Merge pull request #3304 from xianyi/develop

Merge develop int0 0.3.0 for release 0.3.16

Author: martin-frbg

.github/workflows/nightly-Homebrew-build.yml

@@ -43,11 +43,6 @@ jobs:
       - name: Update Homebrew
         if: github.event_name != 'pull_request'
         run: brew update || true
-
-      - name: unlink installed gcc to allow updating
-        run: |
-          brew unlink gcc@8
-          brew unlink gcc@9
 
       - name: Install prerequisites
         run: brew install --fetch-HEAD --HEAD --only-dependencies --keep-tmp openblas

CONTRIBUTORS.md

@@ -194,3 +194,6 @@ In chronological order:
 
 * PingTouGe Semiconductor Co., Ltd.
   * [2020-10] Add RISC-V Vector (0.7.1) support. Optimize BLAS kernels for Xuantie C910
+
+* River Dillon <[email protected]>
+  * [2021-07-10] fix compilation with musl libc

Changelog.txt

@@ -1,4 +1,52 @@
 OpenBLAS ChangeLog
+====================================================================
+Version 0.3.16
+ 11-Jul-2021
+
+common:
+ - drastically reduced the stack size requirements for running the LAPACK
+   testsuite (Reference-LAPACK PR 553)
+ - fixed spurious test failures in the LAPACK testsuite (Reference-LAPACK
+   PR 564)
+ - expressly setting DYNAMIC_ARCH=0 no longer enables dynamic_arch mode 
+ - improved performance of xGER, xSPR, xSPR2, xSYR, xSYR2, xTRSV, SGEMV_N 
+   and DGEMV_N, for small input sizes and consecutive arguments
+ - improved performance of xGETRF, xPORTF and xPOTRI for small input sizes
+   by disabling multithreading
+ - fixed installing with BSD versions of the "install" utility
+
+RISCV:
+ - fixed the implementation of xIMIN
+ - improved the performance of DSDOT
+ - fixed linking of the tests on C910V with current vendor gcc
+
+POWER:
+- fixed SBGEMM computation for some odd value inputs
+- fixed compilation for PPCG4, PPC970, POWER3, POWER4 and POWER5
+
+x86_64:
+ - improved performance of SGEMV_N and SGEMV_T for small N on AVX512-capable cpus
+ - worked around a miscompilation of ZGEMM/ZTRMM on Sandybridge with old gcc
+   versions
+ - fixed compilation with MS Visual Studio versions older than 2017
+ - fixed macro name collision with winnt.h from the latest Win10 SDK
+ - added cpu type autodetection for Intel Ice Lake SP
+ - fixed cpu type autodetection for Intel Tiger Lake
+ - added cpu type autodetection for recent Centaur/Zhaoxin models
+ - fixed compilation with musl libc
+
+ARM64:
+- fixed compilation with gcc/gfortran on the Apple M1
+- fixed linking of the tests on FreeBSD
+- fixed missing restore of a register in the recently rewritten DNRM2 kernel
+  for ThunderX2 and Neoverse N1 that could cause spurious failures in e.g.
+  DGEEV
+- added compiler optimization flags for the EMAG8180
+- added initial support for Cortex A55
+
+ARM:
+- fixed linking of the tests on FreeBSD
+
 ====================================================================
 Version 0.3.15
   2-May-2021

