| Age | Commit message (Collapse) | Author |
|---|
|
Instead of passing a (VFS) inode pointer argument, pass a btrfs_inode
instead, as this is generally what we do for internal APIs, making it
more consistent with most of the code base. This will later allow to
help to remove a lot of BTRFS_I() calls in btrfs_sync_file().
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Instead of using a inode pointer, use a btrfs_inode pointer in the log
context structure, as this is generally what we need and allows for some
internal APIs to take a btrfs_inode instead, making them more consistent
with most of the code base. This will later allow to help to remove a lot
of BTRFS_I() calls in btrfs_sync_file().
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Currently btrfs_finish_ordered_extent() returns a boolean indicating if
the ordered extent was added to the work queue for completion, but none
of its callers cares about it, so make it return void.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
The function btrfs_block_group_root() is declared in disk-io.c; however,
all its callers are in block-group.c. Move it to the latter file and
declare it static.
Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Drop the single-use variable bytenr_orig and instead use btrfs_sb_offset()
in the function argument passing.
Fix a stale comment about not automatically fixing a bad primary
superblock from the backup mirror copies. Also, move the comment closer
to where the primary superblock read occurs.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
We are currently using a cached rb_root (struct rb_root_cached) for the
rb root of struct extent_map_tree. This doesn't offer much of an advantage
here because:
1) It's only advantage over the regular rb_root is that it caches a
pointer to the left most node (first node), so a call to
rb_first_cached() doesn't have to chase pointers until it reaches
the left most node;
2) We only have two scenarios that access left most node with
rb_first_cached():
When dropping all extent maps from an inode, during inode eviction;
When iterating over extent maps during the extent map shrinker;
3) In both cases we keep removing extent maps, which causes deletion of
the left most node so rb_erase_cached() has to call rb_next() to find
out what's the next left most node and assign it to
struct rb_root_cached::rb_leftmost;
4) We can do that ourselves in those two uses cases and stop using a
rb_root_cached rb tree and use instead a regular rb_root rb tree.
This reduces the size of struct extent_map_tree by 8 bytes and, since
this structure is embedded in struct btrfs_inode, it also reduces the
size of that structure by 8 bytes.
So on a 64 bits platform the size of btrfs_inode is reduced from 1032
bytes down to 1024 bytes.
This means we will be able to have 4 inodes per 4K page instead of 3.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Currently we name the rb_root member of struct extent_map_tree as 'map',
which is odd and confusing. Since it's a root node, rename it to 'root'.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
On 64 bits platforms we don't really need to have a dedicated member (the
objectid field) for the inode's number since we store in the VFS inode's
i_ino member, which is an unsigned long and this type is 64 bits wide on
64 bits platforms. We only need that field in case we are on a 32 bits
platform because the unsigned long type is 32 bits wide on such platforms
See commit 33345d01522f ("Btrfs: Always use 64bit inode number") regarding
this 64/32 bits detail.
The objectid field of struct btrfs_inode is also used to store the ID of
a root for directories that are stubs for unreferenced roots. In such
cases the inode is a directory and has the BTRFS_INODE_ROOT_STUB runtime
flag set.
So in order to reduce the size of btrfs_inode structure on 64 bits
platforms we can remove the objectid member and use the VFS inode's i_ino
member instead whenever we need to get the inode number. In case the inode
is a root stub (BTRFS_INODE_ROOT_STUB set) we can use the member
last_reflink_trans to store the ID of the unreferenced root, since such
inode is a directory and reflinks can't be done against directories.
So remove the objectid fields for 64 bits platforms and alias the
last_reflink_trans field with a name of ref_root_id in a union.
On a release kernel config, this reduces the size of struct btrfs_inode
from 1040 bytes down to 1032 bytes.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Currently struct btrfs_inode has a key member, named "location", that is
either:
1) The key of the inode's item. In this case the objectid is the number
of the inode;
2) A key stored in a dir entry with a type of BTRFS_ROOT_ITEM_KEY, for
the case where we have a root that is a snapshot of a subvolume that
points to other subvolumes. In this case the objectid is the ID of
a subvolume inside the snapshotted parent subvolume.
The key is only used to lookup the inode item for the first case, while
for the second it's never used since it corresponds to directory stubs
created with new_simple_dir() and which are marked as dummy, so there's
no actual inode item to ever update. In the second case we only check
the key type at btrfs_ino() for 32 bits platforms and its objectid is
only needed for unlink.
