Status

Changes

Stages

GitHub

Embeddable Build Status

Loading builds...

Changes

#11 (Oct 28, 2026, 4:26:43 PM)

Change MDRangePolicy to class

adrien.taberner at 4:01 AM 10/5/26

Guard with Deprecated code 5

adrien.taberner at 4:01 AM 10/5/26

Rename tile() -> tile_size()

adrien.taberner at 6:24 AM 10/5/26

Fix after rebase

adrien.taberner at 6:34 AM 10/5/26

NextSilicon: use public MDRangePolicy API

cwpears at 3:37 PM 10/6/26

core(graph): node properties enhancements with `get_properties_or`

romin.tomasetti at 3:39 AM 10/14/26

Add support for std::float16_t

Daniel Arndt at 9:40 AM 10/14/26

Use slightly larger mean_eps_expect

Daniel Arndt at 9:40 AM 10/14/26

Fix conversion warning

Daniel Arndt at 9:40 AM 10/14/26

Comment on setup in Kokkos_Macros.hpp

Daniel Arndt at 9:40 AM 10/14/26

KOKKOS_INLINE_FUNCTION -> inline

Daniel Arndt at 9:40 AM 10/14/26

Use consistent guards in Kokkos_Half_FloatingPointWrapper.hpp and Kokkos_Half_MathematicalFunctions.hpp

Daniel Arndt at 9:40 AM 10/14/26

team_scan: respect max team size

patrick.fasano at 2:45 PM 10/14/26

team_scan_ret_val: respect max_team_size

patrick.fasano at 2:45 PM 10/14/26

TestGraph: annotate death test properly for exclusion

patrick.fasano at 2:45 PM 10/14/26

NextSilicon: copy drivers with shared allocation tracking disabled

patrick.fasano at 2:45 PM 10/14/26

NextSilicon: MDRange: add max_tile_size_product

patrick.fasano at 2:45 PM 10/14/26

NextSilicon: add NextSiliconTeamMember and TeamPolicyInternal

patrick.fasano at 2:45 PM 10/14/26

NextSilicon: add league scratch buffer management to instance

patrick.fasano at 2:45 PM 10/14/26

NextSilicon: add team parallel_for, team parallel_reduce, and team parallel_scan

patrick.fasano at 2:45 PM 10/14/26

NextSilicon: ThreadVectorRange scan: match CUDA/HIP/SYCL and remove final()

patrick.fasano at 2:45 PM 10/14/26

NextSilicon: enable unit tests

patrick.fasano at 2:45 PM 10/14/26

NextSilicon: skip 7D/8D MDRange tests

patrick.fasano at 2:45 PM 10/14/26

NextSilicon: skip 4D/5D team MDRange tests

patrick.fasano at 2:45 PM 10/14/26

NextSilicon: disable range_static_batch_size (device printf)

patrick.fasano at 2:45 PM 10/14/26

NextSilicon: disable specific team_vector test configurations

patrick.fasano at 2:46 PM 10/14/26

NextSilicon: document per-test projection times in SNL CI

patrick.fasano at 2:46 PM 10/14/26

Add a comment for the larger tolerance in TestRandom.hpp

Daniel Arndt at 4:40 PM 10/14/26

Cherry-pick 5.2.2 changelog into develop

Damien L-G at 5:32 PM 10/15/26

Add memory-tracking tool to flag potential OOMs on low-memory devices (#9324)

noreply at 8:40 AM 10/16/26

hpx(parallel_reduce): do not force synchronous on parallel_reduce (#9643)

noreply at 2:48 PM 10/19/26

Add _DeathTest to ExcessMemoryAllocation category

cwpears at 4:23 PM 10/19/26

[core] Add ThreadHandle for self-similar RangePolicy for parallel_for (#8953)

noreply at 6:48 PM 10/19/26

build(deps): bump the codeql-action group with 3 updates

noreply at 10:35 PM 10/19/26

Define KOKKOS_HAS_SHARED[_HOST_PINNED]_SPACE in Kokkos_Macros.hpp (#9674)

