deaeb6c5b8
1. Added bfloat16 based dot as new API: shdot
2. Implemented generic kernel and cooperlake-specific (AVX512-BF16) kernel for shdot
3. Added 4 conversion APIs for bfloat16 data type <=> single/double: shstobf16 shdtobf16 sbf16tos dbf16tod
shstobf16 -- convert single float array to bfloat16 array
shdtobf16 -- convert double float array to bfloat16 array
sbf16tos -- convert bfloat16 array to single float array
dbf16tod -- convert bfloat16 array to double float array
4. Implemented generic kernels for all 4 conversion APIs, and cooperlake-specific kernel for shstobf16 and shdtobf16
5. Update level1 thread facilitate functions and macros to support multi-threading for these new APIs
6. Fix Cooperlake platform detection/specify issue when under dynamic-arch building
7. Change the typedef of bfloat16 from unsigned short to more strict uint16_t
Signed-off-by: Chen, Guobing <guobing.chen@intel.com>
87 lines
4.3 KiB
C
87 lines
4.3 KiB
C
/***************************************************************************
|
|
Copyright (c) 2014, The OpenBLAS Project
|
|
All rights reserved.
|
|
Redistribution and use in source and binary forms, with or without
|
|
modification, are permitted provided that the following conditions are
|
|
met:
|
|
1. Redistributions of source code must retain the above copyright
|
|
notice, this list of conditions and the following disclaimer.
|
|
2. Redistributions in binary form must reproduce the above copyright
|
|
notice, this list of conditions and the following disclaimer in
|
|
the documentation and/or other materials provided with the
|
|
distribution.
|
|
3. Neither the name of the OpenBLAS project nor the names of
|
|
its contributors may be used to endorse or promote products
|
|
derived from this software without specific prior written permission.
|
|
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
|
|
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
ARE DISCLAIMED. IN NO EVENT SHALL THE OPENBLAS PROJECT OR CONTRIBUTORS BE
|
|
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
|
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
|
|
SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
|
|
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
|
|
OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
|
|
USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
*****************************************************************************/
|
|
|
|
/* need a new enough GCC for avx512 support */
|
|
#if (( defined(__GNUC__) && __GNUC__ >= 10 && defined(__AVX512BF16__)) || (defined(__clang__) && __clang_major__ >= 9))
|
|
|
|
#define HAVE_TOBF16_ACCL_KERNEL 1
|
|
#include "common.h"
|
|
#include <immintrin.h>
|
|
|
|
static void tobf16_accl_kernel(BLASLONG n, const float * in, bfloat16 * out)
|
|
{
|
|
/* Get the 64-bytes unaligned header number targeting for avx512
|
|
* processing (Assume input float array is natural aligned) */
|
|
int align_header = ((64 - ((uintptr_t)in & (uintptr_t)0x3f)) >> 2) & 0xf;
|
|
|
|
if (n < align_header) {align_header = n;}
|
|
|
|
if (align_header != 0) {
|
|
uint16_t align_mask16 = (((uint16_t)0xffff) >> (16-align_header));
|
|
__m512 a = _mm512_maskz_loadu_ps(*((__mmask16*) &align_mask16), &in[0]);
|
|
_mm256_mask_storeu_epi16(&out[0], *((__mmask16*) &align_mask16), (__m256i) _mm512_cvtneps_pbh(a));
|
|
}
|
|
|
|
if (n == align_header) {
|
|
return;
|
|
} else {
|
|
n -= align_header;
|
|
in += align_header;
|
|
out += align_header;
|
|
}
|
|
|
|
int tail_index_32 = n&(~31);
|
|
int tail_index_128 = n&(~127);
|
|
uint32_t tail_mask32 = (((uint32_t) 0xffffffff) >> (32-(n&31)));
|
|
uint16_t tail_mask16 = (((uint16_t) 0xffff) >> (16-(n&15)));
|
|
|
|
/* Processing the main chunk with 128-elements per round */
|
|
for (int i = 0; i < tail_index_128; i += 128) {
|
|
_mm512_storeu_si512(&out[i+ 0], (__m512i) _mm512_cvtne2ps_pbh(_mm512_load_ps(&in[i+ 16]), _mm512_load_ps(&in[i+ 0])));
|
|
_mm512_storeu_si512(&out[i+32], (__m512i) _mm512_cvtne2ps_pbh(_mm512_load_ps(&in[i+ 48]), _mm512_load_ps(&in[i+32])));
|
|
_mm512_storeu_si512(&out[i+64], (__m512i) _mm512_cvtne2ps_pbh(_mm512_load_ps(&in[i+ 80]), _mm512_load_ps(&in[i+64])));
|
|
_mm512_storeu_si512(&out[i+96], (__m512i) _mm512_cvtne2ps_pbh(_mm512_load_ps(&in[i+112]), _mm512_load_ps(&in[i+96])));
|
|
}
|
|
|
|
/* Processing the remaining <128 chunk with 32-elements per round */
|
|
for (int j = tail_index_128; j < tail_index_32; j += 32) {
|
|
_mm512_storeu_si512(&out[j], (__m512i) _mm512_cvtne2ps_pbh(_mm512_load_ps(&in[j+ 16]), _mm512_load_ps(&in[j])));
|
|
}
|
|
|
|
/* Processing the remaining <32 chunk with masked processing */
|
|
if ((n&31) > 15) {
|
|
__m512 b = _mm512_load_ps(&in[tail_index_32]);
|
|
__m512 a = _mm512_maskz_load_ps(*((__mmask16*) &tail_mask16), &in[tail_index_32+16]);
|
|
_mm512_mask_storeu_epi16(&out[tail_index_32], *((__mmask32*) &tail_mask32), (__m512i) _mm512_cvtne2ps_pbh(a, b));
|
|
} else if ((n&31) > 0) {
|
|
__m512 a = _mm512_maskz_load_ps(*((__mmask16*) &tail_mask16), &in[tail_index_32]);
|
|
_mm256_mask_storeu_epi16(&out[tail_index_32], *((__mmask16*) &tail_mask16), (__m256i) _mm512_cvtneps_pbh(a));
|
|
}
|
|
}
|
|
|
|
#endif
|