https://gcc.gnu.org/bugzilla/show_bug.cgi?id=116157
Bug ID: 116157
Summary: AVX2 _mm256_exp_ps function is missing in the compiler
Product: gcc
Version: 13.2.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: binklings at 163 dot com
Target Milestone: ---
Hello,
Issue:
In both gcc and g++, the _mm256_exp_ps function in Intel AVX2 intrinsics is
missing
gcc/g++ version: x86_64-posix-seh-rev1 13.2.0
Code:
#include <immintrin.h>
int main(){
float a[4] = {0.1, 0.2, 0.3, 0.4};
__m256 vec = _mm256_loadu_ps(a);
vec = _mm256_exp_ps(vec);
_mm256_storeu_ps(a, vec);
}
Compiler Command:
g++ -mavx2 -o test test.cpp
(or gcc)
both output:
test.cpp: In function 'int main()':
test.cpp:6:11: error: '_mm256_exp_ps' was not declared in this scope; did you
mean '_mm256_rcp_ps'?
6 | vec = _mm256_exp_ps(vec);
| ^~~~~~~~~~~~~
| _mm256_rcp_ps
The _mm256_exp_ps function:
https://www.intel.com/content/www/us/en/docs/cpp-compiler/developer-guide-reference/2021-10/mm-exp-ps-mm256-exp-ps.html
It seems that this SVML function can only be used in Intel compiler or MSVC.
Will it be supported in gcc/g++? Is there a way to use this function? Thanks