noreply at 9:05 AM 10/20/26

core(graph): add graph to modules

romin.tomasetti at 9:37 AM 10/20/26

NextSilicon: Symmetrically disable SharedAllocation tracking during destructor call

patrick.fasano at 4:26 PM 10/20/26

Use clang build analyzer in HPSF CI (#9221)

noreply at 4:57 PM 10/20/26

Revert "Disable a new failing OpenACC unit test (SubviewALLFullExtentEquivalence) when nvc++ 26.5 is used."

lees2 at 6:05 PM 10/20/26

Revert "Re-enable OpenACC unit tests fixed by this PR."

lees2 at 6:10 PM 10/20/26

Move mdspan files to new directory structure

Christian Trott at 10:21 AM 10/21/26

Update mdspan files to d1636dadb9f7a57fcca534d4c199b5e1b7666eaf

Christian Trott at 12:04 PM 10/21/26

Exclude .clang-formant and COPYRIGHT.md in mdspan version check

Damien L-G at 12:55 PM 10/21/26

Drop header include

Damien L-G at 1:10 PM 10/21/26

Fix module submdspan_extents -> subextents

Damien L-G at 1:11 PM 10/21/26

Damien L-G at 4:55 PM 10/21/26

core(graph): ensure that the `get_device_handle` query returns the device associated with the node execution policy

maarten.arnst at 6:26 AM 10/23/26

NextSilicon: Fix exported NextAPI CMake target

patrick.fasano at 11:50 AM 10/26/26

[OpenACC, hotfix]: Fix TeamVectorRange indexing (#9691)

noreply at 5:49 PM 10/26/26

Hwm test disable switch (#9689)

noreply at 6:14 PM 10/26/26

Add a new single function to execute serial code on the device (#8123)

noreply at 7:06 PM 10/26/26

NextSilicon: remove redundant tracking guards (#9701)

noreply at 10:06 PM 10/26/26

build(deps): bump the codeql-action group with 3 updates

noreply at 10:35 PM 10/26/26

build(deps): bump actions/setup-python from 7.0.0 to 7.1.0

noreply at 10:36 PM 10/26/26

Use index_type consistently in MDRangePolicy (#9650)

noreply at 10:43 AM 10/27/26

Reuse release workflow (#9710)

noreply at 10:35 AM 10/28/26

Avoid TypeInfo instantiation when label specified (#9653)

noreply at 10:55 AM 10/28/26

Bump version from 5.2.99 to 5.3.99

Damien L-G at 10:58 AM 10/28/26

gha: Add Labeler action (#9685)

noreply at 12:02 PM 10/28/26

HPSF CI: Add NVHPC 26.9 OpenACC build (#9702)

noreply at 12:04 PM 10/28/26

Crs: dispatch to public APIs to avoid refcounted view copies (#9704)

noreply at 12:16 PM 10/28/26

simd: fix svget_neonq_* in the fallback neon sve bridge returning wrong values (#9693)

noreply at 12:29 PM 10/28/26

Re-enable OpenACC unit test fixed by this PR.

lees2 at 4:23 PM 10/28/26

#10 (Oct 15, 2026, 10:30:06 PM)

Disable a new failing OpenACC unit test (SubviewALLFullExtentEquivalence) when nvc++ 26.5 is used.

lees2 at 10:28 PM 10/15/26

#8 (Oct 14, 2026, 3:31:53 PM)

