Skip to contentSkip to conversation

idmapped mounts

PATCH Superseded Christian Brauner 7 participants 82 messages Jan 12, 2021 – Jan 20, 2021 Page 1 of 2

Subsystems (94)

v5 of 6 Superseded Latest is v6 →

lore
Share & tools
lore
#0 lore raw

v5 cover idmapped mounts

Hey everyone,

The only major change is the inclusion of hch's patch to port XFS to support idmapped mounts. Thanks to Christoph for doing that work. (For a full list of major changes between versions see the end of this
cover letter.
Please also note the large xfstests testsuite in patch 42 that has been kept as part of this series. It verifies correct vfs behavior with and without idmapped mounts including covering newer vfs features such as io_uring.
I currently still plan to target the v5.12 merge window.)

With this patchset we make it possible to attach idmappings to mounts, i.e. simply put different bind mounts can expose the same file or directory with different ownership.
Shifting of ownership on a per-mount basis handles a wide range of long standing use-cases. Here are just a few:
- Shifting of a subset of ownership-less filesystems (vfat) for use by multiple users, effectively allowing for DAC on such devices (systemd, Android, ...)
- Allow remapping uid/gid on external filesystems or paths (USB sticks, network filesystem, ...) to match the local system's user and groups. (David Howells intends to port AFS as a first candidate.)
- Shifting of a container rootfs or base image without having to mangle every file (runc, Docker, containerd, k8s, LXD, systemd ...)
- Sharing of data between host or privileged containers with unprivileged containers (runC, Docker, containerd, k8s, LXD, ...)
- Data sharing between multiple user namespaces with incompatible maps (LXD, k8s, ...)

There has been significant interest in this patchset as evidenced by user commenting on previous version of this patchset. They include containerd, ChromeOS, systemd, LXD and a range of others. There is already a patchset up for containerd, the default Kubernetes container runtime https://github.com/containerd/containerd/pull/4734
to make use of this. systemd intends to use it in their systemd-homed implementation for portable home directories. ChromeOS wants to make use of it to share data between the host and the Linux containers they run on Chrome- and Pixelbooks.
(Fwiw, for fun and since I wanted to do this for a long time I've ported
my home directory to be completely portable with a simple service file that now mounts my home directory on an ext4 formatted usb stick with an id mapping mapping all files to the random uid I'm assigned at login.)

Making it possible to share directories and mounts between users with different uids and gids is itself quite an important use-case in distributed systems environments. It's of course especially useful in general for portable usb sticks, sharing data between multiple users in, and sharing home directories between multiple users. The last example is now elegantly expressed in systemd's homed concept for portable home directories. As mentioned above, idmapped mounts also allow data from the host to be shared with unprivileged containers, between privileged and unprivileged containers simultaneously and in addition also between unprivileged containers with different idmappings whenever they are used to isolate one container completely from another container.

We have implemented and proposed multiple solutions to this before. This included the introduction of fsid mappings, a tiny filesystem I've authored with Seth Forshee that is currently carried in Ubuntu that has shown to be the wrong approach, and the conceptual hack of calling override creds directly in the vfs. In addition, to some of these solutions being hacky none of these solutions have covered all of the above use-cases.

Idmappings become a property of struct vfsmount instead of tying it to a process being inside of a user namespace which has been the case for all other proposed approaches. It also allows to pass down the user namespace into the filesystems which is a clean way instead of violating calling conventions by strapping the user namespace information that is a property of the mount to the caller's credentials or similar hacks. Each mount can have a separate idmapping and idmapped mounts can even be created in the initial user namespace unblocking a range of use-cases.

To this end the vfsmount struct gains a new struct user_namespace member. The idmapping of the user namespace becomes the idmapping of the mount. A caller that is privileged with respect to the user namespace of the superblock of the underlying filesystem can create an idmapped mount. In the future, we can enable unprivileged use-cases by checking whether the caller is privileged wrt to the user namespace that an already idmapped mount has been marked with, allowing them to change the idmapping. For now, keep things simple until the need arises. Note, that with syscall interception it is already possible to intercept idmapped mount requests from unprivileged containers and handle them in a sufficiently privileged container manager. Support for this is already available in LXD and will be available in runC where syscall interception is currently in the process of becoming part of the runtime
spec: https://github.com/opencontainers/runtime-spec/pull/1074.

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. By default vfsmounts are marked with the initial user namespace and no behavioral or performance changes are observed. All mapping operations are nops for the initial user namespace. When a file/inode is accessed through an idmapped mount the i_uid and i_gid of the inode will be remapped according to the user namespace the mount has been marked with.

In order to support idmapped mounts, filesystems need to be changed and mark themselves with the FS_ALLOW_IDMAP flag in fs_flags. The initial version contains fat, ext4, and xfs including a list of examples. But patches for other filesystems are actively worked on and will be sent out separately. We are here to see this through and there are multiple people involved in converting filesystems. So filesystem developers are not left alone with this and are provided with a large testsuite to verify that their port is correct.

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. Here are a few illustrations:

1. Create a simple idmapped mount of another user's home directory

u1001@f2-vm:/$ sudo ./mount-idmapped --map-mount b: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--

2. Create mapping of the whole ext4 rootfs without a mapping for uid and gid 0

ubuntu@f2-vm:~$ sudo /mount-idmapped --map-mount b:1:1:65536 / /mnt/
ubuntu@f2-vm:~$ findmnt | grep mnt
└─/mnt                                /dev/sda2  ext4       rw,relatime
  └─/mnt/mnt                          /dev/sda2  ext4       rw,relatime
ubuntu@f2-vm:~$ sudo mkdir /AS-ROOT-CAN-CREATE
ubuntu@f2-vm:~$ sudo mkdir /mnt/AS-ROOT-CANT-CREATE
mkdir: cannot create directory ‘/mnt/AS-ROOT-CANT-CREATE’: Value too large for defined data type
ubuntu@f2-vm:~$ mkdir /mnt/home/ubuntu/AS-USER-1000-CAN-CREATE

