January 2023 - Linux-stable-mirror

FAILED: patch "[PATCH] io_uring/msg_ring: fix remote queue to disabled ring" failed to apply to 6.1-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.1-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. Possible dependencies: 8579538c89e3 ("io_uring/msg_ring: fix remote queue to disabled ring") 56d8e3180c06 ("io_uring/msg_ring: fix flagging remote execution") 423d5081d045 ("io_uring/msg_ring: move double lock/unlock helpers higher up") 761c61c15903 ("io_uring/msg_ring: flag target ring as having task_work, if needed") 6d043ee1164c ("io_uring: do msg_ring in target task via tw") 172113101641 ("io_uring: extract a io_msg_install_complete helper") 11373026f296 ("io_uring: get rid of double locking") 4c979eaefa43 ("io_uring: improve io_double_lock_ctx fail handling") b529c96a896b ("io_uring: remove overflow param from io_post_aux_cqe") a77ab745f28d ("io_uring: make io_fill_cqe_aux static") 9b8c54755a2b ("io_uring: add io_aux_cqe which allows deferred completion") 931147ddfa6e ("io_uring: allow defer completion for aux posted cqes") 1bec951c3809 ("io_uring: iopoll protect complete_post") fa18fa2272c7 ("io_uring: inline __io_req_complete_put()") f9d567c75ec2 ("io_uring: inline __io_req_complete_post()") e2ad599d1ed3 ("io_uring: allow multishot recv CQEs to overflow") 515e26961295 ("io_uring: revert "io_uring fix multishot accept ordering"") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 8579538c89e33ce78be2feb41e07489c8cbf8f31 Mon Sep 17 00:00:00 2001 From: Pavel Begunkov <asml.silence(a)gmail.com> Date: Fri, 20 Jan 2023 16:38:06 +0000 Subject: [PATCH] io_uring/msg_ring: fix remote queue to disabled ring IORING_SETUP_R_DISABLED rings don't have the submitter task set, so it's not always safe to use ->submitter_task. Disallow posting msg_ring messaged to disabled rings. Also add task NULL check for loosy sync around testing for IORING_SETUP_R_DISABLED. Cc: stable(a)vger.kernel.org Fixes: 6d043ee1164ca ("io_uring: do msg_ring in target task via tw") Signed-off-by: Pavel Begunkov <asml.silence(a)gmail.com> Signed-off-by: Jens Axboe <axboe(a)kernel.dk> diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c index 2ac1cd8d23ea..0a4efada9b3c 100644 --- a/io_uring/io_uring.c +++ b/io_uring/io_uring.c @@ -3674,7 +3674,7 @@ static __cold int io_uring_create(unsigned entries, struct io_uring_params *p, if (ctx->flags & IORING_SETUP_SINGLE_ISSUER && !(ctx->flags & IORING_SETUP_R_DISABLED)) - ctx->submitter_task = get_task_struct(current); + WRITE_ONCE(ctx->submitter_task, get_task_struct(current)); file = io_uring_get_file(ctx); if (IS_ERR(file)) { @@ -3868,7 +3868,7 @@ static int io_register_enable_rings(struct io_ring_ctx *ctx) return -EBADFD; if (ctx->flags & IORING_SETUP_SINGLE_ISSUER && !ctx->submitter_task) - ctx->submitter_task = get_task_struct(current); + WRITE_ONCE(ctx->submitter_task, get_task_struct(current)); if (ctx->restrictions.registered) ctx->restricted = 1; diff --git a/io_uring/msg_ring.c b/io_uring/msg_ring.c index bb868447dcdf..15602a136821 100644 --- a/io_uring/msg_ring.c +++ b/io_uring/msg_ring.c @@ -69,6 +69,10 @@ static int io_msg_exec_remote(struct io_kiocb *req, task_work_func_t func) { struct io_ring_ctx *ctx = req->file->private_data; struct io_msg *msg = io_kiocb_to_cmd(req, struct io_msg); + struct task_struct *task = READ_ONCE(ctx->submitter_task); + + if (unlikely(!task)) + return -EOWNERDEAD; init_task_work(&msg->tw, func); if (task_work_add(ctx->submitter_task, &msg->tw, TWA_SIGNAL)) @@ -114,6 +118,8 @@ static int io_msg_ring_data(struct io_kiocb *req, unsigned int issue_flags) if (msg->src_fd || msg->dst_fd || msg->flags) return -EINVAL; + if (target_ctx->flags & IORING_SETUP_R_DISABLED) + return -EBADFD; if (io_msg_need_remote(target_ctx)) return io_msg_exec_remote(req, io_msg_tw_complete); @@ -206,6 +212,8 @@ static int io_msg_send_fd(struct io_kiocb *req, unsigned int issue_flags) if (target_ctx == ctx) return -EINVAL; + if (target_ctx->flags & IORING_SETUP_R_DISABLED) + return -EBADFD; if (!src_file) { src_file = io_msg_grab_file(req, issue_flags); if (!src_file)

2 years, 11 months

1
0
0 0

Please apply ce835dbd04d7 to 5.15

by Arınç ÜNAL

Commit ce835dbd04d7b24f9fd50d9a9c59be46304aaa8a ("staging: mt7621-dts: change some node hex addresses to lower case") upstream. Please apply this commit to 5.15. It solves the regression the kernel test robot has reported. https://lore.kernel.org/all/aa73cfb8-99ed-20b4-071c-9858399aee0a@arinc9.com… Arınç

