| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
The PG_error bit is now unused; delete it and free up a bit in
page->flags.
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Matthew Wilcox (Oracle) <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
|
|
We removed hugetlb_follow_page_mask() in commit 9cb28da54643 ("mm/gup:
handle hugetlb in the generic follow_page_mask code") but forgot to
cleanup some leftovers.
While at it, simplify the hugetlb comment, it's overly detailed and rather
confusing. Stating that we may end up in there during coredumping is
sufficient to explain the PF_DUMPCORE usage.
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: David Hildenbrand <[email protected]>
Reviewed-by: Peter Xu <[email protected]>
Cc: Muchun Song <[email protected]>
Cc: Alexander Viro <[email protected]>
Cc: Christian Brauner <[email protected]>
Cc: Jan Kara <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
|
|
The vma_shrink() and vma_expand() functions are internal VMA manipulation
functions which we ought to abstract for use outside of memory management
code.
To achieve this, we replace shift_arg_pages() in fs/exec.c with an
invocation of a new relocate_vma_down() function implemented in mm/mmap.c,
which enables us to also move move_page_tables() and vma_iter_prev_range()
to internal.h.
The purpose of doing this is to isolate key VMA manipulation functions in
order that we can both abstract them and later render them easily
testable.
Link: https://lkml.kernel.org/r/3cfcd9ec433e032a85f636fdc0d7d98fafbd19c5.1722251717.git.lorenzo.stoakes@oracle.com
Signed-off-by: Lorenzo Stoakes <[email protected]>
Reviewed-by: Vlastimil Babka <[email protected]>
Reviewed-by: Liam R. Howlett <[email protected]>
Cc: Alexander Viro <[email protected]>
Cc: Brendan Higgins <[email protected]>
Cc: Christian Brauner <[email protected]>
Cc: David Gow <[email protected]>
Cc: Eric W. Biederman <[email protected]>
Cc: Jan Kara <[email protected]>
Cc: Kees Cook <[email protected]>
Cc: Matthew Wilcox (Oracle) <[email protected]>
Cc: Rae Moar <[email protected]>
Cc: SeongJae Park <[email protected]>
Cc: Shuah Khan <[email protected]>
Cc: Suren Baghdasaryan <[email protected]>
Cc: Pengfei Xu <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
|
|
Patch series "Make core VMA operations internal and testable", v4.
There are a number of "core" VMA manipulation functions implemented in
mm/mmap.c, notably those concerning VMA merging, splitting, modifying,
expanding and shrinking, which logically don't belong there.
More importantly this functionality represents an internal implementation
detail of memory management and should not be exposed outside of mm/
itself.
This patch series isolates core VMA manipulation functionality into its
own file, mm/vma.c, and provides an API to the rest of the mm code in
mm/vma.h.
Importantly, it also carefully implements mm/vma_internal.h, which
specifies which headers need to be imported by vma.c, leading to the very
useful property that vma.c depends only on mm/vma.h and mm/vma_internal.h.
This means we can then re-implement vma_internal.h in userland, adding
shims for kernel mechanisms as required, allowing us to unit test internal
VMA functionality.
This testing is useful as opposed to an e.g. kunit implementation as this
way we can avoid all external kernel side-effects while testing, run tests
VERY quickly, and iterate on and debug problems quickly.
Excitingly this opens the door to, in the future, recreating precise
problems observed in production in userland and very quickly debugging
problems that might otherwise be very difficult to reproduce.
This patch series takes advantage of existing shim logic and full userland
maple tree support contained in tools/testing/radix-tree/ and
tools/include/linux/, separating out shared components of the radix tree
implementation to provide this testing.
Kernel functionality is stubbed and shimmed as needed in
tools/testing/vma/ which contains a fully functional userland
vma_internal.h file and which imports mm/vma.c and mm/vma.h to be directly
tested from userland.
A simple, skeleton testing implementation is provided in
tools/testing/vma/vma.c as a proof-of-concept, asserting that simple VMA
merge, modify (testing split), expand and shrink functionality work
correctly.
This patch (of 4):
This patch forms part of a patch series intending to separate out VMA
logic and render it testable from userspace, which requires that core
manipulation functions be exposed in an mm/-internal header file.
In order to do this, we must abstract APIs we wish to test, in this
instance functions which ultimately invoke vma_modify().
This patch therefore moves all logic which ultimately invokes vma_modify()
to mm/userfaultfd.c, trying to transfer code at a functional granularity
where possible.
[[email protected]: fix user-after-free in userfaultfd_clear_vma()]
Link: https://lkml.kernel.org/r/[email protected]
Link: https://lkml.kernel.org/r/[email protected]
Link: https://lkml.kernel.org/r/50c3ed995fd81c45876c86304c8a00bf3e396cfd.1722251717.git.lorenzo.stoakes@oracle.com
Signed-off-by: Lorenzo Stoakes <[email protected]>
Reviewed-by: Vlastimil Babka <[email protected]>
Reviewed-by: Liam R. Howlett <[email protected]>
Cc: Alexander Viro <[email protected]>
Cc: Brendan Higgins <[email protected]>
Cc: Christian Brauner <[email protected]>
Cc: David Gow <[email protected]>
Cc: Eric W. Biederman <[email protected]>
Cc: Jan Kara <[email protected]>
Cc: Kees Cook <[email protected]>
Cc: Matthew Wilcox (Oracle) <[email protected]>
Cc: Rae Moar <[email protected]>
Cc: SeongJae Park <[email protected]>
Cc: Shuah Khan <[email protected]>
Cc: Suren Baghdasaryan <[email protected]>
Cc: Pengfei Xu <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
|
|
Sharing page tables between processes but falling back to per-MM page
table locks cannot possibly work.
So, let's make sure that we do have split PMD locks by adding a new
Kconfig option and letting that depend on CONFIG_SPLIT_PMD_PTLOCKS.
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: David Hildenbrand <[email protected]>
Acked-by: Mike Rapoport (Microsoft) <[email protected]>
Cc: Alexander Viro <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: Boris Ostrovsky <[email protected]>
Cc: Christian Brauner <[email protected]>
Cc: Christophe Leroy <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: "H. Peter Anvin" <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Juergen Gross <[email protected]>
Cc: Michael Ellerman <[email protected]>
Cc: Muchun Song <[email protected]>
Cc: "Naveen N. Rao" <[email protected]>
Cc: Nicholas Piggin <[email protected]>
Cc: Oscar Salvador <[email protected]>
Cc: Peter Xu <[email protected]>
Cc: Russell King <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
|
|
Besides the obvious (and desired) difference between krealloc() and
kvrealloc(), there is some inconsistency in their function signatures and
behavior:
- krealloc() frees the memory when the requested size is zero, whereas
kvrealloc() simply returns a pointer to the existing allocation.
