-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCZJU4WgAKCRCRxhvAZXjc
oofvAQDs9RJwQUyWHJmQA+tWz5cUE5DviVWCwwul5dQRRCqgaQEA2OIO0gPFaVoq
1OYOeLyUjl/cpS8e3u4uJtw34jttdQA=
=AwcR
-----END PGP SIGNATURE-----
Merge tag 'v6.5/vfs.file' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull vfs file handling updates from Christian Brauner:
"This contains Amir's work to fix a long-standing problem where an
unprivileged overlayfs mount can be used to avoid fanotify permission
events that were requested for an inode or superblock on the
underlying filesystem.
Some background about files opened in overlayfs. If a file is opened
in overlayfs @file->f_path will refer to a "fake" path. What this
means is that while @file->f_inode will refer to inode of the
underlying layer, @file->f_path refers to an overlayfs
{dentry,vfsmount} pair. The reasons for doing this are out of scope
here but it is the reason why the vfs has been providing the
open_with_fake_path() helper for overlayfs for very long time now. So
nothing new here.
This is for sure not very elegant and everyone including the overlayfs
maintainers agree. Improving this significantly would involve more
fragile and potentially rather invasive changes.
In various codepaths access to the path of the underlying filesystem
is needed for such hybrid file. The best example is fsnotify where
this becomes security relevant. Passing the overlayfs
@file->f_path->dentry will cause fsnotify to skip generating fsnotify
events registered on the underlying inode or superblock.
To fix this we extend the vfs provided open_with_fake_path() concept
for overlayfs to create a backing file container that holds the real
path and to expose a helper that can be used by relevant callers to
get access to the path of the underlying filesystem through the new
file_real_path() helper. This pattern is similar to what we do in
d_real() and d_real_inode().
The first beneficiary is fsnotify and fixes the security sensitive
problem mentioned above.
There's a couple of nice cleanups included as well.
Over time, the old open_with_fake_path() helper added specifically for
overlayfs a long time ago started to get used in other places such as
cachefiles. Even though cachefiles have nothing to do with hybrid
files.
The only reason cachefiles used that concept was that files opened
with open_with_fake_path() aren't charged against the caller's open
file limit by raising FMODE_NOACCOUNT. It's just mere coincidence that
both overlayfs and cachefiles need to ensure to not overcharge the
caller for their internal open calls.
So this work disentangles FMODE_NOACCOUNT use cases and backing file
use-cases by adding the FMODE_BACKING flag which indicates that the
file can be used to retrieve the backing file of another filesystem.
(Fyi, Jens will be sending you a really nice cleanup from Christoph
that gets rid of 3 FMODE_* flags otherwise this would be the last
fmode_t bit we'd be using.)
So now overlayfs becomes the sole user of the renamed
open_with_fake_path() helper which is now named backing_file_open().
For internal kernel users such as cachefiles that are only interested
in FMODE_NOACCOUNT but not in FMODE_BACKING we add a new
kernel_file_open() helper which opens a file without being charged
against the caller's open file limit. All new helpers are properly
documented and clearly annotated to mention their special uses.
We also rename vfs_tmpfile_open() to kernel_tmpfile_open() to clearly
distinguish it from vfs_tmpfile() and align it the other kernel_*()
internal helpers"
* tag 'v6.5/vfs.file' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
ovl: enable fsnotify events on underlying real files
fs: use backing_file container for internal files with "fake" f_path
fs: move kmem_cache_zalloc() into alloc_empty_file*() helpers
fs: use a helper for opening kernel internal files
fs: rename {vfs,kernel}_tmpfile_open()
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCZJU4NwAKCRCRxhvAZXjc
ordqAP0RmZEkUA5p98pK+0FEFIsS2S8qChh6YHQHP+hF606FGgEAivb3UPRm9p58
kRb5yK0/oXDUxGv7A+x+SIMVbcRyLgw=
=pi6N
-----END PGP SIGNATURE-----
Merge tag 'v6.5/vfs.rename.locking' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull vfs rename locking updates from Christian Brauner:
"This contains the work from Jan to fix problems with cross-directory
renames originally reported in [1].
To quickly sum it up some filesystems (so far we know at least about
ext4, udf, f2fs, ocfs2, likely also reiserfs, gfs2 and others) need to
lock the directory when it is being renamed into another directory.
This is because we need to update the parent pointer in the directory
in that case and if that races with other operations on the directory,
in particular a conversion from one directory format into another, bad
things can happen.
So far we've done the locking in the filesystem code but recently
Darrick pointed out in [2] that the RENAME_EXCHANGE case was missing.
That one is particularly nasty because RENAME_EXCHANGE can arbitrarily
mix regular files and directories and proper lock ordering is not
achievable in the filesystems alone.
This patch set adds locking into vfs_rename() so that not only parent
directories but also moved inodes, regardless of whether they are
directories or not, are locked when calling into the filesystem.
This means establishing a locking order for unrelated directories. New
helpers are added for this purpose and our documentation is updated to
cover this in detail.