2 years, 11 months

2
1
0 0

FAILED: patch "[PATCH] btrfs: zoned: enable metadata over-commit for non-ZNS setup" failed to apply to 6.1-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.1-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. Possible dependencies: 85e79ec7b78f ("btrfs: zoned: enable metadata over-commit for non-ZNS setup") c52cc7b7acfb ("btrfs: add a BTRFS_FS_NEED_TRANS_COMMIT flag") 7966a6b5959b ("btrfs: move fs_info::flags enum to fs.h") fc97a410bd78 ("btrfs: move mount option definitions to fs.h") 0d3a9cf8c306 ("btrfs: convert incompat and compat flag test helpers to macros") ec8eb376e271 ("btrfs: move BTRFS_FS_STATE* definitions and helpers to fs.h") 9b569ea0be6f ("btrfs: move the printk helpers out of ctree.h") e118578a8df7 ("btrfs: move assert helpers out of ctree.h") c7f13d428ea1 ("btrfs: move fs wide helpers out of ctree.h") 63a7cb130718 ("btrfs: auto enable discard=async when possible") f1e5c6185ca1 ("btrfs: move flush related definitions to space-info.h") 4300c58f8090 ("btrfs: move btrfs on-disk definitions out of ctree.h") d60d956eb41f ("btrfs: remove unused set/clear_pending_info helpers") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 85e79ec7b78f863178ca488fd8cb5b3de6347756 Mon Sep 17 00:00:00 2001 From: Naohiro Aota <naohiro.aota(a)wdc.com> Date: Tue, 10 Jan 2023 15:04:32 +0900 Subject: [PATCH] btrfs: zoned: enable metadata over-commit for non-ZNS setup The commit 79417d040f4f ("btrfs: zoned: disable metadata overcommit for zoned") disabled the metadata over-commit to track active zones properly. However, it also introduced a heavy overhead by allocating new metadata block groups and/or flushing dirty buffers to release the space reservations. Specifically, a workload (write only without any sync operations) worsen its performance from 343.77 MB/sec (v5.19) to 182.89 MB/sec (v6.0). The performance is still bad on current misc-next which is 187.95 MB/sec. And, with this patch applied, it improves back to 326.70 MB/sec (+73.82%). This patch introduces a new fs_info->flag BTRFS_FS_NO_OVERCOMMIT to indicate it needs to disable the metadata over-commit. The flag is enabled when a device with max active zones limit is loaded into a file-system. Fixes: 79417d040f4f ("btrfs: zoned: disable metadata overcommit for zoned") CC: stable(a)vger.kernel.org # 6.0+ Reviewed-by: Johannes Thumshirn <johannes.thumshirn(a)wdc.com> Signed-off-by: Naohiro Aota <naohiro.aota(a)wdc.com> Signed-off-by: David Sterba <dsterba(a)suse.com> diff --git a/fs/btrfs/fs.h b/fs/btrfs/fs.h index a749367e5ae2..37b86acfcbcf 100644 --- a/fs/btrfs/fs.h +++ b/fs/btrfs/fs.h @@ -119,6 +119,12 @@ enum { /* Indicate that we want to commit the transaction. */ BTRFS_FS_NEED_TRANS_COMMIT, + /* + * Indicate metadata over-commit is disabled. This is set when active + * zone tracking is needed. + */ + BTRFS_FS_NO_OVERCOMMIT, + #if BITS_PER_LONG == 32 /* Indicate if we have error/warn message printed on 32bit systems */ BTRFS_FS_32BIT_ERROR, diff --git a/fs/btrfs/space-info.c b/fs/btrfs/space-info.c index d28ee4e36f3d..69c09508afb5 100644 --- a/fs/btrfs/space-info.c +++ b/fs/btrfs/space-info.c @@ -407,7 +407,8 @@ int btrfs_can_overcommit(struct btrfs_fs_info *fs_info, return 0; used = btrfs_space_info_used(space_info, true); - if (btrfs_is_zoned(fs_info) && (space_info->flags & BTRFS_BLOCK_GROUP_METADATA)) + if (test_bit(BTRFS_FS_NO_OVERCOMMIT, &fs_info->flags) && + (space_info->flags & BTRFS_BLOCK_GROUP_METADATA)) avail = 0; else avail = calc_available_free_space(fs_info, space_info, flush); diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c index a759668477bb..1f503e8e42d4 100644 --- a/fs/btrfs/zoned.c +++ b/fs/btrfs/zoned.c @@ -539,6 +539,8 @@ int btrfs_get_dev_zone_info(struct btrfs_device *device, bool populate_cache) } atomic_set(&zone_info->active_zones_left, max_active_zones - nactive); + /* Overcommit does not work well with active zone tacking. */ + set_bit(BTRFS_FS_NO_OVERCOMMIT, &fs_info->flags); } /* Validate superblock log */

