 x float> %a0) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.cvtss2si(<4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_cvttss2si(<4 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse_cvttss2si(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = extractelement <4 x i32> [[TMP1]], i32 0
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i32 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR5]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.cvttss2si(<4 x float> [[A0:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.cvttss2si(<4 x float> %a0) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.cvttss2si(<4 x float>) nounwind readnone
+
+
+define void @test_x86_sse_ldmxcsr(i8* %a0) #0 {
+; CHECK-LABEL: @test_x86_sse_ldmxcsr(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = ptrtoint i8* [[A0:%.*]] to i64
+; CHECK-NEXT: [[TMP3:%.*]] = xor i64 [[TMP2]], 87960930222080
+; CHECK-NEXT: [[TMP4:%.*]] = inttoptr i64 [[TMP3]] to i32*
+; CHECK-NEXT: [[_LDMXCSR:%.*]] = load i32, i32* [[TMP4]], align 1
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP5:%.*]], label [[TMP6:%.*]], !prof [[PROF0]]
+; CHECK: 5:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR5]]
+; CHECK-NEXT: unreachable
+; CHECK: 6:
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i32 [[_LDMXCSR]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR5]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: call void @llvm.x86.sse.ldmxcsr(i8* [[A0]])
+; CHECK-NEXT: ret void
+;
+ call void @llvm.x86.sse.ldmxcsr(i8* %a0)
+ ret void
+}
+declare void @llvm.x86.sse.ldmxcsr(i8*) nounwind
+
+
+
+define <4 x float> @test_x86_sse_max_ps(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_max_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse.max.ps(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i32> [[_MSPROP]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse.max.ps(<4 x float> %a0, <4 x float> %a1) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse.max.ps(<4 x float>, <4 x float>) nounwind readnone
+
+
+define <4 x float> @test_x86_sse_max_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_max_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse.max.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i32> [[_MSPROP]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse.max.ss(<4 x float> %a0, <4 x float> %a1) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse.max.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define <4 x float> @test_x86_sse_min_ps(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_min_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse.min.ps(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i32> [[_MSPROP]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse.min.ps(<4 x float> %a0, <4 x float> %a1) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse.min.ps(<4 x float>, <4 x float>) nounwind readnone
+
+
+define <4 x float> @test_x86_sse_min_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_min_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse.min.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i32> [[_MSPROP]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse.min.ss(<4 x float> %a0, <4 x float> %a1) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse.min.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_movmsk_ps(<4 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse_movmsk_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR5]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.movmsk.ps(<4 x float> [[A0:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.movmsk.ps(<4 x float> %a0) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.movmsk.ps(<4 x float>) nounwind readnone
+
+
+
+define <4 x float> @test_x86_sse_rcp_ps(<4 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse_rcp_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse.rcp.ps(<4 x float> [[A0:%.*]])
+; CHECK-NEXT: store <4 x i32> [[TMP1]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse.rcp.ps(<4 x float> %a0) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse.rcp.ps(<4 x float>) nounwind readnone
+
+
+define <4 x float> @test_x86_sse_rcp_ss(<4 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse_rcp_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse.rcp.ss(<4 x float> [[A0:%.*]])
+; CHECK-NEXT: store <4 x i32> [[TMP1]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse.rcp.ss(<4 x float> %a0) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse.rcp.ss(<4 x float>) nounwind readnone
+
+
+define <4 x float> @test_x86_sse_rsqrt_ps(<4 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse_rsqrt_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse.rsqrt.ps(<4 x float> [[A0:%.*]])
+; CHECK-NEXT: store <4 x i32> [[TMP1]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse.rsqrt.ps(<4 x float> %a0) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse.rsqrt.ps(<4 x float>) nounwind readnone
+
+
+define <4 x float> @test_x86_sse_rsqrt_ss(<4 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse_rsqrt_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse.rsqrt.ss(<4 x float> [[A0:%.*]])
+; CHECK-NEXT: store <4 x i32> [[TMP1]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse.rsqrt.ss(<4 x float> %a0) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse.rsqrt.ss(<4 x float>) nounwind readnone
+
+
+define void @test_x86_sse_stmxcsr(i8* %a0) #0 {
+; CHECK-LABEL: @test_x86_sse_stmxcsr(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = ptrtoint i8* [[A0:%.*]] to i64
+; CHECK-NEXT: [[TMP3:%.*]] = xor i64 [[TMP2]], 87960930222080
+; CHECK-NEXT: [[TMP4:%.*]] = inttoptr i64 [[TMP3]] to i32*
+; CHECK-NEXT: store i32 0, i32* [[TMP4]], align 4
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP5:%.*]], label [[TMP6:%.*]], !prof [[PROF0]]
+; CHECK: 5:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR5]]
+; CHECK-NEXT: unreachable
+; CHECK: 6:
+; CHECK-NEXT: call void @llvm.x86.sse.stmxcsr(i8* [[A0]])
+; CHECK-NEXT: ret void
+;
+ call void @llvm.x86.sse.stmxcsr(i8* %a0)
+ ret void
+}
+declare void @llvm.x86.sse.stmxcsr(i8*) nounwind
+
+
+define i32 @test_x86_sse_ucomieq_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_ucomieq_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.ucomieq.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.ucomieq.ss(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.ucomieq.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_ucomige_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_ucomige_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.ucomige.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.ucomige.ss(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.ucomige.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_ucomigt_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_ucomigt_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.ucomigt.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.ucomigt.ss(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.ucomigt.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_ucomile_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_ucomile_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.ucomile.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.ucomile.ss(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.ucomile.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_ucomilt_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_ucomilt_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.ucomilt.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.ucomilt.ss(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.ucomilt.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse_ucomineq_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse_ucomineq_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i32 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse.ucomineq.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse.ucomineq.ss(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse.ucomineq.ss(<4 x float>, <4 x float>) nounwind readnone
+
+
+define void @sfence() nounwind {
+; CHECK-LABEL: @sfence(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: tail call void @llvm.x86.sse.sfence()
+; CHECK-NEXT: ret void
+;
+ tail call void @llvm.x86.sse.sfence()
+ ret void
+}
+declare void @llvm.x86.sse.sfence() nounwind
+
+attributes #0 = { sanitize_memory }
diff --git a/llvm/test/Instrumentation/MemorySanitizer/sse2-intrinsics-x86.ll b/llvm/test/Instrumentation/MemorySanitizer/sse2-intrinsics-x86.ll
new file mode 100644
index 0000000000000..ec03c7bf6a577
--- /dev/null