The locking is now actually easier to follow as we now always lock
source and target. We've always locked the target independent of
whether it was a directory or file and we've always locked source if
it was a regular file. The exact details for why this came about can
be found in [3] and [4]"
Link: https://lore.kernel.org/all/20230117123735.un7wbamlbdihninm@quack3 [1]
Link: https://lore.kernel.org/all/20230517045836.GA11594@frogsfrogsfrogs [2]
Link: https://lore.kernel.org/all/20230526-schrebergarten-vortag-9cd89694517e@brauner [3]
Link: https://lore.kernel.org/all/20230530-seenotrettung-allrad-44f4b00139d4@brauner [4]
* tag 'v6.5/vfs.rename.locking' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
fs: Restrict lock_two_nondirectories() to non-directory inodes
fs: Lock moved directories
fs: Establish locking order for unrelated directories
Revert "f2fs: fix potential corruption when moving a directory"
Revert "udf: Protect rename against modification of moved directory"
ext4: Remove ext4 locking of moved directory
Provide helpers to set and clear sb->s_readonly_remount including
appropriate memory barriers. Also use this opportunity to document what
the barriers pair with and why they are needed.
Suggested-by: Dave Chinner <david@fromorbit.com>
Signed-off-by: Jan Kara <jack@suse.cz>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Message-Id: <20230620112832.5158-1-jack@suse.cz>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Overlayfs uses open_with_fake_path() to allocate internal kernel files,
with a "fake" path - whose f_path is not on the same fs as f_inode.
Allocate a container struct backing_file for those internal files, that
is used to hold the "fake" ovl path along with the real path.
backing_file_real_path() can be used to access the stored real path.
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Message-Id: <20230615112229.2143178-5-amir73il@gmail.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Currently the locking order of inode locks for directories that are not
in ancestor relationship is not defined because all operations that
needed to lock two directories like this were serialized by
sb->s_vfs_rename_mutex. However some filesystems need to lock two
subdirectories for RENAME_EXCHANGE operations and for this we need the
locking order established even for two tree-unrelated directories.
Provide a helper function lock_two_inodes() that establishes lock
ordering for any two inodes and use it in lock_two_directories().
CC: stable@vger.kernel.org
Signed-off-by: Jan Kara <jack@suse.cz>
Message-Id: <20230601105830.13168-4-jack@suse.cz>
Signed-off-by: Christian Brauner <brauner@kernel.org>
-----BEGIN PGP SIGNATURE-----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=DweL
-----END PGP SIGNATURE-----
Merge tag '6.4-rc-ksmbd-server-fixes' of git://git.samba.org/ksmbd
Pull ksmbd server updates from Steve French:
- SMB3.1.1 negotiate context fixes and cleanup
- new lock_rename_child VFS helper
- ksmbd fix to avoid unlink race and to use the new VFS helper to avoid
rename race
* tag '6.4-rc-ksmbd-server-fixes' of git://git.samba.org/ksmbd:
ksmbd: fix racy issue from using ->d_parent and ->d_name
ksmbd: remove unused compression negotiate ctx packing
ksmbd: avoid duplicate negotiate ctx offset increments
ksmbd: set NegotiateContextCount once instead of every inc
fs: introduce lock_rename_child() helper
ksmbd: remove internal.h include
Since vfs_path_lookup is exported, It should not be internal.
Move vfs_path_lookup prototype in internal.h to linux/namei.h.
Suggested-by: Al Viro <viro@zeniv.linux.org.uk>
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
We've aligned setgid behavior over multiple kernel releases. The details
can be found in the following two merge messages:
cf619f8919 ("Merge tag 'fs.ovl.setgid.v6.2')
426b4ca2d6 ("Merge tag 'fs.setgid.v6.0')
Consistent setgid stripping behavior is now encapsulated in the
setattr_should_drop_sgid() helper which is used by all filesystems that
strip setgid bits outside of vfs proper. Switch nfs to rely on this
helper as well. Without this patch the setgid stripping tests in
xfstests will fail.
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Message-Id: <20230313-fs-nfs-setgid-v2-1-9a59f436cfc0@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
-----BEGIN PGP SIGNATURE-----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=IU45
-----END PGP SIGNATURE-----
Merge tag 'for-6.3/dio-2023-02-16' of git://git.kernel.dk/linux
Pull legacy dio update from Jens Axboe:
"We only have a few file systems that use the old dio code, make them
select it rather than build it unconditionally"
* tag 'for-6.3/dio-2023-02-16' of git://git.kernel.dk/linux:
fs: build the legacy direct I/O code conditionally
fs: move sb_init_dio_done_wq out of direct-io.c
sb_init_dio_done_wq is also used by the iomap code, so move it to
super.c in preparation for building direct-io.c conditionally.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230125065839.191256-2-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Now that we converted everything to just rely on struct mnt_idmap move it all
into a separate file. This ensure that no code can poke around in struct
mnt_idmap without any dedicated helpers and makes it easier to extend it in the
future. Filesystems will now not be able to conflate mount and filesystem
idmappings as they are two distinct types and require distinct helpers that
cannot be used interchangeably. We are now also able to extend struct mnt_idmap
as we see fit.