2 years, 11 months

1
0
0 0

FAILED: patch "[PATCH] btrfs: fix directory logging due to race with concurrent" failed to apply to 4.14-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.14-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. Possible dependencies: 8bb6898da627 ("btrfs: fix directory logging due to race with concurrent index key deletion") 6d3d970b2735 ("btrfs: fix missing error handling when logging directory items") 732d591a5d6c ("btrfs: stop copying old dir items when logging a directory") a450a4af7433 ("btrfs: don't log unnecessary boundary keys when logging directory") 339d03542484 ("btrfs: only copy dir index keys when logging a directory") 1b2e5e5c7fea ("btrfs: fix missing last dir item offset update when logging directory") 9798ba24cb76 ("btrfs: remove root argument from drop_one_dir_item()") dc2872247ec0 ("btrfs: keep track of the last logged keys when logging a directory") 086dcbfa50d3 ("btrfs: insert items in batches when logging a directory when possible") eb10d85ee77f ("btrfs: factor out the copying loop of dir items from log_dir_items()") 90d04510a774 ("btrfs: remove root argument from btrfs_log_inode() and its callees") 289cffcb0399 ("btrfs: remove no longer needed checks for NULL log context") cfd312695b71 ("btrfs: check for error when looking up inode during dir entry replay") 8dcbc26194eb ("btrfs: unify lookup return value when dir entry is missing") 52db77791fe2 ("btrfs: deal with errors when adding inode reference during log replay") e15ac6413745 ("btrfs: deal with errors when replaying dir entry during log replay") 77a5b9e3d14c ("btrfs: deal with errors when checking if a dir entry exists during log replay") a7d1c5dc8632 ("btrfs: introduce btrfs_lookup_match_dir") b590b839720c ("btrfs: avoid unnecessary logging of xattrs during fast fsyncs") 54a40fc3a1da ("btrfs: fix removed dentries still existing after log is synced") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 8bb6898da6271d82d8e76d8088d66b971a7dcfa6 Mon Sep 17 00:00:00 2001 From: Filipe Manana <fdmanana(a)suse.com> Date: Tue, 10 Jan 2023 14:56:35 +0000 Subject: [PATCH] btrfs: fix directory logging due to race with concurrent index key deletion Sometimes we log a directory without holding its VFS lock, so while we logging it, dir index entries may be added or removed. This typically happens when logging a dentry from a parent directory that points to a new directory, through log_new_dir_dentries(), or when while logging some other inode we also need to log its parent directories (through btrfs_log_all_parents()). This means that while we are at log_dir_items(), we may not find a dir index key we found before, because it was deleted in the meanwhile, so a call to btrfs_search_slot() may return 1 (key not found). In that case we return from log_dir_items() with a success value (the variable 'err' has a value of 0). This can lead to a few problems, specially in the case where the variable 'last_offset' has a value of (u64)-1 (and it's initialized to that when it was declared): 1) By returning from log_dir_items() with success (0) and a value of (u64)-1 for '*last_offset_ret', we end up not logging any other dir index keys that follow the missing, just deleted, index key. The (u64)-1 value makes log_directory_changes() not call log_dir_items() again; 2) Before returning with success (0), log_dir_items(), will log a dir index range item covering a range from the last old dentry index (stored in the variable 'last_old_dentry_offset') to the value of 'last_offset'. If 'last_offset' has a value of (u64)-1, then it means if the log is persisted and replayed after a power failure, it will cause deletion of all the directory entries that have an index number between last_old_dentry_offset + 1 and (u64)-1; 3) We can end up returning from log_dir_items() with ctx->last_dir_item_offset having a lower value than inode->last_dir_index_offset, because the former is set to the current key we are processing at process_dir_items_leaf(), and at the end of log_directory_changes() we set inode->last_dir_index_offset to the current value of ctx->last_dir_item_offset. So if for example a deletion of a lower dir index key happened, we set ctx->last_dir_item_offset to that index value, then if we return from log_dir_items() because btrfs_search_slot() returned 1, we end up returning from log_dir_items() with success (0) and then log_directory_changes() sets inode->last_dir_index_offset to a lower value than it had before. This can result in unpredictable and unexpected behaviour when we need to log again the directory in the same transaction, and can result in ending up with a log tree leaf that has duplicated keys, as we do batch insertions of dir index keys into a log tree. So fix this by making log_dir_items() move on to the next dir index key if it does not find the one it was looking for. Reported-by: David Arendt <admin(a)prnet.org> Link: https://lore.kernel.org/linux-btrfs/ae169fc6-f504-28f0-a098-6fa6a4dfb612@le… CC: stable(a)vger.kernel.org # 4.14+ Reviewed-by: Josef Bacik <josef(a)toxicpanda.com> Signed-off-by: Filipe Manana <fdmanana(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 3ef0266e9527..c09daab3f19e 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -3857,17 +3857,26 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, btrfs_release_path(path); /* - * Find the first key from this transaction again. See the note for - * log_new_dir_dentries, if we're logging a directory recursively we - * won't be holding its i_mutex, which means we can modify the directory - * while we're logging it. If we remove an entry between our first - * search and this search we'll not find the key again and can just - * bail. + * Find the first key from this transaction again or the one we were at + * in the loop below in case we had to reschedule. We may be logging the + * directory without holding its VFS lock, which happen when logging new + * dentries (through log_new_dir_dentries()) or in some cases when we + * need to log the parent directory of an inode. This means a dir index + * key might be deleted from the inode's root, and therefore we may not + * find it anymore. If we can't find it, just move to the next key. We + * can not bail out and ignore, because if we do that we will simply + * not log dir index keys that come after the one that was just deleted + * and we can end up logging a dir index range that ends at (u64)-1 + * (@last_offset is initialized to that), resulting in removing dir + * entries we should not remove at log replay time. */ search: ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0); + if (ret > 0) + ret = btrfs_next_item(root, path); if (ret < 0) err = ret; + /* If ret is 1, there are no more keys in the inode's root. */ if (ret != 0) goto done;

