{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":50436766,"defaultBranch":"master","name":"ROCK-Kernel-Driver","ownerLogin":"ROCm","currentUserCanPush":false,"isFork":false,"isEmpty":false,"createdAt":"2016-01-26T15:11:38.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/21157610?v=4","public":true,"private":false,"isOrgOwned":true},"refInfo":{"name":"","listCacheKey":"v0:1714487495.0","currentOid":""},"activityList":{"items":[{"before":"dad5559abe9d24d2190749e580c32ba681cd19e5","after":"edd10b07b220f05e24fd06ca284e295e3ca2c388","ref":"refs/heads/roc-6.1.x","pushedAt":"2024-04-30T14:31:14.000Z","pushType":"push","commitsCount":32,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdkfd: Fix memory leak in create_process failure\n\nFix memory leak due to a leaked mmget reference on an error handling\ncode path that is triggered when attempting to create KFD processes\nwhile a GPU reset is in progress.\n\nFixes: 0ab2d7532b05 (\"drm/amdkfd: prepare per-process debug enable and disable\")\nCC: Xiaogang Chen \nSigned-off-by: Felix Kuehling \nTested-by: Harish Kasiviswanthan \nReviewed-by: Mukul Joshi \n(cherry picked from commit 8637573fa72a2f6cbbb1242304be85a5dd386b67)\n\nChange-Id: I10270b7616bdc96e04a9d58506152e6862ea3b4b","shortMessageHtmlLink":"drm/amdkfd: Fix memory leak in create_process failure"}},{"before":"dad5559abe9d24d2190749e580c32ba681cd19e5","after":"edd10b07b220f05e24fd06ca284e295e3ca2c388","ref":"refs/heads/master","pushedAt":"2024-04-30T14:31:08.000Z","pushType":"push","commitsCount":32,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdkfd: Fix memory leak in create_process failure\n\nFix memory leak due to a leaked mmget reference on an error handling\ncode path that is triggered when attempting to create KFD processes\nwhile a GPU reset is in progress.\n\nFixes: 0ab2d7532b05 (\"drm/amdkfd: prepare per-process debug enable and disable\")\nCC: Xiaogang Chen \nSigned-off-by: Felix Kuehling \nTested-by: Harish Kasiviswanthan \nReviewed-by: Mukul Joshi \n(cherry picked from commit 8637573fa72a2f6cbbb1242304be85a5dd386b67)\n\nChange-Id: I10270b7616bdc96e04a9d58506152e6862ea3b4b","shortMessageHtmlLink":"drm/amdkfd: Fix memory leak in create_process failure"}},{"before":null,"after":"70a56bb7a029d2a3db5b3e0b293699de056d7b21","ref":"refs/heads/dependabot/pip/drivers/gpu/drm/ci/xfails/idna-3.7","pushedAt":"2024-04-16T18:49:10.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"dependabot[bot]","name":null,"path":"/apps/dependabot","primaryAvatarUrl":"https://avatars.githubusercontent.com/in/29110?s=80&v=4"},"commit":{"message":"build(deps): bump idna from 3.4 to 3.7 in /drivers/gpu/drm/ci/xfails\n\nBumps [idna](https://github.com/kjd/idna) from 3.4 to 3.7.\n- [Release notes](https://github.com/kjd/idna/releases)\n- [Changelog](https://github.com/kjd/idna/blob/master/HISTORY.rst)\n- [Commits](https://github.com/kjd/idna/compare/v3.4...v3.7)\n\n---\nupdated-dependencies:\n- dependency-name: idna\n dependency-type: direct:production\n...\n\nSigned-off-by: dependabot[bot] ","shortMessageHtmlLink":"build(deps): bump idna from 3.4 to 3.7 in /drivers/gpu/drm/ci/xfails"}},{"before":null,"after":"dad5559abe9d24d2190749e580c32ba681cd19e5","ref":"refs/heads/roc-6.1.x","pushedAt":"2024-04-16T18:46:51.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"Revert \"drm/sched: Convert the GPU scheduler to variable number ...\"\n\nRevert submission 1033322\n\nReason for revert: Breaks GPU reset\nReverted Changes:\nI518055d36:drm/amdgpu: drop setting buffer funcs in sdma442\nI3161b3ba3:drm/amdgpu: fix buffer funcs setting order on susp...