Add missing update on kokkos/mdspan/CMakeLists.txt

lees2 at 3:30 PM 10/14/26

#7 (Oct 14, 2026, 3:25:44 PM)

build(deps): bump ossf/scorecard-action from 2.4.4 to 2.5.0

noreply at 9:41 AM 10/14/26

build(deps): bump actions/download-artifact from 8.0.1 to 8.0.2

noreply at 9:41 AM 10/14/26

build(deps): bump actions/upload-artifact from 7.0.1 to 7.0.2

noreply at 9:41 AM 10/14/26

Cumulative kokkos/mdspan updates, which inlcude:

lees2 at 3:24 PM 10/14/26

#7 (Oct 14, 2026, 3:25:44 PM)

build(deps): bump ossf/scorecard-action from 2.4.4 to 2.5.0

noreply at 9:41 AM 10/14/26

build(deps): bump actions/download-artifact from 8.0.1 to 8.0.2

noreply at 9:41 AM 10/14/26

build(deps): bump actions/upload-artifact from 7.0.1 to 7.0.2

noreply at 9:41 AM 10/14/26

Cumulative kokkos/mdspan updates, which inlcude:

lees2 at 3:24 PM 10/14/26

#7 (Oct 14, 2026, 3:25:44 PM)

build(deps): bump ossf/scorecard-action from 2.4.4 to 2.5.0

noreply at 9:41 AM 10/14/26

build(deps): bump actions/download-artifact from 8.0.1 to 8.0.2

noreply at 9:41 AM 10/14/26

build(deps): bump actions/upload-artifact from 7.0.1 to 7.0.2

noreply at 9:41 AM 10/14/26

Cumulative kokkos/mdspan updates, which inlcude:

lees2 at 3:24 PM 10/14/26

#7 (Oct 14, 2026, 3:25:44 PM)

build(deps): bump ossf/scorecard-action from 2.4.4 to 2.5.0

noreply at 9:41 AM 10/14/26

build(deps): bump actions/download-artifact from 8.0.1 to 8.0.2

noreply at 9:41 AM 10/14/26

build(deps): bump actions/upload-artifact from 7.0.1 to 7.0.2

noreply at 9:41 AM 10/14/26

Cumulative kokkos/mdspan updates, which inlcude:

lees2 at 3:24 PM 10/14/26

#6 (Oct 13, 2026, 11:20:37 PM)

HIP: Increase grid size for MDRangePolicy reductions (#9535)

noreply at 2:15 PM 9/15/26

Use ViewArgsToAccessor to implement AccessorFromViewTraits (#9504)

noreply at 2:32 PM 9/15/26

Fix missing and obsolete architecture defines in CMake configuration (#9536)

noreply at 9:59 PM 9/15/26

Reduce the amount of testing for Duplicated specific ScatterView (#9540)

noreply at 3:25 PM 9/16/26

Fixing issue_9511 for crayCC compiler and HIP targets - also works with amdclang++ (#9527)

noreply at 4:50 PM 9/16/26

desul snapshot d602abff752987a45a4da574c0150a6766a39fb5 (#9528)

noreply at 9:43 AM 9/17/26

Add fallback for Kokkos::sort on non-arithmetic data (#9539)

noreply at 10:42 AM 9/17/26

Remove leftover intel classic related code (#9544)

noreply at 10:44 AM 9/17/26

Remove specialize trait (#9478)

noreply at 1:55 PM 9/17/26

Push performance benchmarking result submission into after_script

Daniel Arndt at 2:52 PM 9/17/26

Adjust BitManipuation do_compute functions to return same as public APIs (#9558)

noreply at 7:09 PM 9/17/26

cast __clz to int (#9550)

noreply at 4:20 PM 9/18/26

Silence HostSharedPtr use-after-free warning on GCC 12 on ARM with CUDA (#9551)

noreply at 4:21 PM 9/18/26

Introduce DefaultMemorySpace and DefaultHostMemorySpace (#9560)

noreply at 4:25 PM 9/18/26

Fix GitHub CI - Require Benchmark 1.9.5 (#9562)

noreply at 6:11 PM 9/18/26

Avoid creating core dump files in Jenkin CI (#9554)

