{compiler,lang}[GCCcore/15.2.0,foss/2026.1] numba v0.65.1, LLVM v20.1.8 /w llvmlite

[pavelToman]

(created using eb --new-pr)
extracted from PR #26378

[github-actions]

Updated software LLVM-20.1.8-GCCcore-15.2.0-llvmlite.eb

Diff against LLVM-21.1.8-GCCcore-15.2.0.eb

easybuild/easyconfigs/l/LLVM/LLVM-21.1.8-GCCcore-15.2.0.eb

diff --git a/easybuild/easyconfigs/l/LLVM/LLVM-21.1.8-GCCcore-15.2.0.eb b/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-15.2.0-llvmlite.eb
index 25690f5f75..4d1c370d0d 100644
--- a/easybuild/easyconfigs/l/LLVM/LLVM-21.1.8-GCCcore-15.2.0.eb
+++ b/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-15.2.0-llvmlite.eb
@@ -1,6 +1,6 @@
 name = 'LLVM'
-version = '21.1.8'
-
+version = '20.1.8'
+versionsuffix = '-llvmlite'
 
 homepage = "https://llvm.org/"
 description = """
@@ -14,30 +14,15 @@ to use LLVM as an optimizer and code generator.
 """
 
 toolchain = {'name': 'GCCcore', 'version': '15.2.0'}
-toolchainopts = {
-    'pic': True
-}
+toolchainopts = {'pic': True}
 
 source_urls = ['https://github.com/llvm/llvm-project/releases/download/llvmorg-%(version)s/']
-sources = [
-    'llvm-project-%(version)s.src.tar.xz',
-]
-patches = [
-    'LLVM-18.1.8_envintest.patch',
-    'LLVM-19.1.7_libomptarget_tests.patch',
-    'LLVM-19.1.7_clang_rpathwrap_test.patch',
-    'LLVM-21.1.x_fix-ompt-end-critical-race.patch',
-    'LLVM-21.1.x_fix-sema-overload-deduction.patch',
-]
+sources = ['llvm-project-%(version)s.src.tar.xz']
+patches = ['LLVM-20.1.x_improved-CUDA-13-support.patch']
 checksums = [
-    {'llvm-project-21.1.8.src.tar.xz': '4633a23617fa31a3ea51242586ea7fb1da7140e426bd62fc164261fe036aa142'},
-    {'LLVM-18.1.8_envintest.patch': '8e25dfab8a29a860717b4bd2d8cdd0e795433766d7fffbda32d06a2bde47058d'},
-    {'LLVM-19.1.7_libomptarget_tests.patch': '79a67c118d034cfb74e255696369150c73432d2b422f4834efacb26f7904edbf'},
-    {'LLVM-19.1.7_clang_rpathwrap_test.patch': '5ee6a87ec8ff1c8b736ffe0513aa2098bd2b83a1ffc647a1ad2cf966f567e8a1'},
-    {'LLVM-21.1.x_fix-ompt-end-critical-race.patch':
-     '9dc675f0a53a2308a22c36892eddeeab9c3c736219c04192fc5bb1348f110838'},
-    {'LLVM-21.1.x_fix-sema-overload-deduction.patch':
-     '265a2b5ce0cc439ab98d5a9efb60153b722c91df9db7d18c7b49a159156cb41b'},
+    {'llvm-project-20.1.8.src.tar.xz': '6898f963c8e938981e6c4a302e83ec5beb4630147c7311183cf61069af16333d'},
+    {'LLVM-20.1.x_improved-CUDA-13-support.patch':
+     '78762218c8f34b25533f909533076f195136232b7fca1672af602182dfcdc5bf'},
 ]
 
 builddependencies = [
@@ -54,44 +39,33 @@ dependencies = [
     ('libxml2', '2.15.1'),
     ('ncurses', '6.6'),
     ('zlib', '2.3.2'),
-    ('Z3', '4.15.4'),
     ('zstd', '1.5.7'),
 ]
 
+build_shared_libs = False
 
-build_shared_libs = True
+minimal = True
 
-bootstrap = True
+bootstrap = False
 full_llvm = False
-build_clang_extras = True
-build_runtimes = True
-build_lld = True
-build_lldb = True
-build_bolt = True
-build_openmp = True
-build_openmp_tools = True
-build_openmp_offload = True
-use_polly = True
+build_clang_extras = False
+build_runtimes = False
+build_lld = False
+build_lldb = False
+build_bolt = False
+build_openmp = False
+build_openmp_tools = False
+build_openmp_offload = False
+use_polly = False
 
-python_bindings = True
+python_bindings = False
 
 build_targets = ['all']
 # disable_werror = True
 
 skip_all_tests = False
-skip_sanitizer_tests = True
-test_suite_max_failed = 10
+skip_sanitizer_tests = False
+test_suite_max_failed = 0
 test_suite_timeout_single = 5 * 60
-test_suite_ignore_patterns = [
-    "X86/register-fragments-bolt-symbols.s",
-    "modularize/ProblemsCoverage.modularize",
-    "Driver/atomic.f90",
-    "Driver/gcc-toolchain-install-dir.f90",
-    "api_tests/test_ompd_get_icv_from_scope.c",
-    "std/time/time.syn/formatter",
-    # Those 2 might fail with "tzdb: the path '/etc/localtime' is not a symlink"
-    "std/time/time.zone/time.zone.db/time.zone.db.tzdb",
-    "std/time/time.zone/time.zone.db/time.zone.db.access",
-]
 
 moduleclass = 'compiler'