2 years, 11 months

1
0
0 0

FAILED: patch "[PATCH] btrfs: fix directory logging due to race with concurrent" failed to apply to 4.19-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.19-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. Possible dependencies: 8bb6898da627 ("btrfs: fix directory logging due to race with concurrent index key deletion") 6d3d970b2735 ("btrfs: fix missing error handling when logging directory items") 732d591a5d6c ("btrfs: stop copying old dir items when logging a directory") a450a4af7433 ("btrfs: don't log unnecessary boundary keys when logging directory") 339d03542484 ("btrfs: only copy dir index keys when logging a directory") 1b2e5e5c7fea ("btrfs: fix missing last dir item offset update when logging directory") 9798ba24cb76 ("btrfs: remove root argument from drop_one_dir_item()") dc2872247ec0 ("btrfs: keep track of the last logged keys when logging a directory") 086dcbfa50d3 ("btrfs: insert items in batches when logging a directory when possible") eb10d85ee77f ("btrfs: factor out the copying loop of dir items from log_dir_items()") 90d04510a774 ("btrfs: remove root argument from btrfs_log_inode() and its callees") 289cffcb0399 ("btrfs: remove no longer needed checks for NULL log context") cfd312695b71 ("btrfs: check for error when looking up inode during dir entry replay") 8dcbc26194eb ("btrfs: unify lookup return value when dir entry is missing") 52db77791fe2 ("btrfs: deal with errors when adding inode reference during log replay") e15ac6413745 ("btrfs: deal with errors when replaying dir entry during log replay") 77a5b9e3d14c ("btrfs: deal with errors when checking if a dir entry exists during log replay") a7d1c5dc8632 ("btrfs: introduce btrfs_lookup_match_dir") b590b839720c ("btrfs: avoid unnecessary logging of xattrs during fast fsyncs") 54a40fc3a1da ("btrfs: fix removed dentries still existing after log is synced") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 8bb6898da6271d82d8e76d8088d66b971a7dcfa6 Mon Sep 17 00:00:00 2001 From: Filipe Manana <fdmanana(a)suse.com> Date: Tue, 10 Jan 2023 14:56:35 +0000 Subject: [PATCH] btrfs: fix directory logging due to race with concurrent index key deletion Sometimes we log a directory without holding its VFS lock, so while we logging it, dir index entries may be added or removed. This typically happens when logging a dentry from a parent directory that points to a new directory, through log_new_dir_dentries(), or when while logging some other inode we also need to log its parent directories (through btrfs_log_all_parents()). This means that while we are at log_dir_items(), we may not find a dir index key we found before, because it was deleted in the meanwhile, so a call to btrfs_search_slot() may return 1 (key not found). In that case we return from log_dir_items() with a success value (the variable 'err' has a value of 0). This can lead to a few problems, specially in the case where the variable 'last_offset' has a value of (u64)-1 (and it's initialized to that when it was declared): 1) By returning from log_dir_items() with success (0) and a value of (u64)-1 for '*last_offset_ret', we end up not logging any other dir index keys that follow the missing, just deleted, index key. The (u64)-1 value makes log_directory_changes() not call log_dir_items() again; 2) Before returning with success (0), log_dir_items(), will log a dir index range item covering a range from the last old dentry index (stored in the variable 'last_old_dentry_offset') to the value of 'last_offset'. If 'last_offset' has a value of (u64)-1, then it means if the log is persisted and replayed after a power failure, it will cause deletion of all the directory entries that have an index number between last_old_dentry_offset + 1 and (u64)-1; 3) We can end up returning from log_dir_items() with ctx->last_dir_item_offset having a lower value than inode->last_dir_index_offset, because the former is set to the current key we are processing at process_dir_items_leaf(), and at the end of log_directory_changes() we set inode->last_dir_index_offset to the current value of ctx->last_dir_item_offset. So if for example a deletion of a lower dir index key happened, we set ctx->last_dir_item_offset to that index value, then if we return from log_dir_items() because btrfs_search_slot() returned 1, we end up returning from log_dir_items() with success (0) and then log_directory_changes() sets inode->last_dir_index_offset to a lower value than it had before. This can result in unpredictable and unexpected behaviour when we need to log again the directory in the same transaction, and can result in ending up with a log tree leaf that has duplicated keys, as we do batch insertions of dir index keys into a log tree. So fix this by making log_dir_items() move on to the next dir index key if it does not find the one it was looking for. Reported-by: David Arendt <admin(a)prnet.org> Link: https://lore.kernel.org/linux-btrfs/ae169fc6-f504-28f0-a098-6fa6a4dfb612@le… CC: stable(a)vger.kernel.org # 4.14+ Reviewed-by: Josef Bacik <josef(a)toxicpanda.com> Signed-off-by: Filipe Manana <fdmanana(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 3ef0266e9527..c09daab3f19e 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -3857,17 +3857,26 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, btrfs_release_path(path); /* - * Find the first key from this transaction again. See the note for - * log_new_dir_dentries, if we're logging a directory recursively we - * won't be holding its i_mutex, which means we can modify the directory - * while we're logging it. If we remove an entry between our first - * search and this search we'll not find the key again and can just - * bail. + * Find the first key from this transaction again or the one we were at + * in the loop below in case we had to reschedule. We may be logging the + * directory without holding its VFS lock, which happen when logging new + * dentries (through log_new_dir_dentries()) or in some cases when we + * need to log the parent directory of an inode. This means a dir index + * key might be deleted from the inode's root, and therefore we may not + * find it anymore. If we can't find it, just move to the next key. We + * can not bail out and ignore, because if we do that we will simply + * not log dir index keys that come after the one that was just deleted + * and we can end up logging a dir index range that ends at (u64)-1 + * (@last_offset is initialized to that), resulting in removing dir + * entries we should not remove at log replay time. */ search: ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0); + if (ret > 0) + ret = btrfs_next_item(root, path); if (ret < 0) err = ret; + /* If ret is 1, there are no more keys in the inode's root. */ if (ret != 0) goto done;

