| Age | Commit message (Collapse) | Author |
|---|
|
Bring in the fdtable changes for this cycle.
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
Port files to rely on file_ref reference to improve scaling and gain
overflow protection.
- We continue to WARN during get_file() in case a file that is already
marked dead is revived as get_file() is only valid if the caller
already holds a reference to the file. This hasn't changed just the
check changes.
- The semantics for epoll and ttm's dmabuf usage have changed. Both
epoll and ttm synchronize with __fput() to prevent the underlying file
from beeing freed.
(1) epoll
Explaining epoll is straightforward using a simple diagram.
Essentially, the mutex of the epoll instance needs to be taken in both
__fput() and around epi_fget() preventing the file from being freed
while it is polled or preventing the file from being resurrected.
CPU1 CPU2
fput(file)
-> __fput(file)
-> eventpoll_release(file)
-> eventpoll_release_file(file)
mutex_lock(&ep->mtx)
epi_item_poll()
-> epi_fget()
-> file_ref_get(file)
mutex_unlock(&ep->mtx)
mutex_lock(&ep->mtx);
__ep_remove()
mutex_unlock(&ep->mtx);
-> kmem_cache_free(file)
(2) ttm dmabuf
This explanation is a bit more involved. A regular dmabuf file stashed
the dmabuf in file->private_data and the file in dmabuf->file:
file->private_data = dmabuf;
dmabuf->file = file;
The generic release method of a dmabuf file handles file specific
things:
f_op->release::dma_buf_file_release()
while the generic dentry release method of a dmabuf handles dmabuf
freeing including driver specific things:
dentry->d_release::dma_buf_release()
During ttm dmabuf initialization in ttm_object_device_init() the ttm
driver copies the provided struct dma_buf_ops into a private location:
struct ttm_object_device {
spinlock_t object_lock;
struct dma_buf_ops ops;
void (*dmabuf_release)(struct dma_buf *dma_buf);
struct idr idr;
};
ttm_object_device_init(const struct dma_buf_ops *ops)
{
// copy original dma_buf_ops in private location
tdev->ops = *ops;
// stash the release method of the original struct dma_buf_ops
tdev->dmabuf_release = tdev->ops.release;
// override the release method in the copy of the struct dma_buf_ops
// with ttm's own dmabuf release method
tdev->ops.release = ttm_prime_dmabuf_release;
}
When a new dmabuf is created the struct dma_buf_ops with the overriden
release method set to ttm_prime_dmabuf_release is passed in exp_info.ops:
DEFINE_DMA_BUF_EXPORT_INFO(exp_info);
exp_info.ops = &tdev->ops;
exp_info.size = prime->size;
exp_info.flags = flags;
exp_info.priv = prime;
The call to dma_buf_export() then sets
mutex_lock_interruptible(&prime->mutex);
dma_buf = dma_buf_export(&exp_info)
{
dmabuf->ops = exp_info->ops;
}
mutex_unlock(&prime->mutex);
which creates a new dmabuf file and then install a file descriptor to
it in the callers file descriptor table:
ret = dma_buf_fd(dma_buf, flags);
When that dmabuf file is closed we now get:
fput(file)
-> __fput(file)
-> f_op->release::dma_buf_file_release()
-> dput()
-> d_op->d_release::dma_buf_release()
-> dmabuf->ops->release::ttm_prime_dmabuf_release()
mutex_lock(&prime->mutex);
if (prime->dma_buf == dma_buf)
prime->dma_buf = NULL;
mutex_unlock(&prime->mutex);
Where we can see that prime->dma_buf is set to NULL. So when we have
the following diagram:
CPU1 CPU2
fput(file)
-> __fput(file)
-> f_op->release::dma_buf_file_release()
-> dput()
-> d_op->d_release::dma_buf_release()
-> dmabuf->ops->release::ttm_prime_dmabuf_release()
ttm_prime_handle_to_fd()
mutex_lock_interruptible(&prime->mutex)
dma_buf = prime->dma_buf
dma_buf && get_dma_buf_unless_doomed(dma_buf)
-> file_ref_get(dma_buf->file)
mutex_unlock(&prime->mutex);
mutex_lock(&prime->mutex);
if (prime->dma_buf == dma_buf)
prime->dma_buf = NULL;
mutex_unlock(&prime->mutex);
-> kmem_cache_free(file)
The logic of the mechanism is the same as for epoll: sync with
__fput() preventing the file from being freed. Here the
synchronization happens through the ttm instance's prime->mutex.
Basically, the lifetime of the dma_buf and the file are tighly
coupled.
Both (1) and (2) used to call atomic_inc_not_zero() to check whether
the file has already been marked dead and then refuse to revive it.
This is only safe because both (1) and (2) sync with __fput() and thus
prevent kmem_cache_free() on the file being called and thus prevent
the file from being immediately recycled due to SLAB_TYPESAFE_BY_RCU.
Both (1) and (2) have been ported from atomic_inc_not_zero() to
file_ref_get(). That means a file that is already in the process of
being marked as FILE_REF_DEAD:
file_ref_put()
cnt = atomic_long_dec_return()
-> __file_ref_put(cnt)
if (cnt == FIlE_REF_NOREF)
atomic_long_try_cmpxchg_release(cnt, FILE_REF_DEAD)
can be revived again:
CPU1 CPU2
file_ref_put()
cnt = atomic_long_dec_return()
-> __file_ref_put(cnt)
if (cnt == FIlE_REF_NOREF)
file_ref_get()
// Brings reference back to FILE_REF_ONEREF
atomic_long_add_negative()
atomic_long_try_cmpxchg_release(cnt, FILE_REF_DEAD)
This is fine and inherent to the file_ref_get()/file_ref_put()
semantics. For both (1) and (2) this is safe because __fput() is
prevented from making progress if file_ref_get() fails due to the
aforementioned synchronization mechanisms.
Two cases need to be considered that affect both (1) epoll and (2) ttm
dmabuf:
(i) fput()'s file_ref_put() and marks the file as FILE_REF_NOREF but
before that fput() can mark the file as FILE_REF_DEAD someone
manages to sneak in a file_ref_get() and brings the refcount back
from FILE_REF_NOREF to FILE_REF_ONEREF. In that case the original
fput() doesn't call __fput(). For epoll the poll will finish and
for ttm dmabuf the file can be used again. For ttm dambuf this is
actually an advantage because it avoids immediately allocating
a new dmabuf object.