Diff against LLVM-20.1.8-GCCcore-14.3.0.eb

easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-14.3.0.eb

diff --git a/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-14.3.0.eb b/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-15.2.0-llvmlite.eb
index 62998620ce..4d1c370d0d 100644
--- a/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-14.3.0.eb
+++ b/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-15.2.0-llvmlite.eb
@@ -1,6 +1,6 @@
 name = 'LLVM'
 version = '20.1.8'
-
+versionsuffix = '-llvmlite'
 
 homepage = "https://llvm.org/"
 description = """
@@ -13,135 +13,59 @@ particularly easy to invent your own language (or port an existing compiler)
 to use LLVM as an optimizer and code generator.
 """
 
-toolchain = {'name': 'GCCcore', 'version': '14.3.0'}
-toolchainopts = {
-    # 'cstd': 'gnu++11',
-    'pic': True
-}
+toolchain = {'name': 'GCCcore', 'version': '15.2.0'}
+toolchainopts = {'pic': True}
 
 source_urls = ['https://github.com/llvm/llvm-project/releases/download/llvmorg-%(version)s/']
-sources = [
-    'llvm-project-%(version)s.src.tar.xz',
-]
-patches = [
-    'LLVM-18.1.8_envintest.patch',
-    'LLVM-19.1.7_libomptarget_tests.patch',
-    'LLVM-19.1.7_clang_rpathwrap_test.patch',
-    'LLVM-20.1.x_improved-CUDA-13-support.patch',
-    'LLVM-20.1.5-fix_bindc_commonblocks_fortran.patch',  # This patch is included upstream from >=21.1.x
-    'LLVM-20.1.x_always-link-compiler-rt-to-flang.patch',  # This patch is included upstream from >= 21.1.x
-]
+sources = ['llvm-project-%(version)s.src.tar.xz']
+patches = ['LLVM-20.1.x_improved-CUDA-13-support.patch']
 checksums = [
     {'llvm-project-20.1.8.src.tar.xz': '6898f963c8e938981e6c4a302e83ec5beb4630147c7311183cf61069af16333d'},
-    {'LLVM-18.1.8_envintest.patch': '8e25dfab8a29a860717b4bd2d8cdd0e795433766d7fffbda32d06a2bde47058d'},
-    {'LLVM-19.1.7_libomptarget_tests.patch': '79a67c118d034cfb74e255696369150c73432d2b422f4834efacb26f7904edbf'},
-    {'LLVM-19.1.7_clang_rpathwrap_test.patch': '5ee6a87ec8ff1c8b736ffe0513aa2098bd2b83a1ffc647a1ad2cf966f567e8a1'},
-    {'LLVM-20.1.x_improved-CUDA-13-support.patch': '78762218c8f34b25533f909533076f195136232b7fca1672af602182dfcdc5bf'},
-    {'LLVM-20.1.5-fix_bindc_commonblocks_fortran.patch':
-     '02b1123edfbb0f5c805fb7b7ad9caaf376489a80d723f90bf9366b0c9949dea1'},
-    {'LLVM-20.1.x_always-link-compiler-rt-to-flang.patch':
-     '7ee0152efafdfebfeff01487e85d2aecd38d3c4e892e5ab91a21db02cdc6a3c6'},
+    {'LLVM-20.1.x_improved-CUDA-13-support.patch':
+     '78762218c8f34b25533f909533076f195136232b7fca1672af602182dfcdc5bf'},
 ]
 
-# riscv_MLIR.patch: Disable problematic MLIR Runner tests on RISC-V.
-# Failures on RISC-V are due to missing soft-float helper symbols (like __addsf3),
-# which is a known issue when running LLVM's JIT on platforms like RISC-V.
-
-# riscv_execution_engine.patch: Skip the test on RISC-V, which is appropriate since JIT execution via MLIR's
-# ExecutionEngine is not reliably supported on this architecture in this LLVM version.
-
-# riscv_xray_check.patch: RISC-V has looser execution ordering; relax CHECK-NEXT to CHECK
-
-# riscv_fuzzer.patch: Mark as UNSUPPORTED for RISC-V tests that use libFuzzer + -fsanitize=address,fuzzer
-
-# riscv_ompt.patch: Mark as UNSUPPORTED for RISC-V some ompt tests that are known to be fragile
-# see (https://github.com/llvm/llvm-project/issues/35661#issuecomment-980993177)
-
-if ARCH == 'riscv64':
-    patches += [
-        'LLVM-20.1.5_riscv_optimization_remark.patch',
-        'LLVM-20.1.5_riscv_MLIR.patch',
-        'LLVM-20.1.5_riscv_execution_engine.patch',
-        'LLVM-20.1.5_riscv_xray_check.patch',
-        'LLVM-20.1.5_riscv_fuzzer.patch',
-        'LLVM-20.1.5_riscv_ompt.patch',
-    ]
-    checksums += [
-        {'LLVM-20.1.5_riscv_optimization_remark.patch':
-         'e414a15e0fc48049a6ec75835cd032c8db6fb5c6d2825f0c604969c192aa60e0'},
-        {'LLVM-20.1.5_riscv_MLIR.patch':
-         '2348ee5db243f94d7a8877aee1a4332585e3bd58976af76c4bed01972d7793ca'},
-        {'LLVM-20.1.5_riscv_execution_engine.patch':
-         '7d795920e98c2c65df30ed01cc7e9251d69ae25e2674c251bc283ed44b53d02f'},
-        {'LLVM-20.1.5_riscv_xray_check.patch':
-         '5ca1e5730a3ee7cc1914959d64330f213a58f9818d91dbba346db63c77b47606'},
-        {'LLVM-20.1.5_riscv_fuzzer.patch':
-         '0d28e1acc96ab10daa2f696db01163c81f8a027a6ab8ccd5a770c6b7e91a4c34'},
-        {'LLVM-20.1.5_riscv_ompt.patch':
-         '9ed82093831777cee08629cb68b7e47d2cace3e0b950381b002f9bfd01bae1f6'},
-    ]
-
 builddependencies = [
-    ('binutils', '2.44'),
-    ('Python', '3.13.5'),
-    ('CMake', '4.0.3'),
-    ('psutil', '7.0.0'),  # Needed to enable test timeout in lit
+    ('binutils', '2.45'),
+    ('Python', '3.14.2'),
+    ('CMake', '4.2.1'),
+    ('psutil', '7.2.1'),  # Needed to enable test timeout in lit
     ('lit', '18.1.8'),
-    ('git', '2.50.1'),
+    ('git', '2.52.0'),
 ]
 
