Starting with version 5 the following properties change:
- UBIFS_FLG_DOUBLE_HASH is mandatory
- UBIFS_FLG_ENCRYPTION is optional but depdens on UBIFS_FLG_DOUBLE_HASH
- Filesystems with unknown super block flags will be rejected, this
allows us in future to add new features without raising the UBIFS
write version.
Signed-off-by: Richard Weinberger <richard@nod.at>
This feature flag indicates that all directory entry nodes have a 32bit
cookie set and therefore UBIFS is allowed to perform lookups by hash.
Signed-off-by: Richard Weinberger <richard@nod.at>
UBIFS stores a 32bit hash of every file, for traditional lookups by name
this scheme is fine since UBIFS can first try to find the file by the
hash of the filename and upon collisions it can walk through all entries
with the same hash and do a string compare.
When filesnames are encrypted fscrypto will ask the filesystem for a
unique cookie, based on this cookie the filesystem has to be able to
locate the target file again. With 32bit hashes this is impossible
because the chance for collisions is very high. Do deal with that we
store a 32bit cookie directly in the UBIFS directory entry node such
that we get a 64bit cookie (32bit from filename hash and the dent
cookie). For a lookup by hash UBIFS finds the entry by the first 32bit
and then compares the dent cookie. If it does not match, it has to do a
linear search of the whole directory and compares all dent cookies until
the correct entry is found.
Signed-off-by: Richard Weinberger <richard@nod.at>
As of now all filenames known by UBIFS are strings with a NUL
terminator. With encrypted filenames a filename can be any binary
string and the r5 function cannot search for the NUL terminator.
UBIFS always knows how long a filename is, therefore we can change
the hash function to iterate over the filename length to work
correctly with binary strings.
Signed-off-by: Richard Weinberger <richard@nod.at>
When data of a data node is compressed and encrypted
we need to store the size of the compressed data because
before encryption we may have to add padding bytes.
For the new field we consume the last two padding bytes
in struct ubifs_data_node. Two bytes are fine because
the data length is at most 4096.
Signed-off-by: Richard Weinberger <richard@nod.at>
When we're creating a new inode in UBIFS the inode is not
yet exposed and fscrypto calls ubifs_xattr_set() without
holding the inode mutex. This is okay but ubifs_xattr_set()
has to know about this.
Signed-off-by: Richard Weinberger <richard@nod.at>
When a file is moved or linked into another directory
its current crypto policy has to be compatible with the
target policy.
Signed-off-by: Richard Weinberger <richard@nod.at>
We need ->open() for files to load the crypto key.
If the no key is present and the file is encrypted,
refuse to open.
Signed-off-by: Richard Weinberger <richard@nod.at>
We need the ->open() hook to load the crypto context
which is needed for all crypto operations within that
directory.
Signed-off-by: Richard Weinberger <richard@nod.at>
fscrypto will need this function too. Also get struct ubifs_info
from the provided inode. Not all callers will have a reference to
struct ubifs_info.
Signed-off-by: Richard Weinberger <richard@nod.at>
'ubifs_fast_find_freeable()' can not return an error pointer, so this test
can be removed.
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Richard Weinberger <richard@nod.at>
Right now wbuf timer has hardcoded timeouts and there is no place for
manual adjustments. Some projects / cases many need that though. Few
file systems allow doing that by respecting dirty_writeback_interval
that can be set using sysctl (dirty_writeback_centisecs).
Lowering dirty_writeback_interval could be some way of dealing with user
space apps lacking proper fsyncs. This is definitely *not* a perfect
solution but we don't have ideal (user space) world. There were already
advanced discussions on this matter, mostly when ext4 was introduced and
it wasn't behaving as ext3. Anyway, the final decision was to add some
hacks to the ext4, as trying to fix whole user space or adding new API
was pointless.
We can't (and shouldn't?) just follow ext4. We can't e.g. sync on close
as this would cause too many commits and flash wearing. On the other
hand we still should allow some trade-off between -o sync and default
wbuf timeout. Respecting dirty_writeback_interval should allow some sane
cutomizations if used warily.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Reviewed-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Richard Weinberger <richard@nod.at>
Values of these fields are set during init and never modified. They are
used (read) in a single function only. There isn't really any reason to
keep them in a struct. It only makes struct just a bit bigger without
any visible gain.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Reviewed-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Richard Weinberger <richard@nod.at>
This patch fix a missing size change in f2fs_setattr
Signed-off-by: Yunlei He <heyunlei@huawei.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
The "perf_counter_reset" case has already been handled above.
Moreover "ORANGEFS_PARAM_REQUEST_OP_READAHEAD_COUNT_SIZE" is not a really
consistent.
It is likely that this (dead) code is a cut and paste left over.
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Martin Brandenburg <martin@omnibond.com>
Signed-off-by: Mike Marshall <hubcap@omnibond.com>
allocates string 'new' is not free'd on the exit path when
cdm_element_count <= 0. Fix this by kfree'ing it.
Fixes CoverityScan CID#1375923 "Resource Leak"
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Mike Marshall <hubcap@omnibond.com>
This is exposing an existing deadlock between fsync and AIO. Until we
have the deadlock fixed, I'm pulling this one out.
This reverts commit a23eaa875f.
Signed-off-by: Chris Mason <clm@fb.com>
... to better explain its purpose after introducing in-place encryption
without bounce buffer.
Signed-off-by: David Gstir <david@sigma-star.at>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Since fscrypt users can now indicated if fscrypt_encrypt_page() should
use a bounce page, we can delay the bounce page pool initialization util
it is really needed. That is until fscrypt_operations has no
FS_CFLG_OWN_PAGES flag set.
Signed-off-by: David Gstir <david@sigma-star.at>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Rename the FS_CFLG_INPLACE_ENCRYPTION flag to FS_CFLG_OWN_PAGES which,
when set, indicates that the fs uses pages under its own control as
opposed to writeback pages which require locking and a bounce buffer for
encryption.
Signed-off-by: David Gstir <david@sigma-star.at>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
In case of in-place encryption fscrypt_ctx was allocated but never
released. Since we don't need it for in-place encryption, we skip
allocating it.
Fixes: 1c7dcf69ee ("fscrypt: Add in-place encryption mode")
Signed-off-by: David Gstir <david@sigma-star.at>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Actually use the fs-provided index instead of always using page->index
which is only set for page-cache pages.
Fixes: 9c4bb8a3a9 ("fscrypt: Let fs select encryption index/tweak")
Signed-off-by: David Gstir <david@sigma-star.at>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The fscrypt_initalize() function isn't used outside fs/crypto, so
there's no point making it be an exported symbol.
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Eric Biggers <ebiggers@google.com>
To avoid namespace collisions, rename get_crypt_info() to
fscrypt_get_crypt_info(). The function is only used inside the
fs/crypto directory, so declare it in the new header file,
fscrypt_private.h.
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Multiple bugs were recently fixed in the "set encryption policy" ioctl.
To make it clear that fscrypt_process_policy() and fscrypt_get_policy()
implement ioctls and therefore their implementations must take standard
security and correctness precautions, rename them to
fscrypt_ioctl_set_policy() and fscrypt_ioctl_get_policy(). Make the
latter take in a struct file * to make it consistent with the former.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
SHA256 and ENCRYPTED_KEYS are not needed. CTR shouldn't be needed
either, but I left it for now because it was intentionally added by
commit 71dea01ea2 ("ext4 crypto: require CONFIG_CRYPTO_CTR if ext4
encryption is enabled"). So it sounds like there may be a dependency
problem elsewhere, which I have not been able to identify specifically,
that must be solved before CTR can be removed.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Currently data journalling is incompatible with encryption: enabling both
at the same time has never been supported by design, and would result in
unpredictable behavior. However, users are not precluded from turning on
both features simultaneously. This change programmatically replaces data
journaling for encrypted regular files with ordered data journaling mode.
Background:
Journaling encrypted data has not been supported because it operates on
buffer heads of the page in the page cache. Namely, when the commit
happens, which could be up to five seconds after caching, the commit
thread uses the buffer heads attached to the page to copy the contents of
the page to the journal. With encryption, it would have been required to
keep the bounce buffer with ciphertext for up to the aforementioned five
seconds, since the page cache can only hold plaintext and could not be
used for journaling. Alternatively, it would be required to setup the
journal to initiate a callback at the commit time to perform deferred
encryption - in this case, not only would the data have to be written
twice, but it would also have to be encrypted twice. This level of
complexity was not justified for a mode that in practice is very rarely
used because of the overhead from the data journalling.