Instead of using a key we can do fine with just the objectid, since we
can generate the key whenever we need it having only the objectid, as
in all use cases the type is always BTRFS_INODE_ITEM_KEY and the offset
is always 0.
So use only an objectid instead of a full key. This reduces the size of
struct btrfs_inode from 1048 bytes down to 1040 bytes on a release kernel.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
When not using the NO_HOLES feature we always allocate an io tree for an
inode's file_extent_tree. This is wasteful because that io tree is only
used for regular files, so we allocate more memory than needed for inodes
that represent directories or symlinks for example, or for inodes that
correspond to free space inodes.
So improve on this by allocating the io tree only for inodes of regular
files that are not free space inodes.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
The index_cnt field of struct btrfs_inode is used only for two purposes:
1) To store the index for the next entry added to a directory;
2) For the data relocation inode to track the logical start address of the
block group currently being relocated.
For the relocation case we use index_cnt because it's not used for
anything else in the relocation use case - we could have used other fields
that are not used by relocation such as defrag_bytes, last_unlink_trans
or last_reflink_trans for example (among others).
Since the csum_bytes field is not used for directories, do the following
changes:
1) Put index_cnt and csum_bytes in a union, and index_cnt is only
initialized when the inode is a directory. The csum_bytes is only
accessed in IO paths for regular files, so we're fine here;
2) Use the defrag_bytes field for relocation, since the data relocation
inode is never used for defrag purposes. And to make the naming better,
alias it to reloc_block_group_start by using a union.
This reduces the size of struct btrfs_inode by 8 bytes in a release
kernel, from 1056 bytes down to 1048 bytes.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Currently we use the spinlock inode_lock from struct btrfs_root to
serialize access to two different data structures:
1) The delayed inodes xarray (struct btrfs_root::delayed_nodes);
2) The inodes xarray (struct btrfs_root::inodes).
Instead of using our own lock, we can use the spinlock that is part of the
xarray implementation, by using the xa_lock() and xa_unlock() APIs and
using the xarray APIs with the double underscore prefix that don't take
the xarray locks and assume the caller is using xa_lock() and xa_unlock().
So remove the spinlock inode_lock from struct btrfs_root and use the
corresponding xarray locks. This brings 2 benefits:
1) We reduce the size of struct btrfs_root, from 1336 bytes down to
1328 bytes on a 64 bits release kernel config;
2) We reduce lock contention by not using anymore the same lock for
changing two different and unrelated xarrays.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Make btrfs_iget_path() simpler and easier to read by avoiding nesting of
if-then-else statements and having an error label to do all the error
handling instead of repeating it a couple times.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
When creating a new inode, at btrfs_create_new_inode(), one of the very
last steps is to add the inode to the root's inodes xarray. This often
requires allocating memory which may fail (even though xarrays have a
dedicated kmem_cache which make it less likely to fail), and at that point
we are forced to abort the current transaction (as some, but not all, of
the inode metadata was added to its subvolume btree).
To avoid a transaction abort, preallocate memory for the xarray early at
btrfs_create_new_inode(), so that if we fail we don't need to abort the
transaction and the insertion into the xarray is guaranteed to succeed.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Currently we use a red black tree (rb-tree) to track the currently open
inodes of a root (in struct btrfs_root::inode_tree). This however is not
very efficient when the number of inodes is large since rb-trees are
binary trees. For example for 100K open inodes, the tree has a depth of
17. Besides that, inserting into the tree requires navigating through it
and pulling useless cache lines in the process since the red black tree
nodes are embedded within the btrfs inode - on the other hand, by being
embedded, it requires no extra memory allocations.
We can improve this by using an xarray instead, which is efficient when
indices are densely clustered (such as inode numbers), is more cache
friendly and behaves like a resizable array, with a much better search
and insertion complexity than a red black tree. This only has one small
disadvantage which is that insertion will sometimes require allocating
memory for the xarray - which may fail (not that often since it uses a
kmem_cache) - but on the other hand we can reduce the btrfs inode
structure size by 24 bytes (from 1080 down to 1056 bytes) after removing
the embedded red black tree node, which after the next patches will allow
to reduce the size of the structure to 1024 bytes, meaning we will be able
to store 4 inodes per 4K page instead of 3 inodes.
This change does a straightforward change to use an xarray, and results
in a transaction abort if we can't allocate memory for the xarray when
creating an inode - but the next patch changes things so that we don't
need to abort.
