When lower fs is a nested overlayfs, calling encode_fh() on a lower
directory dentry may trigger copy up and take sb_writers on the upper fs
of the lower nested overlayfs.
The lower nested overlayfs may have the same upper fs as this overlayfs,
so nested sb_writers lock is illegal.
Move all the callers that encode lower fh to before ovl_want_write().
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
overlayfs file open (ovl_maybe_lookup_lowerdata) and overlay file llseek
take the ovl_inode_lock, without holding upper sb_writers.
In case of nested lower overlay that uses same upper fs as this overlay,
lockdep will warn about (possibly false positive) circular lock
dependency when doing open/llseek of lower ovl file during copy up with
our upper sb_writers held, because the locking ordering seems reverse to
the locking order in ovl_copy_up_start():
- lower ovl_inode_lock
- upper sb_writers
Let the copy up "transaction" keeps an elevated mnt write count on upper
mnt, but leaves taking upper sb_writers to lower level helpers only when
they actually need it. This allows to avoid holding upper sb_writers
during lower file open/llseek and prevents the lockdep warning.
Minimizing the scope of upper sb_writers during copy up is also needed
for fixing another possible deadlocks by a following patch.
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Make the locking order of ovl_inode_lock() strictly between the two
vfs stacked layers, i.e.:
- ovl vfs locks: sb_writers, inode_lock, ...
- ovl_inode_lock
- upper vfs locks: sb_writers, inode_lock, ...
To that effect, move ovl_want_write() into the helpers ovl_nlink_start()
and ovl_copy_up_start which currently take the ovl_inode_lock() after
ovl_want_write().
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCZRKHuAAKCRCRxhvAZXjc
ohOLAQDU9Fxq5UdqCdmsyi/b24XJFZlQhcVIZy2Hrhcor9TiVQEAjuECGlxFPSgj
atVOWLdugDJquiHextqTEMgIecJpNw4=
=uINF
-----END PGP SIGNATURE-----
Merge tag 'v6.6-rc4.vfs.fixes' of gitolite.kernel.org:pub/scm/linux/kernel/git/vfs/vfs
Pull vfs fixes from Christian Brauner:
"This contains the usual miscellaneous fixes and cleanups for vfs and
individual fses:
Fixes:
- Revert ki_pos on error from buffered writes for direct io fallback
- Add missing documentation for block device and superblock handling
for changes merged this cycle
- Fix reiserfs flexible array usage
- Ensure that overlayfs sets ctime when setting mtime and atime
- Disable deferred caller completions with overlayfs writes until
proper support exists
Cleanups:
- Remove duplicate initialization in pipe code
- Annotate aio kioctx_table with __counted_by"
* tag 'v6.6-rc4.vfs.fixes' of gitolite.kernel.org:pub/scm/linux/kernel/git/vfs/vfs:
overlayfs: set ctime when setting mtime and atime
ntfs3: put resources during ntfs_fill_super()
ovl: disable IOCB_DIO_CALLER_COMP
porting: document superblock as block device holder
porting: document new block device opening order
fs/pipe: remove duplicate "offset" initializer
fs-writeback: do not requeue a clean inode having skipped pages
aio: Annotate struct kioctx_table with __counted_by
direct_write_fallback(): on error revert the ->ki_pos update from buffered write
reiserfs: Replace 1-element array with C99 style flex-array
Nathan reported that he was seeing the new warning in
setattr_copy_mgtime pop when starting podman containers. Overlayfs is
trying to set the atime and mtime via notify_change without also
setting the ctime.
POSIX states that when the atime and mtime are updated via utimes() that
we must also update the ctime to the current time. The situation with
overlayfs copy-up is analogies, so add ATTR_CTIME to the bitmask.
notify_change will fill in the value.
Reported-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Acked-by: Christian Brauner <brauner@kernel.org>
Acked-by: Amir Goldstein <amir73il@gmail.com>
Message-Id: <20230913-ctime-v1-1-c6bc509cbc27@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Some local filesystems support setting persistent fileattr flags
(e.g. FS_NOATIME_FL) on directories and regular files via ioctl.
Some of those persistent fileattr flags are reflected to vfs as
in-memory inode flags (e.g. S_NOATIME).
Overlayfs uses the in-memory inode flags (e.g. S_NOATIME) on a lower file
as an indication that a the lower file may have persistent inode fileattr
flags (e.g. FS_NOATIME_FL) that need to be copied to upper file.
However, in some cases, the S_NOATIME in-memory flag could be a false
indication for persistent FS_NOATIME_FL fileattr. For example, with NFS
and FUSE lower fs, as was the case in the two bug reports, the S_NOATIME
flag is set unconditionally for all inodes.
Users cannot set persistent fileattr flags on symlinks and special files,
but in some local fs, such as ext4/btrfs/tmpfs, the FS_NOATIME_FL fileattr
flag are inheritted to symlinks and special files from parent directory.
In both cases described above, when lower symlink has the S_NOATIME flag,
overlayfs will try to copy the symlink's fileattrs and fail with error
ENOXIO, because it could not open the symlink for the ioctl security hook.
To solve this failure, do not attempt to copyup fileattrs for anything
other than directories and regular files.