- krealloc() behaves like kmalloc() if a NULL pointer is passed, whereas
kvrealloc() does not accept a NULL pointer at all and, if passed,
would fault instead.
- krealloc() is self-contained, whereas kvrealloc() relies on the caller
to provide the size of the previous allocation.
Inconsistent behavior throughout allocation APIs is error prone, hence
make kvrealloc() behave like krealloc(), which seems superior in all
mentioned aspects.
Besides that, implementing kvrealloc() by making use of krealloc() and
vrealloc() provides oppertunities to grow (and shrink) allocations more
efficiently. For instance, vrealloc() can be optimized to allocate and
map additional pages to grow the allocation or unmap and free unused pages
to shrink the allocation.
[[email protected]: document concurrency restrictions]
Link: https://lkml.kernel.org/r/[email protected]
[[email protected]: disable KASAN when switching to vmalloc]
Link: https://lkml.kernel.org/r/[email protected]
[[email protected]: properly document __GFP_ZERO behavior]
Link: https://lkml.kernel.org/r/[email protected]
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Danilo Krummrich <[email protected]>
Acked-by: Michal Hocko <[email protected]>
Acked-by: Vlastimil Babka <[email protected]>
Cc: Chandan Babu R <[email protected]>
Cc: Christian König <[email protected]>
Cc: Christoph Hellwig <[email protected]>
Cc: Christoph Lameter <[email protected]>
Cc: David Rientjes <[email protected]>
Cc: Hyeonggon Yoo <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Cc: Kees Cook <[email protected]>
Cc: Marc Zyngier <[email protected]>
Cc: Michael Ellerman <[email protected]>
Cc: Miguel Ojeda <[email protected]>
Cc: Oliver Upton <[email protected]>
Cc: Pekka Enberg <[email protected]>
Cc: Roman Gushchin <[email protected]>
Cc: Uladzislau Rezki <[email protected]>
Cc: Wedson Almeida Filho <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
|
|
Pull smb client fixes from Steve French:
- copy_file_range fix
- two read fixes including read past end of file rc fix and read retry
crediting fix
- falloc zero range fix
* tag 'v6.11-rc5-smb-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
cifs: Fix FALLOC_FL_ZERO_RANGE to preflush buffered part of target region
cifs: Fix copy offload to flush destination region
netfs, cifs: Fix handling of short DIO read
cifs: Fix lack of credit renegotiation on read retry
|
|
Push bcachefs fixes from Kent Overstreet:
"The data corruption in the buffered write path is troubling; inode
lock should not have been able to cause that...
- Fix a rare data corruption in the rebalance path, caught as a nonce
inconsistency on encrypted filesystems
- Revert lockless buffered write path
- Mark more errors as autofix"
* tag 'bcachefs-2024-08-21' of https://github.com/koverstreet/bcachefs:
bcachefs: Mark more errors as autofix
bcachefs: Revert lockless buffered IO path
bcachefs: Fix bch2_extents_match() false positive
bcachefs: Fix failure to return error in data_update_index_update()
|
|
errors that are known to always be safe to fix should be autofix: this
should be most errors even at this point, but that will need some
thorough review.
note that errors are still logged in the superblock, so we'll still know
that they happened.
Signed-off-by: Kent Overstreet <[email protected]>
|
|
We had a report of data corruption on nixos when building installer
images.
https://github.com/NixOS/nixpkgs/pull/321055#issuecomment-2184131334
It seems that writes are being dropped, but only when issued by QEMU,
and possibly only in snapshot mode. It's undetermined if it's write
calls are being dropped or dirty folios.
Further testing, via minimizing the original patch to just the change
that skips the inode lock on non appends/truncates, reveals that it
really is just not taking the inode lock that causes the corruption: it
has nothing to do with the other logic changes for preserving write
atomicity in corner cases.
It's also kernel config dependent: it doesn't reproduce with the minimal
kernel config that ktest uses, but it does reproduce with nixos's distro
config. Bisection the kernel config initially pointer the finger at page
migration or compaction, but it appears that was erroneous; we haven't
yet determined what kernel config option actually triggers it.
Sadly it appears this will have to be reverted since we're getting too
close to release and my plate is full, but we'd _really_ like to fully
debug it.
My suspicion is that this patch is exposing a preexisting bug - the
inode lock actually covers very little in IO paths, and we have a
different lock (the pagecache add lock) that guards against races with
truncate here.
Fixes: 7e64c86cdc6c ("bcachefs: Buffered write path now can avoid the inode lock")
Signed-off-by: Kent Overstreet <[email protected]>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux
Pull nfsd fix from Chuck Lever:
- One more write delegation fix
* tag 'nfsd-6.11-3' of git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux:
nfsd: fix nfsd4_deleg_getattr_conflict in presence of third party lease
|
|
Pull xfs fixes from Chandan Babu:
- Do not call out v1 inodes with non-zero di_nlink field as being
corrupt
- Change xfs_finobt_count_blocks() to count "free inode btree" blocks
rather than "inode btree" blocks
- Don't report the number of trimmed bytes via FITRIM because the
underlying storage isn't required to do anything and failed discard
IOs aren't reported to the caller anyway
- Fix incorrect setting of rm_owner field in an rmap query
- Report missing disk offset range in an fsmap query
- Obtain m_growlock when extending realtime section of the filesystem
- Reset rootdir extent size hint after extending realtime section of
the filesystem
* tag 'xfs-6.11-fixes-4' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
xfs: reset rootdir extent size hint after growfsrt
xfs: take m_growlock when running growfsrt
xfs: Fix missing interval for missing_owner in xfs fsmap
xfs: use XFS_BUF_DADDR_NULL for daddrs in getfsmap code
xfs: Fix the owner setting issue for rmap query in xfs fsmap
xfs: don't bother reporting blocks trimmed via FITRIM
xfs: xfs_finobt_count_blocks() walks the wrong btree
xfs: fix folio dirtying for XFILE_ALLOC callers
xfs: fix di_onlink checking for V1/V2 inodes
|
|
It is not safe to dereference fl->c.flc_owner without first confirming
fl->fl_lmops is the expected manager. nfsd4_deleg_getattr_conflict()
tests fl_lmops but largely ignores the result and assumes that flc_owner
is an nfs4_delegation anyway. This is wrong.
With this patch we restore the "!= &nfsd_lease_mng_ops" case to behave
as it did before the change mentioned below. This is the same as the
current code, but without any reference to a possible delegation.
Fixes: c5967721e106 ("NFSD: handle GETATTR conflict with write delegation")
Signed-off-by: NeilBrown <[email protected]>
Reviewed-by: Jeff Layton <[email protected]>
Signed-off-by: Chuck Lever <[email protected]>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux
Pull execve fix from Kees Cook:
- binfmt_elf_fdpic: fix AUXV size with ELF_HWCAP2 (Max Filippov)
* tag 'execve-v6.11-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
binfmt_elf_fdpic: fix AUXV size calculation when ELF_HWCAP2 is defined
|
|
The runtime constant feature removes all the users of these variables,
allowing the compiler to optimize them away. It's quite difficult to
extract their values from the kernel text, and the memory saved by
removing them is tiny, and it was never the point of this optimization.