Running the following fs_mark test showed some improvements:
$ cat test.sh
#!/bin/bash
DEV=/dev/nullb0
MNT=/mnt/nullb0
MOUNT_OPTIONS="-o ssd"
FILES=100000
THREADS=$(nproc --all)
echo "performance" | \
tee /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
mkfs.btrfs -f $DEV
mount $MOUNT_OPTIONS $DEV $MNT
OPTS="-S 0 -L 5 -n $FILES -s 0 -t $THREADS -k"
for ((i = 1; i <= $THREADS; i++)); do
OPTS="$OPTS -d $MNT/d$i"
done
fs_mark $OPTS
umount $MNT
Before this patch:
FSUse% Count Size Files/sec App Overhead
10 1200000 0 92081.6 12505547
16 2400000 0 138222.6 13067072
23 3600000 0 148833.1 13290336
43 4800000 0 97864.7 13931248
53 6000000 0 85597.3 14384313
After this patch:
FSUse% Count Size Files/sec App Overhead
10 1200000 0 93225.1 12571078
16 2400000 0 146720.3 12805007
23 3600000 0 160626.4 13073835
46 4800000 0 116286.2 13802927
53 6000000 0 90087.9 14754892
The test was run with a release kernel config (Debian's default config).
Also capturing the insertion times into the rb tree and into the xarray,
that is measuring the duration of the old function inode_tree_add() and
the duration of the new btrfs_add_inode_to_root() function, gave the
following results (in nanoseconds):
Before this patch, inode_tree_add() execution times:
Count: 5000000
Range: 0.000 - 5536887.000; Mean: 775.674; Median: 729.000; Stddev: 4820.961
Percentiles: 90th: 1015.000; 95th: 1139.000; 99th: 1397.000
0.000 - 7.816: 40 |
7.816 - 37.858: 209 |
37.858 - 170.278: 6059 |
170.278 - 753.961: 2754890 #####################################################
753.961 - 3326.728: 2232312 ###########################################
3326.728 - 14667.018: 4366 |
14667.018 - 64652.943: 852 |
64652.943 - 284981.761: 550 |
284981.761 - 1256150.914: 221 |
1256150.914 - 5536887.000: 7 |
After this patch, btrfs_add_inode_to_root() execution times:
Count: 5000000
Range: 0.000 - 2900652.000; Mean: 272.148; Median: 241.000; Stddev: 2873.369
Percentiles: 90th: 342.000; 95th: 432.000; 99th: 572.000
0.000 - 7.264: 104 |
7.264 - 33.145: 352 |
33.145 - 140.081: 109606 #
140.081 - 581.930: 4840090 #####################################################
581.930 - 2407.590: 43532 |
2407.590 - 9950.979: 2245 |
9950.979 - 41119.278: 514 |
41119.278 - 169902.616: 155 |
169902.616 - 702018.539: 47 |
702018.539 - 2900652.000: 9 |
Average, percentiles, standard deviation, etc, are all much better.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
There are several hard-to-hit ASSERT()s hit inside raid56.
Unfortunately the ASSERT() expression is a little complex, and except
the ASSERT(), there is nothing to provide any clue.
Considering if race is involved, it's pretty hard to reproduce.
Meanwhile sometimes the dump of the rbio structure can provide some
pretty good clues, it's worth to do the extra multi-line dump for
btrfs raid56 related code.
The dump looks like this:
BTRFS critical (device dm-3): bioc logical=4598530048 full_stripe=4598530048 size=0 map_type=0x81 mirror=0 replace_nr_stripes=0 replace_stripe_src=-1 num_stripes=5
BTRFS critical (device dm-3): nr=0 devid=1 physical=1166147584
BTRFS critical (device dm-3): nr=1 devid=2 physical=1145176064
BTRFS critical (device dm-3): nr=2 devid=4 physical=1145176064
BTRFS critical (device dm-3): nr=3 devid=5 physical=1145176064
BTRFS critical (device dm-3): nr=4 devid=3 physical=1145176064
BTRFS critical (device dm-3): rbio flags=0x0 nr_sectors=80 nr_data=4 real_stripes=5 stripe_nsectors=16 scrubp=0 dbitmap=0x0
BTRFS critical (device dm-3): logical=4598530048
assertion failed: orig_logical >= full_stripe_start && orig_logical + orig_len <= full_stripe_start + rbio->nr_data * BTRFS_STRIPE_LEN, in fs/btrfs/raid56.c:1702
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Due to a refactoring introduced by commit 53d9981ca20e ("btrfs: split
btrfs_alloc_ordered_extent to allocation and insertion helpers"), the
function btrfs_alloc_ordered_extent() was renamed to
alloc_ordered_extent(), so the comment at btrfs_remove_ordered_extent()
is no longer very accurate. Update the comment to refer to the new
name "alloc_ordered_extent()".