\nIf96a48811:drm/amdklc: fix a intree build error.\nI370135be6:drm/sched: Don't disturb the entity when in RR-mod...\nI051ca6e0b:drm/sched: Move free worker re-queuing out of the ...\nI25fb2785c:drm/sched: Rename drm_sched_get_cleanup_job to be ...\nI666125051:drm/sched: Rename drm_sched_run_job_queue_if_ready...\nI9a32472a9:drm/sched: Rename drm_sched_free_job_queue to be m...\nI4c2c9bed1:drm/sched: Re-queue run job worker when drm_sched_...\nIe513a85c7:drm/sched: Drain all entities in DRM sched run job...\nI855c12a4e:drm/sched: Split free_job into own work item\nI7a116284b:drm/sched: Convert drm scheduler to use a work que...\nIffc1aa4cd:drm/amdgpu: move buffer funcs setting up a level\nI3753e9dd4:drm/sched: Convert the GPU scheduler to variable n...\n\nChange-Id: I8fa3facd7111e26a08ff77e6345b4437631e9ac0","shortMessageHtmlLink":"Revert \"drm/sched: Convert the GPU scheduler to variable number ...\""}},{"before":"cf8e29ea7ba831dbafe9530dcdad8d5a5682c6d4","after":"dad5559abe9d24d2190749e580c32ba681cd19e5","ref":"refs/heads/master","pushedAt":"2024-04-16T18:46:47.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"Revert \"drm/sched: Convert the GPU scheduler to variable number ...\"\n\nRevert submission 1033322\n\nReason for revert: Breaks GPU reset\nReverted Changes:\nI518055d36:drm/amdgpu: drop setting buffer funcs in sdma442\nI3161b3ba3:drm/amdgpu: fix buffer funcs setting order on susp...\nIf96a48811:drm/amdklc: fix a intree build error.\nI370135be6:drm/sched: Don't disturb the entity when in RR-mod...\nI051ca6e0b:drm/sched: Move free worker re-queuing out of the ...\nI25fb2785c:drm/sched: Rename drm_sched_get_cleanup_job to be ...\nI666125051:drm/sched: Rename drm_sched_run_job_queue_if_ready...\nI9a32472a9:drm/sched: Rename drm_sched_free_job_queue to be m...\nI4c2c9bed1:drm/sched: Re-queue run job worker when drm_sched_...\nIe513a85c7:drm/sched: Drain all entities in DRM sched run job...\nI855c12a4e:drm/sched: Split free_job into own work item\nI7a116284b:drm/sched: Convert drm scheduler to use a work que...\nIffc1aa4cd:drm/amdgpu: move buffer funcs setting up a level\nI3753e9dd4:drm/sched: Convert the GPU scheduler to variable n...\n\nChange-Id: I8fa3facd7111e26a08ff77e6345b4437631e9ac0","shortMessageHtmlLink":"Revert \"drm/sched: Convert the GPU scheduler to variable number ...\""}},{"before":"cc2d8659760f7729b3bed35dc7870eecc5499fd4","after":"cf8e29ea7ba831dbafe9530dcdad8d5a5682c6d4","ref":"refs/heads/roc-6.0.x","pushedAt":"2024-01-31T19:36:11.000Z","pushType":"push","commitsCount":9,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdgpu: Show vram vendor only if available\n\nOny if vram vendor info is available, show in sysfs.\n\nSigned-off-by: Lijo Lazar \nReviewed-by: Hawking Zhang ","shortMessageHtmlLink":"drm/amdgpu: Show vram vendor only if available"}},{"before":"cc2d8659760f7729b3bed35dc7870eecc5499fd4","after":"cf8e29ea7ba831dbafe9530dcdad8d5a5682c6d4","ref":"refs/heads/master","pushedAt":"2024-01-31T19:36:01.000Z","pushType":"push","commitsCount":9,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdgpu: Show vram vendor only if available\n\nOny if vram vendor info is available, show in sysfs.\n\nSigned-off-by: Lijo Lazar \nReviewed-by: Hawking Zhang ","shortMessageHtmlLink":"drm/amdgpu: Show vram vendor only if available"}},{"before":"6a2d04d5e59501d5bcd0b8ec40c8c62eadec6047","after":null,"ref":"refs/heads/abhimeda-adding-issue-template","pushedAt":"2024-01-04T18:27:12.