Solution:
If data=journaled has been set as a mount option for a filesystem, or if
journaling is enabled on a regular file, do not perform journaling if the
file is also encrypted, instead fall back to the data=ordered mode for the
file.
Rationale:
The intent is to allow seamless and proper filesystem operation when
journaling and encryption have both been enabled, and have these two
conflicting features gracefully resolved by the filesystem.
Fixes: 4461471107
Signed-off-by: Sergey Karamov <skaramov@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@vger.kernel.org
Connect the new VFS clone_range, copy_range, and dedupe_range features
to the existing reflink capability of ocfs2. Compared to the existing
ocfs2 reflink ioctl We have to do things a little differently to support
the VFS semantics (we can clone subranges of a file but we don't clone
xattrs), but the VFS ioctls are more broadly supported.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
v2: Convert inline data files to extents files before reflinking,
and fix i_blocks so that stat(2) output is correct.
v3: Make zero-length dedupe consistent with btrfs behavior.
v4: Use VFS double-inode lock routines and remove MAX_DEDUPE_LEN.
When ocfs2 shares blocks from one file to another, it's necessary to
charge that many blocks to the quota because ocfs2 tallies block charges
according to the number of blocks mapped, not the number of physical
blocks used.
Without this patch, reflinking X blocks and then CoWing all of them
causes quota usage to *decrease* by X as seen in generic/305.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
generic/188 triggered a dmesg stack trace because the dio completion
was casting a buffer head to an on-disk inode, which is whacky.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Always unlock the inode when completing dio writes, even if an error
has occurrred. The caller already checks the inode and unlocks it
if needed, so we might as well reduce contention.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
ocfs2_dio_end_io_write eats whatever errors may happen,
which means that write errors do not propagate to userspace.
Fix that.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
When we're adding the refcount flag to an extent, we have to budget
enough space to handle a full extent btree split in addition to
whatever modifications have to be made to the refcount btree. We
don't currently do this, with the result that generic/186 crashes
when we need an extent split but not a refcount split because meta_ac
never gets allocated.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
The swapfile mechanism calls bmap once to find all the swap file
mappings, which means that we cannot properly support CoW remapping.
Therefore, error out if the swap code tries to call bmap on a
refcounted file.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Replace the open-coded inode refcount flag test with a helper function
to reduce the potential for bugs.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
... and don't zero anything on short copy; just unlock
and return 0 if that has happened on non-uptodate page.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
If we had a short copy into an uptodate page, there's no reason
whatsoever to zero anything; OTOH, if that page had _not_ been
uptodate, we must have been trying to overwrite it completely
and got a short copy. In that case, overwriting the end with
zeroes, marking uptodate and sending to server is just plain
wrong. Just unlock, keep it non-uptodate and return 0.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
a) the page is uptodate - ->write_begin() would either fail (in which
case we don't reach ->write_end()), or unstuff the inode, or find the
page already uptodate, or do a successful call of stuffed_readpage(),
which would've made it uptodate
b) zeroing the tail in pagecache is wrong. kill -9 at the right time
while writing unmodified file contents to the same file should _not_
leave us in a situation when read() from the file will be reporting
it full of zeroes. Especially since that effect will be transient -
at some later point the page will be evicted and then we'll be back
to the real file contents.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
don't zero on short copies; if the page was uptodate it's just plain
wrong, and if it wasn't we'll be better off just returning 0 and
buggering off.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
We should set the error code if kzalloc() fails.
Fixes: 67cf5b09a4 ("ext4: add the basic function for inline data support")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@vger.kernel.org
Don't load an inode with a negative size; this causes integer overflow
problems in the VFS.
[ Added EXT4_ERROR_INODE() to mark file system as corrupted. -TYT]
Fixes: a48380f769 (ext4: rename i_dir_acl to i_size_high)
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
Clients can set the umask attribute when creating files to cause the
server to apply it always except when inheriting permissions from the
parent directory. That way, the new files will end up with the same
permissions as files created locally.
See https://tools.ietf.org/html/draft-ietf-nfsv4-umask-02 for more details.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
What matters when deciding if we should make a page uptodate is
not how much we _wanted_ to copy, but how much we actually have
copied. As it is, on architectures that do not zero tail on
short copy we can leave uninitialized data in page marked uptodate.
Cc: stable@vger.kernel.org
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The flexfiles client can piggyback both layout errors and layoutstats
as part of the layoutreturn. Both these payloads can get large, with
20 layout error entries taking up about 1.2K, and 4 layoutstats entries
taking up another 1K.
This patch allows a maximum payload of 4k by allocating a full page.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Hoist both the XFS reflink inode state and preparation code and the XFS
file blocks compare functions into the VFS so that ocfs2 can take
advantage of it for reflink and dedupe.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
A clone is a perfectly fine implementation of a file copy, so most
file systems just implement the copy that way. Instead of duplicating
this logic move it to the VFS. Currently btrfs and XFS implement copies
the same way as clones and there is no behavior change for them, cifs
only implements clones and grow support for copy_file_range with this
patch. NFS implements both, so this will allow copy_file_range to work
on servers that only implement CLONE and be lot more efficient on servers
that implements CLONE and COPY.
Signed-off-by: Christoph Hellwig <hch@lst.de>
kernel crashes. Marked for stable - it goes back to 4.6, but started
popping up only in 4.8.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAABCAAGBQJYSo2yAAoJEEp/3jgCEfOLa1UH/j/nhhmy6bkvNTNQT9PuWmlu
1cGG6JkJq/USEaSO+VXGSAjBCjCngNTYYXBo0IBCnkf11tuwagvz/9LSbvy9P+vu
1IKwcJBFpgcEMEZWsYjVui9uFiDcLYiTPt4pux4tQ4vyj6HEFgioTg/430ApUEOS
ywO1pjRz8RH0FlKhhcTRGOwVcwUzI/aRw7aLeflSwz3mDnh6ajp/8pjvxWf7AN+V
Ih9LygjYNb4IdUcgN2G05z2qKLPfNAoBA+kRdEkOzecX2J0Db8Bu1bfZBxgOK+ui
kpdVlFPkpULbwjlLLpvmOgy7FKgmLfdxuEuQol8hCu0KQ+buP/kZnbjg6QBeCtk=
=1nK/
-----END PGP SIGNATURE-----
Merge tag 'ceph-for-4.9-rc9' of git://github.com/ceph/ceph-client
Pull ceph fix from Ilya Dryomov:
"A fix for an issue with ->d_revalidate() in ceph, causing frequent
kernel crashes.
Marked for stable - it goes back to 4.6, but started popping up only
in 4.8"
* tag 'ceph-for-4.9-rc9' of git://github.com/ceph/ceph-client:
ceph: don't set req->r_locked_dir in ceph_d_revalidate
If .readlink == NULL implies generic_readlink().
Generated by:
to_del="\.readlink.*=.*generic_readlink"
for i in `git grep -l $to_del`; do sed -i "/$to_del"/d $i; done
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
If i_op->readlink is NULL, but i_op->get_link is set then vfs_readlink()
defaults to calling generic_readlink().
The IOP_DEFAULT_READLINK flag indicates that the above conditions are met
and the default action can be taken.
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Also check d_is_symlink() in callers instead of inode->i_op->readlink
because following patches will allow NULL ->readlink for symlinks.
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
The /proc/self and /proc/self-thread symlinks have separate but identical
functionality for reading and following. This cleanup utilizes
generic_readlink to remove the duplication.
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Here again we are copying form one buffer to another, while jumping through
hoops to make kernel memory look like userspace memory.
For no good reason, since vfs_get_link() provides exactly what is needed.
As a bonus, now the security hook for readlink is also called on the
underlying inode.