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Fix typo in the end IO compression callbacks, from "comprssed" to
"compressed".
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
The function btrfs_migrate_to_delayed_refs_rsv() is no longer used.
Its last use was removed in commit 2f6397e448e6 ("btrfs: don't refill
whole delayed refs block reserve when starting transaction").
So remove the function.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
It's not used outside zoned.c, so make it static.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Passing in a 'struct btrfs_io_geometry into handle_ops_on_dev_replace
can reduce the number of arguments by two.
No functional changes otherwise.
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
The ioctls that add relations, create qgroups or set limits start/join
transaction. When quotas are not enabled this is not necessary, there
will be errors reported back anyway but this could be also misleading
and we should really report that quotas are not enabled. For that use
-ENOTCONN.
The helper is meant to do a quick check before any other standard ioctl
checks are done. If quota is disabled meanwhile we still rely on proper
locking inside any active operation changing the qgroup structures.
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Huacai Chen:
The label end_reply is obviously a typo. It should be "replay" in this
context. So rename end_reply to end_replay.
Suggested-by: Huacai Chen <chenhuacai@loongson.cn>
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
|
|
If an NTFS file system is mounted to another system with different
PAGE_SIZE from the original system, log->page_size will change in
log_replay(), but log->page_{mask,bits} don't change correspondingly.
This will cause a panic because "u32 bytes = log->page_size - page_off"
will get a negative value in the later read_log_page().
Cc: stable@vger.kernel.org
Fixes: b46acd6a6a627d876898e ("fs/ntfs3: Add NTFS journal")
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
|
|
fileattr added to support chattr.
Supported attributes: compressed and immutable.
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
|
|
1) Make is_legacy_ntfs static inline.
2) Put legacy file_operations under #if IS_ENABLED(CONFIG_NTFS_FS).
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Christian Brauner <brauner@kernel.org>
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
|
|
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
|
|
For easy internal debugging.
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
|
|
Mark ntfs dirty in this case.
Rename ntfs_filldir to ntfs_dir_emit.
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
|
|
In order not to call copy_to_user (from fiemap_fill_next_extent)
we allocate memory in the kernel, fill it and copy it to user memory
after up_read(run_lock).
Reported-by: syzbot+36bb70085ef6edc2ebb9@syzkaller.appspotmail.com
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
|
|
We skip the run_truncate_head call also for $MFT::$ATTR_BITMAP.
Otherwise wnd_map()/run_lookup_entry will not find the disk position for the bitmap parts.
Fixes: 0e5b044cbf3a ("fs/ntfs3: Refactoring attr_set_size to restore after errors")
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
|
|
Fixes: 3f3b442b5ad2 ("fs/ntfs3: Add bitmap")
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
|
|
The 'nocase' option was mistakenly added as fsparam_flag_no
with the 'no' prefix, causing the case-insensitive mode to require
the 'nonocase' option to be enabled.
Fixes: a3a956c78efa ("fs/ntfs3: Add option "nocase"")
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
|
|
LZ4 always reuses the decompressed buffer as its LZ77 sliding window
(dynamic dictionary) for optimal performance. However, in specific
cases, the output buffer may not fully contain valid page cache pages,
resulting in the use of short-lived pages for temporary purposes.
Due to the limited sliding window size, LZ4 shortlived bounce pages can
also be reused in a sliding manner, so each bounce page can be vmapped
multiple times in different relative positions by design. In order to
avoiding double frees, currently, reuse counts are recorded via page
refcount, but it will no longer be used as-is in the future world of
Memdescs.
Just maintain a lookup table to check if a shortlived page is reused.
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20240711053659.1364989-1-hsiangkao@linux.alibaba.com
|
|
The syzbot constructs a directory that has no dirblock but is non-inline,
i.e. the first directory block is a hole. And no errors are reported when
creating files in this directory in the following flow.
ext4_mknod
...
ext4_add_entry
// Read block 0
ext4_read_dirblock(dir, block, DIRENT)
bh = ext4_bread(NULL, inode, block, 0)
if (!bh && (type == INDEX || type == DIRENT_HTREE))
// The first directory block is a hole
// But type == DIRENT, so no error is reported.
