{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":731160574,"defaultBranch":"main","name":"llvm-project","ownerLogin":"akuegel","currentUserCanPush":false,"isFork":true,"isEmpty":false,"createdAt":"2023-12-13T13:42:00.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/14309772?v=4","public":true,"private":false,"isOrgOwned":false},"refInfo":{"name":"","listCacheKey":"v0:1716873912.0","currentOid":""},"activityList":{"items":[{"before":"290fad3c7d80481abaf2c51ead9143e596dcb550","after":null,"ref":"refs/heads/bazel","pushedAt":"2024-05-28T05:25:12.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"}},{"before":"f1d13bbd6662969ad9e57a7938967217602636c5","after":"290fad3c7d80481abaf2c51ead9143e596dcb550","ref":"refs/heads/bazel","pushedAt":"2024-05-27T09:56:20.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[mlir][Bazel] Add a comment and move deps list into an extra line.\n\nThis is a non-functional change in OSS, but makes it easier for\ndownstream users to run transforms on this target.","shortMessageHtmlLink":"[mlir][Bazel] Add a comment and move deps list into an extra line."}},{"before":null,"after":"f1d13bbd6662969ad9e57a7938967217602636c5","ref":"refs/heads/bazel","pushedAt":"2024-05-27T09:50:53.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[flang] add FIR to FIR pass to lower assumed-rank operations (#93344)\n\nAdd pass to lower assumed-rank operations. The current patch adds\r\ncodegen for fir.rebox_assumed_rank. It will be the pass lowering\r\nfir.select_rank.\r\n \r\nfir.rebox_assumed_rank is lowered to a call to CopyAndUpdateDescriptor\r\nruntime API.\r\n \r\nNote that the lowering ends-up allocating two new descriptors at the\r\nLLVM level (one alloca created by the pass for the CopyAndUpdateDescriptor\r\nresult descriptor argument, the second one is created by the fir.load\r\nof the result descriptor in codegen).\r\nLLVM is currently unable to properly optimize and merge those allocas.\r\nThe \"nocapture\" attribute added to CopyAndUpdateDescriptor arguments\r\ngives part of the information to LLVM, but the fir.load codegen of\r\ndescriptors must be updated to use llvm.memcpy instead of\r\nllvm.load+store to allow LLVM to optimize it. This will be done in later patch.","shortMessageHtmlLink":"[flang] add FIR to FIR pass to lower assumed-rank operations (llvm#93344"}},{"before":"bafbe39778a972d0f2869980de22fb00c03a6a35","after":"f1d13bbd6662969ad9e57a7938967217602636c5","ref":"refs/heads/main","pushedAt":"2024-05-27T09:50:27.000Z","pushType":"push","commitsCount":1796,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[flang] add FIR to FIR pass to lower assumed-rank operations (#93344)\n\nAdd pass to lower assumed-rank operations. The current patch adds\r\ncodegen for fir.rebox_assumed_rank. It will be the pass lowering\r\nfir.select_rank.\r\n \r\nfir.rebox_assumed_rank is lowered to a call to CopyAndUpdateDescriptor\r\nruntime API.\r\n \r\nNote that the lowering ends-up allocating two new descriptors at the\r\nLLVM level (one alloca created by the pass for the CopyAndUpdateDescriptor\r\nresult descriptor argument, the second one is created by the fir.load\r\nof the result descriptor in codegen).\r\nLLVM is currently unable to properly optimize and merge those allocas.\r\nThe \"nocapture\" attribute added to CopyAndUpdateDescriptor arguments\r\ngives part of the information to LLVM, but the fir.load codegen of\r\ndescriptors must be updated to use llvm.memcpy instead of\r\nllvm.load+store to allow LLVM to optimize it. This will be done in later patch.","shortMessageHtmlLink":"[flang] add FIR to FIR pass to lower assumed-rank operations (llvm#93344"}},{"before":"fdc8c5440041ac53726d0b3587762ceeb8cbbb4f","after":"bafbe39778a972d0f2869980de22fb00c03a6a35","ref":"refs/heads/main","pushedAt":"2024-05-08T08:22:35.000Z","pushType":"push","commitsCount":1654,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[NVPTX] Add support for atomic add for bf16 type (#89586)\n\natom.add.noftz.bf16 is supported since SM 9.0 and PTX 7.8","shortMessageHtmlLink":"[NVPTX] Add support for atomic add for bf16 type (llvm#89586)"}},{"before":"aa97489ab36ad0adc396982e03e1b57d948ff582","after":null,"ref":"refs/heads/atomics","pushedAt":"2024-05-08T08:22:15.