amd/blis
1
0
Fork 0
mirror of https://github.com/amd/blis.git synced 2026-05-12 01:59:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

Some improvements to trsm parallelism

Browse Source
This commit is contained in:
Tyler Smith
2014-08-20 14:43:17 -05:00
parent daca500db5
commit 699a8151ca
2 changed files with 6 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
frame/3/trsm/bli_trsm_blk_var2b.c
View File

@@ -83,9 +83,9 @@ void bli_trsm_blk_var2b( obj_t* a,
// Query dimension in partitioning direction.
n_trans = bli_obj_width_after_trans( *b );
dim_t start, end;
bli_get_range_weighted( thread, 0, n_trans,
bli_determine_reg_blocksize( b, cntl_blocksize( cntl ) ),
bli_obj_is_upper( *c ), &start, &end );
bli_get_range( thread, 0, n_trans,
bli_determine_reg_blocksize( b, cntl_blocksize( cntl ) ),
&start, &end );
// Partition along the n dimension.
for ( i = start; i < end; i += b_alg )

6
frame/3/trsm/bli_trsm_blk_var2f.c
View File

@@ -83,9 +83,9 @@ void bli_trsm_blk_var2f( obj_t* a,
// Query dimension in partitioning direction.
n_trans = bli_obj_width_after_trans( *b );
dim_t start, end;
bli_get_range_weighted( thread, 0, n_trans,
bli_determine_reg_blocksize( b, cntl_blocksize( cntl ) ),
bli_obj_is_lower( *c ), &start, &end );
bli_get_range( thread, 0, n_trans,
bli_determine_reg_blocksize( b, cntl_blocksize( cntl ) ),
&start, &end );
// Partition along the n dimension.
for ( i = start; i < end; i += b_alg )