@@ -73,15 +73,15 @@ SBLAS3OBJS = \
7373 strsm.$(SUFFIX ) ssyrk.$(SUFFIX ) ssyr2k.$(SUFFIX ) \
7474 somatcopy.$(SUFFIX ) simatcopy.$(SUFFIX ) \
7575 sgeadd.$(SUFFIX ) sgemmt.$(SUFFIX ) sgemmtr.$(SUFFIX ) \
76- sgemm_batch.$(SUFFIX )
76+ sgemm_batch.$(SUFFIX ) sgemm_batch_strided. $( SUFFIX )
7777
7878ifeq ($(BUILD_BFLOAT16 ) ,1)
7979BBLAS3OBJS = bgemm.$(SUFFIX )
8080BBLAS2OBJS = bgemv.$(SUFFIX )
8181BBLAS1OBJS = bscal.$(SUFFIX )
8282SBBLAS1OBJS = sbdot.$(SUFFIX )
8383SBBLAS2OBJS = sbgemv.$(SUFFIX )
84- SBBLAS3OBJS = sbgemm.$(SUFFIX ) sbgemmt.$(SUFFIX ) sbgemmtr.$(SUFFIX ) sbgemm_batch.$(SUFFIX )
84+ SBBLAS3OBJS = sbgemm.$(SUFFIX ) sbgemmt.$(SUFFIX ) sbgemmtr.$(SUFFIX ) sbgemm_batch.$(SUFFIX ) sbgemm_batch_strided. $( SUFFIX )
8585SBEXTOBJS = sbstobf16.$(SUFFIX ) sbdtobf16.$(SUFFIX ) sbf16tos.$(SUFFIX ) dbf16tod.$(SUFFIX )
8686endif
8787
@@ -113,7 +113,7 @@ DBLAS3OBJS = \
113113 dtrsm.$(SUFFIX ) dsyrk.$(SUFFIX ) dsyr2k.$(SUFFIX ) \
114114 domatcopy.$(SUFFIX ) dimatcopy.$(SUFFIX ) \
115115 dgeadd.$(SUFFIX ) dgemmt.$(SUFFIX ) dgemmtr.$(SUFFIX ) \
116- dgemm_batch.$(SUFFIX )
116+ dgemm_batch.$(SUFFIX ) dgemm_batch_strided. $( SUFFIX )
117117
118118CBLAS1OBJS = \
119119 caxpy.$(SUFFIX ) caxpyc.$(SUFFIX ) cswap.$(SUFFIX ) \
@@ -143,7 +143,7 @@ CBLAS3OBJS = \
143143 chemm.$(SUFFIX ) cherk.$(SUFFIX ) cher2k.$(SUFFIX ) \
144144 comatcopy.$(SUFFIX ) cimatcopy.$(SUFFIX ) \
145145 cgeadd.$(SUFFIX ) cgemmt.$(SUFFIX ) cgemmtr.$(SUFFIX ) \
146- cgemm_batch.$(SUFFIX )
146+ cgemm_batch.$(SUFFIX ) cgemm_batch_strided. $( SUFFIX )
147147
148148ZBLAS1OBJS = \
149149 zaxpy.$(SUFFIX ) zaxpyc.$(SUFFIX ) zswap.$(SUFFIX ) \
@@ -173,7 +173,7 @@ ZBLAS3OBJS = \
173173 zhemm.$(SUFFIX ) zherk.$(SUFFIX ) zher2k.$(SUFFIX ) \
174174 zomatcopy.$(SUFFIX ) zimatcopy.$(SUFFIX ) \
175175 zgeadd.$(SUFFIX ) zgemmt.$(SUFFIX ) zgemmtr.$(SUFFIX ) \
176- zgemm_batch.$(SUFFIX )
176+ zgemm_batch.$(SUFFIX ) zgemm_batch_strided. $( SUFFIX )
177177
178178ifeq ($(SUPPORT_GEMM3M ) , 1)
179179
@@ -321,15 +321,15 @@ CSBLAS2OBJS = \
321321CSBLAS3OBJS = \
322322 cblas_sgemm.$(SUFFIX ) cblas_ssymm.$(SUFFIX ) cblas_strmm.$(SUFFIX ) cblas_strsm.$(SUFFIX ) \
323323 cblas_ssyrk.$(SUFFIX ) cblas_ssyr2k.$(SUFFIX ) cblas_somatcopy.$(SUFFIX ) cblas_simatcopy.$(SUFFIX ) \
324- cblas_sgeadd.$(SUFFIX ) cblas_sgemmt.$(SUFFIX ) cblas_sgemmtr.$(SUFFIX ) cblas_sgemm_batch.$(SUFFIX )
324+ cblas_sgeadd.$(SUFFIX ) cblas_sgemmt.$(SUFFIX ) cblas_sgemmtr.$(SUFFIX ) cblas_sgemm_batch.$(SUFFIX ) cblas_sgemm_batch_strided. $( SUFFIX )