Note: this can be called from link-following context. But this is okay:
- not in RCU mode
- commit e54ad7f1ee ("proc: prevent stacking filesystems on top")
- ecryptfs is *reading* the underlying symlink not following it, so the
right security hook is being called
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Cc: Tyler Hicks <tyhicks@canonical.com>
btrfs_transaction_abort() has a WARN() to help us nail down whatever
problem lead to the abort. But most of the time, we're aborting for EIO,
and the warning just adds noise.
Signed-off-by: Chris Mason <clm@fb.com>
New inode operations were forgotten to be added to bad_inode. Most of the
time the op is checked for NULL before being called but marking the inode
bad and the check can race (very unlikely).
However in case of ->get_link() only DCACHE_SYMLINK_TYPE is checked before
calling the op, so there's no race and will definitely oops when trying to
follow links on such a beast.
Also remove comments about extinct ops.
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Cc: <stable@vger.kernel.org>
This is all unused code, so remove it.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Use NOFS for allocating btree cursors, since they can be called
under the ilock.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Commit 6552321831 ("xfs: remove i_iolock and use i_rwsem in the
VFS inode instead") introduced a regression that truncate(2) doesn't
check on new size, so it succeeds even if the new size exceeds the
current resource limit. Because xfs_setattr_size() was used instead
of xfs_vn_setattr_size(), and the latter calls xfs_vn_change_ok()
first to do sanity check on permission and new size.
This is found by truncate03 test from ltp, and the following is a
simplified reproducer:
#!/bin/bash
dev=/dev/sda5
mnt=/mnt/xfs
mkfs -t xfs -f $dev
mount $dev $mnt
# set max file size to 16k
ulimit -f 16
truncate -s $((16 * 1024 + 1)) /mnt/xfs/testfile
[ $? -eq 0 ] && echo "FAIL: truncate exceeded max file size"
ulimit -f unlimited
umount $mnt
Signed-off-by: Eryu Guan <eguan@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
We always perform integrity operations now, so these mount options
don't do anything. Deprecate them and mark them for removal in
in a year.
Signed-Off-By: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
There is no reason anymore for not issuing device integrity
operations when teh filesystem requires ordering or data integrity
guarantees. We should always issue cache flushes and FUA writes
where necessary and let the underlying storage optimise them as
necessary for correct integrity operation.
Signed-Off-By: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
When we create a new attribute, we first create a shortform
attribute, and try to fit the new attribute into it.
If that fails, we copy the (empty) attribute into a leaf attribute,
and do the copy again. Thus there can be a transient state where
we have an empty leaf attribute.
If we encounter this during log replay, the verifier will fail.
So add a test to ignore this part of the leaf attr verification
during log replay.
Thanks as usual to dchinner for spotting the problem.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
We encountered a deadlock where the SEQUENCE that accompanied the
LAYOUTGET triggered a session drain, while ff_layout_alloc_lseg
triggered a GETDEVICEINFO. The GETDEVICEINFO hung waiting for the
session drain, while the LAYOUTGET held the slot waiting for
alloc_lseg to finish.
Avoid this by moving the call to nfs4_find_get_deviceid out of
ff_layout_alloc_lseg and into nfs4_ff_layout_prepare_ds.
Signed-off-by: Fred Isaman <fred.isaman@gmail.com>
[dros@primarydata.com: pNFS/flexfiles: fix races in ff_layout_mirror_valid]
Signed-off-by: Weston Andros Adamson <dros@primarydata.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
This function sets req->r_locked_dir which is supposed to indicate to
ceph_fill_trace that the parent's i_rwsem is locked for write.
Unfortunately, there is no guarantee that the dir will be locked when
d_revalidate is called, so we really don't want ceph_fill_trace to do
any dcache manipulation from this context. Clear req->r_locked_dir since
it's clearly not safe to do that.
What we really want to know with d_revalidate is whether the dentry
still points to the same inode. ceph_fill_trace installs a pointer to
the inode in req->r_target_inode, so we can just compare that to
d_inode(dentry) to see if it's the same one after the lookup.
Also, since we aren't generally interested in the parent here, we can
switch to using a GETATTR to hint that to the MDS, which also means that
we only need to reserve one cap.
Finally, just remove the d_unhashed check. That's really outside the
purview of a filesystem's d_revalidate. If the thing became unhashed
while we're checking it, then that's up to the VFS to handle anyway.
Fixes: 200fd27c8f ("ceph: use lookup request to revalidate dentry")
Link: http://tracker.ceph.com/issues/18041
Reported-by: Donatas Abraitis <donatas.abraitis@gmail.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Reviewed-by: "Yan, Zheng" <zyan@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
f2fs_sync_file() remount_ro
- f2fs_readonly
- destroy_flush_cmd_control
- f2fs_issue_flush
- no fcc pointer!
So, this patch doesn't free fcc in this case, but just stop its kernel thread
which sends flush commands.
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
put_compat_statfs64() does NOT return -1 and setting errno to EOVERFLOW
when some variables(like: f_bsize) overflowed in the returned struct.
The reason is that the ubuf->f_blocks is __u64 type, it couldn't be
4bits as the judgement in put_comat_statfs64(). Here correct the
__u32 variables(in struct compat_statfs64) for comparison.
reproducer:
step1. mount hugetlbfs with two different pagesize on ppc64 arch.
$ hugeadm --pool-pages-max 16M:0
$ hugeadm --create-mount
$ mount | grep -i hugetlbfs
none on /var/lib/hugetlbfs/pagesize-16MB type hugetlbfs (rw,relatime,seclabel,pagesize=16777216)
none on /var/lib/hugetlbfs/pagesize-16GB type hugetlbfs (rw,relatime,seclabel,pagesize=17179869184)
step2. compile & run this C program.
$ cat statfs64_test.c
#define _LARGEFILE64_SOURCE
#include <stdio.h>
#include <sys/syscall.h>
#include <sys/statfs.h>
int main()
{
struct statfs64 sb;
int err;
err = syscall(SYS_statfs64, "/var/lib/hugetlbfs/pagesize-16GB", sizeof(sb), &sb);
if (err)
return -1;
printf("sizeof f_bsize = %d, f_bsize=%ld\n", sizeof(sb.f_bsize), sb.f_bsize);
return 0;
}
$ gcc -m32 statfs64_test.c
$ ./a.out
sizeof f_bsize = 4, f_bsize=0
Signed-off-by: Li Wang <liwang@redhat.com>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Previous mkfs.f2fs allows small partition inappropriately, so f2fs should detect
that as well.
Refer this in f2fs-tools.
mkfs.f2fs: detect small partition by overprovision ratio and # of segments
Reported-and-Tested-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
The layout-private data may depend on the layout and/or the inode
still existing when it does post-processing and frees its data, so we
need to free them after calling lrp->ld_private.ops->free().
This fixes a mirror list corruption issue in the flexfiles driver.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
When we're merging an old entry into our new entry, we want to ensure that
we add the list entry in the correct place.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Otherwise the lock context won't be freed when we're done with it.
From: NeilBrown <neilb@suse.com>
Fixes: 5bd3f817 ("NFSv4: change nfs4_select_rw_stateid to take a lock_context inplace of lock_owner")
Signed-off-by: Anna Schumaker <Anna.Schumaker@netapp.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Don't load an inode with a negative size; this causes integer overflow
problems in the VFS.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Jan Kara <jack@suse.cz>
On filesystems with a lot of metadata and in metadata intensive workloads
xfs_buf_find() is showing up at the top of the CPU cycles trace. Most of
the CPU time is spent on CPU cache misses while traversing the rbtree.
As the buffer cache does not need any kind of ordering, but fast lookups
a hashtable is the natural data structure to use. The rhashtable
infrastructure provides a self-scaling hashtable implementation and
allows lookups to proceed while the table is going through a resize
operation.
This reduces the CPU-time spent for the lookups to 1/3 even for small
filesystems with a relatively small number of cached buffers, with
possibly much larger gains on higher loaded filesystems.
[dchinner: reduce minimum hash size to an acceptable size for large
filesystems with many AGs with no active use.]
[dchinner: remove stale rbtree asserts.]
[dchinner: use xfs_buf_map for compare function argument.]
[dchinner: make functions static.]
[dchinner: remove redundant comments.]