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"abhimeda","name":null,"path":"/abhimeda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/138710508?s=80&v=4"}},{"before":"cb064fa43f49f4248d7818cb11672f999b67e2a3","after":"6a2d04d5e59501d5bcd0b8ec40c8c62eadec6047","ref":"refs/heads/abhimeda-adding-issue-template","pushedAt":"2023-12-18T19:10:08.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"abhimeda","name":null,"path":"/abhimeda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/138710508?s=80&v=4"},"commit":{"message":"adding ROCm v6, MI300, default component","shortMessageHtmlLink":"adding ROCm v6, MI300, default component"}},{"before":"2a83b3a166ea27e10b5ecb9292965082f31c6b12","after":"cc2d8659760f7729b3bed35dc7870eecc5499fd4","ref":"refs/heads/master","pushedAt":"2023-12-14T02:08:41.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"rajbhar","name":"Rajneesh Bhardwaj","path":"/rajbhar","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27557546?s=80&v=4"},"commit":{"message":"Revert \"drm/amdgpu: Force order between a read and write to the same address\"\n\nThis reverts commit fbfca2c5b9d08b98dbdb98a85911a8488f9f2af4.\n\nReason for revert: Deferred to 6.0.1 apparently\n\nChange-Id: I7eec212d29a97658a6076080e23c7c6b08b2316a","shortMessageHtmlLink":"Revert \"drm/amdgpu: Force order between a read and write to the same …"}},{"before":null,"after":"cc2d8659760f7729b3bed35dc7870eecc5499fd4","ref":"refs/heads/roc-6.0.x","pushedAt":"2023-12-14T01:46:59.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"rajbhar","name":"Rajneesh Bhardwaj","path":"/rajbhar","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27557546?s=80&v=4"},"commit":{"message":"Revert \"drm/amdgpu: Force order between a read and write to the same address\"\n\nThis reverts commit fbfca2c5b9d08b98dbdb98a85911a8488f9f2af4.\n\nReason for revert: Deferred to 6.0.1 apparently\n\nChange-Id: I7eec212d29a97658a6076080e23c7c6b08b2316a","shortMessageHtmlLink":"Revert \"drm/amdgpu: Force order between a read and write to the same …"}},{"before":"d540066add12ada899d902e5488a4af12297b5ae","after":"cb064fa43f49f4248d7818cb11672f999b67e2a3","ref":"refs/heads/abhimeda-adding-issue-template","pushedAt":"2023-12-13T19:38:33.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"abhimeda","name":null,"path":"/abhimeda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/138710508?s=80&v=4"},"commit":{"message":"Add files via upload","shortMessageHtmlLink":"Add files via upload"}},{"before":null,"after":"d540066add12ada899d902e5488a4af12297b5ae","ref":"refs/heads/abhimeda-adding-issue-template","pushedAt":"2023-12-13T19:38:02.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"abhimeda","name":null,"path":"/abhimeda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/138710508?s=80&v=4"},"commit":{"message":"Create config.yml","shortMessageHtmlLink":"Create config.yml"}},{"before":"fb41287cfdce7a47e51d5f16b836cbaf68770114","after":"2a83b3a166ea27e10b5ecb9292965082f31c6b12","ref":"refs/heads/roc-5.7.x","pushedAt":"2023-10-13T14:15:19.000Z","pushType":"push","commitsCount":22,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdkfd: fix some race conditions in vram buffer alloc/free of svm code\n\nThis patch fixes:\n1: ref number of prange's svm_bo got decreased by an async call from hmm. When\nwait svm_bo of prange got released we shoul also wait prang->svm_bo become NULL,\notherwise prange->svm_bo may be set to null after allocate new vram buffer.