2 years, 11 months

1
0
0 0

FAILED: patch "[PATCH] btrfs: fix directory logging due to race with concurrent" failed to apply to 5.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. Possible dependencies: 8bb6898da627 ("btrfs: fix directory logging due to race with concurrent index key deletion") 6d3d970b2735 ("btrfs: fix missing error handling when logging directory items") 732d591a5d6c ("btrfs: stop copying old dir items when logging a directory") a450a4af7433 ("btrfs: don't log unnecessary boundary keys when logging directory") 339d03542484 ("btrfs: only copy dir index keys when logging a directory") 1b2e5e5c7fea ("btrfs: fix missing last dir item offset update when logging directory") 9798ba24cb76 ("btrfs: remove root argument from drop_one_dir_item()") dc2872247ec0 ("btrfs: keep track of the last logged keys when logging a directory") 086dcbfa50d3 ("btrfs: insert items in batches when logging a directory when possible") eb10d85ee77f ("btrfs: factor out the copying loop of dir items from log_dir_items()") 90d04510a774 ("btrfs: remove root argument from btrfs_log_inode() and its callees") 289cffcb0399 ("btrfs: remove no longer needed checks for NULL log context") cfd312695b71 ("btrfs: check for error when looking up inode during dir entry replay") 8dcbc26194eb ("btrfs: unify lookup return value when dir entry is missing") 52db77791fe2 ("btrfs: deal with errors when adding inode reference during log replay") e15ac6413745 ("btrfs: deal with errors when replaying dir entry during log replay") 77a5b9e3d14c ("btrfs: deal with errors when checking if a dir entry exists during log replay") a7d1c5dc8632 ("btrfs: introduce btrfs_lookup_match_dir") b590b839720c ("btrfs: avoid unnecessary logging of xattrs during fast fsyncs") 54a40fc3a1da ("btrfs: fix removed dentries still existing after log is synced") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 8bb6898da6271d82d8e76d8088d66b971a7dcfa6 Mon Sep 17 00:00:00 2001 From: Filipe Manana <fdmanana(a)suse.com> Date: Tue, 10 Jan 2023 14:56:35 +0000 Subject: [PATCH] btrfs: fix directory logging due to race with concurrent index key deletion Sometimes we log a directory without holding its VFS lock, so while we logging it, dir index entries may be added or removed. This typically happens when logging a dentry from a parent directory that points to a new directory, through log_new_dir_dentries(), or when while logging some other inode we also need to log its parent directories (through btrfs_log_all_parents()). This means that while we are at log_dir_items(), we may not find a dir index key we found before, because it was deleted in the meanwhile, so a call to btrfs_search_slot() may return 1 (key not found). In that case we return from log_dir_items() with a success value (the variable 'err' has a value of 0). This can lead to a few problems, specially in the case where the variable 'last_offset' has a value of (u64)-1 (and it's initialized to that when it was declared): 1) By returning from log_dir_items() with success (0) and a value of (u64)-1 for '*last_offset_ret', we end up not logging any other dir index keys that follow the missing, just deleted, index key. The (u64)-1 value makes log_directory_changes() not call log_dir_items() again; 2) Before returning with success (0), log_dir_items(), will log a dir index range item covering a range from the last old dentry index (stored in the variable 'last_old_dentry_offset') to the value of 'last_offset'. If 'last_offset' has a value of (u64)-1, then it means if the log is persisted and replayed after a power failure, it will cause deletion of all the directory entries that have an index number between last_old_dentry_offset + 1 and (u64)-1; 3) We can end up returning from log_dir_items() with ctx->last_dir_item_offset having a lower value than inode->last_dir_index_offset, because the former is set to the current key we are processing at process_dir_items_leaf(), and at the end of log_directory_changes() we set inode->last_dir_index_offset to the current value of ctx->last_dir_item_offset. So if for example a deletion of a lower dir index key happened, we set ctx->last_dir_item_offset to that index value, then if we return from log_dir_items() because btrfs_search_slot() returned 1, we end up returning from log_dir_items() with success (0) and then log_directory_changes() sets inode->last_dir_index_offset to a lower value than it had before. This can result in unpredictable and unexpected behaviour when we need to log again the directory in the same transaction, and can result in ending up with a log tree leaf that has duplicated keys, as we do batch insertions of dir index keys into a log tree. So fix this by making log_dir_items() move on to the next dir index key if it does not find the one it was looking for. Reported-by: David Arendt <admin(a)prnet.org> Link: https://lore.kernel.org/linux-btrfs/ae169fc6-f504-28f0-a098-6fa6a4dfb612@le… CC: stable(a)vger.kernel.org # 4.14+ Reviewed-by: Josef Bacik <josef(a)toxicpanda.com> Signed-off-by: Filipe Manana <fdmanana(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 3ef0266e9527..c09daab3f19e 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -3857,17 +3857,26 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, btrfs_release_path(path); /* - * Find the first key from this transaction again. See the note for - * log_new_dir_dentries, if we're logging a directory recursively we - * won't be holding its i_mutex, which means we can modify the directory - * while we're logging it. If we remove an entry between our first - * search and this search we'll not find the key again and can just - * bail. + * Find the first key from this transaction again or the one we were at + * in the loop below in case we had to reschedule. We may be logging the + * directory without holding its VFS lock, which happen when logging new + * dentries (through log_new_dir_dentries()) or in some cases when we + * need to log the parent directory of an inode. This means a dir index + * key might be deleted from the inode's root, and therefore we may not + * find it anymore. If we can't find it, just move to the next key. We + * can not bail out and ignore, because if we do that we will simply + * not log dir index keys that come after the one that was just deleted + * and we can end up logging a dir index range that ends at (u64)-1 + * (@last_offset is initialized to that), resulting in removing dir + * entries we should not remove at log replay time. */ search: ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0); + if (ret > 0) + ret = btrfs_next_item(root, path); if (ret < 0) err = ret; + /* If ret is 1, there are no more keys in the inode's root. */ if (ret != 0) goto done;

2 years, 11 months

1
0
0 0

FAILED: patch "[PATCH] btrfs: fix directory logging due to race with concurrent" failed to apply to 5.10-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.10-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. Possible dependencies: 8bb6898da627 ("btrfs: fix directory logging due to race with concurrent index key deletion") 6d3d970b2735 ("btrfs: fix missing error handling when logging directory items") 732d591a5d6c ("btrfs: stop copying old dir items when logging a directory") a450a4af7433 ("btrfs: don't log unnecessary boundary keys when logging directory") 339d03542484 ("btrfs: only copy dir index keys when logging a directory") 1b2e5e5c7fea ("btrfs: fix missing last dir item offset update when logging directory") 9798ba24cb76 ("btrfs: remove root argument from drop_one_dir_item()") dc2872247ec0 ("btrfs: keep track of the last logged keys when logging a directory") 086dcbfa50d3 ("btrfs: insert items in batches when logging a directory when possible") eb10d85ee77f ("btrfs: factor out the copying loop of dir items from log_dir_items()") 90d04510a774 ("btrfs: remove root argument from btrfs_log_inode() and its callees") 289cffcb0399 ("btrfs: remove no longer needed checks for NULL log context") cfd312695b71 ("btrfs: check for error when looking up inode during dir entry replay") 8dcbc26194eb ("btrfs: unify lookup return value when dir entry is missing") 52db77791fe2 ("btrfs: deal with errors when adding inode reference during log replay") e15ac6413745 ("btrfs: deal with errors when replaying dir entry during log replay") 77a5b9e3d14c ("btrfs: deal with errors when checking if a dir entry exists during log replay") a7d1c5dc8632 ("btrfs: introduce btrfs_lookup_match_dir") b590b839720c ("btrfs: avoid unnecessary logging of xattrs during fast fsyncs") 54a40fc3a1da ("btrfs: fix removed dentries still existing after log is synced") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 8bb6898da6271d82d8e76d8088d66b971a7dcfa6 Mon Sep 17 00:00:00 2001 From: Filipe Manana <fdmanana(a)suse.com> Date: Tue, 10 Jan 2023 14:56:35 +0000 Subject: [PATCH] btrfs: fix directory logging due to race with concurrent index key deletion Sometimes we log a directory without holding its VFS lock, so while we logging it, dir index entries may be added or removed. This typically happens when logging a dentry from a parent directory that points to a new directory, through log_new_dir_dentries(), or when while logging some other inode we also need to log its parent directories (through btrfs_log_all_parents()). This means that while we are at log_dir_items(), we may not find a dir index key we found before, because it was deleted in the meanwhile, so a call to btrfs_search_slot() may return 1 (key not found). In that case we return from log_dir_items() with a success value (the variable 'err' has a value of 0). This can lead to a few problems, specially in the case where the variable 'last_offset' has a value of (u64)-1 (and it's initialized to that when it was declared): 1) By returning from log_dir_items() with success (0) and a value of (u64)-1 for '*last_offset_ret', we end up not logging any other dir index keys that follow the missing, just deleted, index key. The (u64)-1 value makes log_directory_changes() not call log_dir_items() again; 2) Before returning with success (0), log_dir_items(), will log a dir index range item covering a range from the last old dentry index (stored in the variable 'last_old_dentry_offset') to the value of 'last_offset'. If 'last_offset' has a value of (u64)-1, then it means if the log is persisted and replayed after a power failure, it will cause deletion of all the directory entries that have an index number between last_old_dentry_offset + 1 and (u64)-1; 3) We can end up returning from log_dir_items() with ctx->last_dir_item_offset having a lower value than inode->last_dir_index_offset, because the former is set to the current key we are processing at process_dir_items_leaf(), and at the end of log_directory_changes() we set inode->last_dir_index_offset to the current value of ctx->last_dir_item_offset. So if for example a deletion of a lower dir index key happened, we set ctx->last_dir_item_offset to that index value, then if we return from log_dir_items() because btrfs_search_slot() returned 1, we end up returning from log_dir_items() with success (0) and then log_directory_changes() sets inode->last_dir_index_offset to a lower value than it had before. This can result in unpredictable and unexpected behaviour when we need to log again the directory in the same transaction, and can result in ending up with a log tree leaf that has duplicated keys, as we do batch insertions of dir index keys into a log tree. So fix this by making log_dir_items() move on to the next dir index key if it does not find the one it was looking for. Reported-by: David Arendt <admin(a)prnet.org> Link: https://lore.kernel.org/linux-btrfs/ae169fc6-f504-28f0-a098-6fa6a4dfb612@le… CC: stable(a)vger.kernel.org # 4.14+ Reviewed-by: Josef Bacik <josef(a)toxicpanda.com> Signed-off-by: Filipe Manana <fdmanana(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 3ef0266e9527..c09daab3f19e 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -3857,17 +3857,26 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, btrfs_release_path(path); /* - * Find the first key from this transaction again. See the note for - * log_new_dir_dentries, if we're logging a directory recursively we - * won't be holding its i_mutex, which means we can modify the directory - * while we're logging it. If we remove an entry between our first - * search and this search we'll not find the key again and can just - * bail. + * Find the first key from this transaction again or the one we were at + * in the loop below in case we had to reschedule. We may be logging the + * directory without holding its VFS lock, which happen when logging new + * dentries (through log_new_dir_dentries()) or in some cases when we + * need to log the parent directory of an inode. This means a dir index + * key might be deleted from the inode's root, and therefore we may not + * find it anymore. If we can't find it, just move to the next key. We + * can not bail out and ignore, because if we do that we will simply + * not log dir index keys that come after the one that was just deleted + * and we can end up logging a dir index range that ends at (u64)-1 + * (@last_offset is initialized to that), resulting in removing dir + * entries we should not remove at log replay time. */ search: ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0); + if (ret > 0) + ret = btrfs_next_item(root, path); if (ret < 0) err = ret; + /* If ret is 1, there are no more keys in the inode's root. */ if (ret != 0) goto done;

