Include fs/open.c in filesystems/api-summary.rst to provide its
exported APIs.
Suggested-by: Matthew Wilcox <willy@infradead.org>
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Message-ID: <20260104204530.518206-1-rdunlap@infradead.org>
vsprintf.c uses a mix of the `kernel.kptr_restrict` sysctl and the
`hash_pointers` boot param to control pointer hashing. But that wasn't
possible to tell without looking at the source code.
They have a different focus and purpose. To avoid wasting the time of
users trying to use one instead of the other, simply have them reference
each other in the Documentation.
Signed-off-by: Marc Herbert <marc.herbert@linux.intel.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Message-ID: <20260107-doc-hash-ptr-v2-1-cb4c161218d7@linux.intel.com>
Christian Brauner <brauner@kernel.org> says:
When creating containers the setup usually involves using CLONE_NEWNS
via clone3() or unshare(). This copies the caller's complete mount
namespace. The runtime will also assemble a new rootfs and then use
pivot_root() to switch the old mount tree with the new rootfs. Afterward
it will recursively umount the old mount tree thereby getting rid of all
mounts.
On a basic system here where the mount table isn't particularly large
this still copies about 30 mounts. Copying all of these mounts only to
get rid of them later is pretty wasteful.
This is exacerbated if intermediary mount namespaces are used that only
exist for a very short amount of time and are immediately destroyed
again causing a ton of mounts to be copied and destroyed needlessly.
With a large mount table and a system where thousands or ten-thousands
of namespaces are spawned in parallel this quickly becomes a bottleneck
increasing contention on the semaphore.
Extend open_tree() with a new OPEN_TREE_NAMESPACE flag. Similar to
OPEN_TREE_CLONE only the indicated mount tree is copied. Instead of
returning a file descriptor referring to that mount tree
OPEN_TREE_NAMESPACE will cause open_tree() to return a file descriptor
to a new mount namespace. In that new mount namespace the copied mount
tree has been mounted on top of a copy of the real rootfs.
The caller can setns() into that mount namespace and perform any
additionally setup such as move_mount()ing detached mounts in there.
This allows OPEN_TREE_NAMESPACE to function as a combined
unshare(CLONE_NEWNS) and pivot_root().
A caller may for example choose to create an extremely minimal rootfs:
fd_mntns = open_tree(-EBADF, "/var/lib/containers/wootwoot", OPEN_TREE_NAMESPACE);
This will create a mount namespace where "wootwoot" has become the
rootfs mounted on top of the real rootfs. The caller can now setns()
into this new mount namespace and assemble additional mounts.
This also works with user namespaces:
unshare(CLONE_NEWUSER);
fd_mntns = open_tree(-EBADF, "/var/lib/containers/wootwoot", OPEN_TREE_NAMESPACE);
which creates a new mount namespace owned by the earlier created user
namespace with "wootwoot" as the rootfs mounted on top of the real
rootfs.
This will scale a lot better when creating tons of mount namespaces and
will allow to get rid of a lot of unnecessary mount and umount cycles.
It also allows to create mount namespaces without needing to spawn
throwaway helper processes.
* patches from https://patch.msgid.link/20251229-work-empty-namespace-v1-0-bfb24c7b061f@kernel.org:
selftests/open_tree: add OPEN_TREE_NAMESPACE tests
mount: add OPEN_TREE_NAMESPACE
Link: https://patch.msgid.link/20251229-work-empty-namespace-v1-0-bfb24c7b061f@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>
When creating containers the setup usually involves using CLONE_NEWNS
via clone3() or unshare(). This copies the caller's complete mount
namespace. The runtime will also assemble a new rootfs and then use
pivot_root() to switch the old mount tree with the new rootfs. Afterward
it will recursively umount the old mount tree thereby getting rid of all
mounts.
On a basic system here where the mount table isn't particularly large
this still copies about 30 mounts. Copying all of these mounts only to
get rid of them later is pretty wasteful.
This is exacerbated if intermediary mount namespaces are used that only
exist for a very short amount of time and are immediately destroyed
again causing a ton of mounts to be copied and destroyed needlessly.
With a large mount table and a system where thousands or ten-thousands
of containers are spawned in parallel this quickly becomes a bottleneck
increasing contention on the semaphore.