Since the dentry_hashtable is a core data structure, it's valuable for
debugging tools to be able to read it easily. For instance, scripts
built on drgn, like the dentrycache script[1], rely on it to be able to
perform diagnostics on the contents of the dcache. Annotate it as used,
so the compiler doesn't discard it.
Link: https://github.com/oracle-samples/drgn-tools/blob/3afc56146f54d09dfd1f6d3c1b7436eda7e638be/drgn_tools/dentry.py#L325-L355 [1]
Fixes: e3c92e81711d ("runtime constants: add x86 architecture support")
Signed-off-by: Stephen Brennan <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Under certain conditions, the range to be cleared by FALLOC_FL_ZERO_RANGE
may only be buffered locally and not yet have been flushed to the server.
For example:
xfs_io -f -t -c "pwrite -S 0x41 0 4k" \
-c "pwrite -S 0x42 4k 4k" \
-c "fzero 0 4k" \
-c "pread -v 0 8k" /xfstest.test/foo
will write two 4KiB blocks of data, which get buffered in the pagecache,
and then fallocate() is used to clear the first 4KiB block on the server -
but we don't flush the data first, which means the EOF position on the
server is wrong, and so the FSCTL_SET_ZERO_DATA RPC fails (and xfs_io
ignores the error), but then when we try to read it, we see the old data.
Fix this by preflushing any part of the target region that above the
server's idea of the EOF position to force the server to update its EOF
position.
Note, however, that we don't want to simply expand the file by moving the
EOF before doing the FSCTL_SET_ZERO_DATA[*] because someone else might see
the zeroed region or if the RPC fails we then have to try to clean it up or
risk getting corruption.
[*] And we have to move the EOF first otherwise FSCTL_SET_ZERO_DATA won't
do what we want.
This fixes the generic/008 xfstest.
[!] Note: A better way to do this might be to split the operation into two
parts: we only do FSCTL_SET_ZERO_DATA for the part of the range below the
server's EOF and then, if that worked, invalidate the buffered pages for the
part above the range.
Fixes: 6b69040247e1 ("cifs/smb3: Fix data inconsistent when zero file range")
Signed-off-by: David Howells <[email protected]>
cc: Steve French <[email protected]>
cc: Zhang Xiaoxu <[email protected]>
cc: Pavel Shilovsky <[email protected]>
cc: Paulo Alcantara <[email protected]>
cc: Shyam Prasad N <[email protected]>
cc: Rohith Surabattula <[email protected]>
cc: Jeff Layton <[email protected]>
cc: [email protected]
cc: [email protected]
Signed-off-by: Steve French <[email protected]>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux
Pull nfsd fixes from Chuck Lever:
- Fix a number of crashers
- Update email address for an NFSD reviewer
* tag 'nfsd-6.11-2' of git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux:
fs/nfsd: fix update of inode attrs in CB_GETATTR
nfsd: fix potential UAF in nfsd4_cb_getattr_release
nfsd: hold reference to delegation when updating it for cb_getattr
MAINTAINERS: Update Olga Kornievskaia's email address
nfsd: prevent panic for nfsv4.0 closed files in nfs4_show_open
nfsd: ensure that nfsd4_fattr_args.context is zeroed out
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux
Pull btrfs fixes from David Sterba:
- fix use-after-free when submitting bios for read, after an error and
partially submitted bio the original one is freed while it can be
still be accessed again
- fix fstests case btrfs/301, with enabled quotas wait for delayed
iputs when flushing delalloc
- fix periodic block group reclaim, an unitialized value can be
returned if there are no block groups to reclaim
- fix build warning (-Wmaybe-uninitialized)
* tag 'for-6.11-rc5-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
btrfs: fix uninitialized return value from btrfs_reclaim_sweep()
btrfs: fix a use-after-free when hitting errors inside btrfs_submit_chunk()
btrfs: initialize last_extent_end to fix -Wmaybe-uninitialized warning in extent_fiemap()
btrfs: run delayed iputs when flushing delalloc
|
|
Fix cifs_file_copychunk_range() to flush the destination region before
invalidating it to avoid potential loss of data should the copy fail, in
whole or in part, in some way.
Fixes: 7b2404a886f8 ("cifs: Fix flushing, invalidation and file size with copy_file_range()")
Signed-off-by: David Howells <[email protected]>
cc: Steve French <[email protected]>
cc: Paulo Alcantara <[email protected]>
cc: Shyam Prasad N <[email protected]>
cc: Rohith Surabattula <[email protected]>
cc: Matthew Wilcox <[email protected]>
cc: Jeff Layton <[email protected]>
cc: [email protected]
cc: [email protected]
cc: [email protected]
Signed-off-by: Steve French <[email protected]>
|
|
Short DIO reads, particularly in relation to cifs, are not being handled
correctly by cifs and netfslib. This can be tested by doing a DIO read of
a file where the size of read is larger than the size of the file. When it
crosses the EOF, it gets a short read and this gets retried, and in the
case of cifs, the retry read fails, with the failure being translated to
ENODATA.
Fix this by the following means:
(1) Add a flag, NETFS_SREQ_HIT_EOF, for the filesystem to set when it
detects that the read did hit the EOF.
(2) Make the netfslib read assessment stop processing subrequests when it
encounters one with that flag set.
(3) Return rreq->transferred, the accumulated contiguous amount read to
that point, to userspace for a DIO read.
(4) Make cifs set the flag and clear the error if the read RPC returned
ENODATA.
(5) Make cifs set the flag and clear the error if a short read occurred
without error and the read-to file position is now at the remote inode
size.
Fixes: 69c3c023af25 ("cifs: Implement netfslib hooks")
Signed-off-by: David Howells <[email protected]>
cc: Steve French <[email protected]>
cc: Paulo Alcantara <[email protected]>
cc: Jeff Layton <[email protected]>
cc: [email protected]
cc: [email protected]
cc: [email protected]
Signed-off-by: Steve French <[email protected]>
|
|
When netfslib asks cifs to issue a read operation, it prefaces this with a
call to ->clamp_length() which cifs uses to negotiate credits, providing
receive capacity on the server; however, in the event that a read op needs
reissuing, netfslib doesn't call ->clamp_length() again as that could
shorten the subrequest, leaving a gap.
This causes the retried read to be done with zero credits which causes the
server to reject it with STATUS_INVALID_PARAMETER. This is a problem for a
DIO read that is requested that would go over the EOF. The short read will
be retried, causing EINVAL to be returned to the user when it fails.
Fix this by making cifs_req_issue_read() negotiate new credits if retrying
(NETFS_SREQ_RETRYING now gets set in the read side as well as the write
side in this instance).