CPU1 CPU2
file_ref_put()
cnt = atomic_long_dec_return()
-> __file_ref_put(cnt)
if (cnt == FIlE_REF_NOREF)
file_ref_get()
// Brings reference back to FILE_REF_ONEREF
atomic_long_add_negative()
atomic_long_try_cmpxchg_release(cnt, FILE_REF_DEAD)
(ii) fput()'s file_ref_put() marks the file FILE_REF_NOREF and
also suceeds in actually marking it FILE_REF_DEAD and then calls
into __fput() to free the file.
When either (1) or (2) call file_ref_get() they fail as
atomic_long_add_negative() will return true.
At the same time, both (1) and (2) all file_ref_get() under
mutexes that __fput() must also acquire preventing
kmem_cache_free() from freeing the file.
So while this might be treated as a change in semantics for (1) and
(2) it really isn't. It if should end up causing issues this can be
fixed by adding a helper that does something like:
long cnt = atomic_long_read(&ref->refcnt);
do {
if (cnt < 0)
return false;
} while (!atomic_long_try_cmpxchg(&ref->refcnt, &cnt, cnt + 1));
return true;
which would block FILE_REF_NOREF to FILE_REF_ONEREF transitions.
- Jann correctly pointed out that kmem_cache_zalloc() cannot be used
anymore once files have been ported to file_ref_t.
The kmem_cache_zalloc() call will memset() the whole struct file to
zero when it is reallocated. This will also set file->f_ref to zero
which mens that a concurrent file_ref_get() can return true:
CPU1 CPU2
__get_file_rcu()
rcu_dereference_raw()
close()
[frees file]
alloc_empty_file()
kmem_cache_zalloc()
[reallocates same file]
memset(..., 0, ...)
file_ref_get()
[increments 0->1, returns true]
init_file()
file_ref_init(..., 1)
[sets to 0]
rcu_dereference_raw()
fput()
file_ref_put()
[decrements 0->FILE_REF_NOREF, frees file]
[UAF]
causing a concurrent __get_file_rcu() call to acquire a reference to
the file that is about to be reallocated and immediately freeing it
on realizing that it has been recycled. This causes a UAF for the
task that reallocated/recycled the file.
This is prevented by switching from kmem_cache_zalloc() to
kmem_cache_alloc() and initializing the fields manually. With
file->f_ref initialized last.
Note that a memset() also isn't guaranteed to atomically update an
unsigned long so it's theoretically possible to see torn and
therefore bogus counter values.
Link: https://lore.kernel.org/r/20241007-brauner-file-rcuref-v2-3-387e24dc9163@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next
Kalle Valo says:
====================
wireless-next patches for v6.13
The first -next "new features" pull request for v6.13. This is a big
one as we have not been able to send one earlier. We have also some
patches affecting other subsystems: in staging we deleted the rtl8192e
driver and in debugfs added a new interface to save struct
file_operations memory; both were acked by GregKH.
Because of the lib80211/libipw move there were quite a lot of
conflicts and to solve those we decided to merge net-next into
wireless-next.
Major changes:
cfg80211/mac80211
* stop exporting wext symbols
* new mac80211 op to indicate that a new interface is to be added
* support radio separation of multi-band devices
Wireless Extensions
* move wext spy implementation to libiw
* remove iw_public_data from struct net_device
brcmfmac
* optional LPO clock support
ipw2x00
* move remaining lib80211 code into libiw
wilc1000
* WILC3000 support
rtw89
* RTL8852BE and RTL8852BE-VT BT-coexistence improvements
* tag 'wireless-next-2024-10-25' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next: (126 commits)
mac80211: Remove NOP call to ieee80211_hw_config
wifi: iwlwifi: work around -Wenum-compare-conditional warning
wifi: mac80211: re-order assigning channel in activate links
wifi: mac80211: convert debugfs files to short fops
debugfs: add small file operations for most files
wifi: mac80211: remove misleading j_0 construction parts
wifi: mac80211_hwsim: use hrtimer_active()
wifi: mac80211: refactor BW limitation check for CSA parsing
wifi: mac80211: filter on monitor interfaces based on configured channel
wifi: mac80211: refactor ieee80211_rx_monitor
wifi: mac80211: add support for the monitor SKIP_TX flag
wifi: cfg80211: add monitor SKIP_TX flag
wifi: mac80211: add flag to opt out of virtual monitor support
wifi: cfg80211: pass net_device to .set_monitor_channel
wifi: mac80211: remove status->ampdu_delimiter_crc
wifi: cfg80211: report per wiphy radio antenna mask
wifi: mac80211: use vif radio mask to limit creating chanctx
wifi: mac80211: use vif radio mask to limit ibss scan frequencies
wifi: cfg80211: add option for vif allowed radios
wifi: iwlwifi: allow IWL_FW_CHECK() with just a string
...
====================
Link: https://patch.msgid.link/20241025170705.5F6B2C4CEC3@smtp.kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
When the value of lp is 0 at the beginning of the for loop, it will
become negative in the next assignment and we should bail out.
Reported-by: syzbot+412dea214d8baa3f7483@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=412dea214d8baa3f7483
Tested-by: syzbot+412dea214d8baa3f7483@syzkaller.appspotmail.com
Signed-off-by: Nihar Chaithanya <niharchaithanya@gmail.com>
Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
|
|
Commit 7c55b78818cf ("jfs: xattr: fix buffer overflow for invalid xattr")
also addresses this issue but it only fixes it for positive values, while
ea_size is an integer type and can take negative values, e.g. in case of
a corrupted filesystem. This still breaks validation and would overflow
because of implicit conversion from int to size_t in print_hex_dump().
Fix this issue by clamping the ea_size value instead.
Found by Linux Verification Center (linuxtesting.org) with Syzkaller.
Cc: stable@vger.kernel.org
Signed-off-by: Artem Sadovnikov <ancowi69@gmail.com>
Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
|
|
The stbl might contain some invalid values. Added a check to
return error code in that case.
Reported-by: syzbot+0315f8fe99120601ba88@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=0315f8fe99120601ba88
Signed-off-by: Ghanshyam Agrawal <ghanshyam1898@gmail.com>
Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
|
|
When dmt_budmin is less than zero, it causes errors
in the later stages. Added a check to return an error beforehand
in dbAllocCtl itself.
Reported-by: syzbot+b5ca8a249162c4b9a7d0@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=b5ca8a249162c4b9a7d0
Signed-off-by: Ghanshyam Agrawal <ghanshyam1898@gmail.com>
Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
|
|
The value of stbl can be sometimes out of bounds due
to a bad filesystem. Added a check with appopriate return
of error code in that case.