Reported-by: Ruiwen Zhao <ruiwen@google.com>
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=217850
Fixes: 72db82115d2b ("ovl: copy up sync/noatime fileattr flags")
Cc: <stable@vger.kernel.org> # v5.15
Reviewed-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Always use OVL_FS() to retrieve the corresponding struct ovl_fs from a
struct super_block.
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
The legacy behavior of ovl_statfs() reports the f_fsid filled by
underlying upper fs. This fsid is not unique among overlayfs instances
on the same upper fs.
With mount option uuid=on, generate a non-persistent uuid per overlayfs
instance and use it as the seed for f_fsid, similar to tmpfs.
This is useful for reporting fanotify events with fid info from different
instances of overlayfs over the same upper fs.
The old behavior of null uuid and upper fs fsid is retained with the
mount option uuid=null, which is the default.
The mount option uuid=off that disables uuid checks in underlying layers
also retains the legacy behavior.
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
During regular metacopy, if lowerdata file has fs-verity enabled, and
the verity option is enabled, we add the digest to the metacopy xattr.
If verity is required, and lowerdata does not have fs-verity enabled,
fall back to full copy-up (or the generated metacopy would not
validate).
Signed-off-by: Alexander Larsson <alexl@redhat.com>
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
The new digest field in the metacopy xattr is used during lookup to
record whether the header contained a digest in the OVL_HAS_DIGEST
flags.
When accessing file data the first time, if OVL_HAS_DIGEST is set, we
reload the metadata and check that the source lowerdata inode matches
the specified digest in it (according to the enabled verity
options). If the verity check passes we store this info in the inode
flags as OVL_VERIFIED_DIGEST, so that we can avoid doing it again if
the inode remains in memory.
The verification is done in ovl_maybe_validate_verity() which needs to
be called in the same places as ovl_maybe_lookup_lowerdata(), so there
is a new ovl_verify_lowerdata() helper that calls these in the right
order, and all current callers of ovl_maybe_lookup_lowerdata() are
changed to call it instead.
Signed-off-by: Alexander Larsson <alexl@redhat.com>
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Defer lookup of lowerdata in the data-only layers to first data access
or before copy up.
We perform lowerdata lookup before copy up even if copy up is metadata
only copy up. We can further optimize this lookup later if needed.
We do best effort lazy lookup of lowerdata for d_real_inode(), because
this interface does not expect errors. The only current in-tree caller
of d_real_inode() is trace_uprobe and this caller is likely going to be
followed reading from the file, before placing uprobes on offset within
the file, so lowerdata should be available when setting the uprobe.
Tested-by: kernel test robot <oliver.sang@intel.com>
Reviewed-by: Alexander Larsson <alexl@redhat.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
After copy up, we may need to update d_flags if upper dentry is on a
remote fs and lower dentries are not.
Add helpers to allow incremental update of the revalidate flags.
Fixes: bccece1ead36 ("ovl: allow remote upper")
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
We have decoupled vfs_listxattr() from IOP_XATTR. Instead we just need
to check whether inode->i_op->listxattr is implemented.
Cc: linux-unionfs@vger.kernel.org
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
If st_uid/st_gid doesn't have a mapping in the mounter's user_ns, then
copy-up should fail, just like it would fail if the mounter task was doing
the copy using "cp -a".
There's a corner case where the "cp -a" would succeed but copy up fail: if
there's a mapping of the invalid uid/gid (65534 by default) in the user
namespace. This is because stat(2) will return this value if the mapping
doesn't exist in the current user_ns and "cp -a" will in turn be able to
create a file with this uid/gid.
This behavior would be inconsistent with POSIX ACL's, which return -1 for
invalid uid/gid which result in a failed copy.
For consistency and simplicity fail the copy of the st_uid/st_gid are
invalid.
Fixes: 459c7c565ac3 ("ovl: unprivieged mounts")
Cc: <stable@vger.kernel.org> # v5.11
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Reviewed-by: Christian Brauner <brauner@kernel.org>
Reviewed-by: Seth Forshee <sforshee@kernel.org>
Now that posix acls have a proper api us it to copy them.
All filesystems that can serve as lower or upper layers for overlayfs
have gained support for the new posix acl api in previous patches.
So switch all internal overlayfs codepaths for copying posix acls to the
new posix acl api.
Acked-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQQqUNBr3gm4hGXdBJlZ7Krx/gZQ6wUCY0DP2AAKCRBZ7Krx/gZQ
6/+qAQCEGQWpcC5MB17zylaX7gqzhgAsDrwtpevlno3aIv/1pQD/YWr/E8tf7WTW
ERXRXMRx1cAzBJhUhVgIY+3ANfU2Rg4=
=cko4
-----END PGP SIGNATURE-----
Merge tag 'pull-tmpfile' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull vfs tmpfile updates from Al Viro:
"Miklos' ->tmpfile() signature change; pass an unopened struct file to
it, let it open the damn thing. Allows to add tmpfile support to FUSE"
* tag 'pull-tmpfile' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
fuse: implement ->tmpfile()
vfs: open inside ->tmpfile()
vfs: move open right after ->tmpfile()
vfs: make vfs_tmpfile() static
ovl: use vfs_tmpfile_open() helper
cachefiles: use vfs_tmpfile_open() helper
cachefiles: only pass inode to *mark_inode_inuse() helpers
cachefiles: tmpfile error handling cleanup
hugetlbfs: cleanup mknod and tmpfile
vfs: add vfs_tmpfile_open() helper
If tmpfile is used for copy up, then use this helper to create the tmpfile
and open it at the same time. This will later allow filesystems such as
fuse to do this operation atomically.