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"}},{"before":"8366433e06597316f4dd7bbe610d9caa22f83855","after":"aa97489ab36ad0adc396982e03e1b57d948ff582","ref":"refs/heads/atomics","pushedAt":"2024-05-08T06:45:09.000Z","pushType":"push","commitsCount":1651,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"Merge branch 'llvm:main' into atomics","shortMessageHtmlLink":"Merge branch 'llvm:main' into atomics"}},{"before":"fdc8c5440041ac53726d0b3587762ceeb8cbbb4f","after":"8366433e06597316f4dd7bbe610d9caa22f83855","ref":"refs/heads/atomics","pushedAt":"2024-04-22T10:20:40.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[NVPTX] Add support for atomic add for bf16 type\n\natom.add.noftz.bf16 is supported since SM 9.0 and PTX 7.8","shortMessageHtmlLink":"[NVPTX] Add support for atomic add for bf16 type"}},{"before":null,"after":"fdc8c5440041ac53726d0b3587762ceeb8cbbb4f","ref":"refs/heads/atomics","pushedAt":"2024-04-22T10:15:02.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[flang][driver] Avoid mentions of Clang in Flang's command line reference. (#88932)\n\nThe help text was not updated in #87360.\r\n\r\nClang is also mentioned for the diagnostic warnings reference, which\r\nmostly applies to C/C++/Obj-C, not Fortran. #81726 already tried to fix\r\nthis, and I don't know a better solution.","shortMessageHtmlLink":"[flang][driver] Avoid mentions of Clang in Flang's command line refer…"}},{"before":"9b98692eedb78aa106539c36ba02944f32cae1ff","after":"fdc8c5440041ac53726d0b3587762ceeb8cbbb4f","ref":"refs/heads/main","pushedAt":"2024-04-22T10:14:37.000Z","pushType":"push","commitsCount":3394,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[flang][driver] Avoid mentions of Clang in Flang's command line reference. (#88932)\n\nThe help text was not updated in #87360.\r\n\r\nClang is also mentioned for the diagnostic warnings reference, which\r\nmostly applies to C/C++/Obj-C, not Fortran. #81726 already tried to fix\r\nthis, and I don't know a better solution.","shortMessageHtmlLink":"[flang][driver] Avoid mentions of Clang in Flang's command line refer…"}},{"before":"92bb4aea56d9fee02c6c98f248d767d77cef25cc","after":null,"ref":"refs/heads/fix_msvc_build","pushedAt":"2024-03-18T09:24:25.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"}},{"before":"9b98692eedb78aa106539c36ba02944f32cae1ff","after":"92bb4aea56d9fee02c6c98f248d767d77cef25cc","ref":"refs/heads/fix_msvc_build","pushedAt":"2024-03-18T08:48:24.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[mlir] Adjust split marker flag logic to fix MSVC build.\n\nWe should not try to capture the flag variable inside the callback\npassed when creating the flag variable. As a fix, we use a separate\nboolean variable to track whether the command line flag was specified.","shortMessageHtmlLink":"[mlir] Adjust split marker flag logic to fix MSVC build."}},{"before":null,"after":"9b98692eedb78aa106539c36ba02944f32cae1ff","ref":"refs/heads/fix_msvc_build","pushedAt":"2024-03-18T08:42:38.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[AMDGPU] Run LowerLDS at the end of the fullLTO pipeline (#75333)\n\nThis change allows us to use `--lto-partitions` in some cases (not at\r\nall guaranteed it works perfectly), as LDS is lowered before the module\r\nis split for parallel codegen.\r\n\r\nWe must run LowerLDS before splitting modules as it needs to see all\r\ncallers of functions with LDS to properly lower them.","shortMessageHtmlLink":"[AMDGPU] Run LowerLDS at the end of the fullLTO pipeline (llvm#75333)"}},{"before":"2582965c160486f9e3b0680f1cebc5ffdef9620c","after":"9b98692eedb78aa106539c36ba02944f32cae1ff","ref":"refs/heads/main","pushedAt":"2024-03-18T08:42:19.000Z","pushType":"push","commitsCount":364,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[AMDGPU] Run LowerLDS at the end of the fullLTO pipeline (#75333)\n\nThis change allows us to use `--lto-partitions` in some cases (not at\r\nall guaranteed it works perfectly), as LDS is lowered before the module\r\nis split for parallel codegen.