Extend open_tree() with a new OPEN_TREE_NAMESPACE flag. Similar to
OPEN_TREE_CLONE only the indicated mount tree is copied. Instead of
returning a file descriptor referring to that mount tree
OPEN_TREE_NAMESPACE will cause open_tree() to return a file descriptor
to a new mount namespace. In that new mount namespace the copied mount
tree has been mounted on top of a copy of the real rootfs.
The caller can setns() into that mount namespace and perform any
additionally required setup such as move_mount() detached mounts in
there.
This allows OPEN_TREE_NAMESPACE to function as a combined
unshare(CLONE_NEWNS) and pivot_root().
A caller may for example choose to create an extremely minimal rootfs:
fd_mntns = open_tree(-EBADF, "/var/lib/containers/wootwoot", OPEN_TREE_NAMESPACE);
This will create a mount namespace where "wootwoot" has become the
rootfs mounted on top of the real rootfs. The caller can now setns()
into this new mount namespace and assemble additional mounts.
This also works with user namespaces:
unshare(CLONE_NEWUSER);
fd_mntns = open_tree(-EBADF, "/var/lib/containers/wootwoot", OPEN_TREE_NAMESPACE);
which creates a new mount namespace owned by the earlier created user
namespace with "wootwoot" as the rootfs mounted on top of the real
rootfs.
Link: https://patch.msgid.link/20251229-work-empty-namespace-v1-1-bfb24c7b061f@kernel.org
Tested-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Aleksa Sarai <cyphar@cyphar.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Suggested-by: Christian Brauner <brauner@kernel.org>
Suggested-by: Aleksa Sarai <cyphar@cyphar.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Define man page font as monospace and bold, i.e. the same as what is
used for .code and <pre>.
Signed-off-by: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Message-ID: <20260113113612.315748-4-pvorel@suse.cz>
Removing :manpage: from non-existing man pages (xyzzy(2), xyzzyat(2),
fxyzzy(3) in adding-syscalls.rst, including translations) prevent
adding link to nonexisting man pages when using manpages_url in next
commit.
While at it, add also missing '(2)' in sp_SP translation.
Reviewed-by: Alejandro Colomar <alx@kernel.org>
Signed-off-by: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Message-ID: <20260113113612.315748-2-pvorel@suse.cz>
Miklos Szeredi <mszeredi@redhat.com> says:
This mini series fixes issues with the stale dentry cleanup patches
added in this cycle. In particular commit ab84ad5973 ("fuse: new work
queue to periodically invalidate expired dentries") allowed a race
resulting in UAF.
* patches from https://patch.msgid.link/20260114145344.468856-1-mszeredi@redhat.com:
vfs: document d_dispose_if_unused()
fuse: shrink once after all buckets have been scanned
fuse: clean up fuse_dentry_tree_work()
fuse: add need_resched() before unlocking bucket
fuse: make sure dentry is evicted if stale
fuse: fix race when disposing stale dentries
Link: https://patch.msgid.link/20260114145344.468856-1-mszeredi@redhat.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
Add a warning about the danger of using this function without proper
locking preventing eviction.
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Link: https://patch.msgid.link/20260114145344.468856-7-mszeredi@redhat.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
In fuse_dentry_tree_work() move the shrink_dentry_list() out from the loop.
Suggested-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Link: https://patch.msgid.link/20260114145344.468856-6-mszeredi@redhat.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
- Change time_after64() time_before64(), since the latter is exclusively
used in this file to compare dentry/inode timeout with current time.
- Move the break statement from the else branch to the if branch, reducing
indentation.
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Link: https://patch.msgid.link/20260114145344.468856-5-mszeredi@redhat.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
In fuse_dentry_tree_work() no need to unlock/lock dentry_hash[i].lock on
each iteration.
Suggested-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Link: https://patch.msgid.link/20260114145344.468856-4-mszeredi@redhat.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
d_dispose_if_unused() may find the dentry with a positive refcount, in
which case it won't be put on the dispose list even though it has already
timed out.
"Reinstall" the d_delete() callback, which was optimized out in
fuse_dentry_settime(). This will result in the dentry being evicted as
soon as the refcount hits zero.
