[X86] Fix a few intrinsics that were ignoring their rounding mode argument and hardcoded _MM_FROUND_CUR_DIRECTION internally.

I believe these have been broken since their introduction into clang.

I've enhanced the tests for these intrinsics to using a real rounding mode and checking all the intrinsic arguments instead of just the name.

llvm-svn: 336498

2 files changed, 49 insertions, 49 deletions

diff --git a/clang/lib/Headers/avx512fintrin.h b/clang/lib/Headers/avx512fintrin.h
index a3ca7e730a1..5e86aae70b1 100644
--- a/clang/lib/Headers/avx512fintrin.h
+++ b/clang/lib/Headers/avx512fintrin.h
@@ -6597,7 +6597,7 @@ _mm_maskz_scalef_ss (__mmask8 __U, __m128 __A, __m128 __B)
                                              (__v4sf)(__m128)(B), \
                                              (__v4sf)_mm_setzero_ps(), \
                                              (__mmask8)(U), \
-                                             _MM_FROUND_CUR_DIRECTION)
+                                             (int)(R))
 
 static __inline__ __m512i __DEFAULT_FN_ATTRS
 _mm512_srai_epi32(__m512i __A, int __B)
@@ -7826,7 +7826,7 @@ _mm_maskz_fmadd_ss (__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
   (__m128)__builtin_ia32_vfmaddss3_maskz((__v4sf)(__m128)(A), \
                                          (__v4sf)(__m128)(B), \
                                          (__v4sf)(__m128)(C), (__mmask8)(U), \
-                                         _MM_FROUND_CUR_DIRECTION)
+                                         (int)(R))
 
 static __inline__ __m128 __DEFAULT_FN_ATTRS
 _mm_mask3_fmadd_ss (__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
@@ -7988,7 +7988,7 @@ _mm_maskz_fnmsub_ss (__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
   (__m128)__builtin_ia32_vfmaddss3_maskz((__v4sf)(__m128)(A), \
                                          -(__v4sf)(__m128)(B), \
                                          -(__v4sf)(__m128)(C), (__mmask8)(U), \
-                                         _MM_FROUND_CUR_DIRECTION)
+                                         (int)(R))
 
 static __inline__ __m128 __DEFAULT_FN_ATTRS
 _mm_mask3_fnmsub_ss (__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
@@ -8042,7 +8042,7 @@ _mm_maskz_fmadd_sd (__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
   (__m128d)__builtin_ia32_vfmaddsd3_maskz((__v2df)(__m128d)(A), \
                                           (__v2df)(__m128d)(B), \
                                           (__v2df)(__m128d)(C), (__mmask8)(U), \
-                                          _MM_FROUND_CUR_DIRECTION)
+                                          (int)(R))
 
 static __inline__ __m128d __DEFAULT_FN_ATTRS
 _mm_mask3_fmadd_sd (__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
@@ -8205,7 +8205,7 @@ _mm_maskz_fnmsub_sd (__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
                                           -(__v2df)(__m128d)(B), \
                                           -(__v2df)(__m128d)(C), \
                                           (__mmask8)(U), \
-                                          _MM_FROUND_CUR_DIRECTION)
+                                          (int)(R))
 
 static __inline__ __m128d __DEFAULT_FN_ATTRS
 _mm_mask3_fnmsub_sd (__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
diff --git a/clang/test/CodeGen/avx512f-builtins.c b/clang/test/CodeGen/avx512f-builtins.c
index 0a817938f68..d90f5d03ec1 100644
--- a/clang/test/CodeGen/avx512f-builtins.c
+++ b/clang/test/CodeGen/avx512f-builtins.c
@@ -5433,8 +5433,8 @@ __m512 test_mm512_maskz_scalef_ps(__mmask16 __U, __m512 __A, __m512 __B) {
 