This isn't sufficient, however: the new credits might not be sufficient to
complete the remainder of the read, so also add an additional field,
rreq->actual_len, that holds the actual size of the op we want to perform
without having to alter subreq->len.
We then rely on repeated short reads being retried until we finish the read
or reach the end of file and make a zero-length read.
Also fix a couple of places where the subrequest start and length need to
be altered by the amount so far transferred when being used.
Fixes: 69c3c023af25 ("cifs: Implement netfslib hooks")
Signed-off-by: David Howells <[email protected]>
cc: Steve French <[email protected]>
cc: Paulo Alcantara <[email protected]>
cc: Jeff Layton <[email protected]>
cc: [email protected]
cc: [email protected]
cc: [email protected]
Signed-off-by: Steve French <[email protected]>
|
|
Pull smb client fixes from Steve French:
- two RDMA/smbdirect fixes and a minor cleanup
- punch hole fix
* tag 'v6.11-rc5-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
cifs: Fix FALLOC_FL_PUNCH_HOLE support
smb/client: fix rdma usage in smb2_async_writev()
smb/client: remove unused rq_iter_size from struct smb_rqst
smb/client: avoid dereferencing rdata=NULL in smb2_new_read_req()
|
|
The return variable 'ret' at btrfs_reclaim_sweep() is never assigned if
none of the space infos is reclaimable (for example if periodic reclaim
is disabled, which is the default), so we return an undefined value.
This can be fixed my making btrfs_reclaim_sweep() not return any value
as well as do_reclaim_sweep() because:
1) do_reclaim_sweep() always returns 0, so we can make it return void;
2) The only caller of btrfs_reclaim_sweep() (btrfs_reclaim_bgs()) doesn't
care about its return value, and in its context there's nothing to do
about any errors anyway.
Therefore remove the return value from btrfs_reclaim_sweep() and
do_reclaim_sweep().
Fixes: e4ca3932ae90 ("btrfs: periodic block_group reclaim")
Reviewed-by: Josef Bacik <[email protected]>
Signed-off-by: Filipe Manana <[email protected]>
Reviewed-by: David Sterba <[email protected]>
Signed-off-by: David Sterba <[email protected]>
|
|
If growfsrt is run on a filesystem that doesn't have a rt volume, it's
possible to change the rt extent size. If the root directory was
previously set up with an inherited extent size hint and rtinherit, it's
possible that the hint is no longer a multiple of the rt extent size.
Although the verifiers don't complain about this, xfs_repair will, so if
we detect this situation, log the root directory to clean it up. This
is still racy, but it's better than nothing.
Reviewed-by: Christoph Hellwig <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
Signed-off-by: Chandan Babu R <[email protected]>
|
|
Take the grow lock when we're expanding the realtime volume, like we do
for the other growfs calls.
Reviewed-by: Christoph Hellwig <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
Signed-off-by: Chandan Babu R <[email protected]>
|
|
In the fsmap query of xfs, there is an interval missing problem:
[root@fedora ~]# xfs_io -c 'fsmap -vvvv' /mnt
EXT: DEV BLOCK-RANGE OWNER FILE-OFFSET AG AG-OFFSET TOTAL
0: 253:16 [0..7]: static fs metadata 0 (0..7) 8
1: 253:16 [8..23]: per-AG metadata 0 (8..23) 16
2: 253:16 [24..39]: inode btree 0 (24..39) 16
3: 253:16 [40..47]: per-AG metadata 0 (40..47) 8
4: 253:16 [48..55]: refcount btree 0 (48..55) 8
5: 253:16 [56..103]: per-AG metadata 0 (56..103) 48
6: 253:16 [104..127]: free space 0 (104..127) 24
......
BUG:
[root@fedora ~]# xfs_io -c 'fsmap -vvvv -d 104 107' /mnt
[root@fedora ~]#
Normally, we should be able to get [104, 107), but we got nothing.
The problem is caused by shifting. The query for the problem-triggered
scenario is for the missing_owner interval (e.g. freespace in rmapbt/
unknown space in bnobt), which is obtained by subtraction (gap). For this
scenario, the interval is obtained by info->last. However, rec_daddr is
calculated based on the start_block recorded in key[1], which is converted
by calling XFS_BB_TO_FSBT. Then if rec_daddr does not exceed
info->next_daddr, which means keys[1].fmr_physical >> (mp)->m_blkbb_log
<= info->next_daddr, no records will be displayed. In the above example,
104 >> (mp)->m_blkbb_log = 12 and 107 >> (mp)->m_blkbb_log = 12, so the two
are reduced to 0 and the gap is ignored:
before calculate ----------------> after shifting
104(st) 107(ed) 12(st/ed)
|---------| |
sector size block size
Resolve this issue by introducing the "end_daddr" field in
xfs_getfsmap_info. This records |key[1].fmr_physical + key[1].length| at
the granularity of sector. If the current query is the last, the rec_daddr
is end_daddr to prevent missing interval problems caused by shifting. We
only need to focus on the last query, because xfs disks are internally
aligned with disk blocksize that are powers of two and minimum 512, so
there is no problem with shifting in previous queries.
After applying this patch, the above problem have been solved:
[root@fedora ~]# xfs_io -c 'fsmap -vvvv -d 104 107' /mnt
EXT: DEV BLOCK-RANGE OWNER FILE-OFFSET AG AG-OFFSET TOTAL
0: 253:16 [104..106]: free space 0 (104..106) 3
Fixes: e89c041338ed ("xfs: implement the GETFSMAP ioctl")
Signed-off-by: Zizhi Wo <[email protected]>
Reviewed-by: Darrick J. Wong <[email protected]>
[djwong: limit the range of end_addr correctly]
Signed-off-by: Darrick J. Wong <[email protected]>
Signed-off-by: Chandan Babu R <[email protected]>
|
|
Use XFS_BUF_DADDR_NULL (instead of a magic sentinel value) to mean "this
field is null" like the rest of xfs.
Cc: [email protected]
Fixes: e89c041338ed6 ("xfs: implement the GETFSMAP ioctl")
Reviewed-by: Christoph Hellwig <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
Signed-off-by: Chandan Babu R <[email protected]>
|
|
gitolite.kernel.org:pub/scm/linux/kernel/git/vfs/vfs
Pull vfs fixes from Christian Brauner:
"VFS:
- Ensure that backing files uses file->f_ops->splice_write() for
splice
netfs:
- Revert the removal of PG_private_2 from netfs_release_folio() as
cephfs still relies on this
- When AS_RELEASE_ALWAYS is set on a mapping the folio needs to
always be invalidated during truncation
- Fix losing untruncated data in a folio by making letting
netfs_release_folio() return false if the folio is dirty
- Fix trimming of streaming-write folios in netfs_inval_folio()
- Reset iterator before retrying a short read
- Fix interaction of streaming writes with zero-point tracker
afs:
- During truncation afs currently calls truncate_setsize() which sets
i_size, expands the pagecache and truncates it. The first two
operations aren't needed because they will have already been done.