Reported-by: syzbot+65fa06e29859e41a83f3@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=65fa06e29859e41a83f3
Signed-off-by: Ghanshyam Agrawal <ghanshyam1898@gmail.com>
Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
|
|
Mounting btrfs from two images (which have the same one fsid and two
different dev_uuids) in certain executing order may trigger an UAF for
variable 'device->bdev_file' in __btrfs_free_extra_devids(). And
following are the details:
1. Attach image_1 to loop0, attach image_2 to loop1, and scan btrfs
devices by ioctl(BTRFS_IOC_SCAN_DEV):
/ btrfs_device_1 → loop0
fs_device
\ btrfs_device_2 → loop1
2. mount /dev/loop0 /mnt
btrfs_open_devices
btrfs_device_1->bdev_file = btrfs_get_bdev_and_sb(loop0)
btrfs_device_2->bdev_file = btrfs_get_bdev_and_sb(loop1)
btrfs_fill_super
open_ctree
fail: btrfs_close_devices // -ENOMEM
btrfs_close_bdev(btrfs_device_1)
fput(btrfs_device_1->bdev_file)
// btrfs_device_1->bdev_file is freed
btrfs_close_bdev(btrfs_device_2)
fput(btrfs_device_2->bdev_file)
3. mount /dev/loop1 /mnt
btrfs_open_devices
btrfs_get_bdev_and_sb(&bdev_file)
// EIO, btrfs_device_1->bdev_file is not assigned,
// which points to a freed memory area
btrfs_device_2->bdev_file = btrfs_get_bdev_and_sb(loop1)
btrfs_fill_super
open_ctree
btrfs_free_extra_devids
if (btrfs_device_1->bdev_file)
fput(btrfs_device_1->bdev_file) // UAF !
Fix it by setting 'device->bdev_file' as 'NULL' after closing the
btrfs_device in btrfs_close_one_device().
Fixes: 142388194191 ("btrfs: do not background blkdev_put()")
CC: stable@vger.kernel.org # 4.19+
Link: https://bugzilla.kernel.org/show_bug.cgi?id=219408
Signed-off-by: Zhihao Cheng <chengzhihao1@huawei.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
|
Ensure the refcount and async_copies fields are initialized early.
cleanup_async_copy() will reference these fields if an error occurs
in nfsd4_copy(). If they are not correctly initialized, at the very
least, a refcount underflow occurs.
Reported-by: Olga Kornievskaia <okorniev@redhat.com>
Fixes: aadc3bbea163 ("NFSD: Limit the number of concurrent async COPY operations")
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Tested-by: Olga Kornievskaia <okorniev@redhat.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Add a helper to get the queue_limits from the bdev without having to
poke into the request_queue.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: John Garry <john.g.garry@oracle.com>
Link: https://lore.kernel.org/r/20241029141937.249920-1-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Add NULL check for key returned from bch2_btree_and_journal_iter_peek in
btree_node_iter_and_journal_peek to avoid NULL ptr dereference in
bch2_bkey_buf_reassemble.
When key returned from bch2_btree_and_journal_iter_peek is NULL it means
that btree topology needs repair. Print topology error message with
position at which node wasn't found, its parent node information and
btree_id with level.
Return error code returned by bch2_topology_error to ensure that topology
error is handled properly by recovery.
Reported-by: syzbot+005ef9aa519f30d97657@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=005ef9aa519f30d97657
Fixes: 5222a4607cd8 ("bcachefs: BTREE_ITER_WITH_JOURNAL")
Suggested-by: Alan Huang <mmpgouride@gmail.com>
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>
|
|
The function ec_new_stripe_head_alloc() returns nullptr if kzalloc()
fails. It is crucial to verify its return value before dereferencing
it to avoid a potential nullptr dereference.
Fixes: 035d72f72c91 ("bcachefs: bch2_ec_stripe_head_get() now checks for change in rw devices")
Signed-off-by: Gaosheng Cui <cuigaosheng1@huawei.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
|
|
Open buckets on the partial list should not count as allocated when
we're trying to allocate from the partial list.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
|
|
these are an important source of stranded buckets we need to be able to
watch
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
|
|
We had a bug report where the data update path was creating an extent
that failed to validate because it had too many pointers; almost all of
them were cached.
To fix this, we have:
- want_cached_ptr(), a new helper that checks if we even want a cached
pointer (is on appropriate target, device is readable).
- bch2_extent_set_ptr_cached() now only sets a pointer cached if we want
it.
- bch2_extent_normalize_by_opts() now ensures that we only have a single
cached pointer that we want.
While working on this, it was noticed that this doesn't work well with
reflinked data and per-file options. Another patch series is coming that
plumbs through additional io path options through bch_extent_rebalance,
with improved option handling.
Reported-by: Reed Riley <reed@riley.engineer>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
|
|
Initialize freespace_initialized bits to 0 in member's flags and update
member's cached version for each device in bch2_fs_initialize.
It's possible for the bits to be set to 1 before fs is initialized and if
call to bch2_trans_mark_dev_sbs (just before bch2_fs_freespace_init) fails
bits remain to be 1 which can later indirectly trigger BUG condition in
bch2_bucket_alloc_freelist during shutdown.
Reported-by: syzbot+2b6a17991a6af64f9489@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=2b6a17991a6af64f9489
Fixes: bbe682c76789 ("bcachefs: Ensure devices are always correctly initialized")
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>
|
|
This used to not matter, but now we're being more strict.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
|
|
Syzbot reported that in directory operations after nilfs2 detects
filesystem corruption and degrades to read-only,
__block_write_begin_int(), which is called to prepare block writes, may
fail the BUG_ON check for accesses exceeding the folio/page size,
triggering a kernel bug.
This was found to be because the "checked" flag of a page/folio was not
cleared when it was discarded by nilfs2's own routine, which causes the
sanity check of directory entries to be skipped when the directory
page/folio is reloaded. So, fix that.
This was necessary when the use of nilfs2's own page discard routine was
applied to more than just metadata files.
Link: https://lkml.kernel.org/r/20241017193359.5051-1-konishi.ryusuke@gmail.com
Fixes: 8c26c4e2694a ("nilfs2: fix issue with flush kernel thread after remount in RO mode because of driver's internal error or metadata corruption")
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@gmail.com>
Reported-by: syzbot+d6ca2daf692c7a82f959@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=d6ca2daf692c7a82f959
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Syzbot reported a kernel BUG in ocfs2_truncate_inline. There are two
reasons for this: first, the parameter value passed is greater than
ocfs2_max_inline_data_with_xattr, second, the start and end parameters of
ocfs2_truncate_inline are "unsigned int".