Acked-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Convert to struct mnt_idmap.
Last cycle we merged the necessary infrastructure in
256c8aed2b ("fs: introduce dedicated idmap type for mounts").
This is just the conversion to struct mnt_idmap.
Currently we still pass around the plain namespace that was attached to a
mount. This is in general pretty convenient but it makes it easy to
conflate namespaces that are relevant on the filesystem with namespaces
that are relevent on the mount level. Especially for non-vfs developers
without detailed knowledge in this area this can be a potential source for
bugs.
Once the conversion to struct mnt_idmap is done all helpers down to the
really low-level helpers will take a struct mnt_idmap argument instead of
two namespace arguments. This way it becomes impossible to conflate the two
eliminating the possibility of any bugs. All of the vfs and all filesystems
only operate on struct mnt_idmap.
Acked-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Convert to struct mnt_idmap.
Last cycle we merged the necessary infrastructure in
256c8aed2b ("fs: introduce dedicated idmap type for mounts").
This is just the conversion to struct mnt_idmap.
Currently we still pass around the plain namespace that was attached to a
mount. This is in general pretty convenient but it makes it easy to
conflate namespaces that are relevant on the filesystem with namespaces
that are relevent on the mount level. Especially for non-vfs developers
without detailed knowledge in this area this can be a potential source for
bugs.
Once the conversion to struct mnt_idmap is done all helpers down to the
really low-level helpers will take a struct mnt_idmap argument instead of
two namespace arguments. This way it becomes impossible to conflate the two
eliminating the possibility of any bugs. All of the vfs and all filesystems
only operate on struct mnt_idmap.
Acked-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Convert an initial portion to rely on struct mnt_idmap by converting the
high level xattr helpers.
Reviewed-by: Seth Forshee (DigitalOcean) <sforshee@kernel.org>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
In previous patches we built a new posix api solely around get and set
inode operations. Now that we have all the pieces in place we can switch
the system calls and the vfs over to only rely on this api when
interacting with posix acls. This finally removes all type unsafety and
type conversion issues explained in detail in [1] that we aim to get rid
of.
With the new posix acl api we immediately translate into an appropriate
kernel internal struct posix_acl format both when getting and setting
posix acls. This is a stark contrast to before were we hacked unsafe raw
values into the uapi struct that was stored in a void pointer relying
and having filesystems and security modules hack around in the uapi
struct as well.
Link: https://lore.kernel.org/all/20220801145520.1532837-1-brauner@kernel.org [1]
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Split out the generic checks whether an inode allows writing xattrs. Since
security.* and system.* xattrs don't have any restrictions and we're going
to split out posix acls into a dedicated api we will use this helper to
check whether we can write posix acls.
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Currently setgid stripping in file_remove_privs()'s should_remove_suid()
helper is inconsistent with other parts of the vfs. Specifically, it only
raises ATTR_KILL_SGID if the inode is S_ISGID and S_IXGRP but not if the
inode isn't in the caller's groups and the caller isn't privileged over the
inode although we require this already in setattr_prepare() and
setattr_copy() and so all filesystem implement this requirement implicitly
because they have to use setattr_{prepare,copy}() anyway.
But the inconsistency shows up in setgid stripping bugs for overlayfs in
xfstests (e.g., generic/673, generic/683, generic/685, generic/686,
generic/687). For example, we test whether suid and setgid stripping works
correctly when performing various write-like operations as an unprivileged
user (fallocate, reflink, write, etc.):
echo "Test 1 - qa_user, non-exec file $verb"
setup_testfile
chmod a+rws $junk_file
commit_and_check "$qa_user" "$verb" 64k 64k
The test basically creates a file with 6666 permissions. While the file has
the S_ISUID and S_ISGID bits set it does not have the S_IXGRP set. On a
regular filesystem like xfs what will happen is:
sys_fallocate()
-> vfs_fallocate()
-> xfs_file_fallocate()
-> file_modified()
-> __file_remove_privs()
-> dentry_needs_remove_privs()
-> should_remove_suid()
-> __remove_privs()
newattrs.ia_valid = ATTR_FORCE | kill;
-> notify_change()
-> setattr_copy()
In should_remove_suid() we can see that ATTR_KILL_SUID is raised
unconditionally because the file in the test has S_ISUID set.
But we also see that ATTR_KILL_SGID won't be set because while the file
is S_ISGID it is not S_IXGRP (see above) which is a condition for
ATTR_KILL_SGID being raised.
So by the time we call notify_change() we have attr->ia_valid set to
ATTR_KILL_SUID | ATTR_FORCE. Now notify_change() sees that
ATTR_KILL_SUID is set and does:
ia_valid = attr->ia_valid |= ATTR_MODE
attr->ia_mode = (inode->i_mode & ~S_ISUID);
which means that when we call setattr_copy() later we will definitely
update inode->i_mode. Note that attr->ia_mode still contains S_ISGID.