Signed-off-by: Lucas Stach <dev@lynxeye.de>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Basically, the pjdfstests set the ownership of a file to 06555, and then
chowns it (as root) to a new uid/gid. Prior to commit a09f99edde ("fuse:
fix killing s[ug]id in setattr"), fuse would send down a setattr with both
the uid/gid change and a new mode. Now, it just sends down the uid/gid
change.
Technically this is NOTABUG, since POSIX doesn't _require_ that we clear
these bits for a privileged process, but Linux (wisely) has done that and I
think we don't want to change that behavior here.
This is caused by the use of should_remove_suid(), which will always return
0 when the process has CAP_FSETID.
In fact we really don't need to be calling should_remove_suid() at all,
since we've already been indicated that we should remove the suid, we just
don't want to use a (very) stale mode for that.
This patch should fix the above as well as simplify the logic.
Reported-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Fixes: a09f99edde ("fuse: fix killing s[ug]id in setattr")
Cc: <stable@vger.kernel.org>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Now we only use the root parameter to print the root objectid in
a tracepoint. We can use the root parameter from the transaction
handle for that. It's also used to join the transaction with
async commits, so we remove the comment that it's just for checking.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_write_and_wait_marked_extents and btrfs_sync_log both call
btrfs_wait_marked_extents, which provides a core loop and then handles
errors differently based on whether it's it's a log root or not.
This means that btrfs_write_and_wait_marked_extents needs to take a root
because btrfs_wait_marked_extents requires one, even though it's only
used to determine whether the root is a log root. The log root code
won't ever call into the transaction commit code using a log root, so we
can factor out the core loop and provide the error handling appropriate
to each waiter in new routines. This allows us to eventually remove
the root argument from btrfs_commit_transaction, and as a result,
btrfs_end_transaction.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There are loads of functions in btrfs that accept a root parameter
but only use it to obtain an fs_info pointer. Let's convert those to
just accept an fs_info pointer directly.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
With the exception of the one case where btrfs_wait_cache_io is called
without a block group, it's called with the same arguments. The root
argument is only used in the special case, so let's factor out the core
and simplify the call in the normal case to require a trans, block group,
and path.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The extent-tree tracepoints all operate on the extent root, regardless of
which root is passed in. Let's just use the extent root objectid instead.
If it turns out that nobody is depending on the format of this tracepoint,
we can drop the root printing entirely.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This results in btrfs_assert_delayed_root_empty and
btrfs_destroy_delayed_inode taking an fs_info instead of a root.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In routines where someptr->fs_info is referenced multiple times, we
introduce a convenience variable. This makes the code considerably
more readable.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We track the node sizes per-root, but they never vary from the values
in the superblock. This patch messes with the 80-column style a bit,
but subsequent patches to factor out root->fs_info into a convenience
variable fix it up again.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The io_ctl->root member was only being used to access root->fs_info.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The root is never used. We substitute extent_root in for the
reada_find_extent call, since it's only ever used to obtain the node
size. This call site will be changed to use fs_info in a later patch.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The root member is never used except for obtaining an fs_info pointer.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Even though a separate root is passed in, we're still operating on the
extent root. Let's use that for the trace point.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_init_new_device only uses the root passed in via the ioctl to
start the transaction. Nothing else that happens is related to whatever
root the user used to initiate the ioctl. We can drop the root requirement
and just use fs_info->dev_root instead.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There are many functions that are always called with the same root
argument. Rather than passing the same root every time, we can
pass an fs_info pointer instead and have the function get the root
pointer itself.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There are 11 functions that accept a root parameter and immediately
overwrite it. We can pass those an fs_info pointer instead.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Ensure we release the NFS_LAYOUT_RETURN lock when we invalidate the
layout stateid, so that processes and RPC tasks that are waiting on
the layout return can continue.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
All callers are followed by the same boilerplate - "if it has returned
0, update nd->path/inode/seq - we are not following a symlink here".
Pull it into the function itself, renaming it into step_into().
Rename WALK_GET to WALK_FOLLOW, while we are at it - more descriptive
name.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
... turning the condition for put_link() in walk_component() into
"WALK_MORE not passed and depth is non-zero". Again, makes for
simpler arguments.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The function path_is_under() doesn't modify the paths pointed by its
arguments but only browse them. Constifying this pointers make a cleaner
interface to be used by (future) code which may only have access to
const struct path pointers (e.g. LSM hooks).
Signed-off-by: Mickaël Salaün <mic@digikod.net>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
With the current code it is possible to lock a mutex twice when
a subsequent reconnects are triggered. On the 1st reconnect we
reconnect sessions and tcons and then persistent file handles.
If the 2nd reconnect happens during the reconnecting of persistent
file handles then the following sequence of calls is observed:
cifs_reopen_file -> SMB2_open -> small_smb2_init -> smb2_reconnect
-> cifs_reopen_persistent_file_handles -> cifs_reopen_file (again!).
So, we are trying to acquire the same cfile->fh_mutex twice which
is wrong. Fix this by moving reconnecting of persistent handles to
the delayed work (smb2_reconnect_server) and submitting this work
every time we reconnect tcon in SMB2 commands handling codepath.
This can also lead to corruption of a temporary file list in
cifs_reopen_persistent_file_handles() because we can recursively
call this function twice.
Cc: Stable <stable@vger.kernel.org> # v4.9+
Signed-off-by: Pavel Shilovsky <pshilov@microsoft.com>
We can not unlock/lock cifs_tcp_ses_lock while walking through ses
and tcon lists because it can corrupt list iterator pointers and
a tcon structure can be released if we don't hold an extra reference.
Fix it by moving a reconnect process to a separate delayed work
and acquiring a reference to every tcon that needs to be reconnected.
Also do not send an echo request on newly established connections.
CC: Stable <stable@vger.kernel.org>
Signed-off-by: Pavel Shilovsky <pshilov@microsoft.com>
This reverts commit 1beba1b3a9.
The perpcu_counter doesn't provide atomicity in single core and consume more
DRAM. That incurs fs_mark test failure due to ENOMEM.
Cc: stable@vger.kernel.org # 4.7+
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
copy_from_iter_full(), copy_from_iter_full_nocache() and
csum_and_copy_from_iter_full() - counterparts of copy_from_iter()
et.al., advancing iterator only in case of successful full copy
and returning whether it had been successful or not.
Convert some obvious users. *NOTE* - do not blindly assume that
something is a good candidate for those unless you are sure that
not advancing iov_iter in failure case is the right thing in
this case. Anything that does short read/short write kind of
stuff (or is in a loop, etc.) is unlikely to be a good one.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
If maxBuf is not 0 but less than a size of SMB2 lock structure
we can end up with a memory corruption.
Cc: Stable <stable@vger.kernel.org>
Signed-off-by: Pavel Shilovsky <pshilov@microsoft.com>
Nick Piggin reported that the CRC overhead in an fsync heavy
workload was higher than expected on a Power8 machine. Part of this
was to do with the fact that the power8 CRC implementation is not
efficient for CRC lengths of less than 512 bytes, and so the way we
split the CRCs over the CRC field means a lot of the CRCs are
reduced to being less than than optimal size.
To optimise this, change the CRC update mechanism to zero the CRC
field first, and then compute the CRC in one pass over the buffer
and write the result back into the buffer. We can do this safely
because anything writing a CRC has exclusive access to the buffer
the CRC is being calculated over.
We leave the CRC verify code the same - it still splits the CRC
calculation - because we do not want read-only operations modifying
the underlying buffer. This is because read-only operations may not
have an exclusive access to the buffer guaranteed, and so temporary
modifications could leak out to to other processes accessing the
buffer concurrently.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Embedding a switch statement in every btree stats inc/add adds a lot
of code overhead to the core btree infrastructure paths. Stats are
supposed to be small and lightweight, but the btree stats have
become big and bloated as we've added more btrees. It needs fixing
because the reflink code will just add more overhead again.
Convert the v2 btree stats to arrays instead of independent
variables, and instead use the type to index the specific btree
array via an enum. This allows us to use array based indexing
to update the stats, rather than having to derefence variables
specific to the btree type.
If we then wrap the xfsstats structure in a union and place uint32_t
array beside it, and calculate the correct btree stats array base
array index when creating a btree cursor, we can easily access
entries in the stats structure without having to switch names based
on the btree type.