2 years, 11 months

1
0
0 0

FAILED: patch "[PATCH] btrfs: fix directory logging due to race with concurrent" failed to apply to 5.15-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.15-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. Possible dependencies: 8bb6898da627 ("btrfs: fix directory logging due to race with concurrent index key deletion") 6d3d970b2735 ("btrfs: fix missing error handling when logging directory items") 732d591a5d6c ("btrfs: stop copying old dir items when logging a directory") a450a4af7433 ("btrfs: don't log unnecessary boundary keys when logging directory") 339d03542484 ("btrfs: only copy dir index keys when logging a directory") 1b2e5e5c7fea ("btrfs: fix missing last dir item offset update when logging directory") 9798ba24cb76 ("btrfs: remove root argument from drop_one_dir_item()") dc2872247ec0 ("btrfs: keep track of the last logged keys when logging a directory") 086dcbfa50d3 ("btrfs: insert items in batches when logging a directory when possible") eb10d85ee77f ("btrfs: factor out the copying loop of dir items from log_dir_items()") 90d04510a774 ("btrfs: remove root argument from btrfs_log_inode() and its callees") 289cffcb0399 ("btrfs: remove no longer needed checks for NULL log context") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 8bb6898da6271d82d8e76d8088d66b971a7dcfa6 Mon Sep 17 00:00:00 2001 From: Filipe Manana <fdmanana(a)suse.com> Date: Tue, 10 Jan 2023 14:56:35 +0000 Subject: [PATCH] btrfs: fix directory logging due to race with concurrent index key deletion Sometimes we log a directory without holding its VFS lock, so while we logging it, dir index entries may be added or removed. This typically happens when logging a dentry from a parent directory that points to a new directory, through log_new_dir_dentries(), or when while logging some other inode we also need to log its parent directories (through btrfs_log_all_parents()). This means that while we are at log_dir_items(), we may not find a dir index key we found before, because it was deleted in the meanwhile, so a call to btrfs_search_slot() may return 1 (key not found). In that case we return from log_dir_items() with a success value (the variable 'err' has a value of 0). This can lead to a few problems, specially in the case where the variable 'last_offset' has a value of (u64)-1 (and it's initialized to that when it was declared): 1) By returning from log_dir_items() with success (0) and a value of (u64)-1 for '*last_offset_ret', we end up not logging any other dir index keys that follow the missing, just deleted, index key. The (u64)-1 value makes log_directory_changes() not call log_dir_items() again; 2) Before returning with success (0), log_dir_items(), will log a dir index range item covering a range from the last old dentry index (stored in the variable 'last_old_dentry_offset') to the value of 'last_offset'. If 'last_offset' has a value of (u64)-1, then it means if the log is persisted and replayed after a power failure, it will cause deletion of all the directory entries that have an index number between last_old_dentry_offset + 1 and (u64)-1; 3) We can end up returning from log_dir_items() with ctx->last_dir_item_offset having a lower value than inode->last_dir_index_offset, because the former is set to the current key we are processing at process_dir_items_leaf(), and at the end of log_directory_changes() we set inode->last_dir_index_offset to the current value of ctx->last_dir_item_offset. So if for example a deletion of a lower dir index key happened, we set ctx->last_dir_item_offset to that index value, then if we return from log_dir_items() because btrfs_search_slot() returned 1, we end up returning from log_dir_items() with success (0) and then log_directory_changes() sets inode->last_dir_index_offset to a lower value than it had before. This can result in unpredictable and unexpected behaviour when we need to log again the directory in the same transaction, and can result in ending up with a log tree leaf that has duplicated keys, as we do batch insertions of dir index keys into a log tree. So fix this by making log_dir_items() move on to the next dir index key if it does not find the one it was looking for. Reported-by: David Arendt <admin(a)prnet.org> Link: https://lore.kernel.org/linux-btrfs/ae169fc6-f504-28f0-a098-6fa6a4dfb612@le… CC: stable(a)vger.kernel.org # 4.14+ Reviewed-by: Josef Bacik <josef(a)toxicpanda.com> Signed-off-by: Filipe Manana <fdmanana(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 3ef0266e9527..c09daab3f19e 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -3857,17 +3857,26 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, btrfs_release_path(path); /* - * Find the first key from this transaction again. See the note for - * log_new_dir_dentries, if we're logging a directory recursively we - * won't be holding its i_mutex, which means we can modify the directory - * while we're logging it. If we remove an entry between our first - * search and this search we'll not find the key again and can just - * bail. + * Find the first key from this transaction again or the one we were at + * in the loop below in case we had to reschedule. We may be logging the + * directory without holding its VFS lock, which happen when logging new + * dentries (through log_new_dir_dentries()) or in some cases when we + * need to log the parent directory of an inode. This means a dir index + * key might be deleted from the inode's root, and therefore we may not + * find it anymore. If we can't find it, just move to the next key. We + * can not bail out and ignore, because if we do that we will simply + * not log dir index keys that come after the one that was just deleted + * and we can end up logging a dir index range that ends at (u64)-1 + * (@last_offset is initialized to that), resulting in removing dir + * entries we should not remove at log replay time. */ search: ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0); + if (ret > 0) + ret = btrfs_next_item(root, path); if (ret < 0) err = ret; + /* If ret is 1, there are no more keys in the inode's root. */ if (ret != 0) goto done;

