{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":681688647,"defaultBranch":"develop","name":"rocPRIM","ownerLogin":"umfranzw","currentUserCanPush":false,"isFork":true,"isEmpty":false,"createdAt":"2023-08-22T14:41:47.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/212066?v=4","public":true,"private":false,"isOrgOwned":false},"refInfo":{"name":"","listCacheKey":"v0:1720020266.0","currentOid":""},"activityList":{"items":[{"before":"a81e5d81fb97f2ed1c9a3e10f2314e0f5b264332","after":"32ec07553cd3e20422cdee0f822a5ff26b7de6bd","ref":"refs/heads/set_asan_default_targets","pushedAt":"2024-07-08T13:13:25.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Set default target architectures for ASAN build\n\nDevice-side address sanitizer instrumentation requires xnack+.\nSet the default target GPU architectures to those that provide xnack.","shortMessageHtmlLink":"Set default target architectures for ASAN build"}},{"before":"db30c5b2fe09667d2b13757df384c2f68e29071e","after":"529b8b6e29e23f04b4022b8a207deb4d47050aa9","ref":"refs/heads/develop","pushedAt":"2024-07-08T13:07:17.000Z","pushType":"push","commitsCount":3,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Fixed FileNotFound error when passing path to --install_dir while running rtest.py on windows (#561)\n\n* Fixed FileNotFound error on windows\r\n\r\n* --- (#562)\r\n\r\nupdated-dependencies:\r\n- dependency-name: requests\r\n dependency-type: indirect\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>\r\n\r\n* Enable external CI pipeline triggers (#560)\r\n\r\n* updated changelog\r\n\r\n* updated copyright in rtest.py\r\n\r\n* fixed copyright to 2021-2024\r\n\r\n* Update Read the Docs config to Python 3.10 and latest rocm-docs-core (#564)\r\n\r\n* cmake: don't set CMAKE_C_COMPILER, as rocPRIM is a CXX project (#568)\r\n\r\n* Bump rocm-docs-core from 1.2.0 to 1.4.0 in /docs/sphinx (#569)\r\n\r\nBumps [rocm-docs-core](https://github.com/ROCm/rocm-docs-core) from 1.2.0 to 1.4.0.\r\n- [Release notes](https://github.com/ROCm/rocm-docs-core/releases)\r\n- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/develop/CHANGELOG.md)\r\n- [Commits](https://github.com/ROCm/rocm-docs-core/compare/v1.2.0...v1.4.0)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: rocm-docs-core\r\n dependency-type: direct:production\r\n update-type: version-update:semver-minor\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>\r\n\r\n* Fixed FileNotFound error on windows\r\n\r\n* updated changelog\r\n\r\n* updated copyright in rtest.py\r\n\r\n* fixed copyright to 2021-2024\r\n\r\n---------\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>\r\nCo-authored-by: Joseph Macaranas <145489236+amd-jmacaran@users.noreply.github.com>\r\nCo-authored-by: Sam Wu <22262939+samjwu@users.noreply.github.com>\r\nCo-authored-by: Steve Leung ","shortMessageHtmlLink":"Fixed FileNotFound error when passing path to --install_dir while run…"}},{"before":null,"after":"a81e5d81fb97f2ed1c9a3e10f2314e0f5b264332","ref":"refs/heads/set_asan_default_targets","pushedAt":"2024-07-03T15:24:26.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Set default target architectures for ASAN build\n\nDevice-side address sanitizer instrumentation requires xnack+.\nSet the default target GPU architectures to those that provide xnack.","shortMessageHtmlLink":"Set default target architectures for ASAN build"}},{"before":"d1cba25180cfd1058180bd129e69ea2657602d9e","after":"db30c5b2fe09667d2b13757df384c2f68e29071e","ref":"refs/heads/develop","pushedAt":"2024-07-03T14:27:54.000Z","pushType":"push","commitsCount":3,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Bump rocm-docs-core from 1.4.0 to 1.4.1 in /docs/sphinx (#578)\n\nBumps [rocm-docs-core](https://github.com/ROCm/rocm-docs-core) from 1.4.0 to 1.4.1.\r\n- [Release notes](https://github.com/ROCm/rocm-docs-core/releases)\r\n- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/develop/CHANGELOG.md)\r\n- [Commits](https://github.com/ROCm/rocm-docs-core/compare/v1.4.0...v1.4.1)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: rocm-docs-core\r\n dependency-type: direct:production\r\n update-type: version-update:semver-patch\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>","shortMessageHtmlLink":"Bump rocm-docs-core from 1.4.0 to 1.4.1 in /docs/sphinx (ROCm#578)"}},{"before":"8691783ba2908b3ccb8dcb2648d704581f8ea6da","after":"d1cba25180cfd1058180bd129e69ea2657602d9e","ref":"refs/heads/develop","pushedAt":"2024-06-12T14:48:54.000Z","pushType":"push","commitsCount":5,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Bump rocm-docs-core from 1.2.0 to 1.4.0 in /docs/sphinx (#569)\n\nBumps [rocm-docs-core](https://github.com/ROCm/rocm-docs-core) from 1.2.0 to 1.4.0.\r\n- [Release notes](https://github.com/ROCm/rocm-docs-core/releases)\r\n- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/develop/CHANGELOG.md)\r\n- [Commits](https://github.com/ROCm/rocm-docs-core/compare/v1.2.0...v1.4.0)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: rocm-docs-core\r\n dependency-type: direct:production\r\n update-type: version-update:semver-minor\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>","shortMessageHtmlLink":"Bump rocm-docs-core from 1.2.0 to 1.4.0 in /docs/sphinx (ROCm#569)"}},{"before":"ea1bef6f8b98ff8bc88d5369724d510b22e366a5","after":"8691783ba2908b3ccb8dcb2648d704581f8ea6da","ref":"refs/heads/develop","pushedAt":"2024-05-10T14:48:31.000Z","pushType":"push","commitsCount":11,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Bump jinja2 from 3.1.3 to 3.1.4 in /docs/sphinx (#556)\n\nBumps [jinja2](https://github.com/pallets/jinja) from 3.1.3 to 3.1.4.\r\n- [Release notes](https://github.com/pallets/jinja/releases)\r\n- [Changelog](https://github.com/pallets/jinja/blob/main/CHANGES.rst)\r\n- [Commits](https://github.com/pallets/jinja/compare/3.1.3...3.1.4)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: jinja2\r\n dependency-type: indirect\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>","shortMessageHtmlLink":"Bump jinja2 from 3.1.3 to 3.1.4 in /docs/sphinx (ROCm#556)"}},{"before":"7e343a6424f4e42c0c60c2c8fd565907ed27b799","after":"ea1bef6f8b98ff8bc88d5369724d510b22e366a5","ref":"refs/heads/develop","pushedAt":"2024-04-04T20:18:30.