 dependencies = [
-    ('libffi', '3.5.1'),
-    ('libxml2', '2.14.3'),
-    ('ncurses', '6.5'),
-    ('zlib', '1.3.1'),
-    ('Z3', '4.15.1'),
+    ('libffi', '3.5.2'),
+    ('libxml2', '2.15.1'),
+    ('ncurses', '6.6'),
+    ('zlib', '2.3.2'),
     ('zstd', '1.5.7'),
 ]
 
-build_shared_libs = True
-
-bootstrap = True
-full_llvm = False
-build_clang_extras = True
-build_runtimes = True
-build_lld = True
-build_lldb = True
-build_bolt = True
-build_openmp = True
-build_openmp_tools = True
-build_openmp_offload = True
+build_shared_libs = False
 
-# openmp offload is not supported for riscv64 (https://clang.llvm.org/docs/OffloadingDesign.html)
-# There is a Bug in (at least) testing phase of BOLT for RISC-V. See https://github.com/llvm/llvm-project/issues/134245
-if ARCH == 'riscv64':
-    build_bolt = False
-    build_openmp_offload = False
+minimal = True
 
-use_polly = True
-
-python_bindings = True
+bootstrap = False
+full_llvm = False
+build_clang_extras = False
+build_runtimes = False
+build_lld = False
+build_lldb = False
+build_bolt = False
+build_openmp = False
+build_openmp_tools = False
+build_openmp_offload = False
+use_polly = False
+
+python_bindings = False
 
 build_targets = ['all']
 # disable_werror = True
 
 skip_all_tests = False
-skip_sanitizer_tests = True
-test_suite_max_failed = 10
+skip_sanitizer_tests = False
+test_suite_max_failed = 0
 test_suite_timeout_single = 5 * 60
-test_suite_ignore_patterns = [
-    "X86/register-fragments-bolt-symbols.s",
-    "modularize/ProblemsCoverage.modularize",
-    "Driver/atomic.f90",
-    "Driver/gcc-toolchain-install-dir.f90",
-    "api_tests/test_ompd_get_icv_from_scope.c",
-    "std/time/time.syn/formatter",
-    # Those 2 might fail with "tzdb: the path '/etc/localtime' is not a symlink"
-    "std/time/time.zone/time.zone.db/time.zone.db.tzdb",
-    "std/time/time.zone/time.zone.db/time.zone.db.access",
-    # libFuzzer tests massively fail in RISC-V due to slightly different
-    # behaviour of signal handling and stack unwinding
-    "libFuzzer-riscv64",
-]
 
 moduleclass = 'compiler'

Diff against LLVM-20.1.7-GCCcore-14.2.0.eb

easybuild/easyconfigs/l/LLVM/LLVM-20.1.7-GCCcore-14.2.0.eb

diff --git a/easybuild/easyconfigs/l/LLVM/LLVM-20.1.7-GCCcore-14.2.0.eb b/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-15.2.0-llvmlite.eb
index ba15a15dd1..4d1c370d0d 100644
--- a/easybuild/easyconfigs/l/LLVM/LLVM-20.1.7-GCCcore-14.2.0.eb
+++ b/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-15.2.0-llvmlite.eb
@@ -1,89 +1,71 @@
 name = 'LLVM'
-version = '20.1.7'
-
+version = '20.1.8'
+versionsuffix = '-llvmlite'
 
 homepage = "https://llvm.org/"
-description = """The LLVM Core libraries provide a modern source- and target-independent
- optimizer, along with code generation support for many popular CPUs
- (as well as some less common ones!) These libraries are built around a well
- specified code representation known as the LLVM intermediate representation
- ("LLVM IR"). The LLVM Core libraries are well documented, and it is
- particularly easy to invent your own language (or port an existing compiler)
- to use LLVM as an optimizer and code generator."""
+description = """
+The LLVM Core libraries provide a modern source- and target-independent
+optimizer, along with code generation support for many popular CPUs
+(as well as some less common ones!) These libraries are built around a well
+specified code representation known as the LLVM intermediate representation
+("LLVM IR"). The LLVM Core libraries are well documented, and it is
+particularly easy to invent your own language (or port an existing compiler)
+to use LLVM as an optimizer and code generator.
+"""
 
-toolchain = {'name': 'GCCcore', 'version': '14.2.0'}
-toolchainopts = {
-    # 'cstd': 'gnu++11',
-    'pic': True
-}
+toolchain = {'name': 'GCCcore', 'version': '15.2.0'}
+toolchainopts = {'pic': True}
 
