- Update to version 0.3.14

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)
  RISC V:
  * Fixed compilation on RISCV (missing entry in getarch)
  POWER:
  * Fixed compilation for DYNAMIC_ARCH with clang and with older 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

OBS-URL: https://build.opensuse.org/package/show/science/openblas?expand=0&rev=120

0001-Require-gcc-11-for-builtin_cpu_is-power10.patch

@@ -1,26 +0,0 @@
-From 63fa3c3f8f869c585d8c5aef6f580a967b64405c Mon Sep 17 00:00:00 2001
-From: Martin Kroeker <martin@ruby.chemie.uni-freiburg.de>
-Date: Wed, 20 Jan 2021 15:41:04 +0100
-Subject: [PATCH 1/2] Require gcc 11 for builtin_cpu_is(power10)
-
-fixes #3074
----
- driver/others/dynamic_power.c | 2 +-
- 1 file changed, 1 insertion(+), 1 deletion(-)
-
-diff --git a/driver/others/dynamic_power.c b/driver/others/dynamic_power.c
-index f9feeb6e..18f16f83 100644
---- a/driver/others/dynamic_power.c
-+++ b/driver/others/dynamic_power.c
-@@ -202,7 +202,7 @@ static gotoblas_t *get_coretype(void) {
- 		return &gotoblas_POWER10;
- #endif
- 	/* Fall back to the POWER9 implementation if the toolchain is too old or the MMA feature is not set */
--#if (!defined __GNUC__) || ( __GNUC__ >= 6)
-+#if (!defined __GNUC__) || ( __GNUC__ >= 11)
- 	if (__builtin_cpu_is("power10"))
- 		return &gotoblas_POWER9;
- #endif	
--- 
-2.26.2
-

0002-patch-to-support-power10-in-builtin_cpu_is-was-backp.patch

@@ -1,26 +0,0 @@
-From b94dab5250469d4d30d1a21bf0e0b78eea3cf286 Mon Sep 17 00:00:00 2001
-From: Martin Kroeker <martin@ruby.chemie.uni-freiburg.de>
-Date: Wed, 20 Jan 2021 21:34:36 +0100
-Subject: [PATCH 2/2] patch to support power10 in builtin_cpu_is was backported
- to gcc 10.2, so allow that as wel
-
----
- driver/others/dynamic_power.c | 2 +-
- 1 file changed, 1 insertion(+), 1 deletion(-)
-
-diff --git a/driver/others/dynamic_power.c b/driver/others/dynamic_power.c
-index 18f16f83..b8e5840a 100644
---- a/driver/others/dynamic_power.c
-+++ b/driver/others/dynamic_power.c
-@@ -202,7 +202,7 @@ static gotoblas_t *get_coretype(void) {
- 		return &gotoblas_POWER10;
- #endif
- 	/* Fall back to the POWER9 implementation if the toolchain is too old or the MMA feature is not set */
--#if (!defined __GNUC__) || ( __GNUC__ >= 11)
-+#if (!defined __GNUC__) || ( __GNUC__ >= 11) || (__GNUC__ == 10 && __GNUC_MINOR__ >= 2)
- 	if (__builtin_cpu_is("power10"))
- 		return &gotoblas_POWER9;
- #endif	
--- 
-2.26.2
-

OpenBLAS-0.3.13.tar.gz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:79197543b17cc314b7e43f7a33148c308b0807cd6381ee77f77e15acf3e6459e
-size 12437434

OpenBLAS-0.3.14.tar.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d381935d26f9cae8e4bbd7d7f278435adf8e3a90920edf284bb9ad789ee9ad60
+size 12460465

openblas-noexecstack.patch

@@ -1,7 +1,7 @@
-Index: OpenBLAS-0.3.11/exports/Makefile
+Index: OpenBLAS-0.3.14/exports/Makefile
 ===================================================================
---- OpenBLAS-0.3.11.orig/exports/Makefile
+--- OpenBLAS-0.3.14.orig/exports/Makefile
-+++ OpenBLAS-0.3.11/exports/Makefile
++++ OpenBLAS-0.3.14/exports/Makefile
 @@ -179,6 +179,7 @@ else ifeq ($(F_COMPILER), FLANG)
  else
  ifneq ($(C_COMPILER), LSB)

openblas-ppc64be_up2_p8.patch

@@ -11,36 +11,13 @@ because:
 
 Signed-off-by: Michel Normand <normand@linux.vnet.ibm.com>
 ---
- Makefile.system               |    4 ++++
  driver/others/dynamic_power.c |   11 +++++++++++
- 2 files changed, 15 insertions(+)
+ 1 files changed, 11 insertions(+)
 