So call truncate_pagecache() instead and skip the redundant parts
overlayfs:
- Fix checking of the number of allowed lower layers so 500 layers
can actually be used instead of just 499
- Add missing '\n' to pr_err() output
- Pass string to ovl_parse_layer() and thus allow it to be used for
Opt_lowerdir as well
pidfd:
- Revert blocking the creation of pidfds for kthread as apparently
userspace relies on this. Specifically, it breaks systemd during
shutdown
romfs:
- Fix romfs_read_folio() to use the correct offset with
folio_zero_tail()"
* tag 'vfs-6.11-rc6.fixes' of gitolite.kernel.org:pub/scm/linux/kernel/git/vfs/vfs:
netfs: Fix interaction of streaming writes with zero-point tracker
netfs: Fix missing iterator reset on retry of short read
netfs: Fix trimming of streaming-write folios in netfs_inval_folio()
netfs: Fix netfs_release_folio() to say no if folio dirty
afs: Fix post-setattr file edit to do truncation correctly
mm: Fix missing folio invalidation calls during truncation
ovl: ovl_parse_param_lowerdir: Add missed '\n' for pr_err
ovl: fix wrong lowerdir number check for parameter Opt_lowerdir
ovl: pass string to ovl_parse_layer()
backing-file: convert to using fops->splice_write
Revert "pidfd: prevent creation of pidfds for kthreads"
romfs: fix romfs_read_folio()
netfs, ceph: Partially revert "netfs: Replace PG_fscache by setting folio->private and marking dirty"
|
|
This was caught as a very rare nonce inconsistency, on systems with
encryption and replication (and tiering, or some form of rebalance
operation running):
[Wed Jul 17 13:30:03 2024] about to insert invalid key in data update path
[Wed Jul 17 13:30:03 2024] old: u64s 10 type extent 671283510:6392:U32_MAX len 16 ver 106595503: durability: 2 crc: c_size 8 size 16 offset 0 nonce 0 csum chacha20_poly1305_80 compress zstd ptr: 3:355968:104 gen 7 ptr: 4:513244:48 gen 6 rebalance: target hdd compression zstd
[Wed Jul 17 13:30:03 2024] k: u64s 10 type extent 671283510:6400:U32_MAX len 16 ver 106595508: durability: 2 crc: c_size 8 size 16 offset 0 nonce 0 csum chacha20_poly1305_80 compress zstd ptr: 3:355968:112 gen 7 ptr: 4:513244:56 gen 6 rebalance: target hdd compression zstd
[Wed Jul 17 13:30:03 2024] new: u64s 14 type extent 671283510:6392:U32_MAX len 8 ver 106595508: durability: 2 crc: c_size 8 size 16 offset 0 nonce 0 csum chacha20_poly1305_80 compress zstd ptr: 3:355968:112 gen 7 cached ptr: 4:513244:56 gen 6 cached rebalance: target hdd compression zstd crc: c_size 8 size 16 offset 8 nonce 0 csum chacha20_poly1305_80 compress zstd ptr: 1:10860085:32 gen 0 ptr: 0:17285918:408 gen 0
[Wed Jul 17 13:30:03 2024] bcachefs (cca5bc65-fe77-409d-a9fa-465a6e7f4eae): fatal error - emergency read only
bch2_extents_match() was reporting true for extents that did not
actually point to the same data.
bch2_extent_match() iterates over pairs of pointers, looking for
pointers that point to the same location on disk (with matching
generation numbers). However one or both extents may have been trimmed
(or merged) and they might not have the same disk offset: it corrects
for this by subtracting the key offset and the checksum entry offset.
However, this failed when an extent was immediately partially
overwritten, and the new overwrite was allocated the next adjacent disk
space.
Normally, with compression off, this would never cause a bug, since the
new extent would have to be immediately after the old extent for the
pointer offsets to match, and the rebalance index update path is not
looking for an extent outside the range of the extent it moved.
However with compression enabled, extents take up less space on disk
than they do in the btree index space - and spuriously matching after
partial overwrite is possible.
To fix this, add a secondary check, that strictly checks that the
regions pointed to on disk overlap.
https://github.com/koverstreet/bcachefs/issues/717
Signed-off-by: Kent Overstreet <[email protected]>
|
|
This fixes an assertion pop in io_write.c - if we don't return an error
we're supposed to have completed all the btree updates.
Signed-off-by: Kent Overstreet <[email protected]>
|
|
[BUG]
There is an internal report that KASAN is reporting use-after-free, with
the following backtrace:
BUG: KASAN: slab-use-after-free in btrfs_check_read_bio+0xa68/0xb70 [btrfs]
Read of size 4 at addr ffff8881117cec28 by task kworker/u16:2/45
CPU: 1 UID: 0 PID: 45 Comm: kworker/u16:2 Not tainted 6.11.0-rc2-next-20240805-default+ #76
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.2-3-gd478f380-rebuilt.opensuse.org 04/01/2014
Workqueue: btrfs-endio btrfs_end_bio_work [btrfs]
Call Trace:
dump_stack_lvl+0x61/0x80
print_address_description.constprop.0+0x5e/0x2f0
print_report+0x118/0x216
kasan_report+0x11d/0x1f0
btrfs_check_read_bio+0xa68/0xb70 [btrfs]
process_one_work+0xce0/0x12a0
worker_thread+0x717/0x1250
kthread+0x2e3/0x3c0
ret_from_fork+0x2d/0x70
ret_from_fork_asm+0x11/0x20
Allocated by task 20917:
kasan_save_stack+0x37/0x60
kasan_save_track+0x10/0x30
__kasan_slab_alloc+0x7d/0x80
kmem_cache_alloc_noprof+0x16e/0x3e0
mempool_alloc_noprof+0x12e/0x310
bio_alloc_bioset+0x3f0/0x7a0
btrfs_bio_alloc+0x2e/0x50 [btrfs]
submit_extent_page+0x4d1/0xdb0 [btrfs]
btrfs_do_readpage+0x8b4/0x12a0 [btrfs]
btrfs_readahead+0x29a/0x430 [btrfs]
read_pages+0x1a7/0xc60
page_cache_ra_unbounded+0x2ad/0x560
filemap_get_pages+0x629/0xa20
filemap_read+0x335/0xbf0
vfs_read+0x790/0xcb0
ksys_read+0xfd/0x1d0
do_syscall_64+0x6d/0x140
entry_SYSCALL_64_after_hwframe+0x4b/0x53
Freed by task 20917:
kasan_save_stack+0x37/0x60
kasan_save_track+0x10/0x30
kasan_save_free_info+0x37/0x50
__kasan_slab_free+0x4b/0x60
kmem_cache_free+0x214/0x5d0
bio_free+0xed/0x180
end_bbio_data_read+0x1cc/0x580 [btrfs]
btrfs_submit_chunk+0x98d/0x1880 [btrfs]
btrfs_submit_bio+0x33/0x70 [btrfs]
submit_one_bio+0xd4/0x130 [btrfs]
submit_extent_page+0x3ea/0xdb0 [btrfs]
btrfs_do_readpage+0x8b4/0x12a0 [btrfs]
btrfs_readahead+0x29a/0x430 [btrfs]
read_pages+0x1a7/0xc60
page_cache_ra_unbounded+0x2ad/0x560
filemap_get_pages+0x629/0xa20
filemap_read+0x335/0xbf0
vfs_read+0x790/0xcb0
ksys_read+0xfd/0x1d0
do_syscall_64+0x6d/0x140
entry_SYSCALL_64_after_hwframe+0x4b/0x53
[CAUSE]
Although I cannot reproduce the error, the report itself is good enough
to pin down the cause.
