{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":58414589,"defaultBranch":"main","name":"oneDNN","ownerLogin":"oneapi-src","currentUserCanPush":false,"isFork":false,"isEmpty":false,"createdAt":"2016-05-09T23:26:42.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/60144784?v=4","public":true,"private":false,"isOrgOwned":true},"refInfo":{"name":"","listCacheKey":"v0:1717453903.0","currentOid":""},"activityList":{"items":[{"before":"84adf42c753ab5ce92dd62a7d725fad097732db2","after":"f618fc00c6e4db52ae12f9ae6cccb71c16dfce25","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-08T01:10:52.000Z","pushType":"push","commitsCount":3,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"graph: backend: dnnl: fix sdpa decomp kernel to handle unequal q/k/v seq len and fix last reorder's memory address if in-place","shortMessageHtmlLink":"graph: backend: dnnl: fix sdpa decomp kernel to handle unequal q/k/v …"}},{"before":"1f866c11b2e9224e846ba78507dd725b3f40b89b","after":"628146e606d6955fea2196ac55a4b9fefb339722","ref":"refs/heads/main","pushedAt":"2024-06-08T01:05:49.000Z","pushType":"push","commitsCount":3,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"cpu: x64: update impl verbose","shortMessageHtmlLink":"cpu: x64: update impl verbose"}},{"before":"34a1b5ef04f00fabd99096a46d2bade896400c12","after":"84adf42c753ab5ce92dd62a7d725fad097732db2","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-07T22:40:38.000Z","pushType":"pr_merge","commitsCount":2,"pusher":{"login":"vpirogov","name":"Vadim Pirogov","path":"/vpirogov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19210288?s=80&v=4"},"commit":{"message":"Backport cpu: aarch64: matmul: fuse sum post op in acl matmul\n\nFuse the sum post op in acl matmul by setting the accumulate flag to\ntrue in arm_compute::GEMMInfo. This speeds up the post op and saves\nallocating a temporary dst sized tensor.\n\nWe also added `_for_sum` to `use_dst_acc` flag to stop it being confused\nwith the `dst_acc` used for transposing.\n\nChange the way we deal with fused eltwise (as well as the new sum) to\nfix segfaults when binary ops followed fused ops.\n\nCo-authored-by: Milos Puzovic <milos.puzovic@arm.com>\nCo-authored-by: Jonathan Deakin <jonathan.deakin@arm.com>","shortMessageHtmlLink":"Backport cpu: aarch64: matmul: fuse sum post op in acl matmul"}},{"before":"9fa7f1c3d43a58f17a1fcca8511fe1a82b11d4dd","after":"34a1b5ef04f00fabd99096a46d2bade896400c12","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-07T22:38:47.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"vpirogov","name":"Vadim Pirogov","path":"/vpirogov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19210288?s=80&v=4"},"commit":{"message":"Backport cpu: aarch64: conv: Update direct vs indirect conv heuristics\n\n- Remove fall through to direct conv for low thread counts\n- Do not fall though to direct conv for small convolutions\n  when the datatype is BF16.","shortMessageHtmlLink":"Backport cpu: aarch64: conv: Update direct vs indirect conv heuristics"}},{"before":"19272d1520990fd82bab71fea0f95ea5770f7650","after":"9fa7f1c3d43a58f17a1fcca8511fe1a82b11d4dd","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-07T22:38:27.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"vpirogov","name":"Vadim Pirogov","path":"/vpirogov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19210288?s=80&v=4"},"commit":{"message":"Backport: src: cpu: aarch64: reorder: use jit uni reorder for bf16\n\nadd restriction for bf16 use when beta == 0.f as it is unsupported\n\nuse jit uni reorder for aarch64 src=bf16 and dst=bf16\n\nChange-Id: I330a51fa0aa0e0a2ffd8e08051563bbb49652ae3","shortMessageHtmlLink":"Backport: src: cpu: aarch64: reorder: use jit uni reorder for bf16"}},{"before":"91f073cb94329e7ad3bdf1b2bec1e18504602382","after":"19272d1520990fd82bab71fea0f95ea5770f7650","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-07T22:38:19.