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Update supported data type table (#520)\n\n* Update supported data type table\r\n\r\nUse the type names from the precision landing page of the rocm documentation\r\n\r\n* Update Changelog","shortMessageHtmlLink":"Update supported data type table (ROCm#520)"}},{"before":"eeae9c106d77344c4b02242548f1287295ad3a2d","after":"00be17d5716ef962294e15daf60419e9ead9067b","ref":"refs/heads/SWDEV-447229","pushedAt":"2024-04-01T19:07:37.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Add device partition unit test to check behaviour around shared memory limit\n\nTest the edge case where the data passed to the device partition\nalgorithm will consume the maximum allowable amount of shared memory.\nSince the algorithm itself also requires some shared memory to store\nstate, this should push us over the max limit. In this case,\nthe block size should be reduced to compensate.","shortMessageHtmlLink":"Add device partition unit test to check behaviour around shared memor…"}},{"before":"684c56c8adce8552fe7c0f754e2752ad0b1c5002","after":"7e343a6424f4e42c0c60c2c8fd565907ed27b799","ref":"refs/heads/develop","pushedAt":"2024-04-01T12:41:59.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Bump rocm-docs-core from 0.37.0 to 0.38.0 in /docs/sphinx (#544)\n\nBumps [rocm-docs-core](https://github.com/RadeonOpenCompute/rocm-docs-core) from 0.37.0 to 0.38.0.\r\n- [Release notes](https://github.com/RadeonOpenCompute/rocm-docs-core/releases)\r\n- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/develop/CHANGELOG.md)\r\n- [Commits](https://github.com/RadeonOpenCompute/rocm-docs-core/compare/v0.37.0...v0.38.0)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: rocm-docs-core\r\n dependency-type: direct:production\r\n update-type: version-update:semver-minor\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>","shortMessageHtmlLink":"Bump rocm-docs-core from 0.37.0 to 0.38.0 in /docs/sphinx (ROCm#544)"}},{"before":"d942e8698771cd67142dac3b51ca1880d45f0b17","after":"eeae9c106d77344c4b02242548f1287295ad3a2d","ref":"refs/heads/SWDEV-447229","pushedAt":"2024-03-26T16:02:32.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Update compile-time shared memory usage check for device_partition\n\nThe device partition algorithm uses a default_select_config struct\nto detect which architecture we're running on.\n\nThe default_select_config struct eventually causes the creation of\na struct of type limit_block_size. This struct is used to detect if\nthe launch configuration that's being used (block size and amount of\nshared memory per thread) will cause on the selected device to use\nmore than 32 KiB of shared memory. If so, then limit_block_size\nattempts to reduce the block size (divides it by 2) and checks the\nshared memory usage again.\n\nIf the element type is large enough, it is possible to get into\na situation where, even if we use the minimum block size (a single\nwavefront of threads) and give the threads the minimum possible\nnumber of elements to work on (1 each), we will still use more\nthan 32 KiB of shared memory.\n\nThe limit_block_size struct assumes that the amount of shared memory\nthat will be used is equal to the block size multiplied by the amount\nof memory required per thread. However, the device partition algorithm\nactually requires slightly more shared memory than this, because it\ndoes an extra allocation to store the lookback scan's state.\n\nIt's not really feasible to move this lookback scan state out of\nshared memory because all threads in the block need access to it.\n\nThis change modifies the limit_block_size struct so that it accepts\nan \"ExtraSharedMemory\" template parameter, and updates the shared\nmemory check it performs so that it takes this value into account.\n\nIt also updates the device partition's config-creating code so that\nit passes in the size of the lookback scan state.","shortMessageHtmlLink":"Update compile-time shared memory usage check for device_partition"}},{"before":null,"after":"d942e8698771cd67142dac3b51ca1880d45f0b17","ref":"refs/heads/SWDEV-447229","pushedAt":"2024-03-26T15:58:17.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"}},{"before":"a26745ff840ec095199555990dd6f1e04a6c642b","after":"684c56c8adce8552fe7c0f754e2752ad0b1c5002","ref":"refs/heads/develop","pushedAt":"2024-03-20T19:55:34.000Z","pushType":"push","commitsCount":5,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Switch from using hipcc to using Clang compiler (#514)\n\n* Change default from hipcc to amdclang\r\n\r\n* use clang not amdclang\r\n\r\n* remove compiler from install script\r\n\r\n* remove redundant toolchain argument\r\n\r\n* Switch back to amdclang\r\n\r\n* Update to correct flag behaviour\r\n\r\n* Update CI script","shortMessageHtmlLink":"Switch from using hipcc to using Clang compiler (ROCm#514)"}},{"before":"72ba1aa6a7109934c638e235e27101dcae6e66fd","after":"734088602c178af6352843b688b7b62261c8cc50","ref":"refs/heads/add_contributing_doc","pushedAt":"2024-03-15T20:31:23.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Add Contribution Readme\n\nAdd a CONTRIBUTING.md file that contains information users will need\nif they want to contribute to the repository.\n\nThis is based on the template located here:\nhttps://github.com/ROCm/.github/blob/main/docs/templates/contributing_template.md","shortMessageHtmlLink":"Add Contribution Readme"}},{"before":"9217a14ad37c9d780985160e62a8d710cb3991a8","after":"72ba1aa6a7109934c638e235e27101dcae6e66fd","ref":"refs/heads/add_contributing_doc","pushedAt":"2024-03-13T13:58:16.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"}},{"before":null,"after":"9217a14ad37c9d780985160e62a8d710cb3991a8","ref":"refs/heads/add_contributing_doc","pushedAt":"2024-03-13T13:53:50.