\r\n\r\nWe must run LowerLDS before splitting modules as it needs to see all\r\ncallers of functions with LDS to properly lower them.","shortMessageHtmlLink":"[AMDGPU] Run LowerLDS at the end of the fullLTO pipeline (llvm#75333)"}},{"before":"d09b706e7ecadaf54ad8a1c264b88630eb35e1d8","after":null,"ref":"refs/heads/fp16","pushedAt":"2024-03-15T07:14:07.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"}},{"before":"5e73429a14b640ba89debc718038de013eb35d19","after":"d09b706e7ecadaf54ad8a1c264b88630eb35e1d8","ref":"refs/heads/fp16","pushedAt":"2024-03-15T06:50:22.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"Simplify according to review comments","shortMessageHtmlLink":"Simplify according to review comments"}},{"before":"2582965c160486f9e3b0680f1cebc5ffdef9620c","after":"5e73429a14b640ba89debc718038de013eb35d19","ref":"refs/heads/fp16","pushedAt":"2024-03-14T08:22:18.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"Fix problem with fp16 immediates","shortMessageHtmlLink":"Fix problem with fp16 immediates"}},{"before":null,"after":"2582965c160486f9e3b0680f1cebc5ffdef9620c","ref":"refs/heads/fp16","pushedAt":"2024-03-14T08:08:06.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[C++20] [Modules] [Reduced BMI] Generate the function body from implicitly instantiated class and constant variables\n\nAfter this patch, we will generate the function body from implicitly\ninstantiated class. This is important for consumers with same\ntemplate arguments. Otherwise the consumers won't see the function body.\nSince the consumers won't instantiate the templates again if they find an\ninstantiation.\n\nAlso we will generate the variable definition if the variable is\nnon-inline but known as constant. Such variables may not affect the\nABI, but they may get involved into the compile time constant computation\nin the consumer's code. So we have to generate such definitions.","shortMessageHtmlLink":"[C++20] [Modules] [Reduced BMI] Generate the function body from impli…"}},{"before":"8e0f4b943fee13afc970ca8277a8e76b9da63b96","after":"2582965c160486f9e3b0680f1cebc5ffdef9620c","ref":"refs/heads/main","pushedAt":"2024-03-14T08:07:58.000Z","pushType":"push","commitsCount":272,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[C++20] [Modules] [Reduced BMI] Generate the function body from implicitly instantiated class and constant variables\n\nAfter this patch, we will generate the function body from implicitly\ninstantiated class. This is important for consumers with same\ntemplate arguments. Otherwise the consumers won't see the function body.\nSince the consumers won't instantiate the templates again if they find an\ninstantiation.\n\nAlso we will generate the variable definition if the variable is\nnon-inline but known as constant. Such variables may not affect the\nABI, but they may get involved into the compile time constant computation\nin the consumer's code. So we have to generate such definitions.","shortMessageHtmlLink":"[C++20] [Modules] [Reduced BMI] Generate the function body from impli…"}},{"before":"6e79f77adbbd338848ea770f2f2b110bc57a3990","after":"8e0f4b943fee13afc970ca8277a8e76b9da63b96","ref":"refs/heads/main","pushedAt":"2024-03-12T08:18:38.000Z","pushType":"push","commitsCount":477,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[NVPTX] Add support for atomic add for f16 type (#84295)\n\natom.add.noftz.f16 is supported since SM 7.0","shortMessageHtmlLink":"[NVPTX] Add support for atomic add for f16 type (llvm#84295)"}},{"before":"34ce80d71b8bf6bd0f8755ccbb2b40554a61e706","after":null,"ref":"refs/heads/atomic_add_f16","pushedAt":"2024-03-12T08:18:05.