Merge pull request #3150 from xianyi/develop

Update branch from develop for 0.3.14 release

Author: martin-frbg

.drone.yml

@@ -190,3 +190,27 @@ steps:
     - make -C ctest $COMMON_FLAGS
     - make -C utest $COMMON_FLAGS
     - make -C cpp_thread_test dgemm_tester
+---
+kind: pipeline
+name: arm64_gcc10
+
+platform:
+  os: linux
+  arch: arm64
+
+steps:
+- name: Build and Test
+  image: ubuntu:20.04
+  environment:
+    CC: gcc-10
+    FC: gfortran-10
+    COMMON_FLAGS: 'TARGET=ARMV8 DYNAMIC_ARCH=1'
+  commands:
+    - echo "MAKE_FLAGS:= $COMMON_FLAGS"
+    - apt-get update -y
+    - apt-get install -y make $CC gfortran-10 perl python g++
+    - $CC --version
+    - make QUIET_MAKE=1 $COMMON_FLAGS
+    - make -C utest $COMMON_FLAGS
+    - make -C test $COMMON_FLAGS
+    

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

@@ -44,6 +44,11 @@ jobs:
         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
 

.gitignore

@@ -89,5 +89,7 @@ build.*
 *.swp
 benchmark/*.goto
 benchmark/smallscaling
+.vscode
 CMakeCache.txt
 CMakeFiles/*
+.vscode

CMakeLists.txt

@@ -6,14 +6,17 @@ cmake_minimum_required(VERSION 2.8.5)
 project(OpenBLAS C ASM)
 set(OpenBLAS_MAJOR_VERSION 0)
 set(OpenBLAS_MINOR_VERSION 3)
-set(OpenBLAS_PATCH_VERSION 13)
+set(OpenBLAS_PATCH_VERSION 14)
 set(OpenBLAS_VERSION "${OpenBLAS_MAJOR_VERSION}.${OpenBLAS_MINOR_VERSION}.${OpenBLAS_PATCH_VERSION}")
 
 # Adhere to GNU filesystem layout conventions
 include(GNUInstallDirs)
 
 include(CMakePackageConfigHelpers)
 
+if(MSVC AND NOT DEFINED NOFORTRAN)
+  set(NOFORTRAN ON)
+endif()
 
 #######
 if(MSVC)
@@ -229,7 +232,7 @@ if (NOT NO_CBLAS)
   add_subdirectory(utest)
 endif()
 
-if (NOT MSVC AND NOT NOFORTRAN)
+if (NOT NOFORTRAN)
   # Build test and ctest
   add_subdirectory(test)
   if(NOT NO_CBLAS)

Changelog.txt

@@ -1,4 +1,52 @@
 OpenBLAS ChangeLog
+====================================================================
+Version 0.3.14
+ 17-Mar-2021
+ 
+ common:
+	* Fixed a race condition on thread shutdown in non-OpenMP builds
+	* Fixed custom BUFFERSIZE option getting ignored in gmake builds
+	* Fixed CMAKE compilation of the TRMM kernels for GENERIC platforms
+	* Added CBLAS interfaces for CROTG, ZROTG, CSROT and ZDROT
+	* Improved performance of OMATCOPY_RT across all platforms
+	* Changed perl scripts to use env instead of a hardcoded /usr/bin/perl
+	* Fixed potential misreading of the GCC compiler version in the build scripts
+	* Fixed convergence problems in LAPACK complex GGEV/GGES (Reference-LAPACK #477)
+	* Reduced the stacksize requirements for running the LAPACK testsuite (Reference-LAPACK #335)
+
+ RISCV:
+ 	* Fixed compilation on RISCV (missing entry in getarch)
+	
+ POWER:
+ 	* Fixed compilation for DYNAMIC_ARCH with clang and with old gcc versions
+	* Added support for compilation on FreeBSD/ppc64le
+ 	* Added optimized POWER10 kernels for SSCAL, DSCAL, CSCAL, ZSCAL
+	* Added optimized POWER10 kernels for SROT, DROT, CDOT, SASUM, DASUM
+	* Improved SSWAP, DSWAP, CSWAP, ZSWAP performance on POWER10
+	* Improved SCOPY and CCOPY performance on POWER10
+	* Improved SGEMM and DGEMM performance on POWER10
+	* Added support for compilation with the NVIDIA HPC compiler
+
+ x86_64:
+	* Added an optimized bfloat16 GEMM kernel for Cooperlake
+	* Added CPUID autodetection for Intel Rocket Lake and Tiger Lake cpus
+	* Improved the performance of SASUM,DASUM,SROT,DROT on AMD Ryzen cpus
+	* Added support for compilation with the NAG Fortran compiler
+	* Fixed recognition of the AMD AOCC compiler
+	* Fixed compilation for DYNAMIC_ARCH with clang on Windows
+	* Added support for running the BLAS/CBLAS tests on Windows
+	* Fixed signatures of the tls callback functions for Windows x64
+	* Fixed various issues with fma intrinsics support handling 
+
+ ARM:
+	* Added support for embedded Cortex M targets via a new option EMBEDDED
+
+ ARMV8:
+ 	* Fixed the THUNDERX2T99 and NEOVERSEN1 DNRM2/ZNRM2 kernels for inputs with Inf
+	* Added support for the DYNAMIC_LIST option
+	* Added support for compilation with the NVIDIA HPC compiler
+ 	* Added support for compiling with the NAG Fortran compiler
+
 ====================================================================
 Version 0.3.13
  12-Dec-2020