325325
326326ifeq ($(BUILD_BFLOAT16 ) ,1)
327327CBBLAS3OBJS = cblas_bgemm.$(SUFFIX )
328328CBBLAS2OBJS = cblas_bgemv.$(SUFFIX )
329329CBBLAS1OBJS = cblas_bscal.$(SUFFIX )
330330CSBBLAS1OBJS = cblas_sbdot.$(SUFFIX )
331331CSBBLAS2OBJS = cblas_sbgemv.$(SUFFIX )
332- CSBBLAS3OBJS = cblas_sbgemm.$(SUFFIX ) cblas_sbgemmt.$(SUFFIX ) cblas_sbgemmtr.$(SUFFIX ) cblas_sbgemm_batch.$(SUFFIX )
332+ CSBBLAS3OBJS = cblas_sbgemm.$(SUFFIX ) cblas_sbgemmt.$(SUFFIX ) cblas_sbgemmtr.$(SUFFIX ) cblas_sbgemm_batch.$(SUFFIX ) cblas_sbgemm_batch_strided. $( SUFFIX )
333333CSBEXTOBJS = cblas_sbstobf16.$(SUFFIX ) cblas_sbdtobf16.$(SUFFIX ) cblas_sbf16tos.$(SUFFIX ) cblas_dbf16tod.$(SUFFIX )
334334ifeq ($(ONLY_CBLAS ) ,1)
335335CSBEXTOBJS += sbstobf16.$(SUFFIX ) sbdtobf16.$(SUFFIX ) sbf16tos.$(SUFFIX ) dbf16tod.$(SUFFIX )
@@ -357,7 +357,7 @@ CDBLAS2OBJS = \
357357CDBLAS3OBJS += \
358358 cblas_dgemm.$(SUFFIX ) cblas_dsymm.$(SUFFIX ) cblas_dtrmm.$(SUFFIX ) cblas_dtrsm.$(SUFFIX ) \
359359 cblas_dsyrk.$(SUFFIX ) cblas_dsyr2k.$(SUFFIX ) cblas_domatcopy.$(SUFFIX ) cblas_dimatcopy.$(SUFFIX ) \
360- cblas_dgeadd.$(SUFFIX ) cblas_dgemmt.$(SUFFIX ) cblas_dgemmtr.$(SUFFIX ) cblas_dgemm_batch.$(SUFFIX )
360+ cblas_dgeadd.$(SUFFIX ) cblas_dgemmt.$(SUFFIX ) cblas_dgemmtr.$(SUFFIX ) cblas_dgemm_batch.$(SUFFIX ) cblas_dgemm_batch_strided. $( SUFFIX )
361361
362362CCBLAS1OBJS = \
363363 cblas_icamax.$(SUFFIX ) cblas_icamin.$(SUFFIX ) cblas_scasum.$(SUFFIX ) cblas_caxpy.$(SUFFIX ) \
@@ -382,7 +382,7 @@ CCBLAS3OBJS = \
382382 cblas_csyrk.$(SUFFIX ) cblas_csyr2k.$(SUFFIX ) \
383383 cblas_chemm.$(SUFFIX ) cblas_cherk.$(SUFFIX ) cblas_cher2k.$(SUFFIX ) \
384384 cblas_comatcopy.$(SUFFIX ) cblas_cimatcopy.$(SUFFIX ) \
385- cblas_cgeadd.$(SUFFIX ) cblas_cgemmt.$(SUFFIX ) cblas_cgemmtr.$(SUFFIX ) cblas_cgemm_batch.$(SUFFIX )
385+ cblas_cgeadd.$(SUFFIX ) cblas_cgemmt.$(SUFFIX ) cblas_cgemmtr.$(SUFFIX ) cblas_cgemm_batch.$(SUFFIX ) cblas_cgemm_batch_strided. $( SUFFIX )
386386
387387CXERBLAOBJ = \
388388 cblas_xerbla.$(SUFFIX )
@@ -413,7 +413,7 @@ CZBLAS3OBJS = \
413413 cblas_zsyrk.$(SUFFIX ) cblas_zsyr2k.$(SUFFIX ) \
414414 cblas_zhemm.$(SUFFIX ) cblas_zherk.$(SUFFIX ) cblas_zher2k.$(SUFFIX ) \
415415 cblas_zomatcopy.$(SUFFIX ) cblas_zimatcopy.$(SUFFIX ) \
416- cblas_zgeadd.$(SUFFIX ) cblas_zgemmt.$(SUFFIX ) cblas_zgemmtr.$(SUFFIX ) cblas_zgemm_batch.$(SUFFIX )