The call trace is the regular endio workqueue context, but the
free-by-task trace is showing that during btrfs_submit_chunk() we
already hit a critical error, and is calling btrfs_bio_end_io() to error
out. And the original endio function called bio_put() to free the whole
bio.
This means a double freeing thus causing use-after-free, e.g.:
1. Enter btrfs_submit_bio() with a read bio
The read bio length is 128K, crossing two 64K stripes.
2. The first run of btrfs_submit_chunk()
2.1 Call btrfs_map_block(), which returns 64K
2.2 Call btrfs_split_bio()
Now there are two bios, one referring to the first 64K, the other
referring to the second 64K.
2.3 The first half is submitted.
3. The second run of btrfs_submit_chunk()
3.1 Call btrfs_map_block(), which by somehow failed
Now we call btrfs_bio_end_io() to handle the error
3.2 btrfs_bio_end_io() calls the original endio function
Which is end_bbio_data_read(), and it calls bio_put() for the
original bio.
Now the original bio is freed.
4. The submitted first 64K bio finished
Now we call into btrfs_check_read_bio() and tries to advance the bio
iter.
But since the original bio (thus its iter) is already freed, we
trigger the above use-after free.
And even if the memory is not poisoned/corrupted, we will later call
the original endio function, causing a double freeing.
[FIX]
Instead of calling btrfs_bio_end_io(), call btrfs_orig_bbio_end_io(),
which has the extra check on split bios and do the proper refcounting
for cloned bios.
Furthermore there is already one extra btrfs_cleanup_bio() call, but
that is duplicated to btrfs_orig_bbio_end_io() call, so remove that
label completely.
Reported-by: David Sterba <[email protected]>
Fixes: 852eee62d31a ("btrfs: allow btrfs_submit_bio to split bios")
CC: [email protected] # 6.6+
Reviewed-by: Josef Bacik <[email protected]>
Signed-off-by: Qu Wenruo <[email protected]>
Reviewed-by: David Sterba <[email protected]>
Signed-off-by: David Sterba <[email protected]>
|
|
Currently, we copy the mtime and ctime to the in-core inode and then
mark the inode dirty. This is fine for certain types of filesystems, but
not all. Some require a real setattr to properly change these values
(e.g. ceph or reexported NFS).
Fix this code to call notify_change() instead, which is the proper way
to effect a setattr. There is one problem though:
In this case, the client is holding a write delegation and has sent us
attributes to update our cache. We don't want to break the delegation
for this since that would defeat the purpose. Add a new ATTR_DELEG flag
that makes notify_change bypass the try_break_deleg call.
Fixes: c5967721e106 ("NFSD: handle GETATTR conflict with write delegation")
Reviewed-by: Christian Brauner <[email protected]>
Signed-off-by: Jeff Layton <[email protected]>
Signed-off-by: Chuck Lever <[email protected]>
|
|
create_elf_fdpic_tables() does not correctly account the space for the
AUX vector when an architecture has ELF_HWCAP2 defined. Prior to the
commit 10e29251be0e ("binfmt_elf_fdpic: fix /proc/<pid>/auxv") it
resulted in the last entry of the AUX vector being set to zero, but with
that change it results in a kernel BUG.
Fix that by adding one to the number of AUXV entries (nitems) when
ELF_HWCAP2 is defined.
Fixes: 10e29251be0e ("binfmt_elf_fdpic: fix /proc/<pid>/auxv")
Cc: [email protected]
Reported-by: Greg Ungerer <[email protected]>
Closes: https://lore.kernel.org/lkml/[email protected]/
Signed-off-by: Max Filippov <[email protected]>
Tested-by: Greg Ungerer <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Kees Cook <[email protected]>
|
|
Once we drop the delegation reference, the fields embedded in it are no
longer safe to access. Do that last.
Fixes: c5967721e106 ("NFSD: handle GETATTR conflict with write delegation")
Signed-off-by: Jeff Layton <[email protected]>
Signed-off-by: Chuck Lever <[email protected]>
|
|
Once we've dropped the flc_lock, there is nothing that ensures that the
delegation that was found will still be around later. Take a reference
to it while holding the lock and then drop it when we've finished with
the delegation.
Fixes: c5967721e106 ("NFSD: handle GETATTR conflict with write delegation")
Signed-off-by: Jeff Layton <[email protected]>
Signed-off-by: Chuck Lever <[email protected]>
|
|
extent_fiemap()
There's a warning (probably on some older compiler version):
fs/btrfs/fiemap.c: warning: 'last_extent_end' may be used uninitialized in this function [-Wmaybe-uninitialized]: => 822:19
Initialize the variable to 0 although it's not necessary as it's either
properly set or not used after an error. The called function is in the
same file so this is a false alert but we want to fix all
-Wmaybe-uninitialized reports.
Link: https://lore.kernel.org/all/[email protected]/
Reported-by: Geert Uytterhoeven <[email protected]>
Signed-off-by: David Sterba <[email protected]>
|
|
I notice a rmap query bug in xfs_io fsmap:
[root@fedora ~]# xfs_io -c 'fsmap -vvvv' /mnt
EXT: DEV BLOCK-RANGE OWNER FILE-OFFSET AG AG-OFFSET TOTAL
0: 253:16 [0..7]: static fs metadata 0 (0..7) 8
1: 253:16 [8..23]: per-AG metadata 0 (8..23) 16
2: 253:16 [24..39]: inode btree 0 (24..39) 16
3: 253:16 [40..47]: per-AG metadata 0 (40..47) 8
4: 253:16 [48..55]: refcount btree 0 (48..55) 8
5: 253:16 [56..103]: per-AG metadata 0 (56..103) 48
6: 253:16 [104..127]: free space 0 (104..127) 24
......