So, we need to add a sanity check for byte_start and byte_len right before
ocfs2_truncate_inline() in ocfs2_remove_inode_range(), if they are greater
than ocfs2_max_inline_data_with_xattr return -EINVAL.
Link: https://lkml.kernel.org/r/tencent_D48DB5122ADDAEDDD11918CFB68D93258C07@qq.com
Fixes: 1afc32b95233 ("ocfs2: Write support for inline data")
Signed-off-by: Edward Adam Davis <eadavis@qq.com>
Reported-by: syzbot+81092778aac03460d6b7@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=81092778aac03460d6b7
Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Gang He <ghe@suse.com>
Cc: Jun Piao <piaojun@huawei.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Patch series "fork: do not expose incomplete mm on fork".
During fork we may place the virtual memory address space into an
inconsistent state before the fork operation is complete.
In addition, we may encounter an error during the fork operation that
indicates that the virtual memory address space is invalidated.
As a result, we should not be exposing it in any way to external machinery
that might interact with the mm or VMAs, machinery that is not designed to
deal with incomplete state.
We specifically update the fork logic to defer khugepaged and ksm to the
end of the operation and only to be invoked if no error arose, and
disallow uffd from observing fork events should an error have occurred.
This patch (of 2):
Currently on fork we expose the virtual address space of a process to
userland unconditionally if uffd is registered in VMAs, regardless of
whether an error arose in the fork.
This is performed in dup_userfaultfd_complete() which is invoked
unconditionally, and performs two duties - invoking registered handlers
for the UFFD_EVENT_FORK event via dup_fctx(), and clearing down
userfaultfd_fork_ctx objects established in dup_userfaultfd().
This is problematic, because the virtual address space may not yet be
correctly initialised if an error arose.
The change in commit d24062914837 ("fork: use __mt_dup() to duplicate
maple tree in dup_mmap()") makes this more pertinent as we may be in a
state where entries in the maple tree are not yet consistent.
We address this by, on fork error, ensuring that we roll back state that
we would otherwise expect to clean up through the event being handled by
userland and perform the memory freeing duty otherwise performed by
dup_userfaultfd_complete().
We do this by implementing a new function, dup_userfaultfd_fail(), which
performs the same loop, only decrementing reference counts.
Note that we perform mmgrab() on the parent and child mm's, however
userfaultfd_ctx_put() will mmdrop() this once the reference count drops to
zero, so we will avoid memory leaks correctly here.
Link: https://lkml.kernel.org/r/cover.1729014377.git.lorenzo.stoakes@oracle.com
Link: https://lkml.kernel.org/r/d3691d58bb58712b6fb3df2be441d175bd3cdf07.1729014377.git.lorenzo.stoakes@oracle.com
Fixes: d24062914837 ("fork: use __mt_dup() to duplicate maple tree in dup_mmap()")
Signed-off-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Reported-by: Jann Horn <jannh@google.com>
Reviewed-by: Jann Horn <jannh@google.com>
Reviewed-by: Liam R. Howlett <Liam.Howlett@Oracle.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Linus Torvalds <torvalds@linuxfoundation.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Enable casefold lookup in tmpfs, based on the encoding defined by
userspace. That means that instead of comparing byte per byte a file
name, it compares to a case-insensitive equivalent of the Unicode
string.
* Dcache handling
There's a special need when dealing with case-insensitive dentries.
First of all, we currently invalidated every negative casefold dentries.
That happens because currently VFS code has no proper support to deal
with that, giving that it could incorrectly reuse a previous filename
for a new file that has a casefold match. For instance, this could
happen:
$ mkdir DIR
$ rm -r DIR
$ mkdir dir
$ ls
DIR/
And would be perceived as inconsistency from userspace point of view,
because even that we match files in a case-insensitive manner, we still
honor whatever is the initial filename.
Along with that, tmpfs stores only the first equivalent name dentry used
in the dcache, preventing duplications of dentries in the dcache. The
d_compare() version for casefold files uses a normalized string, so the
filename under lookup will be compared to another normalized string for
the existing file, achieving a casefolded lookup.
* Enabling casefold via mount options
Most filesystems have their data stored in disk, so casefold option need
to be enabled when building a filesystem on a device (via mkfs).
However, as tmpfs is a RAM backed filesystem, there's no disk
information and thus no mkfs to store information about casefold.
For tmpfs, create casefold options for mounting. Userspace can then
enable casefold support for a mount point using:
$ mount -t tmpfs -o casefold=utf8-12.1.0 fs_name mount_dir/
Userspace must set what Unicode standard is aiming to. The available
options depends on what the kernel Unicode subsystem supports.
And for strict encoding:
$ mount -t tmpfs -o casefold=utf8-12.1.0,strict_encoding fs_name mount_dir/
Strict encoding means that tmpfs will refuse to create invalid UTF-8
sequences. When this option is not enabled, any invalid sequence will be
treated as an opaque byte sequence, ignoring the encoding thus not being
able to be looked up in a case-insensitive way.
* Check for casefold dirs on simple_lookup()
On simple_lookup(), do not create dentries for casefold directories.
Currently, VFS does not support case-insensitive negative dentries and
can create inconsistencies in the filesystem. Prevent such dentries to
being created in the first place.
Reviewed-by: Gabriel Krisman Bertazi <gabriel@krisman.be>
Reviewed-by: Gabriel Krisman Bertazi <krisman@suse.de>
Signed-off-by: André Almeida <andrealmeid@igalia.com>
Link: https://lore.kernel.org/r/20241021-tonyk-tmpfs-v8-6-f443d5814194@igalia.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
Export generic_ci_ dentry functions so they can be used by
case-insensitive filesystems that need something more custom than the
default one set by `struct generic_ci_dentry_ops`.
Reviewed-by: Gabriel Krisman Bertazi <gabriel@krisman.be>
Signed-off-by: André Almeida <andrealmeid@igalia.com>
Link: https://lore.kernel.org/r/20241021-tonyk-tmpfs-v8-5-f443d5814194@igalia.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
All filesystems that currently support UTF-8 casefold can fetch the
UTF-8 version from the filesystem metadata stored on disk. They can get
the data stored and directly match it to a integer, so they can skip the
string parsing step, which motivated the removal of this function in the
first place.
