[llvm] [msan] Fix 'Simplify 'maskedCheckAVXIndexShadow' #147839' (PR #148785)
via llvm-commits
llvm-commits at lists.llvm.org
Mon Jul 14 23:38:52 PDT 2025
llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT-->
@llvm/pr-subscribers-llvm-transforms
Author: Thurston Dang (thurstond)
<details>
<summary>Changes</summary>
https://github.com/llvm/llvm-project/pull/147839/ incorrectly checked the (lower bits of the) concrete value rather than the shadow.
---
Patch is 173.42 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/148785.diff
9 Files Affected:
- (modified) llvm/lib/Transforms/Instrumentation/MemorySanitizer.cpp (+2-1)
- (modified) llvm/test/Instrumentation/MemorySanitizer/X86/avx-intrinsics-x86.ll (+29-25)
- (modified) llvm/test/Instrumentation/MemorySanitizer/X86/avx512-intrinsics-upgrade.ll (+112-98)
- (modified) llvm/test/Instrumentation/MemorySanitizer/X86/avx512-intrinsics.ll (+121-105)
- (modified) llvm/test/Instrumentation/MemorySanitizer/X86/avx512bw-intrinsics-upgrade.ll (+31-27)
- (modified) llvm/test/Instrumentation/MemorySanitizer/X86/avx512bw-intrinsics.ll (+31-27)
- (modified) llvm/test/Instrumentation/MemorySanitizer/X86/avx512vl-intrinsics.ll (+113-93)
- (modified) llvm/test/Instrumentation/MemorySanitizer/X86/x86-vpermi2.ll (+13-13)
- (modified) llvm/test/Instrumentation/MemorySanitizer/i386/avx-intrinsics-i386.ll (+29-25)
``````````diff
diff --git a/llvm/lib/Transforms/Instrumentation/MemorySanitizer.cpp b/llvm/lib/Transforms/Instrumentation/MemorySanitizer.cpp
index 5f5200b2c9e62..592db79be0070 100644
--- a/llvm/lib/Transforms/Instrumentation/MemorySanitizer.cpp
+++ b/llvm/lib/Transforms/Instrumentation/MemorySanitizer.cpp
@@ -4322,8 +4322,9 @@ struct MemorySanitizerVisitor : public InstVisitor<MemorySanitizerVisitor> {
if (isa<Constant>(Idx))
return;
+ auto *IdxShadow = getShadow(Idx);
Value *Truncated = IRB.CreateTrunc(
- Idx,
+ IdxShadow,
FixedVectorType::get(Type::getIntNTy(*MS.C, Log2_64(IdxVectorSize)),
IdxVectorSize));
insertCheckShadow(Truncated, getOrigin(Idx), I);
diff --git a/llvm/test/Instrumentation/MemorySanitizer/X86/avx-intrinsics-x86.ll b/llvm/test/Instrumentation/MemorySanitizer/X86/avx-intrinsics-x86.ll
index 5bf529d7d32df..44545685b5121 100644
--- a/llvm/test/Instrumentation/MemorySanitizer/X86/avx-intrinsics-x86.ll
+++ b/llvm/test/Instrumentation/MemorySanitizer/X86/avx-intrinsics-x86.ll
@@ -948,19 +948,20 @@ 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>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: [[A1:%.*]] = load <2 x i64>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 16) to ptr), align 8
; CHECK-NEXT: call void @llvm.donothing()
-; CHECK-NEXT: [[TMP2:%.*]] = trunc <2 x i64> [[A1:%.*]] to <2 x i1>
+; CHECK-NEXT: [[TMP2:%.*]] = trunc <2 x i64> [[A1]] to <2 x i1>
; CHECK-NEXT: [[A0:%.*]] = bitcast <2 x i64> [[TMP1]] to <2 x double>
-; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.avx.vpermilvar.pd(<2 x double> [[A0]], <2 x i64> [[A1]])
+; CHECK-NEXT: [[RES:%.*]] = call <2 x double> @llvm.x86.avx.vpermilvar.pd(<2 x double> [[A0]], <2 x i64> [[A2:%.*]])
; CHECK-NEXT: [[TMP4:%.*]] = bitcast <2 x double> [[RES]] to <2 x i64>
; CHECK-NEXT: [[TMP6:%.*]] = bitcast <2 x i1> [[TMP2]] to i2
; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i2 [[TMP6]], 0
-; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF1]]
-; CHECK: 7:
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF1]]
+; CHECK: 8:
; CHECK-NEXT: call void @__msan_warning_noreturn()
; CHECK-NEXT: unreachable
-; CHECK: 8:
-; CHECK-NEXT: [[RES1:%.*]] = call <2 x double> @llvm.x86.avx.vpermilvar.pd(<2 x double> [[A2:%.*]], <2 x i64> [[A1]])
+; CHECK: 9:
+; CHECK-NEXT: [[RES1:%.*]] = call <2 x double> @llvm.x86.avx.vpermilvar.pd(<2 x double> [[A3:%.*]], <2 x i64> [[A2]])
; CHECK-NEXT: store <2 x i64> [[TMP4]], ptr @__msan_retval_tls, align 8
; CHECK-NEXT: ret <2 x double> [[RES1]]
;
@@ -973,19 +974,20 @@ declare <2 x double> @llvm.x86.avx.vpermilvar.pd(<2 x double>, <2 x i64>) nounwi
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>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: [[A1:%.*]] = load <4 x i64>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 32) to ptr), align 8
; CHECK-NEXT: call void @llvm.donothing()
-; CHECK-NEXT: [[TMP2:%.*]] = trunc <4 x i64> [[A1:%.*]] to <4 x i2>
+; CHECK-NEXT: [[TMP2:%.*]] = trunc <4 x i64> [[A1]] to <4 x i2>
; CHECK-NEXT: [[A0:%.*]] = bitcast <4 x i64> [[TMP1]] to <4 x double>
-; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double> [[A0]], <4 x i64> [[A1]])
+; CHECK-NEXT: [[RES:%.*]] = call <4 x double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double> [[A0]], <4 x i64> [[A2:%.*]])
; CHECK-NEXT: [[TMP4:%.*]] = bitcast <4 x double> [[RES]] to <4 x i64>
; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i2> [[TMP2]] to i8
; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i8 [[TMP6]], 0
-; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF1]]
-; CHECK: 7:
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF1]]
+; CHECK: 8:
; CHECK-NEXT: call void @__msan_warning_noreturn()
; CHECK-NEXT: unreachable
-; CHECK: 8:
-; CHECK-NEXT: [[RES1:%.*]] = call <4 x double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double> [[A2:%.*]], <4 x i64> [[A1]])
+; CHECK: 9:
+; CHECK-NEXT: [[RES1:%.*]] = call <4 x double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double> [[A3:%.*]], <4 x i64> [[A2]])
; CHECK-NEXT: store <4 x i64> [[TMP4]], ptr @__msan_retval_tls, align 8
; CHECK-NEXT: ret <4 x double> [[RES1]]
;
@@ -1012,19 +1014,20 @@ define <4 x double> @test_x86_avx_vpermilvar_pd_256_2(<4 x double> %a0) #0 {
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>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: [[A1:%.*]] = load <4 x i32>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 16) to ptr), align 8