Bug:
[root@fedora ~]# xfs_io -c 'fsmap -vvvv -d 0 3' /mnt
[root@fedora ~]#
Normally, we should be able to get one record, but we got nothing.
The root cause of this problem lies in the incorrect setting of rm_owner in
the rmap query. In the case of the initial query where the owner is not
set, __xfs_getfsmap_datadev() first sets info->high.rm_owner to ULLONG_MAX.
This is done to prevent any omissions when comparing rmap items. However,
if the current ag is detected to be the last one, the function sets info's
high_irec based on the provided key. If high->rm_owner is not specified, it
should continue to be set to ULLONG_MAX; otherwise, there will be issues
with interval omissions. For example, consider "start" and "end" within the
same block. If high->rm_owner == 0, it will be smaller than the founded
record in rmapbt, resulting in a query with no records. The main call stack
is as follows:
xfs_ioc_getfsmap
xfs_getfsmap
xfs_getfsmap_datadev_rmapbt
__xfs_getfsmap_datadev
info->high.rm_owner = ULLONG_MAX
if (pag->pag_agno == end_ag)
xfs_fsmap_owner_to_rmap
// set info->high.rm_owner = 0 because fmr_owner == -1ULL
dest->rm_owner = 0
// get nothing
xfs_getfsmap_datadev_rmapbt_query
The problem can be resolved by simply modify the xfs_fsmap_owner_to_rmap
function internal logic to achieve.
After applying this patch, the above problem have been solved:
[root@fedora ~]# xfs_io -c 'fsmap -vvvv -d 0 3' /mnt
EXT: DEV BLOCK-RANGE OWNER FILE-OFFSET AG AG-OFFSET TOTAL
0: 253:16 [0..7]: static fs metadata 0 (0..7) 8
Fixes: e89c041338ed ("xfs: implement the GETFSMAP ioctl")
Signed-off-by: Zizhi Wo <[email protected]>
Reviewed-by: Darrick J. Wong <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
Signed-off-by: Chandan Babu R <[email protected]>
|
|
Don't bother reporting the number of bytes that we "trimmed" because the
underlying storage isn't required to do anything(!) and failed discard
IOs aren't reported to the caller anyway. It's not like userspace can
use the reported value for anything useful like adjusting the offset
parameter of the next call, and it's not like anyone ever wrote a
manpage about FITRIM's out parameters.
Signed-off-by: Darrick J. Wong <[email protected]>
Reviewed-by: Christoph Hellwig <[email protected]>
Tested-by: Christoph Hellwig <[email protected]>
Signed-off-by: Chandan Babu R <[email protected]>
|
|
As a result of the factoring in commit 14dd46cf31f4 ("xfs: split
xfs_inobt_init_cursor"), mount started taking a long time on a
user's filesystem. For Anders, this made mount times regress from
under a second to over 15 minutes for a filesystem with only 30
million inodes in it.
Anders bisected it down to the above commit, but even then the bug
was not obvious. In this commit, over 20 calls to
xfs_inobt_init_cursor() were modified, and some we modified to call
a new function named xfs_finobt_init_cursor().
If that takes you a moment to reread those function names to see
what the rename was, then you have realised why this bug wasn't
spotted during review. And it wasn't spotted on inspection even
after the bisect pointed at this commit - a single missing "f" isn't
the easiest thing for a human eye to notice....
The result is that xfs_finobt_count_blocks() now incorrectly calls
xfs_inobt_init_cursor() so it is now walking the inobt instead of
the finobt. Hence when there are lots of allocated inodes in a
filesystem, mount takes a -long- time run because it now walks a
massive allocated inode btrees instead of the small, nearly empty
free inode btrees. It also means all the finobt space reservations
are wrong, so mount could potentially given ENOSPC on kernel
upgrade.
In hindsight, commit 14dd46cf31f4 should have been two commits - the
first to convert the finobt callers to the new API, the second to
modify the xfs_inobt_init_cursor() API for the inobt callers. That
would have made the bug very obvious during review.
Fixes: 14dd46cf31f4 ("xfs: split xfs_inobt_init_cursor")
Reported-by: Anders Blomdell <[email protected]>
Signed-off-by: Dave Chinner <[email protected]>
Reviewed-by: Christoph Hellwig <[email protected]>
Reviewed-by: Darrick J. Wong <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
Signed-off-by: Chandan Babu R <[email protected]>
|
|
willy pointed out that folio_mark_dirty is the correct function to use
to mark an xfile folio dirty because it calls out to the mapping's aops
to mark it dirty. For tmpfs this likely doesn't matter much since it
currently uses nop_dirty_folio, but let's use the abstractions properly.
Reported-by: [email protected]
Fixes: 6907e3c00a40 ("xfs: add file_{get,put}_folio")
Signed-off-by: Darrick J. Wong <[email protected]>
Reviewed-by: Christoph Hellwig <[email protected]>
Signed-off-by: Chandan Babu R <[email protected]>
|
|
"KjellR" complained on IRC that an old V4 filesystem suddenly stopped
mounting after upgrading from 6.9.11 to 6.10.3, with the following splat
when trying to read the rt bitmap inode:
00000000: 49 4e 80 00 01 02 00 01 00 00 00 00 00 00 00 00 IN..............
00000010: 00 00 00 01 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000020: 00 00 00 00 00 00 00 00 43 d2 a9 da 21 0f d6 30 ........C...!..0
00000030: 43 d2 a9 da 21 0f d6 30 00 00 00 00 00 00 00 00 C...!..0........
00000040: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000050: 00 00 00 02 00 00 00 00 00 00 00 04 00 00 00 00 ................
00000060: ff ff ff ff 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000070: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
As Dave Chinner points out, this is a V1 inode with both di_onlink and
di_nlink set to 1 and di_flushiter == 0. In other words, this inode was
formatted this way by mkfs and hasn't been touched since then.
Back in the old days of xfsprogs 3.2.3, I observed that libxfs_ialloc
would set di_nlink, but if the filesystem didn't have NLINK, it would
then set di_version = 1. libxfs_iflush_int later sees the V1 inode and
copies the value of di_nlink to di_onlink without zeroing di_onlink.
Eventually this filesystem must have been upgraded to support NLINK
because 6.10 doesn't support !NLINK filesystems, which is how we tripped
over this old behavior. The filesystem doesn't have a realtime section,
so that's why the rtbitmap inode has never been touched.
Fix this by removing the di_onlink/di_nlink checking for all V1/V2
inodes because this is a muddy mess. The V3 inode handling code has
always supported NLINK and written di_onlink==0 so keep that check.
The removal of the V1 inode handling code when we dropped support for
!NLINK obscured this old behavior.