After that, we get a directory block without '.' and '..' but with a valid
dentry. This may cause some code that relies on dot or dotdot (such as
make_indexed_dir()) to crash.
Therefore when ext4_read_dirblock() finds that the first directory block
is a hole report that the filesystem is corrupted and return an error to
avoid loading corrupted data from disk causing something bad.
Reported-by: syzbot+ae688d469e36fb5138d0@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=ae688d469e36fb5138d0
Fixes: 4e19d6b65fb4 ("ext4: allow directory holes")
Cc: stable@kernel.org
Signed-off-by: Baokun Li <libaokun1@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://patch.msgid.link/20240702132349.2600605-3-libaokun@huaweicloud.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
|
|
Syzbot reports a issue as follows:
============================================
BUG: unable to handle page fault for address: ffffed11022e24fe
PGD 23ffee067 P4D 23ffee067 PUD 0
Oops: Oops: 0000 [#1] PREEMPT SMP KASAN PTI
CPU: 0 PID: 5079 Comm: syz-executor306 Not tainted 6.10.0-rc5-g55027e689933 #0
Call Trace:
<TASK>
make_indexed_dir+0xdaf/0x13c0 fs/ext4/namei.c:2341
ext4_add_entry+0x222a/0x25d0 fs/ext4/namei.c:2451
ext4_rename fs/ext4/namei.c:3936 [inline]
ext4_rename2+0x26e5/0x4370 fs/ext4/namei.c:4214
[...]
============================================
The immediate cause of this problem is that there is only one valid dentry
for the block to be split during do_split, so split==0 results in out of
bounds accesses to the map triggering the issue.
do_split
unsigned split
dx_make_map
count = 1
split = count/2 = 0;
continued = hash2 == map[split - 1].hash;
---> map[4294967295]
The maximum length of a filename is 255 and the minimum block size is 1024,
so it is always guaranteed that the number of entries is greater than or
equal to 2 when do_split() is called.
But syzbot's crafted image has no dot and dotdot in dir, and the dentry
distribution in dirblock is as follows:
bus dentry1 hole dentry2 free
|xx--|xx-------------|...............|xx-------------|...............|
0 12 (8+248)=256 268 256 524 (8+256)=264 788 236 1024
So when renaming dentry1 increases its name_len length by 1, neither hole
nor free is sufficient to hold the new dentry, and make_indexed_dir() is
called.
In make_indexed_dir() it is assumed that the first two entries of the
dirblock must be dot and dotdot, so bus and dentry1 are left in dx_root
because they are treated as dot and dotdot, and only dentry2 is moved
to the new leaf block. That's why count is equal to 1.
Therefore add the ext4_check_dx_root() helper function to add more sanity
checks to dot and dotdot before starting the conversion to avoid the above
issue.
Reported-by: syzbot+ae688d469e36fb5138d0@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=ae688d469e36fb5138d0
Fixes: ac27a0ec112a ("[PATCH] ext4: initial copy of files from ext3")
Cc: stable@kernel.org
Signed-off-by: Baokun Li <libaokun1@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://patch.msgid.link/20240702132349.2600605-2-libaokun@huaweicloud.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
|
|
Introduce a new help addrs_per_page() to wrap common code
from addrs_per_inode() and addrs_per_block() for cleanup.
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
|
|
Use temporary variable instead of F2FS_I() for cleanup.
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
|
|
In case of the COW file, new updates and GC writes are already
separated to page caches of the atomic file and COW file. As some cases
that use the meta inode for GC, there are some race issues between a
foreground thread and GC thread.
To handle them, we need to take care when to invalidate and wait
writeback of GC pages in COW files as the case of using the meta inode.
Also, a pointer from the COW inode to the original inode is required to
check the state of original pages.
For the former, we can solve the problem by using the meta inode for GC
of COW files. Then let's get a page from the original inode in
move_data_block when GCing the COW file to avoid race condition.