noreply at 9:36 AM 9/21/26

Update kokkos/mdspan version bundled

Damien L-G at 10:41 AM 9/21/26

Use ulimit -c 0 in HPSF CI (#9567)

noreply at 2:17 PM 9/21/26

Add an aggregate SNL job we can use to manage required jobs

Christian Trott at 6:16 PM 9/21/26

Fix the skip logic for md files

Christian Trott at 6:36 PM 9/21/26

build(deps): bump the codeql-action group with 3 updates

noreply at 10:36 PM 9/21/26

Remove (deprecated) KOKKOS_OPT_RANGE_AGGRESSIVE_VECTORIZATION macro (#9571)

noreply at 1:57 PM 9/22/26

Remove deprecated AMD GPU architecture names (#9570)

noreply at 2:19 PM 9/22/26

Canonicalize Slice arguments for submdspan_mapping

Christian Trott at 3:33 PM 9/22/26

Desul atomics: fix 16-byte MSVC atomic exchange using lock table

Christian Trott at 4:12 PM 9/22/26

Update desul hash

Christian Trott at 4:13 PM 9/22/26

HPSF CI: Cancel pipeline for new commit in same branch

Daniel Arndt at 5:39 PM 9/22/26

make scratch_ptrs unsigned in test to avoid 32bit range issues (#9579)

noreply at 6:09 PM 9/22/26

snl-ci: add rocm-10.0 testing coverage (#9574)

noreply at 6:13 PM 9/22/26

parallel_scan: test with empty work range (#9319)

noreply at 6:31 PM 9/22/26

Pin third-party github action in SNL CI (#9582)

noreply at 10:32 AM 9/23/26

NextSilicon: add scratch reduction buffer management to instance

patrick.fasano at 10:55 AM 9/23/26

NextSilicon: implement parallel_reduce for RangePolicy

patrick.fasano at 10:55 AM 9/23/26

NextSilicon: enable additional tests (now possible with parallel_reduce impl)

patrick.fasano at 10:55 AM 9/23/26

NextSilicon: improve FIXME_NEXTSILICON message

patrick.fasano at 10:55 AM 9/23/26

nextsilicon: skip view_mapping_subview test

patrick.fasano at 10:55 AM 9/23/26

nextsilicon: disable lgamma unit tests

patrick.fasano at 10:55 AM 9/23/26

nextsilicon: disable mathspecialfunc_errorfunc

patrick.fasano at 10:55 AM 9/23/26

NextSilicon: don't run NextSilicon1A test binary

patrick.fasano at 10:55 AM 9/23/26

SYCL: Launch RangePolicy with 2d sycl::range

Daniel Arndt at 1:24 PM 9/23/26

Allow kernels requiring a block size smaller than warp size for HIP, CUDA and SYCL (#9339)

noreply at 2:14 PM 9/23/26

Provide std::common_type specializations for half precision (#9251)

noreply at 5:13 PM 9/23/26

Restrict auto_range to Intel GPUs

Daniel Arndt at 10:27 AM 9/24/26

Remove KOKKOS_IMPL_HOST_INACCESSIBLE_SHARED_ALLOCATION_SPECIALIZATION and KOKKOS_IMPL_SHARED_ALLOCATION_SPECIALIZATION macros

Bruno Turcksin at 11:52 AM 9/24/26

Add missing include for NextSilicon

Bruno Turcksin at 1:18 PM 9/24/26

Adapt test_team_policy_launch_with_maximum_scratch_size for PASCAL

Daniel Arndt at 4:08 PM 9/24/26

Fix test for Pascal and Maxwell and point to explanation

noreply at 4:24 PM 9/24/26

Add a test for dynamic length array scan value type (#9587)

noreply at 4:58 PM 9/24/26

Fix scalar_intra_warp_reduction for low thread count (#9572)