 __m128d test_mm_scalef_round_sd(__m128d __A, __m128d __B) {
   // CHECK-LABEL: @test_mm_scalef_round_sd
-  // CHECK: @llvm.x86.avx512.mask.scalef
-  return _mm_scalef_round_sd(__A, __B, _MM_FROUND_CUR_DIRECTION); 
+  // CHECK: @llvm.x86.avx512.mask.scalef.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %2, i8 -1, i32 8)
+  return _mm_scalef_round_sd(__A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128d test_mm_scalef_sd(__m128d __A, __m128d __B) {
@@ -5451,8 +5451,8 @@ __m128d test_mm_mask_scalef_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d _
 
 __m128d test_mm_mask_scalef_round_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B){
   // CHECK-LABEL: @test_mm_mask_scalef_round_sd
-  // CHECK: @llvm.x86.avx512.mask.scalef.sd
-    return _mm_mask_scalef_round_sd(__W, __U, __A, __B, _MM_FROUND_CUR_DIRECTION);
+  // CHECK: @llvm.x86.avx512.mask.scalef.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
+    return _mm_mask_scalef_round_sd(__W, __U, __A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128d test_mm_maskz_scalef_sd(__mmask8 __U, __m128d __A, __m128d __B){
@@ -5463,14 +5463,14 @@ __m128d test_mm_maskz_scalef_sd(__mmask8 __U, __m128d __A, __m128d __B){
 
 __m128d test_mm_maskz_scalef_round_sd(__mmask8 __U, __m128d __A, __m128d __B){
   // CHECK-LABEL: @test_mm_maskz_scalef_round_sd
-  // CHECK: @llvm.x86.avx512.mask.scalef.sd
-    return _mm_maskz_scalef_round_sd(__U, __A, __B, _MM_FROUND_CUR_DIRECTION);
+  // CHECK: @llvm.x86.avx512.mask.scalef.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
+    return _mm_maskz_scalef_round_sd(__U, __A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128 test_mm_scalef_round_ss(__m128 __A, __m128 __B) {
   // CHECK-LABEL: @test_mm_scalef_round_ss
-  // CHECK: @llvm.x86.avx512.mask.scalef.ss
-  return _mm_scalef_round_ss(__A, __B, _MM_FROUND_CUR_DIRECTION); 
+  // CHECK: @llvm.x86.avx512.mask.scalef.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 -1, i32 8)
+  return _mm_scalef_round_ss(__A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128 test_mm_scalef_ss(__m128 __A, __m128 __B) {
@@ -5487,8 +5487,8 @@ __m128 test_mm_mask_scalef_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B){
 
 __m128 test_mm_mask_scalef_round_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B){
   // CHECK-LABEL: @test_mm_mask_scalef_round_ss
-  // CHECK: @llvm.x86.avx512.mask.scalef.ss
-    return _mm_mask_scalef_round_ss(__W, __U, __A, __B, _MM_FROUND_CUR_DIRECTION);
+  // CHECK: @llvm.x86.avx512.mask.scalef.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
+    return _mm_mask_scalef_round_ss(__W, __U, __A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128 test_mm_maskz_scalef_ss(__mmask8 __U, __m128 __A, __m128 __B){
@@ -5499,8 +5499,8 @@ __m128 test_mm_maskz_scalef_ss(__mmask8 __U, __m128 __A, __m128 __B){
 
 __m128 test_mm_maskz_scalef_round_ss(__mmask8 __U, __m128 __A, __m128 __B){
   // CHECK-LABEL: @test_mm_maskz_scalef_round_ss
-  // CHECK: @llvm.x86.avx512.mask.scalef.ss
-    return _mm_maskz_scalef_round_ss(__U, __A, __B, _MM_FROUND_CUR_DIRECTION);
+  // CHECK: @llvm.x86.avx512.mask.scalef.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
+    return _mm_maskz_scalef_round_ss(__U, __A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m512i test_mm512_srai_epi32(__m512i __A) {
@@ -7232,13 +7232,13 @@ __m128 test_mm_mask_fmadd_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B){
 