However, for tmpfs, the only way to tell the kernel which UTF-8 version
we are about to use is via mount options, using a string. Re-introduce
utf8_parse_version() to be used by tmpfs.
This version differs from the original by skipping the intermediate step
of copying the version string to an auxiliary string before calling
match_token(). This versions calls match_token() in the argument string.
The paramenters are simpler now as well.
utf8_parse_version() was created by 9d53690f0d4 ("unicode: implement
higher level API for string handling") and later removed by 49bd03cc7e9
("unicode: pass a UNICODE_AGE() tripple to utf8_load").
Signed-off-by: André Almeida <andrealmeid@igalia.com>
Link: https://lore.kernel.org/r/20241021-tonyk-tmpfs-v8-4-f443d5814194@igalia.com
Reviewed-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Gabriel Krisman Bertazi <krisman@suse.de>
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
Export latest available UTF-8 version number so filesystems can easily
load the newest one.
Signed-off-by: André Almeida <andrealmeid@igalia.com>
Link: https://lore.kernel.org/r/20241021-tonyk-tmpfs-v8-3-f443d5814194@igalia.com
Acked-by: Gabriel Krisman Bertazi <krisman@suse.de>
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
Use the helper function to check the requirements for casefold
directories using strict encoding.
Suggested-by: Gabriel Krisman Bertazi <krisman@suse.de>
Signed-off-by: André Almeida <andrealmeid@igalia.com>
Link: https://lore.kernel.org/r/20241021-tonyk-tmpfs-v8-2-f443d5814194@igalia.com
Acked-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Gabriel Krisman Bertazi <krisman@suse.de>
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
Most of the callers of wbc_account_cgroup_owner() are converting a folio
to page before calling the function. wbc_account_cgroup_owner() is
converting the page back to a folio to call mem_cgroup_css_from_folio().
Convert wbc_account_cgroup_owner() to take a folio instead of a page,
and convert all callers to pass a folio directly except f2fs.
Convert the page to folio for all the callers from f2fs as they were the
only callers calling wbc_account_cgroup_owner() with a page. As f2fs is
already in the process of converting to folios, these call sites might
also soon be calling wbc_account_cgroup_owner() with a folio directly in
the future.
No functional changes. Only compile tested.
Signed-off-by: Pankaj Raghav <p.raghav@samsung.com>
Link: https://lore.kernel.org/r/20240926140121.203821-1-kernel@pankajraghav.com
Acked-by: David Sterba <dsterba@suse.com>
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
I was so sure the per-dentry expire timeout patch worked ok but my
testing was flawed.
In validate_dev_ioctl() the check for ioctl AUTOFS_DEV_IOCTL_TIMEOUT_CMD
should use the ioctl number not the passed in ioctl command.
Fixes: 433f9d76a010 ("autofs: add per dentry expire timeout")
Cc: <stable@vger.kernel.org> # mainline only
Signed-off-by: Ian Kent <raven@themaw.net>
Link: https://lore.kernel.org/r/20241027224732.5507-1-raven@themaw.net
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
xa_store() can fail, it return xa_err(-EINVAL) if the entry cannot
be stored in an XArray, or xa_err(-ENOMEM) if memory allocation failed,
so check error for xa_store() to fix it.
Cc: stable@vger.kernel.org
Fixes: b685757c7b08 ("ksmbd: Implements sess->rpc_handle_list as xarray")
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Pull xfs fixes from Carlos Maiolino:
- Fix recovery of allocator ops after a growfs
- Do not fail repairs on metadata files with no attr fork
* tag 'xfs-6.12-fixes-5' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
xfs: update the pag for the last AG at recovery time
xfs: don't use __GFP_RETRY_MAYFAIL in xfs_initialize_perag
xfs: error out when a superblock buffer update reduces the agcount
xfs: update the file system geometry after recoverying superblock buffers
xfs: merge the perag freeing helpers
xfs: pass the exact range to initialize to xfs_initialize_perag
xfs: don't fail repairs on metadata files with no attr fork
|
|
Pull more 9p reverts from Dominique Martinet:
"Revert patches causing inode collision problems.
The code simplification introduced significant regressions on servers
that do not remap inode numbers when exporting multiple underlying
filesystems with colliding inodes. See the top-most revert (commit
be2ca3825372) for details.
This problem had been ignored for too long and the reverts will also
head to stable (6.9+).
I'm confident this set of patches gets us back to previous behaviour
(another related patch had already been reverted back in April and
we're almost back to square 1, and the rest didn't touch inode
lifecycle)"
* tag '9p-for-6.12-rc5' of https://github.com/martinetd/linux:
Revert "fs/9p: simplify iget to remove unnecessary paths"
Revert "fs/9p: fix uaf in in v9fs_stat2inode_dotl"
Revert "fs/9p: remove redundant pointer v9ses"
Revert " fs/9p: mitigate inode collisions"
|
|
Pull smb client fixes from Steve French:
- Fix init module error caseb
- Fix memory allocation error path (for passwords) in mount
* tag 'v6.12-rc4-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
cifs: fix warning when destroy 'cifs_io_request_pool'
smb: client: Handle kstrdup failures for passwords
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse
Pull fuse fixes from Miklos Szeredi:
- Fix cached size after passthrough writes
This fix needed a trivial change in the backing-file API, which
resulted in some non-fuse files being touched.
- Revert a commit meant as a cleanup but which triggered a WARNING
- Remove a stray debug line left-over
* tag 'fuse-fixes-6.12-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse:
fuse: remove stray debug line
Revert "fuse: move initialization of fuse_file to fuse_writepages() instead of in callback"
fuse: update inode size after extending passthrough write
fs: pass offset and result to backing_file end_write() callback
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux
Pull nfsd fixes from Chuck Lever:
- Fix a couple of use-after-free bugs
* tag 'nfsd-6.12-2' of git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux:
nfsd: cancel nfsd_shrinker_work using sync mode in nfs4_state_shutdown_net
nfsd: fix race between laundromat and free_stateid
|
|
write_super() -> sb_counters_from_cpu() may reallocate the superblock
Reported-by: syzbot+9fc4dac4775d07bcfe34@syzkaller.appspotmail.com
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
|
|
c->btree_roots_known[i].b can be NULL. In this case, a NULL pointer dereference
occurs, so you need to add code to check the variable.