; CHECK-NEXT: call void @llvm.donothing()
-; CHECK-NEXT: [[TMP2:%.*]] = trunc <4 x i32> [[A1:%.*]] to <4 x i2>
+; CHECK-NEXT: [[TMP2:%.*]] = trunc <4 x i32> [[A1]] to <4 x i2>
; CHECK-NEXT: [[A0:%.*]] = bitcast <4 x i32> [[TMP1]] to <4 x float>
-; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float> [[A0]], <4 x i32> [[A1]])
+; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float> [[A0]], <4 x i32> [[A2:%.*]])
; CHECK-NEXT: [[TMP4:%.*]] = bitcast <4 x float> [[RES]] to <4 x i32>
; CHECK-NEXT: [[TMP6:%.*]] = bitcast <4 x i2> [[TMP2]] to i8
; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i8 [[TMP6]], 0
-; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF1]]
-; CHECK: 7:
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF1]]
+; CHECK: 8:
; CHECK-NEXT: call void @__msan_warning_noreturn()
; CHECK-NEXT: unreachable
-; CHECK: 8:
-; CHECK-NEXT: [[RES1:%.*]] = call <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float> [[A2:%.*]], <4 x i32> [[A1]])
+; CHECK: 9:
+; CHECK-NEXT: [[RES1:%.*]] = call <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float> [[A3:%.*]], <4 x i32> [[A2]])
; CHECK-NEXT: store <4 x i32> [[TMP4]], ptr @__msan_retval_tls, align 8
; CHECK-NEXT: ret <4 x float> [[RES1]]
;
@@ -1047,7 +1050,7 @@ define <4 x float> @test_x86_avx_vpermilvar_ps_load(<4 x float> %a0, ptr %a1) #0
; CHECK-NEXT: [[TMP6:%.*]] = xor i64 [[TMP5]], 87960930222080
; CHECK-NEXT: [[TMP7:%.*]] = inttoptr i64 [[TMP6]] to ptr
; CHECK-NEXT: [[_MSLD:%.*]] = load <4 x i32>, ptr [[TMP7]], align 16
-; CHECK-NEXT: [[TMP8:%.*]] = trunc <4 x i32> [[A2]] to <4 x i2>
+; CHECK-NEXT: [[TMP8:%.*]] = trunc <4 x i32> [[_MSLD]] to <4 x i2>
; CHECK-NEXT: [[A0:%.*]] = bitcast <4 x i32> [[TMP2]] to <4 x float>
; CHECK-NEXT: [[RES:%.*]] = call <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float> [[A0]], <4 x i32> [[A2]])
; CHECK-NEXT: [[TMP10:%.*]] = bitcast <4 x float> [[RES]] to <4 x i32>
@@ -1072,19 +1075,20 @@ declare <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float>, <4 x i32>) nounwind
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>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: [[A1:%.*]] = load <8 x i32>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 32) to ptr), align 8
; CHECK-NEXT: call void @llvm.donothing()
-; CHECK-NEXT: [[TMP2:%.*]] = trunc <8 x i32> [[A1:%.*]] to <8 x i3>
+; CHECK-NEXT: [[TMP2:%.*]] = trunc <8 x i32> [[A1]] to <8 x i3>
; CHECK-NEXT: [[A0:%.*]] = bitcast <8 x i32> [[TMP1]] to <8 x float>
-; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.vpermilvar.ps.256(<8 x float> [[A0]], <8 x i32> [[A1]])
+; CHECK-NEXT: [[RES:%.*]] = call <8 x float> @llvm.x86.avx.vpermilvar.ps.256(<8 x float> [[A0]], <8 x i32> [[A2:%.*]])
; CHECK-NEXT: [[TMP4:%.*]] = bitcast <8 x float> [[RES]] to <8 x i32>
; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i3> [[TMP2]] to i24
; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i24 [[TMP6]], 0
-; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP7:%.*]], label [[TMP8:%.*]], !prof [[PROF1]]
-; CHECK: 7:
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF1]]
+; CHECK: 8:
; CHECK-NEXT: call void @__msan_warning_noreturn()
; CHECK-NEXT: unreachable
-; CHECK: 8:
-; CHECK-NEXT: [[RES1:%.*]] = call <8 x float> @llvm.x86.avx.vpermilvar.ps.256(<8 x float> [[A2:%.*]], <8 x i32> [[A1]])
+; CHECK: 9:
+; CHECK-NEXT: [[RES1:%.*]] = call <8 x float> @llvm.x86.avx.vpermilvar.ps.256(<8 x float> [[A3:%.*]], <8 x i32> [[A2]])
; CHECK-NEXT: store <8 x i32> [[TMP4]], ptr @__msan_retval_tls, align 8
; CHECK-NEXT: ret <8 x float> [[RES1]]
;
diff --git a/llvm/test/Instrumentation/MemorySanitizer/X86/avx512-intrinsics-upgrade.ll b/llvm/test/Instrumentation/MemorySanitizer/X86/avx512-intrinsics-upgrade.ll
index b292a8a9b1d66..74cb49b0f602a 100644
--- a/llvm/test/Instrumentation/MemorySanitizer/X86/avx512-intrinsics-upgrade.ll
+++ b/llvm/test/Instrumentation/MemorySanitizer/X86/avx512-intrinsics-upgrade.ll
@@ -8141,19 +8141,20 @@ declare <8 x double> @llvm.x86.avx512.mask.vpermilvar.pd.512(<8 x double>, <8 x
define <8 x double>@test_int_x86_avx512_vpermilvar_pd_512(<8 x double> %x0, <8 x i64> %x1, <8 x double> %x2) #0 {
; CHECK-LABEL: @test_int_x86_avx512_vpermilvar_pd_512(
; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i64>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: [[X1:%.*]] = load <8 x i64>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 64) to ptr), align 8
; CHECK-NEXT: call void @llvm.donothing()
-; CHECK-NEXT: [[TMP2:%.*]] = trunc <8 x i64> [[X1:%.*]] to <8 x i3>
+; CHECK-NEXT: [[TMP2:%.*]] = trunc <8 x i64> [[X1]] to <8 x i3>
; CHECK-NEXT: [[X0:%.*]] = bitcast <8 x i64> [[TMP1]] to <8 x double>
-; CHECK-NEXT: [[TMP7:%.*]] = call <8 x double> @llvm.x86.avx512.vpermilvar.pd.512(<8 x double> [[X0]], <8 x i64> [[X1]])
+; CHECK-NEXT: [[TMP7:%.*]] = call <8 x double> @llvm.x86.avx512.vpermilvar.pd.512(<8 x double> [[X0]], <8 x i64> [[X2:%.*]])
; CHECK-NEXT: [[TMP4:%.*]] = bitcast <8 x double> [[TMP7]] to <8 x i64>
; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x i3> [[TMP2]] to i24
; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i24 [[TMP6]], 0
-; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP9:%.*]], label [[TMP8:%.*]], !prof [[PROF1]]
-; CHECK: 7:
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF1]]
+; CHECK: 8:
; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR8]]
; CHECK-NEXT: unreachable
-; CHECK: 8:
-; CHECK-NEXT: [[TMP5:%.*]] = call <8 x double> @llvm.x86.avx512.vpermilvar.pd.512(<8 x double> [[X2:%.*]], <8 x i64> [[X1]])
+; CHECK: 9:
+; CHECK-NEXT: [[TMP5:%.*]] = call <8 x double> @llvm.x86.avx512.vpermilvar.pd.512(<8 x double> [[X3:%.*]], <8 x i64> [[X2]])