Now we call into the filesystem's ->setattr() inode operation which will
end up calling setattr_copy(). Since ATTR_MODE is set we will hit:
if (ia_valid & ATTR_MODE) {
umode_t mode = attr->ia_mode;
vfsgid_t vfsgid = i_gid_into_vfsgid(mnt_userns, inode);
if (!vfsgid_in_group_p(vfsgid) &&
!capable_wrt_inode_uidgid(mnt_userns, inode, CAP_FSETID))
mode &= ~S_ISGID;
inode->i_mode = mode;
}
and since the caller in the test is neither capable nor in the group of the
inode the S_ISGID bit is stripped.
But assume the file isn't suid then ATTR_KILL_SUID won't be raised which
has the consequence that neither the setgid nor the suid bits are stripped
even though it should be stripped because the inode isn't in the caller's
groups and the caller isn't privileged over the inode.
If overlayfs is in the mix things become a bit more complicated and the bug
shows up more clearly. When e.g., ovl_setattr() is hit from
ovl_fallocate()'s call to file_remove_privs() then ATTR_KILL_SUID and
ATTR_KILL_SGID might be raised but because the check in notify_change() is
questioning the ATTR_KILL_SGID flag again by requiring S_IXGRP for it to be
stripped the S_ISGID bit isn't removed even though it should be stripped:
sys_fallocate()
-> vfs_fallocate()
-> ovl_fallocate()
-> file_remove_privs()
-> dentry_needs_remove_privs()
-> should_remove_suid()
-> __remove_privs()
newattrs.ia_valid = ATTR_FORCE | kill;
-> notify_change()
-> ovl_setattr()
// TAKE ON MOUNTER'S CREDS
-> ovl_do_notify_change()
-> notify_change()
// GIVE UP MOUNTER'S CREDS
// TAKE ON MOUNTER'S CREDS
-> vfs_fallocate()
-> xfs_file_fallocate()
-> file_modified()
-> __file_remove_privs()
-> dentry_needs_remove_privs()
-> should_remove_suid()
-> __remove_privs()
newattrs.ia_valid = attr_force | kill;
-> notify_change()
The fix for all of this is to make file_remove_privs()'s
should_remove_suid() helper to perform the same checks as we already
require in setattr_prepare() and setattr_copy() and have notify_change()
not pointlessly requiring S_IXGRP again. It doesn't make any sense in the
first place because the caller must calculate the flags via
should_remove_suid() anyway which would raise ATTR_KILL_SGID.
While we're at it we move should_remove_suid() from inode.c to attr.c
where it belongs with the rest of the iattr helpers. Especially since it
returns ATTR_KILL_S{G,U}ID flags. We also rename it to
setattr_should_drop_suidgid() to better reflect that it indicates both
setuid and setgid bit removal and also that it returns attr flags.
Running xfstests with this doesn't report any regressions. We should really
try and use consistent checks.
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
The current setgid stripping logic during write and ownership change
operations is inconsistent and strewn over multiple places. In order to
consolidate it and make more consistent we'll add a new helper
setattr_should_drop_sgid(). The function retains the old behavior where
we remove the S_ISGID bit unconditionally when S_IXGRP is set but also
when it isn't set and the caller is neither in the group of the inode
nor privileged over the inode.
We will use this helper both in write operation permission removal such
as file_remove_privs() as well as in ownership change operations.
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
In setattr_{copy,prepare}() we need to perform the same permission
checks to determine whether we need to drop the setgid bit or not.
Instead of open-coding it twice add a simple helper the encapsulates the
logic. We will reuse this helpers to make dropping the setgid bit during
write operations more consistent in a follow up patch.
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
passing kmap_local_page() result to __kernel_write() is unsafe -
random ->write_iter() might (and 9p one does) get unhappy when
passed ITER_KVEC with pointer that came from kmap_local_page().
Fix by providing a variant of __kernel_write() that takes an iov_iter
from caller (__kernel_write() becomes a trivial wrapper) and adding
dump_emit_page() that parallels dump_emit(), except that instead of
__kernel_write() it uses __kernel_write_iter() with ITER_BVEC source.
Fixes: 3159ed5779 "fs/coredump: use kmap_local_page()"
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Thread A trying to acquire a write lease checks the value of i_readcount
and i_writecount in check_conflicting_open() to verify that its own fd
is the only fd referencing the file.
Thread B trying to open the file for read will call break_lease() in
do_dentry_open() before incrementing i_readcount, which leaves a small
window where thread A can acquire the write lease and then thread B
completes the open of the file for read without breaking the write lease
that was acquired by thread A.
Fix this race by incrementing i_readcount before checking for existing
leases, same as the case with i_writecount.
Use a helper put_file_access() to decrement i_readcount or i_writecount
in do_dentry_open() and __fput().
Fixes: 387e3746d0 ("locks: eliminate false positive conflicts for write lease")
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The fix is usermode_driver.c one - once you've done kern_mount(), you
must kern_unmount(); simple mntput() will end up with a leak. Several
failure exits in there messed up that way... In practice you won't
hit those particular failure exits without fault injection, though.