416+ cblas_zgeadd.$(SUFFIX ) cblas_zgemmt.$(SUFFIX ) cblas_zgemmtr.$(SUFFIX ) cblas_zgemm_batch.$(SUFFIX ) cblas_zgemm_batch_strided. $( SUFFIX )
417417
418418
419419ifeq ($(SUPPORT_GEMM3M ) , 1)
@@ -2544,6 +2544,21 @@ cblas_cgemm_batch.$(SUFFIX) cblas_cgemm_batch.$(PSUFFIX) : gemm_batch.c ../param
25442544cblas_zgemm_batch.$(SUFFIX ) cblas_zgemm_batch.$(PSUFFIX ) : gemm_batch.c ../param.h
25452545 $(CC ) -c $(CFLAGS ) -DCBLAS $< -o $(@F )
25462546
2547+ cblas_sbgemm_batch_strided.$(SUFFIX ) cblas_sbgemm_batch_strided.$(PSUFFIX ) : gemm_batch_strided.c ../param.h
2548+ $(CC ) -c $(CFLAGS ) -DCBLAS $< -o $(@F )
2549+
2550+ cblas_sgemm_batch_strided.$(SUFFIX ) cblas_sgemm_batch_strided.$(PSUFFIX ) : gemm_batch_strided.c ../param.h
2551+ $(CC ) -c $(CFLAGS ) -DCBLAS $< -o $(@F )
2552+
2553+ cblas_dgemm_batch_strided.$(SUFFIX ) cblas_dgemm_batch_strided.$(PSUFFIX ) : gemm_batch_strided.c ../param.h
2554+ $(CC ) -c $(CFLAGS ) -DCBLAS $< -o $(@F )
2555+
2556+ cblas_cgemm_batch_strided.$(SUFFIX ) cblas_cgemm_batch_strided.$(PSUFFIX ) : gemm_batch_strided.c ../param.h
2557+ $(CC ) -c $(CFLAGS ) -DCBLAS $< -o $(@F )
2558+
2559+ cblas_zgemm_batch_strided.$(SUFFIX ) cblas_zgemm_batch_strided.$(PSUFFIX ) : gemm_batch_strided.c ../param.h
2560+ $(CC ) -c $(CFLAGS ) -DCBLAS $< -o $(@F )
2561+
25472562sbgemm_batch.$(SUFFIX ) sbgemm_batch.$(PSUFFIX ) : gemm_batch.c ../param.h
25482563 $(CC ) -c $(CFLAGS ) -UCBLAS $< -o $(@F )
25492564
@@ -2559,3 +2574,17 @@ cgemm_batch.$(SUFFIX) cgemm_batch.$(PSUFFIX) : gemm_batch.c ../param.h
25592574zgemm_batch.$(SUFFIX ) zgemm_batch.$(PSUFFIX ) : gemm_batch.c ../param.h
25602575 $(CC ) -c $(CFLAGS ) -UCBLAS $< -o $(@F )
25612576
2577+ sbgemm_batch_strided.$(SUFFIX ) sbgemm_batch_strided.$(PSUFFIX ) : gemm_batch_strided.c ../param.h
2578+ $(CC ) -c $(CFLAGS ) -UCBLAS $< -o $(@F )
2579+
2580+ sgemm_batch_strided.$(SUFFIX ) sgemm_batch_strided.$(PSUFFIX ) : gemm_batch_strided.c ../param.h
2581+ $(CC ) -c $(CFLAGS ) -UCBLAS $< -o $(@F )
2582+
2583+ dgemm_batch_strided.$(SUFFIX ) dgemm_batch_strided.$(PSUFFIX ) : gemm_batch_strided.c ../param.h
2584+ $(CC ) -c $(CFLAGS ) -UCBLAS $< -o $(@F )
2585+
2586+ cgemm_batch_strided.$(SUFFIX ) cgemm_batch_strided.$(PSUFFIX ) : gemm_batch_strided.c ../param.h
2587+ $(CC ) -c $(CFLAGS ) -UCBLAS $< -o $(@F )
2588+
2589+ zgemm_batch_strided.$(SUFFIX ) zgemm_batch_strided.$(PSUFFIX ) : gemm_batch_strided.c ../param.h
2590+ $(CC ) -c $(CFLAGS ) -UCBLAS $< -o $(@F )
0 commit comments