; CHECK-NEXT: store <8 x i64> [[TMP4]], ptr @__msan_retval_tls, align 8
; CHECK-NEXT: ret <8 x double> [[TMP5]]
;
@@ -8165,21 +8166,22 @@ define <8 x double>@test_int_x86_avx512_mask_vpermilvar_pd_512(<8 x double> %x0,
;
; CHECK-LABEL: @test_int_x86_avx512_mask_vpermilvar_pd_512(
; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i64>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: [[X1:%.*]] = load <8 x i64>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 64) to ptr), align 8
; CHECK-NEXT: [[TMP3:%.*]] = load i8, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 192) to ptr), align 8
; CHECK-NEXT: [[TMP4:%.*]] = load <8 x i64>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 128) to ptr), align 8
; CHECK-NEXT: call void @llvm.donothing()
-; CHECK-NEXT: [[TMP5:%.*]] = trunc <8 x i64> [[X1:%.*]] to <8 x i3>
+; CHECK-NEXT: [[TMP5:%.*]] = trunc <8 x i64> [[X1]] to <8 x i3>
; CHECK-NEXT: [[X0:%.*]] = bitcast <8 x i64> [[TMP1]] to <8 x double>
-; CHECK-NEXT: [[TMP9:%.*]] = call <8 x double> @llvm.x86.avx512.vpermilvar.pd.512(<8 x double> [[X0]], <8 x i64> [[X1]])
+; CHECK-NEXT: [[TMP9:%.*]] = call <8 x double> @llvm.x86.avx512.vpermilvar.pd.512(<8 x double> [[X0]], <8 x i64> [[X4:%.*]])
; CHECK-NEXT: [[TMP6:%.*]] = bitcast <8 x double> [[TMP9]] to <8 x i64>
; CHECK-NEXT: [[TMP8:%.*]] = bitcast <8 x i3> [[TMP5]] to i24
; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i24 [[TMP8]], 0
; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP19:%.*]], label [[TMP20:%.*]], !prof [[PROF1]]
-; CHECK: 9:
+; CHECK: 10:
; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR8]]
; CHECK-NEXT: unreachable
-; CHECK: 10:
-; CHECK-NEXT: [[TMP7:%.*]] = call <8 x double> @llvm.x86.avx512.vpermilvar.pd.512(<8 x double> [[X4:%.*]], <8 x i64> [[X1]])
+; CHECK: 11:
+; CHECK-NEXT: [[TMP7:%.*]] = call <8 x double> @llvm.x86.avx512.vpermilvar.pd.512(<8 x double> [[X5:%.*]], <8 x i64> [[X4]])
; CHECK-NEXT: [[TMP10:%.*]] = bitcast i8 [[TMP3]] to <8 x i1>
; CHECK-NEXT: [[TMP11:%.*]] = bitcast i8 [[X3:%.*]] to <8 x i1>
; CHECK-NEXT: [[TMP12:%.*]] = select <8 x i1> [[TMP11]], <8 x i64> [[TMP6]], <8 x i64> [[TMP4]]
@@ -8201,20 +8203,21 @@ define <8 x double>@test_int_x86_avx512_maskz_vpermilvar_pd_512(<8 x double> %x0
;
; CHECK-LABEL: @test_int_x86_avx512_maskz_vpermilvar_pd_512(
; CHECK-NEXT: [[TMP1:%.*]] = load <8 x i64>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: [[X1:%.*]] = load <8 x i64>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 64) to ptr), align 8
; CHECK-NEXT: [[TMP3:%.*]] = load i8, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 192) to ptr), align 8
; CHECK-NEXT: call void @llvm.donothing()
-; CHECK-NEXT: [[TMP4:%.*]] = trunc <8 x i64> [[X1:%.*]] to <8 x i3>
+; CHECK-NEXT: [[TMP4:%.*]] = trunc <8 x i64> [[X1]] to <8 x i3>
; CHECK-NEXT: [[X0:%.*]] = bitcast <8 x i64> [[TMP1]] to <8 x double>