 source_urls = ['https://github.com/llvm/llvm-project/releases/download/llvmorg-%(version)s/']
-sources = [
-    'llvm-project-%(version)s.src.tar.xz',
-]
-patches = [
-    'LLVM-18.1.8_envintest.patch',
-    'LLVM-19.1.7_libomptarget_tests.patch',
-    'LLVM-19.1.7_clang_rpathwrap_test.patch',
-    'LLVM-20.1.x_improved-CUDA-13-support.patch',
-    'LLVM-20.1.5-fix_bindc_commonblocks_fortran.patch',  # This path is included upstream from >=21.1.x
-]
+sources = ['llvm-project-%(version)s.src.tar.xz']
+patches = ['LLVM-20.1.x_improved-CUDA-13-support.patch']
 checksums = [
-    {'llvm-project-20.1.7.src.tar.xz': 'cd8fd55d97ad3e360b1d5aaf98388d1f70dfffb7df36beee478be3b839ff9008'},
-    {'LLVM-18.1.8_envintest.patch': '8e25dfab8a29a860717b4bd2d8cdd0e795433766d7fffbda32d06a2bde47058d'},
-    {'LLVM-19.1.7_libomptarget_tests.patch': '79a67c118d034cfb74e255696369150c73432d2b422f4834efacb26f7904edbf'},
-    {'LLVM-19.1.7_clang_rpathwrap_test.patch': '5ee6a87ec8ff1c8b736ffe0513aa2098bd2b83a1ffc647a1ad2cf966f567e8a1'},
-    {'LLVM-20.1.x_improved-CUDA-13-support.patch': '78762218c8f34b25533f909533076f195136232b7fca1672af602182dfcdc5bf'},
-    {'LLVM-20.1.5-fix_bindc_commonblocks_fortran.patch':
-     '02b1123edfbb0f5c805fb7b7ad9caaf376489a80d723f90bf9366b0c9949dea1'},
+    {'llvm-project-20.1.8.src.tar.xz': '6898f963c8e938981e6c4a302e83ec5beb4630147c7311183cf61069af16333d'},
+    {'LLVM-20.1.x_improved-CUDA-13-support.patch':
+     '78762218c8f34b25533f909533076f195136232b7fca1672af602182dfcdc5bf'},
 ]
 
 builddependencies = [
-    ('binutils', '2.42'),
-    ('Python', '3.13.1'),
-    ('CMake', '3.31.3'),
-    ('psutil', '7.0.0'),  # Needed to enable test timeout in lit
+    ('binutils', '2.45'),
+    ('Python', '3.14.2'),
+    ('CMake', '4.2.1'),
+    ('psutil', '7.2.1'),  # Needed to enable test timeout in lit
     ('lit', '18.1.8'),
-    ('git', '2.49.0'),
+    ('git', '2.52.0'),
 ]
 
 dependencies = [
-    ('libxml2', '2.13.4'),
-    ('ncurses', '6.5'),
-    ('zlib', '1.3.1'),
-    ('Z3', '4.13.4'),
-    ('zstd', '1.5.6'),
+    ('libffi', '3.5.2'),
+    ('libxml2', '2.15.1'),
+    ('ncurses', '6.6'),
+    ('zlib', '2.3.2'),
+    ('zstd', '1.5.7'),
 ]
 
-build_shared_libs = True
+build_shared_libs = False
 
-bootstrap = True
+minimal = True
+
+bootstrap = False
 full_llvm = False
-build_clang_extras = True
-build_runtimes = True
-build_lld = True
-build_lldb = True
-build_bolt = True
-build_openmp = True
-build_openmp_tools = True
-build_openmp_offload = True
-use_polly = True
+build_clang_extras = False
+build_runtimes = False
+build_lld = False
+build_lldb = False
+build_bolt = False
+build_openmp = False
+build_openmp_tools = False
+build_openmp_offload = False
+use_polly = False
 
-python_bindings = True
+python_bindings = False
 
 build_targets = ['all']
 # disable_werror = True
 
 skip_all_tests = False
-skip_sanitizer_tests = True
-test_suite_max_failed = 10
+skip_sanitizer_tests = False
+test_suite_max_failed = 0
 test_suite_timeout_single = 5 * 60
-test_suite_ignore_patterns = [
-    "X86/register-fragments-bolt-symbols.s",
-    "modularize/ProblemsCoverage.modularize",
-    "Driver/atomic.f90",
-    "Driver/gcc-toolchain-install-dir.f90",
-    "api_tests/test_ompd_get_icv_from_scope.c",
-]
 
 moduleclass = 'compiler'

Updated software numba-0.65.1-foss-2026.1.eb

Diff against numba-0.62.0-foss-2025a.eb

easybuild/easyconfigs/n/numba/numba-0.62.0-foss-2025a.eb

diff --git a/easybuild/easyconfigs/n/numba/numba-0.62.0-foss-2025a.eb b/easybuild/easyconfigs/n/numba/numba-0.65.1-foss-2026.1.eb
index bb31bfe1c6..3e380ba4b2 100644
--- a/easybuild/easyconfigs/n/numba/numba-0.62.0-foss-2025a.eb
+++ b/easybuild/easyconfigs/n/numba/numba-0.65.1-foss-2026.1.eb
@@ -1,43 +1,50 @@
 easyblock = 'PythonBundle'
 
 name = 'numba'
-version = '0.62.0'
+version = '0.65.1'
 
 homepage = 'https://numba.pydata.org/'
 description = """Numba is an Open Source NumPy-aware optimizing compiler for
 Python sponsored by Continuum Analytics, Inc. It uses the remarkable LLVM
 compiler infrastructure to compile Python syntax to machine code."""
 
-toolchain = {'name': 'foss', 'version': '2025a'}
+toolchain = {'name': 'foss', 'version': '2026.1'}
 toolchainopts = {'pic': True}
 
 builddependencies = [
-    ('CMake', '3.31.3'),
-    ('LLVM', '20.1.7'),
+    ('CMake', '4.2.1'),
+    ('LLVM', '20.1.8', '-llvmlite'),  # llvmlite 0.47.0 supports only LLVM 20
 ]
-
 dependencies = [
-    ('Python', '3.13.1'),
-    ('SciPy-bundle', '2025.06'),
-    ('Z3', '4.13.4'),
-    ('zstd', '1.5.6'),
+    ('Python', '3.14.2'),
+    ('Python-bundle-PyPI', '2026.04'),
+    ('SciPy-bundle', '2026.05'),
+    ('Z3', '4.15.4'),
+    ('zstd', '1.5.7'),
 ]
 
 local_llvmlite_preinstallopts = "export LLVM_CONFIG=${EBROOTLLVM}/bin/llvm-config && "
 local_llvmlite_preinstallopts += "sed -i 's/llvm_config, \"--system-libs/llvm_config, \"--link-static --system-libs/' "
 local_llvmlite_preinstallopts += "ffi/build.py && "
 