\n\n2: During waiting svm_bo of prange got released in a while loop should reschedule\ncurrent task to give other tasks oppotunity to run, specially the the workque\ntask that handles svm_bo ref release, otherwise we may enter to softlock.\n\nSigned-off-by: Xiaogang.Chen \nReviewed-by: Felix Kuehling \n(cherry picked from commit 528eed632c4fdc5d2dec90f85f90c6f4cb88aa48)\n\nChange-Id: I80ce4bdc4c79cb8fd43064653b9842b528605dd3","shortMessageHtmlLink":"drm/amdkfd: fix some race conditions in vram buffer alloc/free of svm…"}},{"before":"fb41287cfdce7a47e51d5f16b836cbaf68770114","after":"2a83b3a166ea27e10b5ecb9292965082f31c6b12","ref":"refs/heads/master","pushedAt":"2023-10-13T14:15:17.000Z","pushType":"push","commitsCount":22,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdkfd: fix some race conditions in vram buffer alloc/free of svm code\n\nThis patch fixes:\n1: ref number of prange's svm_bo got decreased by an async call from hmm. When\nwait svm_bo of prange got released we shoul also wait prang->svm_bo become NULL,\notherwise prange->svm_bo may be set to null after allocate new vram buffer.\n\n2: During waiting svm_bo of prange got released in a while loop should reschedule\ncurrent task to give other tasks oppotunity to run, specially the the workque\ntask that handles svm_bo ref release, otherwise we may enter to softlock.\n\nSigned-off-by: Xiaogang.Chen \nReviewed-by: Felix Kuehling \n(cherry picked from commit 528eed632c4fdc5d2dec90f85f90c6f4cb88aa48)\n\nChange-Id: I80ce4bdc4c79cb8fd43064653b9842b528605dd3","shortMessageHtmlLink":"drm/amdkfd: fix some race conditions in vram buffer alloc/free of svm…"}},{"before":null,"after":"fb41287cfdce7a47e51d5f16b836cbaf68770114","ref":"refs/heads/roc-5.7.x","pushedAt":"2023-09-15T16:20:49.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdkfd: use correct method to get clock under SRIOV\n\n[What]\nCurrent SRIOV still using adev->clock.default_XX which gets from\natomfirmware. But these fields are abandoned in atomfirmware long ago.\nWhich may cause function to return a 0 value.\n\n[How]\nWe don't need to check whether SR-IOV. For SR-IOV one-vf-mode,\npm is enabled and VF is able to read dpm clock\nfrom pmfw, so we can use dpm clock interface directly. For\nmulti-VF mode, VF pm is disabled, so driver can just react as pm\ndisabled. One-vf-mode is introduced from GFX9 so it shall not have\nany backward compatibility issue.\n\nSigned-off-by: Horace Chen \nAcked-by: Felix Kuehling \n(cherry picked from commit 6ee9227c7e391acda661bd835d556eb3aedfe838)\n\nChange-Id: I38927b9c19b06b96ccb7013669f27a36154c3034\n(cherry picked from commit 6bbd176400bcdeadc4eef1fee009ec9bd352ed4a)","shortMessageHtmlLink":"drm/amdkfd: use correct method to get clock under SRIOV"}},{"before":"b728ef094ed974e6c7d3288cd97ce4a11519d901","after":"fb41287cfdce7a47e51d5f16b836cbaf68770114","ref":"refs/heads/master","pushedAt":"2023-09-15T16:20:41.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdkfd: use correct method to get clock under SRIOV\n\n[What]\nCurrent SRIOV still using adev->clock.default_XX which gets from\natomfirmware. But these fields are abandoned in atomfirmware long ago.\nWhich may cause function to return a 0 value.\n\n[How]\nWe don't need to check whether SR-IOV. For SR-IOV one-vf-mode,\npm is enabled and VF is able to read dpm clock\nfrom pmfw, so we can use dpm clock interface directly. For\nmulti-VF mode, VF pm is disabled, so driver can just react as pm\ndisabled. One-vf-mode is introduced from GFX9 so it shall not have\nany backward compatibility issue.