-; CHECK-NEXT: [[TMP8:%.*]] = call <8 x double> @llvm.x86.avx512.vpermilvar.pd.512(<8 x double> [[X0]], <8 x i64> [[X1]])
+; CHECK-NEXT: [[TMP8:%.*]] = call <8 x double> @llvm.x86.avx512.vpermilvar.pd.512(<8 x double> [[X0]], <8 x i64> [[X2:%.*]])
; CHECK-NEXT: [[TMP5:%.*]] = bitcast <8 x double> [[TMP8]] to <8 x i64>
; CHECK-NEXT: [[TMP7:%.*]] = bitcast <8 x i3> [[TMP4]] to i24
; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i24 [[TMP7]], 0
; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP17:%.*]], label [[TMP18:%.*]], !prof [[PROF1]]
-; CHECK: 8:
+; CHECK: 9:
; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR8]]
; CHECK-NEXT: unreachable
-; CHECK: 9:
-; CHECK-NEXT: [[TMP6:%.*]] = call <8 x double> @llvm.x86.avx512.vpermilvar.pd.512(<8 x double> [[X2:%.*]], <8 x i64> [[X1]])
+; CHECK: 10:
+; CHECK-NEXT: [[TMP6:%.*]] = call <8 x double> @llvm.x86.avx512.vpermilvar.pd.512(<8 x double> [[X4:%.*]], <8 x i64> [[X2]])
; CHECK-NEXT: [[TMP9:%.*]] = bitcast i8 [[TMP3]] to <8 x i1>
; CHECK-NEXT: [[TMP10:%.*]] = bitcast i8 [[X3:%.*]] to <8 x i1>
; CHECK-NEXT: [[TMP11:%.*]] = select <8 x i1> [[TMP10]], <8 x i64> [[TMP5]], <8 x i64> zeroinitializer
@@ -8236,19 +8239,20 @@ declare <16 x float> @llvm.x86.avx512.mask.vpermilvar.ps.512(<16 x float>, <16 x
define <16 x float>@test_int_x86_avx512_vpermilvar_ps_512(<16 x float> %x0, <16 x i32> %x1, <16 x float> %x2) #0 {
; CHECK-LABEL: @test_int_x86_avx512_vpermilvar_ps_512(
; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: [[X1:%.*]] = load <16 x i32>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 64) to ptr), align 8
; CHECK-NEXT: call void @llvm.donothing()
-; CHECK-NEXT: [[TMP2:%.*]] = trunc <16 x i32> [[X1:%.*]] to <16 x i4>
+; CHECK-NEXT: [[TMP2:%.*]] = trunc <16 x i32> [[X1]] to <16 x i4>
; CHECK-NEXT: [[X0:%.*]] = bitcast <16 x i32> [[TMP1]] to <16 x float>
-; CHECK-NEXT: [[TMP7:%.*]] = call <16 x float> @llvm.x86.avx512.vpermilvar.ps.512(<16 x float> [[X0]], <16 x i32> [[X1]])
+; CHECK-NEXT: [[TMP7:%.*]] = call <16 x float> @llvm.x86.avx512.vpermilvar.ps.512(<16 x float> [[X0]], <16 x i32> [[X2:%.*]])
; CHECK-NEXT: [[TMP4:%.*]] = bitcast <16 x float> [[TMP7]] to <16 x i32>
; CHECK-NEXT: [[TMP6:%.*]] = bitcast <16 x i4> [[TMP2]] to i64
; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP6]], 0
-; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP9:%.*]], label [[TMP8:%.*]], !prof [[PROF1]]
-; CHECK: 7:
+; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP8:%.*]], label [[TMP9:%.*]], !prof [[PROF1]]
+; CHECK: 8:
; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR8]]
; CHECK-NEXT: unreachable
-; CHECK: 8:
-; CHECK-NEXT: [[TMP5:%.*]] = call <16 x float> @llvm.x86.avx512.vpermilvar.ps.512(<16 x float> [[X2:%.*]], <16 x i32> [[X1]])
+; CHECK: 9:
+; CHECK-NEXT: [[TMP5:%.*]] = call <16 x float> @llvm.x86.avx512.vpermilvar.ps.512(<16 x float> [[X3:%.*]], <16 x i32> [[X2]])
