Merge pull request #5773 from OpenMathLib/develop

Merge from develop for 0.3.33 release

Author: martin-frbg

.cirrus.yml

@@ -89,14 +89,16 @@ task:
       type: text/plain
 
 macos_instance:
-  image: ghcr.io/cirruslabs/macos-sonoma-xcode:latest
+  image: ghcr.io/cirruslabs/macos-tahoe-xcode:latest
 task:
   name: AppleM1/LLVM armv7-androidndk xbuild
   compile_script:
   - brew install --cask android-ndk
   - export ANDROID_NDK_HOME="/opt/homebrew/share/android-ndk"
   - export CC=/opt/homebrew/share/android-ndk/toolchains/llvm/prebuilt/darwin-x86_64/bin/armv7a-linux-androideabi23-clang
-  - make TARGET=ARMV7 ARM_SOFTFP_ABI=1 NUM_THREADS=32 HOSTCC=clang NOFORTRAN=1 RANLIB="ls -l"
+  - export AR=/opt/homebrew/share/android-ndk/toolchains/llvm/prebuilt/darwin-x86_64/bin/llvm-ar
+  - export RANLIB=/opt/homebrew/share/android-ndk/toolchains/llvm/prebuilt/darwin-x86_64/bin/llvm-ranlib
+  - make TARGET=ARMV7 ARM_SOFTFP_ABI=1 NUM_THREADS=32 HOSTCC=clang NOFORTRAN=1
   always:
     config_artifacts:
       path: "*conf*"

CMakeLists.txt

@@ -309,8 +309,8 @@ if (USE_OPENMP)
 endif()
 
 # Fix "Argument list too long" for macOS - mostly seen with older OS versions on POWERPC or Intel CPUs 
-if(APPLE)
-  # Use response files
+if(APPLE AND "${CMAKE_GENERATOR}" MATCHES ".*Makefiles")
+  # Use response files to get around the ARG_MAX limit, unless using the Ninja generator
   set(CMAKE_C_USE_RESPONSE_FILE_FOR_OBJECTS 1)
   # Always build static library first
   if(BUILD_STATIC_LIBS)
@@ -333,7 +333,7 @@ if(APPLE)
   endif()
   if(NOT NOFORTRAN)
     set(CMAKE_Fortran_USE_RESPONSE_FILE_FOR_OBJECTS 1)
-    set(CMAKE_Fortran_CREATE_STATIC_LIBRARY ${CREATE_STATIC_LIBRARY_COMMAND})
+	set(CMAKE_Fortran_CREATE_STATIC_LIBRARY ${CREATE_STATIC_LIBRARY_COMMAND})
     if(BUILD_SHARED_LIBS)
       set(CMAKE_Fortran_CREATE_SHARED_LIBRARY
 	"sh -c 'echo \"\" | ${CMAKE_Fortran_COMPILER} -o dummy.o -c -x f95-cpp-input - '"

CONTRIBUTORS.md

@@ -275,3 +275,6 @@ In chronological order:
 
 * Fadi Arafeh <fadi.arafeh@arm.com>
   * [2026-03-05] Accelerate SVE128 SBGEMM/BGEMM
+
+* Nathan Sircombe <nathan.sircombe@arm.com>
+  * [2026-04-16] Add CPU ID for Neoverse V3

