[ Upstream commit 2214ec4bf8 ]
In 'ocfs2_merge_rec_left()', do not reset 'left_path' to NULL after
move, thus allowing 'ocfs2_free_path()' to free it before return.
Link: https://lkml.kernel.org/r/20251205065159.392749-1-dmantipov@yandex.ru
Fixes: 677b975282 ("ocfs2: Add support for cross extent block")
Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru>
Reported-by: syzbot+cfc7cab3bb6eaa7c4de2@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=cfc7cab3bb6eaa7c4de2
Reviewed-by: Heming Zhao <heming.zhao@suse.com>
Acked-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Jun Piao <piaojun@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 0b873de2c0 ]
struct nfs_local_kiocb used ____cacheline_aligned on its iters[] array
and as the structure evolved it caused a 61 byte hole to form. Fix
this by removing ____cacheline_aligned and reordering iters[] before
iter_is_dio_aligned[].
Fixes: 6a218b9c31 ("nfs/localio: do not issue misaligned DIO out-of-order")
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f50d0328d0 ]
This check to ensure dio_offset_align isn't larger than PAGE_SIZE is
no longer relevant (older iterations of NFS Direct was allocating
misaligned head and tail pages but no longer does, so this check isn't
needed).
Fixes: c817248fc8 ("nfs/localio: add proper O_DIRECT support for READ and WRITE")
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4ae4dde6f3 ]
If a DIO read or an unbuffered read request extends beyond the EOF, the
server will return a short read and a status code indicating that EOF was
hit, which gets translated to -ENODATA. Note that the client does not cap
the request at i_size, but asks for the amount requested in case there's a
race on the server with a third party.
Now, on the client side, the request will get split into multiple
subrequests if rsize is smaller than the full request size. A subrequest
that starts before or at the EOF and returns short data up to the EOF will
be correctly handled, with the NETFS_SREQ_HIT_EOF flag being set,
indicating to netfslib that we can't read more.
If a subrequest, however, starts after the EOF and not at it, HIT_EOF will
not be flagged, its error will be set to -ENODATA and it will be abandoned.
This will cause the request as a whole to fail with -ENODATA.
Fix this by setting NETFS_SREQ_HIT_EOF on any subrequest that lies beyond
the EOF marker.
Fixes: 1da29f2c39 ("netfs, cifs: Fix handling of short DIO read")
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>
cc: Shyam Prasad N <sprasad@microsoft.com>
cc: linux-cifs@vger.kernel.org
cc: netfs@lists.linux.dev
cc: linux-fsdevel@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9d85ac939d ]
If a DIO read or an unbuffered read request extends beyond the EOF, the
server will return a short read and a status code indicating that EOF was
hit, which gets translated to -ENODATA. Note that the client does not cap
the request at i_size, but asks for the amount requested in case there's a
race on the server with a third party.
Now, on the client side, the request will get split into multiple
subrequests if rsize is smaller than the full request size. A subrequest
that starts before or at the EOF and returns short data up to the EOF will
be correctly handled, with the NETFS_SREQ_HIT_EOF flag being set,
indicating to netfslib that we can't read more.
If a subrequest, however, starts after the EOF and not at it, HIT_EOF will
not be flagged, its error will be set to -ENODATA and it will be abandoned.
This will cause the request as a whole to fail with -ENODATA.
Fix this by setting NETFS_SREQ_HIT_EOF on any subrequest that lies beyond
the EOF marker.
This can be reproduced by mounting with "cache=none,sign,vers=1.0" and
doing a read of a file that's significantly bigger than the size of the
file (e.g. attempting to read 64KiB from a 16KiB file).
Fixes: a68c74865f ("cifs: Fix SMB1 readv/writev callback in the same way as SMB2/3")
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>
cc: Shyam Prasad N <sprasad@microsoft.com>
cc: linux-cifs@vger.kernel.org
cc: netfs@lists.linux.dev
cc: linux-fsdevel@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7ee8bc3942 ]
The recent increase in the number of Segment Summary Area (SSA) entries
from 512 to 2048 was an unintentional change in logic of 16kb block
support. This commit corrects the issue.
