amd/blis
1
0
Fork 0
mirror of https://github.com/amd/blis.git synced 2026-04-20 15:48:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

Additional bug-fix for AOCL-BLAS bench

Browse Source 
- Corrected the format specifier setting(as macro) to not
  include additional spaces, since this would cause incorrect
  parsing of input files(in case they have exactly the expected
  number of parameters and not more).

- Updated the inputgemm.txt file to contain some inputs that
  have the exact parameters, to validate this fix.

AMD-Internal: [CPUPL-6365]
Change-Id: Ie9a83d4ed7e750ff1380d00c9c182b0c9ed42c49
This commit is contained in:
Vignesh Balasubramanian
2025-01-30 12:25:40 +05:30
committed by Vignesh Balasubramanian
parent 4ade159800
commit 0e71d28c01
2 changed files with 8 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
bench/blis_int_type.h
View File

@@ -36,9 +36,9 @@
// NOTE : When blis-int-size is not explicitly specified, the underlying
// hardware is detected at compile time and the width is set.
#if BLIS_INT_TYPE_SIZE == 32
#define INT_FS " %d "
#define UINT_FS " %u "
#define INT_FS "%d"
#define UINT_FS "%u"
#else
#define INT_FS " %ld "
#define UINT_FS " %lu "
#define INT_FS "%ld"
#define UINT_FS "%lu"
#endif

8
bench/inputgemm.txt
View File

@@ -26,7 +26,7 @@ sgemm_ S N N 1000 3000 2000 0.900000 0.000000 4000 5000 -1.100000 0.000000 6000
sgemm_ S N N 100 100 100 0.900000 0.000000 104 104 -1.100000 0.000000 104 nt=4 0.362 ms 5.525 GFLOPS
sgemm_ S N N 500 500 500 0.900000 0.000000 504 504 -1.100000 0.000000 504 nt=4 1.688 ms 148.104 GFLOPS
sgemm_ S N N 900 900 900 0.900000 0.000000 904 904 -1.100000 0.000000 904 nt=4 147.791 ms 9.865 GFLOPS
sgemm_ S N N 1300 1300 1300 0.900000 0.000000 1304 1304 -1.100000 0.000000 1304 nt=4 451.156 ms 9.739 GFLOPS
sgemm_ S N T 1700 1700 1700 0.900000 0.000000 1704 1704 -1.100000 0.000000 1704 nt=4 873.577 ms 11.248 GFLOPS
sgemm_ S T N 2100 2100 2100 0.900000 0.000000 2104 2104 -1.100000 0.000000 2104 nt=4 1699.278 ms 10.900 GFLOPS
sgemm_ S T T 2500 2500 2500 0.900000 0.000000 2504 2504 -1.100000 0.000000 2504 nt=4 2651.917 ms 11.784 GFLOPS
sgemm_ S N N 1300 1300 1300 0.900000 0.000000 1304 1304 -1.100000 0.000000 1304
sgemm_ S N T 1700 1700 1700 0.900000 0.000000 1704 1704 -1.100000 0.000000 1704
sgemm_ S T N 2100 2100 2100 0.900000 0.000000 2104 2104 -1.100000 0.000000 2104
sgemm_ S T T 2500 2500 2500 0.900000 0.000000 2504 2504 -1.100000 0.000000 2504