noreply at 5:46 PM 9/24/26

Swap expected values

Daniel Arndt at 5:46 PM 9/24/26

NextSilicon: implement parallel_scan for RangePolicy

patrick.fasano at 5:26 PM 9/25/26

NextSilicon: enable parallel_scan unit tests

patrick.fasano at 5:26 PM 9/25/26

NextSilicon: Add FIXME comment for reduce/scan work-per-thread limits

patrick.fasano at 5:27 PM 9/25/26

Check that ExecutionSpace::index_type exists and is consistent with ::size_type

Damien L-G at 2:41 PM 9/28/26

build(deps): bump the codeql-action group with 3 updates

noreply at 10:35 PM 9/28/26

Add a missing #include <unordered_map>

trevis.morvany at 4:50 AM 9/29/26

Add Hygon DCU architecture support (#9403)

noreply at 11:27 AM 9/29/26

Test for ALL equivalent to full_extent as slice

Christian Trott at 3:26 PM 9/29/26

Fix use of full_extent as subview slice specifier with bounds checking enabled

Christian Trott at 3:26 PM 9/29/26

Fix test_parallel_scan_dynamic_array (#9605)

noreply at 12:04 PM 9/30/26

Alternative to Vendor mdspan through fetchcontent (#9604)

noreply at 1:29 PM 9/30/26

Remove the Clacc compiler support for the OpenACC backend. (#9608)

noreply at 5:47 PM 9/30/26

Drop workaround for invalid use of View of Views that causes program hang on destruction (#9612)

noreply at 5:51 PM 9/30/26

Workaround erroneous std::[b]float16_t Cuda behavior (#9613)

noreply at 8:52 PM 9/30/26

Fix response file handling in nvcc_wrapper by removing quotes (#9573)

noreply at 10:43 PM 9/30/26

fully quality StdAlgoLessThanBinaryPredicate & fix ScatterView namespace inconsistency (#9609)

noreply at 10:43 PM 9/30/26

Use cleanWs in Jenkins CI

Daniel Arndt at 12:27 PM 10/1/26

SYCL: Use the same team scratch memory guards as Cuda/HIP

Daniel Arndt at 12:41 PM 10/1/26

Add tuple metafunction support to pair (#9583)

noreply at 7:12 PM 10/1/26

NextSilicon: switch from std::mutex to std::recursive_mutex

patrick.fasano at 3:42 PM 10/2/26

NextSilicon: add MDRange policy specialization and MDRange parallel_for/reduce

patrick.fasano at 3:42 PM 10/2/26

NextSilicon: enable parallel_for MDRange unit tests

patrick.fasano at 3:42 PM 10/2/26

NextSilicon: Add NextSilicon LLVM to TestCompilerMacros

patrick.fasano at 3:42 PM 10/2/26

NextSilicon: disable TestHalfOperators

patrick.fasano at 3:42 PM 10/2/26

Remove unused TestLogicalSpaces.cpp

cwpears at 7:44 AM 10/5/26

NextSilicon: disable nextsilicon.range_dynamic_policy

cwpears at 11:18 AM 10/5/26

Desul atomics: Relax semantics of atomic floating-point min/max

Damien L-G at 1:54 PM 10/5/26

Update desul commit hash value

Damien L-G at 1:54 PM 10/5/26

Cuda: Prefer block sizes above 32 threads

Daniel Arndt at 5:08 PM 10/5/26

NextSilicon: MDRange parallel_{for,reduce}: pass through exec instance (#9622)

noreply at 11:31 PM 10/5/26

Host: parallel_scan(range, functor, const Reducer&) contributes final result

cwpears at 6:59 PM 10/6/26

Remove CUDA constant memory dispatch when `__grid_constant__` is available at compile time

noreply at 9:49 PM 10/6/26

Update security vulnerability reporting guidance

Damien L-G at 10:13 PM 10/6/26

make `final` anonymous

noreply at 10:29 AM 10/7/26

remove HostSpace argument from mirror view

cwpears at 10:52 AM 10/7/26

Fix #9477: signedness inconsistency in scan/reduce (#9633)

