summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2024-09-28bcachefs: memset bounce buffer portion to 0 after key_sort_fix_overlappingPiotr Zalewski1-0/+4
Zero-initialize part of allocated bounce buffer which wasn't touched by subsequent bch2_key_sort_fix_overlapping to mitigate later uinit-value use KMSAN bug[1]. After applying the patch reproducer still triggers stack overflow[2] but it seems unrelated to the uninit-value use warning. After further investigation it was found that stack overflow occurs because KMSAN adds too many function calls[3]. Backtrace of where the stack magic number gets smashed was added as a reply to syzkaller thread[3]. It was confirmed that task's stack magic number gets smashed after the code path where KSMAN detects uninit-value use is executed, so it can be assumed that it doesn't contribute in any way to uninit-value use detection. [1] https://syzkaller.appspot.com/bug?extid=6f655a60d3244d0c6718 [2] https://lore.kernel.org/lkml/66e57e46.050a0220.115905.0002.GAE@google.com [3] https://lore.kernel.org/all/rVaWgPULej8K7HqMPNIu8kVNyXNjjCiTB-QBtItLFBmk0alH6fV2tk4joVPk97Evnuv4ZRDd8HB5uDCkiFG6u81xKdzDj-KrtIMJSlF6Kt8=@proton.me Reported-by: syzbot+6f655a60d3244d0c6718@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=6f655a60d3244d0c6718 Fixes: ec4edd7b9d20 ("bcachefs: Prep work for variable size btree node buffers") Suggested-by: Kent Overstreet <kent.overstreet@linux.dev> Signed-off-by: Piotr Zalewski <pZ010001011111@proton.me> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-28bcachefs: Improve bch2_is_inode_open() warning messageKent Overstreet1-3/+3
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-28bcachefs: Add extra padding in bkey_make_mut_noupdate()Kent Overstreet1-1/+2
This fixes a kasan splat in propagate_key_to_snapshot_leaves() - varint_decode_fast() does reads (that it never uses) up to 7 bytes past the end of the integer. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-28bcachefs: Mark inode errors as autofixKent Overstreet1-16/+16
Most or all errors will be autofix in the future, we're currently just doing the ones that we know are well tested. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-24bcachefs: Fix infinite loop in propagate_key_to_snapshot_leaves()Kent Overstreet1-0/+1
As we iterate we need to mark that we no longer need iterators - otherwise we'll infinite loop via the "too many iters" check when there's many snapshots. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-24bcachefs: Ensure BCH_FS_accounting_replay_done is always setKent Overstreet1-0/+3
if it doesn't get set we'll never be able to flush the btree write buffer; this only happens in fake rw mode, but prevents us from shutting down. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Hold read lock in bch2_snapshot_tree_oldest_subvol()Ahmed Ehab1-0/+2
Syzbot reports a problem that a warning is triggered due to suspicious use of rcu_dereference_check(). That is triggered by a call of bch2_snapshot_tree_oldest_subvol(). The cause of the warning is that inside bch2_snapshot_tree_oldest_subvol(), snapshot_t() is called which calls rcu_dereference() that requires a read lock to be held. Also, the call of bch2_snapshot_tree_next() eventually calls snapshot_t(). To fix this, call rcu_read_lock() before calling snapshot_t(). Then, release the lock after the termination of the while loop. Reported-by: <syzbot+f7c41a878676b72c16a6@syzkaller.appspotmail.com> Signed-off-by: Ahmed Ehab <bottaawesome633@gmail.com> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: return err ptr instead of null in read sb cleanDiogo Jahchan Koike1-1/+1
syzbot reported a null-ptr-deref in bch2_fs_start. [0] When a sb is marked clear but doesn't have a clean section bch2_read_superblock_clean returns NULL which PTR_ERR_OR_ZERO lets through, eventually leading to a null ptr dereference down the line. Adjust read sb clean to return an ERR_PTR indicating the invalid clean section. [0] https://syzkaller.appspot.com/bug?extid=1cecc37d87c4286e5543 Reported-by: syzbot+1cecc37d87c4286e5543@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=1cecc37d87c4286e5543 Signed-off-by: Diogo Jahchan Koike <djahchankoike@gmail.com> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Remove duplicated include in backpointers.cYang Li1-1/+0
The header files bbpos.h is included twice in backpointers.c, so one inclusion of each can be removed. Reported-by: Abaci Robot <abaci@linux.alibaba.com> Closes: https://bugzilla.openanolis.cn/show_bug.cgi?id=10783 Signed-off-by: Yang Li <yang.lee@linux.alibaba.com> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Don't drop devices with stripe pointersKent Overstreet4-9/+32
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: bch2_ec_stripe_head_get() now checks for change in rw devicesKent Overstreet2-27/+60
