Skip to content

Commit 838bb57

Browse files
michalowski-armmichalowski-arm
authored
Merge branch 'develop' into develop
2 parents 4d5b13f + 876ba58 commit 838bb57

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

50 files changed

+852
-164
lines changed

‎CONTRIBUTORS.md‎

Lines changed: 4 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -233,5 +233,8 @@ In chronological order:
233233
* Aniket P. Garade <https://github.com/garadeaniket> Sushil Pratap Singh <https://github.com/SushilPratap04> Juliya James <https://github.com/Juliya32>
234234
* [2024-12-13] Optimized swap and rot Level-1 BLAS routines with ARM SVE
235235

236+
* Annop Wongwathanarat <annop.wongwathanarat@arm.com>
237+
* [2025-01-10] Add thread throttling profile for SGEMM on NEOVERSEV1
238+
236239
* Marek Michalowski <https://github.com/michalowski-arm>
237-
* [2025-01-21] Add thread throttling profile for SGEMV on `NEOVERSEV1`
240+
* [2025-01-21] Add thread throttling profile for SGEMV on `NEOVERSEV1`

‎Makefile.install‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -315,8 +315,8 @@ endif
315315

316316
endif
317317
ifeq ($(CPP_THREAD_SAFETY_TEST), 1)
318-
@install -m 666 cpp_thread_test/dgemm_tester $(DESTDIR)$(OPENBLAS_BINARY_DIR)
319-
@install -m 666 cpp_thread_test/dgemv_tester $(DESTDIR)$(OPENBLAS_BINARY_DIR)
318+
@install -m 666 cpp_thread_test/dgemm_tester $(DESTDIR)$(OPENBLAS_BINARY_DIR)
319+
@install -m 666 cpp_thread_test/dgemv_tester $(DESTDIR)$(OPENBLAS_BINARY_DIR)
320320
endif
321321
endif
322322

‎cmake/kernel.cmake‎

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -79,6 +79,9 @@ macro(SetDefaultL1)
7979
SetFallback(CROTKERNEL zrot.S)
8080
SetFallback(ZROTKERNEL zrot.S)
8181
SetFallback(XROTKERNEL zrot.S)
82+
SetFallback(SROTMKERNEL rotm.S)
83+
SetFallback(DROTMKERNEL rotm.S)
84+
SetFallback(QROTMKERNEL rotm.S)
8285
SetFallback(SSCALKERNEL scal.S)
8386
SetFallback(DSCALKERNEL scal.S)
8487
SetFallback(CSCALKERNEL zscal.S)

‎cmake/system.cmake‎

Lines changed: 9 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -21,7 +21,15 @@ endif()
2121
# Other files expect CORE, which is actually TARGET and will become TARGET_CORE for kernel build. Confused yet?
2222
# It seems we are meant to use TARGET as input and CORE internally as kernel.
2323
if(NOT DEFINED CORE AND DEFINED TARGET)
24-
set(CORE ${TARGET})
24+
if (${TARGET} STREQUAL "LOONGSON3R5")
25+
set(CORE "LA464")
26+
elseif (${TARGET} STREQUAL "LOONGSON2K1000")
27+
set(CORE "LA264")
28+
elseif (${TARGET} STREQUAL "LOONGSONGENERIC")
29+
set(CORE "LA64_GENERIC)")
30+
else ()
31+
set(CORE ${TARGET})
32+
endif()
2533
endif()
2634

2735
# TARGET_CORE will override TARGET which is used in DYNAMIC_ARCH=1.

‎cmake/utils.cmake‎

Lines changed: 8 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -16,6 +16,14 @@ endfunction ()
1616
macro(ParseMakefileVars MAKEFILE_IN)
1717
message(STATUS "Reading vars from ${MAKEFILE_IN}...")
1818
set (C_COMPILER ${CMAKE_C_COMPILER_ID})
19+
set (OSNAME ${CMAKE_SYSTEM_NAME})
20+
if (${C_COMPILER} MATCHES Clang)
21+
set (C_COMPILER CLANG)
22+
endif ()
23+
if (${OSNAME} STREQUAL Windows)
24+
set (OSNAME WINNT)
25+
endif ()
26+
message(STATUS OS ${OSNAME} COMPILER ${C_COMPILER})
1927
set (IfElse 0)
2028
set (ElseSeen 0)
2129
set (SkipIfs 0)

‎common_d.h‎

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -22,6 +22,7 @@
2222
#define DSUM_K dsum_k
2323
#define DSWAP_K dswap_k
2424
#define DROT_K drot_k
25+
#define DROTM_K drotm_k
2526

2627
#define DGEMV_N dgemv_n
2728
#define DGEMV_T dgemv_t
@@ -180,6 +181,7 @@
180181
#define DSUM_K gotoblas -> dsum_k
181182
#define DSWAP_K gotoblas -> dswap_k
182183
#define DROT_K gotoblas -> drot_k
184+
#define DROTM_K gotoblas -> drotm_k
183185

184186
#define DGEMV_N gotoblas -> dgemv_n
185187
#define DGEMV_T gotoblas -> dgemv_t

‎common_level1.h‎

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -213,9 +213,9 @@ int srotmg_k(float *, float *, float *, float *, float *);
213213
int drotmg_k(double *, double *, double *, double *, double *);
214214
int qrotmg_k(xdouble *, xdouble *, xdouble *, xdouble *, xdouble *);
215215

