{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":215874442,"defaultBranch":"master","name":"pytorch","ownerLogin":"EscapeZero","currentUserCanPush":false,"isFork":true,"isEmpty":false,"createdAt":"2019-10-17T19:53:11.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/16230210?v=4","public":true,"private":false,"isOrgOwned":false},"refInfo":{"name":"","listCacheKey":"v0:1715204795.0","currentOid":""},"activityList":{"items":[{"before":null,"after":"f9c89aacc2b99281be434df6ab558f39f27e0ae1","ref":"refs/heads/export-D57087985","pushedAt":"2024-05-08T21:46:35.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"EscapeZero","name":"Michael Ranieri","path":"/EscapeZero","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/16230210?s=80&v=4"},"commit":{"message":"[pytorch] fix blasLt on windows\n\nSummary:\nIt seems like required functions are not available due to `_MSC_VER` guard. Does anyone have more context why this functionality has been disabled for windows?\n\nI'm also unsure how this currently compiles in OSS land on windows, as there doesn't seem to be any preprocessor protection around `scaled_gemm` getting pulled in.\n\nTest Plan:\nFix compilation errors like this\n```\nC:\\open\\fbsource\\xplat\\caffe2\\aten\\src\\ATen\\cuda\\tunable\\TunableGemm.h(74): error C2039: 'scaled_gemm': is not a member of 'at::cuda::blas'\nC:\\open\\fbsource\\xplat\\caffe2\\aten\\src\\ATen\\cuda\\CUDABlas.h(19): note: see declaration of 'at::cuda::blas'\nC:\\open\\fbsource\\xplat\\caffe2\\aten\\src\\ATen\\cuda\\tunable\\TunableGemm.h(74): note: the template instantiation context (the oldest one first) is\nC:\\open\\fbsource\\xplat\\caffe2\\aten\\src\\ATen\\cuda\\tunable\\TunableGemm.h(71): note: while compiling class template 'at::cuda::tunable::DefaultScaledGemmOp'\nAction failed: fbsource//xplat/caffe2:ATen_cuda_lib_ovrsource (cxx_compile aten/src/ATen/native/cuda/Blas.cpp)\n```\n\nDifferential Revision: D57087985","shortMessageHtmlLink":"[pytorch] fix blasLt on windows"}},{"before":null,"after":"1329f929cc4f3292e27c27641a4f1f8cec6f413d","ref":"refs/heads/export-D54699849","pushedAt":"2024-03-11T19:30:17.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"EscapeZero","name":"Michael Ranieri","path":"/EscapeZero","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/16230210?s=80&v=4"},"commit":{"message":"avoid aten dispatch shadowing type with variable\n\nSummary:\n`DECLARE_DISPATCH` is shadowing the variable data with the data type:\n`extern TORCH_API struct name name` -> `extern TORCH_API struct gemm_stub gemm_stub` for instance.\nThis is probably dangerous behavior to rely on, as the compiler needs to always resolve to type and/or data based on context. Previous macro fails with VS2022.\n\nTest Plan: `buck2 build arvr/mode/win/vs2022/cpp20/opt //xplat/caffe2:aten_pow_ovrsource`\n\nDifferential Revision: D54699849","shortMessageHtmlLink":"avoid aten dispatch shadowing type with variable"}}],"hasNextPage":false,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAERREcFwA","startCursor":null,"endCursor":null}},"title":"Activity · EscapeZero/pytorch"}