vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

- rearranged defines, SSSE3 function only compiled if used

Browse source
This commit is contained in:
3ooabkhxtn 2023-05-12 20:48:41 +00:00
parent fc26f54e74
commit 25b448a32f
1 changed files with 12 additions and 12 deletions
Download patch file Download diff file Expand all files Collapse all files

24
ggml.c
View file

@ -485,18 +485,6 @@ static inline __m128i mul_sum_i8_pairs(const __m128i x, const __m128i y) {
return _mm_madd_epi16(ones, dot); return _mm_madd_epi16(ones, dot);
} }
// horizontally add 4x4 floats
static inline float hsum_float_4x4(const __m128 a, const __m128 b, const __m128 c, const __m128 d) {
__m128 res_0 =_mm_hadd_ps(a, b);
__m128 res_1 =_mm_hadd_ps(c, d);
__m128 res =_mm_hadd_ps(res_0, res_1);
res =_mm_hadd_ps(res, res);
res =_mm_hadd_ps(res, res);
return _mm_cvtss_f32(res);
}
#endif
#if __AVX__ || __AVX2__ || __AVX512F__ #if __AVX__ || __AVX2__ || __AVX512F__
// horizontally add 8 floats // horizontally add 8 floats
static inline float hsum_float_8(const __m256 x) { static inline float hsum_float_8(const __m256 x) {
@ -609,7 +597,19 @@ static inline __m128i packNibbles( __m128i bytes1, __m128i bytes2 )
return _mm_packus_epi16( bytes1, bytes2); return _mm_packus_epi16( bytes1, bytes2);
} }
#endif #endif
#elif defined(__SSSE3__)
// horizontally add 4x4 floats
static inline float hsum_float_4x4(const __m128 a, const __m128 b, const __m128 c, const __m128 d) {
__m128 res_0 =_mm_hadd_ps(a, b);
__m128 res_1 =_mm_hadd_ps(c, d);
__m128 res =_mm_hadd_ps(res_0, res_1);
res =_mm_hadd_ps(res, res);
res =_mm_hadd_ps(res, res);
return _mm_cvtss_f32(res);
}
#endif // __AVX__ || __AVX2__ || __AVX512F__ #endif // __AVX__ || __AVX2__ || __AVX512F__
#endif // defined(__AVX__) || defined(__AVX2__) || defined(__AVX512F__) || defined(__SSSE3__)
#if __ARM_NEON #if __ARM_NEON