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

don't apply switch_ratio to tail of loop

This commit is contained in:
Martin Kroeker 2024-10-06 20:03:32 +02:00 committed by GitHub
parent 624e9d110e
commit 8a1710dd0d
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
driver/level3/level3_thread.c
View File

@ -742,7 +742,7 @@ static int gemm_driver(blas_arg_t *args, BLASLONG *range_m, BLASLONG
num_parts = 0;
while (n > 0){
width = blas_quickdivide(n + nthreads - num_parts - 1, nthreads - num_parts);
if (width < switch_ratio) {
if (width < switch_ratio && width > 1) {
width = switch_ratio;
}
width = round_up(n, width, GEMM_PREFERED_SIZE);