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"}},{"before":"0e8c448f3b8b66a7a565c92754d3650c2e2208d8","after":"a26745ff840ec095199555990dd6f1e04a6c642b","ref":"refs/heads/develop","pushedAt":"2024-03-13T12:53:51.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Bump rocm-docs-core from 0.35.1 to 0.36.0 in /docs/sphinx (#536)\n\nBumps [rocm-docs-core](https://github.com/RadeonOpenCompute/rocm-docs-core) from 0.35.1 to 0.36.0.\r\n- [Release notes](https://github.com/RadeonOpenCompute/rocm-docs-core/releases)\r\n- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/develop/CHANGELOG.md)\r\n- [Commits](https://github.com/RadeonOpenCompute/rocm-docs-core/compare/v0.35.1...v0.36.0)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: rocm-docs-core\r\n dependency-type: direct:production\r\n update-type: version-update:semver-minor\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>","shortMessageHtmlLink":"Bump rocm-docs-core from 0.35.1 to 0.36.0 in /docs/sphinx (ROCm#536)"}},{"before":"3cf3144004eb4d5afd37160167ffafe6661f0ef8","after":"7603348222577bb96276754ebdd0916ebee40929","ref":"refs/heads/re-enable_dev_adj_diff_test","pushedAt":"2024-03-11T19:34:18.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Re-enable hipGraph version of device_adjacent_difference test\n\nPreviously, we disabled the hipGraph version of the LargeIndices test\nfor the device_adjacent_difference algorithm because it was failing on\nNavi3x and MI300.\n\nThe root cause of this failure was a call to hipMemset that occasionally\ncontinued to run while the hipGraph was instantiated. I've now learned\nthat this is expected behaviour - hipMemset runs asynchronously unless the\npointer it is passed refers to pinned host memory. This is true even if\nthe call to hipMemset is issued into the default stream, like it is in\nthis case. hipMemset*Async*'s behaviour differs only with respect to\npinned host memory (it's non-blocking in that case). This behaviour is\nnot currently mentioned in the HIP documentation, so I'll put in a\nrequest to add it there.\n\nThis change just inserts a call to hipDeviceSynchronize to ensure that\nthe hipMemset call completes before the graph is instantiated.\nIt also removes the GTEST_SKIP() call that disabled the hipGraph version\nof the test.","shortMessageHtmlLink":"Re-enable hipGraph version of device_adjacent_difference test"}},{"before":"6325547d514b46d1ab51aff0195851b3fcc626d1","after":"0e8c448f3b8b66a7a565c92754d3650c2e2208d8","ref":"refs/heads/develop","pushedAt":"2024-03-11T19:33:15.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Bump rocm-docs-core from 0.35.0 to 0.35.1 in /docs/sphinx (#535)\n\nBumps [rocm-docs-core](https://github.com/RadeonOpenCompute/rocm-docs-core) from 0.35.0 to 0.35.1.\r\n- [Release notes](https://github.com/RadeonOpenCompute/rocm-docs-core/releases)\r\n- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/develop/CHANGELOG.md)\r\n- [Commits](https://github.com/RadeonOpenCompute/rocm-docs-core/compare/v0.35.0...v0.35.1)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: rocm-docs-core\r\n dependency-type: direct:production\r\n update-type: version-update:semver-patch\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>","shortMessageHtmlLink":"Bump rocm-docs-core from 0.35.0 to 0.35.1 in /docs/sphinx (ROCm#535)"}},{"before":"d1dc46b602b982c7aafc59e4636fddc03feea215","after":"3cf3144004eb4d5afd37160167ffafe6661f0ef8","ref":"refs/heads/re-enable_dev_adj_diff_test","pushedAt":"2024-03-07T13:56:49.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"}},{"before":null,"after":"d1dc46b602b982c7aafc59e4636fddc03feea215","ref":"refs/heads/re-enable_dev_adj_diff_test","pushedAt":"2024-03-01T21:44:42.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"}},{"before":"2939260ad8a4f9f25615e595645ae9e83ee5357e","after":"6325547d514b46d1ab51aff0195851b3fcc626d1","ref":"refs/heads/develop","pushedAt":"2024-03-01T21:19:56.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Bump rocm-docs-core from 0.34.2 to 0.35.0 in /docs/sphinx (#533)\n\nBumps [rocm-docs-core](https://github.com/RadeonOpenCompute/rocm-docs-core) from 0.34.2 to 0.35.0.\r\n- [Release notes](https://github.com/RadeonOpenCompute/rocm-docs-core/releases)\r\n- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/develop/CHANGELOG.md)\r\n- [Commits](https://github.com/RadeonOpenCompute/rocm-docs-core/compare/v0.34.2...v0.35.0)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: rocm-docs-core\r\n dependency-type: direct:production\r\n update-type: version-update:semver-minor\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>","shortMessageHtmlLink":"Bump rocm-docs-core from 0.34.2 to 0.35.0 in /docs/sphinx (ROCm#533)"}},{"before":null,"after":"3e7b71071f42faa2668e15f56e0fe83777cf8e72","ref":"refs/heads/dev_adj_diff_reproducer","pushedAt":"2024-02-28T16:14:47.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Reproducer for device_adjacent_difference hipGraph unit test bug\n\nThe device_adjacent_difference test fails intermittently when\nrun using hipGraphs on gfx1101. The issue appears to be related\nto the hipMemset calls the are made into the default stream\njust before stream capture is started. If I insert a call to\nhipStreamSynchronize(0) between the hipMemset calls and the\nbeginning of the graph capture, the test consistently passes.