216-
int srotm_k (BLASLONG, float, BLASLONG, float, BLASLONG, float);
217-
int drotm_k (BLASLONG, double, BLASLONG, double, BLASLONG, double);
218-
int qrotm_k (BLASLONG, xdouble, BLASLONG, xdouble, BLASLONG, xdouble);
216+
int srotm_k (BLASLONG, float *, BLASLONG, float *, BLASLONG, float *);
217+
int drotm_k (BLASLONG, double *, BLASLONG, double *, BLASLONG, double *);
218+
int qrotm_k (BLASLONG, xdouble *, BLASLONG, xdouble *, BLASLONG, xdouble *);
219219

220220

221221
int saxpby_k (BLASLONG, float, float *, BLASLONG, float, float *, BLASLONG);

‎common_macro.h‎

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -70,6 +70,7 @@
7070
#define SUM_K QSUM_K
7171
#define SWAP_K QSWAP_K
7272
#define ROT_K QROT_K
73+
#define ROTM_K QROTM_K
7374

7475
#define GEMV_N QGEMV_N
7576
#define GEMV_T QGEMV_T
@@ -361,6 +362,7 @@
361362
#define SUM_K DSUM_K
362363
#define SWAP_K DSWAP_K
363364
#define ROT_K DROT_K
365+
#define ROTM_K DROTM_K
364366

365367
#define GEMV_N DGEMV_N
366368
#define GEMV_T DGEMV_T
@@ -977,6 +979,7 @@
977979
#define SUM_K SSUM_K
978980
#define SWAP_K SSWAP_K
979981
#define ROT_K SROT_K
982+
#define ROTM_K SROTM_K
980983

981984
#define GEMV_N SGEMV_N
982985
#define GEMV_T SGEMV_T

‎common_param.h‎

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -197,6 +197,7 @@ BLASLONG (*ismin_k) (BLASLONG, float *, BLASLONG);
197197
//double (*dsdot_k) (BLASLONG, float *, BLASLONG, float *, BLASLONG);
198198

199199
int (*srot_k) (BLASLONG, float *, BLASLONG, float *, BLASLONG, float, float);
200+
int (*srotm_k) (BLASLONG, float *, BLASLONG, float *, BLASLONG, float *);
200201
#endif
201202
#if (BUILD_SINGLE==1) || (BUILD_DOUBLE==1) || (BUILD_COMPLEX==1)
202203
int (*saxpy_k) (BLASLONG, BLASLONG, BLASLONG, float, float *, BLASLONG, float *, BLASLONG, float *, BLASLONG);
@@ -330,6 +331,7 @@ BLASLONG (*idmin_k) (BLASLONG, double *, BLASLONG);
330331
#endif
331332
#if (BUILD_DOUBLE==1) || (BUILD_COMPLEX16==1)
332333
int (*drot_k) (BLASLONG, double *, BLASLONG, double *, BLASLONG, double, double);
334+
int (*drotm_k) (BLASLONG, double *, BLASLONG, double *, BLASLONG, double *);
333335
int (*daxpy_k) (BLASLONG, BLASLONG, BLASLONG, double, double *, BLASLONG, double *, BLASLONG, double *, BLASLONG);
334336
int (*dscal_k) (BLASLONG, BLASLONG, BLASLONG, double, double *, BLASLONG, double *, BLASLONG, double *, BLASLONG);
335337
int (*dswap_k) (BLASLONG, BLASLONG, BLASLONG, double, double *, BLASLONG, double *, BLASLONG, double *, BLASLONG);
@@ -439,6 +441,7 @@ BLASLONG (*iqmin_k) (BLASLONG, xdouble *, BLASLONG);
439441
int (*qcopy_k) (BLASLONG, xdouble *, BLASLONG, xdouble *, BLASLONG);
440442
xdouble (*qdot_k) (BLASLONG, xdouble *, BLASLONG, xdouble *, BLASLONG);
441443
int (*qrot_k) (BLASLONG, xdouble *, BLASLONG, xdouble *, BLASLONG, xdouble, xdouble);
444+
int (*qrotm_k) (BLASLONG, xdouble *, BLASLONG, xdouble *, BLASLONG, xdouble *);
442445

443446
int (*qaxpy_k) (BLASLONG, BLASLONG, BLASLONG, xdouble, xdouble *, BLASLONG, xdouble *, BLASLONG, xdouble *, BLASLONG);
444447
int (*qscal_k) (BLASLONG, BLASLONG, BLASLONG, xdouble, xdouble *, BLASLONG, xdouble *, BLASLONG, xdouble *, BLASLONG);

‎common_q.h‎

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -22,6 +22,7 @@
2222
#define QSUM_K qsum_k
2323
#define QSWAP_K qswap_k
2424
#define QROT_K qrot_k
25+
#define QROTM_K qrotm_k
2526

2627
#define QGEMV_N qgemv_n
2728
#define QGEMV_T qgemv_t
@@ -165,6 +166,7 @@
165166
#define QSUM_K gotoblas -> qsum_k
166167
#define QSWAP_K gotoblas -> qswap_k
167168
#define QROT_K gotoblas -> qrot_k
169+
#define QROTM_K gotoblas -> qrotm_k
168170

169171
#define QGEMV_N gotoblas -> qgemv_n
170172
#define QGEMV_T gotoblas -> qgemv_t

0 commit comments

Comments
 (0)