2 years, 11 months

1
0
0 0

FAILED: patch "[PATCH] btrfs: fix missing error handling when logging directory" failed to apply to 4.14-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.14-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. Possible dependencies: 6d3d970b2735 ("btrfs: fix missing error handling when logging directory items") 732d591a5d6c ("btrfs: stop copying old dir items when logging a directory") a450a4af7433 ("btrfs: don't log unnecessary boundary keys when logging directory") 339d03542484 ("btrfs: only copy dir index keys when logging a directory") 1b2e5e5c7fea ("btrfs: fix missing last dir item offset update when logging directory") 9798ba24cb76 ("btrfs: remove root argument from drop_one_dir_item()") dc2872247ec0 ("btrfs: keep track of the last logged keys when logging a directory") 086dcbfa50d3 ("btrfs: insert items in batches when logging a directory when possible") eb10d85ee77f ("btrfs: factor out the copying loop of dir items from log_dir_items()") 90d04510a774 ("btrfs: remove root argument from btrfs_log_inode() and its callees") 289cffcb0399 ("btrfs: remove no longer needed checks for NULL log context") cfd312695b71 ("btrfs: check for error when looking up inode during dir entry replay") 8dcbc26194eb ("btrfs: unify lookup return value when dir entry is missing") 52db77791fe2 ("btrfs: deal with errors when adding inode reference during log replay") e15ac6413745 ("btrfs: deal with errors when replaying dir entry during log replay") 77a5b9e3d14c ("btrfs: deal with errors when checking if a dir entry exists during log replay") a7d1c5dc8632 ("btrfs: introduce btrfs_lookup_match_dir") b590b839720c ("btrfs: avoid unnecessary logging of xattrs during fast fsyncs") 54a40fc3a1da ("btrfs: fix removed dentries still existing after log is synced") 64d6b281ba4d ("btrfs: remove unnecessary check_parent_dirs_for_sync()") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 6d3d970b2735b967650d319be27268fedc5598d1 Mon Sep 17 00:00:00 2001 From: Filipe Manana <fdmanana(a)suse.com> Date: Tue, 10 Jan 2023 14:56:34 +0000 Subject: [PATCH] btrfs: fix missing error handling when logging directory items When logging a directory, at log_dir_items(), if we get an error when attempting to search the subvolume tree for a dir index item, we end up returning 0 (success) from log_dir_items() because 'err' is left with a value of 0. This can lead to a few problems, specially in the case the variable 'last_offset' has a value of (u64)-1 (and it's initialized to that when it was declared): 1) By returning from log_dir_items() with success (0) and a value of (u64)-1 for '*last_offset_ret', we end up not logging any other dir index keys that follow the missing, just deleted, index key. The (u64)-1 value makes log_directory_changes() not call log_dir_items() again; 2) Before returning with success (0), log_dir_items(), will log a dir index range item covering a range from the last old dentry index (stored in the variable 'last_old_dentry_offset') to the value of 'last_offset'. If 'last_offset' has a value of (u64)-1, then it means if the log is persisted and replayed after a power failure, it will cause deletion of all the directory entries that have an index number between last_old_dentry_offset + 1 and (u64)-1; 3) We can end up returning from log_dir_items() with ctx->last_dir_item_offset having a lower value than inode->last_dir_index_offset, because the former is set to the current key we are processing at process_dir_items_leaf(), and at the end of log_directory_changes() we set inode->last_dir_index_offset to the current value of ctx->last_dir_item_offset. So if for example a deletion of a lower dir index key happened, we set ctx->last_dir_item_offset to that index value, then if we return from log_dir_items() because btrfs_search_slot() returned an error, we end up returning without any error from log_dir_items() and then log_directory_changes() sets inode->last_dir_index_offset to a lower value than it had before. This can result in unpredictable and unexpected behaviour when we need to log again the directory in the same transaction, and can result in ending up with a log tree leaf that has duplicated keys, as we do batch insertions of dir index keys into a log tree. Fix this by setting 'err' to the value of 'ret' in case btrfs_search_slot() or btrfs_previous_item() returned an error. That will result in falling back to a full transaction commit. Reported-by: David Arendt <admin(a)prnet.org> Link: https://lore.kernel.org/linux-btrfs/ae169fc6-f504-28f0-a098-6fa6a4dfb612@le… Fixes: e02119d5a7b4 ("Btrfs: Add a write ahead tree log to optimize synchronous operations") CC: stable(a)vger.kernel.org # 4.14+ Reviewed-by: Josef Bacik <josef(a)toxicpanda.com> Signed-off-by: Filipe Manana <fdmanana(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index fb52aa060093..3ef0266e9527 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -3826,7 +3826,10 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, path->slots[0]); if (tmp.type == BTRFS_DIR_INDEX_KEY) last_old_dentry_offset = tmp.offset; + } else if (ret < 0) { + err = ret; } + goto done; } @@ -3846,7 +3849,11 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, */ if (tmp.type == BTRFS_DIR_INDEX_KEY) last_old_dentry_offset = tmp.offset; + } else if (ret < 0) { + err = ret; + goto done; } + btrfs_release_path(path); /* @@ -3859,6 +3866,8 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, */ search: ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0); + if (ret < 0) + err = ret; if (ret != 0) goto done;