-Index: OpenBLAS-0.3.13/Makefile.system
+Index: OpenBLAS-0.3.14/driver/others/dynamic_power.c
 ===================================================================
---- OpenBLAS-0.3.13.orig/Makefile.system
+--- OpenBLAS-0.3.14.orig/driver/others/dynamic_power.c
-+++ OpenBLAS-0.3.13/Makefile.system
++++ OpenBLAS-0.3.14/driver/others/dynamic_power.c
-@@ -665,6 +665,9 @@ endif # ARCH zarch
- ifeq ($(ARCH), power)
- DYNAMIC_CORE = POWER6
- DYNAMIC_CORE += POWER8
-+ifeq ($(__BYTE_ORDER__),__ORDER_BIG_ENDIAN__)
-+$(info, OpenBLAS: for big endian limit to POWER8 kernels.)
-+else
- ifneq ($(C_COMPILER), GCC)
- DYNAMIC_CORE += POWER9
- DYNAMIC_CORE += POWER10
-@@ -690,6 +693,7 @@ $(info, OpenBLAS: Your gcc version is to
- endif
- endif
- endif
-+endif
- 
- # If DYNAMIC_CORE is not set, DYNAMIC_ARCH cannot do anything, so force it to empty
- ifndef DYNAMIC_CORE
-Index: OpenBLAS-0.3.13/driver/others/dynamic_power.c
-===================================================================
---- OpenBLAS-0.3.13.orig/driver/others/dynamic_power.c
-+++ OpenBLAS-0.3.13/driver/others/dynamic_power.c
 @@ -3,6 +3,7 @@
  
  extern gotoblas_t gotoblas_POWER6;
@@ -57,9 +34,9 @@ Index: OpenBLAS-0.3.13/driver/others/dynamic_power.c
  
  extern void openblas_warning(int verbose, const char *msg);
  
-@@ -29,12 +31,14 @@ static char *corename[] = {
+@@ -31,12 +33,14 @@ char *gotoblas_corename(void) {
- char *gotoblas_corename(void) {
  	if (gotoblas == &gotoblas_POWER6)	return corename[1];
+ #endif
  	if (gotoblas == &gotoblas_POWER8)	return corename[2];
 +#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
  #if (!defined __GNUC__) || ( __GNUC__ >= 6)
@@ -72,8 +49,8 @@ Index: OpenBLAS-0.3.13/driver/others/dynamic_power.c
  	return corename[0];
  }
  
-@@ -44,6 +48,10 @@ static gotoblas_t *get_coretype(void) {
+@@ -200,6 +204,10 @@ static gotoblas_t *get_coretype(void) {
- 		return &gotoblas_POWER6;
+ #endif
  	if (__builtin_cpu_is("power8"))
  		return &gotoblas_POWER8;
 +	/* Fall back to the POWER8 implementation for big endian */
@@ -83,7 +60,7 @@ Index: OpenBLAS-0.3.13/driver/others/dynamic_power.c
  #if (!defined __GNUC__) || ( __GNUC__ >= 6)
  	if (__builtin_cpu_is("power9"))
  		return &gotoblas_POWER9;
-@@ -57,6 +65,7 @@ static gotoblas_t *get_coretype(void) {
+@@ -213,6 +221,7 @@ static gotoblas_t *get_coretype(void) {
  	if (__builtin_cpu_is("power10"))
  		return &gotoblas_POWER9;
  #endif	
@@ -91,9 +68,9 @@ Index: OpenBLAS-0.3.13/driver/others/dynamic_power.c
  	return NULL;
  }
  
-@@ -79,12 +88,14 @@ static gotoblas_t *force_coretype(char *
+@@ -237,12 +246,14 @@ static gotoblas_t *force_coretype(char *
- 	{
  	case  1: return (&gotoblas_POWER6);
+ #endif
  	case  2: return (&gotoblas_POWER8);
 +#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
  #if (!defined __GNUC__) || ( __GNUC__ >= 6)

openblas-s390.patch

@@ -1,8 +1,8 @@
-Index: OpenBLAS-0.3.11/c_check
+Index: OpenBLAS-0.3.14/c_check
 ===================================================================
---- OpenBLAS-0.3.11.orig/c_check
+--- OpenBLAS-0.3.14.orig/c_check
-+++ OpenBLAS-0.3.11/c_check
++++ OpenBLAS-0.3.14/c_check
-@@ -11,7 +11,7 @@ $hostarch = "x86_64" if ($hostarch eq "a
+@@ -12,7 +12,7 @@ $hostarch = "x86_64" if ($hostarch eq "a
  $hostarch = "arm" if ($hostarch ne "arm64" && $hostarch =~ /^arm.*/);
  $hostarch = "arm64" if ($hostarch eq "aarch64");
  $hostarch = "power" if ($hostarch =~ /^(powerpc|ppc).*/);
@@ -11,10 +11,10 @@ Index: OpenBLAS-0.3.11/c_check
  