To better utilize the space available from the erroneous 2048-entry
calculation, we are implementing a solution to share the currently
unused SSA space with neighboring segments. This enhances overall
SSA utilization without impacting the established 8MB segment size.
Fixes: d7e9a9037d ("f2fs: Support Block Size == Page Size")
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit c36f9d7b28 ]
After commit 25524b6190 ("fs/nls: Fix utf16 to utf8 conversion"),
the return values of utf8_to_utf32() and utf32_to_utf8() are
inconsistent when encountering an error: utf8_to_utf32() returns -1,
while utf32_to_utf8() returns errno codes. Fix this inconsistency
by modifying utf8_to_utf32() to return errno codes as well.
Fixes: 25524b6190 ("fs/nls: Fix utf16 to utf8 conversion")
Suggested-by: Andy Shevchenko <andriy.shevchenko@intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com>
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://patch.msgid.link/20251129111535.8984-1-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2b092175f5 ]
Only inherit the block sizes that were actually specified as mount
parameters for the parent mount.
Fixes: 62a55d088c ("NFS: Additional refactoring for fs_context conversion")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8675c69816 ]
When a filesystem is being automounted, it needs to preserve the
user-set superblock mount options, such as the "ro" flag.
Reported-by: Li Lingfeng <lilingfeng3@huawei.com>
Link: https://lore.kernel.org/all/20240604112636.236517-3-lilingfeng@huaweicloud.com/
Fixes: f2aedb713c ("NFS: Add fs_context support.")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e0f8058f2c ]
Fixes a crash when layout is null during this call stack:
write_inode
-> nfs4_write_inode
-> pnfs_layoutcommit_inode
pnfs_set_layoutcommit relies on the lseg refcount to keep the layout
around. Need to clear NFS_INO_LAYOUTCOMMIT otherwise we might attempt
to reference a null layout.
Fixes: fe1cf9469d ("pNFS: Clear all layout segment state in pnfs_mark_layout_stateid_invalid")
Signed-off-by: Jonathan Curley <jcurley@purestorage.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 0f900f1100 ]
Ensure that the verifiers are initialised before calling
d_splice_alias() in _nfs4_open_and_get_state().
Reported-by: Michael Stoler <michael.stoler@vastdata.com>
Fixes: cf5b4059ba ("NFSv4: Fix races between open and dentry revalidation")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 518c32a1bc ]
Ensure that the verifiers are initialised before calling
d_splice_alias() in nfs_atomic_open().
Reported-by: Michael Stoler <michael.stoler@vastdata.com>
Fixes: 809fd143de ("NFSv4: Ensure nfs_atomic_open set the dentry verifier on ENOENT")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9bd545539b ]
Ensure that the verifiers are initialised before calling
d_splice_alias() in both nfs_prime_dcache() and nfs_lookup().
Reported-by: Michael Stoler <michael.stoler@vastdata.com>
Fixes: a1147b8281 ("NFS: Fix up directory verifier races")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 25524b6190 ]
Currently the function responsible for converting between utf16 and
utf8 strings will ignore any characters that cannot be converted. This
however also includes multi-byte characters that do not fit into the
provided string buffer.
This can cause problems if such a multi-byte character is followed by
a single-byte character. In such a case the multi-byte character might
be ignored when the provided string buffer is too small, but the
single-byte character might fit and is thus still copied into the
resulting string.
Fix this by stop filling the provided string buffer once a character
does not fit. In order to be able to do this extend utf32_to_utf8()
to return useful errno codes instead of -1.
Fixes: 74675a5850 ("NLS: update handling of Unicode")
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://patch.msgid.link/20251111131125.3379-2-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit bd4928ec79 ]
If a file removal races with another operation that updates its
attributes, then skip the change to nlink, and just mark the attributes
as being stale.
Reported-by: Aiden Lambert <alambert48@gatech.edu>
Fixes: 59a707b0d4 ("NFS: Ensure we revalidate the inode correctly after remove or rename")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e462fc48ce ]
The current version missed setting one time GC for normal zoned GC
cycle. So, valid threshold control is not working. Need to fix it to
prevent excessive GC for zoned devices.