+++ b/llvm/test/Instrumentation/MemorySanitizer/sse2-intrinsics-x86.ll
@@ -0,0 +1,1381 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt %s -S -passes=msan 2>&1 | FileCheck %s
+
+target datalayout = "e-m:o-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+define <2 x double> @test_x86_sse2_cmp_pd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_cmp_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = icmp ne <2 x i64> [[TMP3]], zeroinitializer
+; CHECK-NEXT: [[TMP5:%.*]] = sext <2 x i1> [[TMP4]] to <2 x i64>
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.sse2.cmp.pd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <2 x i64> [[TMP5]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.sse2.cmp.pd(<2 x double> %a0, <2 x double> %a1, i8 7) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.sse2.cmp.pd(<2 x double>, <2 x double>, i8) nounwind readnone
+
+
+define <2 x double> @test_x86_sse2_cmp_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_cmp_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i128
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i128 [[TMP6]] to <2 x i64>
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <2 x i64> [[TMP7]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double> %a0, <2 x double> %a1, i8 7) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double>, <2 x double>, i8) nounwind readnone
+
+
+define i32 @test_x86_sse2_comieq_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_comieq_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.comieq.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.comieq.sd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.comieq.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_sse2_comige_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_comige_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.comige.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.comige.sd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.comige.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_sse2_comigt_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_comigt_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.comigt.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.comigt.sd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.comigt.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_sse2_comile_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_comile_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.comile.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.comile.sd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.comile.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_sse2_comilt_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_comilt_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.comilt.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.comilt.sd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.comilt.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_sse2_comineq_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_comineq_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.comineq.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.comineq.sd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.comineq.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define <4 x i32> @test_x86_sse2_cvtpd2dq(<2 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_cvtpd2dq(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0:![0-9]+]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6:[0-9]+]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.sse2.cvtpd2dq(<2 x double> [[A0:%.*]])
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.sse2.cvtpd2dq(<2 x double> %a0) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.sse2.cvtpd2dq(<2 x double>) nounwind readnone
+
+
+define <2 x i64> @test_mm_cvtpd_epi32_zext(<2 x double> %a0) nounwind #0 {
+; CHECK-LABEL: @test_mm_cvtpd_epi32_zext(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[CVT:%.*]] = call <4 x i32> @llvm.x86.sse2.cvtpd2dq(<2 x double> [[A0:%.*]])
+; CHECK-NEXT: [[RES:%.*]] = shufflevector <4 x i32> [[CVT]], <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+; CHECK-NEXT: [[BC:%.*]] = bitcast <4 x i32> [[RES]] to <2 x i64>
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[BC]]
+;
+ %cvt = call <4 x i32> @llvm.x86.sse2.cvtpd2dq(<2 x double> %a0)
+ %res = shufflevector <4 x i32> %cvt, <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+ %bc = bitcast <4 x i32> %res to <2 x i64>
+ ret <2 x i64> %bc
+}
+
+
+define <2 x i64> @test_mm_cvtpd_epi32_zext_load(<2 x double>* %p0) nounwind #0 {
+; CHECK-LABEL: @test_mm_cvtpd_epi32_zext_load(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF0]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[A0:%.*]] = load <2 x double>, <2 x double>* [[P0:%.*]], align 16
+; CHECK-NEXT: [[TMP4:%.*]] = ptrtoint <2 x double>* [[P0]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = xor i64 [[TMP4]], 87960930222080
+; CHECK-NEXT: [[TMP6:%.*]] = inttoptr i64 [[TMP5]] to <2 x i64>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <2 x i64>, <2 x i64>* [[TMP6]], align 16
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast <2 x i64> [[_MSLD]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP7]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[CVT:%.*]] = call <4 x i32> @llvm.x86.sse2.cvtpd2dq(<2 x double> [[A0]])
+; CHECK-NEXT: [[RES:%.*]] = shufflevector <4 x i32> [[CVT]], <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+; CHECK-NEXT: [[BC:%.*]] = bitcast <4 x i32> [[RES]] to <2 x i64>
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[BC]]
+;
+ %a0 = load <2 x double>, <2 x double>* %p0
+ %cvt = call <4 x i32> @llvm.x86.sse2.cvtpd2dq(<2 x double> %a0)
+ %res = shufflevector <4 x i32> %cvt, <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+ %bc = bitcast <4 x i32> %res to <2 x i64>
+ ret <2 x i64> %bc
+}
+
+
+define <4 x float> @test_x86_sse2_cvtpd2ps(<2 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_cvtpd2ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse2.cvtpd2ps(<2 x double> [[A0:%.*]])
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse2.cvtpd2ps(<2 x double> %a0) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse2.cvtpd2ps(<2 x double>) nounwind readnone
+
+define <4 x float> @test_x86_sse2_cvtpd2ps_zext(<2 x double> %a0) nounwind #0 {
+; CHECK-LABEL: @test_x86_sse2_cvtpd2ps_zext(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[CVT:%.*]] = call <4 x float> @llvm.x86.sse2.cvtpd2ps(<2 x double> [[A0:%.*]])
+; CHECK-NEXT: [[RES:%.*]] = shufflevector <4 x float> [[CVT]], <4 x float> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %cvt = call <4 x float> @llvm.x86.sse2.cvtpd2ps(<2 x double> %a0)
+ %res = shufflevector <4 x float> %cvt, <4 x float> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+ ret <4 x float> %res
+}
+
+define <4 x float> @test_x86_sse2_cvtpd2ps_zext_load(<2 x double>* %p0) nounwind #0 {
+; CHECK-LABEL: @test_x86_sse2_cvtpd2ps_zext_load(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF0]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[A0:%.*]] = load <2 x double>, <2 x double>* [[P0:%.*]], align 16
+; CHECK-NEXT: [[TMP4:%.*]] = ptrtoint <2 x double>* [[P0]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = xor i64 [[TMP4]], 87960930222080
+; CHECK-NEXT: [[TMP6:%.*]] = inttoptr i64 [[TMP5]] to <2 x i64>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <2 x i64>, <2 x i64>* [[TMP6]], align 16
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast <2 x i64> [[_MSLD]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP7]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[CVT:%.*]] = call <4 x float> @llvm.x86.sse2.cvtpd2ps(<2 x double> [[A0]])
+; CHECK-NEXT: [[RES:%.*]] = shufflevector <4 x float> [[CVT]], <4 x float> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %a0 = load <2 x double>, <2 x double>* %p0
+ %cvt = call <4 x float> @llvm.x86.sse2.cvtpd2ps(<2 x double> %a0)
+ %res = shufflevector <4 x float> %cvt, <4 x float> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+ ret <4 x float> %res
+}
+
+define <4 x i32> @test_x86_sse2_cvtps2dq(<4 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_cvtps2dq(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.sse2.cvtps2dq(<4 x float> [[A0:%.*]])
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.sse2.cvtps2dq(<4 x float> %a0) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.sse2.cvtps2dq(<4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse2_cvtsd2si(<2 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_cvtsd2si(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = extractelement <2 x i64> [[TMP1]], i32 0