  #$tmpf = new File::Temp( UNLINK => 1 );
  $binary = $ENV{"BINARY"};
-Index: OpenBLAS-0.3.11/ctest.c
+Index: OpenBLAS-0.3.14/ctest.c
 ===================================================================
---- OpenBLAS-0.3.11.orig/ctest.c
+--- OpenBLAS-0.3.14.orig/ctest.c
-+++ OpenBLAS-0.3.11/ctest.c
++++ OpenBLAS-0.3.14/ctest.c
 @@ -117,7 +117,7 @@ ARCH_X86_64
  ARCH_POWER
  #endif
@@ -24,11 +24,11 @@ Index: OpenBLAS-0.3.11/ctest.c
  ARCH_ZARCH
  #endif
  
-Index: OpenBLAS-0.3.11/getarch.c
+Index: OpenBLAS-0.3.14/getarch.c
 ===================================================================
---- OpenBLAS-0.3.11.orig/getarch.c
+--- OpenBLAS-0.3.14.orig/getarch.c
-+++ OpenBLAS-0.3.11/getarch.c
++++ OpenBLAS-0.3.14/getarch.c
-@@ -1266,7 +1266,7 @@ USE OF THIS SOFTWARE, EVEN IF ADVISED OF
+@@ -1333,7 +1333,7 @@ USE OF THIS SOFTWARE, EVEN IF ADVISED OF
  #define OPENBLAS_SUPPORTED
  #endif
  

openblas.changes

@@ -1,3 +1,55 @@
+-------------------------------------------------------------------
+Thu Mar 18 08:05:58 UTC 2021 - Ismail Dönmez <idonmez@suse.com>
+
+- Update to version 0.3.14
+  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)
+
+  RISC V:
+  * Fixed compilation on RISCV (missing entry in getarch)
+
+  POWER:
+  * Fixed compilation for DYNAMIC_ARCH with clang and with older 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:
+  * Support compilation for embedded Cortex M4 targets via a new option EMBEDDED
+
+  ARM64:
+  * 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
+
+- Remove 0001-Require-gcc-11-for-builtin_cpu_is-power10.patch
+  0002-patch-to-support-power10-in-builtin_cpu_is-was-backp.patch
+  Upstream fixed in a different way.
+
 -------------------------------------------------------------------
 Thu Feb  4 11:49:11 UTC 2021 - Michel Normand <normand@linux.vnet.ibm.com>
 

openblas.spec

@@ -18,8 +18,8 @@
 
 %global flavor @BUILD_FLAVOR@%{nil}
 
-%define _vers 0_3_13
+%define _vers 0_3_14
-%define vers 0.3.13
+%define vers 0.3.14
 %define pname openblas
 
 %bcond_with ringdisabled
@@ -167,9 +167,6 @@ URL:            http://www.openblas.net
 Source0:        https://github.com/xianyi/OpenBLAS/archive/v%{version}.tar.gz#/OpenBLAS-%{version}.tar.gz
 Source1:        README.SUSE
 Source2:        README.HPC.SUSE
-# Temporarily - delete with next version update
-Patch1:         0001-Require-gcc-11-for-builtin_cpu_is-power10.patch
-Patch2:         0002-patch-to-support-power10-in-builtin_cpu_is-was-backp.patch
 # PATCH-FIX-UPSTREAM openblas-noexecstack.patch
 Patch101:       openblas-noexecstack.patch
 # PATCH port
@@ -182,7 +179,7 @@ BuildRoot:      %{_tmppath}/%{name}-%{version}-build
 BuildRequires:  gcc-fortran
 BuildRequires:  update-alternatives
 Requires(post): update-alternatives
-Requires(preun): update-alternatives
+Requires(preun):update-alternatives
 %else
 BuildRequires:  %{compiler_family}%{?c_f_ver}-compilers-hpc-macros-devel
 BuildRequires:  lua-lmod
@@ -198,7 +195,7 @@ Summary:        An optimized BLAS library based on GotoBLAS2, %{flavor} version
 Group:          System/Libraries
 %if %{without hpc}
 Requires(post): update-alternatives
-Requires(preun): update-alternatives
+Requires(preun):update-alternatives
  %if "%flavor" == "serial"
 Obsoletes:      lib%{pname}%{so_v} < %{version}
 Provides:       lib%{pname}%{so_v} = %{version}
@@ -284,8 +281,6 @@ This package contains headers for OpenBLAS.
 %prep
 
 %setup -q -n OpenBLAS-%{version}
-%patch1 -p1
-%patch2 -p1
 %patch101 -p1
 %patch102 -p1
 %patch103 -p1