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQQqUNBr3gm4hGXdBJlZ7Krx/gZQ6wUCYpvrWQAKCRBZ7Krx/gZQ
6z29AP9EZVSyIvnwXleehpa2mEZhsp+KAKgV/ENaKHMn7jiH0wD/bfgnhxIDNuc5
108E2R5RWEYTynW5k7nnP5PsTsMq5Qc=
=b3Wc
-----END PGP SIGNATURE-----
Merge tag 'pull-18-rc1-work.mount' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull mount handling updates from Al Viro:
"Cleanups (and one fix) around struct mount handling.
The fix is usermode_driver.c one - once you've done kern_mount(), you
must kern_unmount(); simple mntput() will end up with a leak. Several
failure exits in there messed up that way... In practice you won't hit
those particular failure exits without fault injection, though"
* tag 'pull-18-rc1-work.mount' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
move mount-related externs from fs.h to mount.h
blob_to_mnt(): kern_unmount() is needed to undo kern_mount()
m->mnt_root->d_inode->i_sb is a weird way to spell m->mnt_sb...
linux/mount.h: trim includes
uninline may_mount() and don't opencode it in fspick(2)/fsopen(2)
Currently we have 3 primitives for removing an opened file from descriptor
table - pick_file(), __close_fd_get_file() and close_fd_get_file(). Their
calling conventions are rather odd and there's a code duplication for no
good reason. They can be unified -
1) have __range_close() cap max_fd in the very beginning; that way
we don't need separate way for pick_file() to report being past the end
of descriptor table.
2) make {__,}close_fd_get_file() return file (or NULL) directly, rather
than returning it via struct file ** argument. Don't bother with
(bogus) return value - nobody wants that -ENOENT.
3) make pick_file() return NULL on unopened descriptor - the only caller
that used to care about the distinction between descriptor past the end
of descriptor table and finding NULL in descriptor table doesn't give
a damn after (1).
4) lift ->files_lock out of pick_file()
That actually simplifies the callers, as well as the primitives themselves.
Code duplication is also gone...
Reviewed-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
This splits off do_getxattr function from the getxattr function. This will
allow io_uring to call it from its io worker.
Signed-off-by: Stefan Roesch <shr@fb.com>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Link: https://lore.kernel.org/r/20220323154420.3301504-3-shr@fb.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This splits of the setup part of the function setxattr in its own
dedicated function called setxattr_copy. In addition it also exposes a new
function called do_setxattr for making the setxattr call.
This makes it possible to call these two functions from io_uring in the
processing of an xattr request.
Signed-off-by: Stefan Roesch <shr@fb.com>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Link: https://lore.kernel.org/r/20220323154420.3301504-2-shr@fb.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Pull vfs updates from Al Viro:
"Assorted bits and pieces"
* 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
aio: drop needless assignment in aio_read()
clean overflow checks in count_mounts() a bit
seq_file: fix NULL pointer arithmetic warning
uml/x86: use x86 load_unaligned_zeropad()
asm/user.h: killed unused macros
constify struct path argument of finish_automount()/do_add_mount()
fs: Remove FIXME comment in generic_write_checks()
-----BEGIN PGP SIGNATURE-----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=YARd
-----END PGP SIGNATURE-----
Merge tag 'for-5.18-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux
Pull btrfs updates from David Sterba:
"This contains feature updates, performance improvements, preparatory
and core work and some related VFS updates:
Features:
- encoded read/write ioctls, allows user space to read or write raw
data directly to extents (now compressed, encrypted in the future),
will be used by send/receive v2 where it saves processing time
- zoned mode now works with metadata DUP (the mkfs.btrfs default)
- error message header updates:
- print error state: transaction abort, other error, log tree
errors
- print transient filesystem state: remount, device replace,
ignored checksum verifications
- tree-checker: verify the transaction id of the to-be-written dirty
extent buffer
Performance improvements for fsync:
- directory logging speedups (up to -90% run time)
- avoid logging all directory changes during renames (up to -60% run
time)
- avoid inode logging during rename and link when possible (up to
-60% run time)
- prepare extents to be logged before locking a log tree path
(throughput +7%)
- stop copying old file extents when doing a full fsync()
- improved logging of old extents after truncate
Core, fixes:
- improved stale device identification by dev_t and not just path
(for devices that are behind other layers like device mapper)
- continued extent tree v2 preparatory work
- disable features that won't work yet
- add wrappers and abstractions for new tree roots
- improved error handling
- add super block write annotations around background block group
reclaim
- fix device scanning messages potentially accessing stale pointer
- cleanups and refactoring
VFS:
- allow reflinks/deduplication from two different mounts of the same
filesystem
- export and add helpers for read/write range verification, for the
encoded ioctls"
* tag 'for-5.18-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux: (98 commits)
btrfs: zoned: put block group after final usage
btrfs: don't access possibly stale fs_info data in device_list_add
btrfs: add lockdep_assert_held to need_preemptive_reclaim
btrfs: verify the tranisd of the to-be-written dirty extent buffer
btrfs: unify the error handling of btrfs_read_buffer()
btrfs: unify the error handling pattern for read_tree_block()
btrfs: factor out do_free_extent_accounting helper
btrfs: remove last_ref from the extent freeing code
btrfs: add a alloc_reserved_extent helper
btrfs: remove BUG_ON(ret) in alloc_reserved_tree_block
btrfs: add and use helper for unlinking inode during log replay
btrfs: extend locking to all space_info members accesses
btrfs: zoned: mark relocation as writing
fs: allow cross-vfsmount reflink/dedupe
btrfs: remove the cross file system checks from remap
btrfs: pass btrfs_fs_info to btrfs_recover_relocation
btrfs: pass btrfs_fs_info for deleting snapshots and cleaner
btrfs: add filesystems state details to error messages
btrfs: deal with unexpected extent type during reflinking
btrfs: fix unexpected error path when reflinking an inline extent
...