Reviewed-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
magical no_llseek thing and makes checks consistent. In particular,
ad-hoc "can we do splice via internal pipe" checks got saner (and
somewhat more permissive, which is what Jason had been after, AFAICT)
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQQqUNBr3gm4hGXdBJlZ7Krx/gZQ6wUCYug2xgAKCRBZ7Krx/gZQ
6wxWAQDqeg+xMq2FGPXmgjCa+Cp3PXH96Lp6f3hHzakIDx+t8gEAxvuiXAD22Mct
6S1SKuGj0iDIuM4L7hUiWTiY/bDXSAc=
=3EC/
-----END PGP SIGNATURE-----
Merge tag 'pull-work.lseek' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull vfs lseek updates from Al Viro:
"Jason's lseek series.
Saner handling of 'lseek should fail with ESPIPE' - this gets rid of
the magical no_llseek thing and makes checks consistent.
In particular, the ad-hoc "can we do splice via internal pipe" checks
got saner (and somewhat more permissive, which is what Jason had been
after, AFAICT)"
* tag 'pull-work.lseek' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
fs: remove no_llseek
fs: check FMODE_LSEEK to control internal pipe splicing
vfio: do not set FMODE_LSEEK flag
dma-buf: remove useless FMODE_LSEEK flag
fs: do not compare against ->llseek
fs: clear or set FMODE_LSEEK based on llseek function
Now vfs_llseek() can simply check for FMODE_LSEEK; if it's set,
we know that ->llseek() won't be NULL and if it's not we should
just fail with -ESPIPE.
A couple of other places where we used to check for special
values of ->llseek() (somewhat inconsistently) switched to
checking FMODE_LSEEK.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Now that we introduced new infrastructure to increase the type safety
for filesystems supporting idmapped mounts port the first part of the
vfs over to them.
This ports the attribute changes codepaths to rely on the new better
helpers using a dedicated type.
Before this change we used to take a shortcut and place the actual
values that would be written to inode->i_{g,u}id into struct iattr. This
had the advantage that we moved idmappings mostly out of the picture
early on but it made reasoning about changes more difficult than it
should be.
The filesystem was never explicitly told that it dealt with an idmapped
mount. The transition to the value that needed to be stored in
inode->i_{g,u}id appeared way too early and increased the probability of
bugs in various codepaths.
We know place the same value in struct iattr no matter if this is an
idmapped mount or not. The vfs will only deal with type safe
vfs{g,u}id_t. This makes it massively safer to perform permission checks
as the type will tell us what checks we need to perform and what helpers
we need to use.
Fileystems raising FS_ALLOW_IDMAP can't simply write ia_vfs{g,u}id to
inode->i_{g,u}id since they are different types. Instead they need to
use the dedicated vfs{g,u}id_to_k{g,u}id() helpers that map the
vfs{g,u}id into the filesystem.
The other nice effect is that filesystems like overlayfs don't need to
care about idmappings explicitly anymore and can simply set up struct
iattr accordingly directly.
Link: https://lore.kernel.org/lkml/CAHk-=win6+ahs1EwLkcq8apqLi_1wXFWbrPf340zYEhObpz4jA@mail.gmail.com [1]
Link: https://lore.kernel.org/r/20220621141454.2914719-9-brauner@kernel.org
Cc: Seth Forshee <sforshee@digitalocean.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
CC: linux-fsdevel@vger.kernel.org
Reviewed-by: Seth Forshee <sforshee@digitalocean.com>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Add a helper that allows to retrieve ovl xattrs from either lower or
upper layers. To stop passing mnt and dentry separately everywhere use
struct path which more accurately reflects the tight coupling between
mount and dentry in this helper. Swich over all places to pass a path
argument that can operate on either upper or lower layers. This is
needed to support idmapped base layers with overlayfs.
Some helpers are always called with an upper dentry, which is now utilized
by these helpers to create the path. Make this usage explicit by renaming
the argument to "upperdentry" and by renaming the function as well in some
cases. Also add a check in ovl_do_getxattr() to catch misuse of these
functions.
Cc: <linux-unionfs@vger.kernel.org>
Tested-by: Giuseppe Scrivano <gscrivan@redhat.com>
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Introduce ovl_lookup_upper() as a simple wrapper around lookup_one().
Make it clear in the helper's name that this only operates on the upper
layer. The wrapper will take upper layer's idmapping into account when
checking permission in lookup_one().
Cc: <linux-unionfs@vger.kernel.org>
Tested-by: Giuseppe Scrivano <gscrivan@redhat.com>
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Introduce ovl_do_notify_change() as a simple wrapper around
notify_change() to support idmapped layers. The helper mirrors other
ovl_do_*() helpers that operate on the upper layers.