This factors out ec_strie_head_devs_update(), which initializes the bitmap of devices we're allocating from, and runs it every time c->rw_devs_change_count changes. We also cancel pending, not allocated stripes, since they may refer to devices that are no longer available. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: bch_fs.rw_devs_change_countKent Overstreet2-4/+9
Add a counter that's incremented whenever rw devices change; this will be used for erasure coding so that it can keep ec_stripe_head in sync and not deadlock on a new stripe when a device it wants goes away. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: bch2_dev_remove_stripes()Kent Overstreet4-3/+74
We can now correctly force-remove a device that has stripes on it; this uses the new BCH_SB_MEMBER_INVALID sentinal value. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: bch2_trigger_ptr() calculates sectors even when no deviceKent Overstreet2-10/+21
This is necessary for erasure coded pointers to devices that have been removed. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: improve error messages in bch2_ec_read_extent()Kent Overstreet3-19/+23
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: improve error message on too few devices for ecKent Overstreet1-3/+16
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: improve bch2_new_stripe_to_text()Kent Overstreet1-0/+2
also print out the new stripe key Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: ec_stripe_head.nr_createdKent Overstreet2-2/+6
additional debug stat Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: bch_stripe.disk_labelKent Overstreet4-16/+43
When reshaping existing stripes, we should keep them on the same target that they were allocated on; to do this, we need to add a field to the btree stripe type. This is a tad awkward, because we only have 8 bits left, and targets are 16 bits - but we only need to store a label, not a full target. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: stripe_to_mem()Kent Overstreet1-18/+15
factor out a common helper Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: EIO errcode cleanupKent Overstreet5-27/+33
We want to be using private errcodes whenever possible, for better error messages. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Rework btree node pinningKent Overstreet7-75/+150
In backpointers fsck, we do a seqential scan of one btree, and check references to another: extents <-> backpointers Checking references generates random lookups, so we want to pin that btree in memory (or only a range, if it doesn't fit in ram). Previously, this was done with a simple check in the shrinker - "if btree node is in range being pinned, don't free it" - but this generated OOMs, as our shrinker wasn't well behaved if there was less memory available than expected. Instead, we now have two different shrinkers and lru lists; the second shrinker being for pinned nodes, with seeks set much higher than normal - so they can still be freed if necessary, but we'll prefer not to. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: split up btree cache counters for live, freeableKent Overstreet6-32/+47
this is prep for introducing a second live list and shrinker for pinned nodes Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: btree cache counters should be size_tKent Overstreet6-36/+37
32 bits won't overflow any time soon, but size_t is the correct type for counting objects in memory. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Don't count "skipped access bit" as touched in btree cache scanKent Overstreet1-0/+1
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Failed devices no longer require mounting in degraded modeKent Overstreet1-1/+1
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: bch2_dev_rcu_noerror()Kent Overstreet6-13/+22
bch2_dev_rcu() now properly errors if the device is invalid Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Progress indicator for extents_to_backpointersKent Overstreet1-6/+82
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: bch2_opts_to_text()Kent Overstreet3-21/+35
Factor out bch2_show_options() into a generic helper, for debugging option passing issues. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: improve "no device to read from" messageKent Overstreet1-1/+7
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Fix compilation error for bch2_sb_member_allocHongbo Li1-6/+10
Fix the following compilation error: ``` fs/bcachefs/sb-members.c: In function ‘bch2_sb_member_alloc’: fs/bcachefs/sb-members.c:508:2: error: a label can only be part of a statement and a declaration is not a statement 508 | unsigned nr_devices = max_t(unsigned, dev_idx + 1, c->sb.nr_devices); ``` Fixes: a7d364a133c7 ("bcachefs: bch2_sb_member_alloc()") Signed-off-by: Hongbo Li <lihongbo22@huawei.com> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: bch2_sb_member_alloc()Kent Overstreet3-46/+53
refactoring Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: bch2_dev_remove_alloc() -> alloc_background.cKent Overstreet3-27/+30
