floraachy
/
OpenBLAS
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

For the sake of windows compatible, used "unsigned long long" to ensure 64-bit length 

Signed-off-by: Wang, Long <long1.wang@intel.com>
This commit is contained in:
Wang, Long 2019-11-20 21:30:16 +08:00
parent 0caf1434c9
commit 1191db1a49
2 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
kernel/x86_64/sgemm_kernel_16x4_skylakex.c
View File

@ -762,7 +762,7 @@ USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
int __attribute__ ((noinline))
CNAME(BLASLONG m, BLASLONG n, BLASLONG k, float alpha, float * __restrict A, float * __restrict B, float * __restrict C, BLASLONG ldc)
{
unsigned long M = m, N = n, K = k;
unsigned long long M = m, N = n, K = k;
if (M == 0)
return 0;
if (N == 0)
@ -1639,4 +1639,4 @@ void sgemm_kernel_direct (BLASLONG M, BLASLONG N, BLASLONG K, float * __restrict
STORE_SCALAR(0, 0);
}
}
}
}

2
kernel/x86_64/sgemm_kernel_16x4_skylakex_2.c
View File

@ -452,7 +452,7 @@ CNAME(BLASLONG m, BLASLONG n, BLASLONG k, float alpha, float * __restrict__ A, f
int sgemm_kernel_direct_performant(BLASLONG M, BLASLONG N, BLASLONG K)
{
unsigned long mnk = M * N * K;
unsigned long long mnk = M * N * K;
/* large matrixes -> not performant */
if (mnk >= 28 * 512 * 512)
return 0;