Reported-by: syzbot+b468b9fef56949c3b528@syzkaller.appspotmail.com
Fixes: 7773df19c35f ("bcachefs: metadata version bucket_stripe_sectors")
Signed-off-by: Jeongjun Park <aha310510@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
|
|
It wasn't there when the patch was posted for review, but somehow made it
into the pull.
Link: https://lore.kernel.org/all/20240913104703.1673180-1-mszeredi@redhat.com/
Fixes: efad7153bf93 ("fuse: allow O_PATH fd for FUSE_DEV_IOC_BACKING_OPEN")
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
|
|
The size of a.data_type is set abnormally large, causing shift-out-of-bounds.
To fix this, we need to add validation on a.data_type in
alloc_lru_idx_fragmentation().
Reported-by: syzbot+7f45fa9805c40db3f108@syzkaller.appspotmail.com
Fixes: 260af1562ec1 ("bcachefs: Kill alloc_v4.fragmentation_lru")
Signed-off-by: Jeongjun Park <aha310510@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
|
|
Add check on layout->sb_max_size_bits against BCH_SB_LAYOUT_SIZE_BITS_MAX
to prevent UBSAN shift-out-of-bounds in validate_sb_layout().
Reported-by: syzbot+089fad5a3a5e77825426@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=089fad5a3a5e77825426
Fixes: 03ef80b469d5 ("bcachefs: Ignore unknown mount options")
Tested-by: syzbot+089fad5a3a5e77825426@syzkaller.appspotmail.com
Signed-off-by: Gianfranco Trad <gianf.trad@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
|
|
This reverts commit 724a08450f74b02bd89078a596fd24857827c012.
This code simplification introduced significant regressions on servers
that do not remap inode numbers when exporting multiple underlying
filesystems with colliding inodes, as can be illustrated with simple
tmpfs exports in qemu with remapping disabled:
```
# host side
cd /tmp/linux-test
mkdir m1 m2
mount -t tmpfs tmpfs m1
mount -t tmpfs tmpfs m2
mkdir m1/dir m2/dir
echo foo > m1/dir/foo
echo bar > m2/dir/bar
# guest side
# started with -virtfs local,path=/tmp/linux-test,mount_tag=tmp,security_model=mapped-file
mount -t 9p -o trans=virtio,debug=1 tmp /mnt/t
ls /mnt/t/m1/dir
# foo
ls /mnt/t/m2/dir
# bar (works ok if directry isn't open)
# cd to keep first dir's inode alive
cd /mnt/t/m1/dir
ls /mnt/t/m2/dir
# foo (should be bar)
```
Other examples can be crafted with regular files with fscache enabled,
in which case I/Os just happen to the wrong file leading to
corruptions, or guest failing to boot with:
| VFS: Lookup of 'com.android.runtime' in 9p 9p would have caused loop
In theory, we'd want the servers to be smart enough and ensure they
never send us two different files with the same 'qid.path', but while
qemu has an option to remap that is recommended (and qemu prints a
warning if this case happens), there are many other servers which do
not (kvmtool, nfs-ganesha, probably diod...), we should at least ensure
we don't cause regressions on this:
- assume servers can't be trusted and operations that should get a 'new'
inode properly do so. commit d05dcfdf5e16 (" fs/9p: mitigate inode
collisions") attempted to do this, but v9fs_fid_iget_dotl() was not
called so some higher level of caching got in the way; this needs to be
fixed properly before we can re-apply the patches.
- if we ever want to really simplify this code, we will need to add some
negotiation with the server at mount time where the server could claim
they handle this properly, at which point we could optimize this out.
(but that might not be needed at all if we properly handle the 'new'
check?)
Fixes: 724a08450f74 ("fs/9p: simplify iget to remove unnecessary paths")
Reported-by: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/all/20240408141436.GA17022@redhat.com/
Link: https://lkml.kernel.org/r/20240923100508.GA32066@willie-the-truck
Cc: stable@vger.kernel.org # v6.9+
Message-ID: <20241024-revert_iget-v1-4-4cac63d25f72@codewreck.org>
Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
|
|
This reverts commit 11763a8598f888dec631a8a903f7ada32181001f.
This is a requirement to revert commit 724a08450f74 ("fs/9p: simplify
iget to remove unnecessary paths"), see that revert for details.
Fixes: 724a08450f74 ("fs/9p: simplify iget to remove unnecessary paths")
Reported-by: Will Deacon <will@kernel.org>
Link: https://lkml.kernel.org/r/20240923100508.GA32066@willie-the-truck
Cc: stable@vger.kernel.org # v6.9+
Message-ID: <20241024-revert_iget-v1-3-4cac63d25f72@codewreck.org>
Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
|
|
This reverts commit 10211b4a23cf4a3df5c11a10e5b3d371f16a906f.
This is a requirement to revert commit 724a08450f74 ("fs/9p: simplify
iget to remove unnecessary paths"), see that revert for details.
Fixes: 724a08450f74 ("fs/9p: simplify iget to remove unnecessary paths")
Reported-by: Will Deacon <will@kernel.org>
Link: https://lkml.kernel.org/r/20240923100508.GA32066@willie-the-truck
Cc: stable@vger.kernel.org # v6.9+
Message-ID: <20241024-revert_iget-v1-2-4cac63d25f72@codewreck.org>
Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
|
|
This reverts commit d05dcfdf5e1659b2949d13060284eff3888b644e.
This is a requirement to revert commit 724a08450f74 ("fs/9p: simplify
iget to remove unnecessary paths"), see that revert for details.