When changing ownership of an upper object the intended ownership needs
to be mapped according to the upper layer's idmapping. This mapping is
the inverse to the mapping applied when copying inode information from
an upper layer to the corresponding overlay inode. So e.g., when an
upper mount maps files that are stored on-disk as owned by id 1001 to
1000 this means that calling stat on this object from an idmapped mount
will report the file as being owned by id 1000. Consequently in order to
change ownership of an object in this filesystem so it appears as being
owned by id 1000 in the upper idmapped layer it needs to store id 1001
on disk. The mnt mapping helpers take care of this.
All idmapping helpers are nops when no idmapped base layers are used.
Cc: <linux-unionfs@vger.kernel.org>
Tested-by: Giuseppe Scrivano <gscrivan@redhat.com>
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Pass down struct ovl_fs to setattr operations so we can ultimately
retrieve the relevant upper mount and take the mount's idmapping into
account when creating new filesystem objects. This is needed to support
idmapped base layers with overlay.
Cc: <linux-unionfs@vger.kernel.org>
Tested-by: Giuseppe Scrivano <gscrivan@redhat.com>
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Pass down struct ovl_fs to all creation helpers so we can ultimately
retrieve the relevant upper mount and take the mount's idmapping into
account when creating new filesystem objects. This is needed to support
idmapped base layers with overlay.
Cc: <linux-unionfs@vger.kernel.org>
Tested-by: Giuseppe Scrivano <gscrivan@redhat.com>
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Use helpers ovl_*xattr() to access user/trusted.overlay.* xattrs
and use helpers ovl_do_*xattr() to access generic xattrs. This is a
preparatory patch for using idmapped base layers with overlay.
Note that a few of those places called vfs_*xattr() calls directly to
reduce the amount of debug output. But as Miklos pointed out since
overlayfs has been stable for quite some time the debug output isn't all
that relevant anymore and the additional debug in all locations was
actually quite helpful when developing this patch series.
Cc: <linux-unionfs@vger.kernel.org>
Tested-by: Giuseppe Scrivano <gscrivan@redhat.com>
Reviewed-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Christoph Fritz is reporting that failure to copy up fileattr when upper
doesn't support fileattr or xattr results in a regression.
Return success in these failure cases; this reverts overlayfs to the old
behavior.
Add a pr_warn_once() in these cases to still let the user know about the
copy up failures.
Reported-by: Christoph Fritz <chf.fritz@googlemail.com>
Fixes: 72db82115d2b ("ovl: copy up sync/noatime fileattr flags")
Cc: <stable@vger.kernel.org> # v5.15
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
This patch is fixing a NULL pointer dereference to get a recently
introduced warning message working.
Fixes: 5b0a414d06c3 ("ovl: fix filattr copy-up failure")
Signed-off-by: Christoph Fritz <chf.fritz@googlemail.com>
Cc: <stable@vger.kernel.org> # v5.15
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
This regression can be reproduced with ntfs-3g and overlayfs:
mkdir lower upper work overlay
dd if=/dev/zero of=ntfs.raw bs=1M count=2
mkntfs -F ntfs.raw
mount ntfs.raw lower
touch lower/file.txt
mount -t overlay -o lowerdir=lower,upperdir=upper,workdir=work - overlay
mv overlay/file.txt overlay/file2.txt
mv fails and (misleadingly) prints
mv: cannot move 'overlay/file.txt' to a subdirectory of itself, 'overlay/file2.txt'
The reason is that ovl_copy_fileattr() is triggered due to S_NOATIME being
set on all inodes (by fuse) regardless of fileattr.
ovl_copy_fileattr() tries to retrieve file attributes from lower file, but
that fails because filesystem does not support this ioctl (this should fail
with ENOTTY, but ntfs-3g return EINVAL instead). This failure is
propagated to origial operation (in this case rename) that triggered the
copy-up.
The fix is to ignore ENOTTY and EINVAL errors from fileattr_get() in copy
up. This also requires turning the internal ENOIOCTLCMD into ENOTTY.
As a further measure to prevent unnecessary failures, only try the
fileattr_get/set on upper if there are any flags to copy up.
Side note: a number of filesystems set S_NOATIME (and sometimes other inode
flags) irrespective of fileattr flags. This causes unnecessary calls
during copy up, which might lead to a performance issue, especially if
latency is high. To fix this, the kernel would need to differentiate
between the two cases. E.g. introduce SB_NOATIME_UPDATE, a per-sb variant
of S_NOATIME. SB_NOATIME doesn't work, because that's interpreted as
"filesystem doesn't store an atime attribute"
Reported-and-tested-by: Kevin Locke <kevin@kevinlocke.name>
Fixes: 72db82115d2b ("ovl: copy up sync/noatime fileattr flags")
Cc: <stable@vger.kernel.org> # v5.15
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Extended attributes are usually small, but could be up to 64k in size, so
use the most efficient method for doing the allocation.
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
When a lower file has immutable/append-only fileattr flags, the behavior of
overlayfs post copy up is inconsistent.