I'm adding btrfs ioctls to read and write compressed data, and rather
than duplicating the checks in rw_verify_area(), let's just export it.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
One of the key architectual tenets is to keep the parameters for
io-uring stable. After the call has been submitted, its value can
be changed. Unfortunaltely this is not the case for the current statx
implementation.
IO-Uring change:
This changes replaces the const char * filename pointer in the io_statx
structure with a struct filename *. In addition it also creates the
filename object during the prepare phase.
With this change, the opcode also needs to invoke cleanup, so the
filename object gets freed after processing the request.
fs change:
This replaces the const char* __user filename parameter in the two
functions do_statx and vfs_statx with a struct filename *. In addition
to be able to correctly construct a filename object a new helper
function getname_statx_lookup_flags is introduced. The function makes
sure that do_statx and vfs_statx is invoked with the correct lookup flags.
Signed-off-by: Stefan Roesch <shr@fb.com>
Link: https://lore.kernel.org/r/20220225185326.1373304-2-shr@fb.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
There are no plans to convert buffer_head infrastructure to use large
folios, but __block_write_begin_int() is called from iomap, and it's
more convenient and less error-prone if we pass in a folio from iomap.
It also has a nice saving of almost 200 bytes of code from removing
repeated calls to compound_head().
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Merge more updates from Andrew Morton:
"87 patches.
Subsystems affected by this patch series: mm (pagecache and hugetlb),
procfs, misc, MAINTAINERS, lib, checkpatch, binfmt, kallsyms, ramfs,
init, codafs, nilfs2, hfs, crash_dump, signals, seq_file, fork,
sysvfs, kcov, gdb, resource, selftests, and ipc"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (87 commits)
ipc/ipc_sysctl.c: remove fallback for !CONFIG_PROC_SYSCTL
ipc: check checkpoint_restore_ns_capable() to modify C/R proc files
selftests/kselftest/runner/run_one(): allow running non-executable files
virtio-mem: disallow mapping virtio-mem memory via /dev/mem
kernel/resource: disallow access to exclusive system RAM regions
kernel/resource: clean up and optimize iomem_is_exclusive()
scripts/gdb: handle split debug for vmlinux
kcov: replace local_irq_save() with a local_lock_t
kcov: avoid enable+disable interrupts if !in_task()
kcov: allocate per-CPU memory on the relevant node
Documentation/kcov: define `ip' in the example
Documentation/kcov: include types.h in the example
sysv: use BUILD_BUG_ON instead of runtime check
kernel/fork.c: unshare(): use swap() to make code cleaner
seq_file: fix passing wrong private data
seq_file: move seq_escape() to a header
signal: remove duplicate include in signal.h
crash_dump: remove duplicate include in crash_dump.h
crash_dump: fix boolreturn.cocci warning
hfs/hfsplus: use WARN_ON for sanity check
...
Historically (pre-2.5), the inode shrinker used to reclaim only empty
inodes and skip over those that still contained page cache. This caused
problems on highmem hosts: struct inode could put fill lowmem zones
before the cache was getting reclaimed in the highmem zones.
To address this, the inode shrinker started to strip page cache to
facilitate reclaiming lowmem. However, this comes with its own set of
problems: the shrinkers may drop actively used page cache just because
the inodes are not currently open or dirty - think working with a large
git tree. It further doesn't respect cgroup memory protection settings
and can cause priority inversions between containers.
Nowadays, the page cache also holds non-resident info for evicted cache
pages in order to detect refaults. We've come to rely heavily on this
data inside reclaim for protecting the cache workingset and driving swap
behavior. We also use it to quantify and report workload health through
psi. The latter in turn is used for fleet health monitoring, as well as
driving automated memory sizing of workloads and containers, proactive
reclaim and memory offloading schemes.
The consequences of dropping page cache prematurely is that we're seeing
subtle and not-so-subtle failures in all of the above-mentioned
scenarios, with the workload generally entering unexpected thrashing
states while losing the ability to reliably detect it.
