amd/blis
1
0
Fork 0
mirror of https://github.com/amd/blis.git synced 2026-05-11 17:50:00 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
f7b9bc734e83752d2215a63dbaa874626bd3a2ec
blis/frame/base
History
mkadavil 7dca25d056 Fixed decision logic in choosing between sup vs native GEMM code paths 
-Currently the storage preference of native kernel is used to determine
appropriate "m" and "n" which are used in native/SUP threshold logic.
This assumption works only when "sup" & native kernel preferences are
same. However, this need not have to be the case. This fix calculates
the correct "m" and "n" irrespective of the native/sup kernel
preferences thereby improving accuracy of this decision logic.

AMD-Internal: [CPUPL-3039]
Change-Id: I88c5a6838aeed867effdbbdfc41fbf2e88d4e52a
2023-03-02 02:12:42 -05:00
..
cast
AOCL Windows: 3.1 BLIS changes
2021-03-08 19:04:17 +05:30
check
AOCL Windows: 3.1 BLIS changes
2021-03-08 19:04:17 +05:30
noopt
AOCL Windows: 3.1 BLIS changes
2021-03-08 19:04:17 +05:30
proj
AOCL Windows: 3.1 BLIS changes
2021-03-08 19:04:17 +05:30
bli_apool.c
Merge branch 'master' into amd
2019-10-11 11:53:51 -05:00
bli_apool.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_arch.c
BLIS: Improve architecture selection at runtime
2022-08-19 10:59:35 -04:00
bli_arch.h
Replaced use of bool_t type with C99 bool.
2020-08-03 11:27:13 +05:30
bli_array.c
This check in has changes w.r.t Copyright information, which is changed to (start year) - 2019
2019-05-27 16:24:43 +05:30
bli_array.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_auxinfo.h
Declare/define static functions via BLIS_INLINE.
2020-08-03 11:23:40 +05:30
bli_blksz.c
Add AVX2 path for TRSM+GEMM combination.
2022-06-29 10:16:24 +00:00
bli_blksz.h
Removed dependency on AOCL_BLIS_ZEN for TRSM blocksizes.
2021-08-16 00:12:33 -04:00
bli_check.c
Implemented runtime subconfig selection (#451).
2020-10-18 18:04:03 -05:00
bli_check.h
Implemented runtime subconfig selection (#451).
2020-10-18 18:04:03 -05:00
bli_clock.c
Added support for systemless build (no pthreads).
2020-11-16 15:55:45 -06:00
bli_clock.h
Removed export macros from all internal prototypes.
2019-08-23 14:18:07 +05:30
bli_cntl.c
Support multithreading within the sup framework.
2020-03-13 01:09:29 -04:00
bli_cntl.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_cntx.c
BLIS : Compiler warning fixes
2021-11-12 08:58:52 +05:30
bli_cntx.h
Fixed decision logic in choosing between sup vs native GEMM code paths
2023-03-02 02:12:42 -05:00
bli_const.c
Remove UT-Austin from copyright headers' clause 3.
2018-12-04 14:31:06 -06:00
bli_const.h
Handling zgemm real(+/-1) alpha and beta
2021-02-10 02:58:58 -05:00
bli_cpuid.c
Update zen4 test in bli_cpuid.c
2022-09-01 05:12:24 -04:00
bli_cpuid.h
Added API to check for BF16 ISA support
2022-08-25 11:00:35 -04:00
bli_env.c
BLIS: Improve architecture selection at runtime
2022-08-19 10:59:35 -04:00
bli_env.h
BLIS: Improve architecture selection at runtime
2022-08-10 08:26:49 -04:00
bli_error.c
Implemented runtime subconfig selection (#451).
2020-10-18 18:04:03 -05:00
bli_error.h
BLIS:merge:
2021-04-27 11:09:48 +05:30
bli_func.c
Replaced use of bool_t type with C99 bool.
2020-08-03 11:27:13 +05:30
bli_func.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_getopt.c
Export functions without def file (#303)
2020-08-03 11:46:07 +05:30
bli_getopt.h
Add symbol export macro for all functions (#302)
2019-08-23 14:18:07 +05:30
bli_gks.c
Added support for zen4 architecture
2021-11-23 10:29:15 +05:30
bli_gks.h
BLIS:merge:
2021-04-27 11:09:48 +05:30
bli_info.c
Updated BLIS version string format
2022-08-12 05:53:35 +00:00
bli_info.h
Merge branch 'master' into amd
2019-10-11 11:53:51 -05:00
bli_init.c
BLIS: Nested parallelism issues
2022-10-21 07:38:39 -04:00
bli_init.h
Removed export macros from all internal prototypes.