Fixes: e791d00bd0 ("f2fs: add valid block ratio not to do excessive GC for one time GC")
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f044561331 ]
commit 4eb3117888 changed the cache= option to accept either string
shortcuts or bitfield values. It also changed /proc/mounts to emit the
option as the hexadecimal numeric value rather than the shortcut string.
However, by printing "cache=%x" without the leading 0x, shortcuts such
as "cache=loose" will emit "cache=f" and 'f' is not a string that is
parseable by kstrtoint(), so remounting may fail if a remount with
"cache=f" is attempted.
debug=%x has had the same problem since options have been displayed in
c4fac91004 ("9p: Implement show_options")
Fix these by adding the 0x prefix to the hexadecimal value shown in
/proc/mounts.
Fixes: 4eb3117888 ("fs/9p: Rework cache modes and add new options to Documentation")
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Message-ID: <54b93378-dcf1-4b04-922d-c8b4393da299@redhat.com>
[Dominique: use %#x at Al Viro's suggestion, also handle debug]
Tested-by: Remi Pommarel <repk@triplefau.lt>
Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d70a5804c5 ]
The variable max_ra_count can be 0 in exfat_allocate_bitmap(),
which causes a divide-by-zero error in the subsequent modulo operation
(i % max_ra_count), leading to a system crash.
When max_ra_count is 0, it means that readahead is not used. This patch
load the bitmap without readahead.
Fixes: 9fd688678d ("exfat: optimize allocation bitmap loading time")
Reported-by: Jiaming Zhang <r772577952@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9aee8de970 ]
Fix refcount leaks in `exfat_find` related to `exfat_get_dentry_set`.
Function `exfat_get_dentry_set` would increase the reference counter of
`es->bh` on success. Therefore, `exfat_put_dentry_set` must be called
after `exfat_get_dentry_set` to ensure refcount consistency. This patch
relocate two checks to avoid possible leaks.
Fixes: 82ebecdc74 ("exfat: fix improper check of dentry.stream.valid_size")
Fixes: 13940cef95 ("exfat: add a check for invalid data size")
Signed-off-by: Shuhao Fu <sfual@cse.ust.hk>
Reviewed-by: Yuezhang Mo <Yuezhang.Mo@sony.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit dc10cf1368 ]
sc->first_error might already be set and sc->status
is thus unexpected, so this should avoid the WARN[_ON]_ONCE()
if sc->first_error is already set and have a usable error path.
While there set sc->first_error as soon as possible.
This is done based on a problem seen in similar places on
the server. And there it was already very useful in order
to find the problem when we have a meaningful WARN_ONCE()
that prints details about the connection.
This is much more useful:
[ 309.560973] expected[NEGOTIATE_NEEDED] != RDMA_CONNECT_RUNNING
first_error=0 local=192.168.0.200:445 remote=192.168.0.100:60445
[ 309.561034] WARNING: CPU: 2 PID: 78 at transport_rdma.c:643
recv_done+0x2fa/0x3d0 [ksmbd]
than what we had before (only):
[ 894.140316] WARNING: CPU: 1 PID: 116 at
fs/smb/server/transport_rdma.c:642 recv_done+0x308/0x360 [ksmbd]
Fixes: 58dfba8a2d ("smb: client/smbdirect: replace SMBDIRECT_SOCKET_CONNECTING with more detailed states")
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Long Li <longli@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: Paulo Alcantara <pc@manguebit.org>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1adb2dab97 ]
These will be used in various places in order to assert
the current status mostly during the connect and negotiation
phase. It will replace the WARN_ON_ONCE(sc->status != ...)
calls, which are very useless in order to identify the
problem that happened.
As a start client and server will need to define their own
__SMBDIRECT_SOCKET_DISCONNECT(__sc) macro in order to use
SMBDIRECT_CHECK_STATUS_DISCONNECT().