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Move tabstop setup to bch2_dev_usage_to_text()Kent Overstreet2-7/+9
No reason for it not to be where it's needed. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Options for recovery_passes, recovery_passes_excludeKent Overstreet8-20/+33
This adds mount options for specifying recovery passes to run, or exclude; the immediate need for this is that backpointers fsck is having trouble completing, so we need a way to skip it. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Use mm_account_reclaimed_pages() when freeing btree nodesKent Overstreet1-0/+11
When freeing in a shrinker callback, we need to notify memory reclaim, so it knows forward progress has been made. Normally this is done in e.g. slab code, but we're not freeing through slab - or rather we are, but these allocations are big, and use the kmalloc_large() path. This is really a bug in the slub code, but we're working around it here for now. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Use __GFP_ACCOUNT for reclaimable memoryKent Overstreet2-0/+4
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Hook up RENAME_WHITEOUT in rename.Sasha Finkelstein4-14/+52
This is needed for overlayfs, which is used by container managers. Signed-off-by: Sasha Finkelstein <fnkl.kernel@gmail.com> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: rebalance writes use BCH_WRITE_ONLY_SPECIFIED_DEVSKent Overstreet2-2/+3
this was an oversight: rebalance is moving data to a specific device, so we don't want it falling back to the full filesystem Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: BCH_WRITE_ALLOC_NOWAIT no longer applies to open bucket allocationKent Overstreet3-12/+16
rebalance writes must be BCH_WRITE_ALLOC_NOWAIT because they don't allocate from the full filesystem - but we don't want spurious allocation failures due to open buckets. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: fix prototype to bch2_alloc_sectors_start_trans()Kent Overstreet4-17/+18
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: kill redundant is_vmalloc_addr()Kent Overstreet1-8/+4
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: convert __bch2_encrypt_bio() to darrayKent Overstreet1-19/+22
like the previous patch, kill use of bare arrays; the encryption code likes to work in big batches, so this is a small performance improvement. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: do_encrypt() now handles allocation failuresKent Overstreet1-18/+30
convert to darray, and add a fallback when allocation fails Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: Add pinned to btree cache not freed countersKent Overstreet2-21/+36
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-09bcachefs: Annotate bch_replicas_entry_{v0,v1} with __counted_by()Thorsten Blum4-7/+13
Add the __counted_by compiler attribute to the flexible array members devs to improve access bounds-checking via CONFIG_UBSAN_BOUNDS and CONFIG_FORTIFY_SOURCE. Increment nr_devs before adding a new device to the devs array and adjust the array indexes accordingly. Add a helper macro for adding a new device. In bch2_journal_read(), explicitly set nr_devs to 0. Signed-off-by: Thorsten Blum <thorsten.blum@toblux.com> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-09bcachefs: support idmap mountsHongbo Li1-16/+34
We enable idmapped mounts for bcachefs. Here, we just pass down the user_namespace argument from the VFS methods to the relevant helpers. The idmap test in bcachefs is as following: ``` 1. losetup /dev/loop1 bcachefs.img 2. ./bcachefs format /dev/loop1 3. mount -t bcachefs /dev/loop1 /mnt/bcachefs/ 4. ./mount-idmapped --map-mount b:0:1000:1 /mnt/bcachefs /mnt/idmapped1/ ll /mnt/bcachefs total 2 drwx------. 2 root root 0 Jun 14 14:10 lost+found -rw-r--r--. 1 root root 1945 Jun 14 14:12 profile ll /mnt/idmapped1/ total 2 drwx------. 2 1000 1000 0 Jun 14 14:10 lost+found -rw-r--r--. 1 1000 1000 1945 Jun 14 14:12 profile Signed-off-by: Hongbo Li <lihongbo22@huawei.com> Reviewed-by: Christian Brauner <brauner@kernel.org> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-09bcachefs: Annotate struct bch_xattr with __counted_by()Thorsten Blum1-1/+1
Add the __counted_by compiler attribute to the flexible array member x_name to improve access bounds-checking via CONFIG_UBSAN_BOUNDS and CONFIG_FORTIFY_SOURCE. Signed-off-by: Thorsten Blum <thorsten.blum@toblux.com> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-09bcachefs: Switch gc bucket array to a genradixKent Overstreet4-37/+6
A user with a 30 tb device is overflowing the INT_MAX limit on vmalloc allocations... Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-09bcachefs: darray: convert to alloc_hooks()Kent Overstreet2-18/+12
better memory allocation profiling support Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
generated by cgit v1.2.3 (git 2.39.0) at 2025-09-29 00:26:56 +0300