2019-08-23 14:18:07 +05:30
bli_machval.c
Export functions without def file (#303)
2020-08-03 11:46:07 +05:30
bli_machval.h
Removed export macros from all internal prototypes.
2019-08-23 14:18:07 +05:30
bli_malloc.c
Export functions without def file (#303)
2020-08-03 11:46:07 +05:30
bli_malloc.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_mbool.c
Replaced use of bool_t type with C99 bool.
2020-08-03 11:27:13 +05:30
bli_mbool.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_mem.h
Squash-merge 'pr' into 'squash'. (#457)
2020-11-14 09:39:48 -06:00
bli_membrk.c
Merge branch 'master' into amd
2019-10-11 11:53:51 -05:00
bli_membrk.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_memsys.c
This check in has changes w.r.t Copyright information, which is changed to (start year) - 2019
2019-05-27 16:24:43 +05:30
bli_memsys.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_obj_scalar.c
Replaced use of bool_t type with C99 bool.
2020-08-03 11:27:13 +05:30
bli_obj_scalar.h
Replaced use of bool_t type with C99 bool.
2020-08-03 11:27:13 +05:30
bli_obj.c
Code review + tweaks to AMD's AOCL 2.0 PR (#349).
2019-10-11 10:24:24 -05:00
bli_obj.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_opid.h
Replaced use of bool_t type with C99 bool.
2020-08-03 11:27:13 +05:30
bli_pack.c
Fixed bug in packing API
2021-11-12 08:58:50 +05:30
bli_pack.h
Rebased amd-staging-milan-3.0 branch on master
2020-08-06 10:09:29 +05:30
bli_param_map.c
Updated level-3 BLAS to call object API directly.
2019-08-23 14:18:07 +05:30
bli_param_map.h
Declare/define static functions via BLIS_INLINE.
2020-08-03 11:23:40 +05:30
bli_part.c
Fixed obscure bug in bli_acquire_mpart_[mn]dim().
2019-11-14 16:48:14 -06:00
bli_part.h
Removed export macros from all internal prototypes.
2019-08-23 14:18:07 +05:30
bli_pool.c
Copyright modification
2022-10-14 12:43:35 +05:30
bli_pool.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_prune.c
This check in has changes w.r.t Copyright information, which is changed to (start year) - 2019
2019-05-27 16:24:43 +05:30
bli_prune.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_query.c
Replaced use of bool_t type with C99 bool.
2020-08-03 11:27:13 +05:30
bli_query.h
Replaced use of bool_t type with C99 bool.
2020-08-03 11:27:13 +05:30
bli_rntm.c
BLIS: Nested Parallelism issues (3)
2022-12-13 12:38:40 -05:00
bli_rntm.h
BLIS: Nested parallelism issues
2022-10-21 07:38:39 -04:00
bli_sba.c
This check in has changes w.r.t Copyright information, which is changed to (start year) - 2019
2019-05-27 16:24:43 +05:30
bli_sba.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_setgetij.c
Export functions without def file (#303)
2020-08-03 11:46:07 +05:30
bli_setgetij.h
Add symbol export macro for all functions (#302)
2019-08-23 14:18:07 +05:30
bli_setri.c
Remove UT-Austin from copyright headers' clause 3.
2018-12-04 14:31:06 -06:00
bli_setri.h
Add symbol export macro for all functions (#302)
2019-08-23 14:18:07 +05:30
bli_string.c
Remove UT-Austin from copyright headers' clause 3.
2018-12-04 14:31:06 -06:00
bli_string.h
Removed export macros from all internal prototypes.
2020-08-03 11:47:18 +05:30
bli_winsys.c
Remove UT-Austin from copyright headers' clause 3.
2018-12-04 14:31:06 -06:00
bli_winsys.h
Annotated additional symbols for export.
2020-08-03 11:47:18 +05:30
CMakeLists.txt
AOCL Windows BLIS : Windows build for dynamic dispatch library
2021-11-12 08:58:57 +05:30