2 years, 11 months

1
0
0 0

FAILED: patch "[PATCH] btrfs: fix missing error handling when logging directory" failed to apply to 4.19-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.19-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. Possible dependencies: 6d3d970b2735 ("btrfs: fix missing error handling when logging directory items") 732d591a5d6c ("btrfs: stop copying old dir items when logging a directory") a450a4af7433 ("btrfs: don't log unnecessary boundary keys when logging directory") 339d03542484 ("btrfs: only copy dir index keys when logging a directory") 1b2e5e5c7fea ("btrfs: fix missing last dir item offset update when logging directory") 9798ba24cb76 ("btrfs: remove root argument from drop_one_dir_item()") dc2872247ec0 ("btrfs: keep track of the last logged keys when logging a directory") 086dcbfa50d3 ("btrfs: insert items in batches when logging a directory when possible") eb10d85ee77f ("btrfs: factor out the copying loop of dir items from log_dir_items()") 90d04510a774 ("btrfs: remove root argument from btrfs_log_inode() and its callees") 289cffcb0399 ("btrfs: remove no longer needed checks for NULL log context") cfd312695b71 ("btrfs: check for error when looking up inode during dir entry replay") 8dcbc26194eb ("btrfs: unify lookup return value when dir entry is missing") 52db77791fe2 ("btrfs: deal with errors when adding inode reference during log replay") e15ac6413745 ("btrfs: deal with errors when replaying dir entry during log replay") 77a5b9e3d14c ("btrfs: deal with errors when checking if a dir entry exists during log replay") a7d1c5dc8632 ("btrfs: introduce btrfs_lookup_match_dir") b590b839720c ("btrfs: avoid unnecessary logging of xattrs during fast fsyncs") 54a40fc3a1da ("btrfs: fix removed dentries still existing after log is synced") 64d6b281ba4d ("btrfs: remove unnecessary check_parent_dirs_for_sync()") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 6d3d970b2735b967650d319be27268fedc5598d1 Mon Sep 17 00:00:00 2001 From: Filipe Manana <fdmanana(a)suse.com> Date: Tue, 10 Jan 2023 14:56:34 +0000 Subject: [PATCH] btrfs: fix missing error handling when logging directory items When logging a directory, at log_dir_items(), if we get an error when attempting to search the subvolume tree for a dir index item, we end up returning 0 (success) from log_dir_items() because 'err' is left with a value of 0. This can lead to a few problems, specially in the case the variable 'last_offset' has a value of (u64)-1 (and it's initialized to that when it was declared): 1) By returning from log_dir_items() with success (0) and a value of (u64)-1 for '*last_offset_ret', we end up not logging any other dir index keys that follow the missing, just deleted, index key. The (u64)-1 value makes log_directory_changes() not call log_dir_items() again; 2) Before returning with success (0), log_dir_items(), will log a dir index range item covering a range from the last old dentry index (stored in the variable 'last_old_dentry_offset') to the value of 'last_offset'. If 'last_offset' has a value of (u64)-1, then it means if the log is persisted and replayed after a power failure, it will cause deletion of all the directory entries that have an index number between last_old_dentry_offset + 1 and (u64)-1; 3) We can end up returning from log_dir_items() with ctx->last_dir_item_offset having a lower value than inode->last_dir_index_offset, because the former is set to the current key we are processing at process_dir_items_leaf(), and at the end of log_directory_changes() we set inode->last_dir_index_offset to the current value of ctx->last_dir_item_offset. So if for example a deletion of a lower dir index key happened, we set ctx->last_dir_item_offset to that index value, then if we return from log_dir_items() because btrfs_search_slot() returned an error, we end up returning without any error from log_dir_items() and then log_directory_changes() sets inode->last_dir_index_offset to a lower value than it had before. This can result in unpredictable and unexpected behaviour when we need to log again the directory in the same transaction, and can result in ending up with a log tree leaf that has duplicated keys, as we do batch insertions of dir index keys into a log tree. Fix this by setting 'err' to the value of 'ret' in case btrfs_search_slot() or btrfs_previous_item() returned an error. That will result in falling back to a full transaction commit. Reported-by: David Arendt <admin(a)prnet.org> Link: https://lore.kernel.org/linux-btrfs/ae169fc6-f504-28f0-a098-6fa6a4dfb612@le… Fixes: e02119d5a7b4 ("Btrfs: Add a write ahead tree log to optimize synchronous operations") CC: stable(a)vger.kernel.org # 4.14+ Reviewed-by: Josef Bacik <josef(a)toxicpanda.com> Signed-off-by: Filipe Manana <fdmanana(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index fb52aa060093..3ef0266e9527 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -3826,7 +3826,10 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, path->slots[0]); if (tmp.type == BTRFS_DIR_INDEX_KEY) last_old_dentry_offset = tmp.offset; + } else if (ret < 0) { + err = ret; } + goto done; } @@ -3846,7 +3849,11 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, */ if (tmp.type == BTRFS_DIR_INDEX_KEY) last_old_dentry_offset = tmp.offset; + } else if (ret < 0) { + err = ret; + goto done; } + btrfs_release_path(path); /* @@ -3859,6 +3866,8 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, */ search: ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0); + if (ret < 0) + err = ret; if (ret != 0) goto done;

2 years, 11 months

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror January 2023