{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":2343211,"defaultBranch":"for-linus","name":"ceph-client","ownerLogin":"ceph","currentUserCanPush":false,"isFork":false,"isEmpty":false,"createdAt":"2011-09-07T17:26:50.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/1015767?v=4","public":true,"private":false,"isOrgOwned":true},"refInfo":{"name":"","listCacheKey":"v0:1722617053.0","currentOid":""},"activityList":{"items":[{"before":"ed98f3985b626a58b7f51c3b826775f7e541ff4b","after":"2ca44a2f35ca8f36983aaaa46d49312239ef519a","ref":"refs/heads/testing","pushedAt":"2024-09-06T10:58:29.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lxbsz","name":"Xiubo Li","path":"/lxbsz","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27719678?s=80&v=4"},"commit":{"message":"ceph: Remove empty definition in header file\n\nThe real definition of ceph_acl_chmod() has been removed since\ncommit 4db658ea0ca2 (\"ceph: Fix up after semantic merge conflict\"),\nremain the empty definition untouched in the header files. Let's\nremove the empty definition.\n\nSigned-off-by: Zhang Zekun \nReviewed-by: Xiubo Li ","shortMessageHtmlLink":"ceph: Remove empty definition in header file"}},{"before":"8ad8b8b160987a14eb199bb67cc00e23a08d43f6","after":"ed98f3985b626a58b7f51c3b826775f7e541ff4b","ref":"refs/heads/testing","pushedAt":"2024-09-05T12:29:36.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"lxbsz","name":"Xiubo Li","path":"/lxbsz","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27719678?s=80&v=4"},"commit":{"message":"ceph: remove the incorrect Fw reference check when dirtying pages\n\nWhen doing the direct-io reads it will also try to mark pages dirty,\nbut for the read path it won't hold the Fw caps and there is case\nwill it get the Fw reference.\n\nFixes: 5dda377cf0a6 (\"ceph: set i_head_snapc when getting CEPH_CAP_FILE_WR reference\")\nSigned-off-by: Xiubo Li ","shortMessageHtmlLink":"ceph: remove the incorrect Fw reference check when dirtying pages"}},{"before":"6b05d4783986ea7a978621e6aecdd5311718203d","after":"8ad8b8b160987a14eb199bb67cc00e23a08d43f6","ref":"refs/heads/testing","pushedAt":"2024-08-27T07:34:47.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"ceph: fix memory in MDS client cap_auths\n\nThe cap_auths that are allocated during an MDS session opening are never\nreleased, causing a memory leak detected by kmemleak. Fix this by freeing\nthe memory allocated when shutting down the mds client.\n\nFixes: 1d17de9534cb (\"ceph: save cap_auths in MDS client when session is opened\")\nSigned-off-by: Luis Henriques (SUSE) \nReviewed-by: Xiubo Li ","shortMessageHtmlLink":"ceph: fix memory in MDS client cap_auths"}},{"before":"43f5b8b50fb547fafcdc21154623f85a02bfa81c","after":"ede0b1d30b82829d6bc7924be18c7ae09cb1eb33","ref":"refs/heads/master","pushedAt":"2024-08-27T07:32:31.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"libceph: use min() to simplify code in ceph_dns_resolve_name()\n\nWhen resolving name in ceph_dns_resolve_name(), the end address of name\nis determined by the minimum value of delim_p and colon_p. So using min()\nhere is more in line with the context.\n\nSigned-off-by: Li Zetao \nReviewed-by: Simon Horman \nSigned-off-by: Ilya Dryomov ","shortMessageHtmlLink":"libceph: use min() to simplify code in ceph_dns_resolve_name()"}},{"before":"8618c856c9a644ad80ca70c748bebe0c9b6227d4","after":"6b05d4783986ea7a978621e6aecdd5311718203d","ref":"refs/heads/testing","pushedAt":"2024-08-27T07:27:27.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"libceph: use min() to simplify code in ceph_dns_resolve_name()\n\nWhen resolving name in ceph_dns_resolve_name(), the end address of name\nis determined by the minimum value of delim_p and colon_p. So using min()\nhere is more in line with the context.\n\nSigned-off-by: Li Zetao \nReviewed-by: Simon Horman \nSigned-off-by: Ilya Dryomov ","shortMessageHtmlLink":"libceph: use min() to simplify code in ceph_dns_resolve_name()"}},{"before":"7e35574567e19d2ca345caad671b98560d512bea","after":"8618c856c9a644ad80ca70c748bebe0c9b6227d4","ref":"refs/heads/testing","pushedAt":"2024-08-23T01:23:07.