3. Create a vfat usb mount and expose to user 1001 and 5000

ubuntu@f2-vm:/$ sudo mount /dev/sdb /mnt
ubuntu@f2-vm:/$ findmnt  | grep mnt
└─/mnt                                /dev/sdb vfat       rw,relatime,fmask=0022,dmask=0022,codepage=437,iocharset=iso8859-1,shortname=mixed,errors=remount-ro
ubuntu@f2-vm:/$ ls -al /mnt
total 12
drwxr-xr-x  2 root root 4096 Jan  1  1970 .
drwxr-xr-x 34 root root 4096 Oct 28 22:24 ..
-rwxr-xr-x  1 root root    4 Oct 28 03:44 aaa
-rwxr-xr-x  1 root root    0 Oct 28 01:09 bbb
ubuntu@f2-vm:/$ sudo /mount-idmapped --map-mount b:0:1001:1 /mnt /mnt-1001/
ubuntu@f2-vm:/$ ls -al /mnt-1001/
total 12
drwxr-xr-x  2 u1001 u1001 4096 Jan  1  1970 .
drwxr-xr-x 34 root  root  4096 Oct 28 22:24 ..
-rwxr-xr-x  1 u1001 u1001    4 Oct 28 03:44 aaa
-rwxr-xr-x  1 u1001 u1001    0 Oct 28 01:09 bbb
ubuntu@f2-vm:/$ sudo /mount-idmapped --map-mount b:0:5000:1 /mnt /mnt-5000/
ubuntu@f2-vm:/$ ls -al /mnt-5000/
total 12
drwxr-xr-x  2 5000 5000 4096 Jan  1  1970 .
drwxr-xr-x 34 root root 4096 Oct 28 22:24 ..
-rwxr-xr-x  1 5000 5000    4 Oct 28 03:44 aaa
-rwxr-xr-x  1 5000 5000    0 Oct 28 01:09 bbb

4. Create an idmapped rootfs mount for a container

root@f2-vm:~# ls -al /var/lib/lxc/f2/rootfs/
total 68
drwxr-xr-x 17 20000 20000 4096 Sep 24 07:48 .
drwxrwx---  3 20000 20000 4096 Oct 16 19:26 ..
lrwxrwxrwx  1 20000 20000    7 Sep 24 07:43 bin -> usr/bin
drwxr-xr-x  2 20000 20000 4096 Apr 15  2020 boot
drwxr-xr-x  3 20000 20000 4096 Oct 16 19:26 dev
drwxr-xr-x 61 20000 20000 4096 Oct 16 19:26 etc
drwxr-xr-x  3 20000 20000 4096 Sep 24 07:45 home
lrwxrwxrwx  1 20000 20000    7 Sep 24 07:43 lib -> usr/lib
lrwxrwxrwx  1 20000 20000    9 Sep 24 07:43 lib32 -> usr/lib32
lrwxrwxrwx  1 20000 20000    9 Sep 24 07:43 lib64 -> usr/lib64
lrwxrwxrwx  1 20000 20000   10 Sep 24 07:43 libx32 -> usr/libx32
drwxr-xr-x  2 20000 20000 4096 Sep 24 07:43 media
drwxr-xr-x  2 20000 20000 4096 Sep 24 07:43 mnt
drwxr-xr-x  2 20000 20000 4096 Sep 24 07:43 opt
drwxr-xr-x  2 20000 20000 4096 Apr 15  2020 proc
drwx------  2 20000 20000 4096 Sep 24 07:43 root
drwxr-xr-x  2 20000 20000 4096 Sep 24 07:45 run
lrwxrwxrwx  1 20000 20000    8 Sep 24 07:43 sbin -> usr/sbin
drwxr-xr-x  2 20000 20000 4096 Sep 24 07:43 srv
drwxr-xr-x  2 20000 20000 4096 Apr 15  2020 sys
drwxrwxrwt  2 20000 20000 4096 Sep 24 07:44 tmp
drwxr-xr-x 13 20000 20000 4096 Sep 24 07:43 usr
drwxr-xr-x 12 20000 20000 4096 Sep 24 07:44 var
root@f2-vm:~# /mount-idmapped --map-mount b:20000:10000:100000 /var/lib/lxc/f2/rootfs/ /mnt
root@f2-vm:~# ls -al /mnt
total 68
drwxr-xr-x 17 10000 10000 4096 Sep 24 07:48 .
drwxr-xr-x 34 root  root  4096 Oct 28 22:24 ..
lrwxrwxrwx  1 10000 10000    7 Sep 24 07:43 bin -> usr/bin
drwxr-xr-x  2 10000 10000 4096 Apr 15  2020 boot
drwxr-xr-x  3 10000 10000 4096 Oct 16 19:26 dev
drwxr-xr-x 61 10000 10000 4096 Oct 16 19:26 etc
drwxr-xr-x  3 10000 10000 4096 Sep 24 07:45 home
lrwxrwxrwx  1 10000 10000    7 Sep 24 07:43 lib -> usr/lib
lrwxrwxrwx  1 10000 10000    9 Sep 24 07:43 lib32 -> usr/lib32
lrwxrwxrwx  1 10000 10000    9 Sep 24 07:43 lib64 -> usr/lib64
lrwxrwxrwx  1 10000 10000   10 Sep 24 07:43 libx32 -> usr/libx32
drwxr-xr-x  2 10000 10000 4096 Sep 24 07:43 media
drwxr-xr-x  2 10000 10000 4096 Sep 24 07:43 mnt
drwxr-xr-x  2 10000 10000 4096 Sep 24 07:43 opt
drwxr-xr-x  2 10000 10000 4096 Apr 15  2020 proc
drwx------  2 10000 10000 4096 Sep 24 07:43 root
drwxr-xr-x  2 10000 10000 4096 Sep 24 07:45 run
lrwxrwxrwx  1 10000 10000    8 Sep 24 07:43 sbin -> usr/sbin
drwxr-xr-x  2 10000 10000 4096 Sep 24 07:43 srv
drwxr-xr-x  2 10000 10000 4096 Apr 15  2020 sys
drwxrwxrwt  2 10000 10000 4096 Sep 24 07:44 tmp
drwxr-xr-x 13 10000 10000 4096 Sep 24 07:43 usr
drwxr-xr-x 12 10000 10000 4096 Sep 24 07:44 var
root@f2-vm:~# lxc-start f2 # uses /mnt as rootfs
root@f2-vm:~# lxc-attach f2 -- cat /proc/1/uid_map
         0      10000      10000
root@f2-vm:~# lxc-attach f2 -- cat /proc/1/gid_map
         0      10000      10000
root@f2-vm:~# lxc-attach f2 -- ls -al /
total 52
drwxr-xr-x  17 root   root    4096 Sep 24 07:48 .
drwxr-xr-x  17 root   root    4096 Sep 24 07:48 ..
lrwxrwxrwx   1 root   root       7 Sep 24 07:43 bin -> usr/bin
drwxr-xr-x   2 root   root    4096 Apr 15  2020 boot
drwxr-xr-x   5 root   root     500 Oct 28 23:39 dev
drwxr-xr-x  61 root   root    4096 Oct 28 23:39 etc
drwxr-xr-x   3 root   root    4096 Sep 24 07:45 home
lrwxrwxrwx   1 root   root       7 Sep 24 07:43 lib -> usr/lib
lrwxrwxrwx   1 root   root       9 Sep 24 07:43 lib32 -> usr/lib32
lrwxrwxrwx   1 root   root       9 Sep 24 07:43 lib64 -> usr/lib64
lrwxrwxrwx   1 root   root      10 Sep 24 07:43 libx32 -> usr/libx32
drwxr-xr-x   2 root   root    4096 Sep 24 07:43 media
drwxr-xr-x   2 root   root    4096 Sep 24 07:43 mnt
drwxr-xr-x   2 root   root    4096 Sep 24 07:43 opt
dr-xr-xr-x 232 nobody nogroup    0 Oct 28 23:39 proc
drwx------   2 root   root    4096 Oct 28 23:41 root
drwxr-xr-x  12 root   root     360 Oct 28 23:39 run
lrwxrwxrwx   1 root   root       8 Sep 24 07:43 sbin -> usr/sbin
drwxr-xr-x   2 root   root    4096 Sep 24 07:43 srv
dr-xr-xr-x  13 nobody nogroup    0 Oct 28 23:39 sys
drwxrwxrwt  11 root   root    4096 Oct 28 23:40 tmp
drwxr-xr-x  13 root   root    4096 Sep 24 07:43 usr
drwxr-xr-x  12 root   root    4096 Sep 24 07:44 var
root@f2-vm:~# lxc-attach f2 -- ls -al /my-file
-rw-r--r-- 1 root root 0 Oct 28 23:43 /my-file
root@f2-vm:~# ls -al /var/lib/lxc/f2/rootfs/my-file
-rw-r--r-- 1 20000 20000 0 Oct 28 23:43 /var/lib/lxc/f2/rootfs/my-file