Makefile

@@ -59,6 +59,9 @@ endif
 	@$(CC) --version > /dev/null 2>&1;\
 	if [ $$? -eq 0 ]; then \
 	   cverinfo=`$(CC) --version | sed -n '1p'`; \
+	   if [ -z "$${cverinfo}" ]; then \
+	   cverinfo=`$(CC) --version | sed -n '2p'`; \
+	   fi; \
 	   echo "  C compiler       ... $(C_COMPILER)  (cmd & version : $${cverinfo})";\
 	else  \
 	   echo "  C compiler       ... $(C_COMPILER)  (command line : $(CC))";\
@@ -67,6 +70,9 @@ ifeq ($(NOFORTRAN), $(filter 0,$(NOFORTRAN)))
 	@$(FC) --version > /dev/null 2>&1;\
 	if [ $$? -eq 0 ]; then \
 	   fverinfo=`$(FC) --version | sed -n '1p'`; \
+	   if [ -z "$${fverinfo}" ]; then \
+	   fverinfo=`$(FC) --version | sed -n '2p'`; \
+	   fi; \
 	   echo "  Fortran compiler ... $(F_COMPILER)  (cmd & version : $${fverinfo})";\
 	else \
 	   echo "  Fortran compiler ... $(F_COMPILER)  (command line : $(FC))";\

Makefile.arm64

@@ -1,80 +1,110 @@
-
+ifneq ($(C_COMPILER), PGI)
 ifeq ($(CORE), ARMV8)
 CCOMMON_OPT += -march=armv8-a
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8-a
 endif
+endif
 
 ifeq ($(CORE), CORTEXA53)
 CCOMMON_OPT += -march=armv8-a -mtune=cortex-a53
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8-a -mtune=cortex-a53
 endif
+endif
 
 ifeq ($(CORE), CORTEXA57)
 CCOMMON_OPT += -march=armv8-a -mtune=cortex-a57
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8-a -mtune=cortex-a57
 endif
+endif
 
 ifeq ($(CORE), CORTEXA72)
 CCOMMON_OPT += -march=armv8-a -mtune=cortex-a72
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8-a -mtune=cortex-a72
 endif
+endif
 
 ifeq ($(CORE), CORTEXA73)
 CCOMMON_OPT += -march=armv8-a -mtune=cortex-a73
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8-a -mtune=cortex-a73
 endif
+endif
 
 # Use a72 tunings because Neoverse-N1 is only available
 # in GCC>=9
 ifeq ($(CORE), NEOVERSEN1)
 ifeq ($(GCCVERSIONGTEQ7), 1)
 ifeq ($(GCCVERSIONGTEQ9), 1)
 CCOMMON_OPT += -march=armv8.2-a -mtune=neoverse-n1
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8.2-a -mtune=neoverse-n1
+endif
 else
 CCOMMON_OPT += -march=armv8.2-a -mtune=cortex-a72
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8.2-a -mtune=cortex-a72
 endif
+endif
 else
 CCOMMON_OPT += -march=armv8-a -mtune=cortex-a72
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8-a -mtune=cortex-a72
 endif
 endif
+endif
 
 ifeq ($(CORE), THUNDERX)
 CCOMMON_OPT += -march=armv8-a -mtune=thunderx
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8-a -mtune=thunderx
 endif
+endif
 
 ifeq ($(CORE), FALKOR)
 CCOMMON_OPT += -march=armv8-a -mtune=falkor
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8-a -mtune=falkor
 endif
+endif
 
 ifeq ($(CORE), THUNDERX2T99)
 CCOMMON_OPT += -march=armv8.1-a -mtune=thunderx2t99
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8.1-a -mtune=thunderx2t99
 endif
+endif
 
 ifeq ($(CORE), THUNDERX3T110)
 ifeq ($(GCCVERSIONGTEQ10), 1)
 CCOMMON_OPT += -march=armv8.3-a -mtune=thunderx3t110
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8.3-a -mtune=thunderx3t110
+endif
 else
 CCOMMON_OPT += -march=armv8.1-a -mtune=thunderx2t99
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8.1-a -mtune=thunderx2t99
 endif
 endif
+endif
 
 ifeq ($(CORE), VORTEX)
 CCOMMON_OPT += -march=armv8.3-a
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8.3-a
 endif
+endif
 
 ifeq ($(GCCVERSIONGTEQ9), 1)
 ifeq ($(CORE), TSV110)
 CCOMMON_OPT += -march=armv8.2-a -mtune=tsv110
+ifneq ($(F_COMPILER), NAG)
 FCOMMON_OPT += -march=armv8.2-a -mtune=tsv110
 endif
 endif
-
+endif
+endif

Makefile.power

@@ -10,9 +10,11 @@ USE_OPENMP = 1
 endif
 
 ifeq ($(CORE), POWER10)
+ifneq ($(C_COMPILER), PGI)
 CCOMMON_OPT += -Ofast -mcpu=power10 -mtune=power10 -mvsx -fno-fast-math
 FCOMMON_OPT += -O2 -frecursive -mcpu=power10 -mtune=power10  -fno-fast-math
 endif
+endif
 
 ifeq ($(CORE), POWER9)
 ifneq ($(C_COMPILER), PGI)

Makefile.rule

@@ -3,7 +3,7 @@
 #
 
 # This library's version
-VERSION = 0.3.13
+VERSION = 0.3.14
 
 # If you set the suffix, the library name will be libopenblas_$(LIBNAMESUFFIX).a
 # and libopenblas_$(LIBNAMESUFFIX).so. Meanwhile, the soname in shared library