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Long Li <longli@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: Paulo Alcantara <pc@manguebit.org>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
Stable-dep-of: 425c32750b ("smb: server: relax WARN_ON_ONCE(SMBDIRECT_SOCKET_*) checks in recv_done() and smb_direct_cm_handler()")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1f3fd108c5 ]
This can be used like this:
int err = somefunc();
pr_warn("err=%1pe\n", SMBDIRECT_DEBUG_ERR_PTR(err));
This will be used in the following fixes in order
to be prepared to identify real world problems
more easily.
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Long Li <longli@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: Paulo Alcantara <pc@manguebit.org>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
Stable-dep-of: 425c32750b ("smb: server: relax WARN_ON_ONCE(SMBDIRECT_SOCKET_*) checks in recv_done() and smb_direct_cm_handler()")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d9ee3ff810 ]
When the MB_CHECK_ASSERT macro is enabled, we found that the
current validation logic in __mb_check_buddy has a gap in
detecting certain invalid buddy states, particularly related
to order-0 (bitmap) bits.
The original logic consists of three steps:
1. Validates higher-order buddies: if a higher-order bit is
set, at most one of the two corresponding lower-order bits
may be free; if a higher-order bit is clear, both lower-order
bits must be allocated (and their bitmap bits must be 0).
2. For any set bit in order-0, ensures all corresponding
higher-order bits are not free.
3. Verifies that all preallocated blocks (pa) in the group
have pa_pstart within bounds and their bitmap bits marked as
allocated.
However, this approach fails to properly validate cases where
order-0 bits are incorrectly cleared (0), allowing some invalid
configurations to pass:
corrupt integral
order 3 1 1
order 2 1 1 1 1
order 1 1 1 1 1 1 1 1 1
order 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Here we get two adjacent free blocks at order-0 with inconsistent
higher-order state, and the right one shows the correct scenario.
The root cause is insufficient validation of order-0 zero bits.
To fix this and improve completeness without significant performance
cost, we refine the logic:
1. Maintain the top-down higher-order validation, but we no longer
check the cases where the higher-order bit is 0, as this case will
be covered in step 2.
2. Enhance order-0 checking by examining pairs of bits:
- If either bit in a pair is set (1), all corresponding
higher-order bits must not be free.
- If both bits are clear (0), then exactly one of the
corresponding higher-order bits must be free
3. Keep the preallocation (pa) validation unchanged.
This change closes the validation gap, ensuring illegal buddy states
involving order-0 are correctly detected, while removing redundant
checks and maintaining efficiency.
Fixes: c9de560ded ("ext4: Add multi block allocator for ext4")
Suggested-by: Jan Kara <jack@suse.cz>
Signed-off-by: Yongjian Sun <sunyongjian1@huawei.com>
Reviewed-by: Baokun Li <libaokun1@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Message-ID: <20251106060614.631382-3-sunyongjian@huaweicloud.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7f30e7a423 ]
FMODE_NOCMTIME used to be just a hack for the legacy XFS handle-based
"invisible I/O", but commit e5e9b24ab8 ("nfsd: freeze c/mtime updates
with outstanding WRITE_ATTRS delegation") started using it from
generic callers.
I'm not sure other file systems are actually read for this in general,
so the above commit should get a closer look, but for it to make any
sense, file_update_time needs to respect the flag.
Lift the check from file_modified_flags to file_update_time so that
users of file_update_time inherit the behavior and so that all the
checks are done in one place.
Fixes: e5e9b24ab8 ("nfsd: freeze c/mtime updates with outstanding WRITE_ATTRS delegation")
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://patch.msgid.link/20251120064859.2911749-3-hch@lst.de
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3cd9a42f1b ]
Currently the two high-level APIs use two helper functions to implement
almost all of the logic. Refactor the two helpers and the common logic
into a new file_update_time_flags routine that gets the iocb flags or
0 in case of file_update_time passed so that the entire logic is
contained in a single function and can be easily understood and modified.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://patch.msgid.link/20251120064859.2911749-2-hch@lst.de
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Stable-dep-of: 7f30e7a423 ("fs: lift the FMODE_NOCMTIME check into file_update_time_flags")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2c5f4a5347 ]
Function new_inode() returns a new inode with inode->i_mapping->gfp_mask
set to GFP_HIGHUSER_MOVABLE. This value includes the __GFP_FS flag, so
allocations in that address space can recurse into filesystem memory
reclaim. We don't want that to happen because it can consume a
significant amount of stack memory.
