Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[AVX10.2] Fix wrong intrinsic names after rename #126390

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

[AVX10.2] Fix wrong intrinsic names after rename #126390

wants to merge 2 commits into from
+38 −39

Conversation

mikolaj-pirog
Copy link
Contributor

In my previous PR (#123656) to update the names of AVX10.2 intrinsics and mnemonics, I have erroneously deleted _ph from few intrinsics. This PR corrects this.

@llvmbot llvmbot added clang Clang issues not falling into any other category backend:X86 clang:headers Headers provided by Clang, e.g. for intrinsics labels Feb 9, 2025
@llvmbot
Copy link
Member

llvmbot commented Feb 9, 2025 •
edited
Loading

@llvm/pr-subscribers-clang

@llvm/pr-subscribers-backend-x86

Author: Mikołaj Piróg (mikolaj-pirog)

Changes

In my previous PR (#123656) to update the names of AVX10.2 intrinsics and mnemonics, I have erroneously deleted _ph from few intrinsics. This PR corrects this.

Full diff: https://github.com/llvm/llvm-project/pull/126390.diff

4 Files Affected:

  • (modified) clang/lib/Headers/avx10_2_512convertintrin.h (+3-3)
  • (modified) clang/lib/Headers/avx10_2convertintrin.h (+6-6)
  • (modified) clang/test/CodeGen/X86/avx10_2_512convert-builtins.c (+9-9)
  • (modified) clang/test/CodeGen/X86/avx10_2convert-builtins.c (+18-18)
diff --git a/clang/lib/Headers/avx10_2_512convertintrin.h b/clang/lib/Headers/avx10_2_512convertintrin.h
index 0b5fca5cda5228f..516ccc68672d636 100644
--- a/clang/lib/Headers/avx10_2_512convertintrin.h
+++ b/clang/lib/Headers/avx10_2_512convertintrin.h
@@ -213,19 +213,19 @@ _mm512_maskz_cvts2ph_hf8(__mmask64 __U, __m512h __A, __m512h __B) {
       (__v64qi)(__m512i)_mm512_setzero_si512());
 }
 
-static __inline__ __m512h __DEFAULT_FN_ATTRS512 _mm512_cvthf8(__m256i __A) {
+static __inline__ __m512h __DEFAULT_FN_ATTRS512 _mm512_cvthf8_ph(__m256i __A) {
   return (__m512h)__builtin_ia32_vcvthf8_2ph512_mask(
       (__v32qi)__A, (__v32hf)(__m512h)_mm512_undefined_ph(), (__mmask32)-1);
 }
 
 static __inline__ __m512h __DEFAULT_FN_ATTRS512
-_mm512_mask_cvthf8(__m512h __W, __mmask32 __U, __m256i __A) {
+_mm512_mask_cvthf8_ph(__m512h __W, __mmask32 __U, __m256i __A) {
   return (__m512h)__builtin_ia32_vcvthf8_2ph512_mask(
       (__v32qi)__A, (__v32hf)(__m512h)__W, (__mmask32)__U);
 }
 
 static __inline__ __m512h __DEFAULT_FN_ATTRS512
-_mm512_maskz_cvthf8(__mmask32 __U, __m256i __A) {
+_mm512_maskz_cvthf8_ph(__mmask32 __U, __m256i __A) {
   return (__m512h)__builtin_ia32_vcvthf8_2ph512_mask(
       (__v32qi)__A, (__v32hf)(__m512h)_mm512_setzero_ph(), (__mmask32)__U);
 }
diff --git a/clang/lib/Headers/avx10_2convertintrin.h b/clang/lib/Headers/avx10_2convertintrin.h
index c67a5b890f1957d..cbc02e37b0cc245 100644
--- a/clang/lib/Headers/avx10_2convertintrin.h
+++ b/clang/lib/Headers/avx10_2convertintrin.h
@@ -381,37 +381,37 @@ _mm256_maskz_cvts2ph_hf8(__mmask32 __U, __m256h __A, __m256h __B) {
       (__v32qi)(__m256i)_mm256_setzero_si256());
 }
 