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lxbsz","name":"Xiubo Li","path":"/lxbsz","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27719678?s=80&v=4"},"commit":{"message":"ceph: Convert to use jiffies macro\n\nUse time_after_eq macro instead of using\njiffies directly to handle wraparound.\nThe modifications made compared to the previous version are as follows:\n1. Remove extra '+' in the header.\n\n[ xiubli: adjust the header files order ]\n\nSigned-off-by: Chen Yufan \nReviewed-by: Xiubo Li ","shortMessageHtmlLink":"ceph: Convert to use jiffies macro"}},{"before":"7503b862ff5bc4e556b3daee1c7881aa3e90ed62","after":"7e35574567e19d2ca345caad671b98560d512bea","ref":"refs/heads/testing","pushedAt":"2024-08-19T12:29:49.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lxbsz","name":"Xiubo Li","path":"/lxbsz","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27719678?s=80&v=4"},"commit":{"message":"ceph: fix memory in MDS client cap_auths\n\nThe cap_auths that are allocated during an MDS session opening are never\nreleased, causing a memory leak detected by kmemleak. Fix this by freeing\nthe memory allocated when shutting down the mds client.\n\nFixes: 1d17de9534cb (\"ceph: save cap_auths in MDS client when session is opened\")\nSigned-off-by: Luis Henriques (SUSE) \nReviewed-by: Xiubo Li ","shortMessageHtmlLink":"ceph: fix memory in MDS client cap_auths"}},{"before":"ef231aaae597a12bd3e2479b633f31ed1fcc23a8","after":"7503b862ff5bc4e556b3daee1c7881aa3e90ed62","ref":"refs/heads/testing","pushedAt":"2024-08-16T07:33:28.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"ceph: flush all the caps release when syncing the whole filesystem\n\nWe have hit a race between cap releases and cap revoke request\nthat will cause the check_caps() to miss sending a cap revoke ack\nto MDS. And the client will depend on the cap release to release\nthat revoking caps, which could be delayed for some unknown reasons.\n\nIn Kclient we have figured out the RCA about race and we need\na way to explictly trigger this manually could help to get rid\nof the caps revoke stuck issue.\n\nURL: https://tracker.ceph.com/issues/67221\nSigned-off-by: Xiubo Li ","shortMessageHtmlLink":"ceph: flush all the caps release when syncing the whole filesystem"}},{"before":"31634d7597d8c57894b6c98eeefc9e58cf842993","after":"43f5b8b50fb547fafcdc21154623f85a02bfa81c","ref":"refs/heads/master","pushedAt":"2024-08-16T07:26:34.000Z","pushType":"push","commitsCount":652,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"ceph: Remove unused declarations\n\nThese functions is never implemented and used.\n\nSigned-off-by: Yue Haibing \nReviewed-by: Jeff Layton \nSigned-off-by: Ilya Dryomov ","shortMessageHtmlLink":"ceph: Remove unused declarations"}},{"before":"3ceccb14f5576e02b81cc8b105ab81f224bd87f6","after":"31634d7597d8c57894b6c98eeefc9e58cf842993","ref":"refs/heads/for-linus","pushedAt":"2024-08-02T16:43:56.000Z","pushType":"push","commitsCount":10000,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"ceph: force sending a cap update msg back to MDS for revoke op\n\nIf a client sends out a cap update dropping caps with the prior 'seq'\njust before an incoming cap revoke request, then the client may drop\nthe revoke because it believes it's already released the requested\ncapabilities.\n\nThis causes the MDS to wait indefinitely for the client to respond\nto the revoke. It's therefore always a good idea to ack the cap\nrevoke request with the bumped up 'seq'.\n\nCurrently if the cap->issued equals to the newcaps the check_caps()\nwill do nothing, we should force flush the caps.