\n\nSigned-off-by: Horace Chen \nAcked-by: Felix Kuehling \n(cherry picked from commit 6ee9227c7e391acda661bd835d556eb3aedfe838)\n\nChange-Id: I38927b9c19b06b96ccb7013669f27a36154c3034\n(cherry picked from commit 6bbd176400bcdeadc4eef1fee009ec9bd352ed4a)","shortMessageHtmlLink":"drm/amdkfd: use correct method to get clock under SRIOV"}},{"before":"795c38ee2a5de8659273232830c280f08a9ddc64","after":"b728ef094ed974e6c7d3288cd97ce4a11519d901","ref":"refs/heads/master","pushedAt":"2023-08-29T20:56:38.000Z","pushType":"push","commitsCount":4,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdkcl: test whether drm_dp_mst_topology_cbs->poll_hpd_irq is available\n\nIt's caused by ed8496801ab71fdfb9c9fdcbef058aa20a549ebd\n\"drm/amd/display: Add polling method to handle MST reply packet\"\n\nSigned-off-by: Bob Zhou \nReviewed-by: Leslie Shi ","shortMessageHtmlLink":"drm/amdkcl: test whether drm_dp_mst_topology_cbs->poll_hpd_irq is ava…"}},{"before":"795c38ee2a5de8659273232830c280f08a9ddc64","after":"b728ef094ed974e6c7d3288cd97ce4a11519d901","ref":"refs/heads/roc-5.6.x","pushedAt":"2023-08-29T20:56:34.000Z","pushType":"push","commitsCount":4,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdkcl: test whether drm_dp_mst_topology_cbs->poll_hpd_irq is available\n\nIt's caused by ed8496801ab71fdfb9c9fdcbef058aa20a549ebd\n\"drm/amd/display: Add polling method to handle MST reply packet\"\n\nSigned-off-by: Bob Zhou \nReviewed-by: Leslie Shi ","shortMessageHtmlLink":"drm/amdkcl: test whether drm_dp_mst_topology_cbs->poll_hpd_irq is ava…"}},{"before":null,"after":"795c38ee2a5de8659273232830c280f08a9ddc64","ref":"refs/heads/roc-5.6.x","pushedAt":"2023-06-28T13:53:40.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdkfd: Don't trigger evictions unmapping dmabuf attachments\n\nDon't move DMABuf attachments for PCIe P2P mappings to the SYSTEM domain\nwhen unmapping. This avoids triggering eviction fences unnecessarily.\nInstead do the move to SYSTEM and back to GTT when mapping these\nattachments to ensure the SG table gets updated after evictions.\n\nThis may still trigger unnecessary evictions if user mode unmaps and\nremaps the same BO. However, this is unlikely in real applications.\n\nSigned-off-by: Felix Kuehling \nReviewed-by: Eric Huang \n(cherry picked from commit b52649cf41d0c2ac98a835a518d36a03cdaf1087)\n\nChange-Id: I1c9cc6c3332e18a90484215d901d6b24f8668cf3","shortMessageHtmlLink":"drm/amdkfd: Don't trigger evictions unmapping dmabuf attachments"}},{"before":"0b579de9622f5c93021dcb7927d13926313740a2","after":"795c38ee2a5de8659273232830c280f08a9ddc64","ref":"refs/heads/master","pushedAt":"2023-06-28T13:53:31.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdkfd: Don't trigger evictions unmapping dmabuf attachments\n\nDon't move DMABuf attachments for PCIe P2P mappings to the SYSTEM domain\nwhen unmapping. This avoids triggering eviction fences unnecessarily.\nInstead do the move to SYSTEM and back to GTT when mapping these\nattachments to ensure the SG table gets updated after evictions.\n\nThis may still trigger unnecessary evictions if user mode unmaps and\nremaps the same BO. However, this is unlikely in real applications.\n\nSigned-off-by: Felix Kuehling \nReviewed-by: Eric Huang \n(cherry picked from commit b52649cf41d0c2ac98a835a518d36a03cdaf1087)\n\nChange-Id: I1c9cc6c3332e18a90484215d901d6b24f8668cf3","shortMessageHtmlLink":"drm/amdkfd: Don't trigger evictions unmapping dmabuf attachments"}},{"before":null,"after":"0b579de9622f5c93021dcb7927d13926313740a2","ref":"refs/heads/roc-5.5.x","pushedAt":"2023-04-24T13:23:44.