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.cvtsd2si(<2 x double> [[A0:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.cvtsd2si(<2 x double> %a0) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.cvtsd2si(<2 x double>) nounwind readnone
+
+
+define <4 x float> @test_x86_sse2_cvtsd2ss(<4 x float> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_cvtsd2ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = extractelement <2 x i64> [[TMP1]], i32 0
+; CHECK-NEXT: [[TMP4:%.*]] = insertelement <4 x i32> [[TMP2]], i32 0, i32 0
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP5:%.*]], label [[TMP6:%.*]], !prof [[PROF0]]
+; CHECK: 5:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 6:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse2.cvtsd2ss(<4 x float> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i32> [[TMP4]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse2.cvtsd2ss(<4 x float> %a0, <2 x double> %a1) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse2.cvtsd2ss(<4 x float>, <2 x double>) nounwind readnone
+
+
+define <4 x float> @test_x86_sse2_cvtsd2ss_load(<4 x float> %a0, <2 x double>* %p1) #0 {
+; CHECK-LABEL: @test_x86_sse2_cvtsd2ss_load(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[A1:%.*]] = load <2 x double>, <2 x double>* [[P1:%.*]], align 16
+; CHECK-NEXT: [[TMP5:%.*]] = ptrtoint <2 x double>* [[P1]] to i64
+; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
+; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to <2 x i64>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <2 x i64>, <2 x i64>* [[TMP7]], align 16
+; CHECK-NEXT: [[TMP8:%.*]] = extractelement <2 x i64> [[_MSLD]], i32 0
+; CHECK-NEXT: [[TMP9:%.*]] = insertelement <4 x i32> [[TMP2]], i32 0, i32 0
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP8]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP10:%.*]], label [[TMP11:%.*]], !prof [[PROF0]]
+; CHECK: 10:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 11:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse2.cvtsd2ss(<4 x float> [[A0:%.*]], <2 x double> [[A1]])
+; CHECK-NEXT: store <4 x i32> [[TMP9]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %a1 = load <2 x double>, <2 x double>* %p1
+ %res = call <4 x float> @llvm.x86.sse2.cvtsd2ss(<4 x float> %a0, <2 x double> %a1) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+
+
+define <4 x float> @test_x86_sse2_cvtsd2ss_load_optsize(<4 x float> %a0, <2 x double>* %p1) optsize #0 {
+; CHECK-LABEL: @test_x86_sse2_cvtsd2ss_load_optsize(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[A1:%.*]] = load <2 x double>, <2 x double>* [[P1:%.*]], align 16
+; CHECK-NEXT: [[TMP5:%.*]] = ptrtoint <2 x double>* [[P1]] to i64
+; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
+; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to <2 x i64>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <2 x i64>, <2 x i64>* [[TMP7]], align 16
+; CHECK-NEXT: [[TMP8:%.*]] = extractelement <2 x i64> [[_MSLD]], i32 0
+; CHECK-NEXT: [[TMP9:%.*]] = insertelement <4 x i32> [[TMP2]], i32 0, i32 0
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i64 [[TMP8]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP10:%.*]], label [[TMP11:%.*]], !prof [[PROF0]]
+; CHECK: 10:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 11:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse2.cvtsd2ss(<4 x float> [[A0:%.*]], <2 x double> [[A1]])
+; CHECK-NEXT: store <4 x i32> [[TMP9]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %a1 = load <2 x double>, <2 x double>* %p1
+ %res = call <4 x float> @llvm.x86.sse2.cvtsd2ss(<4 x float> %a0, <2 x double> %a1) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+
+
+define <4 x i32> @test_x86_sse2_cvttpd2dq(<2 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_cvttpd2dq(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.sse2.cvttpd2dq(<2 x double> [[A0:%.*]])
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.sse2.cvttpd2dq(<2 x double> %a0) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.sse2.cvttpd2dq(<2 x double>) nounwind readnone
+
+
+define <2 x i64> @test_mm_cvttpd_epi32_zext(<2 x double> %a0) nounwind #0 {
+; CHECK-LABEL: @test_mm_cvttpd_epi32_zext(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[CVT:%.*]] = call <4 x i32> @llvm.x86.sse2.cvttpd2dq(<2 x double> [[A0:%.*]])
+; CHECK-NEXT: [[RES:%.*]] = shufflevector <4 x i32> [[CVT]], <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+; CHECK-NEXT: [[BC:%.*]] = bitcast <4 x i32> [[RES]] to <2 x i64>
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[BC]]
+;
+ %cvt = call <4 x i32> @llvm.x86.sse2.cvttpd2dq(<2 x double> %a0)
+ %res = shufflevector <4 x i32> %cvt, <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+ %bc = bitcast <4 x i32> %res to <2 x i64>
+ ret <2 x i64> %bc
+}
+
+
+define <2 x i64> @test_mm_cvttpd_epi32_zext_load(<2 x double>* %p0) nounwind #0 {
+; CHECK-LABEL: @test_mm_cvttpd_epi32_zext_load(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF0]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: [[A0:%.*]] = load <2 x double>, <2 x double>* [[P0:%.*]], align 16
+; CHECK-NEXT: [[TMP4:%.*]] = ptrtoint <2 x double>* [[P0]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = xor i64 [[TMP4]], 87960930222080
+; CHECK-NEXT: [[TMP6:%.*]] = inttoptr i64 [[TMP5]] to <2 x i64>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <2 x i64>, <2 x i64>* [[TMP6]], align 16
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast <2 x i64> [[_MSLD]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP7]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF0]]
+; CHECK: 8:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 9:
+; CHECK-NEXT: [[CVT:%.*]] = call <4 x i32> @llvm.x86.sse2.cvttpd2dq(<2 x double> [[A0]])
+; CHECK-NEXT: [[RES:%.*]] = shufflevector <4 x i32> [[CVT]], <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+; CHECK-NEXT: [[BC:%.*]] = bitcast <4 x i32> [[RES]] to <2 x i64>
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[BC]]
+;
+ %a0 = load <2 x double>, <2 x double>* %p0
+ %cvt = call <4 x i32> @llvm.x86.sse2.cvttpd2dq(<2 x double> %a0)
+ %res = shufflevector <4 x i32> %cvt, <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+ %bc = bitcast <4 x i32> %res to <2 x i64>
+ ret <2 x i64> %bc
+}
+
+
+define <4 x i32> @test_x86_sse2_cvttps2dq(<4 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_cvttps2dq(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.sse2.cvttps2dq(<4 x float> [[A0:%.*]])
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.sse2.cvttps2dq(<4 x float> %a0) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.sse2.cvttps2dq(<4 x float>) nounwind readnone
+
+
+define i32 @test_x86_sse2_cvttsd2si(<2 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_cvttsd2si(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = extractelement <2 x i64> [[TMP1]], i32 0
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.cvttsd2si(<2 x double> [[A0:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.cvttsd2si(<2 x double> %a0) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.cvttsd2si(<2 x double>) nounwind readnone
+
+
+define <2 x double> @test_x86_sse2_max_pd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_max_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.sse2.max.pd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store <2 x i64> [[_MSPROP]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.sse2.max.pd(<2 x double> %a0, <2 x double> %a1) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.sse2.max.pd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define <2 x double> @test_x86_sse2_max_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_max_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = shufflevector <2 x i64> [[TMP1]], <2 x i64> [[TMP3]], <2 x i32> <i32 2, i32 1>
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.sse2.max.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store <2 x i64> [[TMP4]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.sse2.max.sd(<2 x double> %a0, <2 x double> %a1) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.sse2.max.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define <2 x double> @test_x86_sse2_min_pd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_min_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.sse2.min.pd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store <2 x i64> [[_MSPROP]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.sse2.min.pd(<2 x double> %a0, <2 x double> %a1) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.sse2.min.pd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define <2 x double> @test_x86_sse2_min_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_min_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = shufflevector <2 x i64> [[TMP1]], <2 x i64> [[TMP3]], <2 x i32> <i32 2, i32 1>
