https://github.com/hpkfft created https://github.com/llvm/llvm-project/pull/99691
Intrinsics such as `_mm512_add_round_ps` take a rounding mode argument to specify the floating point rounding mode. This, and similar instructions, do NOT round their result to an integer. Thus it is inappropriate for user code to specify the existing `_MM_FROUND_TO_NEAREST_INT` when desiring to round to the nearest floating point number. This commit adds a suitable macro definition. >From b4fcf05f7df9365cd8289a20badad6ff485ecbd5 Mon Sep 17 00:00:00 2001 From: "hpkfft.com" <p...@hpkfft.com> Date: Fri, 19 Jul 2024 12:35:33 -0700 Subject: [PATCH] Add _MM_FROUND_TIES_TO_EVEN to avx512fintrin.h Intrinsics such as `_mm512_add_round_ps` take a rounding mode argument to specify the floating point rounding mode. This, and similar instructions, do NOT round their result to an integer. Thus it is inappropriate for user code to specify the existing `_MM_FROUND_TO_NEAREST_INT` when desiring to round to the nearest floating point number. This commit adds a suitable macro definition. --- clang/lib/Headers/avx512fintrin.h | 1 + 1 file changed, 1 insertion(+) diff --git a/clang/lib/Headers/avx512fintrin.h b/clang/lib/Headers/avx512fintrin.h index 4f172c74b31cb..2d7b5c534e554 100644 --- a/clang/lib/Headers/avx512fintrin.h +++ b/clang/lib/Headers/avx512fintrin.h @@ -43,6 +43,7 @@ typedef unsigned short __mmask16; /* Rounding mode macros. */ #define _MM_FROUND_TO_NEAREST_INT 0x00 +#define _MM_FROUND_TIES_TO_EVEN 0x00 #define _MM_FROUND_TO_NEG_INF 0x01 #define _MM_FROUND_TO_POS_INF 0x02 #define _MM_FROUND_TO_ZERO 0x03 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