I'd like to say thanks to:
Al for pointing me into the direction to avoid inode alias issues during lookup. David for various discussions around this. Christoph for porting xfs, providing good reviews and for being involved in the original idea. Tycho for helping with this series and on future patches to convert filesystems. Alban Crequy and the Kinvolk peeps located just a few streets away from me in Berlin for providing use-case discussions and writing patches for containerd. Stéphane for his invaluable input on many things and level head and enabling me to work on this. Amir for explaining and discussing aspects of overlayfs with me. I'd like to especially thank Seth Forshee. He provided a lot of good analysis, suggestions, and participated in short-notice discussions in both chat and video for some nitty-gritty technical details.

This series can be found and pulled from the three usual locations:
https://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux.git/log/?h=idmapped_mounts
https://github.com/brauner/linux/tree/idmapped_mounts
https://gitlab.com/brauner/linux/-/commits/idmapped_mounts

/* v5 */
- Adress Christoph's feedback.
- Use v5.11-rc3 as new base.
- Add Christoph's xfs port.

/* v4 */
- Split out several preparatory patches from the initial mount_setattr patch as requested by Christoph.
- Add new tests for file/directory creation in directories with the setgid bit set. Specifically, verify that the setgid bit is correctly ignored when creating a file with the setgid bit and the parent directory's i_gid isn't in_group_p() and the caller isn't capable_wrt_inode_uidgid() over the parent directory's inode when inode_init_owner() is called.
Conversely, verify that the setgid bit is set when creating a file with the setgid bit and the parent's i_gid is either in_group_p() or the caller is capable_wrt_inode_uidgid() over the parent directory's inode. In additiona, verify that the setgid bit is always inherited when creating directories.
Test all of this on regular mounts, idmapped mounts, and on idmapped mounts in user namespaces.
- Add new tests to verify that the i_gid of newly created files or directories is correctly set to the parent directory's i_gid when the parent directory has the setgid bit set.
- Use "mnt_userns" as the de facto name for a vfsmount's user namespace everywhere as suggested by Serge.
- Reuse existing propagation flags instead of introducing new ones as suggested by Christoph. (This is in line with Linus request to not introduce too many new flags as evidenced by prior discussions on other patchsets such as openat2().)
- Add first set of Acked-bys from Serge and Reviewed-bys from Christoph.
- Fix commit messages to reflect the fact that we modify existing vfs helpers but do not introduce new ones like we did in the first version. Some commit messages still implied we were adding new helpers.
- Reformat all commit messages to adhere to 73 char length limit and wrap all lines in commits at 80 chars whenever this doesn't hinder legibility.
- Simplify various codepaths with Christoph's suggestions.

/* v3 */
- The major change is the port of the test-suite from the kernel-internal selftests framework to xfstests as requested by Darrick and Christoph. The test-suite for xfstests is patch 38 in this series. It has been kept as part of this series even though it belongs to xfstests so it's easier to see what is tested and to keep it in-sync.
- Note, the test-suite now has been extended to cover io_uring and idmapped mounts. The IORING_REGISTER_PERSONALITY feature allows to register the caller's credentials with io_uring and returns an id associated with these credentials. This is useful for applications that wish to share a ring between separate users/processes. Callers can pass in the credential id in the sqe personality field. If set, that particular sqe will be issued with these credentials. The test-suite now tests that the openat* operations with different registered credentials work correctly and safely on regular mounts, on regular mounts inside user namespaces, on idmapped mounts, and on idmapped mounts inside user namespaces.