We then replace with the switch statement with a simple set of stats
wrapper macros, resulting in a significant simplification of the
btree stats code, and:
text data bss dec hex filename
48905 144 8 49057 bfa1 fs/xfs/libxfs/xfs_btree.o.old
36793 144 8 36945 9051 fs/xfs/libxfs/xfs_btree.o
it reduces the core btree infrastructure code size by close to 25%!
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
After various discussions on linux-fsdevel, it has been decided that it
is not necessary to cap the length of a dedupe request, and that
correctly-written userspace client programs will be able to absorb the
change. Therefore, remove the length clamping behavior.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
The on-disk field di_size is used to set i_size, which is a signed
integer of loff_t. If the high bit of di_size is set, we'll end up with
a negative i_size, which will cause all sorts of problems. Since the
VFS won't let us create a file with such length, we should catch them
here in the verifier too.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
We shouldn't assert if somehow we end up trying to add an attr fork to
an inode that apparently already has attr extents because this is an
indication of on-disk corruption. Instead, return an error code to
userspace.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
In xfs_dir3_data_read, we can encounter the situation where err == 0 and
*bpp == NULL if the given bno offset happens to be a hole; this leads to
a crash if we try to set the buffer type after the _da_read_buf call.
Holes can happen due to corrupt or malicious entries in the bmbt data,
so be a little more careful when we're handling buffers.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
When reading into memory all extents of a btree-format inode fork,
complain if the number of extents we find is not the same as the number
of extents reported in the inode core. This is needed to stop an IO
action from accessing the garbage areas of the in-core fork.
[dchinner: removed redundant assert]
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
When we're reading a btree block, make sure that what we retrieved
matches the owner and level; and has a plausible number of records.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
There is no such thing as a zero-level AG btree since even a single-node
zero-records btree has one level. Btree cursor constructors read
cur_nlevels straight from disk and then access things like
cur_bufs[cur_nlevels - 1] which is /really/ bad if cur_nlevels is zero!
Therefore, strengthen the verifiers to prevent this possibility.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
There are a handful of xattr functions which now return
nothing but zero. They can be made void, chased through calling
functions, and error handling etc can be removed.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
By inspection, xfs_bmap_trace_exlist isn't handling cow forks,
and will trace the data fork instead.
Fix this by setting state appropriately if whichfork
== XFS_COW_FORK.
()___()
< @ @ >
| |
{o_o}
(|)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
When xfs_bmap_trace_exlist called trace_xfs_extlist,
it sent in the "whichfork" var instead of the bmap "state"
as expected (even though state was already set up for this
purpose).
As a result, the xfs_bmap_class in tracing code used
"whichfork" not state in xfs_iext_state_to_fork(), and got
the wrong ifork pointer. It all goes downhill from
there, including an ASSERT when ifp_bytes is empty
by the time it reaches xfs_iext_get_ext():
XFS: Assertion failed: idx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
We've missed properly setting the buffer type for
an AGI transaction in 3 spots now, so just move it
into xfs_read_agi() and set it if we are in a transaction
to avoid the problem in the future.
This is similar to how it is done in i.e. the dir3
and attr3 read functions.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
xlog_recover_clear_agi_bucket didn't set the
type to XFS_BLFT_AGI_BUF, so we got a warning during log
replay (or an ASSERT on a debug build).
XFS (md0): Unknown buffer type 0!
XFS (md0): _xfs_buf_ioapply: no ops on block 0xaea8802/0x1
Fix this, as was done in f19b872b for 2 other locations
with the same problem.
cc: <stable@vger.kernel.org> # 3.10 to current
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
If the session has an error, then we want to start by recovering the
session, as any SEQUENCE we send is going to fail with a session
error.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
In the case where SEQUENCE receives a NFS4ERR_BADSESSION or
NFS4ERR_DEADSESSION error, we just want to report the session as needing
recovery, and then we want to retry the operation.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
When looking at whether or not our dcache is valid, we really don't care
about the general state of the directory attribute cache. Instead, we
we only care about the state of the change attribute.
This fixes a performance issue when the client is responsible for
changing the directory contents; a number of NFSv4 operations will
atomically update the directory change attribute, but may not return
all the other attributes.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
We should only care about checking the attributes if the page cache
is marked as dubious (using NFS_INO_REVAL_PAGECACHE) and the
NFS_INO_REVAL_FORCED flag is set.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
We should only care about checking the attributes if the page cache
is marked as dubious (using NFS_INO_REVAL_PAGECACHE) and the
NFS_INO_REVAL_FORCED flag is set.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Commit 7cbdb4a286 altered the autofs indirect mount expire to
not hold a spin lock during the expire check.
The direct mount expire needs the same treatment because to
make autofs expires namespace aware may_umount_tree() needs to
to use a similar method to may_umount() when checking if a mount
tree is in use.
This means may_umount_tree() will end up taking the namespace_sem
for the check so the autofs direct mount expire won't be allowed
to hold a spin lock over the check.
Signed-off-by: Ian Kent <raven@themaw.net>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Omar Sandoval <osandov@osandov.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Ian Kent <raven@themaw.net>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Omar Sandoval <osandov@osandov.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Now that path_has_submounts() has been added have_submounts() is no
longer used so remove it.
Link: http://lkml.kernel.org/r/20161011053428.27645.12310.stgit@pluto.themaw.net
Signed-off-by: Ian Kent <raven@themaw.net>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Omar Sandoval <osandov@osandov.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
If an automount mount is clone(2)ed into a file system that is propagation
private, when it later expires in the originating namespace, subsequent
calls to autofs ->d_automount() for that dentry in the original namespace
will return ELOOP until the mount is umounted in the cloned namespace.
Now that a struct path is available where needed use path_has_submounts()
instead of have_submounts() so we don't get false positives when checking
if a dentry is a mount point or contains mounts in the current namespace.
Link: http://lkml.kernel.org/r/20161011053423.27645.91233.stgit@pluto.themaw.net
Signed-off-by: Ian Kent <raven@themaw.net>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Omar Sandoval <osandov@osandov.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
If an automount mount is clone(2)ed into a file system that is propagation
private, when it later expires in the originating namespace, subsequent
calls to autofs ->d_automount() for that dentry in the original namespace
will return ELOOP until the mount is umounted in the cloned namespace.
Now that a struct path is available where needed use path_is_mountpoint()
instead of d_mountpoint() so we don't get false positives when checking if
a dentry is a mount point in the current namespace.
Link: http://lkml.kernel.org/r/20161011053418.27645.15241.stgit@pluto.themaw.net
Signed-off-by: Ian Kent <raven@themaw.net>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Omar Sandoval <osandov@osandov.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
In order to use the functions path_is_mountpoint() and path_has_submounts()
autofs needs to pass a struct path in several places.
Now change autofs4_wait() to take a struct path instead of a struct
dentry.
Link: http://lkml.kernel.org/r/20161011053413.27645.84666.stgit@pluto.themaw.net
Signed-off-by: Ian Kent <raven@themaw.net>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Omar Sandoval <osandov@osandov.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
In order to use the functions path_is_mountpoint() and path_has_submounts()
autofs needs to pass a struct path in several places.
Start by changing autofs4_expire_wait() and do_expire_wait() to take
a struct path instead of a struct dentry.
Link: http://lkml.kernel.org/r/20161011053408.27645.40091.stgit@pluto.themaw.net
Signed-off-by: Ian Kent <raven@themaw.net>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Omar Sandoval <osandov@osandov.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
d_mountpoint() can only be used reliably to establish if a dentry is
not mounted in any namespace. It isn't aware of the possibility there
may be multiple mounts using the given dentry, possibly in a different
namespace.
Add function, path_has_submounts(), that checks is a struct path contains
mounts (or is a mountpoint itself) to handle this case.
Link: http://lkml.kernel.org/r/20161011053403.27645.55242.stgit@pluto.themaw.net
Signed-off-by: Ian Kent <raven@themaw.net>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Omar Sandoval <osandov@osandov.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
d_mountpoint() can only be used reliably to establish if a dentry is
not mounted in any namespace. It isn't aware of the possibility there
may be multiple mounts using a given dentry that may be in a different
namespace.
Add helper functions, path_is_mountpoint(), that checks if a struct path
is a mountpoint for this case.