To fix this on non-highmem systems at least, going back to rotating
inodes on the LRU isn't feasible. We've tried (commit a76cf1a474
("mm: don't reclaim inodes with many attached pages")) and failed
(commit 69056ee6a8 ("Revert "mm: don't reclaim inodes with many
attached pages"")).
The issue is mostly that shrinker pools attract pressure based on their
size, and when objects get skipped the shrinkers remember this as
deferred reclaim work. This accumulates excessive pressure on the
remaining inodes, and we can quickly eat into heavily used ones, or
dirty ones that require IO to reclaim, when there potentially is plenty
of cold, clean cache around still.
Instead, this patch keeps populated inodes off the inode LRU in the
first place - just like an open file or dirty state would. An otherwise
clean and unused inode then gets queued when the last cache entry
disappears. This solves the problem without reintroducing the reclaim
issues, and generally is a bit more scalable than having to wade through
potentially hundreds of thousands of busy inodes.
Locking is a bit tricky because the locks protecting the inode state
(i_lock) and the inode LRU (lru_list.lock) don't nest inside the
irq-safe page cache lock (i_pages.xa_lock). Page cache deletions are
serialized through i_lock, taken before the i_pages lock, to make sure
depopulated inodes are queued reliably. Additions may race with
deletions, but we'll check again in the shrinker. If additions race
with the shrinker itself, we're protected by the i_lock: if find_inode()
or iput() win, the shrinker will bail on the elevated i_count or
I_REFERENCED; if the shrinker wins and goes ahead with the inode, it
will set I_FREEING and inhibit further igets(), which will cause the
other side to create a new instance of the inode instead.
Link: https://lkml.kernel.org/r/20210614211904.14420-4-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Get rid of the indirections and just provide a sync_bdevs
helper for the generic sync code.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20211019062530.2174626-8-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Instead offer a new sync_blockdev_nowait helper for the !wait case.
This new helper is exported as it will grow modular callers in a bit.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20211019062530.2174626-3-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
- Simplify the bio_end_page usage in the buffered IO code.
- Support reading inline data at nonzero offsets for erofs.
- Fix some typos and bad grammar.
- Convert kmap_atomic usage in the inline data read path.
- Add some extra inline data input checking.
- Fix a memory corruption bug stemming from iomap_swapfile_activate
trying to activate more pages than mm was expecting.
- Pass errnos through the page writeback code so that writeback errors
are reported correctly instead of being munged to EIO.
- Replace iomap_apply with a open-coded iterator loops to reduce the
number of indirect calls by a third to a half.
- Refactor the fsdax code to use iomap iterators instead of the
open-coded iomap_apply code that it had before.
- Format file range iomap tracepoint data in hexadecimal and
standardize the names used in the pretty-print string.
-----BEGIN PGP SIGNATURE-----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=iVns
-----END PGP SIGNATURE-----
Merge tag 'iomap-5.15-merge-4' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux
Pull iomap updates from Darrick Wong:
"The most notable externally visible change for this cycle is the
addition of support for reads to inline tail fragments of files, which
was requested by the erofs developers; and a correction for a kernel
memory corruption bug if the sysadmin tries to activate a swapfile
with more pages than the swapfile header suggests.
We also now report writeback completion errors to the file mapping
correctly, instead of munging all errors into EIO.
Internally, the bulk of the changes are Christoph's patchset to reduce
the indirect function call count by a third to a half by converting
iomap iteration from a loop pattern to a generator/consumer pattern.
As an added bonus, fsdax no longer open-codes iomap apply loops.
Summary:
- Simplify the bio_end_page usage in the buffered IO code.
- Support reading inline data at nonzero offsets for erofs.
- Fix some typos and bad grammar.
- Convert kmap_atomic usage in the inline data read path.
- Add some extra inline data input checking.
- Fix a memory corruption bug stemming from iomap_swapfile_activate
trying to activate more pages than mm was expecting.
- Pass errnos through the page writeback code so that writeback
errors are reported correctly instead of being munged to EIO.
- Replace iomap_apply with a open-coded iterator loops to reduce the
number of indirect calls by a third to a half.
- Refactor the fsdax code to use iomap iterators instead of the
open-coded iomap_apply code that it had before.
- Format file range iomap tracepoint data in hexadecimal and
standardize the names used in the pretty-print string"
* tag 'iomap-5.15-merge-4' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux: (41 commits)
iomap: standardize tracepoint formatting and storage
mm/swap: consider max pages in iomap_swapfile_add_extent
iomap: move loop control code to iter.c
iomap: constify iomap_iter_srcmap
fsdax: switch the fault handlers to use iomap_iter
fsdax: factor out a dax_fault_actor() helper
fsdax: factor out helpers to simplify the dax fault code
iomap: rework unshare flag
iomap: pass an iomap_iter to various buffered I/O helpers
iomap: remove iomap_apply
fsdax: switch dax_iomap_rw to use iomap_iter
iomap: switch iomap_swapfile_activate to use iomap_iter
iomap: switch iomap_seek_data to use iomap_iter
iomap: switch iomap_seek_hole to use iomap_iter
iomap: switch iomap_bmap to use iomap_iter
iomap: switch iomap_fiemap to use iomap_iter
iomap: switch __iomap_dio_rw to use iomap_iter
iomap: switch iomap_page_mkwrite to use iomap_iter
iomap: switch iomap_zero_range to use iomap_iter
iomap: switch iomap_file_unshare to use iomap_iter
...