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"vpirogov","name":"Vadim Pirogov","path":"/vpirogov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19210288?s=80&v=4"},"commit":{"message":"Backport: src: cpu: aarch64: conv: Use acl_indirect_gemm for bf16 convolutions\n\nperformance improvements:\n\nTotal benchdnn tests: 57\nMin: 15x\nAverage: 131x\nMax: 320x\n\nChange-Id: I6266fa72491a03dd96f0c6d51334d7ab376a0e26","shortMessageHtmlLink":"Backport: src: cpu: aarch64: conv: Use acl_indirect_gemm for bf16 con…"}},{"before":"921008c09096051d426aaf3a27db0991f9a7fe7f","after":"91f073cb94329e7ad3bdf1b2bec1e18504602382","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-07T22:37:42.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"vpirogov","name":"Vadim Pirogov","path":"/vpirogov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19210288?s=80&v=4"},"commit":{"message":"cpu: x64: update impl verbose","shortMessageHtmlLink":"cpu: x64: update impl verbose"}},{"before":"777d028c737680c9ba74f3de4211785f78f49b5a","after":"921008c09096051d426aaf3a27db0991f9a7fe7f","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-07T19:11:16.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"x64: reorder: fix modulo by zero simd_w","shortMessageHtmlLink":"x64: reorder: fix modulo by zero simd_w"}},{"before":"45db71b77ea4b65098675b2709436af85c3c4ba5","after":"cecafbdb110f48e7a4952b99aab9215bd8493339","ref":"refs/heads/gh-pages","pushedAt":"2024-06-07T18:28:49.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"gh-pages: update documentation","shortMessageHtmlLink":"gh-pages: update documentation"}},{"before":"d7a52de12c7428d37cc9a1faf35ebd94a9b5b5fa","after":"aa9feb63b395a2bc27aaeb693d38c9f4a826044a","ref":"refs/heads/amanerik/rfcs/coo-sparse-encoding","pushedAt":"2024-06-07T18:18:53.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"avmanerikar","name":"Ankit Manerikar","path":"/avmanerikar","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/106178660?s=80&v=4"},"commit":{"message":"rfcs: add proposal on coo sparse encoding","shortMessageHtmlLink":"rfcs: add proposal on coo sparse encoding"}},{"before":"3b8db6465aa993b9bacb2b1654dce24b823a9d83","after":"d7a52de12c7428d37cc9a1faf35ebd94a9b5b5fa","ref":"refs/heads/amanerik/rfcs/coo-sparse-encoding","pushedAt":"2024-06-07T18:14:39.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"avmanerikar","name":"Ankit Manerikar","path":"/avmanerikar","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/106178660?s=80&v=4"},"commit":{"message":"rfcs: add proposal on coo sparse encoding","shortMessageHtmlLink":"rfcs: add proposal on coo sparse encoding"}},{"before":"612ccf3e2d9963efbe58fad6b6338bcf04dad57f","after":"1f866c11b2e9224e846ba78507dd725b3f40b89b","ref":"refs/heads/main","pushedAt":"2024-06-07T18:07:11.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"densamoilov","name":"Denis Samoilov","path":"/densamoilov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/11770112?s=80&v=4"},"commit":{"message":"gpu: nvidia: fix stream init\n\nThe `!` seems to have been dropped by mistake during refactoring.","shortMessageHtmlLink":"gpu: nvidia: fix stream init"}},{"before":"cd1c2c761884ff021332a26ef5c83f68d9be2b6b","after":"777d028c737680c9ba74f3de4211785f78f49b5a","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-07T11:59:19.