/* v2 */
- The major change is the rework requested by Christoph and others to adapt all relevant helpers and inode_operations methods to account for idmapped mounts instead of introducing new helpers and methods specific to idmapped mounts like we did before. We've also moved the overlayfs conversion to handle idmapped mounts into a separate patchset that will be sent out separately after the core changes landed. The converted filesytems in this series include fat and ext4. As per Christoph's request the vfs-wide config option to disable idmapped mounts has been removed. Instead the filesystems can decide whether or not they want to allow idmap mounts through a config option. These config options default to off. Having a config option allows us to gain some confidence in the patchset over multiple kernel releases.
- This version introduces a large test-suite to test current vfs behavior and idmapped mounts behavior. This test-suite is intended to grow over time.
- While while working on adapting this patchset to the requested changes, the runC and containerd crowd was nice enough to adapt containerd to this patchset to make use of idmapped mounts in one of the most widely used container runtimes:
https://github.com/containerd/containerd/pull/4734

The solution proposed here has it's origins in multiple discussions during Linux Plumbers 2017 during and after the end of the containers microconference.
To the best of my knowledge this involved Aleksa, Stéphane, Eric, David, James, and myself.The original idea or a variant thereof has been discussed, again to the best of my knowledge, after a Linux conference in St. Petersburg in Russia in 2017 between Christoph, Tycho, and myself.
We've taken the time to implement a working version of this solution over the last weeks to the best of my abilities. Tycho has signed up for this sligthly crazy endeavour as well and he has helped with the conversion of the xattr codepaths and will be involved with others in converting additional filesystems.

Thanks!
Christian

Christian Brauner (39):
namespace: take lock_mount_hash() directly when changing flags
mount: make {lock,unlock}_mount_hash() static
namespace: only take read lock in do_reconfigure_mnt()
fs: split out functions to hold writers
fs: add attr_flags_to_mnt_flags helper
fs: add mount_setattr()
tests: add mount_setattr() selftests
fs: add id translation helpers
mount: attach mappings to mounts
capability: handle idmapped mounts
namei: make permission helpers idmapped mount aware
inode: make init and permission helpers idmapped mount aware
attr: handle idmapped mounts
acl: handle idmapped mounts
fs: add file_user_ns() helper
commoncap: handle idmapped mounts
stat: handle idmapped mounts
namei: handle idmapped mounts in may_*() helpers
namei: introduce struct renamedata
namei: prepare for idmapped mounts
open: handle idmapped mounts in do_truncate()
open: handle idmapped mounts
af_unix: handle idmapped mounts
utimes: handle idmapped mounts
fcntl: handle idmapped mounts
notify: handle idmapped mounts
init: handle idmapped mounts
ioctl: handle idmapped mounts
would_dump: handle idmapped mounts
exec: handle idmapped mounts
fs: make helpers idmap mount aware
apparmor: handle idmapped mounts
ima: handle idmapped mounts
fat: handle idmapped mounts
ext4: support idmapped mounts
ecryptfs: do not mount on top of idmapped mounts
overlayfs: do not mount on top of idmapped mounts
fs: introduce MOUNT_ATTR_IDMAP
tests: extend mount_setattr tests