Changelog.txt

@@ -1,4 +1,54 @@
 OpenBLAS ChangeLog
+====================================================================
+Version 0.3.33
+23-Apr-2026
+
+general:
+ - fixed an incorrect cast in the SBGEMM test case that could lead to spurious test failures
+ - fixed an invalid memory access in the converted C version of the CBLAS tests
+ - made the BIGNUMA setting automatic when the number of cores exceeds 256
+ - Imported recent updates from Reference-LAPACK to realign with its upcoming 3.13.0 release:
+   - Implement ?LARF1F and ?ORM2R (Reference-LAPACK PRs 1019,1020,1196,1257)
+   - Change loop order in ?GETC2 to improve performance (Reference-LAPACK PR 1023)
+   - Change WORK array dimension in ?GELQS/?GEQRS (Reference-LAPACK PR 1094)
+   - Add NaN checks for input matrix A in ?GEEV (Reference-LAPACK PR 1136)
+   - Fix support for jobu/v in LAPACKE_?GESVDQ_WORK (Reference-LAPACK PRs 1146,1221)
+   - Fix display of version number in LAPACK testsuite (Reference-LAPACK PR 1149)
+   - Fix DGGES test seed to avoid bad matrix cases (Reference-LAPACK PR 1187)
+   - Fix truncation of large WORK array sizes in ZHE (Reference-LAPACK PR 1195)
+   - Fix overwriting of LDSWORK parameter in ?TRSYL3 (Reference-LAPACK PR 1206)
+   - Fix overwriting of error states in some EIG tests (Reference-LAPACK PR 1207) 
+   - Remove unused parameter in DORBDB3/ZUNBDB3 (Reference-LAPACK PR 1209)
+   - Re-enable testing of ?BB and ?GG driver functions (Reference-LAPACK PR 1211)
+   - Fix workspace size calculation in ?TGSEN (Reference-LAPACK PR 774)
+   - Fix typos in the EIG DMD tests and initialized the cutoff variable (PR 1212,1228)
+   - Optimized looping in ?LACPY/?LASCL/?LANTR with fat matrix and UPLO=L (PR 1251)
+
+arm64:
+ - worked around a serious miscompilation of the DDOT kernel by GCC15, affecting
+   most non-SVE targets, and SVE targets in the case of non-unit array stride)
+ - fixed an accuracy issue in the GEMV kernel for Neoverse V1 and other SVE targets
+ - fixed broken STRMM and SSYMM in DYNAMIC_ARCH builds when running on non-SME hardware
+ - added an optimized SHGEMM kernel for Neoverse N2
+ - fixed DYNAMIC_ARCH builds under Windows on Arm
+ - Added autodetection of Cortex A75/A76 in DYNAMIC_ARCH builds
+ - Added autodetection of Neoverse V3, currently supported through V2 kernels
+ - Re-added support for the "VORTEX" target in DYNAMIC_ARCH builds with DYNAMIC_LIST
+ - Fixed CMake-based builds that use the "Ninja" generator
+
+loongarch64:
+ - fixed a build failure due to missing support for the new half-precision float type
+ - fixed a long-standing bug in asserting 64bit capability in the c_check helper script
+
+x86_64:
+ - added a workaround for miscompilation of the AVX512 GEMM kernels by LLVM on Windows
+ - fixed a build failure in the LAED3 code when compiling with MinGW on Windows
+ - fixed CMake-based compilation with the NVIDIA HPC compiler
+ - Fixed CMake-based builds that use the "Ninja" generator
+
+wasm:
+ - added optimized kernels for STRSM and DTRSM
+
 ====================================================================
 Version 0.3.32
 23-Mar-2026

Makefile.rule

@@ -3,7 +3,7 @@
 #
 
 # This library's version
-VERSION = 0.3.32
+VERSION = 0.3.32.dev
 
 # If you set this prefix, the library name will be lib$(LIBNAMESUFFIX)openblas.a
 # and lib$(LIBNAMESUFFIX)openblas.so, with a matching soname in the shared library

README.md

@@ -176,7 +176,7 @@ Please read `GotoBLAS_01Readme.txt` for older CPU models already supported by th
 - **Cortex-A53**: same as ARMV8 (different cpu specifications)
 - **Cortex-A55**: same as ARMV8 (different cpu specifications)
 - **Cortex A57**: Optimized Level-3 and Level-2 functions
-- **Cortex A72**: same as A57 ( different cpu specifications)
+- **Cortex A72**: same as A57 (different cpu specifications)
 - **Cortex A73**: same as A57 (different cpu specifications)
 - **Cortex A76**: same as A57 (different cpu specifications)
 - **Falkor**: same as A57 (different cpu specifications)
@@ -189,6 +189,8 @@ Please read `GotoBLAS_01Readme.txt` for older CPU models already supported by th
 - **Neoverse V1**: (AWS Graviton3) optimized Level-3 BLAS
 - **Neoverse N2**: preliminary support
 - **Neoverse V2**: preliminary support
+- **Neoverse V3**: preliminary support
+- **Neoverse V3AE**: preliminary support
 - **Apple Vortex**: preliminary support based on ThunderX2/3
 - **Apple VortexM4**: preliminary support based on ThunderX2/3, SME kernels for SGEMM,SSYMM,STRMM,SSYRK,SSYR2K
 - **A64FX**:  preliminary support, optimized Level-3 BLAS

