linux-stable.git/fs/bcachefs, branch linux-6.9.y Linux kernel stable tree bcachefs: btree_gc can now handle unknown btrees 2024-07-05T07:38:17+00:00 Kent Overstreet kent.overstreet@linux.dev 2024-05-27T22:40:50+00:00 6580811459fa224b0524516084df232ad119bf41 commit 088d0de81220a74d7d553febb81656927f10bb16 upstream. Compatibility fix - we no longer have a separate table for which order gc walks btrees in, and special case the stripes btree directly. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit 088d0de81220a74d7d553febb81656927f10bb16 upstream.

Compatibility fix - we no longer have a separate table for which order
gc walks btrees in, and special case the stripes btree directly.

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
bcachefs: Fix setting of downgrade recovery passes/errors 2024-07-05T07:38:17+00:00 Kent Overstreet kent.overstreet@linux.dev 2024-05-27T20:30:19+00:00 2c2cd3213084e762f879d25020ed1d755335968d commit 247c056bde2ebc9fad2fc62332dc7cc99b58d720 upstream. bch2_check_version_downgrade() was setting c->sb.version, which bch2_sb_set_downgrade() expects to be at the previous version; and it shouldn't even have been set directly because c->sb.version is updated by write_super(). Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit 247c056bde2ebc9fad2fc62332dc7cc99b58d720 upstream.

bch2_check_version_downgrade() was setting c->sb.version, which
bch2_sb_set_downgrade() expects to be at the previous version; and it
shouldn't even have been set directly because c->sb.version is updated
by write_super().

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
bcachefs: Fix bch2_sb_downgrade_update() 2024-07-05T07:38:16+00:00 Kent Overstreet kent.overstreet@linux.dev 2024-06-17T15:31:00+00:00 5da1f2a83951ec05f2dc9b5b7650218356b3c999 commit ddd118ab45e848b1956ef8c8ef84963a554b5b58 upstream. Missing enum conversion Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit ddd118ab45e848b1956ef8c8ef84963a554b5b58 upstream.

Missing enum conversion

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
bcachefs: Fix sb-downgrade validation 2024-07-05T07:38:16+00:00 Kent Overstreet kent.overstreet@linux.dev 2024-05-25T16:38:53+00:00 3989465f47b732b081e49336f4bd0d547c5d6a68 commit 9242a34b760648b722f4958749ad83ef7d0f7525 upstream. Superblock downgrade entries are only two byte aligned, but section sizes are 8 byte aligned, which means we have to be careful about overrun checks; an entry that crosses the end of the section is allowed (and ignored) as long as it has zero errors. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit 9242a34b760648b722f4958749ad83ef7d0f7525 upstream.

Superblock downgrade entries are only two byte aligned, but section
sizes are 8 byte aligned, which means we have to be careful about
overrun checks; an entry that crosses the end of the section is allowed
(and ignored) as long as it has zero errors.

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
bcachefs: Fix sb_field_downgrade validation 2024-07-05T07:38:16+00:00 Kent Overstreet kent.overstreet@linux.dev 2024-05-06T13:16:33+00:00 bf920ed92ef24dcd6970c88881cd4700b3acf05b commit 692aa7a54b2b28d59f24b3bf8250837805484b99 upstream. - bch2_sb_downgrade_validate() wasn't checking for a downgrade entry extending past the end of the superblock section - for_each_downgrade_entry() is used in to_text() and needs to work on malformed input; it also was missing a check for a field extending past the end of the section Reported-by: syzbot+e49ccab73449180bc9be@syzkaller.appspotmail.com Fixes: 84f1638795da ("bcachefs: bch_sb_field_downgrade") Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit 692aa7a54b2b28d59f24b3bf8250837805484b99 upstream.

- bch2_sb_downgrade_validate() wasn't checking for a downgrade entry
  extending past the end of the superblock section

- for_each_downgrade_entry() is used in to_text() and needs to work on
  malformed input; it also was missing a check for a field extending
  past the end of the section

Reported-by: syzbot+e49ccab73449180bc9be@syzkaller.appspotmail.com
Fixes: 84f1638795da ("bcachefs: bch_sb_field_downgrade")
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
bcachefs: Add missing sched_annotate_sleep() in bch2_journal_flush_seq_async() 2024-05-07T15:02:37+00:00 Kent Overstreet kent.overstreet@linux.dev 2024-05-07T03:11:43+00:00 6e297a73bccf852e7716207caa8eb868737c7155 Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev> 
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
bcachefs: Fix race in bch2_write_super() 2024-05-07T15:02:36+00:00 Kent Overstreet kent.overstreet@linux.dev 2024-05-07T00:49:24+00:00 54541c1f78e12a78487ae63e2e199a7d4f6dbd26 bch2_write_super() was looping over online devices multiple times - dropping and retaking io_ref each time. This meant it could race with device removal; it could increment the sequence number on a device but fail to write it - and then if the device was re-added, it would get confused the next time around thinking a superblock write was silently dropped. Fix this by taking io_ref once, and stashing pointers to online devices in a darray. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev> 
bch2_write_super() was looping over online devices multiple times -
dropping and retaking io_ref each time.

This meant it could race with device removal; it could increment the
sequence number on a device but fail to write it - and then if the
device was re-added, it would get confused the next time around thinking
a superblock write was silently dropped.

Fix this by taking io_ref once, and stashing pointers to online devices
in a darray.

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
bcachefs: BCH_SB_LAYOUT_SIZE_BITS_MAX 2024-05-06T14:58:17+00:00 Kent Overstreet kent.overstreet@linux.dev 2024-05-06T13:10:29+00:00 71dac2482ad3c8d4a8b8998a96751f009bad895f Define a constant for the max superblock size, to avoid a too-large shift. Reported-by: syzbot+a8b0fb419355c91dda7f@syzkaller.appspotmail.com Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev> 
Define a constant for the max superblock size, to avoid a too-large
shift.

Reported-by: syzbot+a8b0fb419355c91dda7f@syzkaller.appspotmail.com
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
bcachefs: Add missing skcipher_request_set_callback() call 2024-05-06T14:58:17+00:00 Kent Overstreet kent.overstreet@linux.dev 2024-05-06T12:40:46+00:00 88ab10186c44d0a8b90842beab8648b5fd14432d Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev> 
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
bcachefs: Fix snapshot_t() usage in bch2_fs_quota_read_inode() 2024-05-06T14:58:17+00:00 Kent Overstreet kent.overstreet@linux.dev 2024-05-06T02:56:54+00:00 8060bf1d83f7d404bacb0e5a38f2d4d8f4c9dfb7 bch2_fs_quota_read_inode() wasn't entirely updated to the bch2_snapshot_tree() helper, which takes rcu lock. Reported-by: syzbot+a3a9a61224ed3b7f0010@syzkaller.appspotmail.com Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev> 
bch2_fs_quota_read_inode() wasn't entirely updated to the
bch2_snapshot_tree() helper, which takes rcu lock.

Reported-by: syzbot+a3a9a61224ed3b7f0010@syzkaller.appspotmail.com
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>