-static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_cvthf8(__m128i __A) {
+static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_cvthf8_ph(__m128i __A) {
   return (__m128h)__builtin_ia32_vcvthf8_2ph128_mask(
       (__v16qi)__A, (__v8hf)(__m128h)_mm_undefined_ph(), (__mmask8)-1);
 }
 
-static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_mask_cvthf8(__m128h __W,
+static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_mask_cvthf8_ph(__m128h __W,
                                                                 __mmask8 __U,
                                                                 __m128i __A) {
   return (__m128h)__builtin_ia32_vcvthf8_2ph128_mask(
       (__v16qi)__A, (__v8hf)(__m128h)__W, (__mmask8)__U);
 }
 
-static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_maskz_cvthf8(__mmask8 __U,
+static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_maskz_cvthf8_ph(__mmask8 __U,
                                                                  __m128i __A) {
   return (__m128h)__builtin_ia32_vcvthf8_2ph128_mask(
       (__v16qi)__A, (__v8hf)(__m128h)_mm_setzero_ph(), (__mmask8)__U);
 }
 
-static __inline__ __m256h __DEFAULT_FN_ATTRS256 _mm256_cvthf8(__m128i __A) {
+static __inline__ __m256h __DEFAULT_FN_ATTRS256 _mm256_cvthf8_ph(__m128i __A) {
   return (__m256h)__builtin_ia32_vcvthf8_2ph256_mask(
       (__v16qi)__A, (__v16hf)(__m256h)_mm256_undefined_ph(), (__mmask16)-1);
 }
 
 static __inline__ __m256h __DEFAULT_FN_ATTRS256
-_mm256_mask_cvthf8(__m256h __W, __mmask16 __U, __m128i __A) {
+_mm256_mask_cvthf8_ph(__m256h __W, __mmask16 __U, __m128i __A) {
   return (__m256h)__builtin_ia32_vcvthf8_2ph256_mask(
       (__v16qi)__A, (__v16hf)(__m256h)__W, (__mmask16)__U);
 }
 
 static __inline__ __m256h __DEFAULT_FN_ATTRS256
-_mm256_maskz_cvthf8(__mmask16 __U, __m128i __A) {
+_mm256_maskz_cvthf8_ph(__mmask16 __U, __m128i __A) {
   return (__m256h)__builtin_ia32_vcvthf8_2ph256_mask(
       (__v16qi)__A, (__v16hf)(__m256h)_mm256_setzero_ph(), (__mmask16)__U);
 }
diff --git a/clang/test/CodeGen/X86/avx10_2_512convert-builtins.c b/clang/test/CodeGen/X86/avx10_2_512convert-builtins.c
index 22503c640a727f2..dcf7bbc005a7c6c 100644
--- a/clang/test/CodeGen/X86/avx10_2_512convert-builtins.c
+++ b/clang/test/CodeGen/X86/avx10_2_512convert-builtins.c
@@ -201,22 +201,22 @@ __m512i test_mm512_maskz_cvts2ph_hf8(__mmask64 __U, __m512h __A, __m512h __B) {
   return _mm512_maskz_cvts2ph_hf8(__U, __A, __B);
 }
 
-__m512h test_mm512_cvthf8(__m256i __A) {
-  // CHECK-LABEL: @test_mm512_cvthf8(
+__m512h test_mm512_cvthf8_ph(__m256i __A) {
+  // CHECK-LABEL: @test_mm512_cvthf8_ph(
   // CHECK: call <32 x half> @llvm.x86.avx10.mask.vcvthf82ph512(
-  return _mm512_cvthf8(__A);
+  return _mm512_cvthf8_ph(__A);
 }
 