\n\nCc: stable@vger.kernel.org\nLink: https://tracker.ceph.com/issues/61782\nSigned-off-by: Xiubo Li \nReviewed-by: Venky Shankar \nSigned-off-by: Ilya Dryomov ","shortMessageHtmlLink":"ceph: force sending a cap update msg back to MDS for revoke op"}},{"before":"34096973aec0bfbe2bb86f34701acc6cd4133fd9","after":"ef231aaae597a12bd3e2479b633f31ed1fcc23a8","ref":"refs/heads/testing","pushedAt":"2024-08-01T11:19:37.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"ceph: flush all the caps release when syncing the whole filesystem\n\nWe have hit a race between cap releases and cap revoke request\nthat will cause the check_caps() to miss sending a cap revoke ack\nto MDS. And the client will depend on the cap release to release\nthat revoking caps, which could be delayed for some unknown reasons.\n\nIn Kclient we have figured out the RCA about race and we need\na way to explictly trigger this manually could help to get rid\nof the caps revoke stuck issue.\n\nURL: https://tracker.ceph.com/issues/67221\nSigned-off-by: Xiubo Li ","shortMessageHtmlLink":"ceph: flush all the caps release when syncing the whole filesystem"}},{"before":"3ceccb14f5576e02b81cc8b105ab81f224bd87f6","after":"31634d7597d8c57894b6c98eeefc9e58cf842993","ref":"refs/heads/master","pushedAt":"2024-08-01T11:17:27.000Z","pushType":"push","commitsCount":10000,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"ceph: force sending a cap update msg back to MDS for revoke op\n\nIf a client sends out a cap update dropping caps with the prior 'seq'\njust before an incoming cap revoke request, then the client may drop\nthe revoke because it believes it's already released the requested\ncapabilities.\n\nThis causes the MDS to wait indefinitely for the client to respond\nto the revoke. It's therefore always a good idea to ack the cap\nrevoke request with the bumped up 'seq'.\n\nCurrently if the cap->issued equals to the newcaps the check_caps()\nwill do nothing, we should force flush the caps.\n\nCc: stable@vger.kernel.org\nLink: https://tracker.ceph.com/issues/61782\nSigned-off-by: Xiubo Li \nReviewed-by: Venky Shankar \nSigned-off-by: Ilya Dryomov ","shortMessageHtmlLink":"ceph: force sending a cap update msg back to MDS for revoke op"}},{"before":"3e9255e907100f88049f320ca33d977e29155d12","after":"34096973aec0bfbe2bb86f34701acc6cd4133fd9","ref":"refs/heads/testing","pushedAt":"2024-07-30T05:42:47.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"lxbsz","name":"Xiubo Li","path":"/lxbsz","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27719678?s=80&v=4"},"commit":{"message":"ceph: flush all the caps release when syncing the whole filesystem\n\nWe have hit a race between cap releases and cap revoke request\nthat will cause the check_caps() to miss sending a cap revoke ack\nto MDS. And the client will depend on the cap release to release\nthat revoking caps, which could be delayed for some unknown reasons.\n\nIn Kclient we have figured out the RCA about race and we need\na way to explictly trigger this manually could help to get rid\nof the caps revoke stuck issue.\n\nURL: https://tracker.ceph.com/issues/67221\nSigned-off-by: Xiubo Li ","shortMessageHtmlLink":"ceph: flush all the caps release when syncing the whole filesystem"}},{"before":"359bc01d2ecc9093216d21cfa03a545c44413cb6","after":"3ceccb14f5576e02b81cc8b105ab81f224bd87f6","ref":"refs/heads/for-linus","pushedAt":"2024-07-26T14:38:48.000Z","pushType":"push","commitsCount":281,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"rbd: don't assume rbd_is_lock_owner() for exclusive mappings\n\nExpanding on the previous commit, assuming that rbd_is_lock_owner()\nalways returns true (i.e. that we are either in RBD_LOCK_STATE_LOCKED\nor RBD_LOCK_STATE_QUIESCING) if the mapping is exclusive is wrong too.\nIn case ceph_cls_set_cookie() fails, the lock would be temporarily\nreleased even if the mapping is exclusive, meaning that we can end up\neven in RBD_LOCK_STATE_UNLOCKED.\n\nIOW, exclusive mappings are really \"just\" about disabling automatic\nlock transitions (as documented in the man page), not about grabbing\nthe lock and holding on to it whatever it takes.