IORING_OP_LINKAT behaves like linkat(2) and takes the same flags and
arguments.
In some internal places 'hardlink' is used instead of 'link' to avoid
confusion with the SQE links. Name 'link' conflicts with the existing
'link' member of io_kiocb.
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Suggested-by: Christian Brauner <christian.brauner@ubuntu.com>
Link: https://lore.kernel.org/io-uring/20210514145259.wtl4xcsp52woi6ab@wittgenstein/
Signed-off-by: Dmitry Kadashev <dkadashev@gmail.com>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Link: https://lore.kernel.org/r/20210708063447.3556403-12-dkadashev@gmail.com
[axboe: add splice_fd_in check]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Update the following to return int rather than long, for uniformity with
the rest of the do_* helpers in namei.c:
* do_rmdir()
* do_unlinkat()
* do_mkdirat()
* do_mknodat()
* do_symlinkat()
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christian Brauner <christian.brauner@ubuntu.com>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Link: https://lore.kernel.org/io-uring/20210514143202.dmzfcgz5hnauy7ze@wittgenstein/
Signed-off-by: Dmitry Kadashev <dkadashev@gmail.com>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Link: https://lore.kernel.org/r/20210708063447.3556403-9-dkadashev@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Pass in the struct filename pointers instead of the user string, and
update the three callers to do the same. This is heavily based on
commit dbea8d345177 ("fs: make do_renameat2() take struct filename").
This behaves like do_unlinkat() and do_renameat2().
Cc: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Dmitry Kadashev <dkadashev@gmail.com>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Link: https://lore.kernel.org/r/20210708063447.3556403-4-dkadashev@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
__block_write_begin_int never modifies the passed in iomap, so mark it
const.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Richard reported sporadic (roughly one in 10 or so) null dereferences and
other strange behaviour for a set of automated LTP tests. Things like:
BUG: kernel NULL pointer dereference, address: 0000000000000008
#PF: supervisor read access in kernel mode
#PF: error_code(0x0000) - not-present page
PGD 0 P4D 0
Oops: 0000 [#1] PREEMPT SMP PTI
CPU: 0 PID: 1516 Comm: umount Not tainted 5.10.0-yocto-standard #1
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-48-gd9c812dda519-prebuilt.qemu.org 04/01/2014
RIP: 0010:kernfs_sop_show_path+0x1b/0x60
...or these others:
RIP: 0010:do_mkdirat+0x6a/0xf0
RIP: 0010:d_alloc_parallel+0x98/0x510
RIP: 0010:do_readlinkat+0x86/0x120
There were other less common instances of some kind of a general scribble
but the common theme was mount and cgroup and a dubious dentry triggering
the NULL dereference. I was only able to reproduce it under qemu by
replicating Richard's setup as closely as possible - I never did get it
to happen on bare metal, even while keeping everything else the same.
In commit 71d883c37e ("cgroup_do_mount(): massage calling conventions")
we see this as a part of the overall change:
--------------
struct cgroup_subsys *ss;
- struct dentry *dentry;
[...]
- dentry = cgroup_do_mount(&cgroup_fs_type, fc->sb_flags, root,
- CGROUP_SUPER_MAGIC, ns);
[...]
- if (percpu_ref_is_dying(&root->cgrp.self.refcnt)) {
- struct super_block *sb = dentry->d_sb;
- dput(dentry);
+ ret = cgroup_do_mount(fc, CGROUP_SUPER_MAGIC, ns);
+ if (!ret && percpu_ref_is_dying(&root->cgrp.self.refcnt)) {
+ struct super_block *sb = fc->root->d_sb;
+ dput(fc->root);
deactivate_locked_super(sb);
msleep(10);
return restart_syscall();
}
--------------
In changing from the local "*dentry" variable to using fc->root, we now
export/leave that dentry pointer in the file context after doing the dput()
in the unlikely "is_dying" case. With LTP doing a crazy amount of back to
back mount/unmount [testcases/bin/cgroup_regression_5_1.sh] the unlikely
becomes slightly likely and then bad things happen.
A fix would be to not leave the stale reference in fc->root as follows:
--------------
dput(fc->root);
+ fc->root = NULL;
deactivate_locked_super(sb);
--------------
...but then we are just open-coding a duplicate of fc_drop_locked() so we
simply use that instead.
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Tejun Heo <tj@kernel.org>
Cc: Zefan Li <lizefan.x@bytedance.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: stable@vger.kernel.org # v5.1+
Reported-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fixes: 71d883c37e ("cgroup_do_mount(): massage calling conventions")
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