Christoph Hellwig (1):
xfs: support idmapped mounts

Tycho Andersen (1):
xattr: handle idmapped mounts

 Documentation/filesystems/locking.rst         |    6 +-
 Documentation/filesystems/porting.rst         |    2 +
 Documentation/filesystems/vfs.rst             |   19 +-
 arch/alpha/kernel/syscalls/syscall.tbl        |    1 +
 arch/arm/tools/syscall.tbl                    |    1 +
 arch/arm64/include/asm/unistd32.h             |    2 +
 arch/ia64/kernel/syscalls/syscall.tbl         |    1 +
 arch/m68k/kernel/syscalls/syscall.tbl         |    1 +
 arch/microblaze/kernel/syscalls/syscall.tbl   |    1 +
 arch/mips/kernel/syscalls/syscall_n32.tbl     |    1 +
 arch/mips/kernel/syscalls/syscall_n64.tbl     |    1 +
 arch/mips/kernel/syscalls/syscall_o32.tbl     |    1 +
 arch/parisc/kernel/syscalls/syscall.tbl       |    1 +
 arch/powerpc/kernel/syscalls/syscall.tbl      |    1 +
 arch/powerpc/platforms/cell/spufs/inode.c     |    5 +-
 arch/s390/kernel/syscalls/syscall.tbl         |    1 +
 arch/sh/kernel/syscalls/syscall.tbl           |    1 +
 arch/sparc/kernel/syscalls/syscall.tbl        |    1 +
 arch/x86/entry/syscalls/syscall_32.tbl        |    1 +
 arch/x86/entry/syscalls/syscall_64.tbl        |    1 +
 arch/xtensa/kernel/syscalls/syscall.tbl       |    1 +
 drivers/android/binderfs.c                    |    6 +-
 drivers/base/devtmpfs.c                       |   12 +-
 fs/9p/acl.c                                   |    8 +-
 fs/9p/v9fs.h                                  |    3 +-
 fs/9p/v9fs_vfs.h                              |    2 +-
 fs/9p/vfs_inode.c                             |   36 +-
 fs/9p/vfs_inode_dotl.c                        |   39 +-
 fs/9p/xattr.c                                 |    1 +
 fs/adfs/adfs.h                                |    3 +-
 fs/adfs/inode.c                               |    5 +-
 fs/affs/affs.h                                |   10 +-
 fs/affs/inode.c                               |    7 +-
 fs/affs/namei.c                               |   15 +-
 fs/afs/dir.c                                  |   34 +-
 fs/afs/inode.c                                |    9 +-
 fs/afs/internal.h                             |    7 +-
 fs/afs/security.c                             |    2 +-
 fs/afs/xattr.c                                |    2 +
 fs/attr.c                                     |  124 +-
 fs/autofs/root.c                              |   13 +-
 fs/bad_inode.c                                |   36 +-
 fs/bfs/dir.c                                  |   12 +-
 fs/btrfs/acl.c                                |    5 +-
 fs/btrfs/ctree.h                              |    3 +-
 fs/btrfs/inode.c                              |   45 +-
 fs/btrfs/ioctl.c                              |   25 +-
 fs/btrfs/tests/btrfs-tests.c                  |    2 +-
 fs/btrfs/xattr.c                              |    2 +
 fs/cachefiles/interface.c                     |    4 +-
 fs/cachefiles/namei.c                         |   19 +-
 fs/cachefiles/xattr.c                         |   16 +-
 fs/ceph/acl.c                                 |    5 +-
 fs/ceph/dir.c                                 |   23 +-
 fs/ceph/inode.c                               |   17 +-
 fs/ceph/super.h                               |   12 +-
 fs/ceph/xattr.c                               |    1 +
 fs/cifs/cifsfs.c                              |    4 +-
 fs/cifs/cifsfs.h                              |   21 +-
 fs/cifs/dir.c                                 |    8 +-
 fs/cifs/inode.c                               |   26 +-
 fs/cifs/link.c                                |    3 +-
 fs/cifs/xattr.c                               |    1 +
 fs/coda/coda_linux.h                          |    6 +-
 fs/coda/dir.c                                 |   17 +-
 fs/coda/inode.c                               |    9 +-
 fs/coda/pioctl.c                              |    6 +-
 fs/configfs/configfs_internal.h               |    7 +-
 fs/configfs/dir.c                             |    3 +-
 fs/configfs/inode.c                           |    5 +-
 fs/configfs/symlink.c                         |    5 +-
 fs/coredump.c                                 |   14 +-
 fs/crypto/policy.c                            |    2 +-
 fs/debugfs/inode.c                            |    9 +-
 fs/ecryptfs/crypto.c                          |    4 +-
 fs/ecryptfs/inode.c                           |   80 +-
 fs/ecryptfs/main.c                            |    6 +
 fs/ecryptfs/mmap.c                            |    4 +-
 fs/efivarfs/file.c                            |    2 +-
 fs/efivarfs/inode.c                           |    4 +-
 fs/erofs/inode.c                              |    7 +-
 fs/erofs/internal.h                           |    5 +-
 fs/exec.c                                     |   12 +-
 fs/exfat/exfat_fs.h                           |    8 +-
 fs/exfat/file.c                               |   14 +-
 fs/exfat/namei.c                              |   14 +-
 fs/ext2/acl.c                                 |    5 +-
 fs/ext2/acl.h                                 |    3 +-
 fs/ext2/ext2.h                                |    5 +-
 fs/ext2/ialloc.c                              |    2 +-
 fs/ext2/inode.c                               |   15 +-
 fs/ext2/ioctl.c                               |    6 +-
 fs/ext2/namei.c                               |   22 +-
 fs/ext2/xattr_security.c                      |    1 +
 fs/ext2/xattr_trusted.c                       |    1 +
 fs/ext2/xattr_user.c                          |    1 +
 fs/ext4/acl.c                                 |    5 +-
 fs/ext4/acl.h                                 |    3 +-
 fs/ext4/ext4.h                                |   21 +-
 fs/ext4/ialloc.c                              |    7 +-
 fs/ext4/inode.c                               |   21 +-
 fs/ext4/ioctl.c                               |   19 +-
 fs/ext4/namei.c                               |   49 +-
 fs/ext4/super.c                               |    2 +-
 fs/ext4/xattr_hurd.c                          |    1 +
 fs/ext4/xattr_security.c                      |    1 +
 fs/ext4/xattr_trusted.c                       |    1 +
 fs/ext4/xattr_user.c                          |    1 +
 fs/f2fs/acl.c                                 |    5 +-
 fs/f2fs/acl.h                                 |    3 +-
 fs/f2fs/f2fs.h                                |    7 +-
 fs/f2fs/file.c                                |   35 +-
 fs/f2fs/namei.c                               |   23 +-
 fs/f2fs/xattr.c                               |    4 +-
 fs/fat/fat.h                                  |    6 +-
 fs/fat/file.c                                 |   24 +-
 fs/fat/namei_msdos.c                          |   12 +-
 fs/fat/namei_vfat.c                           |   15 +-
 fs/fcntl.c                                    |    3 +-
 fs/fuse/acl.c                                 |    3 +-
 fs/fuse/dir.c                                 |   45 +-
 fs/fuse/fuse_i.h                              |    4 +-
 fs/fuse/xattr.c                               |    2 +
 fs/gfs2/acl.c                                 |    5 +-
 fs/gfs2/acl.h                                 |    3 +-
 fs/gfs2/file.c                                |    4 +-
 fs/gfs2/inode.c                               |   59 +-
 fs/gfs2/inode.h                               |    3 +-
 fs/gfs2/xattr.c                               |    1 +
 fs/hfs/attr.c                                 |    1 +
 fs/hfs/dir.c                                  |   13 +-
 fs/hfs/hfs_fs.h                               |    2 +-
 fs/hfs/inode.c                                |    7 +-
 fs/hfsplus/dir.c                              |   25 +-
 fs/hfsplus/hfsplus_fs.h                       |    5 +-
 fs/hfsplus/inode.c                            |   16 +-
 fs/hfsplus/ioctl.c                            |    2 +-
 fs/hfsplus/xattr.c                            |    1 +
 fs/hfsplus/xattr_security.c                   |    1 +
 fs/hfsplus/xattr_trusted.c                    |    1 +
 fs/hfsplus/xattr_user.c                       |    1 +
 fs/hostfs/hostfs_kern.c                       |   29 +-
 fs/hpfs/hpfs_fn.h                             |    2 +-
 fs/hpfs/inode.c                               |    7 +-
 fs/hpfs/namei.c                               |   20 +-
 fs/hugetlbfs/inode.c                          |   31 +-
 fs/init.c                                     |   27 +-
 fs/inode.c                                    |   50 +-
 fs/internal.h                                 |    2 +-
 fs/jffs2/acl.c                                |    5 +-
 fs/jffs2/acl.h                                |    3 +-
 fs/jffs2/dir.c                                |   32 +-
 fs/jffs2/fs.c                                 |    7 +-
 fs/jffs2/os-linux.h                           |    2 +-
 fs/jffs2/security.c                           |    1 +
 fs/jffs2/xattr_trusted.c                      |    1 +
 fs/jffs2/xattr_user.c                         |    1 +
 fs/jfs/acl.c                                  |    5 +-
 fs/jfs/file.c                                 |    9 +-
 fs/jfs/ioctl.c                                |    2 +-
 fs/jfs/jfs_acl.h                              |    3 +-
 fs/jfs/jfs_inode.c                            |    2 +-
 fs/jfs/jfs_inode.h                            |    2 +-
 fs/jfs/namei.c                                |   21 +-
 fs/jfs/xattr.c                                |    2 +
 fs/kernfs/dir.c                               |    7 +-
 fs/kernfs/inode.c                             |   19 +-
 fs/kernfs/kernfs-internal.h                   |    9 +-
 fs/libfs.c                                    |   28 +-
 fs/minix/bitmap.c                             |    2 +-
 fs/minix/file.c                               |    7 +-
 fs/minix/inode.c                              |    6 +-