Reported-by: [email protected]
Fixes: 40cb8613d612 ("xfs: check unused nlink fields in the ondisk inode")
Signed-off-by: Darrick J. Wong <[email protected]>
Reviewed-by: Christoph Hellwig <[email protected]>
Signed-off-by: Chandan Babu R <[email protected]>
|
|
We have transient failures with btrfs/301, specifically in the part
where we do
for i in $(seq 0 10); do
write 50m to file
rm -f file
done
Sometimes this will result in a transient quota error, and it's because
sometimes we start writeback on the file which results in a delayed
iput, and thus the rm doesn't actually clean the file up. When we're
flushing the quota space we need to run the delayed iputs to make sure
all the unlinks that we think have completed have actually completed.
This removes the small window where we could fail to find enough space
in our quota.
CC: [email protected] # 5.15+
Reviewed-by: Qu Wenruo <[email protected]>
Signed-off-by: Josef Bacik <[email protected]>
Signed-off-by: David Sterba <[email protected]>
|
|
The cifs filesystem doesn't quite emulate FALLOC_FL_PUNCH_HOLE correctly
(note that due to lack of protocol support, it can't actually implement it
directly). Whilst it will (partially) invalidate dirty folios in the
pagecache, it doesn't write them back first, and so the EOF marker on the
server may be lower than inode->i_size.
This presents a problem, however, as if the punched hole invalidates the
tail of the locally cached dirty data, writeback won't know it needs to
move the EOF over to account for the hole punch (which isn't supposed to
move the EOF). We could just write zeroes over the punched out region of
the pagecache and write that back - but this is supposed to be a
deallocatory operation.
Fix this by manually moving the EOF over on the server after the operation
if the hole punched would corrupt it.
Note that the FSCTL_SET_ZERO_DATA RPC and the setting of the EOF should
probably be compounded to stop a third party interfering (or, at least,
massively reduce the chance).
This was reproducible occasionally by using fsx with the following script:
truncate 0x0 0x375e2 0x0
punch_hole 0x2f6d3 0x6ab5 0x375e2
truncate 0x0 0x3a71f 0x375e2
mapread 0xee05 0xcf12 0x3a71f
write 0x2078e 0x5604 0x3a71f
write 0x3ebdf 0x1421 0x3a71f *
punch_hole 0x379d0 0x8630 0x40000 *
mapread 0x2aaa2 0x85b 0x40000
fallocate 0x1b401 0x9ada 0x40000
read 0x15f2 0x7d32 0x40000
read 0x32f37 0x7a3b 0x40000 *
The second "write" should extend the EOF to 0x40000, and the "punch_hole"
should operate inside of that - but that depends on whether the VM gets in
and writes back the data first. If it doesn't, the file ends up 0x3a71f in
size, not 0x40000.
Fixes: 31742c5a3317 ("enable fallocate punch hole ("fallocate -p") for SMB3")
Signed-off-by: David Howells <[email protected]>
cc: Steve French <[email protected]>
cc: Paulo Alcantara <[email protected]>
cc: Shyam Prasad N <[email protected]>
cc: Jeff Layton <[email protected]>
cc: [email protected]
cc: [email protected]
Signed-off-by: Steve French <[email protected]>
|
|
rqst.rq_iter needs to be truncated otherwise we'll
also send the bytes into the stream socket...
This is the logic behind rqst.rq_npages = 0, which was removed in
"cifs: Change the I/O paths to use an iterator rather than a page list"
(d08089f649a0cfb2099c8551ac47eef0cc23fdf2).
Cc: [email protected]
Fixes: d08089f649a0 ("cifs: Change the I/O paths to use an iterator rather than a page list")
Reviewed-by: David Howells <[email protected]>
Signed-off-by: Stefan Metzmacher <[email protected]>
Signed-off-by: Steve French <[email protected]>
|
|
Reviewed-by: David Howells <[email protected]>
Fixes: d08089f649a0 ("cifs: Change the I/O paths to use an iterator rather than a page list")
Signed-off-by: Stefan Metzmacher <[email protected]>
Signed-off-by: Steve French <[email protected]>
|
|
This happens when called from SMB2_read() while using rdma
and reaching the rdma_readwrite_threshold.
Cc: [email protected]
Fixes: a6559cc1d35d ("cifs: split out smb3_use_rdma_offload() helper")
Reviewed-by: David Howells <[email protected]>
Signed-off-by: Stefan Metzmacher <[email protected]>
Signed-off-by: Steve French <[email protected]>
|
|
Pull bcachefs fixes from Kent Overstreet:
- assorted syzbot fixes
- some upgrade fixes for old (pre 1.0) filesystems
- fix for moving data off a device that was switched to durability=0
after data had been written to it.
- nocow deadlock fix
- fix for new rebalance_work accounting
* tag 'bcachefs-2024-08-24' of git://evilpiepirate.org/bcachefs: (28 commits)
bcachefs: Fix rebalance_work accounting
bcachefs: Fix failure to flush moves before sleeping in copygc
bcachefs: don't use rht_bucket() in btree_key_cache_scan()
bcachefs: add missing inode_walker_exit()
bcachefs: clear path->should_be_locked in bch2_btree_key_cache_drop()
bcachefs: Fix double assignment in check_dirent_to_subvol()
bcachefs: Fix refcounting in discard path
bcachefs: Fix compat issue with old alloc_v4 keys
bcachefs: Fix warning in bch2_fs_journal_stop()
fs/super.c: improve get_tree() error message
bcachefs: Fix missing validation in bch2_sb_journal_v2_validate()
bcachefs: Fix replay_now_at() assert
bcachefs: Fix locking in bch2_ioc_setlabel()
bcachefs: fix failure to relock in btree_node_fill()
bcachefs: fix failure to relock in bch2_btree_node_mem_alloc()
bcachefs: unlock_long() before resort in journal replay
bcachefs: fix missing bch2_err_str()
bcachefs: fix time_stats_to_text()
bcachefs: Fix bch2_bucket_gens_init()
bcachefs: Fix bch2_trigger_alloc assert
...
|
|
Pull smb server fixes from Steve French:
- query directory flex array fix
- fix potential null ptr reference in open
- fix error message in some open cases
- two minor cleanups
* tag '6.11-rc5-server-fixes' of git://git.samba.org/ksmbd:
smb/server: update misguided comment of smb2_allocate_rsp_buf()
smb/server: remove useless assignment of 'file_present' in smb2_open()
smb/server: fix potential null-ptr-deref of lease_ctx_info in smb2_open()
smb/server: fix return value of smb2_open()
ksmbd: the buffer of smb2 query dir response has at least 1 byte
|
|
rebalance_work was keying off of the presence of rebelance_opts in the
extent - but that was incorrect, we keep those around after rebalance
for indirect extents since the inode's options are not directly
available
Fixes: 20ac515a9cc7 ("bcachefs: bch_acct_rebalance_work")
Signed-off-by: Kent Overstreet <[email protected]>
|
|
This fixes an apparent deadlock - rebalance would get stuck trying to
take nocow locks because they weren't being released by copygc.
Signed-off-by: Kent Overstreet <[email protected]>
|