+# Limit numba.runtests worker processes to avoid multiprocessing/forkserver hangs
+local_test_workers = '$(( %(parallel)s > 8 ? 4 : (%(parallel)s + 1) / 2 ))'
+
 exts_list = [
-    ('llvmlite', '0.45.0', {
+    ('llvmlite', '0.47.0', {
         'preinstallopts': local_llvmlite_preinstallopts,
-        'checksums': ['ceb0bcd20da949178bd7ab78af8de73e9f3c483ac46b5bef39f06a4862aa8336'],
+        'checksums': ['62031ce968ec74e95092184d4b0e857e444f8fdff0b8f9213707699570c33ccc'],
         'testinstall': True,
         'runtest': 'python -m llvmlite.tests',
     }),
     (name, version, {
-        'checksums': ['2afcc7899dc93fefecbb274a19c592170bc2dbfae02b00f83e305332a9857a5a'],
+        'checksums': ['19357146c32fe9ed25059ab915e8465fb13951cf6b0aace3826b76886373ab23'],
         'testinstall': True,
-        'runtest': 'cd %(builddir)s && python -m numba.runtests -m %(parallel)s',
+        'runtest': (
+            'cd %(builddir)s && '
+            # NUMBA_NUM_THREADS=2 is needed for tests that call set_num_threads(2)
+            f'NUMBA_NUM_THREADS=2 python -m numba.runtests -m {local_test_workers}'
+        ),
     }),
 ]
 
@@ -48,8 +55,6 @@ sanity_check_paths = {
     'dirs': ['lib/python%(pyshortver)s/site-packages'],
 }
 
-sanity_check_commands = [
-    'numba --help',
-]
+sanity_check_commands = ['numba --help']
 
 moduleclass = 'lang'

Diff against numba-0.62.0-foss-2025b.eb

easybuild/easyconfigs/n/numba/numba-0.62.0-foss-2025b.eb

diff --git a/easybuild/easyconfigs/n/numba/numba-0.62.0-foss-2025b.eb b/easybuild/easyconfigs/n/numba/numba-0.65.1-foss-2026.1.eb
index b5fae7b510..3e380ba4b2 100644
--- a/easybuild/easyconfigs/n/numba/numba-0.62.0-foss-2025b.eb
+++ b/easybuild/easyconfigs/n/numba/numba-0.65.1-foss-2026.1.eb
@@ -1,25 +1,25 @@
 easyblock = 'PythonBundle'
 
 name = 'numba'
-version = '0.62.0'
+version = '0.65.1'
 
 homepage = 'https://numba.pydata.org/'
 description = """Numba is an Open Source NumPy-aware optimizing compiler for
 Python sponsored by Continuum Analytics, Inc. It uses the remarkable LLVM
 compiler infrastructure to compile Python syntax to machine code."""
 
-toolchain = {'name': 'foss', 'version': '2025b'}
+toolchain = {'name': 'foss', 'version': '2026.1'}
 toolchainopts = {'pic': True}
 
 builddependencies = [
-    ('CMake', '4.0.3'),
-    ('LLVM', '20.1.8'),
+    ('CMake', '4.2.1'),
+    ('LLVM', '20.1.8', '-llvmlite'),  # llvmlite 0.47.0 supports only LLVM 20
 ]
-
 dependencies = [
-    ('Python', '3.13.5'),
-    ('SciPy-bundle', '2025.07'),
-    ('Z3', '4.15.1'),
+    ('Python', '3.14.2'),
+    ('Python-bundle-PyPI', '2026.04'),
+    ('SciPy-bundle', '2026.05'),
+    ('Z3', '4.15.4'),
     ('zstd', '1.5.7'),
 ]
 
@@ -27,17 +27,24 @@ local_llvmlite_preinstallopts = "export LLVM_CONFIG=${EBROOTLLVM}/bin/llvm-confi
 local_llvmlite_preinstallopts += "sed -i 's/llvm_config, \"--system-libs/llvm_config, \"--link-static --system-libs/' "
 local_llvmlite_preinstallopts += "ffi/build.py && "
 
+# Limit numba.runtests worker processes to avoid multiprocessing/forkserver hangs
+local_test_workers = '$(( %(parallel)s > 8 ? 4 : (%(parallel)s + 1) / 2 ))'
+
 exts_list = [
-    ('llvmlite', '0.45.0', {
+    ('llvmlite', '0.47.0', {
         'preinstallopts': local_llvmlite_preinstallopts,
-        'checksums': ['ceb0bcd20da949178bd7ab78af8de73e9f3c483ac46b5bef39f06a4862aa8336'],
+        'checksums': ['62031ce968ec74e95092184d4b0e857e444f8fdff0b8f9213707699570c33ccc'],
         'testinstall': True,
         'runtest': 'python -m llvmlite.tests',
     }),
     (name, version, {
-        'checksums': ['2afcc7899dc93fefecbb274a19c592170bc2dbfae02b00f83e305332a9857a5a'],
+        'checksums': ['19357146c32fe9ed25059ab915e8465fb13951cf6b0aace3826b76886373ab23'],
         'testinstall': True,
-        'runtest': 'cd %(builddir)s && python -m numba.runtests -m %(parallel)s',
+        'runtest': (
+            'cd %(builddir)s && '
+            # NUMBA_NUM_THREADS=2 is needed for tests that call set_num_threads(2)
+            f'NUMBA_NUM_THREADS=2 python -m numba.runtests -m {local_test_workers}'
+        ),
     }),
 ]
 