 __m128 test_mm_fmadd_round_ss(__m128 __A, __m128 __B, __m128 __C){
   // CHECK-LABEL: @test_mm_fmadd_round_ss
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 -1, i32 8)
   return _mm_fmadd_round_ss(__A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128 test_mm_mask_fmadd_round_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B){
   // CHECK-LABEL: @test_mm_mask_fmadd_round_ss
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask_fmadd_round_ss(__W, __U, __A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7266,7 +7266,7 @@ __m128 test_mm_maskz_fmadd_ss(__mmask8 __U, __m128 __A, __m128 __B, __m128 __C){
 
 __m128 test_mm_maskz_fmadd_round_ss(__mmask8 __U, __m128 __A, __m128 __B, __m128 __C){
   // CHECK-LABEL: @test_mm_maskz_fmadd_round_ss
-  // CHECK: @llvm.x86.avx512.maskz.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.maskz.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_maskz_fmadd_round_ss(__U, __A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7295,7 +7295,7 @@ __m128 test_mm_mask3_fmadd_ss(__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U){
 
 __m128 test_mm_mask3_fmadd_round_ss(__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U){
   // CHECK-LABEL: @test_mm_mask3_fmadd_round_ss
-  // CHECK: @llvm.x86.avx512.mask3.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.mask3.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask3_fmadd_round_ss(__W, __X, __Y, __U, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7325,13 +7325,13 @@ __m128 test_mm_mask_fmsub_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B){
 
 __m128 test_mm_fmsub_round_ss(__m128 __A, __m128 __B, __m128 __C){
   // CHECK-LABEL: @test_mm_fmsub_round_ss
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 -1, i32 8)
   return _mm_fmsub_round_ss(__A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128 test_mm_mask_fmsub_round_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B){
   // CHECK-LABEL: @test_mm_mask_fmsub_round_ss
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask_fmsub_round_ss(__W, __U, __A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7360,7 +7360,7 @@ __m128 test_mm_maskz_fmsub_ss(__mmask8 __U, __m128 __A, __m128 __B, __m128 __C){
 
 __m128 test_mm_maskz_fmsub_round_ss(__mmask8 __U, __m128 __A, __m128 __B, __m128 __C){
   // CHECK-LABEL: @test_mm_maskz_fmsub_round_ss
-  // CHECK: @llvm.x86.avx512.maskz.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.maskz.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_maskz_fmsub_round_ss(__U, __A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7390,7 +7390,7 @@ __m128 test_mm_mask3_fmsub_ss(__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U){
 
 __m128 test_mm_mask3_fmsub_round_ss(__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U){
   // CHECK-LABEL: @test_mm_mask3_fmsub_round_ss
-  // CHECK: @llvm.x86.avx512.mask3.vfmsub.ss
+  // CHECK: @llvm.x86.avx512.mask3.vfmsub.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask3_fmsub_round_ss(__W, __X, __Y, __U, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7420,13 +7420,13 @@ __m128 test_mm_mask_fnmadd_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B){
 
 __m128 test_mm_fnmadd_round_ss(__m128 __A, __m128 __B, __m128 __C){
   // CHECK-LABEL: @test_mm_fnmadd_round_ss
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 -1, i32 8)
   return _mm_fnmadd_round_ss(__A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128 test_mm_mask_fnmadd_round_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B){
   // CHECK-LABEL: @test_mm_mask_fnmadd_round_ss
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask_fnmadd_round_ss(__W, __U, __A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7455,7 +7455,7 @@ __m128 test_mm_maskz_fnmadd_ss(__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
 
 __m128 test_mm_maskz_fnmadd_round_ss(__mmask8 __U, __m128 __A, __m128 __B, __m128 __C){
   // CHECK-LABEL: @test_mm_maskz_fnmadd_round_ss
-  // CHECK: @llvm.x86.avx512.maskz.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.maskz.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_maskz_fnmadd_round_ss(__U, __A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7485,7 +7485,7 @@ __m128 test_mm_mask3_fnmadd_ss(__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
 
 __m128 test_mm_mask3_fnmadd_round_ss(__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U){
   // CHECK-LABEL: @test_mm_mask3_fnmadd_round_ss
-  // CHECK: @llvm.x86.avx512.mask3.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.mask3.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask3_fnmadd_round_ss(__W, __X, __Y, __U, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7516,13 +7516,13 @@ __m128 test_mm_mask_fnmsub_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B){
 