Worse than that is that it can also deadlock: for example, in several
places, gfs2_unstuff_dinode() is called inside filesystem transactions.
This calls filemap_grab_folio(), which can allocate a new folio, which
can trigger memory reclaim. If memory reclaim recurses into the
filesystem and starts another transaction, a deadlock will ensue.
To fix these kinds of problems, prevent memory reclaim from recursing
into filesystem code by making sure that the gfp_mask of inode address
spaces doesn't include __GFP_FS.
The "meta" and resource group address spaces were already using GFP_NOFS
as their gfp_mask (which doesn't include __GFP_FS). The default value
of GFP_HIGHUSER_MOVABLE is less restrictive than GFP_NOFS, though. To
avoid being overly limiting, use the default value and only knock off
the __GFP_FS flag. I'm not sure if this will actually make a
difference, but it also shouldn't hurt.
This patch is loosely based on commit ad22c7a043 ("xfs: prevent stack
overflows from page cache allocation").
Fixes xfstest generic/273.
Fixes: dc0b943523 ("gfs: Don't use GFP_NOFS in gfs2_unstuff_dinode")
Reviewed-by: Andrew Price <anprice@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7fd8720dff ]
Since commit 222f2c7c6d14 ("iomap: always run error completions in user
context"), read error completions are deferred to s_dio_done_wq. This
means the workqueue also needs to be allocated for async reads.
Fixes: 222f2c7c6d14 ("iomap: always run error completions in user context")
Reported-by: syzbot+a2b9a4ed0d61b1efb3f5@syzkaller.appspotmail.com
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://patch.msgid.link/20251124140013.902853-1-hch@lst.de
Tested-by: syzbot+a2b9a4ed0d61b1efb3f5@syzkaller.appspotmail.com
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ddb4873286 ]
At least zonefs expects error completions to be able to sleep. Because
error completions aren't performance critical, just defer them to workqueue
context unconditionally.
Fixes: 8dcc1a9d90 ("fs: New zonefs file system")
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://patch.msgid.link/20251113170633.1453259-3-hch@lst.de
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e7dd1182fc ]
If the call to btrfs_del_leaf() fails we return without decrementing the
extra ref we took on the leaf, therefore leaking it. Fix this by ensuring
we drop the ref count before returning the error.
Fixes: 751a27615d ("btrfs: do not BUG_ON() on tree mod log failures at btrfs_del_ptr()")
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>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d435c51365 ]
Unlike queue_scrub_stripe() which uses the global sctx->extent_path and
sctx->csum_path which are always released at the end of scrub_stripe(),
scrub_raid56_parity_stripe() uses local extent_path and csum_path, as
that function is going to handle the full stripe, whose bytenr may be
smaller than the bytenr in the global sctx paths.
However the cleanup of local extent/csum paths is only happening after
we have successfully submitted an rbio.
There are several error routes that we didn't release those two paths:
- scrub_find_fill_first_stripe() errored out at csum tree search
In that case extent_path is still valid, and that function itself will
not release the extent_path passed in.
And the function returns directly without releasing both paths.
- The full stripe is empty
- Some blocks failed to be recovered
- btrfs_map_block() failed
- raid56_parity_alloc_scrub_rbio() failed
The function returns directly without releasing both paths.
Fix it by covering btrfs_release_path() calls inside the out: tag.
This is just a hot fix, in the long run we will go scoped based auto
freeing for both local paths.
Fixes: 1dc4888e72 ("btrfs: scrub: avoid unnecessary extent tree search preparing stripes")
Fixes: 3c771c1944 ("btrfs: scrub: avoid unnecessary csum tree search preparing stripes")
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 38e818718c ]
From the memory-barriers.txt document regarding memory barrier ordering
guarantees:
(*) These guarantees do not apply to bitfields, because compilers often
generate code to modify these using non-atomic read-modify-write
sequences. Do not attempt to use bitfields to synchronize parallel
algorithms.