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"}},{"before":"0e4b917aef24adf80959d9c7b2fc052c74140bd3","after":"34ce80d71b8bf6bd0f8755ccbb2b40554a61e706","ref":"refs/heads/atomic_add_f16","pushedAt":"2024-03-12T07:13:38.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"Autogenerate tests.","shortMessageHtmlLink":"Autogenerate tests."}},{"before":"3bd7abaa7f1582f6f2706aac14bd8026925d7730","after":"0e4b917aef24adf80959d9c7b2fc052c74140bd3","ref":"refs/heads/atomic_add_f16","pushedAt":"2024-03-12T06:37:59.000Z","pushType":"push","commitsCount":474,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"Merge branch 'main' into atomic_add_f16","shortMessageHtmlLink":"Merge branch 'main' into atomic_add_f16"}},{"before":"9765a9a1362abe3536c6888b7d44ab1afbaecfb8","after":"3bd7abaa7f1582f6f2706aac14bd8026925d7730","ref":"refs/heads/atomic_add_f16","pushedAt":"2024-03-11T07:11:56.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"Remove dump-input=always left over from debugging.","shortMessageHtmlLink":"Remove dump-input=always left over from debugging."}},{"before":"dbd3f9a2b5707246932a17c3271298ba2e53fa1a","after":"9765a9a1362abe3536c6888b7d44ab1afbaecfb8","ref":"refs/heads/atomic_add_f16","pushedAt":"2024-03-11T07:06:11.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"Update tests and add more tests","shortMessageHtmlLink":"Update tests and add more tests"}},{"before":"9aefbbee78df14eedac9ffce4238ed520c7a2fe2","after":"dbd3f9a2b5707246932a17c3271298ba2e53fa1a","ref":"refs/heads/atomic_add_f16","pushedAt":"2024-03-08T07:38:32.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"Also check PTX version.","shortMessageHtmlLink":"Also check PTX version."}},{"before":"7bac5a88e37b016d4797072aa40cb791792c8749","after":"9aefbbee78df14eedac9ffce4238ed520c7a2fe2","ref":"refs/heads/atomic_add_f16","pushedAt":"2024-03-07T10:00:47.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[NVPTX] Add atomic add test.","shortMessageHtmlLink":"[NVPTX] Add atomic add test."}},{"before":"6e79f77adbbd338848ea770f2f2b110bc57a3990","after":"7bac5a88e37b016d4797072aa40cb791792c8749","ref":"refs/heads/atomic_add_f16","pushedAt":"2024-03-07T09:59:29.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[NVPTX] Add support for atomic add for f16 type.\n\natom.add.noftz.f16 is supported since SM 7.0","shortMessageHtmlLink":"[NVPTX] Add support for atomic add for f16 type."}},{"before":null,"after":"6e79f77adbbd338848ea770f2f2b110bc57a3990","ref":"refs/heads/atomic_add_f16","pushedAt":"2024-03-07T09:52:18.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[dataflow][nfc] Fix u8 string usage with c++20 (#84291)\n\nClang returns an error when compiling this file with c++20\r\n```\r\nerror: ISO C++20 does not permit initialization of char array with UTF-8 string literal\r\n```\r\nIt seems like c++20 treats u8strings differently than strings (probably\r\nneeds char8_t).\r\nMake this a string to fix the error.","shortMessageHtmlLink":"[dataflow][nfc] Fix u8 string usage with c++20 (llvm#84291)"}},{"before":"02362b1ad1c07a01714b195d769400dd40dbfd04","after":"6e79f77adbbd338848ea770f2f2b110bc57a3990","ref":"refs/heads/main","pushedAt":"2024-03-07T09:49:40.000Z","pushType":"push","commitsCount":2696,"pusher":{"login":"akuegel","name":"Adrian Kuegel","path":"/akuegel","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/14309772?s=80&v=4"},"commit":{"message":"[dataflow][nfc] Fix u8 string usage with c++20 (#84291)\n\nClang returns an error when compiling this file with c++20\r\n```\r\nerror: ISO C++20 does not permit initialization of char array with UTF-8 string literal\r\n```\r\nIt seems like c++20 treats u8strings differently than strings (probably\r\nneeds char8_t).\r\nMake this a string to fix the error.","shortMessageHtmlLink":"[dataflow][nfc] Fix u8 string usage with c++20 (llvm#84291)"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAEVXQqAgA","startCursor":null,"endCursor":null}},"title":"Activity · akuegel/llvm-project"}