 __m128 test_mm_fnmsub_round_ss(__m128 __A, __m128 __B, __m128 __C){
   // CHECK-LABEL: @test_mm_fnmsub_round_ss
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 -1, i32 8)
   return _mm_fnmsub_round_ss(__A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128 test_mm_mask_fnmsub_round_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B){
   // CHECK-LABEL: @test_mm_mask_fnmsub_round_ss
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask_fnmsub_round_ss(__W, __U, __A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7552,7 +7552,7 @@ __m128 test_mm_maskz_fnmsub_ss(__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
 
 __m128 test_mm_maskz_fnmsub_round_ss(__mmask8 __U, __m128 __A, __m128 __B, __m128 __C){
   // CHECK-LABEL: @test_mm_maskz_fnmsub_round_ss
-  // CHECK: @llvm.x86.avx512.maskz.vfmadd.ss
+  // CHECK: @llvm.x86.avx512.maskz.vfmadd.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_maskz_fnmsub_round_ss(__U, __A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7583,7 +7583,7 @@ __m128 test_mm_mask3_fnmsub_ss(__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
 
 __m128 test_mm_mask3_fnmsub_round_ss(__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U){
   // CHECK-LABEL: @test_mm_mask3_fnmsub_round_ss
-  // CHECK: @llvm.x86.avx512.mask3.vfmsub.ss
+  // CHECK: @llvm.x86.avx512.mask3.vfmsub.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask3_fnmsub_round_ss(__W, __X, __Y, __U, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7612,13 +7612,13 @@ __m128d test_mm_mask_fmadd_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __
 
 __m128d test_mm_fmadd_round_sd(__m128d __A, __m128d __B, __m128d __C){
   // CHECK-LABEL: @test_mm_fmadd_round_sd
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 -1, i32 8)
   return _mm_fmadd_round_sd(__A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128d test_mm_mask_fmadd_round_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B){
   // CHECK-LABEL: @test_mm_mask_fmadd_round_sd
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask_fmadd_round_sd(__W, __U, __A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7646,7 +7646,7 @@ __m128d test_mm_maskz_fmadd_sd(__mmask8 __U, __m128d __A, __m128d __B, __m128d _
 
 __m128d test_mm_maskz_fmadd_round_sd(__mmask8 __U, __m128d __A, __m128d __B, __m128d __C){
   // CHECK-LABEL: @test_mm_maskz_fmadd_round_sd
-  // CHECK: @llvm.x86.avx512.maskz.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.maskz.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_maskz_fmadd_round_sd(__U, __A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7675,7 +7675,7 @@ __m128d test_mm_mask3_fmadd_sd(__m128d __W, __m128d __X, __m128d __Y, __mmask8 _
 
 __m128d test_mm_mask3_fmadd_round_sd(__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U){
   // CHECK-LABEL: @test_mm_mask3_fmadd_round_sd
-  // CHECK: @llvm.x86.avx512.mask3.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.mask3.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask3_fmadd_round_sd(__W, __X, __Y, __U, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7705,13 +7705,13 @@ __m128d test_mm_mask_fmsub_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __
 
 __m128d test_mm_fmsub_round_sd(__m128d __A, __m128d __B, __m128d __C){
   // CHECK-LABEL: @test_mm_fmsub_round_sd
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 -1, i32 8)
   return _mm_fmsub_round_sd(__A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128d test_mm_mask_fmsub_round_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B){
   // CHECK-LABEL: @test_mm_mask_fmsub_round_sd
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask_fmsub_round_sd(__W, __U, __A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7740,7 +7740,7 @@ __m128d test_mm_maskz_fmsub_sd(__mmask8 __U, __m128d __A, __m128d __B, __m128d _
 