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.sse2.min.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store <2 x i64> [[TMP4]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.sse2.min.sd(<2 x double> %a0, <2 x double> %a1) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.sse2.min.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_sse2_movmsk_pd(<2 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_movmsk_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.movmsk.pd(<2 x double> [[A0:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.movmsk.pd(<2 x double> %a0) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.movmsk.pd(<2 x double>) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse2_packssdw_128(<4 x i32> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_packssdw_128(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <4 x i32> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <4 x i1> [[TMP3]] to <4 x i32>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <4 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <4 x i1> [[TMP5]] to <4 x i32>
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <8 x i16> @llvm.x86.sse2.packssdw.128(<4 x i32> [[TMP4]], <4 x i32> [[TMP6]])
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse2.packssdw.128(<4 x i32> [[A0:%.*]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i16> [[_MSPROP_VECTOR_PACK]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse2.packssdw.128(<4 x i32> %a0, <4 x i32> %a1) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse2.packssdw.128(<4 x i32>, <4 x i32>) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse2_packssdw_128_fold() #0 {
+; CHECK-LABEL: @test_x86_sse2_packssdw_128_fold(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <8 x i16> @llvm.x86.sse2.packssdw.128(<4 x i32> zeroinitializer, <4 x i32> zeroinitializer)
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse2.packssdw.128(<4 x i32> zeroinitializer, <4 x i32> <i32 65535, i32 65536, i32 -1, i32 -131072>)
+; CHECK-NEXT: store <8 x i16> [[_MSPROP_VECTOR_PACK]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse2.packssdw.128(<4 x i32> zeroinitializer, <4 x i32> <i32 65535, i32 65536, i32 -1, i32 -131072>)
+ ret <8 x i16> %res
+}
+
+
+define <16 x i8> @test_x86_sse2_packsswb_128(<8 x i16> %a0, <8 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_packsswb_128(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <8 x i16> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <8 x i1> [[TMP3]] to <8 x i16>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <8 x i16> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <8 x i1> [[TMP5]] to <8 x i16>
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <16 x i8> @llvm.x86.sse2.packsswb.128(<8 x i16> [[TMP4]], <8 x i16> [[TMP6]])
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i8> @llvm.x86.sse2.packsswb.128(<8 x i16> [[A0:%.*]], <8 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i8> [[_MSPROP_VECTOR_PACK]], <16 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i8>*), align 8
+; CHECK-NEXT: ret <16 x i8> [[RES]]
+;
+ %res = call <16 x i8> @llvm.x86.sse2.packsswb.128(<8 x i16> %a0, <8 x i16> %a1) ; <<16 x i8>> [#uses=1]
+ ret <16 x i8> %res
+}
+declare <16 x i8> @llvm.x86.sse2.packsswb.128(<8 x i16>, <8 x i16>) nounwind readnone
+
+
+define <16 x i8> @test_x86_sse2_packsswb_128_fold() #0 {
+; CHECK-LABEL: @test_x86_sse2_packsswb_128_fold(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <16 x i8> @llvm.x86.sse2.packsswb.128(<8 x i16> zeroinitializer, <8 x i16> zeroinitializer)
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i8> @llvm.x86.sse2.packsswb.128(<8 x i16> <i16 0, i16 255, i16 256, i16 -1, i16 -1, i16 -255, i16 -256, i16 -32678>, <8 x i16> zeroinitializer)
+; CHECK-NEXT: store <16 x i8> [[_MSPROP_VECTOR_PACK]], <16 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i8>*), align 8
+; CHECK-NEXT: ret <16 x i8> [[RES]]
+;
+ %res = call <16 x i8> @llvm.x86.sse2.packsswb.128(<8 x i16> <i16 0, i16 255, i16 256, i16 65535, i16 -1, i16 -255, i16 -256, i16 -32678>, <8 x i16> zeroinitializer)
+ ret <16 x i8> %res
+}
+
+
+define <16 x i8> @test_x86_sse2_packuswb_128(<8 x i16> %a0, <8 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_packuswb_128(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <8 x i16> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <8 x i1> [[TMP3]] to <8 x i16>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <8 x i16> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <8 x i1> [[TMP5]] to <8 x i16>
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <16 x i8> @llvm.x86.sse2.packsswb.128(<8 x i16> [[TMP4]], <8 x i16> [[TMP6]])
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i8> @llvm.x86.sse2.packuswb.128(<8 x i16> [[A0:%.*]], <8 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i8> [[_MSPROP_VECTOR_PACK]], <16 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i8>*), align 8
+; CHECK-NEXT: ret <16 x i8> [[RES]]
+;
+ %res = call <16 x i8> @llvm.x86.sse2.packuswb.128(<8 x i16> %a0, <8 x i16> %a1) ; <<16 x i8>> [#uses=1]
+ ret <16 x i8> %res
+}
+declare <16 x i8> @llvm.x86.sse2.packuswb.128(<8 x i16>, <8 x i16>) nounwind readnone
+
+
+define <16 x i8> @test_x86_sse2_packuswb_128_fold() #0 {
+; CHECK-LABEL: @test_x86_sse2_packuswb_128_fold(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <16 x i8> @llvm.x86.sse2.packsswb.128(<8 x i16> zeroinitializer, <8 x i16> zeroinitializer)
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i8> @llvm.x86.sse2.packuswb.128(<8 x i16> <i16 0, i16 255, i16 256, i16 -1, i16 -1, i16 -255, i16 -256, i16 -32678>, <8 x i16> zeroinitializer)
+; CHECK-NEXT: store <16 x i8> [[_MSPROP_VECTOR_PACK]], <16 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i8>*), align 8
+; CHECK-NEXT: ret <16 x i8> [[RES]]
+;
+ %res = call <16 x i8> @llvm.x86.sse2.packuswb.128(<8 x i16> <i16 0, i16 255, i16 256, i16 65535, i16 -1, i16 -255, i16 -256, i16 -32678>, <8 x i16> zeroinitializer)
+ ret <16 x i8> %res
+}
+
+
+define <16 x i8> @test_x86_sse2_pavg_b(<16 x i8> %a0, <16 x i8> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_pavg_b(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i8>, <16 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i8>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i8>, <16 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <16 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <16 x i8> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i8> @llvm.x86.sse2.pavg.b(<16 x i8> [[A0:%.*]], <16 x i8> [[A1:%.*]])
+; CHECK-NEXT: store <16 x i8> [[_MSPROP]], <16 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i8>*), align 8
+; CHECK-NEXT: ret <16 x i8> [[RES]]
+;
+ %res = call <16 x i8> @llvm.x86.sse2.pavg.b(<16 x i8> %a0, <16 x i8> %a1) ; <<16 x i8>> [#uses=1]
+ ret <16 x i8> %res
+}
+declare <16 x i8> @llvm.x86.sse2.pavg.b(<16 x i8>, <16 x i8>) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse2_pavg_w(<8 x i16> %a0, <8 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_pavg_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse2.pavg.w(<8 x i16> [[A0:%.*]], <8 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i16> [[_MSPROP]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse2.pavg.w(<8 x i16> %a0, <8 x i16> %a1) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse2.pavg.w(<8 x i16>, <8 x i16>) nounwind readnone
+
+
+define <4 x i32> @test_x86_sse2_pmadd_wd(<8 x i16> %a0, <8 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_pmadd_wd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <8 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = bitcast <8 x i16> [[TMP3]] to <4 x i32>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <4 x i32> [[TMP4]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <4 x i1> [[TMP5]] to <4 x i32>
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.sse2.pmadd.wd(<8 x i16> [[A0:%.*]], <8 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <4 x i32> [[TMP6]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.sse2.pmadd.wd(<8 x i16> %a0, <8 x i16> %a1) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.sse2.pmadd.wd(<8 x i16>, <8 x i16>) nounwind readnone
+
+