Link: http://lkml.kernel.org/r/20161011053358.27645.9729.stgit@pluto.themaw.net
Signed-off-by: Ian Kent <raven@themaw.net>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Omar Sandoval <osandov@osandov.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Before commit c3fe493ccd ('ext4: remove unneeded test in
ext4_alloc_file_blocks()') then it was possible for "depth" to be -1
but now, it's not possible that it is negative.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
Combination of data=ordered mode and journal_async_commit mount option
is invalid. However the check in parse_options() fails to detect the
case where we simply end up defaulting to data=ordered mode and we
detect the problem only on remount which triggers hard to understand
failure to remount the filesystem.
Fix the checking of mount options to take into account also the default
mode by moving the check somewhat later in the mount sequence.
Reported-by: Wolfgang Walter <linux@stwm.de>
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
mb_cache_entry_find_first() and mb_cache_entry_find_next() only return
cache entries with the 'e_reusable' bit set. This should be documented.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
mbcache used several different types to represent the number of entries
in the cache. For consistency within mbcache and with the shrinker API,
always use unsigned long.
This does not change behavior for current mbcache users (ext2 and ext4)
since they limit the entry count to a value which easily fits in an int.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
When mbcache is built as a module, any modules that use it (ext2 and/or
ext4) will depend on its symbols directly, incrementing its reference
count. Therefore, there is no need to do module_get/module_put.
Also note that since the module_get/module_put were in the mbcache
module itself, executing those lines of code was already dependent on
another reference to the mbcache module being held.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
Add the layout error payload to the flexfiles layoutreturn private
data, and set up the encoding mechanisms. This is a refactoring in
preparation for adding the layout iostats payload.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
mbcache can be a module that is loaded long after startup, when someone
asks to mount an ext2 or ext4 filesystem. Therefore it should not BUG()
if kmem_cache_create() fails, but rather just fail the module load.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
mbcache entries have an 'e_referenced' bit which users can set with
mb_cache_entry_touch() to indicate that an entry should be given another
pass through the LRU list before the shrinker can delete it. However,
mb_cache_shrink() actually would, when seeing an e_referenced entry at
the front of the list (the least-recently used end), place it right at
the front of the list again. The next iteration would then remove the
entry from the list and delete it. Consequently, e_referenced had
essentially no effect, so ext2/ext4 xattr blocks would sometimes not be
reused as often as expected.
Fix this by making the shrinker move e_referenced entries to the back of
the list rather than the front.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
Add a callback to allow the flexfiles layout driver to initialise the
layout private payload.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Couple conflicts resolved here:
1) In the MACB driver, a bug fix to properly initialize the
RX tail pointer properly overlapped with some changes
to support variable sized rings.
2) In XGBE we had a "CONFIG_PM" --> "CONFIG_PM_SLEEP" fix
overlapping with a reorganization of the driver to support
ACPI, OF, as well as PCI variants of the chip.
3) In 'net' we had several probe error path bug fixes to the
stmmac driver, meanwhile a lot of this code was cleaned up
and reorganized in 'net-next'.
4) The cls_flower classifier obtained a helper function in
'net-next' called __fl_delete() and this overlapped with
Daniel Borkamann's bug fix to use RCU for object destruction
in 'net'. It also overlapped with Jiri's change to guard
the rhashtable_remove_fast() call with a check against
tc_skip_sw().
5) In mlx4, a revert bug fix in 'net' overlapped with some
unrelated changes in 'net-next'.
6) In geneve, a stale header pointer after pskb_expand_head()
bug fix in 'net' overlapped with a large reorganization of
the same code in 'net-next'. Since the 'net-next' code no
longer had the bug in question, there was nothing to do
other than to simply take the 'net-next' hunks.
Signed-off-by: David S. Miller <davem@davemloft.net>
Cleanup to allow layout drivers to attach private data to layoutreturn,
and manage the data.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
For the autofs module to be able to reliably check if a dentry is a
mountpoint in a multiple namespace environment the ->d_manage() dentry
operation will need to take a path argument instead of a dentry.
Link: http://lkml.kernel.org/r/20161011053352.27645.83962.stgit@pluto.themaw.net
Signed-off-by: Ian Kent <raven@themaw.net>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Omar Sandoval <osandov@osandov.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
On a filesystem with no journal, a symlink longer than about 32
characters (exact length depending on padding for encryption) could not
be followed or read immediately after being created in an encrypted
directory. This happened because when the symlink data went through the
delayed allocation path instead of the journaling path, the symlink was
incorrectly detected as a "fast" symlink rather than a "slow" symlink
until its data was written out.
To fix this, disable delayed allocation for symlinks, since there is
no benefit for delayed allocation anyway.
Reported-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
If there have been no reads or writes to a given mirror since the last
layoutstats update, then don't resend the same data.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
If the use called stat() on an 'ls -l' workload, and the attribute
cache was successfully revalidate by READDIRPLUS, then we want to
report that back so that the readdir code continues to use
readdirplus.
Reviewed-by: Benjamin Coddington <bcodding@redhat.com>
Tested-by: Benjamin Coddington <bcodding@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
There is little point in setting NFS_INO_ADVISE_RDPLUS in nfs_lookup and
nfs_lookup_revalidate() unless a process is actually doing readdir on the
parent directory.
Furthermore, there is little point in using readdirplus if we're trying
to revalidate a negative dentry.
Reviewed-by: Benjamin Coddington <bcodding@redhat.com>
Tested-by: Benjamin Coddington <bcodding@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Ben Coddington reports that commit 311324ad17, by adding the function
nfs_dir_mapping_need_revalidate() that checks page cache validity on
each call to nfs_readdir() causes a performance regression when
the directory is being modified.
If the directory is changing while we're iterating through the directory,
POSIX does not require us to invalidate the page cache unless the user
calls rewinddir(). However, we still do want to ensure that we use
readdirplus in order to avoid a load of stat() calls when the user
is doing an 'ls -l' workload.
The fix should be to invalidate the page cache immediately when we're
setting the NFS_INO_ADVISE_RDPLUS bit.
Reported-by: Benjamin Coddington <bcodding@redhat.com>
Fixes: 311324ad17 ("NFS: Be more aggressive in using readdirplus...")
Reviewed-by: Benjamin Coddington <bcodding@redhat.com>
Tested-by: Benjamin Coddington <bcodding@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
It now has only one field and is only used in one structure.
So replaced it in that structure by the field it contains.
Signed-off-by: NeilBrown <neilb@suse.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
A process can have two possible lock owner for a given open file:
a per-process Posix lock owner and a per-open-file flock owner
Use both of these when searching for a suitable stateid to use.
With this patch, READ/WRITE requests will use the correct stateid
if a flock lock is active.
Signed-off-by: NeilBrown <neilb@suse.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
The only time that a lock_context is not immediately available is in
setattr, and now that it has an open_context, it can easily find one
with nfs_get_lock_context.
This removes the need for the on-stack nfs_lockowner.
This change is preparation for correctly support flock stateids.
Signed-off-by: NeilBrown <neilb@suse.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
The open_context can always lead directly to the state, and is always easily
available, so this is a straightforward change.
Doing this makes more information available to _nfs4_do_setattr() for use
in the next patch.
Signed-off-by: NeilBrown <neilb@suse.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
An open file description (struct file) in a given process can be
associated with two different lock owners.
It can have a Posix lock owner which will be different in each process
that has a fd on the file.
It can have a Flock owner which will be the same in all processes.
When searching for a lock stateid to use, we need to consider both of these
owners
So add a new "flock_owner" to the "nfs_open_context" (of which there
is one for each open file description).
This flock_owner does not need to be reference-counted as there is a
1-1 relation between 'struct file' and nfs open contexts,
and it will never be part of a list of contexts. So there is no need
for a 'flock_context' - just the owner is enough.
The io_count included in the (Posix) lock_context provides no
guarantee that all read-aheads that could use the state have
completed, so not supporting it for flock locks in not a serious
problem. Synchronization between flock and read-ahead can be added
later if needed.
When creating an open_context for a non-openning create call, we don't have
a 'struct file' to pass in, so the lock context gets initialized with
a NULL owner, but this will never be used.
The flock_owner is not used at all in this patch, that will come later.