@@ -48,8 +55,6 @@ sanity_check_paths = {
     'dirs': ['lib/python%(pyshortver)s/site-packages'],
 }
 
-sanity_check_commands = [
-    'numba --help',
-]
+sanity_check_commands = ['numba --help']
 
 moduleclass = 'lang'

Diff against numba-0.60.0-foss-2024a.eb

easybuild/easyconfigs/n/numba/numba-0.60.0-foss-2024a.eb

diff --git a/easybuild/easyconfigs/n/numba/numba-0.60.0-foss-2024a.eb b/easybuild/easyconfigs/n/numba/numba-0.65.1-foss-2026.1.eb
index f138e3f9ac..3e380ba4b2 100644
--- a/easybuild/easyconfigs/n/numba/numba-0.60.0-foss-2024a.eb
+++ b/easybuild/easyconfigs/n/numba/numba-0.65.1-foss-2026.1.eb
@@ -1,41 +1,50 @@
 easyblock = 'PythonBundle'
 
 name = 'numba'
-version = '0.60.0'
+version = '0.65.1'
 
 homepage = 'https://numba.pydata.org/'
 description = """Numba is an Open Source NumPy-aware optimizing compiler for
 Python sponsored by Continuum Analytics, Inc. It uses the remarkable LLVM
 compiler infrastructure to compile Python syntax to machine code."""
 
-toolchain = {'name': 'foss', 'version': '2024a'}
+toolchain = {'name': 'foss', 'version': '2026.1'}
 toolchainopts = {'pic': True}
 
 builddependencies = [
-    ('LLVM', '14.0.6', '-llvmlite'),
+    ('CMake', '4.2.1'),
+    ('LLVM', '20.1.8', '-llvmlite'),  # llvmlite 0.47.0 supports only LLVM 20
 ]
-
 dependencies = [
-    ('Python', '3.12.3'),
-    ('SciPy-bundle', '2024.05'),
+    ('Python', '3.14.2'),
+    ('Python-bundle-PyPI', '2026.04'),
+    ('SciPy-bundle', '2026.05'),
+    ('Z3', '4.15.4'),
+    ('zstd', '1.5.7'),
 ]
 
 local_llvmlite_preinstallopts = "export LLVM_CONFIG=${EBROOTLLVM}/bin/llvm-config && "
 local_llvmlite_preinstallopts += "sed -i 's/llvm_config, \"--system-libs/llvm_config, \"--link-static --system-libs/' "
 local_llvmlite_preinstallopts += "ffi/build.py && "
 
+# Limit numba.runtests worker processes to avoid multiprocessing/forkserver hangs
+local_test_workers = '$(( %(parallel)s > 8 ? 4 : (%(parallel)s + 1) / 2 ))'
+
 exts_list = [
-    ('llvmlite', '0.43.0', {
+    ('llvmlite', '0.47.0', {
         'preinstallopts': local_llvmlite_preinstallopts,
-        'checksums': ['ae2b5b5c3ef67354824fb75517c8db5fbe93bc02cd9671f3c62271626bc041d5'],
+        'checksums': ['62031ce968ec74e95092184d4b0e857e444f8fdff0b8f9213707699570c33ccc'],
+        'testinstall': True,
+        'runtest': 'python -m llvmlite.tests',
     }),
     (name, version, {
-        'patches': ['numba-0.60.0_fix-parfor-with-buildmap.patch'],
-        'checksums': [
-            {'numba-0.60.0.tar.gz': '5df6158e5584eece5fc83294b949fd30b9f1125df7708862205217e068aabf16'},
-            {'numba-0.60.0_fix-parfor-with-buildmap.patch':
-             '0d8ffa14f33d0c777f0a5a81655fd0d92276289d2810ad8b28726d3588d7f9bf'},
-        ],
+        'checksums': ['19357146c32fe9ed25059ab915e8465fb13951cf6b0aace3826b76886373ab23'],
+        'testinstall': True,
+        'runtest': (
+            'cd %(builddir)s && '
+            # NUMBA_NUM_THREADS=2 is needed for tests that call set_num_threads(2)
+            f'NUMBA_NUM_THREADS=2 python -m numba.runtests -m {local_test_workers}'
+        ),
     }),
 ]
 
@@ -46,9 +55,6 @@ sanity_check_paths = {
     'dirs': ['lib/python%(pyshortver)s/site-packages'],
 }
 
-sanity_check_commands = [
-    "python -m llvmlite.tests",
-    "numba --help",
-]
+sanity_check_commands = ['numba --help']
 
 moduleclass = 'lang'

[Thyre]

As mentioned in #26378, I'd heavily suggest to go with a minimal LLVM using a -llvmlite versionsuffix, similar to LLVM-14.0.6-GCCcore-13.3.0-llvmlite.

While that one still contains some shared libraries (partially even without a version), it's very unlikely that one uses this to build something else by mistake.
We could even catch this by the EasyConfig CI if we want to.

$ find $EBROOTLLVM/ -name "*.so"
LLVM/14.0.6-GCCcore-13.3.0-llvmlite/lib/libLTO.so
LLVM/14.0.6-GCCcore-13.3.0-llvmlite/lib/libLLVM-14.0.6.so
LLVM/14.0.6-GCCcore-13.3.0-llvmlite/lib/libLLVM.so
LLVM/14.0.6-GCCcore-13.3.0-llvmlite/lib/libLLVM-14.so
LLVM/14.0.6-GCCcore-13.3.0-llvmlite/lib/libRemarks.so 

[Thyre]

This wasn't required in prior versions of numba.
Did you encounter these issues with only this numba version?

Also, why hard limit to 8 threads?
Does this affect users using numba later on?

[pavelToman]

There are some tests which needs NUMBA_NUM_THREADS=2 (tests that call set_num_threads(2)) and it does not work with %(parallel)s in runtest for me.