+define i32 @test_x86_sse2_pmovmskb_128(<16 x i8> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_pmovmskb_128(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i8>, <16 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <16 x i8> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.pmovmskb.128(<16 x i8> [[A0:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.pmovmskb.128(<16 x i8> %a0) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.pmovmskb.128(<16 x i8>) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse2_pmulh_w(<8 x i16> %a0, <8 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_pmulh_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse2.pmulh.w(<8 x i16> [[A0:%.*]], <8 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i16> [[_MSPROP]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse2.pmulh.w(<8 x i16> %a0, <8 x i16> %a1) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse2.pmulh.w(<8 x i16>, <8 x i16>) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse2_pmulhu_w(<8 x i16> %a0, <8 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_pmulhu_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <8 x i16> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse2.pmulhu.w(<8 x i16> [[A0:%.*]], <8 x i16> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i16> [[_MSPROP]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse2.pmulhu.w(<8 x i16> %a0, <8 x i16> %a1) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse2.pmulhu.w(<8 x i16>, <8 x i16>) nounwind readnone
+
+
+define <2 x i64> @test_x86_sse2_psad_bw(<16 x i8> %a0, <16 x i8> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_psad_bw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i8>, <16 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i8>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i8>, <16 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <16 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <16 x i8> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = bitcast <16 x i8> [[TMP3]] to <2 x i64>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <2 x i64> [[TMP4]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <2 x i1> [[TMP5]] to <2 x i64>
+; CHECK-NEXT: [[TMP7:%.*]] = lshr <2 x i64> [[TMP6]], <i64 48, i64 48>
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.sse2.psad.bw(<16 x i8> [[A0:%.*]], <16 x i8> [[A1:%.*]])
+; CHECK-NEXT: store <2 x i64> [[TMP7]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.sse2.psad.bw(<16 x i8> %a0, <16 x i8> %a1) ; <<2 x i64>> [#uses=1]
+ ret <2 x i64> %res
+}
+declare <2 x i64> @llvm.x86.sse2.psad.bw(<16 x i8>, <16 x i8>) nounwind readnone
+
+
+define <4 x i32> @test_x86_sse2_psll_d(<4 x i32> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_psll_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i128
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i128 [[TMP6]] to <4 x i32>
+; CHECK-NEXT: [[TMP8:%.*]] = call <4 x i32> @llvm.x86.sse2.psll.d(<4 x i32> [[TMP1]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <4 x i32> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.sse2.psll.d(<4 x i32> [[A0:%.*]], <4 x i32> [[A1]])
+; CHECK-NEXT: store <4 x i32> [[TMP9]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.sse2.psll.d(<4 x i32> %a0, <4 x i32> %a1) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.sse2.psll.d(<4 x i32>, <4 x i32>) nounwind readnone
+
+
+define <2 x i64> @test_x86_sse2_psll_q(<2 x i64> %a0, <2 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_psll_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i128
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i128 [[TMP6]] to <2 x i64>
+; CHECK-NEXT: [[TMP8:%.*]] = call <2 x i64> @llvm.x86.sse2.psll.q(<2 x i64> [[TMP1]], <2 x i64> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <2 x i64> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.sse2.psll.q(<2 x i64> [[A0:%.*]], <2 x i64> [[A1]])
+; CHECK-NEXT: store <2 x i64> [[TMP9]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.sse2.psll.q(<2 x i64> %a0, <2 x i64> %a1) ; <<2 x i64>> [#uses=1]
+ ret <2 x i64> %res
+}
+declare <2 x i64> @llvm.x86.sse2.psll.q(<2 x i64>, <2 x i64>) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse2_psll_w(<8 x i16> %a0, <8 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_psll_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i16> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i128
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i128 [[TMP6]] to <8 x i16>
+; CHECK-NEXT: [[TMP8:%.*]] = call <8 x i16> @llvm.x86.sse2.psll.w(<8 x i16> [[TMP1]], <8 x i16> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <8 x i16> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse2.psll.w(<8 x i16> [[A0:%.*]], <8 x i16> [[A1]])
+; CHECK-NEXT: store <8 x i16> [[TMP9]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse2.psll.w(<8 x i16> %a0, <8 x i16> %a1) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse2.psll.w(<8 x i16>, <8 x i16>) nounwind readnone
+
+
+define <4 x i32> @test_x86_sse2_pslli_d(<4 x i32> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_pslli_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <4 x i32> @llvm.x86.sse2.pslli.d(<4 x i32> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.sse2.pslli.d(<4 x i32> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <4 x i32> [[TMP3]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.sse2.pslli.d(<4 x i32> %a0, i32 7) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.sse2.pslli.d(<4 x i32>, i32) nounwind readnone
+
+
+define <2 x i64> @test_x86_sse2_pslli_q(<2 x i64> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_pslli_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <2 x i64> @llvm.x86.sse2.pslli.q(<2 x i64> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.sse2.pslli.q(<2 x i64> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <2 x i64> [[TMP3]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.sse2.pslli.q(<2 x i64> %a0, i32 7) ; <<2 x i64>> [#uses=1]
+ ret <2 x i64> %res
+}
+declare <2 x i64> @llvm.x86.sse2.pslli.q(<2 x i64>, i32) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse2_pslli_w(<8 x i16> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_pslli_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <8 x i16> @llvm.x86.sse2.pslli.w(<8 x i16> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <8 x i16> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse2.pslli.w(<8 x i16> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <8 x i16> [[TMP3]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse2.pslli.w(<8 x i16> %a0, i32 7) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse2.pslli.w(<8 x i16>, i32) nounwind readnone
+
+
+define <4 x i32> @test_x86_sse2_psra_d(<4 x i32> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_psra_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i128
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i128 [[TMP6]] to <4 x i32>
+; CHECK-NEXT: [[TMP8:%.*]] = call <4 x i32> @llvm.x86.sse2.psra.d(<4 x i32> [[TMP1]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <4 x i32> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.sse2.psra.d(<4 x i32> [[A0:%.*]], <4 x i32> [[A1]])
+; CHECK-NEXT: store <4 x i32> [[TMP9]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.sse2.psra.d(<4 x i32> %a0, <4 x i32> %a1) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.sse2.psra.d(<4 x i32>, <4 x i32>) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse2_psra_w(<8 x i16> %a0, <8 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_psra_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i16> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i128
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i128 [[TMP6]] to <8 x i16>
+; CHECK-NEXT: [[TMP8:%.*]] = call <8 x i16> @llvm.x86.sse2.psra.w(<8 x i16> [[TMP1]], <8 x i16> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <8 x i16> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse2.psra.w(<8 x i16> [[A0:%.*]], <8 x i16> [[A1]])
+; CHECK-NEXT: store <8 x i16> [[TMP9]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse2.psra.w(<8 x i16> %a0, <8 x i16> %a1) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse2.psra.w(<8 x i16>, <8 x i16>) nounwind readnone
+
+