Immediattely after copy up, ovl inode still has the S_IMMUTABLE/S_APPEND
inode flags copied from lower inode, so vfs code still treats the ovl inode
as immutable/append-only. After ovl inode evict or mount cycle, the ovl
inode does not have these inode flags anymore.
We cannot copy up the immutable and append-only fileattr flags, because
immutable/append-only inodes cannot be linked and because overlayfs will
not be able to set overlay.* xattr on the upper inodes.
Instead, if any of the fileattr flags of interest exist on the lower inode,
we store them in overlay.protattr xattr on the upper inode and we read the
flags from xattr on lookup and on fileattr_get().
This gives consistent behavior post copy up regardless of inode eviction
from cache.
When user sets new fileattr flags, we update or remove the overlay.protattr
xattr.
Storing immutable/append-only fileattr flags in an xattr instead of upper
fileattr also solves other non-standard behavior issues - overlayfs can now
copy up children of "ovl-immutable" directories and lower aliases of
"ovl-immutable" hardlinks.
Reported-by: Chengguang Xu <cgxu519@mykernel.net>
Link: https://lore.kernel.org/linux-unionfs/20201226104618.239739-1-cgxu519@mykernel.net/
Link: https://lore.kernel.org/linux-unionfs/20210210190334.1212210-5-amir73il@gmail.com/
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
When a lower file has sync/noatime fileattr flags, the behavior of
overlayfs post copy up is inconsistent.
Immediately after copy up, ovl inode still has the S_SYNC/S_NOATIME
inode flags copied from lower inode, so vfs code still treats the ovl
inode as sync/noatime. After ovl inode evict or mount cycle,
the ovl inode does not have these inode flags anymore.
To fix this inconsistency, try to copy the fileattr flags on copy up
if the upper fs supports the fileattr_set() method.
This gives consistent behavior post copy up regardless of inode eviction
from cache.
We cannot copy up the immutable/append-only inode flags in a similar
manner, because immutable/append-only inodes cannot be linked and because
overlayfs will not be able to set overlay.* xattr on the upper inodes.
Those flags will be addressed by a followup patch.
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Smatch complains about missing that the ovl_override_creds() doesn't
have a matching revert_creds() if the dentry is disconnected. Fix this
by moving the ovl_override_creds() until after the disconnected check.
Fixes: aa3ff3c152ff ("ovl: copy up of disconnected dentries")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCYCegywAKCRCRxhvAZXjc
ouJ6AQDlf+7jCQlQdeKKoN9QDFfMzG1ooemat36EpRRTONaGuAD8D9A4sUsG4+5f
4IU5Lj9oY4DEmF8HenbWK2ZHsesL2Qg=
=yPaw
-----END PGP SIGNATURE-----
Merge tag 'idmapped-mounts-v5.12' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux
Pull idmapped mounts from Christian Brauner:
"This introduces idmapped mounts which has been in the making for some
time. Simply put, different mounts can expose the same file or
directory with different ownership. This initial implementation comes
with ports for fat, ext4 and with Christoph's port for xfs with more
filesystems being actively worked on by independent people and
maintainers.
Idmapping mounts handle a wide range of long standing use-cases. Here
are just a few:
- Idmapped mounts make it possible to easily share files between
multiple users or multiple machines especially in complex
scenarios. For example, idmapped mounts will be used in the
implementation of portable home directories in
systemd-homed.service(8) where they allow users to move their home
directory to an external storage device and use it on multiple
computers where they are assigned different uids and gids. This
effectively makes it possible to assign random uids and gids at
login time.
- It is possible to share files from the host with unprivileged
containers without having to change ownership permanently through
chown(2).
- It is possible to idmap a container's rootfs and without having to
mangle every file. For example, Chromebooks use it to share the
user's Download folder with their unprivileged containers in their
Linux subsystem.
- It is possible to share files between containers with
non-overlapping idmappings.
- Filesystem that lack a proper concept of ownership such as fat can
use idmapped mounts to implement discretionary access (DAC)
permission checking.
- They allow users to efficiently changing ownership on a per-mount
basis without having to (recursively) chown(2) all files. In
contrast to chown (2) changing ownership of large sets of files is
instantenous with idmapped mounts. This is especially useful when
ownership of a whole root filesystem of a virtual machine or
container is changed. With idmapped mounts a single syscall
mount_setattr syscall will be sufficient to change the ownership of
all files.
- Idmapped mounts always take the current ownership into account as
idmappings specify what a given uid or gid is supposed to be mapped
to. This contrasts with the chown(2) syscall which cannot by itself
take the current ownership of the files it changes into account. It
simply changes the ownership to the specified uid and gid. This is
especially problematic when recursively chown(2)ing a large set of
files which is commong with the aforementioned portable home
directory and container and vm scenario.
- Idmapped mounts allow to change ownership locally, restricting it
to specific mounts, and temporarily as the ownership changes only
apply as long as the mount exists.
Several userspace projects have either already put up patches and
pull-requests for this feature or will do so should you decide to pull
this:
- systemd: In a wide variety of scenarios but especially right away
in their implementation of portable home directories.
https://systemd.io/HOME_DIRECTORY/
- container runtimes: containerd, runC, LXD:To share data between
host and unprivileged containers, unprivileged and privileged
containers, etc. The pull request for idmapped mounts support in
containerd, the default Kubernetes runtime is already up for quite
a while now: https://github.com/containerd/containerd/pull/4734
- The virtio-fs developers and several users have expressed interest
in using this feature with virtual machines once virtio-fs is
ported.