 fs/minix/minix.h                              |    3 +-
 fs/minix/namei.c                              |   24 +-
 fs/mount.h                                    |   10 -
 fs/namei.c                                    |  513 ++++--
 fs/namespace.c                                |  484 +++++-
 fs/nfs/dir.c                                  |   25 +-
 fs/nfs/inode.c                                |    9 +-
 fs/nfs/internal.h                             |   10 +-
 fs/nfs/namespace.c                            |   14 +-
 fs/nfs/nfs3_fs.h                              |    3 +-
 fs/nfs/nfs3acl.c                              |    3 +-
 fs/nfs/nfs4proc.c                             |    3 +
 fs/nfsd/nfs2acl.c                             |    4 +-
 fs/nfsd/nfs3acl.c                             |    4 +-
 fs/nfsd/nfs4acl.c                             |    4 +-
 fs/nfsd/nfs4recover.c                         |    6 +-
 fs/nfsd/nfsfh.c                               |    2 +-
 fs/nfsd/nfsproc.c                             |    2 +-
 fs/nfsd/vfs.c                                 |   47 +-
 fs/nilfs2/inode.c                             |   13 +-
 fs/nilfs2/ioctl.c                             |    2 +-
 fs/nilfs2/namei.c                             |   19 +-
 fs/nilfs2/nilfs.h                             |    4 +-
 fs/notify/fanotify/fanotify_user.c            |    2 +-
 fs/notify/inotify/inotify_user.c              |    3 +-
 fs/ntfs/inode.c                               |    6 +-
 fs/ntfs/inode.h                               |    3 +-
 fs/ocfs2/acl.c                                |    5 +-
 fs/ocfs2/acl.h                                |    3 +-
 fs/ocfs2/dlmfs/dlmfs.c                        |   17 +-
 fs/ocfs2/file.c                               |   17 +-
 fs/ocfs2/file.h                               |   11 +-
 fs/ocfs2/ioctl.c                              |    2 +-
 fs/ocfs2/namei.c                              |   21 +-
 fs/ocfs2/refcounttree.c                       |    4 +-
 fs/ocfs2/xattr.c                              |    3 +
 fs/omfs/dir.c                                 |   13 +-
 fs/omfs/file.c                                |    7 +-
 fs/omfs/inode.c                               |    2 +-
 fs/open.c                                     |   50 +-
 fs/orangefs/acl.c                             |    5 +-
 fs/orangefs/inode.c                           |   20 +-
 fs/orangefs/namei.c                           |   12 +-
 fs/orangefs/orangefs-kernel.h                 |   13 +-
 fs/orangefs/xattr.c                           |    1 +
 fs/overlayfs/copy_up.c                        |   20 +-
 fs/overlayfs/dir.c                            |   31 +-
 fs/overlayfs/file.c                           |    6 +-
 fs/overlayfs/inode.c                          |   26 +-
 fs/overlayfs/overlayfs.h                      |   44 +-
 fs/overlayfs/super.c                          |   19 +-
 fs/overlayfs/util.c                           |    4 +-
 fs/posix_acl.c                                |  101 +-
 fs/proc/base.c                                |   28 +-
 fs/proc/fd.c                                  |    5 +-
 fs/proc/fd.h                                  |    3 +-
 fs/proc/generic.c                             |   12 +-
 fs/proc/internal.h                            |    5 +-
 fs/proc/proc_net.c                            |    5 +-
 fs/proc/proc_sysctl.c                         |   15 +-
 fs/proc/root.c                                |    5 +-
 fs/proc_namespace.c                           |    3 +
 fs/ramfs/file-nommu.c                         |    9 +-
 fs/ramfs/inode.c                              |   18 +-
 fs/reiserfs/acl.h                             |    3 +-
 fs/reiserfs/inode.c                           |    7 +-
 fs/reiserfs/ioctl.c                           |    4 +-
 fs/reiserfs/namei.c                           |   21 +-
 fs/reiserfs/reiserfs.h                        |    3 +-
 fs/reiserfs/xattr.c                           |   12 +-
 fs/reiserfs/xattr.h                           |    3 +-
 fs/reiserfs/xattr_acl.c                       |    7 +-
 fs/reiserfs/xattr_security.c                  |    3 +-
 fs/reiserfs/xattr_trusted.c                   |    3 +-
 fs/reiserfs/xattr_user.c                      |    3 +-
 fs/remap_range.c                              |    7 +-
 fs/stat.c                                     |   26 +-
 fs/sysv/file.c                                |    7 +-
 fs/sysv/ialloc.c                              |    2 +-
 fs/sysv/itree.c                               |    6 +-
 fs/sysv/namei.c                               |   21 +-
 fs/sysv/sysv.h                                |    3 +-
 fs/tracefs/inode.c                            |    4 +-
 fs/ubifs/dir.c                                |   30 +-
 fs/ubifs/file.c                               |    5 +-
 fs/ubifs/ioctl.c                              |    2 +-
 fs/ubifs/ubifs.h                              |    5 +-
 fs/ubifs/xattr.c                              |    1 +
 fs/udf/file.c                                 |    9 +-
 fs/udf/ialloc.c                               |    2 +-
 fs/udf/namei.c                                |   24 +-
 fs/udf/symlink.c                              |    7 +-
 fs/ufs/ialloc.c                               |    2 +-
 fs/ufs/inode.c                                |    7 +-
 fs/ufs/namei.c                                |   19 +-
 fs/ufs/ufs.h                                  |    3 +-
 fs/utimes.c                                   |    4 +-
 fs/vboxsf/dir.c                               |   12 +-
 fs/vboxsf/utils.c                             |    9 +-
 fs/vboxsf/vfsmod.h                            |    8 +-
 fs/verity/enable.c                            |    2 +-
 fs/xattr.c                                    |  136 +-
 fs/xfs/xfs_acl.c                              |    5 +-
 fs/xfs/xfs_acl.h                              |    3 +-
 fs/xfs/xfs_file.c                             |    4 +-
 fs/xfs/xfs_inode.c                            |   26 +-
 fs/xfs/xfs_inode.h                            |   16 +-
 fs/xfs/xfs_ioctl.c                            |   23 +-
 fs/xfs/xfs_iops.c                             |   98 +-
 fs/xfs/xfs_iops.h                             |    3 +-
 fs/xfs/xfs_qm.c                               |    3 +-
 fs/xfs/xfs_super.c                            |    2 +-
 fs/xfs/xfs_symlink.c                          |    5 +-
 fs/xfs/xfs_symlink.h                          |    5 +-
 fs/xfs/xfs_xattr.c                            |    3 +-
 fs/zonefs/super.c                             |    9 +-
 include/linux/capability.h                    |   15 +-
 include/linux/fs.h                            |  158 +-
 include/linux/ima.h                           |   17 +-
 include/linux/lsm_hook_defs.h                 |   15 +-
 include/linux/lsm_hooks.h                     |    1 +
 include/linux/mount.h                         |    7 +
 include/linux/nfs_fs.h                        |    7 +-
 include/linux/posix_acl.h                     |   15 +-
 include/linux/posix_acl_xattr.h               |   12 +-
 include/linux/security.h                      |   46 +-
 include/linux/syscalls.h                      |    4 +
 include/linux/xattr.h                         |   30 +-
 include/uapi/asm-generic/unistd.h             |    4 +-
 include/uapi/linux/mount.h                    |   17 +
 ipc/mqueue.c                                  |    8 +-
 kernel/auditsc.c                              |    5 +-
 kernel/bpf/inode.c                            |   13 +-
 kernel/capability.c                           |   14 +-
 kernel/cgroup/cgroup.c                        |    2 +-
 kernel/sys.c                                  |    2 +-
 mm/madvise.c                                  |    4 +-
 mm/memcontrol.c                               |    2 +-
 mm/mincore.c                                  |    4 +-
 mm/shmem.c                                    |   48 +-
 net/socket.c                                  |    6 +-
 net/unix/af_unix.c                            |    4 +-
 security/apparmor/apparmorfs.c                |    3 +-
 security/apparmor/domain.c                    |   13 +-
 security/apparmor/file.c                      |    5 +-
 security/apparmor/lsm.c                       |   12 +-
 security/commoncap.c                          |  109 +-
 security/integrity/evm/evm_crypto.c           |   11 +-
 security/integrity/evm/evm_main.c             |    4 +-
 security/integrity/evm/evm_secfs.c            |    2 +-
 security/integrity/ima/ima.h                  |   19 +-
 security/integrity/ima/ima_api.c              |   10 +-
 security/integrity/ima/ima_appraise.c         |   22 +-
 security/integrity/ima/ima_asymmetric_keys.c  |    2 +-
 security/integrity/ima/ima_main.c             |   31 +-
 security/integrity/ima/ima_policy.c           |   19 +-
 security/integrity/ima/ima_queue_keys.c       |    2 +-
 security/security.c                           |   25 +-
 security/selinux/hooks.c                      |   22 +-
 security/smack/smack_lsm.c                    |   18 +-
 tools/include/uapi/asm-generic/unistd.h       |    4 +-
 tools/testing/selftests/Makefile              |    1 +
 .../selftests/mount_setattr/.gitignore        |    1 +
 .../testing/selftests/mount_setattr/Makefile  |    7 +
 tools/testing/selftests/mount_setattr/config  |    1 +
 .../mount_setattr/mount_setattr_test.c        | 1424 +++++++++++++++++
 338 files changed, 4718 insertions(+), 1731 deletions(-)
 create mode 100644 tools/testing/selftests/mount_setattr/.gitignore
 create mode 100644 tools/testing/selftests/mount_setattr/Makefile
 create mode 100644 tools/testing/selftests/mount_setattr/config
 create mode 100644 tools/testing/selftests/mount_setattr/mount_setattr_test.c