Makefile.arm64

@@ -57,6 +57,28 @@ endif
 endif
 endif
 
+# Use a53 tunings because a55 is only available in GCC>=8.1
+ifeq ($(CORE), CORTEXA55)
+ifeq ($(GCCVERSIONGTEQ7), 1)
+ifeq ($(GCCVERSIONGTEQ8), 1)
+CCOMMON_OPT += -march=armv8.2-a -mtune=cortex-a55
+ifneq ($(F_COMPILER), NAG)
+FCOMMON_OPT += -march=armv8.2-a -mtune=cortex-a55
+endif
+else
+CCOMMON_OPT += -march=armv8.2-a -mtune=cortex-a53
+ifneq ($(F_COMPILER), NAG)
+FCOMMON_OPT += -march=armv8.2-a -mtune=cortex-a53
+endif
+endif
+else
+CCOMMON_OPT += -march=armv8-a -mtune=cortex-a53
+ifneq ($(F_COMPILER), NAG)
+FCOMMON_OPT += -march=armv8-a -mtune=cortex-a53
+endif
+endif
+endif
+
 ifeq ($(CORE), THUNDERX)
 CCOMMON_OPT += -march=armv8-a -mtune=thunderx
 ifneq ($(F_COMPILER), NAG)
@@ -107,4 +129,13 @@ FCOMMON_OPT += -march=armv8.2-a -mtune=tsv110
 endif
 endif
 endif
+
+ifeq ($(GCCVERSIONGTEQ9), 1)
+ifeq ($(CORE), EMAG8180)
+CCOMMON_OPT += -march=armv8-a -mtune=emag
+ifneq ($(F_COMPILER), NAG)
+FCOMMON_OPT += -march=armv8-a -mtune=emag
+endif
+endif
+endif
 endif

Makefile.install

@@ -74,25 +74,25 @@ endif
 ifneq ($(OSNAME), AIX)
 ifndef NO_LAPACKE
 	@echo Copying LAPACKE header files to $(DESTDIR)$(OPENBLAS_INCLUDE_DIR)
-	@-install -pm644 $(NETLIB_LAPACK_DIR)/LAPACKE/include/lapack.h "$(DESTDIR)$(OPENBLAS_INCLUDE_DIR)/lapack.h"
-	@-install -pm644 $(NETLIB_LAPACK_DIR)/LAPACKE/include/lapacke.h "$(DESTDIR)$(OPENBLAS_INCLUDE_DIR)/lapacke.h"
-	@-install -pm644 $(NETLIB_LAPACK_DIR)/LAPACKE/include/lapacke_config.h "$(DESTDIR)$(OPENBLAS_INCLUDE_DIR)/lapacke_config.h"
-	@-install -pm644 $(NETLIB_LAPACK_DIR)/LAPACKE/include/lapacke_mangling_with_flags.h.in "$(DESTDIR)$(OPENBLAS_INCLUDE_DIR)/lapacke_mangling.h"
-	@-install -pm644 $(NETLIB_LAPACK_DIR)/LAPACKE/include/lapacke_utils.h "$(DESTDIR)$(OPENBLAS_INCLUDE_DIR)/lapacke_utils.h"
+	@-install -m644 $(NETLIB_LAPACK_DIR)/LAPACKE/include/lapack.h "$(DESTDIR)$(OPENBLAS_INCLUDE_DIR)/lapack.h"
+	@-install -m644 $(NETLIB_LAPACK_DIR)/LAPACKE/include/lapacke.h "$(DESTDIR)$(OPENBLAS_INCLUDE_DIR)/lapacke.h"
+	@-install -m644 $(NETLIB_LAPACK_DIR)/LAPACKE/include/lapacke_config.h "$(DESTDIR)$(OPENBLAS_INCLUDE_DIR)/lapacke_config.h"
+	@-install -m644 $(NETLIB_LAPACK_DIR)/LAPACKE/include/lapacke_mangling_with_flags.h.in "$(DESTDIR)$(OPENBLAS_INCLUDE_DIR)/lapacke_mangling.h"
+	@-install -m644 $(NETLIB_LAPACK_DIR)/LAPACKE/include/lapacke_utils.h "$(DESTDIR)$(OPENBLAS_INCLUDE_DIR)/lapacke_utils.h"
 endif
 