-__m512h test_mm512_mask_cvthf8(__m512h __A, __mmask32 __B, __m256i __C) {
-  // CHECK-LABEL: @test_mm512_mask_cvthf8(
+__m512h test_mm512_mask_cvthf8_ph(__m512h __A, __mmask32 __B, __m256i __C) {
+  // CHECK-LABEL: @test_mm512_mask_cvthf8_ph(
   // CHECK: call <32 x half> @llvm.x86.avx10.mask.vcvthf82ph512(
-  return _mm512_mask_cvthf8(__A, __B, __C);
+  return _mm512_mask_cvthf8_ph(__A, __B, __C);
 }
 
-__m512h test_mm512_maskz_cvthf8(__mmask32 __A, __m256i __B) {
-  // CHECK-LABEL: @test_mm512_maskz_cvthf8(
+__m512h test_mm512_maskz_cvthf8_ph(__mmask32 __A, __m256i __B) {
+  // CHECK-LABEL: @test_mm512_maskz_cvthf8_ph(
   // CHECK: call <32 x half> @llvm.x86.avx10.mask.vcvthf82ph512(
-  return _mm512_maskz_cvthf8(__A, __B);
+  return _mm512_maskz_cvthf8_ph(__A, __B);
 }
 
 __m256i test_mm512_cvtph_bf8(__m512h __A) {
diff --git a/clang/test/CodeGen/X86/avx10_2convert-builtins.c b/clang/test/CodeGen/X86/avx10_2convert-builtins.c
index efd9a31c40875ab..d73a223922ce03d 100644
--- a/clang/test/CodeGen/X86/avx10_2convert-builtins.c
+++ b/clang/test/CodeGen/X86/avx10_2convert-builtins.c
@@ -379,40 +379,40 @@ __m256i test_mm256_maskz_cvts2ph_hf8(__mmask16 __U, __m256h __A, __m256h __B) {
   return _mm256_maskz_cvts2ph_hf8(__U, __A, __B);
 }
 
-__m128h test_mm_cvthf8(__m128i __A) {
-  // CHECK-LABEL: @test_mm_cvthf8(
+__m128h test_mm_cvthf8_ph(__m128i __A) {
+  // CHECK-LABEL: @test_mm_cvthf8_ph(
   // CHECK: call <8 x half> @llvm.x86.avx10.mask.vcvthf82ph128(
-  return _mm_cvthf8(__A);
+  return _mm_cvthf8_ph(__A);
 }
 
-__m128h test_mm_mask_cvthf8(__m128h __A, __mmask8 __B, __m128i __C) {
-  // CHECK-LABEL: @test_mm_mask_cvthf8(
+__m128h test_mm_mask_cvthf8_ph(__m128h __A, __mmask8 __B, __m128i __C) {
+  // CHECK-LABEL: @test_mm_mask_cvthf8_ph(
   // CHECK: call <8 x half> @llvm.x86.avx10.mask.vcvthf82ph128(
-  return _mm_mask_cvthf8(__A, __B, __C);
+  return _mm_mask_cvthf8_ph(__A, __B, __C);
 }
 
-__m128h test_mm_maskz_cvthf8(__mmask8 __A, __m128i __B) {
-  // CHECK-LABEL: @test_mm_maskz_cvthf8(
+__m128h test_mm_maskz_cvthf8_ph(__mmask8 __A, __m128i __B) {
+  // CHECK-LABEL: @test_mm_maskz_cvthf8_ph(
   // CHECK: call <8 x half> @llvm.x86.avx10.mask.vcvthf82ph128(
-  return _mm_maskz_cvthf8(__A, __B);
+  return _mm_maskz_cvthf8_ph(__A, __B);
 }
 
-__m256h test_mm256_cvthf8(__m128i __A) {
-  // CHECK-LABEL: @test_mm256_cvthf8(
+__m256h test_mm256_cvthf8_ph(__m128i __A) {
+  // CHECK-LABEL: @test_mm256_cvthf8_ph(
   // CHECK: call <16 x half> @llvm.x86.avx10.mask.vcvthf82ph256(
-  return _mm256_cvthf8(__A);
+  return _mm256_cvthf8_ph(__A);
 }
 