Fixes: ab84ad5973 ("fuse: new work queue to periodically invalidate expired dentries")
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Link: https://patch.msgid.link/20260114145344.468856-3-mszeredi@redhat.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
In fuse_dentry_tree_work() just before d_dispose_if_unused() the dentry
could get evicted, resulting in UAF.
Move unlocking dentry_hash[i].lock to after the dispose. To do this,
fuse_dentry_tree_del_node() needs to be moved from fuse_dentry_prune() to
fuse_dentry_release() to prevent an ABBA deadlock.
The lock ordering becomes:
-> dentry_bucket.lock
-> dentry.d_lock
Reported-by: Al Viro <viro@zeniv.linux.org.uk>
Closes: https://lore.kernel.org/all/20251206014242.GO1712166@ZenIV/
Fixes: ab84ad5973 ("fuse: new work queue to periodically invalidate expired dentries")
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Link: https://patch.msgid.link/20260114145344.468856-2-mszeredi@redhat.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
The driver binding documentation (still) mentions that "When a driver is
attached to a device, the device is inserted into the driver's list of
devices.".
While it is true that the driver-core keeps track of all the devices
that are attached to a driver, this is purely for internal purposes
(i.e. it is an implementation detail) and has no relevance for user
facing documentation.
In fact, it is even misleading, since it could be read as if it were
valid for driver implementations to keep track of all the devices bound
to it.
Instead, drivers operate on a per-device basis, with a separate
per-device instance created when the driver is bound to a device.
Hence, remove the mention of a driver's list of devices and instead add
some documentation of the relationship between drivers and devices.
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Message-ID: <20260115215718.6405-1-dakr@kernel.org>
Fix various typos and grammatical errors across documentation files:
- Fix missing preposition 'in' in process/changes.rst
- Correct 'result by' to 'result from' in admin-guide/README.rst
- Fix 'before hand' to 'beforehand' in cgroup-v1/hugetlb.rst
- Correct 'allows to limit' to 'allows limiting' in hugetlb.rst,
cgroup-v2.rst, and kconfig-language.rst
- Fix 'needs precisely know' to 'needs to precisely know'
- Correct 'overcommited' to 'overcommitted' in hugetlb.rst
- Fix subject-verb agreement: 'never causes' to 'never cause'
- Fix 'there is enough' to 'there are enough' in hugetlb.rst
- Fix 'metadatas' to 'metadata' in filesystems/erofs.rst
- Fix 'hardwares' to 'hardware' in scsi/ChangeLog.sym53c8xx
Signed-off-by: Nauman Sabir <officialnaumansabir@gmail.com>
Acked-by: Tejun Heo <tj@kernel.org>
Message-ID: <20260115230110.7734-1-officialnaumansabir@gmail.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Puranjay Mohan says:
====================
bpf: Fix linked register tracking
This patch fixes the linked register tracking when multiple links from
the same register are created with a sync between the creation of these
links. The sync corrupts the id of the register and therefore the second
link is not created properly. See the patch description to understand
more.
The fix is to preserve the id while doing the sync similar to the off.
====================
Link: https://patch.msgid.link/20260115151143.1344724-1-puranjay@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Before the last commit, sync_linked_regs() corrupted the register whose
bounds are being updated by copying known_reg's id to it. The ids are
the same in value but known_reg has the BPF_ADD_CONST flag which is
wrongly copied to reg.
This later causes issues when creating new links to this reg.
assign_scalar_id_before_mov() sees this BPF_ADD_CONST and gives a new id
to this register and breaks the old links. This is exposed by the added
selftest.
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Tested-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20260115151143.1344724-3-puranjay@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
sync_linked_regs() copies the id of known_reg to reg when propagating
bounds of known_reg to reg using the off of known_reg, but when
known_reg was linked to reg like:
known_reg = reg ; both known_reg and reg get same id
known_reg += 4 ; known_reg gets off = 4, and its id gets BPF_ADD_CONST
now when a call to sync_linked_regs() happens, let's say with the following:
if known_reg >= 10 goto pc+2
known_reg's new bounds are propagated to reg but now reg gets
BPF_ADD_CONST from the copy.
This means if another link to reg is created like:
another_reg = reg ; another_reg should get the id of reg but
assign_scalar_id_before_mov() sees
BPF_ADD_CONST on reg and assigns a new id to it.