 #for install static library
 ifneq ($(NO_STATIC),1)
 	@echo Copying the static library to $(DESTDIR)$(OPENBLAS_LIBRARY_DIR)
-	@install -pm644 $(LIBNAME) "$(DESTDIR)$(OPENBLAS_LIBRARY_DIR)"
+	@install -m644 $(LIBNAME) "$(DESTDIR)$(OPENBLAS_LIBRARY_DIR)"
 	@cd "$(DESTDIR)$(OPENBLAS_LIBRARY_DIR)" ; \
 	ln -fs $(LIBNAME) $(LIBPREFIX).$(LIBSUFFIX)
 endif
 #for install shared library
 ifneq ($(NO_SHARED),1)
 	@echo Copying the shared library to $(DESTDIR)$(OPENBLAS_LIBRARY_DIR)
 ifeq ($(OSNAME), $(filter $(OSNAME),Linux SunOS Android Haiku FreeBSD DragonFly))
-	@install -pm755 $(LIBSONAME) "$(DESTDIR)$(OPENBLAS_LIBRARY_DIR)"
+	@install -m755 $(LIBSONAME) "$(DESTDIR)$(OPENBLAS_LIBRARY_DIR)"
 	@cd "$(DESTDIR)$(OPENBLAS_LIBRARY_DIR)" ; \
 	ln -fs $(LIBSONAME) $(LIBPREFIX).so ; \
 	ln -fs $(LIBSONAME) $(LIBPREFIX).so.$(MAJOR_VERSION)

Makefile.system

@@ -333,6 +333,7 @@ GCCVERSIONGTEQ4 := $(shell expr `$(CC) -dumpversion | cut -f1 -d.` \>= 4)
 GCCVERSIONGT4 := $(shell expr `$(CC) -dumpversion | cut -f1 -d.` \> 4)
 GCCVERSIONGT5 := $(shell expr `$(CC) -dumpversion | cut -f1 -d.` \> 5)
 GCCVERSIONGTEQ7 := $(shell expr `$(CC) -dumpversion | cut -f1 -d.` \>= 7)
+GCCVERSIONGTEQ8 := $(shell expr `$(CC) -dumpversion | cut -f1 -d.` \>= 8)
 GCCVERSIONGTEQ9 := $(shell expr `$(CC) -dumpversion | cut -f1 -d.` \>= 9)
 GCCVERSIONGTEQ11 := $(shell expr `$(CC) -dumpversion | cut -f1 -d.` \>= 11)
 GCCVERSIONGTEQ10 := $(shell expr `$(CC) -dumpversion | cut -f1 -d.` \>= 10)
@@ -380,6 +381,12 @@ ifeq ($(OSNAME), AIX)
 EXTRALIB	+= -lm
 endif
 
+ifeq ($(OSNAME), FreeBSD)
+ifeq ($(ARCH), $(filter $(ARCH),arm arm64))
+EXTRALIB	+= -lm
+endif
+endif
+
 ifeq ($(OSNAME), WINNT)
 NEED_PIC = 0
 NO_EXPRECISION = 1
@@ -619,6 +626,7 @@ DYNAMIC_CORE += CORTEXA57
 DYNAMIC_CORE += CORTEXA72
 DYNAMIC_CORE += CORTEXA73
 DYNAMIC_CORE += NEOVERSEN1
+DYNAMIC_CORE += CORTEXA55
 DYNAMIC_CORE += FALKOR
 DYNAMIC_CORE += THUNDERX
 DYNAMIC_CORE += THUNDERX2T99

Makefile.x86

@@ -1,6 +1,6 @@
 # COMPILER_PREFIX = mingw32-
 
-ifndef DYNAMIC_ARCH
+ifneq ($(DYNAMIC_ARCH),1)
 ADD_CPUFLAGS = 1
 else
 ifdef TARGET_CORE

Makefile.x86_64

@@ -9,7 +9,7 @@ endif
 endif
 
 
-ifndef DYNAMIC_ARCH
+ifneq ($(DYNAMIC_ARCH),1)
 ADD_CPUFLAGS = 1
 else
 ifdef TARGET_CORE

README.md

@@ -27,7 +27,7 @@ We provide official binary packages for the following platform:
 
   * Windows x86/x86_64
 
-You can download them from [file hosting on sourceforge.net](https://sourceforge.net/projects/openblas/files/).
+You can download them from [file hosting on sourceforge.net](https://sourceforge.net/projects/openblas/files/) or from the Releases section of the github project page, [https://github.com/xianyi/OpenBLAS/releases](https://github.com/xianyi/OpenBLAS/releases).
 
 ## Installation from Source
 

TargetList.txt

@@ -92,6 +92,7 @@ CORTEXA57
 CORTEXA72
 CORTEXA73
 NEOVERSEN1
+CORTEXA55
 EMAG8180
 FALKOR
 THUNDERX