\n\nCc: stable@vger.kernel.org\nFixes: 637cd060537d (\"rbd: new exclusive lock wait/wake code\")\nSigned-off-by: Ilya Dryomov \nReviewed-by: Dongsheng Yang ","shortMessageHtmlLink":"rbd: don't assume rbd_is_lock_owner() for exclusive mappings"}},{"before":"7bc81acc946e36bfc3ec8313b59200005ddf3f73","after":"3e9255e907100f88049f320ca33d977e29155d12","ref":"refs/heads/testing","pushedAt":"2024-07-25T12:48:41.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"ceph: force sending a cap update msg back to MDS for revoke op\n\nIf a client sends out a cap update dropping caps with the prior 'seq'\njust before an incoming cap revoke request, then the client may drop\nthe revoke because it believes it's already released the requested\ncapabilities.\n\nThis causes the MDS to wait indefinitely for the client to respond\nto the revoke. It's therefore always a good idea to ack the cap\nrevoke request with the bumped up 'seq'.\n\nCurrently if the cap->issued equals to the newcaps the check_caps()\nwill do nothing, we should force flush the caps.\n\nCc: stable@vger.kernel.org\nLink: https://tracker.ceph.com/issues/61782\nSigned-off-by: Xiubo Li ","shortMessageHtmlLink":"ceph: force sending a cap update msg back to MDS for revoke op"}},{"before":"75104e41ebf16190bdbcbec869cc94d33c281797","after":"3ceccb14f5576e02b81cc8b105ab81f224bd87f6","ref":"refs/heads/master","pushedAt":"2024-07-25T10:18:50.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"rbd: don't assume rbd_is_lock_owner() for exclusive mappings\n\nExpanding on the previous commit, assuming that rbd_is_lock_owner()\nalways returns true (i.e. that we are either in RBD_LOCK_STATE_LOCKED\nor RBD_LOCK_STATE_QUIESCING) if the mapping is exclusive is wrong too.\nIn case ceph_cls_set_cookie() fails, the lock would be temporarily\nreleased even if the mapping is exclusive, meaning that we can end up\neven in RBD_LOCK_STATE_UNLOCKED.\n\nIOW, exclusive mappings are really \"just\" about disabling automatic\nlock transitions (as documented in the man page), not about grabbing\nthe lock and holding on to it whatever it takes.\n\nCc: stable@vger.kernel.org\nFixes: 637cd060537d (\"rbd: new exclusive lock wait/wake code\")\nSigned-off-by: Ilya Dryomov \nReviewed-by: Dongsheng Yang ","shortMessageHtmlLink":"rbd: don't assume rbd_is_lock_owner() for exclusive mappings"}},{"before":"5c9e62c1afd3eb0b1f41e3a1abdfbc2c1ba07615","after":"75104e41ebf16190bdbcbec869cc94d33c281797","ref":"refs/heads/master","pushedAt":"2024-07-25T10:17:29.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"rbd: don't assume rbd_is_lock_owner() for exclusive mappings\n\nExpanding on the previous commit, assuming that rbd_is_lock_owner()\nalways returns true (i.e. that we are either in RBD_LOCK_STATE_LOCKED\nor RBD_LOCK_STATE_QUIESCING) if the mapping is exclusive is wrong too.\nIn case ceph_cls_set_cookie() fails, the lock would be temporarily\nreleased even if the mapping is exclusive, meaning that we can end up\neven in RBD_LOCK_STATE_UNLOCKED.\n\nIOW, exclusive mappings are really \"just\" about disabling automatic\nlock transitions (as documented in the man page), not about grabbing\nthe lock and holding on to it whatever it takes.\n\nCc: stable@vger.kernel.org\nFixes: 637cd060537d (\"rbd: new exclusive lock wait/wake code\")\nSigned-off-by: Ilya Dryomov \nReviewed-by: Dongsheng Yang ","shortMessageHtmlLink":"rbd: don't assume rbd_is_lock_owner() for exclusive mappings"}},{"before":"03230edb0bd831662a7c08b6fef66b2a9a817774","after":"5c9e62c1afd3eb0b1f41e3a1abdfbc2c1ba07615","ref":"refs/heads/master","pushedAt":"2024-07-24T09:19:12.000Z","pushType":"push","commitsCount":3,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"rbd: don't assume rbd_is_lock_owner() for exclusive mappings\n\nExpanding on the previous commit, assuming that rbd_is_lock_owner()\nalways returns true (i.e. that we are either in RBD_LOCK_STATE_LOCKED\nor RBD_LOCK_STATE_QUIESCING) if the mapping is exclusive is wrong too.