noreply at 12:02 PM 10/7/26

Test TeamPolicy reduce on uint8_t[] and uint16_t[]

bmkelle at 12:31 PM 10/7/26

Remove redundant casts

bmkelle at 1:03 PM 10/7/26

snl-ci: update cuda jobs to test with external mdspan

ndellin at 1:26 PM 10/7/26

Update .github/workflows/snl-ci.yml

noreply at 1:37 PM 10/7/26

Add test for large uint8_t[] reduction

bmkelle at 1:40 PM 10/7/26

bmkelle at 1:41 PM 10/7/26

Update license to match closer the original LLVM Project one (#9636)

noreply at 1:59 PM 10/7/26

use fetch-depth: 0 with mdspan checkout to pull full history

ndellin at 3:22 PM 10/7/26

Remove redundant init/join in functor

bmkelle at 3:51 PM 10/7/26

Store mdspan hash as env variable to use as ref

ndellin at 4:53 PM 10/7/26

Add Nvidia Vera CPU support (#9630)

noreply at 10:02 PM 10/7/26

Cuda, SYCL: Remove (unused) usage of z-launch coordinate in Team kernels (#9640)

noreply at 9:57 AM 10/8/26

Move storage of mdspan hash to its own step

ndellin at 12:13 PM 10/8/26

Disable large array reduce test on SYCL

bmkelle at 1:02 PM 10/8/26

bmkelle at 1:29 PM 10/8/26

Allow DualView with mdspan-style template args (#9638)

noreply at 3:06 PM 10/8/26

Apply suggestion from @crtrott

noreply at 4:13 PM 10/8/26

Update .github/workflows/snl-ci.yml

noreply at 5:27 PM 10/8/26

Fix inverted memory space order in access trait

maarten.arnst at 3:06 AM 10/9/26

NextSilicon: enable more MDRange tests (#9623)

noreply at 11:19 AM 10/9/26

Fix mistaken removal of tests from default test list (#9646)

noreply at 11:51 AM 10/9/26

simd: Skip masked-off lane in scalar masked scatter

thomas.rothe at 1:18 PM 10/9/26

simd: Use the plain array in the masked-off scatter test

thomas.rothe at 12:08 PM 10/12/26

Fix DynRankView rank when wrapping a pointer with an AccessorArg (#9651)

noreply at 4:34 PM 10/12/26

Change global tag variables accessed in mdspan/subview-related functions to local variables.

lees2 at 11:17 PM 10/13/26

Re-enable OpenACC unit tests fixed by this PR.

lees2 at 11:17 PM 10/13/26

Update mdspan-hash.txt

lees2 at 11:18 PM 10/13/26

#3 (Sep 15, 2026, 11:57:58 AM)

Reduce scatterview test time (#9537)

noreply at 3:18 PM 9/14/26

Fix warnings for self similar RangePolicy interface when instantiating for a host execution space in a CUDA/HIP build (#9487)

noreply at 3:22 PM 9/14/26

Change KOKKOS_IMPL_[B]HALF_TYPE_DEFINED to KOKKOS_HAS_NATIVE_[B]HALF_TYPE (#9519)

noreply at 5:52 PM 9/14/26

skip exec_space_thread_safety_range_scan_different_sizes for SYCL with NVIDIA GPUs (#9534)

noreply at 6:48 PM 9/14/26

Disable failing OpenACC unit tests when NVHPC > 24.5 (#9406)

noreply at 9:50 AM 9/15/26

Fix OpenACC CI for unreacheble code in tests (#9538)

noreply at 9:59 AM 9/15/26

build(deps): bump the codeql-action group with 3 updates (#9541)

noreply at 10:02 AM 9/15/26

Change global tag variables accessed in mdspan/subview-related functions to local variables.

lees2 at 11:37 AM 9/15/26

Re-enable OpenACC unit tests fixed by this PR.

lees2 at 11:56 AM 9/15/26