(*) Even in cases where bitfields are protected by locks, all fields
in a given bitfield must be protected by one lock. If two fields
in a given bitfield are protected by different locks, the compiler's
non-atomic read-modify-write sequences can cause an update to one
field to corrupt the value of an adjacent field.
btrfs_space_info has a bitfield sharing an underlying word consisting of
the fields full, chunk_alloc, and flush:
struct btrfs_space_info {
struct btrfs_fs_info * fs_info; /* 0 8 */
struct btrfs_space_info * parent; /* 8 8 */
...
int clamp; /* 172 4 */
unsigned int full:1; /* 176: 0 4 */
unsigned int chunk_alloc:1; /* 176: 1 4 */
unsigned int flush:1; /* 176: 2 4 */
...
Therefore, to be safe from parallel read-modify-writes losing a write to
one of the bitfield members protected by a lock, all writes to all the
bitfields must use the lock. They almost universally do, except for
btrfs_clear_space_info_full() which iterates over the space_infos and
writes out found->full = 0 without a lock.
Imagine that we have one thread completing a transaction in which we
finished deleting a block_group and are thus calling
btrfs_clear_space_info_full() while simultaneously the data reclaim
ticket infrastructure is running do_async_reclaim_data_space():
T1 T2
btrfs_commit_transaction
btrfs_clear_space_info_full
data_sinfo->full = 0
READ: full:0, chunk_alloc:0, flush:1
do_async_reclaim_data_space(data_sinfo)
spin_lock(&space_info->lock);
if(list_empty(tickets))
space_info->flush = 0;
READ: full: 0, chunk_alloc:0, flush:1
MOD/WRITE: full: 0, chunk_alloc:0, flush:0
spin_unlock(&space_info->lock);
return;
MOD/WRITE: full:0, chunk_alloc:0, flush:1
and now data_sinfo->flush is 1 but the reclaim worker has exited. This
breaks the invariant that flush is 0 iff there is no work queued or
running. Once this invariant is violated, future allocations that go
into __reserve_bytes() will add tickets to space_info->tickets but will
see space_info->flush is set to 1 and not queue the work. After this,
they will block forever on the resulting ticket, as it is now impossible
to kick the worker again.
I also confirmed by looking at the assembly of the affected kernel that
it is doing RMW operations. For example, to set the flush (3rd) bit to 0,
the assembly is:
andb $0xfb,0x60(%rbx)
and similarly for setting the full (1st) bit to 0:
andb $0xfe,-0x20(%rax)
So I think this is really a bug on practical systems. I have observed
a number of systems in this exact state, but am currently unable to
reproduce it.
Rather than leaving this footgun lying around for the future, take
advantage of the fact that there is room in the struct anyway, and that
it is already quite large and simply change the three bitfield members to
bools. This avoids writes to space_info->full having any effect on
writes to space_info->flush, regardless of locking.
Fixes: 957780eb27 ("Btrfs: introduce ticketed enospc infrastructure")
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Boris Burkov <boris@bur.io>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 725e462988 ]
In the previous code it was possible to incur into a double kfree()
scenario when calling add_delayed_ref_head(). This could happen if the
record was reported to already exist in the
btrfs_qgroup_trace_extent_nolock() call, but then there was an error
later on add_delayed_ref_head(). In this case, since
add_delayed_ref_head() returned an error, the caller went to free the
record. Since add_delayed_ref_head() couldn't set this kfree'd pointer
to NULL, then kfree() would have acted on a non-NULL 'record' object
which was pointing to memory already freed by the callee.
The problem comes from the fact that the responsibility to kfree the
object is on both the caller and the callee at the same time. Hence, the
fix for this is to shift the ownership of the 'qrecord' object out of
the add_delayed_ref_head(). That is, we will never attempt to kfree()
the given object inside of this function, and will expect the caller to
act on the 'qrecord' object on its own. The only exception where the
'qrecord' object cannot be kfree'd is if it was inserted into the
tracing logic, for which we already have the 'qrecord_inserted_ret'
boolean to account for this. Hence, the caller has to kfree the object
only if add_delayed_ref_head() reports not to have inserted it on the
tracing logic.