[Thyre]

Hm... lets keep it for now.
I'll also do some testing with that once we use the minimal LLVM instead of a full one.

Running with max. 8 threads should probably still be fine. The tests do not take too long.
We should really make sure that there is not some kind of regression impacting users of the module later on. That would be unfortunate...

[boegel]

As mentioned in #26378, I'd heavily suggest to go with a minimal LLVM using a -llvmlite versionsuffix, similar to LLVM-14.0.6-GCCcore-13.3.0-llvmlite.

While that one still contains some shared libraries (partially even without a version), it's very unlikely that one uses this to build something else by mistake. We could even catch this by the EasyConfig CI if we want to.

$ find $EBROOTLLVM/ -name "*.so"
LLVM/14.0.6-GCCcore-13.3.0-llvmlite/lib/libLTO.so
LLVM/14.0.6-GCCcore-13.3.0-llvmlite/lib/libLLVM-14.0.6.so
LLVM/14.0.6-GCCcore-13.3.0-llvmlite/lib/libLLVM.so
LLVM/14.0.6-GCCcore-13.3.0-llvmlite/lib/libLLVM-14.so
LLVM/14.0.6-GCCcore-13.3.0-llvmlite/lib/libRemarks.so 

We've quickly discussed this in today's EasyBuild conf call, and I agree with @Thyre: introducing an older LLVM version is a necessary evil here, but it should be a stripped down minimal easyconfig, to avoid that it would cause trouble in other contexts (like combining software X that requires LLVM 21.x as runtime dep with software Y which requires an older LLVM 20.x as build dependency, for example).

@pavelToman Can you look into reworking the LLVM 20.x easyconfig included in here to be simular to the LLVM-14.0.6-GCCcore-13.3.0-llvmlite.eb easyconfig, which was used as build dependency for numba-0.60.0-foss-2024a.eb ?
We were lucky enough to not need an approach like this for foss/2025a and foss/2025b, but we need to resort back to that approach here imho...

[pavelToman]

@Thyre could you please check the easyconfig for LLVM-llvmlite? I am not sure about skip all patches and also if the build_targets = ['all'] could be removed.

[pavelToman]

@boegelbot please test @ jsc-zen3

[boegelbot]

@pavelToman: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de

PR test command 'if [[ develop != 'develop' ]]; then EB_BRANCH=develop ./easybuild_develop.sh 2> /dev/null 1>&2; EB_PREFIX=/home/boegelbot/easybuild/develop source init_env_easybuild_develop.sh; fi; EB_PR=26383 EB_ARGS= EB_CONTAINER= EB_REPO=easybuild-easyconfigs EB_BRANCH=develop /opt/software/slurm/bin/sbatch --job-name test_PR_26383 --ntasks=8 ~/boegelbot/eb_from_pr_upload_jsc-zen3.sh' executed!

• exit code: 0
• output:

Submitted batch job 10978

Test results coming soon (I hope)...

Details

- notification for comment with ID 4869615436 processed

Message to humans: this is just bookkeeping information for me,
it is of no use to you (unless you think I have a bug, which I don't).

[boegelbot]

Test report by @boegelbot
SUCCESS
Build succeeded for 2 out of 2 (total: 1 hour 27 mins 4 secs) (2 easyconfigs in total)
jsczen3c1.int.jsc-zen3.fz-juelich.de - Linux Rocky Linux 9.7, x86_64, AMD EPYC-Milan Processor (zen3), Python 3.9.25
See https://gist.github.com/boegelbot/2c490800945c3aaf53b08c5610c9e02f for a full test report.

[pavelToman]

Test report by @pavelToman
SUCCESS
Build succeeded for 2 out of 2 (total: 2 hours 24 mins 51 secs) (2 easyconfigs in total)
node4203.shinx.os - Linux RHEL 9.6, x86_64, AMD EPYC 9654 96-Core Processor, Python 3.9.21
See https://gist.github.com/pavelToman/35c12321908d6a841eab26ef1d437230 for a full test report.

[pavelToman]

Test report by @pavelToman
SUCCESS
Build succeeded for 2 out of 2 (total: 2 hours 59 mins 59 secs) (2 easyconfigs in total)
node4002.donphan.os - Linux RHEL 9.6, x86_64, Intel(R) Xeon(R) Gold 6240 CPU @ 2.60GHz, 1 x NVIDIA NVIDIA A2, 595.71.05, Python 3.9.21
See https://gist.github.com/pavelToman/e78c7e1e367d81873a09394d108acdb9 for a full test report.

[pavelToman]

@boegelbot please test @ jsc-zen3

[boegelbot]

@pavelToman: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de

PR test command 'if [[ develop != 'develop' ]]; then EB_BRANCH=develop ./easybuild_develop.sh 2> /dev/null 1>&2; EB_PREFIX=/home/boegelbot/easybuild/develop source init_env_easybuild_develop.sh; fi; EB_PR=26383 EB_ARGS= EB_CONTAINER= EB_REPO=easybuild-easyconfigs EB_BRANCH=develop /opt/software/slurm/bin/sbatch --job-name test_PR_26383 --ntasks=8 ~/boegelbot/eb_from_pr_upload_jsc-zen3.sh' executed!

• exit code: 0
• output:

Submitted batch job 10986

Test results coming soon (I hope)...

Details

- notification for comment with ID 4877775926 processed

Message to humans: this is just bookkeeping information for me,
it is of no use to you (unless you think I have a bug, which I don't).

[boegelbot]

Test report by @boegelbot
SUCCESS
Build succeeded for 2 out of 2 (total: 1 hour 41 mins 1 secs) (2 easyconfigs in total)
jsczen3c1.int.jsc-zen3.fz-juelich.de - Linux Rocky Linux 9.8, x86_64, AMD EPYC-Milan Processor (zen3), Python 3.9.25
See https://gist.github.com/boegelbot/52cc20406dcaf8d5260f0d7a820b374f for a full test report.