c_check

@@ -131,7 +131,7 @@ case "$architecture" in
     	defined=1
     	;;
     arm|arm64) defined=1 ;;
-    zarch|e2k|alpha|ia64|riscv64|loonarch64|wasm)
+    zarch|e2k|alpha|ia64|riscv64|loongarch64|wasm)
     	defined=1
     	BINARY=64
     	;;

cmake/arch.cmake

@@ -48,9 +48,11 @@ if (DYNAMIC_ARCH)
       if (${CMAKE_C_COMPILER_VERSION} VERSION_GREATER_EQUAL 11) # SVE ACLE supported in LLVM >= 11
         set(DYNAMIC_CORE ${DYNAMIC_CORE} NEOVERSEV1 NEOVERSEN2 ARMV8SVE A64FX)
       endif ()
+	  if (NOT ${CMAKE_SYSTEM_NAME} STREQUAL "Windows")
       if (${CMAKE_C_COMPILER_VERSION} VERSION_GREATER_EQUAL 19 OR (${CMAKE_C_COMPILER_ID} MATCHES AppleClang AND ${CMAKE_C_COMPILER_VERSION} VERSION_GREATER_EQUAL 17) ) # SME ACLE supported in LLVM >= 19 and AppleClang >= 17
         set(DYNAMIC_CORE ${DYNAMIC_CORE} ARMV9SME VORTEXM4)
       endif()
+	  endif()
     endif ()
     if (DYNAMIC_LIST)
 	  set(DYNAMIC_CORE ARMV8 ${DYNAMIC_LIST})

cmake/cc.cmake

@@ -113,6 +113,7 @@ if (${CMAKE_C_COMPILER_ID} STREQUAL "PGI")
 endif ()
 
 if (${CMAKE_C_COMPILER_ID} STREQUAL "NVHPC")
+  set (GCC_VERSION 100)
   if (POWER)
     set(CCOMMON_OPT "${CCOMMON_OPT} -tp pwr8")
   elseif (X86_64)