Acked-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: NeilBrown <neilb@suse.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
this field is not used in any important way and probably should
have been removed by
Commit: 8003d3c4aa ("nfs4: treat lock owners as opaque values")
which removed the pid argument from nfs4_get_lock_state.
Except in unusual and uninteresting cases, two threads with the same
->tgid will have the same ->files pointer, so keeping them both
for comparison brings no benefit.
Acked-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: NeilBrown <neilb@suse.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
This parameter hasn't been used since 2a009ec9 (Linux 3.13-rc3), so
let's remove it from this function.
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
This parameter hasn't been used since f8407299 (Linux 3.11-rc2), so
let's remove it from this function and callers.
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
It's possible that two different servers can return the same (clientid,
verifier) pair purely by coincidence. Both are 64-bit values, but
depending on the server implementation, they can be highly predictable
and collisions may be quite likely, especially when there are lots of
servers.
So, check for this case. If the clientid and verifier both match, then
we actually know they *can't* be the same server, since a new
SETCLIENTID to an already-known server should have changed the verifier.
This helps fix a bug that could cause the client to mount a filesystem
from the wrong server.
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Tested-by: Yongcheng Yang <yoyang@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Ensure that the layout state bits are synced when we cache a layout
segment for layoutreturn using an appropriate call to
pnfs_set_plh_return_info.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
We need to honour the NFS_LAYOUT_RETURN_REQUESTED bit regardless of
whether or not there are layout segments pending.
Furthermore, we should ensure that we leave the plh_return_segs list
empty.
This patch fixes a memory leak of the layout segments on plh_return_segs.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
When the layout state is invalidated, then so is the layout segment
state, and hence we do need to clean up the state bits.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
If we cannot grab the inode or superblock, then we cannot pin the
layout header, and so we cannot send a layoutreturn as part of an
async delegreturn call. In this case, we currently end up sending
an extra layoutreturn after the delegreturn. Since the layout was
implicitly returned by the delegreturn, that just gets a BAD_STATEID.
The fix is to simply complete the return-on-close immediately.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Amend the pnfs return on close helper functions to enable sending the
layoutreturn op in CLOSE/DELEGRETURN. This closes a potential race between
CLOSE/DELEGRETURN and parallel OPEN calls to the same file, and allows the
client and the server to agree on whether or not there is an outstanding
layout.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Add XDR encoding for the layoutreturn op, and storage for the layoutreturn
arguments to the DELEGRETURN compound.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Add XDR encoding for the layoutreturn op, and storage for the layoutreturn
arguments to the CLOSE compound.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
The layoutreturn call will take care of invalidating the layout segments
once the call is successful.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
There is no change to the value of NFS_LAYOUT_RETURN, so we should
not be waking up the RPC call.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Fix a potential race with CB_LAYOUTRECALL in which the server recalls the
remaining layout segments while our LAYOUTRETURN is still in transit.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
We may want to process and transmit layout stat information for the
layout segments that are being returned, so we should defer freeing
them until after the layoutreturn has completed.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Instead of grabbing the layout, we want to get the inode so that we
can reduce races between layoutget and layoutrecall when the server
does not support call referring.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Both pnfs.c and the flexfiles code have their own versions of the
range intersection testing, and the "end_offset" helper.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
We must put the task to sleep while holding the inode->i_lock in order
to ensure atomicity with the test for NFS_LAYOUT_RETURN.
Fixes: 500d701f33 ("NFS41: make close wait for layoutreturn")
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
If there is an I/O error, we should not call LAYOUTGET until the
LAYOUTRETURN that reports the error is complete.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Cc: stable@vger.kernel.org # v4.8+
If the server sends us a completely new stateid, and the client thinks
it already holds a layout, then force a retry of the LAYOUTGET after
invalidating the existing layout in order to avoid corruption due to
races.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
We must ensure that we don't schedule a layoutreturn if the layout stateid
has been marked as invalid.
Fixes: 2a59a04116 ("pNFS: Fix pnfs_set_layout_stateid() to clear...")
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Cc: stable@vger.kernel.org # v4.8+
If we no longer hold any layout segments, we're normally expected to
consider the layout stateid to be invalid. However we cannot assume this
if we're about to, or in the process of sending a layoutreturn.
Fixes: 334a8f3711 ("pNFS: Don't forget the layout stateid if...")
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Cc: stable@vger.kernel.org # v4.8+
We must not call nfs_pageio_init_read() on a new nfs_pageio_descriptor
while holding a reference to a layout segment, as that can deadlock
pnfs_update_layout().
Fixes: d67ae825a5 ("pnfs/flexfiles: Add the FlexFile Layout Driver")
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Cc: stable@vger.kernel.org # v4.0+
When initializing a freshly created slot for the calllback channel,
the seq_nr needs to be 0, not 1. Otherwise validate_seqid
and nfs4_slot_wait_on_seqid get confused and believe that the
mpty slot corresponds to a previously sent reply.
Signed-off-by: Fred Isaman <fred.isaman@gmail.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
The NFS_INO_REVAL_FORCED flag needs to be set if we just got a delegation,
and we see that there might still be some ambiguity as to whether or not
our attribute or data cache are valid.
In practice, this means that a call to nfs_check_inode_attributes() will
have noticed a discrepancy between cached attributes and measured ones,
so let's move the setting of NFS_INO_REVAL_FORCED to there.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
If holding a delegation, we do not need to ask the server to return
close-to-open cache consistency attributes as part of the CLOSE
compound.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
If we're not closing the file completely, there is no need to request
close-to-open attributes.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
We don't need to ask for the change attribute when returning a delegation
or recovering from a server reboot, and it could actually cause us to
obtain an incorrect value if we're using a pNFS flavour that requires
LAYOUTCOMMIT.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
If we're reclaiming state after a reboot, or as part of returning a
delegation, we don't need to check access modes again.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Ralf Spenneberg reported that he hit a kernel crash when mounting a
modified ext4 image. And it turns out that kernel crashed when
calculating fs overhead (ext4_calculate_overhead()), this is because
the image has very large s_first_meta_bg (debug code shows it's
842150400), and ext4 overruns the memory in count_overhead() when
setting bitmap buffer, which is PAGE_SIZE.
ext4_calculate_overhead():
buf = get_zeroed_page(GFP_NOFS); <=== PAGE_SIZE buffer
blks = count_overhead(sb, i, buf);
count_overhead():
for (j = ext4_bg_num_gdb(sb, grp); j > 0; j--) { <=== j = 842150400
ext4_set_bit(EXT4_B2C(sbi, s++), buf); <=== buffer overrun
count++;
}
This can be reproduced easily for me by this script:
#!/bin/bash
rm -f fs.img
mkdir -p /mnt/ext4
fallocate -l 16M fs.img
mke2fs -t ext4 -O bigalloc,meta_bg,^resize_inode -F fs.img
debugfs -w -R "ssv first_meta_bg 842150400" fs.img
mount -o loop fs.img /mnt/ext4
Fix it by validating s_first_meta_bg first at mount time, and
refusing to mount if its value exceeds the largest possible meta_bg
number.
Reported-by: Ralf Spenneberg <ralf@os-t.de>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
It was possible for an xattr value to have a very large size, which
would then pass validation on 32-bit architectures due to a pointer
wraparound. Fix this by validating the size in a way which avoids
pointer wraparound.
It was also possible that a value's size would fit in the available
space but its padded size would not. This would cause an out-of-bounds
memory write in ext4_xattr_set_entry when replacing the xattr value.
For example, if an xattr value of unpadded size 253 bytes went until the
very end of the inode or block, then using setxattr(2) to replace this
xattr's value with 256 bytes would cause a write to the 3 bytes past the
end of the inode or buffer, and the new xattr value would be incorrectly
truncated. Fix this by requiring that the padded size fit in the
available space rather than the unpadded size.
This patch shouldn't have any noticeable effect on
non-corrupted/non-malicious filesystems.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
With i_extra_isize equal to or close to the available space, it was
possible for us to read past the end of the inode when trying to detect
or validate in-inode xattrs. Fix this by checking for the needed extra
space first.
This patch shouldn't have any noticeable effect on
non-corrupted/non-malicious filesystems.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
i_extra_isize not divisible by 4 is problematic for several reasons:
- It causes the in-inode xattr space to be misaligned, but the xattr
header and entries are not declared __packed to express this
possibility. This may cause poor performance or incorrect code
generation on some platforms.