+define <4 x i32> @test_x86_sse2_psrai_d(<4 x i32> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_psrai_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <4 x i32> @llvm.x86.sse2.psrai.d(<4 x i32> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.sse2.psrai.d(<4 x i32> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <4 x i32> [[TMP3]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.sse2.psrai.d(<4 x i32> %a0, i32 7) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.sse2.psrai.d(<4 x i32>, i32) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse2_psrai_w(<8 x i16> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_psrai_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <8 x i16> @llvm.x86.sse2.psrai.w(<8 x i16> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <8 x i16> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse2.psrai.w(<8 x i16> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <8 x i16> [[TMP3]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse2.psrai.w(<8 x i16> %a0, i32 7) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse2.psrai.w(<8 x i16>, i32) nounwind readnone
+
+
+define <4 x i32> @test_x86_sse2_psrl_d(<4 x i32> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_psrl_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i128
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i128 [[TMP6]] to <4 x i32>
+; CHECK-NEXT: [[TMP8:%.*]] = call <4 x i32> @llvm.x86.sse2.psrl.d(<4 x i32> [[TMP1]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <4 x i32> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.sse2.psrl.d(<4 x i32> [[A0:%.*]], <4 x i32> [[A1]])
+; CHECK-NEXT: store <4 x i32> [[TMP9]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.sse2.psrl.d(<4 x i32> %a0, <4 x i32> %a1) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.sse2.psrl.d(<4 x i32>, <4 x i32>) nounwind readnone
+
+
+define <2 x i64> @test_x86_sse2_psrl_q(<2 x i64> %a0, <2 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_psrl_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i128
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i128 [[TMP6]] to <2 x i64>
+; CHECK-NEXT: [[TMP8:%.*]] = call <2 x i64> @llvm.x86.sse2.psrl.q(<2 x i64> [[TMP1]], <2 x i64> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <2 x i64> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.sse2.psrl.q(<2 x i64> [[A0:%.*]], <2 x i64> [[A1]])
+; CHECK-NEXT: store <2 x i64> [[TMP9]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.sse2.psrl.q(<2 x i64> %a0, <2 x i64> %a1) ; <<2 x i64>> [#uses=1]
+ ret <2 x i64> %res
+}
+declare <2 x i64> @llvm.x86.sse2.psrl.q(<2 x i64>, <2 x i64>) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse2_psrl_w(<8 x i16> %a0, <8 x i16> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_psrl_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <8 x i16> [[TMP2]] to i128
+; CHECK-NEXT: [[TMP4:%.*]] = trunc i128 [[TMP3]] to i64
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i128
+; CHECK-NEXT: [[TMP7:%.*]] = bitcast i128 [[TMP6]] to <8 x i16>
+; CHECK-NEXT: [[TMP8:%.*]] = call <8 x i16> @llvm.x86.sse2.psrl.w(<8 x i16> [[TMP1]], <8 x i16> [[A1:%.*]])
+; CHECK-NEXT: [[TMP9:%.*]] = or <8 x i16> [[TMP8]], [[TMP7]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse2.psrl.w(<8 x i16> [[A0:%.*]], <8 x i16> [[A1]])
+; CHECK-NEXT: store <8 x i16> [[TMP9]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse2.psrl.w(<8 x i16> %a0, <8 x i16> %a1) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse2.psrl.w(<8 x i16>, <8 x i16>) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse2_psrl_w_load(<8 x i16> %a0, <8 x i16>* %p) #0 {
+; CHECK-LABEL: @test_x86_sse2_psrl_w_load(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[A1:%.*]] = load <8 x i16>, <8 x i16>* [[P:%.*]], align 16
+; CHECK-NEXT: [[TMP5:%.*]] = ptrtoint <8 x i16>* [[P]] to i64
+; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
+; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to <8 x i16>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <8 x i16>, <8 x i16>* [[TMP7]], align 16
+; CHECK-NEXT: [[TMP8:%.*]] = bitcast <8 x i16> [[_MSLD]] to i128
+; CHECK-NEXT: [[TMP9:%.*]] = trunc i128 [[TMP8]] to i64
+; CHECK-NEXT: [[TMP10:%.*]] = icmp ne i64 [[TMP9]], 0
+; CHECK-NEXT: [[TMP11:%.*]] = sext i1 [[TMP10]] to i128
+; CHECK-NEXT: [[TMP12:%.*]] = bitcast i128 [[TMP11]] to <8 x i16>
+; CHECK-NEXT: [[TMP13:%.*]] = call <8 x i16> @llvm.x86.sse2.psrl.w(<8 x i16> [[TMP2]], <8 x i16> [[A1]])
+; CHECK-NEXT: [[TMP14:%.*]] = or <8 x i16> [[TMP13]], [[TMP12]]
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse2.psrl.w(<8 x i16> [[A0:%.*]], <8 x i16> [[A1]])
+; CHECK-NEXT: store <8 x i16> [[TMP14]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %a1 = load <8 x i16>, <8 x i16>* %p
+ %res = call <8 x i16> @llvm.x86.sse2.psrl.w(<8 x i16> %a0, <8 x i16> %a1) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+
+
+define <4 x i32> @test_x86_sse2_psrli_d(<4 x i32> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_psrli_d(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <4 x i32> @llvm.x86.sse2.psrli.d(<4 x i32> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <4 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <4 x i32> @llvm.x86.sse2.psrli.d(<4 x i32> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <4 x i32> [[TMP3]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x i32> [[RES]]
+;
+ %res = call <4 x i32> @llvm.x86.sse2.psrli.d(<4 x i32> %a0, i32 7) ; <<4 x i32>> [#uses=1]
+ ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.sse2.psrli.d(<4 x i32>, i32) nounwind readnone
+
+
+define <2 x i64> @test_x86_sse2_psrli_q(<2 x i64> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_psrli_q(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <2 x i64> @llvm.x86.sse2.psrli.q(<2 x i64> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <2 x i64> @llvm.x86.sse2.psrli.q(<2 x i64> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <2 x i64> [[TMP3]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x i64> [[RES]]
+;
+ %res = call <2 x i64> @llvm.x86.sse2.psrli.q(<2 x i64> %a0, i32 7) ; <<2 x i64>> [#uses=1]
+ ret <2 x i64> %res
+}
+declare <2 x i64> @llvm.x86.sse2.psrli.q(<2 x i64>, i32) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse2_psrli_w(<8 x i16> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse2_psrli_w(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = call <8 x i16> @llvm.x86.sse2.psrli.w(<8 x i16> [[TMP1]], i32 7)
+; CHECK-NEXT: [[TMP3:%.*]] = or <8 x i16> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse2.psrli.w(<8 x i16> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <8 x i16> [[TMP3]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse2.psrli.w(<8 x i16> %a0, i32 7) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse2.psrli.w(<8 x i16>, i32) nounwind readnone
+
+
+define i32 @test_x86_sse2_ucomieq_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_ucomieq_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.ucomieq.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.ucomieq.sd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.ucomieq.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_sse2_ucomige_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_ucomige_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.ucomige.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.ucomige.sd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.ucomige.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_sse2_ucomigt_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_ucomigt_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.ucomigt.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.ucomigt.sd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.ucomigt.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_sse2_ucomile_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_ucomile_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.ucomile.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.ucomile.sd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.ucomile.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_sse2_ucomilt_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_ucomilt_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.ucomilt.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.ucomilt.sd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.ucomilt.sd(<2 x double>, <2 x double>) nounwind readnone
+
+
+define i32 @test_x86_sse2_ucomineq_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse2_ucomineq_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP3]], i64 0
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne i64 [[TMP4]], 0
+; CHECK-NEXT: [[TMP6:%.*]] = sext i1 [[TMP5]] to i32
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse2.ucomineq.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]])
+; CHECK-NEXT: store i32 [[TMP6]], i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse2.ucomineq.sd(<2 x double> %a0, <2 x double> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse2.ucomineq.sd(<2 x double>, <2 x double>) nounwind readnone
+
+define void @test_x86_sse2_pause() #0 {