cmake/lapack.cmake

@@ -71,7 +71,7 @@ set(SLASRC
    slaqr0.f slaqr1.f slaqr2.f slaqr3.f slaqr4.f slaqr5.f
    slaqtr.f slar1v.f slar2v.f ilaslr.f ilaslc.f
    slarf.f  slarfb.f slarfb_gett.f slarfg.f slarfgp.f slarft.f slarfx.f slarfy.f slargv.f
-   slarrv.f slartv.f
+   slarf1f.f slarf1l.f slarrv.f slartv.f
    slarz.f  slarzb.f slarzt.f slasy2.f
    slasyf.f slasyf_rook.f slasyf_rk.f slasyf_aa.f
    slatbs.f slatdf.f slatps.f slatrd.f slatrs.f slatrz.f
@@ -178,6 +178,7 @@ set(CLASRC
    claqz0.f claqz1.f claqz2.f claqz3.f
    claqsp.f claqsy.f clar1v.f clar2v.f ilaclr.f ilaclc.f
    clarf.f  clarfb.f clarfb_gett.f clarfg.f clarfgp.f clarft.f
+   clarf1f.f clarf1l.f
    clarfx.f clarfy.f clargv.f clarnv.f clarrv.f clartg.f90 clartv.f
    clarz.f  clarzb.f clarzt.f clascl.f claset.f clasr.f  classq.f90
    clasyf.f clasyf_rook.f clasyf_rk.f clasyf_aa.f
@@ -262,7 +263,7 @@ set(DLASRC
    dlaqr0.f dlaqr1.f dlaqr2.f dlaqr3.f dlaqr4.f dlaqr5.f
    dlaqtr.f dlar1v.f dlar2v.f iladlr.f iladlc.f
    dlarf.f  dlarfb.f dlarfb_gett.f dlarfg.f dlarfgp.f dlarft.f dlarfx.f dlarfy.f
-   dlargv.f dlarrv.f dlartv.f
+   dlarf1f.f dlarf1l.f dlargv.f dlarrv.f dlartv.f
    dlarz.f  dlarzb.f dlarzt.f dlasy2.f
    dlasyf.f dlasyf_rook.f dlasyf_rk.f dlasyf_aa.f
    dlatbs.f dlatdf.f dlatps.f dlatrd.f dlatrs.f dlatrz.f
@@ -371,7 +372,7 @@ set(ZLASRC
    zlaqr0.f zlaqr1.f zlaqr2.f zlaqr3.f zlaqr4.f zlaqr5.f
    zlaqsp.f zlaqsy.f zlar1v.f zlar2v.f ilazlr.f ilazlc.f
    zlarcm.f zlarf.f  zlarfb.f zlarfb_gett.f
-   zlarfg.f zlarfgp.f zlarft.f
+   zlarfg.f zlarfgp.f zlarft.f zlarf1f.f zlarf1l.f
    zlarfx.f zlarfy.f zlargv.f zlarnv.f zlarrv.f zlartg.f90 zlartv.f
    zlarz.f  zlarzb.f zlarzt.f zlascl.f zlaset.f zlasr.f
    zlassq.f90 zlasyf.f zlasyf_rook.f zlasyf_rk.f zlasyf_aa.f
@@ -575,7 +576,7 @@ set(SLASRC
    slaqr0.c slaqr1.c slaqr2.c slaqr3.c slaqr4.c slaqr5.c
    slaqtr.c slar1v.c slar2v.c ilaslr.c ilaslc.c
    slarf.c  slarfb.c slarfb_gett.c slarfg.c slarfgp.c slarft.c slarfx.c slarfy.c slargv.c
-   slarrv.c slartv.c
+   slarf1f.c slarf1l.c slarrv.c slartv.c
    slarz.c  slarzb.c slarzt.c slasy2.c
    slasyf.c slasyf_rook.c slasyf_rk.c slasyf_aa.c
    slatbs.c slatdf.c slatps.c slatrd.c slatrs.c slatrz.c
@@ -681,6 +682,7 @@ set(CLASRC
    claqr0.c claqr1.c claqr2.c claqr3.c claqr4.c claqr5.c
    claqsp.c claqsy.c clar1v.c clar2v.c ilaclr.c ilaclc.c
    clarf.c  clarfb.c clarfb_gett.c clarfg.c clarfgp.c clarft.c
+   clarf1f.c clarf1l.c
    clarfx.c clarfy.c clargv.c clarnv.c clarrv.c clartg.c clartv.c
    clarz.c  clarzb.c clarzt.c clascl.c claset.c clasr.c  classq.c
    clasyf.c clasyf_rook.c clasyf_rk.c clasyf_aa.c
@@ -764,7 +766,7 @@ set(DLASRC
    dlaqr0.c dlaqr1.c dlaqr2.c dlaqr3.c dlaqr4.c dlaqr5.c
    dlaqtr.c dlar1v.c dlar2v.c iladlr.c iladlc.c
    dlarf.c  dlarfb.c dlarfb_gett.c dlarfg.c dlarfgp.c dlarft.c dlarfx.c dlarfy.c
-   dlargv.c dlarrv.c dlartv.c
+   dlarf1f.c dlarf1l.c dlargv.c dlarrv.c dlartv.c
    dlarz.c  dlarzb.c dlarzt.c dlasy2.c
    dlasyf.c dlasyf_rook.c dlasyf_rk.c dlasyf_aa.c
    dlatbs.c dlatdf.c dlatps.c dlatrd.c dlatrs.c dlatrz.c
@@ -871,7 +873,7 @@ set(ZLASRC
    zlaqhb.c zlaqhe.c zlaqhp.c zlaqp2.c zlaqp2rk.c zlaqp3rk.c zlaqps.c zlaqsb.c
    zlaqr0.c zlaqr1.c zlaqr2.c zlaqr3.c zlaqr4.c zlaqr5.c
    zlaqsp.c zlaqsy.c zlar1v.c zlar2v.c ilazlr.c ilazlc.c
-   zlarcm.c zlarf.c  zlarfb.c zlarfb_gett.c
+   zlarcm.c zlarf.c  zlarfb.c zlarfb_gett.c zlarf1f.c zlarf1l.c
    zlarfg.c zlarfgp.c zlarft.c
    zlarfx.c zlarfy.c zlargv.c zlarnv.c zlarrv.c zlartg.c zlartv.c
    zlarz.c  zlarzb.c zlarzt.c zlascl.c zlaset.c zlasr.c