; CHECK-NEXT: store <16 x i32> [[TMP4]], ptr @__msan_retval_tls, align 8
; CHECK-NEXT: ret <16 x float> [[TMP5]]
;
@@ -8260,21 +8264,22 @@ define <16 x float>@test_int_x86_avx512_mask_vpermilvar_ps_512(<16 x float> %x0,
;
; CHECK-LABEL: @test_int_x86_avx512_mask_vpermilvar_ps_512(
; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: [[X1:%.*]] = load <16 x i32>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 64) to ptr), align 8
; CHECK-NEXT: [[TMP3:%.*]] = load i16, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 192) to ptr), align 8
; CHECK-NEXT: [[TMP4:%.*]] = load <16 x i32>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 128) to ptr), align 8
; CHECK-NEXT: call void @llvm.donothing()
-; CHECK-NEXT: [[TMP5:%.*]] = trunc <16 x i32> [[X1:%.*]] to <16 x i4>
+; CHECK-NEXT: [[TMP5:%.*]] = trunc <16 x i32> [[X1]] to <16 x i4>
; CHECK-NEXT: [[X0:%.*]] = bitcast <16 x i32> [[TMP1]] to <16 x float>
-; CHECK-NEXT: [[TMP9:%.*]] = call <16 x float> @llvm.x86.avx512.vpermilvar.ps.512(<16 x float> [[X0]], <16 x i32> [[X1]])
+; CHECK-NEXT: [[TMP9:%.*]] = call <16 x float> @llvm.x86.avx512.vpermilvar.ps.512(<16 x float> [[X0]], <16 x i32> [[X4:%.*]])
; CHECK-NEXT: [[TMP6:%.*]] = bitcast <16 x float> [[TMP9]] to <16 x i32>
; CHECK-NEXT: [[TMP8:%.*]] = bitcast <16 x i4> [[TMP5]] to i64
; CHECK-NEXT: [[_MSCMP:%.*]] = icmp ne i64 [[TMP8]], 0
; CHECK-NEXT: br i1 [[_MSCMP]], label [[TMP19:%.*]], label [[TMP20:%.*]], !prof [[PROF1]]
-; CHECK: 9:
+; CHECK: 10:
; CHECK-NEXT: call void @__msan_warning_noreturn() #[[ATTR8]]
; CHECK-NEXT: unreachable
-; CHECK: 10:
-; CHECK-NEXT: [[TMP7:%.*]] = call <16 x float> @llvm.x86.avx512.vpermilvar.ps.512(<16 x float> [[X4:%.*]], <16 x i32> [[X1]])
+; CHECK: 11:
+; CHECK-NEXT: [[TMP7:%.*]] = call <16 x float> @llvm.x86.avx512.vpermilvar.ps.512(<16 x float> [[X5:%.*]], <16 x i32> [[X4]])
; CHECK-NEXT: [[TMP10:%.*]] = bitcast i16 [[TMP3]] to <16 x i1>
; CHECK-NEXT: [[TMP11:%.*]] = bitcast i16 [[X3:%.*]] to <16 x i1>
; CHECK-NEXT: [[TMP12:%.*]] = select <16 x i1> [[TMP11]], <16 x i32> [[TMP6]], <16 x i32> [[TMP4]]
@@ -8297,20 +8302,21 @@ define <16 x float>@test_int_x86_avx512_maskz_vpermilvar_ps_512(<16 x float> %x0
;
; CHECK-LABEL: @test_int_x86_avx512_maskz_vpermilvar_ps_512(
; CHECK-NEXT: [[TMP1:%.*]] = load <16 x i32>, ptr @__msan_param_tls, align 8
+; CHECK-NEXT: [[X1:%.*]] = load <16 x i32>, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 64) to ptr), align 8
; CHECK-NEXT: [[TMP3:%.*]] = load i16, ptr inttoptr (i64 add (i64 ptrtoint (ptr @__msan_param_tls to i64), i64 128) to ptr), align 8
; CHECK-NEXT: call void @llvm.donothing()
-; CHECK-NEXT: [[TMP4:%.*]] = trunc <16 x i32> [[X1:%....
[truncated]
``````````
</details>
https://github.com/llvm/llvm-project/pull/148785
More information about the llvm-commits
mailing list