- When validating the xattr entries we can read past the end of the
inode if the size available for xattrs is not a multiple of 4.
- It allows the nonsensical i_extra_isize=1, which doesn't even leave
enough room for i_extra_isize itself.
Therefore, update ext4_iget() to consider i_extra_isize not divisible by
4 to be an error, like the case where i_extra_isize is too large.
This also matches the rule recently added to e2fsck for determining
whether an inode has valid i_extra_isize.
This patch shouldn't have any noticeable effect on
non-corrupted/non-malicious filesystems, since the size of ext4_inode
has always been a multiple of 4.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
Pull overlayfs fix from Miklos Szeredi:
"This fixes a regression introduced in 4.8"
* 'overlayfs-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/vfs:
ovl: fix d_real() for stacked fs
On a CONFIG_EXT4_FS_ENCRYPTION=n kernel, the ioctls to get and set
encryption policies were disabled but EXT4_IOC_GET_ENCRYPTION_PWSALT was
not. But there's no good reason to expose the pwsalt ioctl if the
kernel doesn't support encryption. The pwsalt ioctl was also disabled
pre-4.8 (via ext4_sb_has_crypto() previously returning 0 when encryption
was disabled by config) and seems to have been enabled by mistake when
ext4 encryption was refactored to use fs/crypto/. So let's disable it
again.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
ext4_sb_has_crypto() just called through to ext4_has_feature_encrypt(),
and all callers except one were already using the latter. So remove it
and switch its one caller to ext4_has_feature_encrypt().
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
We've fixed the race condition problem in calculating ext4 checksum
value in commit b47820edd1 ("ext4: avoid modifying checksum fields
directly during checksum veficationon"). However, by this change,
when calculating the checksum value of inode whose i_extra_size is
less than 4, we couldn't calculate the checksum value in a proper way.
This problem was found and reported by Nix, Thank you.
Reported-by: Nix <nix@esperi.org.uk>
Signed-off-by: Daeho Jeong <daeho.jeong@samsung.com>
Signed-off-by: Youngjin Gil <youngjin.gil@samsung.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Warn when a page is dirtied without buffers (as that will likely lead to
a crash in ext4_writepages()) or when it gets newly dirtied without the
page being locked (as there is nothing that prevents buffers to get
stripped just before calling set_page_dirty() under memory pressure).
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
New mount option "snapshot=<time>" to allow mounting an earlier
version of the remote volume (if such a snapshot exists on
the server).
Note that eventually specifying a snapshot time of 1 will allow
the user to mount the oldest snapshot. A subsequent patch
add the processing for that and another for actually specifying
the "time warp" create context on SMB2/SMB3 open.
Check to make sure SMB2 negotiated, and ensure that
we use a different tcon if mount same share twice
but with different snaphshot times
Signed-off-by: Steve French <smfrench@gmail.com>
Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com>
The ER records are printed without explicit log level presuming line
continuation until "\n". After the commit 4bcc595ccd (printk:
reinstate KERN_CONT for printing continuation lines), the ER records are
printed a character per line.
Adding KERN_CONT to appropriate printk statements restores the printout
behavior.
Signed-off-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
If a log tree has a layout like the following:
leaf N:
...
item 240 key (282 DIR_LOG_ITEM 0) itemoff 8189 itemsize 8
dir log end 1275809046
leaf N + 1:
item 0 key (282 DIR_LOG_ITEM 3936149215) itemoff 16275 itemsize 8
dir log end 18446744073709551615
...
When we pass the value 1275809046 + 1 as the parameter start_ret to the
function tree-log.c:find_dir_range() (done by replay_dir_deletes()), we
end up with path->slots[0] having the value 239 (points to the last item
of leaf N, item 240). Because the dir log item in that position has an
offset value smaller than *start_ret (1275809046 + 1) we need to move on
to the next leaf, however the logic for that is wrong since it compares
the current slot to the number of items in the leaf, which is smaller
and therefore we don't lookup for the next leaf but instead we set the
slot to point to an item that does not exist, at slot 240, and we later
operate on that slot which has unexpected content or in the worst case
can result in an invalid memory access (accessing beyond the last page
of leaf N's extent buffer).
So fix the logic that checks when we need to lookup at the next leaf
by first incrementing the slot and only after to check if that slot
is beyond the last item of the current leaf.
Signed-off-by: Robbie Ko <robbieko@synology.com>
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Fixes: e02119d5a7 (Btrfs: Add a write ahead tree log to optimize synchronous operations)
Cc: stable@vger.kernel.org # 2.6.29+
Signed-off-by: Filipe Manana <fdmanana@suse.com>
[Modified changelog for clarity and correctness]
The hole punching can result in adding new leafs (and as a consequence
new nodes) to the tree because when we find file extent items that span
beyond the hole range we may end up not deleting them (just adjusting
them, reducing their range by reducing their length or increasing their
offset field) and add new file extent items representing holes.
So after splitting a leaf (therefore creating a new one) to insert a new
file extent item representing a hole, a new node might be added to each
level of the tree in the worst case scenario (since there's a new key
and every parent node was full).
For example if a file has an extent item representing the range 0 to 64Mb
and we punch a hole in the range 1Mb to 20Mb, the existing extent item is
duplicated and one of the copies is adjusted to represent the range 0 to
1Mb, the other copy adjusted to represent the range 20Mb to 64Mb, and a
new file extent item representing a hole in the range 1Mb to 20Mb is
inserted.
Fix this by using btrfs_calc_trans_metadata_size() instead of
btrfs_calc_trunc_metadata_size(), so that enough metadata space is
reserved for the worst possible case.
Signed-off-by: Robbie Ko <robbieko@synology.com>
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
[Modified changelog for clarity and correctness]
This issue was found when I tried to delete a heavily reflinked file,
when deleting such files, other transaction operation will not have a
chance to make progress, for example, start_transaction() will blocked
in wait_current_trans(root) for long time, sometimes it even triggers
soft lockups, and the time taken to delete such heavily reflinked file
is also very large, often hundreds of seconds. Using perf top, it reports
that:
PerfTop: 7416 irqs/sec kernel:99.8% exact: 0.0% [4000Hz cpu-clock], (all, 4 CPUs)
---------------------------------------------------------------------------------------
84.37% [btrfs] [k] __btrfs_run_delayed_refs.constprop.80
11.02% [kernel] [k] delay_tsc
0.79% [kernel] [k] _raw_spin_unlock_irq
0.78% [kernel] [k] _raw_spin_unlock_irqrestore
0.45% [kernel] [k] do_raw_spin_lock
0.18% [kernel] [k] __slab_alloc
It seems __btrfs_run_delayed_refs() took most cpu time, after some debug
work, I found it's select_delayed_ref() causing this issue, for a delayed
head, in our case, it'll be full of BTRFS_DROP_DELAYED_REF nodes, but
select_delayed_ref() will firstly try to iterate node list to find
BTRFS_ADD_DELAYED_REF nodes, obviously it's a disaster in this case, and
waste much time.
To fix this issue, we introduce a new ref_add_list in struct btrfs_delayed_ref_head,
then in select_delayed_ref(), if this list is not empty, we can directly use
nodes in this list. With this patch, it just took about 10~15 seconds to
delte the same file. Now using perf top, it reports that:
PerfTop: 2734 irqs/sec kernel:99.5% exact: 0.0% [4000Hz cpu-clock], (all, 4 CPUs)
----------------------------------------------------------------------------------------
20.74% [kernel] [k] _raw_spin_unlock_irqrestore
16.33% [kernel] [k] __slab_alloc
5.41% [kernel] [k] lock_acquired
4.42% [kernel] [k] lock_acquire
4.05% [kernel] [k] lock_release
3.37% [kernel] [k] _raw_spin_unlock_irq
For normal files, this patch also gives help, at least we do not need to
iterate whole list to found BTRFS_ADD_DELAYED_REF nodes.
Signed-off-by: Wang Xiaoguang <wangxg.fnst@cn.fujitsu.com>
Reviewed-by: Liu Bo <bo.li.liu@oracle.com>
Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Richard Weinberger