[Clang][X86] Add tests for AVX512 integer comparison intrinsics to be used in constexpr #169452
+512
−9
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Thank you for submitting a Pull Request (PR) to the LLVM Project! This PR will be automatically labeled and the relevant teams will be notified. If you wish to, you can add reviewers by using the "Reviewers" section on this page. If this is not working for you, it is probably because you do not have write permissions for the repository. In which case you can instead tag reviewers by name in a comment by using If you have received no comments on your PR for a week, you can request a review by "ping"ing the PR by adding a comment “Ping”. The common courtesy "ping" rate is once a week. Please remember that you are asking for valuable time from other developers. If you have further questions, they may be answered by the LLVM GitHub User Guide. You can also ask questions in a comment on this PR, on the LLVM Discord or on the forums. |
llvmbot
added
clang
Member
|
@llvm/pr-subscribers-backend-x86 @llvm/pr-subscribers-clang Author: irhs0 (shri-acha) ChangesCovers the test for AVX512 comparison intrinsics mentioned in: Full diff: https://github.com/llvm/llvm-project/pull/169452.diff 1 Files Affected:
diff --git a/clang/test/CodeGen/X86/avx512vlbw-builtins.c b/clang/test/CodeGen/X86/avx512vlbw-builtins.c
index 16e113031bfda..cb69205664a3d 100644
--- a/clang/test/CodeGen/X86/avx512vlbw-builtins.c
+++ b/clang/test/CodeGen/X86/avx512vlbw-builtins.c
@@ -645,20 +645,123 @@ __mmask16 test_mm_cmp_epi8_mask(__m128i __a, __m128i __b) {
return (__mmask16)_mm_cmp_epi8_mask(__a, __b, 0);
}
+// cmpeq tests
TEST_CONSTEXPR(_mm_cmpeq_epi8_mask(
- ((__m128i)(__v16qi){5, 3, 7, 2, 9, 3, 7, 1, 5, 4, 8, 2, 9, 6, 7, 5}),
- ((__m128i)(__v16qi){5, 2, 7, 3, 9, 4, 6, 1, 5, 3, 8, 1, 9, 5, 7, 5})
-) == (__mmask16)0xd595);
+ ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}),
+ ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0})
+) == (__mmask16)0xffff);
+
+TEST_CONSTEXPR(_mm_cmpeq_epi8_mask(
+ ((__m128i)(__v16qi){1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}),
+ ((__m128i)(__v16qi){1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1})
+) == (__mmask16)0xffff);
+
+TEST_CONSTEXPR(_mm_cmpeq_epi8_mask(
+ ((__m128i)(__v16qi){127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127}),
+ ((__m128i)(__v16qi){-128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128})
+) == (__mmask16)0x0000);
+
+TEST_CONSTEXPR(_mm_cmpeq_epi8_mask(
+ ((__m128i)(__v16qi){-31, 90, -66, 3, 27, -22, -64, 111, -87, 105, -19, 0, 26, -111, 24, -72}),
+ ((__m128i)(__v16qi){-84, -98, 20, -83, -98, 80, -46, -9, 22, -120, -123, 53, 117, -85, 50, 94})
+) == (__mmask16)0x0000);
+
+// cmpneq tests
+TEST_CONSTEXPR(_mm_cmpneq_epi8_mask(
+ ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}),
+ ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0})
+) == (__mmask16)0x0000);
+
+TEST_CONSTEXPR(_mm_cmpneq_epi8_mask(
+ ((__m128i)(__v16qi){1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}),
+ ((__m128i)(__v16qi){1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1})
+) == (__mmask16)0x0000);
+
+TEST_CONSTEXPR(_mm_cmpneq_epi8_mask(
+ ((__m128i)(__v16qi){127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127}),
+ ((__m128i)(__v16qi){-128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128})
+) == (__mmask16)0xffff);
+
+TEST_CONSTEXPR(_mm_cmpneq_epi8_mask(
+ ((__m128i)(__v16qi){-2, 49, -5, -11, 21, -70, 121, -111, 20, 112, -121, 18, -34, -73, 89, 122}),
+ ((__m128i)(__v16qi){14, 36, 2, 3, 118, 88, -126, -21, 104, -125, -1, 39, 99, -12, 35, -126})
+) == (__mmask16)0xffff);
+
+// cmplt tests
+TEST_CONSTEXPR(_mm_cmplt_epi8_mask(
+ ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}),
+ ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0})
+) == (__mmask16)0x0000);
TEST_CONSTEXPR(_mm_cmplt_epi8_mask(
- ((__m128i)(__v16qi){1, 5, 3, 7, 2, 8, 4, 6, 9, 5, 3, 11, 2, 6, 15, 8}),
- ((__m128i)(__v16qi){2, 4, 6, 8, 3, 5, 7, 9, 4, 6, 8, 10, 5, 7, 9, 11})
-) == (__mmask16)0xb6dd);
+ ((__m128i)(__v16qi){1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}),
+ ((__m128i)(__v16qi){1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1})
+) == (__mmask16)0x0000);
+TEST_CONSTEXPR(_mm_cmplt_epi8_mask(
+ ((__m128i)(__v16qi){127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127}),
+ ((__m128i)(__v16qi){-128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128})
+) == (__mmask16)0x0000);
+
+TEST_CONSTEXPR(_mm_cmplt_epi8_mask(
+ ((__m128i)(__v16qi){-111, -10, -60, -123, -6, -110, -43, -32, -58, -7, 42, -128, -21, 24, 8, -101}),