base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

Signature · 1 line
2.30.0
#1 lore raw

PATCH v5 1/42 namespace: take lock_mount_hash() directly when changing flags

Changing mount options always ends up taking lock_mount_hash() but when MNT_READONLY is requested and neither the mount nor the superblock are MNT_READONLY we end up taking the lock, dropping it, and retaking it to change the other mount attributes. Instead, let's acquire the lock once when changing the mount attributes. This simplifies the locking in these codepath, makes them easier to reason about and avoids having to reacquire the lock right after dropping it.

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Remove pointless __mnt_unmake_readonly() helper.
- Even though Christoph suggested to lockdep_assert_held() into places that require {lock,unlock}_mount_hash() it seems that seqlock's don't support it.

/* v3 */
unchanged

/* v4 */
unchanged

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/namespace.c | 22 ++++++++--------------
 1 file changed, 8 insertions(+), 14 deletions(-)
fs/namespace.c +8 -14
index d2db7dfe232b..dc782ddf603e 100644
@@ -463,7 +463,6 @@ static int mnt_make_readonly(struct mount *mnt)
 {
 	int ret = 0;
 
-	lock_mount_hash();
 	mnt->mnt.mnt_flags |= MNT_WRITE_HOLD;
 	/*
 	 * After storing MNT_WRITE_HOLD, we'll read the counters. This store
@@ -497,18 +496,9 @@ static int mnt_make_readonly(struct mount *mnt)
 	 */
 	smp_wmb();
 	mnt->mnt.mnt_flags &= ~MNT_WRITE_HOLD;
-	unlock_mount_hash();
 	return ret;
 }
 