\nIn case ceph_cls_set_cookie() fails, the lock would be temporarily\nreleased even if the mapping is exclusive, meaning that we can end up\neven in RBD_LOCK_STATE_UNLOCKED.\n\nIOW, exclusive mappings are really \"just\" about disabling automatic\nlock transitions (as documented in the man page), not about grabbing\nthe lock and holding on to it whatever it takes.\n\nCc: stable@vger.kernel.org\nFixes: 637cd060537d (\"rbd: new exclusive lock wait/wake code\")\nSigned-off-by: Ilya Dryomov ","shortMessageHtmlLink":"rbd: don't assume rbd_is_lock_owner() for exclusive mappings"}},{"before":null,"after":"23aa762179fa36ecac7a54037392ed141b2090ff","ref":"refs/heads/wip-exclusive-option-states","pushedAt":"2024-07-23T18:44:47.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"rbd: don't assume rbd_is_lock_owner() for exclusive mappings\n\nExpanding on the previous commit, assuming that rbd_is_lock_owner()\nalways returns true (i.e. that we are either in RBD_LOCK_STATE_LOCKED\nor RBD_LOCK_STATE_QUIESCING) if the mapping is exclusive is wrong too.\nIn case ceph_cls_set_cookie() fails, the lock would be temporarily\nreleased even if the mapping is exclusive, meaning that we can end up\neven in RBD_LOCK_STATE_UNLOCKED.\n\nIOW, exclusive mappings are really \"just\" about disabling automatic\nlock transitions (as documented in the man page), not about grabbing\nthe lock and holding on to it whatever it takes.\n\nSigned-off-by: Ilya Dryomov ","shortMessageHtmlLink":"rbd: don't assume rbd_is_lock_owner() for exclusive mappings"}},{"before":"c30fd5683ea85fa16b09ca6926db2200d11edd1d","after":"7bc81acc946e36bfc3ec8313b59200005ddf3f73","ref":"refs/heads/testing","pushedAt":"2024-07-23T17:40:12.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"ceph: force sending a cap update msg back to MDS for revoke op\n\nIf a client sends out a cap update dropping caps with the prior 'seq'\njust before an incoming cap revoke request, then the client may drop\nthe revoke because it believes it's already released the requested\ncapabilities.\n\nThis causes the MDS to wait indefinitely for the client to respond\nto the revoke. It's therefore always a good idea to ack the cap\nrevoke request with the bumped up 'seq'.\n\nCurrently if the cap->issued equals to the newcaps the check_caps()\nwill do nothing, we should force flush the caps.\n\nCc: stable@vger.kernel.org\nLink: https://tracker.ceph.com/issues/61782\nSigned-off-by: Xiubo Li ","shortMessageHtmlLink":"ceph: force sending a cap update msg back to MDS for revoke op"}},{"before":"359bc01d2ecc9093216d21cfa03a545c44413cb6","after":"03230edb0bd831662a7c08b6fef66b2a9a817774","ref":"refs/heads/master","pushedAt":"2024-07-23T08:31:55.000Z","pushType":"push","commitsCount":278,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"ceph: fix incorrect kmalloc size of pagevec mempool\n\nThe kmalloc size of pagevec mempool is incorrectly calculated.\nIt misses the size of page pointer and only accounts the number for the array.\n\nFixes: a0102bda5bc0 (\"ceph: move sb->wb_pagevec_pool to be a global mempool\")\nSigned-off-by: ethanwu \nReviewed-by: Xiubo Li \nSigned-off-by: Ilya Dryomov ","shortMessageHtmlLink":"ceph: fix incorrect kmalloc size of pagevec mempool"}},{"before":"7413610fe7f3d299b2889cc32b8b541267df7913","after":"c30fd5683ea85fa16b09ca6926db2200d11edd1d","ref":"refs/heads/testing","pushedAt":"2024-07-16T08:48:47.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"lxbsz","name":"Xiubo Li","path":"/lxbsz","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27719678?s=80&v=4"},"commit":{"message":"ceph: force sending a cap update msg back to MDS for revoke op\n\nIf a client sends out a cap update dropping caps with the prior 'seq'\njust before an incoming cap revoke request, then the client may drop\nthe revoke because it believes it's already released the requested\ncapabilities.\n\nThis causes the MDS to wait indefinitely for the client to respond\nto the revoke. It's therefore always a good idea to ack the cap\nrevoke request with the bumped up 'seq'.