As a side-effect of the above, we must guarantee that
'qrecord_inserted_ret' is properly initialized at the start of the
function, not at the end, and then set when an actual insert
happens. This way we avoid 'qrecord_inserted_ret' having an invalid
value on an early exit.
The documentation from the add_delayed_ref_head() has also been updated
to reflect on the exact ownership of the 'qrecord' object.
Fixes: 6ef8fbce01 ("btrfs: fix missing error handling when adding delayed ref with qgroups enabled")
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Miquel Sabaté Solà <mssola@mssola.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ebe4f3f6eb ]
The detection of the primary device is skipped incorrectly
if the multiple or flattened feature is enabled.
It also fixes the FSDAX misdetection for non-block extra blobs.
Fixes: c6993c4cb9 ("erofs: Fallback to normal access if DAX is not supported on extra device")
Reported-and-tested-by: syzbot+31b8fb02cb8a25bd5e78@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/r/691af9f6.a70a0220.3124cb.0097.GAE@google.com
Cc: Yuezhang Mo <Yuezhang.Mo@sony.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ccc4e86d1c ]
If a rb node with the same ino already exists in the rb tree, the newly
alloced mft_inode in ni_add_subrecord() will not have its memory cleaned
up, which leads to the memory leak issue reported by syzbot.
The best option to avoid this issue is to put the newly alloced mft node
when a rb node with the same ino already exists in the rb tree and return
the rb node found in the rb tree to the parent layer.
syzbot reported:
BUG: memory leak
unreferenced object 0xffff888110bef280 (size 128):
backtrace (crc 126a088f):
ni_add_subrecord+0x31/0x180 fs/ntfs3/frecord.c:317
ntfs_look_free_mft+0xf0/0x790 fs/ntfs3/fsntfs.c:715
BUG: memory leak
unreferenced object 0xffff888109093400 (size 1024):
backtrace (crc 7197c55e):
mi_init+0x2b/0x50 fs/ntfs3/record.c:105
mi_format_new+0x40/0x220 fs/ntfs3/record.c:422
Fixes: 4342306f0f ("fs/ntfs3: Add file operations and implementation")
Reported-by: syzbot+3932ccb896e06f7414c9@syzkaller.appspotmail.com
Signed-off-by: Edward Adam Davis <eadavis@qq.com>
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4d78d1173a ]
After ntfs_look_free_mft() executes successfully, all subsequent code
that fails to execute must put mi.
Fixes: 4342306f0f ("fs/ntfs3: Add file operations and implementation")
Signed-off-by: Edward Adam Davis <eadavis@qq.com>
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3524b021b0 ]
The extent returned by the file system may have a smaller offset than
the segment offset requested by the client. In this case, the minimum
segment length must be checked against the requested range. Otherwise,
the client may not be able to continue the read/write operation.
Fixes: 8650b8a058 ("nfsd: pNFS block layout driver")
Signed-off-by: Sergey Bashirov <sergeybashirov@gmail.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 798a401660 ]
If we have LOCKDOWN_TRACEFS, the function bails out - *after*
having locked the parent directory and without bothering to
undo that. Just check it before tracefs_start_creating()...
Fixes: e24709454c "tracefs/eventfs: Add missing lockdown checks"
Acked-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit c460192aae ]
fuse_ctl_remove_conn() used to decrement the link count of root
manually; that got subsumed by simple_recursive_removal(), but
in case when subdirectory creation has failed the latter won't
get called.
Just move the modification of parent's link count into
fuse_ctl_add_dentry() to keep the things simple. Allows to
get rid of the nlink argument as well...
Fixes: fcaac5b427 "fuse_ctl: use simple_recursive_removal()"
Acked-by: Miklos Szeredi <mszeredi@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit c9dff86eb7 ]
Fields in ocfs2_dinode is little endian, covert to host endian when
checking those contents.