+; CHECK-LABEL: @test_x86_sse2_pause(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: tail call void @llvm.x86.sse2.pause()
+; CHECK-NEXT: ret void
+;
+ tail call void @llvm.x86.sse2.pause()
+ ret void
+}
+declare void @llvm.x86.sse2.pause() nounwind
+
+define void @lfence() nounwind #0 {
+; CHECK-LABEL: @lfence(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: tail call void @llvm.x86.sse2.lfence()
+; CHECK-NEXT: ret void
+;
+ tail call void @llvm.x86.sse2.lfence()
+ ret void
+}
+declare void @llvm.x86.sse2.lfence() nounwind
+
+define void @mfence() nounwind #0 {
+; CHECK-LABEL: @mfence(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: tail call void @llvm.x86.sse2.mfence()
+; CHECK-NEXT: ret void
+;
+ tail call void @llvm.x86.sse2.mfence()
+ ret void
+}
+declare void @llvm.x86.sse2.mfence() nounwind
+
+define void @clflush(i8* %p) nounwind #0 {
+; CHECK-LABEL: @clflush(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP2:%.*]], label [[TMP3:%.*]], !prof [[PROF0]]
+; CHECK: 2:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR6]]
+; CHECK-NEXT: unreachable
+; CHECK: 3:
+; CHECK-NEXT: tail call void @llvm.x86.sse2.clflush(i8* [[P:%.*]])
+; CHECK-NEXT: ret void
+;
+ tail call void @llvm.x86.sse2.clflush(i8* %p)
+ ret void
+}
+declare void @llvm.x86.sse2.clflush(i8*) nounwind
+
+attributes #0 = { sanitize_memory }
diff --git a/llvm/test/Instrumentation/MemorySanitizer/sse41-intrinsics-x86.ll b/llvm/test/Instrumentation/MemorySanitizer/sse41-intrinsics-x86.ll
new file mode 100644
index 0000000000000..53f902ee695b7
--- /dev/null
+++ b/llvm/test/Instrumentation/MemorySanitizer/sse41-intrinsics-x86.ll
@@ -0,0 +1,463 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt %s -S -passes=msan 2>&1 | FileCheck %s
+
+target datalayout = "e-m:o-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+define <2 x double> @test_x86_sse41_blendvpd(<2 x double> %a0, <2 x double> %a1, <2 x double> %a2) #0 {
+; CHECK-LABEL: @test_x86_sse41_blendvpd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <2 x i64> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[_MSPROP1:%.*]] = or <2 x i64> [[_MSPROP]], [[TMP3]]
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.sse41.blendvpd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]], <2 x double> [[A2:%.*]])
+; CHECK-NEXT: store <2 x i64> [[_MSPROP1]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.sse41.blendvpd(<2 x double> %a0, <2 x double> %a1, <2 x double> %a2) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.sse41.blendvpd(<2 x double>, <2 x double>, <2 x double>) nounwind readnone
+
+
+define <4 x float> @test_x86_sse41_blendvps(<4 x float> %a0, <4 x float> %a1, <4 x float> %a2) #0 {
+; CHECK-LABEL: @test_x86_sse41_blendvps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <4 x i32> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[_MSPROP1:%.*]] = or <4 x i32> [[_MSPROP]], [[TMP3]]
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse41.blendvps(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]], <4 x float> [[A2:%.*]])
+; CHECK-NEXT: store <4 x i32> [[_MSPROP1]], <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse41.blendvps(<4 x float> %a0, <4 x float> %a1, <4 x float> %a2) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse41.blendvps(<4 x float>, <4 x float>, <4 x float>) nounwind readnone
+
+
+define <2 x double> @test_x86_sse41_dppd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse41_dppd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0:![0-9]+]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3:[0-9]+]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.sse41.dppd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.sse41.dppd(<2 x double> %a0, <2 x double> %a1, i8 7) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.sse41.dppd(<2 x double>, <2 x double>, i8) nounwind readnone
+
+
+define <4 x float> @test_x86_sse41_dpps(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse41_dpps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse41.dpps(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse41.dpps(<4 x float> %a0, <4 x float> %a1, i8 7) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse41.dpps(<4 x float>, <4 x float>, i8) nounwind readnone
+
+
+define <4 x float> @test_x86_sse41_insertps(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse41_insertps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse41.insertps(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]], i8 17)
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse41.insertps(<4 x float> %a0, <4 x float> %a1, i8 17) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse41.insertps(<4 x float>, <4 x float>, i8) nounwind readnone
+
+
+
+define <8 x i16> @test_x86_sse41_mpsadbw(<16 x i8> %a0, <16 x i8> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse41_mpsadbw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i8>, <16 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i8>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i8>, <16 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <16 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <16 x i8> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <16 x i8> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse41.mpsadbw(<16 x i8> [[A0:%.*]], <16 x i8> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <8 x i16> zeroinitializer, <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse41.mpsadbw(<16 x i8> %a0, <16 x i8> %a1, i8 7) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse41.mpsadbw(<16 x i8>, <16 x i8>, i8) nounwind readnone
+
+define <8 x i16> @test_x86_sse41_mpsadbw_load_op0(<16 x i8>* %ptr, <16 x i8> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse41_mpsadbw_load_op0(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* getelementptr inbounds ([100 x i64], [100 x i64]* @__msan_param_tls, i32 0, i32 0), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i8>, <16 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 8) to <16 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[A0:%.*]] = load <16 x i8>, <16 x i8>* [[PTR:%.*]], align 16
+; CHECK-NEXT: [[TMP5:%.*]] = ptrtoint <16 x i8>* [[PTR]] to i64
+; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
+; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to <16 x i8>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <16 x i8>, <16 x i8>* [[TMP7]], align 16
+; CHECK-NEXT: [[TMP8:%.*]] = bitcast <16 x i8> [[_MSLD]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP8]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP9:%.*]], label [[TMP10:%.*]], !prof [[PROF0]]
+; CHECK: 9:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 10:
+; CHECK-NEXT: [[TMP11:%.*]] = bitcast <16 x i8> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP2:%.*]] = icmp ne i128 [[TMP11]], 0
+; CHECK-NEXT: br i1 [[_MSCMP2]], label [[TMP12:%.*]], label [[TMP13:%.*]], !prof [[PROF0]]
+; CHECK: 12:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 13:
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse41.mpsadbw(<16 x i8> [[A0]], <16 x i8> [[A1:%.*]], i8 7)
+; CHECK-NEXT: store <8 x i16> zeroinitializer, <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %a0 = load <16 x i8>, <16 x i8>* %ptr
+ %res = call <8 x i16> @llvm.x86.sse41.mpsadbw(<16 x i8> %a0, <16 x i8> %a1, i8 7) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+
+define <8 x i16> @test_x86_sse41_packusdw(<4 x i32> %a0, <4 x i32> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse41_packusdw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ne <4 x i32> [[TMP1]], zeroinitializer
+; CHECK-NEXT: [[TMP4:%.*]] = sext <4 x i1> [[TMP3]] to <4 x i32>
+; CHECK-NEXT: [[TMP5:%.*]] = icmp ne <4 x i32> [[TMP2]], zeroinitializer
+; CHECK-NEXT: [[TMP6:%.*]] = sext <4 x i1> [[TMP5]] to <4 x i32>
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <8 x i16> @llvm.x86.sse2.packssdw.128(<4 x i32> [[TMP4]], <4 x i32> [[TMP6]])
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse41.packusdw(<4 x i32> [[A0:%.*]], <4 x i32> [[A1:%.*]])
+; CHECK-NEXT: store <8 x i16> [[_MSPROP_VECTOR_PACK]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse41.packusdw(<4 x i32> %a0, <4 x i32> %a1) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse41.packusdw(<4 x i32>, <4 x i32>) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse41_packusdw_fold() #0 {
+; CHECK-LABEL: @test_x86_sse41_packusdw_fold(
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP_VECTOR_PACK:%.*]] = call <8 x i16> @llvm.x86.sse2.packssdw.128(<4 x i32> zeroinitializer, <4 x i32> zeroinitializer)
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse41.packusdw(<4 x i32> zeroinitializer, <4 x i32> <i32 65535, i32 65536, i32 -1, i32 -131072>)