\n\nCurrently if the cap->issued equals to the newcaps the check_caps()\nwill do nothing, we should force flush the caps.\n\nCc: stable@vger.kernel.org\nLink: https://tracker.ceph.com/issues/61782\nSigned-off-by: Xiubo Li ","shortMessageHtmlLink":"ceph: force sending a cap update msg back to MDS for revoke op"}},{"before":"67cda83601c15d40f669af21ed3d5646e9ccfb0e","after":"7413610fe7f3d299b2889cc32b8b541267df7913","ref":"refs/heads/testing","pushedAt":"2024-07-15T02:12:04.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lxbsz","name":"Xiubo Li","path":"/lxbsz","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27719678?s=80&v=4"},"commit":{"message":"ceph: fix incorrect kmalloc size of pagevec mempool\n\nThe kmalloc size of pagevec mempool is incorrectly calculated.\nIt misses the size of page pointer and only accounts the number for the array.\n\nFixes: a0102bda5bc0 (\"ceph: move sb->wb_pagevec_pool to be a global mempool\")\nSigned-off-by: ethanwu \nReviewed-by: Xiubo Li ","shortMessageHtmlLink":"ceph: fix incorrect kmalloc size of pagevec mempool"}},{"before":"9a71ad4237626f12236c921b19d996b3a72af338","after":"67cda83601c15d40f669af21ed3d5646e9ccfb0e","ref":"refs/heads/testing","pushedAt":"2024-07-15T01:42:16.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"lxbsz","name":"Xiubo Li","path":"/lxbsz","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27719678?s=80&v=4"},"commit":{"message":"ceph: force sending a cap update msg back to MDS for revoke op\n\nIf a client sends out a cap update dropping caps with the prior 'seq'\njust before an incoming cap revoke request, then the client may drop\nthe revoke because it believes it's already released the requested\ncapabilities.\n\nThis causes the MDS to wait indefinitely for the client to respond\nto the revoke. It's therefore always a good idea to ack the cap\nrevoke request with the bumped up 'seq'.\n\nCurrently if the cap->issued equals to the newcaps the check_caps()\nwill do nothing, we should force flush the caps.\n\nCc: stable@vger.kernel.org\nLink: https://tracker.ceph.com/issues/61782\nSigned-off-by: Xiubo Li ","shortMessageHtmlLink":"ceph: force sending a cap update msg back to MDS for revoke op"}},{"before":"93a2221c9c1ae32643df67c482dc4c4c591b7514","after":"359bc01d2ecc9093216d21cfa03a545c44413cb6","ref":"refs/heads/for-linus","pushedAt":"2024-07-12T16:43:30.000Z","pushType":"push","commitsCount":10000,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"libceph: fix crush_choose_firstn() kernel-doc warnings\n\nCurrently, when built with \"make W=1\", the following warnings are\ngenerated:\n\nnet/ceph/crush/mapper.c:466: warning: Function parameter or struct member 'work' not described in 'crush_choose_firstn'\nnet/ceph/crush/mapper.c:466: warning: Function parameter or struct member 'weight' not described in 'crush_choose_firstn'\nnet/ceph/crush/mapper.c:466: warning: Function parameter or struct member 'weight_max' not described in 'crush_choose_firstn'\nnet/ceph/crush/mapper.c:466: warning: Function parameter or struct member 'choose_args' not described in 'crush_choose_firstn'\n\nUpdate the crush_choose_firstn() kernel-doc to document these\nparameters.\n\nSigned-off-by: Jeff Johnson \nReviewed-by: Simon Horman \nSigned-off-by: Ilya Dryomov ","shortMessageHtmlLink":"libceph: fix crush_choose_firstn() kernel-doc warnings"}},{"before":"69c7b2fe4c9cc1d3b1186d1c5606627ecf0de883","after":"359bc01d2ecc9093216d21cfa03a545c44413cb6","ref":"refs/heads/master","pushedAt":"2024-07-11T14:34:43.