[pavelToman]

Test report by @pavelToman
SUCCESS
Build succeeded for 2 out of 2 (total: 2 hours 50 mins 31 secs) (2 easyconfigs in total)
node4203.shinx.os - Linux RHEL 9.6, x86_64, AMD EPYC 9654 96-Core Processor, Python 3.9.21
See https://gist.github.com/pavelToman/95d29a335b901e36d1c1f91cb1c65374 for a full test report.

[Thyre]

Just FYI -- numba v0.66 (using LLVM 22) is out now: https://github.com/numba/numba/releases/tag/0.66.0

[Thyre]

@pavelToman, can you clean up the LLVM EasyConfig and fix the CI failures?

[9](https://github.com/easybuilders/easybuild-easyconfigs/actions/runs/28670721608/job/85033122331?pr=26383#step:8:160)
FAIL: test_style_conformance (test.easyconfigs.styletests.StyleTest)
Check the easyconfigs for style
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/runner/work/easybuild-easyconfigs/easybuild-easyconfigs/test/easyconfigs/styletests.py", line 64, in test_style_conformance
    stderr,
AssertionError: There shouldn't be any code style errors (and/or warnings), found 5:
/home/runner/work/easybuild-easyconfigs/easybuild-easyconfigs/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-15.2.0-llvmlite.eb:22:1: E122 continuation line missing indentation or outdented
/home/runner/work/easybuild-easyconfigs/easybuild-easyconfigs/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-15.2.0-llvmlite.eb:23:1: E122 continuation line missing indentation or outdented
/home/runner/work/easybuild-easyconfigs/easybuild-easyconfigs/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-15.2.0-llvmlite.eb:24:1: E122 continuation line missing indentation or outdented
/home/runner/work/easybuild-easyconfigs/easybuild-easyconfigs/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-15.2.0-llvmlite.eb:26:1: E122 continuation line missing indentation or outdented
/home/runner/work/easybuild-easyconfigs/easybuild-easyconfigs/easybuild/easyconfigs/l/LLVM/LLVM-20.1.8-GCCcore-15.2.0-llvmlite.eb:27:1: E122 continuation line missing indentation or outdented

[Thyre]

Test report by @Thyre
SUCCESS
Build succeeded for 2 out of 2 (total: 53 mins 21 secs) (2 easyconfigs in total)
Linux - Linux Arch Linux UNKNOWN, x86_64, AMD Ryzen 7 7800X3D 8-Core Processor (zen4), 1 x AMD Navi 48 [Radeon RX 9070/9070 XT/9070 GRE] (device id: 0x7550, gfx: gfx1201, driver: Linuxversion7.1.2-arch3-1(linux@archlinux)(gcc(GCC)16.1.120260625,GNUld(GNUBinutils)2.46.1)#1SMPPREEMPT_DYNAMICFri,03Jul202623:25:36+0000), 1 x AMD Raphael (device id: 0x164e, gfx: gfx1036, driver: Linuxversion7.1.2-arch3-1(linux@archlinux)(gcc(GCC)16.1.120260625,GNUld(GNUBinutils)2.46.1)#1SMPPREEMPT_DYNAMICFri,03Jul202623:25:36+0000), Python 3.14.6
See https://gist.github.com/Thyre/113df95f9dcfd014a7ce963c2b258a5d for a full test report.

[Thyre]

@boegelbot please test @ jsc-zen3

[boegelbot]

@Thyre: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de

PR test command 'if [[ develop != 'develop' ]]; then EB_BRANCH=develop ./easybuild_develop.sh 2> /dev/null 1>&2; EB_PREFIX=/home/boegelbot/easybuild/develop source init_env_easybuild_develop.sh; fi; EB_PR=26383 EB_ARGS= EB_CONTAINER= EB_REPO=easybuild-easyconfigs EB_BRANCH=develop /opt/software/slurm/bin/sbatch --job-name test_PR_26383 --ntasks=8 ~/boegelbot/eb_from_pr_upload_jsc-zen3.sh' executed!

• exit code: 0
• output:

Submitted batch job 10990

Test results coming soon (I hope)...

Details

- notification for comment with ID 4894069651 processed

Message to humans: this is just bookkeeping information for me,
it is of no use to you (unless you think I have a bug, which I don't).

[boegelbot]

Test report by @boegelbot
SUCCESS
Build succeeded for 2 out of 2 (total: 1 hour 38 mins 43 secs) (2 easyconfigs in total)
jsczen3c1.int.jsc-zen3.fz-juelich.de - Linux Rocky Linux 9.8, x86_64, AMD EPYC-Milan Processor (zen3), Python 3.9.25
See https://gist.github.com/boegelbot/39b903157f24df9fa0a4aa708147f07b for a full test report.

[Thyre]

LGTM, if we go for this numba version.
@boegel, what do you think? Should we stick with 0.65.1 or update to 0.66.0?

[pavelToman]

@Thyre I have discussed this with Kenneth on today meeting, and we should stick to numba-0.65.1

[boegel]

@Thyre I have discussed this with Kenneth on today meeting, and we should stick to numba-0.65.1

Indeed, I see strong need to bump this to numba 0.66.0, which was just released 2 days ago (and which requires LLVM 22.x, so would be a significant extra effort)

[boegel]

lgtm

[boegel]

Test report by @boegel
SUCCESS
Build succeeded for 2 out of 2 (total: 1 hour 50 mins 41 secs) (2 easyconfigs in total)
node4401.skiddo.os - Linux RHEL 9.6, x86_64, AMD EPYC 9755 128-Core Processor (zen5), Python 3.9.21
See https://gist.github.com/boegel/571dd946aa60a23e018879b696fccca7 for a full test report.

[boegel]

Going in, thanks @pavelToman!