Fixes: 3db1de0e582c ("f2fs: change the current atomic write way")
Cc: stable@vger.kernel.org #v5.19+
Reviewed-by: Sungjong Seo <sj1557.seo@samsung.com>
Reviewed-by: Yeongjin Gil <youngjin.gil@samsung.com>
Signed-off-by: Sunmin Jeong <s_min.jeong@samsung.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
|
|
The page cache of the atomic file keeps new data pages which will be
stored in the COW file. It can also keep old data pages when GCing the
atomic file. In this case, new data can be overwritten by old data if a
GC thread sets the old data page as dirty after new data page was
evicted.
Also, since all writes to the atomic file are redirected to COW inodes,
GC for the atomic file is not working well as below.
f2fs_gc(gc_type=FG_GC)
- select A as a victim segment
do_garbage_collect
- iget atomic file's inode for block B
move_data_page
f2fs_do_write_data_page
- use dn of cow inode
- set fio->old_blkaddr from cow inode
- seg_freed is 0 since block B is still valid
- goto gc_more and A is selected as victim again
To solve the problem, let's separate GC writes and updates in the atomic
file by using the meta inode for GC writes.
Fixes: 3db1de0e582c ("f2fs: change the current atomic write way")
Cc: stable@vger.kernel.org #v5.19+
Reviewed-by: Sungjong Seo <sj1557.seo@samsung.com>
Reviewed-by: Yeongjin Gil <youngjin.gil@samsung.com>
Signed-off-by: Sunmin Jeong <s_min.jeong@samsung.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
|
|
When new_curseg() is allocating a new segment, if mode=fragment:xxx is
switched on in large section scenario, __get_next_segno() will select
the next segno randomly in the range of [0, maxsegno] in order to
fragment segments.
If the candidate segno is free, get_new_segment() will use it directly
as the new segment.
However, if the section of the candidate is not empty, and some other
segments have already been used, and have a different type (e.g NODE)
with the candidate (e.g DATA), GC will complain inconsistent segment
type later.
This could be reproduced by the following steps:
dd if=/dev/zero of=test.img bs=1M count=10240
mkfs.f2fs -s 128 test.img
mount -t f2fs test.img /mnt -o mode=fragment:block
echo 1 > /sys/fs/f2fs/loop0/max_fragment_chunk
echo 512 > /sys/fs/f2fs/loop0/max_fragment_hole
dd if=/dev/zero of=/mnt/testfile bs=4K count=100
umount /mnt
F2FS-fs (loop0): Inconsistent segment (4377) type [0, 1] in SSA and SIT
In order to allow simulating segment fragmentation in large section
scenario, this patch reduces the candidate range:
* if curseg is the last segment in the section, return curseg->segno
to make get_new_segment() itself find the next free segment.
* if curseg is in the middle of the section, select candicate randomly
in the range of [curseg + 1, last_seg_in_the_same_section] to keep
type consistent.
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Sheng Yong <shengyong@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
|
|
Commit 59c9081bc86e ("f2fs: allow write page cache when writting cp")
allows write() to write data to page cache during checkpoint, so block
count fields like .total_valid_block_count, .alloc_valid_block_count
and .rf_node_block_count may encounter race condition as below:
CP Thread A
- write_checkpoint
- block_operations
- f2fs_down_write(&sbi->node_change)
- __prepare_cp_block
: ckpt->valid_block_count = .total_valid_block_count
- f2fs_up_write(&sbi->node_change)
- write
- f2fs_preallocate_blocks
- f2fs_map_blocks(,F2FS_GET_BLOCK_PRE_AIO)
- f2fs_map_lock
- f2fs_down_read(&sbi->node_change)
- f2fs_reserve_new_blocks
- inc_valid_block_count
: percpu_counter_add(&sbi->alloc_valid_block_count, count)
: sbi->total_valid_block_count += count
- f2fs_up_read(&sbi->node_change)
- do_checkpoint
: sbi->last_valid_block_count = sbi->total_valid_block_count
: percpu_counter_set(&sbi->alloc_valid_block_count, 0)
: percpu_counter_set(&sbi->rf_node_block_count, 0)
- fsync
- need_do_checkpoint
- f2fs_space_for_roll_forward
: alloc_valid_block_count was reset to zero,
so, it may missed last data during checkpoint
Let's change to update .total_valid_block_count, .alloc_valid_block_count
and .rf_node_block_count in block_operations(), then their access can be
protected by .node_change and .cp_rwsem lock, so that it can avoid above
race condition.
Fixes: 59c9081bc86e ("f2fs: allow write page cache when writting cp")
Cc: Yunlei He <heyunlei@oppo.com>
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
|
|
The lazytime/nolazytime options are now handled in the VFS, and are
never seen in filesystem parsers, so remove handling of these
options from f2fs.