As reg has a new id now, known_reg's link to reg is broken. If we find
new bounds for known_reg, they will not be propagated to reg.
This can be seen in the selftest added in the next commit:
0: (85) call bpf_get_prandom_u32#7 ; R0=scalar()
1: (57) r0 &= 255 ; R0=scalar(smin=smin32=0,smax=umax=smax32=umax32=255,var_off=(0x0; 0xff))
2: (bf) r1 = r0 ; R0=scalar(id=1,smin=smin32=0,smax=umax=smax32=umax32=255,var_off=(0x0; 0xff)) R1=scalar(id=1,smin=smin32=0,smax=umax=smax32=umax32=255,var_off=(0x0; 0xff))
3: (07) r1 += 4 ; R1=scalar(id=1+4,smin=umin=smin32=umin32=4,smax=umax=smax32=umax32=259,var_off=(0x0; 0x1ff))
4: (a5) if r1 < 0xa goto pc+4 ; R1=scalar(id=1+4,smin=umin=smin32=umin32=10,smax=umax=smax32=umax32=259,var_off=(0x0; 0x1ff))
5: (bf) r2 = r0 ; R0=scalar(id=2,smin=umin=smin32=umin32=6,smax=umax=smax32=umax32=255) R2=scalar(id=2,smin=umin=smin32=umin32=6,smax=umax=smax32=umax32=255)
6: (a5) if r1 < 0xe goto pc+2 ; R1=scalar(id=1+4,smin=umin=smin32=umin32=14,smax=umax=smax32=umax32=259,var_off=(0x0; 0x1ff))
7: (35) if r0 >= 0xa goto pc+1 ; R0=scalar(id=2,smin=umin=smin32=umin32=6,smax=umax=smax32=umax32=9,var_off=(0x0; 0xf))
8: (37) r0 /= 0
div by zero
When 4 is verified, r1's bounds are propagated to r0 but r0 also gets
BPF_ADD_CONST (bug).
When 5 is verified, r0 gets a new id (2) and its link with r1 is broken.
After 6 we know r1 has bounds [14, 259] and therefore r0 should have
bounds [10, 255], therefore the branch at 7 is always taken. But because
r0's id was changed to 2, r1's new bounds are not propagated to r0.
The verifier still thinks r0 has bounds [6, 255] before 7 and execution
can reach div by zero.
Fix this by preserving id in sync_linked_regs() like off and subreg_def.
Fixes: 98d7ca374b ("bpf: Track delta between "linked" registers.")
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20260115151143.1344724-2-puranjay@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
- implement the missing .get_direction() callback for gpio-davinci
- remove redundant check in GPIO core which can also propagate an
invalid errno to user-space
-----BEGIN PGP SIGNATURE-----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=jof/
-----END PGP SIGNATURE-----
Merge tag 'gpio-fixes-for-v6.19-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux
Pull gpio fixes from Bartosz Golaszewski:
"Two more GPIO fixes addressing an issue uncovered by the shared GPIO
management changes in v6.19:
- implement the missing .get_direction() callback for gpio-davinci
- remove redundant check in GPIO core which can also propagate an
invalid errno to user-space"
* tag 'gpio-fixes-for-v6.19-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux:
gpiolib: remove redundant callback check
gpio: davinci: implement .get_direction()
... same as native mount(2) is doing for devname argument. While we
are at it, fix misspelling ufs_args as cdfs_args in osf_ufs_mount() -
layouts are identical, so it doesn't change anything, but the current
variant is confusing for no reason.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
filename_flags is used by user_path_at(). I suspect that mixing
LOOKUP_EMPTY with real lookup flags had been a mistake all along; the
former belongs to pathname import, the latter - to pathwalk. Right now
none of the remaining in-tree callers of user_path_at() are getting
LOOKUP_EMPTY in flags, so user_path_at() could probably be switched
to CLASS(filename)...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Note that failures from filename_lookup() are final - ESTALE returned
by it means that retry had been done by filename_lookup() and it failed
there.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Callers switched to CLASS(filename_maybe_null) (in fs/xattr.c)
and CLASS(filename_complete_delayed) (in io_uring/xattr.c).
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