\n\nI believe this behaviour should not be occurring because calls\nsubmitted to the default stream are supposed to be run synchronously.","shortMessageHtmlLink":"Reproducer for device_adjacent_difference hipGraph unit test bug"}},{"before":"3e31a1513a8ed3f7306bd02f0e45bf0e26defc02","after":"2939260ad8a4f9f25615e595645ae9e83ee5357e","ref":"refs/heads/develop","pushedAt":"2024-02-26T14:13:13.000Z","pushType":"push","commitsCount":6,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Merge back change from ROCm 6.1 (#532)\n\n* 6.1 bulk update from develop branch 2024-1-16 (#510)\r\n\r\n* StreamHPC 2023-10-18 (#480)\r\n\r\n* Excessive shared memory usage in block_shuffle fix\r\n\r\n* remove block_sort_algorithm template param from block_sort_kernel_impl and block_sort_impl\r\n\r\n* fixed compile errors\r\n\r\n* Updated ChangeLog.md\r\n\r\n* remove unnecessary code\r\n\r\n* fixed CHANGELOG.md to not be so verbose about non public api changes\r\n\r\n* Add dynamic dispatch and autotuning to device_adjacent_difference\r\n\r\n* Fix device_adjacent_difference storage type\r\n\r\n* ci: remove autotune dependency from build:benchmark\r\n\r\nThe workaround needed to make this work is has major disadvantages,\r\nand our current workflow does not make use of this dependency anyway\r\n(Currently the generated configs are checked into the repository, so\r\nthe CI would run the benchmarks on them on the next push to the\r\nmerge-request).\r\n\r\nWhen we improve automation around autotuning this could be implemented\r\nwith conditional jobs, but lets just drop the dependency for now.\r\n\r\n* test: fix indexing error test_type_helper::get_random_data\r\n\r\nIndexing was 4 based when the type has 3 variables, therefore it was\r\noverflowing. Caught with address sanitizer.\r\n\r\n* fixes for compilation in debug for radix_sort\r\n\r\n- Add force inline to onesweep kernel, to avoid too much shared memory\r\n errors\r\n- Declare `block_radix_sort::radix_bits_per_pass` to fix linker errors\r\n\r\n* fix: Detect DPP & DPP broadcast support with __GFX__ macros\r\n\r\nThe amdgpu target in clang now provides the GFX generation as a\r\npredefined macro, so we no longer need to explicitly list all targets,\r\nwhich was bad for maintenance.\r\n\r\nAlso replace the use of the generic `ROCPRIM_NAVI` which signals navi support,\r\nwith `ROCPRIM_DETAIL_HAS_DPP_BROADCAST`, a macro that explicitly\r\nstates what we're after.\r\n\r\nAlso also makes sure that `ROCPRIM_DETAIL_USE_DPP` is always defined\r\n(to 0 when DPP is disabled), previously it was undefined when\r\n`ROCPRIM_DISABLE_DPP` was set.\r\n\r\n* refactor: Use __GFX__ to detect NAVI cards\r\n\r\n* docs: Update CHANGELOG for DPP & ROCPRIM_NAVI fixes\r\n\r\n* remove deprecated structs and functions\r\n\r\n* rename scan_by_key_config_v2 to scan_by_key_config\r\nremove the option to use custom implemented config for scan_by_key\r\nupdate tests to not use custom implemented config for scan_by_key\r\n\r\n* remove the option to use custom implemented config for histogram\r\nupdate tests to not use custom implemented config for histogram\r\n\r\n* update config compile time check to a different pattern\r\n\r\n* update documentation comments for configs\r\n\r\n* change documentation comments\r\n\r\n* change documentation comments on device_radix_sort\r\nrename radix_sort_config_v2 to radix_sort_config\r\n\r\n* change documentation comment\r\nadd static_assert to check type for reduce_config\r\n\r\n* update documentation comments\r\nremove wrap_scan_config function\r\nadd static_assert to disallow custom scan_config type\r\nrename scan_config_v2 to scan_config\r\n\r\n* update documentation comments\r\n\r\n* update documentation comments\r\nmake transform_config inherit from detail::transfomr_config_params\r\nremove wrap_transform_config\r\nadd static assert to test for Config type in device_transform\r\n\r\n* remove wrap_adjacent_difference_config function\r\nadd static_assert to test config type\r\ncreate default ctor for adjacent_difference_config\r\n\r\n* add missing transform_config ctor\r\nrewrite adjacent_difference_config ctor to match other config structs\r\n\r\n* fix binary search still using wrap_transform_config\r\n\r\n* implement static_asset to make binary_search only use binary search configs, but also work with the underlying transform\r\n\r\n* update changelog\r\n\r\n* remove some *_v2s that went under the radar\r\n\r\n* remove unnecessary default values\r\n\r\n* Add binary search, lower_bound and upper_bound documentation\r\n\r\n* host_warp_size() is replaced with two different versions with parameters.\r\nthe new versions use either a device id or a stream to figure out the warp size of the device\r\n\r\n* comment out unused param names\r\n\r\n* fix typos in the documentation\r\n\r\n* move host_warp_size to config_type.hpp\r\nchanged host_warp_size signatures to fit other similar functions\r\n\r\n* add error checks to host_warp_size calls in tests and benchmarks\r\n\r\n* fix format\r\n\r\n* add missing comment\r\n\r\n* fix error handling in lookback_scan_state.hpp\r\n\r\n* fix compilation error\r\n\r\n* change block_radix_rank_match and block_histogram_atomic to use rocprim::match_any instead of implementing same functionality\r\n\r\n* change radix_digit_count_helper to use rocprim::match_any instead of implementing same functionality\r\nadded predicate param to rocprim::match_any to set invalid lanes and added tests for this functionality\r\n\r\n* add elect function to warp intrinsics\r\nadd test for elect\r\nchange block_histogram_atomic, block_radix_rank_match, device_histogram, device_radix_sort to use elect instead of copy-paste code\r\n\r\n* update match_any to return 0 when predicate is false\r\n\r\n* fix the bit check in elect function\r\n\r\n* update changelog.md\r\n\r\n* fix hard coded warps per block value to come from param in kernel\r\n\r\n* remove unused variables\r\n\r\n* fix review comments\r\nminor name changes\r\nupdate test\r\nupdate comments\r\n\r\n* update group_elect test\r\ntests multiple groups per warp\r\ndoesn't check which exact thread is elected in a group, only that one is elected\r\n\r\n* remove unnecessary comments\r\n\r\n* remove expected from group_elect test\r\nfix compile error\r\n\r\n* fix overindexing\r\n\r\n* fix review comments\r\nupdate group_elect_test to have better coverage\r\n\r\n* format\r\n\r\n* fix review comments\r\n\r\n* fix perf regression\r\n\r\n* undo group_elect in block_histogram_atomic.hpp, because