- ChromeOS: Sharing host-directories with unprivileged containers.
I've tightly synced with all those projects and all of those listed
here have also expressed their need/desire for this feature on the
mailing list. For more info on how people use this there's a bunch of
talks about this too. Here's just two recent ones:
https://www.cncf.io/wp-content/uploads/2020/12/Rootless-Containers-in-Gitpod.pdfhttps://fosdem.org/2021/schedule/event/containers_idmap/
This comes with an extensive xfstests suite covering both ext4 and
xfs:
https://git.kernel.org/brauner/xfstests-dev/h/idmapped_mounts
It covers truncation, creation, opening, xattrs, vfscaps, setid
execution, setgid inheritance and more both with idmapped and
non-idmapped mounts. It already helped to discover an unrelated xfs
setgid inheritance bug which has since been fixed in mainline. It will
be sent for inclusion with the xfstests project should you decide to
merge this.
In order to support per-mount idmappings vfsmounts are marked with
user namespaces. The idmapping of the user namespace will be used to
map the ids of vfs objects when they are accessed through that mount.
By default all vfsmounts are marked with the initial user namespace.
The initial user namespace is used to indicate that a mount is not
idmapped. All operations behave as before and this is verified in the
testsuite.
Based on prior discussions we want to attach the whole user namespace
and not just a dedicated idmapping struct. This allows us to reuse all
the helpers that already exist for dealing with idmappings instead of
introducing a whole new range of helpers. In addition, if we decide in
the future that we are confident enough to enable unprivileged users
to setup idmapped mounts the permission checking can take into account
whether the caller is privileged in the user namespace the mount is
currently marked with.
The user namespace the mount will be marked with can be specified by
passing a file descriptor refering to the user namespace as an
argument to the new mount_setattr() syscall together with the new
MOUNT_ATTR_IDMAP flag. The system call follows the openat2() pattern
of extensibility.
The following conditions must be met in order to create an idmapped
mount:
- The caller must currently have the CAP_SYS_ADMIN capability in the
user namespace the underlying filesystem has been mounted in.
- The underlying filesystem must support idmapped mounts.
- The mount must not already be idmapped. This also implies that the
idmapping of a mount cannot be altered once it has been idmapped.
- The mount must be a detached/anonymous mount, i.e. it must have
been created by calling open_tree() with the OPEN_TREE_CLONE flag
and it must not already have been visible in the filesystem.
The last two points guarantee easier semantics for userspace and the
kernel and make the implementation significantly simpler.
By default vfsmounts are marked with the initial user namespace and no
behavioral or performance changes are observed.
The manpage with a detailed description can be found here:
1d7b902e28
In order to support idmapped mounts, filesystems need to be changed
and mark themselves with the FS_ALLOW_IDMAP flag in fs_flags. The
patches to convert individual filesystem are not very large or
complicated overall as can be seen from the included fat, ext4, and
xfs ports. Patches for other filesystems are actively worked on and
will be sent out separately. The xfstestsuite can be used to verify
that port has been done correctly.
The mount_setattr() syscall is motivated independent of the idmapped
mounts patches and it's been around since July 2019. One of the most
valuable features of the new mount api is the ability to perform
mounts based on file descriptors only.
Together with the lookup restrictions available in the openat2()
RESOLVE_* flag namespace which we added in v5.6 this is the first time
we are close to hardened and race-free (e.g. symlinks) mounting and
path resolution.
While userspace has started porting to the new mount api to mount
proper filesystems and create new bind-mounts it is currently not
possible to change mount options of an already existing bind mount in
the new mount api since the mount_setattr() syscall is missing.
With the addition of the mount_setattr() syscall we remove this last
restriction and userspace can now fully port to the new mount api,
covering every use-case the old mount api could. We also add the
crucial ability to recursively change mount options for a whole mount
tree, both removing and adding mount options at the same time. This
syscall has been requested multiple times by various people and
projects.
There is a simple tool available at
https://github.com/brauner/mount-idmapped
that allows to create idmapped mounts so people can play with this
patch series. I'll add support for the regular mount binary should you
decide to pull this in the following weeks:
Here's an example to a simple idmapped mount of another user's home
directory:
u1001@f2-vm:/$ sudo ./mount --idmap both:1000:1001:1 /home/ubuntu/ /mnt
u1001@f2-vm:/$ ls -al /home/ubuntu/
total 28
drwxr-xr-x 2 ubuntu ubuntu 4096 Oct 28 22:07 .
drwxr-xr-x 4 root root 4096 Oct 28 04:00 ..