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"gtest: graph: unit: release the strm pointer during code quit, fix memory leaky issue","shortMessageHtmlLink":"gtest: graph: unit: release the strm pointer during code quit, fix me…"}},{"before":"525e7b7791e5fbc973e294d79b8e57b395f7007f","after":"cd1c2c761884ff021332a26ef5c83f68d9be2b6b","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-07T01:16:43.000Z","pushType":"push","commitsCount":6,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"src: gpu: intel: fix coverity issues for april and may 2024","shortMessageHtmlLink":"src: gpu: intel: fix coverity issues for april and may 2024"}},{"before":"a3f716cc5b103657f219d2c5975678d910955f99","after":"612ccf3e2d9963efbe58fad6b6338bcf04dad57f","ref":"refs/heads/main","pushedAt":"2024-06-07T01:11:40.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"gtest: graph: unit: release the strm pointer during code quit, fix memory leaky issue","shortMessageHtmlLink":"gtest: graph: unit: release the strm pointer during code quit, fix me…"}},{"before":"ec3e754bba864665f0dbaf684b0ce2552ebe58b6","after":"a3f716cc5b103657f219d2c5975678d910955f99","ref":"refs/heads/main","pushedAt":"2024-06-06T19:17:59.000Z","pushType":"push","commitsCount":57,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"x64: reorder: fix modulo by zero simd_w","shortMessageHtmlLink":"x64: reorder: fix modulo by zero simd_w"}},{"before":"e5f0384f4c75b2a3623f5dffdab54696c5020f19","after":"ec3e754bba864665f0dbaf684b0ce2552ebe58b6","ref":"refs/heads/main","pushedAt":"2024-06-06T16:46:43.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"vpirogov","name":"Vadim Pirogov","path":"/vpirogov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19210288?s=80&v=4"},"commit":{"message":"cpu: fix coverity: avoid division by zero\n\nSet defaults to avoid division by zero in brgemm matmul.","shortMessageHtmlLink":"cpu: fix coverity: avoid division by zero"}},{"before":"134256fc7f4a486b728583e10e65e36b943e1556","after":"525e7b7791e5fbc973e294d79b8e57b395f7007f","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-06T16:46:04.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"vpirogov","name":"Vadim Pirogov","path":"/vpirogov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19210288?s=80&v=4"},"commit":{"message":"graph: backend: compiler: ops: support transpose with no permutation axis","shortMessageHtmlLink":"graph: backend: compiler: ops: support transpose with no permutation …"}},{"before":"1d0e97f4efcfa936c5af16b0e22446f05483c56c","after":"e5f0384f4c75b2a3623f5dffdab54696c5020f19","ref":"refs/heads/main","pushedAt":"2024-06-06T07:13:05.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"x64: brgemm: fix coverity","shortMessageHtmlLink":"x64: brgemm: fix coverity"}},{"before":"85f3652cd64762a642ba93a47019e123cd21221f","after":"134256fc7f4a486b728583e10e65e36b943e1556","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-06T01:11:57.000Z","pushType":"push","commitsCount":4,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"doc: verbose: fixes for verbose message table","shortMessageHtmlLink":"doc: verbose: fixes for verbose message table"}},{"before":"732e3dd34802ef358bba76179f5276fac7b0b83a","after":"1d0e97f4efcfa936c5af16b0e22446f05483c56c","ref":"refs/heads/main","pushedAt":"2024-06-06T01:06:39.000Z","pushType":"push","commitsCount":3,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"doc: verbose: fixes for verbose message table","shortMessageHtmlLink":"doc: verbose: fixes for verbose message table"}},{"before":"9c1dfd9baee33d054d557bd065097bcc252a4718","after":"85f3652cd64762a642ba93a47019e123cd21221f","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-05T19:16:27.