Fixes: 724a08450f74 ("fs/9p: simplify iget to remove unnecessary paths")
Reported-by: Will Deacon <will@kernel.org>
Link: https://lkml.kernel.org/r/20240923100508.GA32066@willie-the-truck
Cc: stable@vger.kernel.org # v6.9+
Message-ID: <20241024-revert_iget-v1-1-4cac63d25f72@codewreck.org>
Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux
Pull btrfs fixes from David Sterba:
- mount option fixes:
- fix handling of compression mount options on remount
- reject rw remount in case there are options that don't work
in read-write mode (like rescue options)
- fix zone accounting of unusable space
- fix in-memory corruption when merging extent maps
- fix delalloc range locking for sector < page
- use more convenient default value of drop subtree threshold, clean
more subvolumes without the fallback to marking quotas inconsistent
- fix smatch warning about incorrect value passed to ERR_PTR
* tag 'for-6.12-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
btrfs: fix passing 0 to ERR_PTR in btrfs_search_dir_index_item()
btrfs: reject ro->rw reconfiguration if there are hard ro requirements
btrfs: fix read corruption due to race with extent map merging
btrfs: fix the delalloc range locking if sector size < page size
btrfs: qgroup: set a more sane default value for subtree drop threshold
btrfs: clear force-compress on remount when compress mount option is given
btrfs: zoned: fix zone unusable accounting for freed reserved extent
|
|
Pull jfs fix from David Kleikamp:
"Fix a regression introduced in 6.12-rc1"
* tag 'jfs-6.12-rc5' of github.com:kleikamp/linux-shaggy:
jfs: Fix sanity check in dbMount
|
|
Pull bcachefs fixes from Kent Overstreet:
"Lots of hotfixes:
- transaction restart injection has been shaking out a few things
- fix a data corruption in the buffered write path on -ENOSPC, found
by xfstests generic/299
- Some small show_options fixes
- Repair mismatches in inode hash type, seed: different snapshot
versions of an inode must have the same hash/type seed, used for
directory entries and xattrs. We were checking the hash seed, but
not the type, and a user contributed a filesystem where the hash
type on one inode had somehow been flipped; these fixes allow his
filesystem to repair.
Additionally, the hash type flip made some directory entries
invisible, which were then recreated by userspace; so the hash
check code now checks for duplicate non dangling dirents, and
renames one of them if necessary.
- Don't use wait_event_interruptible() in recovery: this fixes some
filesystems failing to mount with -ERESTARTSYS
- Workaround for kvmalloc not supporting > INT_MAX allocations,
causing an -ENOMEM when allocating the sorted array of journal
keys: this allows a 75 TB filesystem to mount
- Make sure bch_inode_unpacked.bi_snapshot is set in the old inode
compat path: this alllows Marcin's filesystem (in use since before
6.7) to repair and mount"
* tag 'bcachefs-2024-10-22' of https://github.com/koverstreet/bcachefs: (26 commits)
bcachefs: Set bch_inode_unpacked.bi_snapshot in old inode path
bcachefs: Mark more errors as AUTOFIX
bcachefs: Workaround for kvmalloc() not supporting > INT_MAX allocations
bcachefs: Don't use wait_event_interruptible() in recovery
bcachefs: Fix __bch2_fsck_err() warning
bcachefs: fsck: Improve hash_check_key()
bcachefs: bch2_hash_set_or_get_in_snapshot()
bcachefs: Repair mismatches in inode hash seed, type
bcachefs: Add hash seed, type to inode_to_text()
bcachefs: INODE_STR_HASH() for bch_inode_unpacked
bcachefs: Run in-kernel offline fsck without ratelimit errors
bcachefs: skip mount option handle for empty string.
bcachefs: fix incorrect show_options results
bcachefs: Fix data corruption on -ENOSPC in buffered write path
bcachefs: bch2_folio_reservation_get_partial() is now better behaved
bcachefs: fix disk reservation accounting in bch2_folio_reservation_get()
bcachefS: ec: fix data type on stripe deletion
bcachefs: Don't use commit_do() unnecessarily
bcachefs: handle restarts in bch2_bucket_io_time_reset()
bcachefs: fix restart handling in __bch2_resume_logged_op_finsert()
...
|
|
This reverts commit 1325e4a91a405f88f1b18626904d37860a4f9069.
using multipage folios apparently break some madvise operations like
MADV_PAGEOUT which do not reliably unload the specified page anymore,
Revert the patch until that is figured out.
Reported-by: Andrii Nakryiko <andrii@kernel.org>
Fixes: 1325e4a91a40 ("9p: Enable multipage folios")
Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
A common pattern when using pid fds is having to get information
about the process, which currently requires /proc being mounted,
resolving the fd to a pid, and then do manual string parsing of
/proc/N/status and friends. This needs to be reimplemented over
and over in all userspace projects (e.g.: I have reimplemented
resolving in systemd, dbus, dbus-daemon, polkit so far), and
requires additional care in checking that the fd is still valid
after having parsed the data, to avoid races.
Having a programmatic API that can be used directly removes all
these requirements, including having /proc mounted.
As discussed at LPC24, add an ioctl with an extensible struct
so that more parameters can be added later if needed. Start with
returning pid/tgid/ppid and creds unconditionally, and cgroupid
optionally.
Signed-off-by: Luca Boccassi <luca.boccassi@gmail.com>
Link: https://lore.kernel.org/r/20241010155401.2268522-1-luca.boccassi@gmail.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
When rename moves an AFS subdirectory between parent directories, the
subdir also needs a bit of editing: the ".." entry needs updating to point
to the new parent (though I don't make use of the info) and the DV needs
incrementing by 1 to reflect the change of content. The server also sends
a callback break notification on the subdirectory if we have one, but we
can take care of recovering the promise next time we access the subdir.
This can be triggered by something like:
mount -t afs %example.com:xfstest.test20 /xfstest.test/
mkdir /xfstest.test/{aaa,bbb,aaa/ccc}
touch /xfstest.test/bbb/ccc/d
mv /xfstest.test/{aaa/ccc,bbb/ccc}
touch /xfstest.test/bbb/ccc/e
When the pathwalk for the second touch hits "ccc", kafs spots that the DV
is incorrect and downloads it again (so the fix is not critical).
Fix this, if the rename target is a directory and the old and new
parents are different, by:
(1) Incrementing the DV number of the target locally.
(2) Editing the ".." entry in the target to refer to its new parent's
vnode ID and uniquifier.
Link: https://lore.kernel.org/r/3340431.1729680010@warthog.procyon.org.uk
Fixes: 63a4681ff39c ("afs: Locally edit directory data for mkdir/create/unlink/...")
cc: David Howells <dhowells@redhat.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
The purpose of btrfs_bbio_propagate_error() shall be propagating an error
of split bio to its original btrfs_bio, and tell the error to the upper
layer. However, it's not working well on some cases.
* Case 1. Immediate (or quick) end_bio with an error
When btrfs sends btrfs_bio to mirrored devices, btrfs calls
btrfs_bio_end_io() when all the mirroring bios are completed. If that
btrfs_bio was split, it is from btrfs_clone_bioset and its end_io function
is btrfs_orig_write_end_io. For this case, btrfs_bbio_propagate_error()
accesses the orig_bbio's bio context to increase the error count.
That works well in most cases. However, if the end_io is called enough
fast, orig_bbio's (remaining part after split) bio context may not be
properly set at that time. Since the bio context is set when the orig_bbio
(the last btrfs_bio) is sent to devices, that might be too late for earlier
split btrfs_bio's completion. That will result in NULL pointer
dereference.