-rw------- 1 ubuntu ubuntu 3154 Oct 28 22:12 .bash_history
-rw-r--r-- 1 ubuntu ubuntu 220 Feb 25 2020 .bash_logout
-rw-r--r-- 1 ubuntu ubuntu 3771 Feb 25 2020 .bashrc
-rw-r--r-- 1 ubuntu ubuntu 807 Feb 25 2020 .profile
-rw-r--r-- 1 ubuntu ubuntu 0 Oct 16 16:11 .sudo_as_admin_successful
-rw------- 1 ubuntu ubuntu 1144 Oct 28 00:43 .viminfo
u1001@f2-vm:/$ ls -al /mnt/
total 28
drwxr-xr-x 2 u1001 u1001 4096 Oct 28 22:07 .
drwxr-xr-x 29 root root 4096 Oct 28 22:01 ..
-rw------- 1 u1001 u1001 3154 Oct 28 22:12 .bash_history
-rw-r--r-- 1 u1001 u1001 220 Feb 25 2020 .bash_logout
-rw-r--r-- 1 u1001 u1001 3771 Feb 25 2020 .bashrc
-rw-r--r-- 1 u1001 u1001 807 Feb 25 2020 .profile
-rw-r--r-- 1 u1001 u1001 0 Oct 16 16:11 .sudo_as_admin_successful
-rw------- 1 u1001 u1001 1144 Oct 28 00:43 .viminfo
u1001@f2-vm:/$ touch /mnt/my-file
u1001@f2-vm:/$ setfacl -m u:1001:rwx /mnt/my-file
u1001@f2-vm:/$ sudo setcap -n 1001 cap_net_raw+ep /mnt/my-file
u1001@f2-vm:/$ ls -al /mnt/my-file
-rw-rwxr--+ 1 u1001 u1001 0 Oct 28 22:14 /mnt/my-file
u1001@f2-vm:/$ ls -al /home/ubuntu/my-file
-rw-rwxr--+ 1 ubuntu ubuntu 0 Oct 28 22:14 /home/ubuntu/my-file
u1001@f2-vm:/$ getfacl /mnt/my-file
getfacl: Removing leading '/' from absolute path names
# file: mnt/my-file
# owner: u1001
# group: u1001
user::rw-
user:u1001:rwx
group::rw-
mask::rwx
other::r--
u1001@f2-vm:/$ getfacl /home/ubuntu/my-file
getfacl: Removing leading '/' from absolute path names
# file: home/ubuntu/my-file
# owner: ubuntu
# group: ubuntu
user::rw-
user:ubuntu:rwx
group::rw-
mask::rwx
other::r--"
* tag 'idmapped-mounts-v5.12' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux: (41 commits)
xfs: remove the possibly unused mp variable in xfs_file_compat_ioctl
xfs: support idmapped mounts
ext4: support idmapped mounts
fat: handle idmapped mounts
tests: add mount_setattr() selftests
fs: introduce MOUNT_ATTR_IDMAP
fs: add mount_setattr()
fs: add attr_flags_to_mnt_flags helper
fs: split out functions to hold writers
namespace: only take read lock in do_reconfigure_mnt()
mount: make {lock,unlock}_mount_hash() static
namespace: take lock_mount_hash() directly when changing flags
nfs: do not export idmapped mounts
overlayfs: do not mount on top of idmapped mounts
ecryptfs: do not mount on top of idmapped mounts
ima: handle idmapped mounts
apparmor: handle idmapped mounts
fs: make helpers idmap mount aware
exec: handle idmapped mounts
would_dump: handle idmapped mounts
...
When inode has no listxattr op of its own (e.g. squashfs) vfs_listxattr
calls the LSM inode_listsecurity hooks to list the xattrs that LSMs will
intercept in inode_getxattr hooks.
When selinux LSM is installed but not initialized, it will list the
security.selinux xattr in inode_listsecurity, but will not intercept it
in inode_getxattr. This results in -ENODATA for a getxattr call for an
xattr returned by listxattr.
This situation was manifested as overlayfs failure to copy up lower
files from squashfs when selinux is built-in but not initialized,
because ovl_copy_xattr() iterates the lower inode xattrs by
vfs_listxattr() and vfs_getxattr().
ovl_copy_xattr() skips copy up of security labels that are indentified by
inode_copy_up_xattr LSM hooks, but it does that after vfs_getxattr().
Since we are not going to copy them, skip vfs_getxattr() of the security
labels.
Reported-by: Michael Labriola <michael.d.labriola@gmail.com>
Tested-by: Michael Labriola <michael.d.labriola@gmail.com>
Link: https://lore.kernel.org/linux-unionfs/2nv9d47zt7.fsf@aldarion.sourceruckus.org/
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
When interacting with extended attributes the vfs verifies that the
caller is privileged over the inode with which the extended attribute is
associated. For posix access and posix default extended attributes a uid
or gid can be stored on-disk. Let the functions handle posix extended
attributes on idmapped mounts. If the inode is accessed through an
idmapped mount we need to map it according to the mount's user
namespace. Afterwards the checks are identical to non-idmapped mounts.
This has no effect for e.g. security xattrs since they don't store uids
or gids and don't perform permission checks on them like posix acls do.
Link: https://lore.kernel.org/r/20210121131959.646623-10-christian.brauner@ubuntu.com
Cc: Christoph Hellwig <hch@lst.de>
Cc: David Howells <dhowells@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: linux-fsdevel@vger.kernel.org
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: James Morris <jamorris@linux.microsoft.com>
Signed-off-by: Tycho Andersen <tycho@tycho.pizza>
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
When file attributes are changed most filesystems rely on the
setattr_prepare(), setattr_copy(), and notify_change() helpers for
initialization and permission checking. Let them handle idmapped mounts.