of perf impact\r\n\r\n* fix bad func name in CHANGELOG.md\r\n\r\n* fix merge errors\r\n\r\n* Fix reduce_by_key algorithm so keys[0] is not flagged as a new run when is nan\r\n\r\n* make device_radix_sort compatible with compiler provided __int128_t and __uint128_t\r\n\r\n* add ifdefs to only compile int128 parts on clang/gcc\r\n\r\n* update changelog\r\n\r\n* fix for int128 to_string labdas\r\n\r\n* add test for block_radix_sort int128 support\r\n\r\n* Implement block run length decode\r\n\r\n* Fix reduce_by_key algorithm so out of bounds items are not flagged as new runs for NaNs\r\n\r\n* Add reduce_by_key test to check that flagging is correct when keys are all different\r\n\r\n* Fix performance regression observed during tuning for gfx1030 and gfx1102\r\n\r\n* Block Runlength Decode: Fix incorrect offsets and improve test\r\n\r\n* Remove duplicate key from .clang-format\r\n\r\n* Remove additional duplicates from clang-format\r\n\r\n* Fix binary_search upper/lower_bound config tuning\r\n\r\nUse specialized configurations for upper, lower, and binary search\r\nalgorithms when preforming tuning\r\n\r\n* unify language around config params in documentation\r\n\r\n* Make the autotune build job run nightly\r\n\r\n* remove radix_sort_onesweep autotuning workaround\r\n\r\n* Resolve doxygen warnings for upstream PR\r\n\r\n* Enable get_device_from_stream for Windows\r\n\r\n* Use _ENABLE_EXTENDED_ALIGNED_STORAGE for windows build in rmake.py\r\n\r\n* Bump unreleased ROCm version\r\n\r\n---------\r\n\r\nCo-authored-by: Ivan Siutsou \r\nCo-authored-by: Bence Parajdi \r\nCo-authored-by: Bálint Soproni \r\nCo-authored-by: Gergely Meszaros \r\nCo-authored-by: Beatriz Navidad Vilches \r\nCo-authored-by: Mátyás Aradi \r\n\r\n* StreamHPC 2023-11-17 (batch memcpy) (#485)\r\n\r\n* Implemented batch memcpy algorithm and relevant tests and benchmarks\r\n\r\n* Optimize match_any by using arithmetic shifts\r\n\r\nThe compiler seems to see through these much better than the conditional,\r\ngenerating bit-field extract instructions, and recognizing that the loop\r\nis a reduction.\r\n\r\n* Pedantic / consistency changes for batch memcpy\r\n\r\n* Improve interface and implementation of align_(up|down)\r\n\r\n- Use the alignment of the destination type instead of its size\r\n- Rename to emphasize that this does a form of reinterpret_cast\r\n- Use the same type as the return type and template parameter, to\r\n match the interface of built-in casts\r\n- Pedantic: use uintptr_t instead of size_t for the numerical value\r\n of a pointer\r\n- Use clangs __builtin_align_(up|down) when available\r\n\r\n* Take parameters as explicit const-ref in test_utils::bit_equal\r\n\r\nBecause these are templates this already works for non-copyable types,\r\n(as `T` will be deduced to `Type&`) but its confusing, and wouldn't work\r\nfor r-values. Because we are comparing object representations taking a copy\r\nisn't okay as that only guarantees that the value representation is copied.\r\n(I.e. padding bytes are not required to be copied when taking a parameter\r\n by copy)\r\n\r\n* Actually make custom_non(copyable|moveable)_type non (copy|move)-able\r\n\r\n* Allow passing rocprim::default_config to batch_memcpy\r\n\r\nAs all the other device functions do too.\r\n\r\n* Fix typo in cast_align_down documentation\r\n\r\n* Fixup accidentally deleted constructor of custom_non_moveable_type\r\n\r\nThis was accidentally deleted, it was meant to be defaulted.\r\nCurrently no test calls this as batch-memcpy tests only use this type\r\nat the device side.\r\n\r\n* Improve error message of test_rocprim_package\r\n\r\nThe error message of the package test wasn't very nice, improve it\r\nfor easier debugging in the future.\r\n\r\nBefore:\r\n```console\r\n❯ ./a.out\r\n98\r\n```\r\n\r\nAfter:\r\n```console\r\n❯ ./a.out\r\nError hipErrorInvalidDeviceFunction(98): invalid device function in main at test_rocprim_package.cpp:90\r\n```\r\n\r\n* Refactor test_utils::get_random_data into generate_random_data_n\r\n\r\n- Writes the output into an output iterator instead of creating &\r\n returning a vector. This allows greater flexibility for users\r\n i.e. writing random values with differing options into the same\r\n container.\r\n- Accepts a generator instead of a seed. This is more efficient, because\r\n creating an instance of an rng engine might be costly. It's also\r\n more consistent with how the standard library operates.\r\n- The naming and interface tries to mirror the stl (i.e. `std::generate_n`)\r\n- Backwards compatibility is maintained by adding test_utils::get_random_data\r\n that uses `generate_random_data_n` internally.\r\n\r\n* Refactor get_random_data into generate_random_data_n in benchmark_utils\r\n\r\nThis mirrors the test changes in the previous commit\r\n\r\n* Unify segmnented generation from test generate_random_data_n overloads\r\n\r\n* Add missing include for iterator traits to benchmark_utils\r\n\r\n* ci: use build instead rocm-build tag\r\n\r\nThis allows the build job to be performed by any runner configured\r\nfor building, instead of the ROCm-specialized builder. As the\r\ntarget architectures are specified ahead of time, the GPU is not\r\nneeded during the build process, and may be performed by any builder.\r\n\r\n* fix: Fixed doxygen warning in device_memcpy_config.hpp\r\n\r\n* Speed up / Improve data-generation in test_device_batch_memcpy\r\n\r\nDo bulk data-generation instead of individual calls, especially of\r\nindividual bytes for the data to copy.\r\nAlso changes the verification to do bulk memcmp instead of item-wise\r\ntest_utils::bit_equals for each buffer.\r\nOverall this reduces the time it takes to run the test to ~1s from\r\naround 3s.\r\n\r\n* Refactor & Speedup benchmark_device_batch_memcpy\r\n\r\n- Share the data generation between the naive and uut benchmarks\r\n- Make the data-generation be bulk using a fast random number engine\r\n (mt19937) to significantly speed it up.