 __m128d test_mm_maskz_fmsub_round_sd(__mmask8 __U, __m128d __A, __m128d __B, __m128d __C){
   // CHECK-LABEL: @test_mm_maskz_fmsub_round_sd
-  // CHECK: @llvm.x86.avx512.maskz.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.maskz.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_maskz_fmsub_round_sd(__U, __A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7770,7 +7770,7 @@ __m128d test_mm_mask3_fmsub_sd(__m128d __W, __m128d __X, __m128d __Y, __mmask8 _
 
 __m128d test_mm_mask3_fmsub_round_sd(__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U){
   // CHECK-LABEL: @test_mm_mask3_fmsub_round_sd
-  // CHECK: @llvm.x86.avx512.mask3.vfmsub.sd
+  // CHECK: @llvm.x86.avx512.mask3.vfmsub.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask3_fmsub_round_sd(__W, __X, __Y, __U, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7800,13 +7800,13 @@ __m128d test_mm_mask_fnmadd_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d _
 
 __m128d test_mm_fnmadd_round_sd(__m128d __A, __m128d __B, __m128d __C){
   // CHECK-LABEL: @test_mm_fnmadd_round_sd
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 -1, i32 8)
   return _mm_fnmadd_round_sd(__A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128d test_mm_mask_fnmadd_round_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B){
   // CHECK-LABEL: @test_mm_mask_fnmadd_round_sd
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask_fnmadd_round_sd(__W, __U, __A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7835,7 +7835,7 @@ __m128d test_mm_maskz_fnmadd_sd(__mmask8 __U, __m128d __A, __m128d __B, __m128d
 
 __m128d test_mm_maskz_fnmadd_round_sd(__mmask8 __U, __m128d __A, __m128d __B, __m128d __C){
   // CHECK-LABEL: @test_mm_maskz_fnmadd_round_sd
-  // CHECK: @llvm.x86.avx512.maskz.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.maskz.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_maskz_fnmadd_round_sd(__U, __A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7865,7 +7865,7 @@ __m128d test_mm_mask3_fnmadd_sd(__m128d __W, __m128d __X, __m128d __Y, __mmask8
 
 __m128d test_mm_mask3_fnmadd_round_sd(__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U){
   // CHECK-LABEL: @test_mm_mask3_fnmadd_round_sd
-  // CHECK: @llvm.x86.avx512.mask3.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.mask3.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask3_fnmadd_round_sd(__W, __X, __Y, __U, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7896,13 +7896,13 @@ __m128d test_mm_mask_fnmsub_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d _
 
 __m128d test_mm_fnmsub_round_sd(__m128d __A, __m128d __B, __m128d __C){
   // CHECK-LABEL: @test_mm_fnmsub_round_sd
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 -1, i32 8)
   return _mm_fnmsub_round_sd(__A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
 __m128d test_mm_mask_fnmsub_round_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B){
   // CHECK-LABEL: @test_mm_mask_fnmsub_round_sd
-  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.mask.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask_fnmsub_round_sd(__W, __U, __A, __B, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7932,7 +7932,7 @@ __m128d test_mm_maskz_fnmsub_sd(__mmask8 __U, __m128d __A, __m128d __B, __m128d
 
 __m128d test_mm_maskz_fnmsub_round_sd(__mmask8 __U, __m128d __A, __m128d __B, __m128d __C){
   // CHECK-LABEL: @test_mm_maskz_fnmsub_round_sd
-  // CHECK: @llvm.x86.avx512.maskz.vfmadd.sd
+  // CHECK: @llvm.x86.avx512.maskz.vfmadd.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_maskz_fnmsub_round_sd(__U, __A, __B, __C, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }
 
@@ -7963,7 +7963,7 @@ __m128d test_mm_mask3_fnmsub_sd(__m128d __W, __m128d __X, __m128d __Y, __mmask8
 
 __m128d test_mm_mask3_fnmsub_round_sd(__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U){
   // CHECK-LABEL: @test_mm_mask3_fnmsub_round_sd
-  // CHECK: @llvm.x86.avx512.mask3.vfmsub.sd
+  // CHECK: @llvm.x86.avx512.mask3.vfmsub.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 %{{.*}}, i32 8)
   return _mm_mask3_fnmsub_round_sd(__W, __X, __Y, __U, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
 }