-__m256h test_mm256_mask_cvthf8(__m256h __A, __mmask16 __B, __m128i __C) {
-  // CHECK-LABEL: @test_mm256_mask_cvthf8(
+__m256h test_mm256_mask_cvthf8_ph(__m256h __A, __mmask16 __B, __m128i __C) {
+  // CHECK-LABEL: @test_mm256_mask_cvthf8_ph(
   // CHECK: call <16 x half> @llvm.x86.avx10.mask.vcvthf82ph256(
-  return _mm256_mask_cvthf8(__A, __B, __C);
+  return _mm256_mask_cvthf8_ph(__A, __B, __C);
 }
 
-__m256h test_mm256_maskz_cvthf8(__mmask16 __A, __m128i __B) {
-  // CHECK-LABEL: @test_mm256_maskz_cvthf8(
+__m256h test_mm256_maskz_cvthf8_ph(__mmask16 __A, __m128i __B) {
+  // CHECK-LABEL: @test_mm256_maskz_cvthf8_ph(
   // CHECK: call <16 x half> @llvm.x86.avx10.mask.vcvthf82ph256(
-  return _mm256_maskz_cvthf8(__A, __B);
+  return _mm256_maskz_cvthf8_ph(__A, __B);
 }
 
 __m128i test_mm_cvtph_bf8(__m128h __A) {

@github-actions GitHub Actions
Copy link

github-actions bot commented Feb 9, 2025 •
edited
Loading

✅ With the latest revision this PR passed the C/C++ code formatter.

phoebewang
phoebewang approved these changes Feb 9, 2025
View reviewed changes
Copy link
Contributor

@phoebewang phoebewang left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Good catch!

@phoebewang
Copy link
Contributor

⚠️ C/C++ code formatter, clang-format found issues in your code. ⚠️

You can test this locally with the following command:
View the diff from clang-format here.

diff --git a/clang/lib/Headers/avx10_2convertintrin.h b/clang/lib/Headers/avx10_2convertintrin.h
index cbc02e37b0..c419323910 100644
--- a/clang/lib/Headers/avx10_2convertintrin.h
+++ b/clang/lib/Headers/avx10_2convertintrin.h
@@ -386,15 +386,14 @@ static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_cvthf8_ph(__m128i __A) {
       (__v16qi)__A, (__v8hf)(__m128h)_mm_undefined_ph(), (__mmask8)-1);
 }
 
-static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_mask_cvthf8_ph(__m128h __W,
-                                                                __mmask8 __U,
-                                                                __m128i __A) {
+static __inline__ __m128h __DEFAULT_FN_ATTRS128
+_mm_mask_cvthf8_ph(__m128h __W, __mmask8 __U, __m128i __A) {
   return (__m128h)__builtin_ia32_vcvthf8_2ph128_mask(
       (__v16qi)__A, (__v8hf)(__m128h)__W, (__mmask8)__U);
 }
 
-static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_maskz_cvthf8_ph(__mmask8 __U,
-                                                                 __m128i __A) {
+static __inline__ __m128h __DEFAULT_FN_ATTRS128
+_mm_maskz_cvthf8_ph(__mmask8 __U, __m128i __A) {
   return (__m128h)__builtin_ia32_vcvthf8_2ph128_mask(
       (__v16qi)__A, (__v8hf)(__m128h)_mm_setzero_ph(), (__mmask8)__U);
 }

Probably need to fix the format. You may apply the diff directly.

@mikolaj-pirog
Copy link
Contributor Author

Good catch!

Thanks! I have fixed the formatting

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@phoebewang phoebewang phoebewang approved these changes

Assignees
No one assigned
Labels
backend:X86 clang:headers Headers provided by Clang, e.g. for intrinsics clang Clang issues not falling into any other category
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@mikolaj-pirog @llvmbot @phoebewang