\r\n\r\nThe overall runtime of the benchmark decreased from 14 minutes (!) to\r\naround 2 minutes.\r\n\r\n* Fix explanation comment in batch_memcpy test/benchmark\r\n\r\n* fix include order in benchmark_device_batch_memcpy\r\n\r\n* doc: add batch memcpy to changelog\r\n\r\n---------\r\n\r\nCo-authored-by: Gergely Meszaros \r\nCo-authored-by: Robin Voetter \r\n\r\n* Add unit testing to verify that algorithms work with hipGraphs (#478)\r\n\r\n* Basic hipGraph tests\r\n\r\n* Add basic tests for graph creation, instantiation, and execution using:\r\n * stream capture\r\n * manual construction\r\n\r\n* hipGraph test for device_reduce algorithms\r\n* Added new unit tests for device_reduce, device_reduce_by_key algorithms\r\nto verify basic support for hipGraphs (no synchronous API functions are\r\ncalled within the algorithms).\r\n* Fixed up CMakeLists compile issue for tests in the test/hipgraph folder\r\n* Updated code documentation\r\n\r\n* Add hipGraph unit tests for device level algorithms\r\n\r\n* Added unit tests that run the following algorithms inside of a graph\r\n(in isolation):\r\n - device_adjacent_difference\r\n - device_binary_search\r\n - device_histogram\r\n - device_merge\r\n - device_merge_sort\r\n - device_partition\r\n - device_radix_sort\r\n - device_scan\r\n - device_segmented_reduce\r\n - device_segmented_scan\r\n - device_select\r\n - device_transform\r\n\r\n* Updated existing tests for:\r\n - device_reduce\r\n - device_reduce_by_key\r\n\r\n* Moved graph test helper functions to a separate file\r\n\r\n* Add hipGraph unit tests\r\n\r\n* Added remaining device level hipGraph unit tests\r\n\r\n* Note: currently, there are two device level algorithms that\r\ndo no work with hipGraphs because they contain synchronization\r\nbarriers. No hipGraph unit tests have been added for these\r\nalgorithms:\r\n * device_run_length_encode\r\n * device_segmented_radix_sort\r\n\r\n* Added a functional integration test for hipGraphs, which\r\nruns several algorithms back-to-back within a graph.\r\n\r\n* Refactored test helper code to remove unnecessary parameter\r\n\r\n* Set hipgraph test pointers to nullptr\r\n\r\n* Set key_type device pointers to nullptr when they are declared, for\r\n safety.\r\n\r\n* Several minor fixes for hipGraph tests\r\n* Fixed up spelling error in comments\r\n* Moved call to hipGetLastError to a more appropriate position\r\n* Removed old commented test code\r\n\r\n* Minor fixes for hipgraph unit tests\r\n* Moved several synchronization barriers so they are now outside of graph capture blocks\r\n in the test_device_partition source\r\n* Changed several loop counters to unsigned type\r\n* Updatedpgraph cmake files - removed test/hipgraph\r\n directory's CMakeLists.txt\r\n\r\n* Additional test and bugfix for hipgraph tests\r\n* Removed syncrhonization barrier in test_device_scan\r\n* Added basic test to exercise atomic function within a hipgraph\r\n* Rebased and resolved merge conflicts\r\n\r\n* readme and changelog updates (#486)\r\n\r\n* Skip device_adjacent_difference hipGraph test on Windows for Navi3x (#490)\r\n\r\n* Currently, the LargeIndices hipGraphs test for gfx1030 on Windows is skipped\r\n* This change causes this test case to also get skiped on gfx1100, gfx1101, gfx1102 on Windows\r\n* The reason this test fails on Navi on Windows appears to be related to\r\n the check_output class (used by OutputIterator in the test).\r\n * this may be releated to using atomics inside of graphs, but further\r\n investigation is needed\r\n\r\n* Bump cryptography from 41.0.4 to 41.0.6 in /docs/.sphinx (#488)\r\n\r\nBumps [cryptography](https://github.com/pyca/cryptography) from 41.0.4 to 41.0.6.\r\n- [Changelog](https://github.com/pyca/cryptography/blob/main/CHANGELOG.rst)\r\n- [Commits](https://github.com/pyca/cryptography/compare/41.0.4...41.0.6)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: cryptography\r\n dependency-type: indirect\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>\r\n\r\n* Bump rocm-docs-core from 0.27.0 to 0.30.0 in /docs/.sphinx (#489)\r\n\r\nBumps [rocm-docs-core](https://github.com/RadeonOpenCompute/rocm-docs-core) from 0.27.0 to 0.30.0.\r\n- [Release notes](https://github.com/RadeonOpenCompute/rocm-docs-core/releases)\r\n- [Changelog](https://github.com/RadeonOpenCompute/rocm-docs-core/blob/develop/CHANGELOG.md)\r\n- [Commits](https://github.com/RadeonOpenCompute/rocm-docs-core/compare/v0.27.0...v0.30.0)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: rocm-docs-core\r\n dependency-type: direct:production\r\n update-type: version-update:semver-minor\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>\r\n\r\n* Lookback state fixes (#491)\r\n\r\n* Do not call fence in the wait loop\r\n\r\n* Use __hip_atomic_load/store instead of atomicExch/atomicAdd\r\n\r\natomicExch is compiled to global_atomic_swap even when the results is not\r\nused.\r\n\r\n* Use faster fences in lookback algorithms on gfx94*\r\n\r\nThis version is specific for devices with slow __threadfence (\"agent\" fence which does\r\nL2 cache flushing and invalidation).\r\nFences with \"workgroup\" scope are used instead to ensure ordering only but not coherence,\r\nthey do not flush and invalidate cache.\r\nGlobal coherence of prefixes_*_values is ensured by atomic_load/atomic_store that bypass\r\ncache.\r\n\r\n* Rename ROCPRIM_DETAIL_LOOKBACK_SCAN_STATE_WITHOUT_SLOW_FENCES\r\n\r\nfrom ROCPRIM_LOOKBACK_WITHOUT_SLOW_FENCES.\r\nThis is more verbose to communicates that it is implementation detail\r\n\r\nIt uses 0 and 1 instead of the presence of the macro now, and won't\r\nbe overriden if set by a developer on the command line.\r\n\r\n* Add WITHOUT_SLOW_FENCES version to lookback_scan_state::get_complete_value\r\n\r\n* refactor: lookback_scan_state WITHOUT_SLOW_FENCES misc changes\r\n\r\n- use sizeof(variable)\r\n- use auto* and const auto* instead of just auto\r\n- use void* instead of char* to avoid yet another cast\r\n- make the atomic order fence a separate function and add docs &\r\n warning\r\n\r\n* fix: Restore removed interfaces of lookback_scan_state\r\n\r\nEven though these are in the detail namespace and as such explicitly\r\nnot meant for usage by users, some projects did start depending on them.