Link: https://lkml.kernel.org/r/20251025123218.3997866-1-joseph.qi@linux.alibaba.com
Fixes: fdbb6cd96e ("ocfs2: correct l_next_free_rec in online check")
Signed-off-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Reviewed-by: Heming Zhao <heming.zhao@suse.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Jun Piao <piaojun@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8a7d58845f ]
In '__ocfs2_move_extent()', relax 'BUG()' to 'ocfs2_error()' just
to avoid crashing the whole kernel due to a filesystem corruption.
Fixes: 8f603e567a ("Ocfs2/move_extents: move a range of extent.")
Link: https://lkml.kernel.org/r/20251009102349.181126-2-dmantipov@yandex.ru
Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru>
Closes: https://syzkaller.appspot.com/bug?extid=727d161855d11d81e411
Reported-by: syzbot+727d161855d11d81e411@syzkaller.appspotmail.com
Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Jun Piao <piaojun@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit be99c62ac7 ]
After setting the inode mode of $Extend to a regular file, executing the
truncate system call will enter the do_truncate() routine, causing the
run_lock uninitialized error reported by syzbot.
Prior to patch 4e8011ffec, if the inode mode of $Extend was not set to
a regular file, the do_truncate() routine would not be entered.
Add the run_lock initialization when loading $Extend.
syzbot reported:
INFO: trying to register non-static key.
Call Trace:
dump_stack_lvl+0x189/0x250 lib/dump_stack.c:120
assign_lock_key+0x133/0x150 kernel/locking/lockdep.c:984
register_lock_class+0x105/0x320 kernel/locking/lockdep.c:1299
__lock_acquire+0x99/0xd20 kernel/locking/lockdep.c:5112
lock_acquire+0x120/0x360 kernel/locking/lockdep.c:5868
down_write+0x96/0x1f0 kernel/locking/rwsem.c:1590
ntfs_set_size+0x140/0x200 fs/ntfs3/inode.c:860
ntfs_extend+0x1d9/0x970 fs/ntfs3/file.c:387
ntfs_setattr+0x2e8/0xbe0 fs/ntfs3/file.c:808
Fixes: 4e8011ffec ("ntfs3: pretend $Extend records as regular files")
Reported-by: syzbot+bdeb22a4b9a09ab9aa45@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=bdeb22a4b9a09ab9aa45
Tested-by: syzbot+bdeb22a4b9a09ab9aa45@syzkaller.appspotmail.com
Signed-off-by: Edward Adam Davis <eadavis@qq.com>
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit c105e76bb1 ]
This code calls hfs_bnode_put(node) which drops the refcount and then
dreferences "node" on the next line. It's only safe to use "node"
when we're holding a reference so flip these two lines around.
Fixes: a06ec283e1 ("hfs: add logic of correcting a next unused CNID")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Signed-off-by: Viacheslav Dubeyko <slava@dubeyko.com>
Link: https://lore.kernel.org/r/aN-Xw8KnbSnuIcLk@stanley.mountain
Signed-off-by: Viacheslav Dubeyko <slava@dubeyko.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit a2e5a3cea4 ]
The move extent operation should return -EOPNOTSUPP if any of the inodes
is a quota inode, rather than requiring both to be quota inodes.
Fixes: 02749a4c20 ("ext4: add ext4_is_quota_file()")
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Message-ID: <20251013015128.499308-2-yi.zhang@huaweicloud.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit a63dd8fd13 ]
When page cache is used, writebacks are done on a page granularity, and it
is expected that the underlying filesystem (such as v9fs) should respect
the write position. However, currently v9fs will passthrough O_APPEND to
the server even on cached mode. This causes data corruption if a sync or
fstat gets between two writes to the same file.
This patch removes the APPEND flag from the open request we send to the
server when writeback caching is involved. I believe keeping server-side
APPEND is probably fine for uncached mode (even if two fds are opened, one
without O_APPEND and one with it, this should still be fine since they
would use separate fid for the writes).
Signed-off-by: Tingmao Wang <m@maowtm.org>
Fixes: 4eb3117888 ("fs/9p: Rework cache modes and add new options to Documentation")
Message-ID: <20251102235631.8724-1-m@maowtm.org>
Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