That bug is easily reproducible by running btrfs/146 on zoned devices [1]
and it shows the following trace.
[1] You need raid-stripe-tree feature as it create "-d raid0 -m raid1" FS.
BUG: kernel NULL pointer dereference, address: 0000000000000020
#PF: supervisor read access in kernel mode
#PF: error_code(0x0000) - not-present page
PGD 0 P4D 0
Oops: Oops: 0000 [#1] PREEMPT SMP PTI
CPU: 1 UID: 0 PID: 13 Comm: kworker/u32:1 Not tainted 6.11.0-rc7-BTRFS-ZNS+ #474
Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
Workqueue: writeback wb_workfn (flush-btrfs-5)
RIP: 0010:btrfs_bio_end_io+0xae/0xc0 [btrfs]
BTRFS error (device dm-0): bdev /dev/mapper/error-test errs: wr 2, rd 0, flush 0, corrupt 0, gen 0
RSP: 0018:ffffc9000006f248 EFLAGS: 00010246
RAX: 0000000000000000 RBX: ffff888005a7f080 RCX: ffffc9000006f1dc
RDX: 0000000000000000 RSI: 000000000000000a RDI: ffff888005a7f080
RBP: ffff888011dfc540 R08: 0000000000000000 R09: 0000000000000001
R10: ffffffff82e508e0 R11: 0000000000000005 R12: ffff88800ddfbe58
R13: ffff888005a7f080 R14: ffff888005a7f158 R15: ffff888005a7f158
FS: 0000000000000000(0000) GS:ffff88803ea80000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000020 CR3: 0000000002e22006 CR4: 0000000000370ef0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
<TASK>
? __die_body.cold+0x19/0x26
? page_fault_oops+0x13e/0x2b0
? _printk+0x58/0x73
? do_user_addr_fault+0x5f/0x750
? exc_page_fault+0x76/0x240
? asm_exc_page_fault+0x22/0x30
? btrfs_bio_end_io+0xae/0xc0 [btrfs]
? btrfs_log_dev_io_error+0x7f/0x90 [btrfs]
btrfs_orig_write_end_io+0x51/0x90 [btrfs]
dm_submit_bio+0x5c2/0xa50 [dm_mod]
? find_held_lock+0x2b/0x80
? blk_try_enter_queue+0x90/0x1e0
__submit_bio+0xe0/0x130
? ktime_get+0x10a/0x160
? lockdep_hardirqs_on+0x74/0x100
submit_bio_noacct_nocheck+0x199/0x410
btrfs_submit_bio+0x7d/0x150 [btrfs]
btrfs_submit_chunk+0x1a1/0x6d0 [btrfs]
? lockdep_hardirqs_on+0x74/0x100
? __folio_start_writeback+0x10/0x2c0
btrfs_submit_bbio+0x1c/0x40 [btrfs]
submit_one_bio+0x44/0x60 [btrfs]
submit_extent_folio+0x13f/0x330 [btrfs]
? btrfs_set_range_writeback+0xa3/0xd0 [btrfs]
extent_writepage_io+0x18b/0x360 [btrfs]
extent_write_locked_range+0x17c/0x340 [btrfs]
? __pfx_end_bbio_data_write+0x10/0x10 [btrfs]
run_delalloc_cow+0x71/0xd0 [btrfs]
btrfs_run_delalloc_range+0x176/0x500 [btrfs]
? find_lock_delalloc_range+0x119/0x260 [btrfs]
writepage_delalloc+0x2ab/0x480 [btrfs]
extent_write_cache_pages+0x236/0x7d0 [btrfs]
btrfs_writepages+0x72/0x130 [btrfs]
do_writepages+0xd4/0x240
? find_held_lock+0x2b/0x80
? wbc_attach_and_unlock_inode+0x12c/0x290
? wbc_attach_and_unlock_inode+0x12c/0x290
__writeback_single_inode+0x5c/0x4c0
? do_raw_spin_unlock+0x49/0xb0
writeback_sb_inodes+0x22c/0x560
__writeback_inodes_wb+0x4c/0xe0
wb_writeback+0x1d6/0x3f0
wb_workfn+0x334/0x520
process_one_work+0x1ee/0x570
? lock_is_held_type+0xc6/0x130
worker_thread+0x1d1/0x3b0
? __pfx_worker_thread+0x10/0x10
kthread+0xee/0x120
? __pfx_kthread+0x10/0x10
ret_from_fork+0x30/0x50
? __pfx_kthread+0x10/0x10
ret_from_fork_asm+0x1a/0x30
</TASK>
Modules linked in: dm_mod btrfs blake2b_generic xor raid6_pq rapl
CR2: 0000000000000020
* Case 2. Earlier completion of orig_bbio for mirrored btrfs_bios
btrfs_bbio_propagate_error() assumes the end_io function for orig_bbio is
called last among split bios. In that case, btrfs_orig_write_end_io() sets
the bio->bi_status to BLK_STS_IOERR by seeing the bioc->error [2].
Otherwise, the increased orig_bio's bioc->error is not checked by anyone
and return BLK_STS_OK to the upper layer.
[2] Actually, this is not true. Because we only increases orig_bioc->errors
by max_errors, the condition "atomic_read(&bioc->error) > bioc->max_errors"
is still not met if only one split btrfs_bio fails.
* Case 3. Later completion of orig_bbio for un-mirrored btrfs_bios
In contrast to the above case, btrfs_bbio_propagate_error() is not working
well if un-mirrored orig_bbio is completed last. It sets
orig_bbio->bio.bi_status to the btrfs_bio's error. But, that is easily
over-written by orig_bbio's completion status. If the status is BLK_STS_OK,
the upper layer would not know the failure.
* Solution
Considering the above cases, we can only save the error status in the
orig_bbio (remaining part after split) itself as it is always
available. Also, the saved error status should be propagated when all the
split btrfs_bios are finished (i.e, bbio->pending_ios == 0).
This commit introduces "status" to btrfs_bbio and saves the first error of
split bios to original btrfs_bio's "status" variable. When all the split
bios are finished, the saved status is loaded into original btrfs_bio's
status.
With this commit, btrfs/146 on zoned devices does not hit the NULL pointer
dereference anymore.
Fixes: 852eee62d31a ("btrfs: allow btrfs_submit_bio to split bios")
CC: stable@vger.kernel.org # 6.6+
Reviewed-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
|