\r\n\r\nThe interfaces for these are slightly broken and rocPRIM developers\r\ndiscourage any users from using them (or the newer interfaces for that\r\nmatter) because they are implementation details. No further guarantees\r\nare provided for these APIs.\r\n\r\nIn the future a public interface is planned for lookback_scan_state\r\nas we have recognized that this is a useful primitive, and it's\r\nunreasonable to expect users to implement for themselves.\r\n\r\n* refactor: rename __builtin_amdgcn_fence as atomic_fence_acquire_order_only\r\n\r\n---------\r\n\r\nCo-authored-by: Anton Gorenko \r\n\r\n* Bump rocm-docs-core from 0.30.0 to 0.30.3 in /docs/.sphinx (#496)\r\n\r\nBumps [rocm-docs-core](https://github.com/RadeonOpenCompute/rocm-docs-core) from 0.30.0 to 0.30.3.\r\n- [Release notes](https://github.com/RadeonOpenCompute/rocm-docs-core/releases)\r\n- [Changelog](https://github.com/RadeonOpenCompute/rocm-docs-core/blob/develop/CHANGELOG.md)\r\n- [Commits](https://github.com/RadeonOpenCompute/rocm-docs-core/compare/v0.30.0...v0.30.3)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: rocm-docs-core\r\n dependency-type: direct:production\r\n update-type: version-update:semver-patch\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>\r\n\r\n* 6.0 final mergeback to develop (#498)\r\n\r\n* Fix cpp-check reported issues\r\n\r\nFixed a number of issues that static the analysis tool picked up:\r\n - Made some functions const since they don't modify member state\r\n - Made some parameters const, since they're never modified\r\n - Made some functions static (for performance), since they don't require access to the class instance\r\n - Fixes for several benchmark/test functions\r\n - Removed unused variable declarations\r\n - Added missing input data transfer from host to device\r\n - Added default return value for one overlooked execution path\r\n - Added some member variables to constructor initializer list\r\n - Added override keyword in several places\r\n - Fixed up item placeholders in some printf statements\r\n\r\n* Separate gfx942 specific code (#468)\r\n\r\nCo-authored-by: Stanley Tsang \r\n\r\n* Fix cpp-check reported issues\r\n* Removed host to data transfer from memcpy benchmark.\r\nSince this benchmark only tests memcpy performance between device buffers,\r\nwe don't really need to copy data into these from the host.\r\n\r\n* Remove Unnecessary Newline & Re-trigger Performance Checks\r\n\r\n* Update comment for extra clarification\r\n\r\n* Updated comment in memcpy benchmark to make the purpose of the code a little clearer.\r\n\r\n* Update googlebenchmark version (#477)\r\n\r\n* 6.0 cherry pick for changelog and version update (#483)\r\n\r\n* Fix changelog for 6.0\r\n\r\n* Fix version\r\n\r\n* Fix up changelog\r\n\r\n---------\r\n\r\nCo-authored-by: Wayne Franz \r\nCo-authored-by: Eiden Yoshida <47196116+eidenyoshida@users.noreply.github.com>\r\nCo-authored-by: Lauren Wrubleski \r\n\r\n* Add CODEOWNERS file (#504)\r\n\r\n* Standardize documentation for ReadtheDocs (#497)\r\n\r\n* Bump jinja2 from 3.1.2 to 3.1.3 in /docs/sphinx (#506)\r\n\r\nBumps [jinja2](https://github.com/pallets/jinja) from 3.1.2 to 3.1.3.\r\n- [Release notes](https://github.com/pallets/jinja/releases)\r\n- [Changelog](https://github.com/pallets/jinja/blob/main/CHANGES.rst)\r\n- [Commits](https://github.com/pallets/jinja/compare/3.1.2...3.1.3)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: jinja2\r\n dependency-type: indirect\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>\r\n\r\n* Bump gitpython from 3.1.37 to 3.1.41 in /docs/sphinx (#508)\r\n\r\nBumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.\r\n- [Release notes](https://github.com/gitpython-developers/GitPython/releases)\r\n- [Changelog](https://github.com/gitpython-developers/GitPython/blob/main/CHANGES)\r\n- [Commits](https://github.com/gitpython-developers/GitPython/compare/3.1.37...3.1.41)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: gitpython\r\n dependency-type: indirect\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>\r\n\r\n* Bump rocm-docs-core from 0.30.3 to 0.31.0 in /docs/sphinx (#507)\r\n\r\nBumps [rocm-docs-core](https://github.com/RadeonOpenCompute/rocm-docs-core) from 0.30.3 to 0.31.0.\r\n- [Release notes](https://github.com/RadeonOpenCompute/rocm-docs-core/releases)\r\n- [Changelog](https://github.com/RadeonOpenCompute/rocm-docs-core/blob/develop/CHANGELOG.md)\r\n- [Commits](https://github.com/RadeonOpenCompute/rocm-docs-core/compare/v0.30.3...v0.31.0)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: rocm-docs-core\r\n dependency-type: direct:production\r\n update-type: version-update:semver-minor\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>\r\n\r\n* Update links in README.md\r\n\r\n- Update links to other ROCm repositories.\r\n\r\n* Update package version\r\n\r\n---------\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: Nara \r\nCo-authored-by: Ivan Siutsou \r\nCo-authored-by: Bence Parajdi \r\nCo-authored-by: Bálint Soproni \r\nCo-authored-by: Gergely Meszaros \r\nCo-authored-by: Beatriz Navidad Vilches \r\nCo-authored-by: Mátyás Aradi \r\nCo-authored-by: Robin Voetter \r\nCo-authored-by: Wayne Franz \r\nCo-authored-by: Lisa \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>\r\nCo-authored-by: Anton Gorenko \r\nCo-authored-by: Eiden Yoshida <47196116+eidenyoshida@users.noreply.github.com>\r\nCo-authored-by: Lauren Wrubleski \r\nCo-authored-by: Sam Wu \r\nCo-authored-by: David Galiffi \r\n\r\n* Temporarily skip hipGraph version of LargeIndices test\r\n\r\nCurrently there appear to be some problems calling the atomicAdd function within hipGraphs on some architectures. The atomicAdd function is called within the device_adjacent_difference algorithm's LargeIndices test (check_output class) to increment a counter value. As a result, the hipGraph version of the test fails when called on affected architectures.