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"Revert \"drm/amdgpu: TA unload messages are not actually sent to psp when amdgpu is uninstalled\"\n\nThis reverts commit fac53471d0ea9693d314aa2df08d62b2e7e3a0f8.\nThe following change: move the drm_dev_unplug call after\namdgpu_driver_unload_kms in amdgpu_pci_remove. The reason is\nthe following: amdgpu_pci_remove calls drm_dev_unregister\nand it should be called first to ensure userspace can't access the\ndevice instance anymore. If we call drm_dev_unplug after\namdgpu_driver_unload_kms then we observe IGT PCI software unplug\ntest failure (kernel hung) for all ASICs. This is how this\nregression was found.\n\nAfter this revert, the following commands do work not, but it would\nbe fixed in the next commit:\n - sudo modprobe -r amdgpu\n - sudo modprobe amdgpu\n\nSigned-off-by: Vitaly Prosyak \nReviewed-by Alex Deucher ","shortMessageHtmlLink":"Revert \"drm/amdgpu: TA unload messages are not actually sent to psp w…"}},{"before":"8ab8a55bc43060286d9a3c92bbf9f2dd90eedd5a","after":"0b579de9622f5c93021dcb7927d13926313740a2","ref":"refs/heads/master","pushedAt":"2023-04-24T13:23:37.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"Revert \"drm/amdgpu: TA unload messages are not actually sent to psp when amdgpu is uninstalled\"\n\nThis reverts commit fac53471d0ea9693d314aa2df08d62b2e7e3a0f8.\nThe following change: move the drm_dev_unplug call after\namdgpu_driver_unload_kms in amdgpu_pci_remove. The reason is\nthe following: amdgpu_pci_remove calls drm_dev_unregister\nand it should be called first to ensure userspace can't access the\ndevice instance anymore. If we call drm_dev_unplug after\namdgpu_driver_unload_kms then we observe IGT PCI software unplug\ntest failure (kernel hung) for all ASICs. This is how this\nregression was found.\n\nAfter this revert, the following commands do work not, but it would\nbe fixed in the next commit:\n - sudo modprobe -r amdgpu\n - sudo modprobe amdgpu\n\nSigned-off-by: Vitaly Prosyak \nReviewed-by Alex Deucher ","shortMessageHtmlLink":"Revert \"drm/amdgpu: TA unload messages are not actually sent to psp w…"}},{"before":"e2d59fe1b94a3e5a4264467118f391425efbd47c","after":"8ab8a55bc43060286d9a3c92bbf9f2dd90eedd5a","ref":"refs/heads/roc-5.4.x","pushedAt":"2023-03-22T13:23:52.000Z","pushType":"push","commitsCount":43,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdgpu: Add umc channel index mapping table for umc_v8_10\n\nAdd umc channel index mapping table for umc_v8_10.\n\nSigned-off-by: YiPeng Chai \nReviewed-by: Hawking Zhang \nReviewed-by: Tao Zhou ","shortMessageHtmlLink":"drm/amdgpu: Add umc channel index mapping table for umc_v8_10"}},{"before":"e2d59fe1b94a3e5a4264467118f391425efbd47c","after":"8ab8a55bc43060286d9a3c92bbf9f2dd90eedd5a","ref":"refs/heads/master","pushedAt":"2023-03-22T13:23:45.000Z","pushType":"push","commitsCount":43,"pusher":{"login":"kentrussell","name":"Kent Russell","path":"/kentrussell","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/19553272?s=80&v=4"},"commit":{"message":"drm/amdgpu: Add umc channel index mapping table for umc_v8_10\n\nAdd umc channel index mapping table for umc_v8_10.\n\nSigned-off-by: YiPeng Chai \nReviewed-by: Hawking Zhang \nReviewed-by: Tao Zhou ","shortMessageHtmlLink":"drm/amdgpu: Add umc channel index mapping table for umc_v8_10"}}],"hasNextPage":false,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAEPhwKcwA","startCursor":null,"endCursor":null}},"title":"Activity · ROCm/ROCK-Kernel-Driver"}