+ ((__m128i)(__v16qi){-108, 30, 71, 73, 20, 117, 63, -93, 79, -30, 99, -100, 34, 49, 83, 68})
+) == (__mmask16)0xfd7f);
+
+// cmple tests
+TEST_CONSTEXPR(_mm_cmple_epi8_mask(
+ ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}),
+ ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0})
+) == (__mmask16)0xffff);
TEST_CONSTEXPR(_mm_cmple_epi8_mask(
- ((__m128i)(__v16qi){1, 3, 5, 7, 2, 6, 6, 8, 1, 3, 9, 7, 2, 4, 6, 10}),
- ((__m128i)(__v16qi){2, 3, 4, 7, 3, 4, 5, 8, 2, 3, 4, 7, 3, 4, 5, 8})
-) == (__mmask16)0x3b9b);
+ ((__m128i)(__v16qi){1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}),
+ ((__m128i)(__v16qi){1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1})
+) == (__mmask16)0xffff);
+TEST_CONSTEXPR(_mm_cmple_epi8_mask(
+ ((__m128i)(__v16qi){127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127}),
+ ((__m128i)(__v16qi){-128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128})
+) == (__mmask16)0x0000);
+TEST_CONSTEXPR(_mm_cmple_epi8_mask(
+ ((__m128i)(__v16qi){122, 6, -22, -94, 78, -35, -43, -16, -69, 124, -2, 24, -117, 8, -17, 118}),
+ ((__m128i)(__v16qi){53, -50, 104, 11, 63, -77, -25, 102, 46, 62, 27, -28, -61, 68, 40, -65})
+) == (__mmask16)0x75cc);
+
+// cmpge tests
+TEST_CONSTEXPR(_mm_cmpge_epi8_mask(
+ ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}),
+ ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0})
+) == (__mmask16)0xffff);
+TEST_CONSTEXPR(_mm_cmpge_epi8_mask(
+ ((__m128i)(__v16qi){1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}),
+ ((__m128i)(__v16qi){1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1})
+) == (__mmask16)0xffff);
+TEST_CONSTEXPR(_mm_cmpge_epi8_mask(
+ ((__m128i)(__v16qi){127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127}),
+ ((__m128i)(__v16qi){-128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128})
+) == (__mmask16)0xffff);
+TEST_CONSTEXPR(_mm_cmpge_epi8_mask(
+ ((__m128i)(__v16qi){-11, 0, 97, 123, -48, 105, 26, -118, 62, -86, -94, -32, 14, -4, -50, 72}),
+ ((__m128i)(__v16qi){-33, 49, 22, 31, -4, -81, 6, -22, 40, 127, -1, -106, 6, -64, 12, 8})
+) == (__mmask16)0xb96d);
+
+// cmpgt tests
+TEST_CONSTEXPR(_mm_cmpgt_epi8_mask(
+ ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}),
+ ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0})
+) == (__mmask16)0x0000);
+TEST_CONSTEXPR(_mm_cmpgt_epi8_mask(
+ ((__m128i)(__v16qi){1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}),
+ ((__m128i)(__v16qi){1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1})
+) == (__mmask16)0x0000);
+TEST_CONSTEXPR(_mm_cmpgt_epi8_mask(
+ ((__m128i)(__v16qi){127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127}),
+ ((__m128i)(__v16qi){-128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128, -128})
+) == (__mmask16)0xffff);
+TEST_CONSTEXPR(_mm_cmpgt_epi8_mask(
+ ((__m128i)(__v16qi){-127, 37, -123, -60, 98, -68, -115, 96, 80, -27, -105, 64, -6, -51, -45, -81}),
+ ((__m128i)(__v16qi){-124, 59, 8, 59, 122, 47, -74, 120, 19, -43, -33, -76, 7, -22, -24, -14})
+) == (__mmask16)0x0b00);
+
__mmask16 test_mm_mask_cmp_epi8_mask(__mmask16 __u, __m128i __a, __m128i __b) {
// CHECK-LABEL: test_mm_mask_cmp_epi8_mask
|
Author
|
@RKSimon could you verify the nature of the tests?, after i'll add the other generated tests from the script |
|
|
RKSimon
requested changes
Nov 25, 2025
| ) == (__mmask16)0xd595); | ||
| ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}), | ||
| ((__m128i)(__v16qi){0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}) | ||
| ) == (__mmask16)0xffff); |
Collaborator
There was a problem hiding this comment.
avoid 'splat' tests like this - all ones/zero results always make me suspicious :)
Author
There was a problem hiding this comment.
:) okay ,will modify the 'splat' cases here onwards.
| ((__m128i)(__v16qi){-31, 90, -66, 3, 27, -22, -64, 111, -87, 105, -19, 0, 26, -111, 24, -72}), | ||
| ((__m128i)(__v16qi){-84, -98, 20, -83, -98, 80, -46, -9, 22, -120, -123, 53, 117, -85, 50, 94}) | ||
| ) == (__mmask16)0x0000); | ||
|
|
Collaborator
There was a problem hiding this comment.
You need _mm_mask_cmpeq_epi8_mask (et al) test coverage as well
Author
There was a problem hiding this comment.
yes, I do have those in mind aswell, i will cover those after all of the _mm_mask_cmp*_epi* are complete.
🐧 Linux x64 Test Results
Failed Tests(click on a test name to see its output) ClangClang.CodeGen/X86/avx512vlbw-builtins.cIf these failures are unrelated to your changes (for example tests are broken or flaky at HEAD), please open an issue at https://github.com/llvm/llvm-project/issues and add the |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Covers the test for AVX512 comparison intrinsics mentioned in:
#162054