\r\n\r\nThis change temporarily skips the hipGraph version of the test. We can re-enable it when the root cause has been addressed.\r\n\r\n---------\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: Stanley Tsang \r\nCo-authored-by: Nara \r\nCo-authored-by: Ivan Siutsou \r\nCo-authored-by: Bence Parajdi \r\nCo-authored-by: Bálint Soproni \r\nCo-authored-by: Gergely Meszaros \r\nCo-authored-by: Beatriz Navidad Vilches \r\nCo-authored-by: Mátyás Aradi \r\nCo-authored-by: Robin Voetter \r\nCo-authored-by: Wayne Franz \r\nCo-authored-by: Lisa \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>\r\nCo-authored-by: Anton Gorenko \r\nCo-authored-by: Eiden Yoshida <47196116+eidenyoshida@users.noreply.github.com>\r\nCo-authored-by: Lauren Wrubleski \r\nCo-authored-by: Sam Wu \r\nCo-authored-by: David Galiffi ","shortMessageHtmlLink":"Merge back change from ROCm 6.1 (ROCm#532)"}},{"before":"788a99a0da716cc29451a35ab0667e5c5dee4827","after":"3e31a1513a8ed3f7306bd02f0e45bf0e26defc02","ref":"refs/heads/develop","pushedAt":"2024-02-13T14:27:20.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Update exclusion list for CI testing (#519)\n\n* Update exclusion list for CI testing; device_scan due to navi3x compiler bug\r\n\r\n* Fix typo","shortMessageHtmlLink":"Update exclusion list for CI testing (ROCm#519)"}},{"before":null,"after":"4eb64fefbbee031f3db4c4591656b10fe10ed3b0","ref":"refs/heads/disable_hipgraph_dev_adj_diff_test","pushedAt":"2024-02-08T17:27:41.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Temporarily skip hipGraph version of LargeIndices test\n\nCurrently there appear to be some problems calling the atomicAdd function within hipGraphs on some architectures. The atomicAdd function is called within the device_adjacent_difference algorithm's LargeIndices test (check_output class) to increment a counter value. As a result, the hipGraph version of the test fails when called on affected architectures.\n\nThis change temporarily skips the hipGraph version of the test. We can re-enable it when the root cause has been addressed.","shortMessageHtmlLink":"Temporarily skip hipGraph version of LargeIndices test"}},{"before":"6bfe1bb2e4610a32b2a225c8230479566ad753a4","after":null,"ref":"refs/heads/disable_hipgraph_dev_adj_diff_test","pushedAt":"2024-02-08T17:27:15.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"}},{"before":"ea3430ceb14415fac638753b80bd6cb276a6b079","after":null,"ref":"refs/heads/fix_device_adjacent_difference_test","pushedAt":"2024-02-08T17:09:54.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"}},{"before":null,"after":"6bfe1bb2e4610a32b2a225c8230479566ad753a4","ref":"refs/heads/disable_hipgraph_dev_adj_diff_test","pushedAt":"2024-02-08T14:34:08.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Temporarily skip hipGraph version of LargeIndices test in test_device_adjacent_difference\n\nCurrently there appear to be some problems calling the atomicAdd function within hipGraphs on some architectures.\nThe atomicAdd function is called within the device_adjacent_difference algorithm's LargeIndices test\n(check_output class) to increment a counter value. As a result, the hipGraph version of the test fails\nwhen called on affected architectures.\n\nThis change temporarily skips the hipGraph version of the test. We can re-enable it when the root cause\nhas been addressed.","shortMessageHtmlLink":"Temporarily skip hipGraph version of LargeIndices test in test_device…"}},{"before":"ce39f3027ee7c568d02a277e5373511a054db1a3","after":"788a99a0da716cc29451a35ab0667e5c5dee4827","ref":"refs/heads/develop","pushedAt":"2024-02-08T13:40:40.000Z","pushType":"push","commitsCount":13,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Bump rocm-docs-core from 0.33.1 to 0.33.2 in /docs/sphinx (#523)\n\nBumps [rocm-docs-core](https://github.com/RadeonOpenCompute/rocm-docs-core) from 0.33.1 to 0.33.2.\r\n- [Release notes](https://github.com/RadeonOpenCompute/rocm-docs-core/releases)\r\n- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/develop/CHANGELOG.md)\r\n- [Commits](https://github.com/RadeonOpenCompute/rocm-docs-core/compare/v0.33.1...v0.33.2)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: rocm-docs-core\r\n dependency-type: direct:production\r\n update-type: version-update:semver-patch\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>","shortMessageHtmlLink":"Bump rocm-docs-core from 0.33.1 to 0.33.2 in /docs/sphinx (ROCm#523)"}},{"before":"57c30d8d1feab4f8204fd2d9c751b468bfec8241","after":"ce39f3027ee7c568d02a277e5373511a054db1a3","ref":"refs/heads/develop","pushedAt":"2024-01-08T13:51:06.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"umfranzw","name":"Wayne Franz","path":"/umfranzw","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/212066?s=80&v=4"},"commit":{"message":"Bump rocm-docs-core from 0.30.0 to 0.30.3 in /docs/.sphinx (#496)\n\nBumps [rocm-docs-core](https://github.com/RadeonOpenCompute/rocm-docs-core) from 0.30.0 to 0.30.3.\r\n- [Release notes](https://github.com/RadeonOpenCompute/rocm-docs-core/releases)\r\n- [Changelog](https://github.com/RadeonOpenCompute/rocm-docs-core/blob/develop/CHANGELOG.md)\r\n- [Commits](https://github.com/RadeonOpenCompute/rocm-docs-core/compare/v0.30.0...v0.30.3)\r\n\r\n---\r\nupdated-dependencies:\r\n- dependency-name: rocm-docs-core\r\n dependency-type: direct:production\r\n update-type: version-update:semver-patch\r\n...\r\n\r\nSigned-off-by: dependabot[bot] \r\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>","shortMessageHtmlLink":"Bump rocm-docs-core from 0.30.0 to 0.30.3 in /docs/.sphinx (ROCm#496)"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAEeZ44kAA","startCursor":null,"endCursor":null}},"title":"Activity · umfranzw/rocPRIM"}