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"libceph: fix crush_choose_firstn() kernel-doc warnings\n\nCurrently, when built with \"make W=1\", the following warnings are\ngenerated:\n\nnet/ceph/crush/mapper.c:466: warning: Function parameter or struct member 'work' not described in 'crush_choose_firstn'\nnet/ceph/crush/mapper.c:466: warning: Function parameter or struct member 'weight' not described in 'crush_choose_firstn'\nnet/ceph/crush/mapper.c:466: warning: Function parameter or struct member 'weight_max' not described in 'crush_choose_firstn'\nnet/ceph/crush/mapper.c:466: warning: Function parameter or struct member 'choose_args' not described in 'crush_choose_firstn'\n\nUpdate the crush_choose_firstn() kernel-doc to document these\nparameters.\n\nSigned-off-by: Jeff Johnson \nReviewed-by: Simon Horman \nSigned-off-by: Ilya Dryomov ","shortMessageHtmlLink":"libceph: fix crush_choose_firstn() kernel-doc warnings"}},{"before":"e092818aaa88764fa61bab25bb186f5169c0f6a3","after":"9a71ad4237626f12236c921b19d996b3a72af338","ref":"refs/heads/testing","pushedAt":"2024-07-11T10:43:11.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lxbsz","name":"Xiubo Li","path":"/lxbsz","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27719678?s=80&v=4"},"commit":{"message":"ceph: force sending a cap update msg back to MDS for revoke op\n\nIf a client sends out a cap update dropping caps with the prior 'seq'\njust before an incoming cap revoke request, then the client may drop\nthe revoke because it believes it's already released the requested\ncapabilities.\n\nThis causes the MDS to wait indefinitely for the client to respond\nto the revoke. It's therefore always a good idea to ack the cap\nrevoke request with the bumped up 'seq'.\n\nCurrently if the cap->issued equals to the newcaps the check_caps()\nwill do nothing, we should force flush the caps.\n\nCc: stable@vger.kernel.org\nLink: https://tracker.ceph.com/issues/61782\nSigned-off-by: Xiubo Li ","shortMessageHtmlLink":"ceph: force sending a cap update msg back to MDS for revoke op"}},{"before":"d25801b6defdaef655b98b6d7ba00106d6567e5d","after":"e092818aaa88764fa61bab25bb186f5169c0f6a3","ref":"refs/heads/testing","pushedAt":"2024-07-11T05:27:49.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"lxbsz","name":"Xiubo Li","path":"/lxbsz","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27719678?s=80&v=4"},"commit":{"message":"ceph: periodically flush the cap releases\n\nThe MDS could be waiting the caps releases infinitely in some corner\ncase and then reporting the caps revoke stuck warning. To fix this\nwe should periodically flush the cap releases.\n\nURL: https://tracker.ceph.com/issues/57244\nSigned-off-by: Xiubo Li ","shortMessageHtmlLink":"ceph: periodically flush the cap releases"}},{"before":"43ff43b7f44ad7dfcdb25a0dfc9f78c5d4b482cf","after":"d25801b6defdaef655b98b6d7ba00106d6567e5d","ref":"refs/heads/testing","pushedAt":"2024-07-10T12:40:12.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lxbsz","name":"Xiubo Li","path":"/lxbsz","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/27719678?s=80&v=4"},"commit":{"message":"ceph: periodically flush the cap releases\n\nThe MDS could be waiting the caps releases infinitely in some corner\ncase and then reporting the caps revoke stuck warning. To fix this\nwe should periodically flush the cap releases.\n\nURL: https://tracker.ceph.com/issues/57244\nSigned-off-by: Xiubo Li ","shortMessageHtmlLink":"ceph: periodically flush the cap releases"}},{"before":"3739909cfccdcf4ec24bc40a8bdaf6b0f28ab28b","after":"43ff43b7f44ad7dfcdb25a0dfc9f78c5d4b482cf","ref":"refs/heads/testing","pushedAt":"2024-07-10T08:16:34.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"idryomov","name":"Ilya Dryomov","path":"/idryomov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/567341?s=80&v=4"},"commit":{"message":"ceph: convert comma to semicolon in __ceph_dentry_dir_lease_touch()\n\nReplace a comma between expression statements by a semicolon.\n\nSigned-off-by: Chen Ni \nReviewed-by: Xiubo Li \nSigned-off-by: Ilya Dryomov ","shortMessageHtmlLink":"ceph: convert comma to semicolon in __ceph_dentry_dir_lease_touch()"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAErtNwqAA","startCursor":null,"endCursor":null}},"title":"Activity ยท ceph/ceph-client"}