+; CHECK-NEXT: store <8 x i16> [[_MSPROP_VECTOR_PACK]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse41.packusdw(<4 x i32> zeroinitializer, <4 x i32> <i32 65535, i32 65536, i32 -1, i32 -131072>)
+ ret <8 x i16> %res
+}
+
+
+define <16 x i8> @test_x86_sse41_pblendvb(<16 x i8> %a0, <16 x i8> %a1, <16 x i8> %a2) #0 {
+; CHECK-LABEL: @test_x86_sse41_pblendvb(
+; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i8>, <16 x i8>* bitcast ([100 x i64]* @__msan_param_tls to <16 x i8>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <16 x i8>, <16 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <16 x i8>*), align 8
+; CHECK-NEXT: [[TMP3:%.*]] = load <16 x i8>, <16 x i8>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 32) to <16 x i8>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSPROP:%.*]] = or <16 x i8> [[TMP1]], [[TMP2]]
+; CHECK-NEXT: [[_MSPROP1:%.*]] = or <16 x i8> [[_MSPROP]], [[TMP3]]
+; CHECK-NEXT: [[RES:%.*]] = call <16 x i8> @llvm.x86.sse41.pblendvb(<16 x i8> [[A0:%.*]], <16 x i8> [[A1:%.*]], <16 x i8> [[A2:%.*]])
+; CHECK-NEXT: store <16 x i8> [[_MSPROP1]], <16 x i8>* bitcast ([100 x i64]* @__msan_retval_tls to <16 x i8>*), align 8
+; CHECK-NEXT: ret <16 x i8> [[RES]]
+;
+ %res = call <16 x i8> @llvm.x86.sse41.pblendvb(<16 x i8> %a0, <16 x i8> %a1, <16 x i8> %a2) ; <<16 x i8>> [#uses=1]
+ ret <16 x i8> %res
+}
+declare <16 x i8> @llvm.x86.sse41.pblendvb(<16 x i8>, <16 x i8>, <16 x i8>) nounwind readnone
+
+
+define <8 x i16> @test_x86_sse41_phminposuw(<8 x i16> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse41_phminposuw(
+; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* bitcast ([100 x i64]* @__msan_param_tls to <8 x i16>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[RES:%.*]] = call <8 x i16> @llvm.x86.sse41.phminposuw(<8 x i16> [[A0:%.*]])
+; CHECK-NEXT: store <8 x i16> [[TMP1]], <8 x i16>* bitcast ([100 x i64]* @__msan_retval_tls to <8 x i16>*), align 8
+; CHECK-NEXT: ret <8 x i16> [[RES]]
+;
+ %res = call <8 x i16> @llvm.x86.sse41.phminposuw(<8 x i16> %a0) ; <<8 x i16>> [#uses=1]
+ ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.sse41.phminposuw(<8 x i16>) nounwind readnone
+
+
+define i32 @test_x86_sse41_ptestc(<2 x i64> %a0, <2 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse41_ptestc(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse41.ptestc(<2 x i64> [[A0:%.*]], <2 x i64> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse41.ptestc(<2 x i64> %a0, <2 x i64> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse41.ptestc(<2 x i64>, <2 x i64>) nounwind readnone
+
+
+define i32 @test_x86_sse41_ptestnzc(<2 x i64> %a0, <2 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse41_ptestnzc(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse41.ptestnzc(<2 x i64> [[A0:%.*]], <2 x i64> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse41.ptestnzc(<2 x i64> %a0, <2 x i64> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse41.ptestnzc(<2 x i64>, <2 x i64>) nounwind readnone
+
+
+define i32 @test_x86_sse41_ptestz(<2 x i64> %a0, <2 x i64> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse41_ptestz(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call i32 @llvm.x86.sse41.ptestz(<2 x i64> [[A0:%.*]], <2 x i64> [[A1:%.*]])
+; CHECK-NEXT: store i32 0, i32* bitcast ([100 x i64]* @__msan_retval_tls to i32*), align 8
+; CHECK-NEXT: ret i32 [[RES]]
+;
+ %res = call i32 @llvm.x86.sse41.ptestz(<2 x i64> %a0, <2 x i64> %a1) ; <i32> [#uses=1]
+ ret i32 %res
+}
+declare i32 @llvm.x86.sse41.ptestz(<2 x i64>, <2 x i64>) nounwind readnone
+
+
+define <2 x double> @test_x86_sse41_round_pd(<2 x double> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse41_round_pd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <2 x i64> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.sse41.round.pd(<2 x double> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <2 x i64> zeroinitializer, <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.sse41.round.pd(<2 x double> %a0, i32 7) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.sse41.round.pd(<2 x double>, i32) nounwind readnone
+
+
+define <4 x float> @test_x86_sse41_round_ps(<4 x float> %a0) #0 {
+; CHECK-LABEL: @test_x86_sse41_round_ps(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP2:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP2]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse41.round.ps(<4 x float> [[A0:%.*]], i32 7)
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse41.round.ps(<4 x float> %a0, i32 7) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse41.round.ps(<4 x float>, i32) nounwind readnone
+
+
+define <2 x double> @test_x86_sse41_round_sd(<2 x double> %a0, <2 x double> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse41_round_sd(
+; CHECK-NEXT: [[TMP1:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <2 x i64> [[TMP1]], <2 x i64> [[TMP2]], <2 x i32> <i32 2, i32 1>
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.sse41.round.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1:%.*]], i32 7)
+; CHECK-NEXT: store <2 x i64> [[TMP3]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %res = call <2 x double> @llvm.x86.sse41.round.sd(<2 x double> %a0, <2 x double> %a1, i32 7) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.sse41.round.sd(<2 x double>, <2 x double>, i32) nounwind readnone
+
+
+define <2 x double> @test_x86_sse41_round_sd_load(<2 x double> %a0, <2 x double>* %a1) #0 {
+; CHECK-LABEL: @test_x86_sse41_round_sd_load(
+; CHECK-NEXT: [[TMP1:%.*]] = load i64, i64* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to i64*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* bitcast ([100 x i64]* @__msan_param_tls to <2 x i64>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP1]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP3:%.*]], label [[TMP4:%.*]], !prof [[PROF0]]
+; CHECK: 3:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 4:
+; CHECK-NEXT: [[A1B:%.*]] = load <2 x double>, <2 x double>* [[A1:%.*]], align 16
+; CHECK-NEXT: [[TMP5:%.*]] = ptrtoint <2 x double>* [[A1]] to i64
+; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
+; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to <2 x i64>*
+; CHECK-NEXT: [[_MSLD:%.*]] = load <2 x i64>, <2 x i64>* [[TMP7]], align 16
+; CHECK-NEXT: [[TMP8:%.*]] = shufflevector <2 x i64> [[TMP2]], <2 x i64> [[_MSLD]], <2 x i32> <i32 2, i32 1>
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.sse41.round.sd(<2 x double> [[A0:%.*]], <2 x double> [[A1B]], i32 7)
+; CHECK-NEXT: store <2 x i64> [[TMP8]], <2 x i64>* bitcast ([100 x i64]* @__msan_retval_tls to <2 x i64>*), align 8
+; CHECK-NEXT: ret <2 x double> [[RES]]
+;
+ %a1b = load <2 x double>, <2 x double>* %a1
+ %res = call <2 x double> @llvm.x86.sse41.round.sd(<2 x double> %a0, <2 x double> %a1b, i32 7) ; <<2 x double>> [#uses=1]
+ ret <2 x double> %res
+}
+
+
+define <4 x float> @test_x86_sse41_round_ss(<4 x float> %a0, <4 x float> %a1) #0 {
+; CHECK-LABEL: @test_x86_sse41_round_ss(
+; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([100 x i64]* @__msan_param_tls to <4 x i32>*), align 8
+; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* inttoptr (i64 add (i64 ptrtoint ([100 x i64]* @__msan_param_tls to i64), i64 16) to <4 x i32>*), align 8
+; CHECK-NEXT: call void @llvm.donothing()
+; CHECK-NEXT: [[TMP3:%.*]] = bitcast <4 x i32> [[TMP1]] to i128
+; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i128 [[TMP3]], 0
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP4:%.*]], label [[TMP5:%.*]], !prof [[PROF0]]
+; CHECK: 4:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 5:
+; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i32> [[TMP2]] to i128
+; CHECK-NEXT: [[_MSCMP1:%.*]] = icmp ne i128 [[TMP6]], 0
+; CHECK-NEXT: br i1 [[_MSCMP1]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF0]]
+; CHECK: 7:
+; CHECK-NEXT: call void @__msan_warning_with_origin_noreturn(i32 0) #[[ATTR3]]
+; CHECK-NEXT: unreachable
+; CHECK: 8:
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.sse41.round.ss(<4 x float> [[A0:%.*]], <4 x float> [[A1:%.*]], i32 7)
+; CHECK-NEXT: store <4 x i32> zeroinitializer, <4 x i32>* bitcast ([100 x i64]* @__msan_retval_tls to <4 x i32>*), align 8
+; CHECK-NEXT: ret <4 x float> [[RES]]
+;
+ %res = call <4 x float> @llvm.x86.sse41.round.ss(<4 x float> %a0, <4 x float> %a1, i32 7) ; <<4 x float>> [#uses=1]
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse41.round.ss(<4 x float>, <4 x float>, i32) nounwind readnone
+
+attributes #0 = { sanitize_memory }
More information about the llvm-commits
mailing list