Note: when lazytime support was added in 6d94c74ab85f it made
lazytime the default in default_options() - as a result, lazytime
cannot be disabled (because Opt_nolazytime is never seen in f2fs
parsing).
If lazytime is desired to be configurable, and default off is OK,
default_options() could be updated to stop setting it by default
and allow mount option control.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
|
|
This patch fixes a potentially null pointer being accessed by
is_end_zone_blkaddr() that checks the last block of a zone
when f2fs is mounted as a single device.
Fixes: e067dc3c6b9c ("f2fs: maintain six open zones for zoned devices")
Signed-off-by: Daejun Park <daejun7.park@samsung.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Reviewed-by: Daeho Jeong <daehojeong@google.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Pull misc fixes from Andrew Morton:
"21 hotfixes, 15 of which are cc:stable.
No identifiable theme here - all are singleton patches, 19 are for MM"
* tag 'mm-hotfixes-stable-2024-07-10-13-19' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (21 commits)
mm/hugetlb: fix kernel NULL pointer dereference when migrating hugetlb folio
mm/hugetlb: fix potential race in __update_and_free_hugetlb_folio()
filemap: replace pte_offset_map() with pte_offset_map_nolock()
arch/xtensa: always_inline get_current() and current_thread_info()
sched.h: always_inline alloc_tag_{save|restore} to fix modpost warnings
MAINTAINERS: mailmap: update Lorenzo Stoakes's email address
mm: fix crashes from deferred split racing folio migration
lib/build_OID_registry: avoid non-destructive substitution for Perl < 5.13.2 compat
mm: gup: stop abusing try_grab_folio
nilfs2: fix kernel bug on rename operation of broken directory
mm/hugetlb_vmemmap: fix race with speculative PFN walkers
cachestat: do not flush stats in recency check
mm/shmem: disable PMD-sized page cache if needed
mm/filemap: skip to create PMD-sized page cache if needed
mm/readahead: limit page cache size in page_cache_ra_order()
mm/filemap: make MAX_PAGECACHE_ORDER acceptable to xarray
mm/damon/core: merge regions aggressively when max_nr_regions is unmet
Fix userfaultfd_api to return EINVAL as expected
mm: vmalloc: check if a hash-index is in cpu_possible_mask
mm: prevent derefencing NULL ptr in pfn_section_valid()
...
|
|
Pull bcachefs fixes from Kent Overstreet:
- Switch some asserts to WARN()
- Fix a few "transaction not locked" asserts in the data read retry
paths and backpointers gc
- Fix a race that would cause the journal to get stuck on a flush
commit
- Add missing fsck checks for the fragmentation LRU
- The usual assorted ssorted syzbot fixes
* tag 'bcachefs-2024-07-10' of https://evilpiepirate.org/git/bcachefs: (22 commits)
bcachefs: Add missing bch2_trans_begin()
bcachefs: Fix missing error check in journal_entry_btree_keys_validate()
bcachefs: Warn on attempting a move with no replicas
bcachefs: bch2_data_update_to_text()
bcachefs: Log mount failure error code
bcachefs: Fix undefined behaviour in eytzinger1_first()
bcachefs: Mark bch_inode_info as SLAB_ACCOUNT
bcachefs: Fix bch2_inode_insert() race path for tmpfiles
closures: fix closure_sync + closure debugging
bcachefs: Fix journal getting stuck on a flush commit
bcachefs: io clock: run timer fns under clock lock
bcachefs: Repair fragmentation_lru in alloc_write_key()
bcachefs: add check for missing fragmentation in check_alloc_to_lru_ref()
bcachefs: bch2_btree_write_buffer_maybe_flush()
bcachefs: Add missing printbuf_tabstops_reset() calls
bcachefs: Fix loop restart in bch2_btree_transactions_read()
bcachefs: Fix bch2_read_retry_nodecode()
bcachefs: Don't use the new_fs() bucket alloc path on an initialized fs
bcachefs: Fix shift greater than integer size
bcachefs: Change bch2_fs_journal_stop() BUG_ON() to warning
...
|
|
The nfs_page_length is not used anywhere, remove it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
pnfs_layoutdriver_type
max_deviceinfo_size is not set anywhere, remove it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
|
|
Reported-by: syzbot+e74fea078710bbca6f4b@syzkaller.appspotmail.com
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
|
