Skip to content

loongarch64: fix segfaults in copy kernels and adjust dsyrk block size #5942

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
HecaiYuan wants to merge 4 commits into
base: develop
Choose a base branch
from
+12 −0
Open

loongarch64: fix segfaults in copy kernels and adjust dsyrk block size #5942

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
018891a
LoongArch64: fix sgemm_ncopy_8_lasx.S segfault for small columns
HecaiYuan Jul 24, 2026
4c78593
LoongArch64: fix sgemm_ncopy_16_lasx.S segfault for small columns
HecaiYuan Jul 24, 2026
1b3358b
LoongArch64: fix dgemm_ncopy_4_lsx.S segfault for small columns
HecaiYuan Jul 24, 2026
d16316e
LoongArch64: adjust dsyrk GEMM_P block size to avoid failure
HecaiYuan Jul 24, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions driver/level3/level3_syrk_threaded.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,6 +37,11 @@
/* or implied, of The University of Texas at Austin. */
/*********************************************************************/

#if defined(__loongarch__) && defined(LA464) && defined(DOUBLE)
#undef GEMM_P
#define GEMM_P 96
#endif

#ifndef CACHE_LINE_SIZE
#define CACHE_LINE_SIZE 8
#endif
Expand Down
2 changes: 2 additions & 0 deletions kernel/loongarch64/dgemm_ncopy_4_lsx.S
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -172,6 +172,8 @@ USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
blt ZERO, I, .L_2II1
.L_N1:
move S1, TS
andi J, N, 0x01
beq ZERO, J, .L_N0
beq ZERO, M, .L_N0
.L_M1:
fld.d F0, S1, 0x00
Expand Down
2 changes: 2 additions & 0 deletions kernel/loongarch64/sgemm_ncopy_16_lasx.S
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -453,6 +453,8 @@ USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
.align 5
.L_N1:
move S1, TS
andi J, N, 0x01
beq ZERO, J, .L_N0
beq ZERO, M, .L_N0
.L_N1_M1:
fld.s F0, S1, 0x00
Expand Down
3 changes: 3 additions & 0 deletions kernel/loongarch64/sgemm_ncopy_8_lasx.S
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -288,7 +288,10 @@ USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
.align 5
.L_N1:
move S1, TS
andi J, N, 0x01
beq ZERO, J, .L_N0
beq ZERO, M, .L_N0

.L_N1_M1:
fld.s F0, S1, 0x00
PTR_ADDI S1, S1, 0x04
Expand Down
Loading