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"gpu: ocl: gemm: fixup ref gemm coverity warning; add check for batch runtime dim","shortMessageHtmlLink":"gpu: ocl: gemm: fixup ref gemm coverity warning; add check for batch …"}},{"before":"eb013e386e47d021b2ab169500ce6ab1147142e1","after":"732e3dd34802ef358bba76179f5276fac7b0b83a","ref":"refs/heads/main","pushedAt":"2024-06-05T19:11:22.000Z","pushType":"push","commitsCount":13,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"src: gpu: intel: fix coverity issues for april and may 2024","shortMessageHtmlLink":"src: gpu: intel: fix coverity issues for april and may 2024"}},{"before":"800690a0121bbc03834815bc04fcb51e57154b24","after":"eb013e386e47d021b2ab169500ce6ab1147142e1","ref":"refs/heads/main","pushedAt":"2024-06-05T13:06:58.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"graph: backend: compiler: ops: support transpose with no permutation axis","shortMessageHtmlLink":"graph: backend: compiler: ops: support transpose with no permutation …"}},{"before":"5db5d72caf453b6df30ff795771608f75bb1a4ad","after":"d998a1732b577c03ff649795db39d160a1229cd2","ref":"refs/heads/rfcs","pushedAt":"2024-06-05T07:41:33.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"TaoLv","name":"Tao Lv","path":"/TaoLv","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/22437510?s=80&v=4"},"commit":{"message":"rfcs: graph: propose ocl runtime support","shortMessageHtmlLink":"rfcs: graph: propose ocl runtime support"}},{"before":"c0816f95d4b4448ab161d28acce3c1d85b657c8b","after":"800690a0121bbc03834815bc04fcb51e57154b24","ref":"refs/heads/main","pushedAt":"2024-06-05T07:03:56.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"graph: backend: compiler: align graph compiler mha pattern with sdpa","shortMessageHtmlLink":"graph: backend: compiler: align graph compiler mha pattern with sdpa"}},{"before":"825898ee338cb4154c931a5c083e5350cb535e8c","after":"9c1dfd9baee33d054d557bd065097bcc252a4718","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-05T01:11:38.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"gpu: intel: ocl: reorder: avoid extraneous copies","shortMessageHtmlLink":"gpu: intel: ocl: reorder: avoid extraneous copies"}},{"before":"e6cd123e00b618c5c3d1947f255a445064c81f60","after":"c0816f95d4b4448ab161d28acce3c1d85b657c8b","ref":"refs/heads/main","pushedAt":"2024-06-05T01:06:36.000Z","pushType":"push","commitsCount":3,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"gpu: intel: ocl: rnn: initialized ref_rnn_common_t member variables","shortMessageHtmlLink":"gpu: intel: ocl: rnn: initialized ref_rnn_common_t member variables"}},{"before":"8a26446d6bc3383779773c5c1365945ca2562c2d","after":"e6cd123e00b618c5c3d1947f255a445064c81f60","ref":"refs/heads/main","pushedAt":"2024-06-04T20:18:13.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"vpirogov","name":"Vadim Pirogov","path":"/vpirogov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19210288?s=80&v=4"},"commit":{"message":"github: workflows: removed RFC notifications from main branch\n\nRFC PRs are targeting rfcs branch, not main","shortMessageHtmlLink":"github: workflows: removed RFC notifications from main branch"}},{"before":"2383e0c4cefab9362c44b1142cecc45aec5505d3","after":"825898ee338cb4154c931a5c083e5350cb535e8c","ref":"refs/heads/rls-v3.5","pushedAt":"2024-06-04T19:11:03.000Z","pushType":"push","commitsCount":5,"pusher":{"login":"perflibs","name":null,"path":"/perflibs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73841347?s=80&v=4"},"commit":{"message":"cpu: convolution: fix for verbose ctest segfault for 1x1 conv","shortMessageHtmlLink":"cpu: convolution: fix for verbose ctest segfault for 1x1 conv"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAEX7ImKQA","startCursor":null,"endCursor":null}},"title":"Activity · oneapi-src/oneDNN"}