If the inode is accessed through an idmapped mount map it into the
mount's user namespace. Afterwards the checks are identical to
non-idmapped mounts. If the initial user namespace is passed nothing
changes so non-idmapped mounts will see identical behavior as before.
Helpers that perform checks on the ia_uid and ia_gid fields in struct
iattr assume that ia_uid and ia_gid are intended values and have already
been mapped correctly at the userspace-kernelspace boundary as we
already do today. If the initial user namespace is passed nothing
changes so non-idmapped mounts will see identical behavior as before.
Link: https://lore.kernel.org/r/20210121131959.646623-8-christian.brauner@ubuntu.com
Cc: Christoph Hellwig <hch@lst.de>
Cc: David Howells <dhowells@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: linux-fsdevel@vger.kernel.org
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Comment above call already says this, but only EOPNOTSUPP is ignored, other
failures are not.
For example setting "user.*" will fail with EPERM on symlink/special.
Ignore this error as well.
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
This replaces uuid with null in overlayfs file handles and thus relaxes
uuid checks for overlay index feature. It is only possible in case there is
only one filesystem for all the work/upper/lower directories and bare file
handles from this backing filesystem are unique. In other case when we have
multiple filesystems lets just fallback to "uuid=on" which is and
equivalent of how it worked before with all uuid checks.
This is needed when overlayfs is/was mounted in a container with index
enabled (e.g.: to be able to resolve inotify watch file handles on it to
paths in CRIU), and this container is copied and started alongside with the
original one. This way the "copy" container can't have the same uuid on the
superblock and mounting the overlayfs from it later would fail.
That is an example of the problem on top of loop+ext4:
dd if=/dev/zero of=loopbackfile.img bs=100M count=10
losetup -fP loopbackfile.img
losetup -a
#/dev/loop0: [64768]:35 (/loop-test/loopbackfile.img)
mkfs.ext4 loopbackfile.img
mkdir loop-mp
mount -o loop /dev/loop0 loop-mp
mkdir loop-mp/{lower,upper,work,merged}
mount -t overlay overlay -oindex=on,lowerdir=loop-mp/lower,\
upperdir=loop-mp/upper,workdir=loop-mp/work loop-mp/merged
umount loop-mp/merged
umount loop-mp
e2fsck -f /dev/loop0
tune2fs -U random /dev/loop0
mount -o loop /dev/loop0 loop-mp
mount -t overlay overlay -oindex=on,lowerdir=loop-mp/lower,\
upperdir=loop-mp/upper,workdir=loop-mp/work loop-mp/merged
#mount: /loop-test/loop-mp/merged:
#mount(2) system call failed: Stale file handle.
If you just change the uuid of the backing filesystem, overlay is not
mounting any more. In Virtuozzo we copy container disks (ploops) when
create the copy of container and we require fs uuid to be unique for a new
container.
Signed-off-by: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
This will be used in next patch to be able to change uuid checks and add
uuid nullification based on ofs->config.index for a new "uuid=off" mode.
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Call ovl_do_*xattr() when accessing an overlay private xattr, vfs_*xattr()
otherwise.
This has an effect on debug output, which is made more consistent by this
patch.
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Lose the padding and the failure message (in line with other parts of the
copy up process). Return zero for both nonexistent or empty xattr.
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
ovl_getattr() returns the value of an xattr in a kmalloced buffer. There
are two callers:
ovl_copy_up_meta_inode_data() (copy_up.c)
ovl_get_redirect_xattr() (util.c)
This patch just copies ovl_getxattr() to copy_up.c, the following patches
will deal with the differences in idividual callers.
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Container folks are complaining that dnf/yum issues too many sync while
installing packages and this slows down the image build. Build requirement
is such that they don't care if a node goes down while build was still
going on. In that case, they will simply throw away unfinished layer and
start new build. So they don't care about syncing intermediate state to the
disk and hence don't want to pay the price associated with sync.
So they are asking for mount options where they can disable sync on overlay
mount point.
They primarily seem to have two use cases.
- For building images, they will mount overlay with nosync and then sync
upper layer after unmounting overlay and reuse upper as lower for next
layer.
- For running containers, they don't seem to care about syncing upper layer
because if node goes down, they will simply throw away upper layer and
create a fresh one.
So this patch provides a mount option "volatile" which disables all forms
of sync. Now it is caller's responsibility to throw away upper if system
crashes or shuts down and start fresh.
With "volatile", I am seeing roughly 20% speed up in my VM where I am just
installing emacs in an image. Installation time drops from 31 seconds to 25
seconds when nosync option is used. This is for the case of building on top
of an image where all packages are already cached. That way I take out the
network operations latency out of the measurement.
Giuseppe is also looking to cut down on number of iops done on the disk. He
is complaining that often in cloud their VMs are throttled if they cross
the limit. This option can help them where they reduce number of iops (by
cutting down on frequent sync and writebacks).
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