-static int __mnt_unmake_readonly(struct mount *mnt)
-{
-	lock_mount_hash();
-	mnt->mnt.mnt_flags &= ~MNT_READONLY;
-	unlock_mount_hash();
-	return 0;
-}
-
 int sb_prepare_remount_readonly(struct super_block *sb)
 {
 	struct mount *mnt;
@@ -2511,7 +2501,8 @@ static int change_mount_ro_state(struct mount *mnt, unsigned int mnt_flags)
 	if (readonly_request)
 		return mnt_make_readonly(mnt);
 
-	return __mnt_unmake_readonly(mnt);
+	mnt->mnt.mnt_flags &= ~MNT_READONLY;
+	return 0;
 }
 
 /*
@@ -2520,11 +2511,9 @@ static int change_mount_ro_state(struct mount *mnt, unsigned int mnt_flags)
  */
 static void set_mount_attributes(struct mount *mnt, unsigned int mnt_flags)
 {
-	lock_mount_hash();
 	mnt_flags |= mnt->mnt.mnt_flags & ~MNT_USER_SETTABLE_MASK;
 	mnt->mnt.mnt_flags = mnt_flags;
 	touch_mnt_namespace(mnt->mnt_ns);
-	unlock_mount_hash();
 }
 
 static void mnt_warn_timestamp_expiry(struct path *mountpoint, struct vfsmount *mnt)
@@ -2570,9 +2559,11 @@ static int do_reconfigure_mnt(struct path *path, unsigned int mnt_flags)
 		return -EPERM;
 
 	down_write(&sb->s_umount);
+	lock_mount_hash();
 	ret = change_mount_ro_state(mnt, mnt_flags);
 	if (ret == 0)
 		set_mount_attributes(mnt, mnt_flags);
+	unlock_mount_hash();
 	up_write(&sb->s_umount);
 
 	mnt_warn_timestamp_expiry(path, &mnt->mnt);
@@ -2613,8 +2604,11 @@ static int do_remount(struct path *path, int ms_flags, int sb_flags,
 		err = -EPERM;
 		if (ns_capable(sb->s_user_ns, CAP_SYS_ADMIN)) {
 			err = reconfigure_super(fc);
-			if (!err)
+			if (!err) {
+				lock_mount_hash();
 				set_mount_attributes(mnt, mnt_flags);
+				unlock_mount_hash();
+			}
 		}
 		up_write(&sb->s_umount);
 	}
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
#2 lore raw

PATCH v5 2/42 mount: make {lock,unlock}_mount_hash() static

In reply to Christian Brauner

The lock_mount_hash() and unlock_mount_hash() helpers are never called outside a single file. Remove them from the header and make them static to reflect this fact. There's no need to have them callable from other places right now, as Christoph observed.

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Add a patch to make {lock,unlock)_mount_hash() static.

/* v3 */
unchanged

/* v4 */
unchanged

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/mount.h     | 10 ----------
 fs/namespace.c | 10 ++++++++++
 2 files changed, 10 insertions(+), 10 deletions(-)
fs/mount.h +0 -10
index ce6c376e0bc2..0b6e08cf8afb 100644
@@ -124,16 +124,6 @@ static inline void get_mnt_ns(struct mnt_namespace *ns)
 
 extern seqlock_t mount_lock;
 
-static inline void lock_mount_hash(void)
-{
-	write_seqlock(&mount_lock);
-}
-
-static inline void unlock_mount_hash(void)
-{
-	write_sequnlock(&mount_lock);
-}
-
 struct proc_mounts {
 	struct mnt_namespace *ns;
 	struct path root;
fs/namespace.c +10 -0
index dc782ddf603e..8cb2c132525a 100644
@@ -87,6 +87,16 @@ EXPORT_SYMBOL_GPL(fs_kobj);
  */
 __cacheline_aligned_in_smp DEFINE_SEQLOCK(mount_lock);
 
+static inline void lock_mount_hash(void)
+{
+	write_seqlock(&mount_lock);
+}
+
+static inline void unlock_mount_hash(void)
+{
+	write_sequnlock(&mount_lock);
+}
+
 static inline struct hlist_head *m_hash(struct vfsmount *mnt, struct dentry *dentry)
 {
 	unsigned long tmp = ((unsigned long)mnt / L1_CACHE_BYTES);
Signature · 1 line
2.30.0
#3 lore raw

PATCH v5 3/42 namespace: only take read lock in do_reconfigure_mnt()

In reply to Christian Brauner

do_reconfigure_mnt() used to take the down_write(&sb->s_umount) lock which seems unnecessary since we're not changing the superblock. We're only checking whether it is already read-only. Setting other mount attributes is protected by lock_mount_hash() afaict and not by s_umount.

So I think the history of down_write(&sb->s_umount) lock being taken when setting mount attributes dates back to the introduction of MNT_READONLY in [2]. Afaict, this introduced the concept of having read-only mounts in contrast to just having a read-only superblock. When it got introduced it was simply plumbed into do_remount() which already took down_write(&sb->s_umount) because it was only used to actually change the superblock before [2]. Afaict, it would've already been possible back then to only use down_read(&sb->s_umount) for MS_BIND | MS_REMOUNT since actual mount options were protected by the vfsmount lock already. But that would've meant special casing the locking for MS_BIND | MS_REMOUNT in do_remount() which people might not have considered worth it.
Then in [1] MS_BIND | MS_REMOUNT mount option changes were split out of do_remount() into do_reconfigure_mnt() but the down_write(&sb->s_umount) lock was simply copied over.
Now that we have this be a separate helper only take the down_read(&sb->s_umount) lock since we're only interested in checking whether the super block is currently read-only and blocking any writers from changing it. Essentially, checking that the super block is read-only has the advantage that we can avoid having to go into the slowpath and through MNT_WRITE_HOLD and can simply set the read-only flag on the mount in set_mount_attributes().

[1]: commit 43f5e655eff7 ("vfs: Separate changing mount flags full remount") [2]: commit 2e4b7fcd9260 ("[PATCH] r/o bind mounts: honor mount writer counts at remount")

---
/* v2 */
unchanged

/* v3 */
unchanged

/* v4 */
unchanged

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/namespace.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
fs/namespace.c +6 -6
index 8cb2c132525a..5177bf6d810e 100644
@@ -2515,10 +2515,6 @@ static int change_mount_ro_state(struct mount *mnt, unsigned int mnt_flags)
 	return 0;
 }
 
-/*
- * Update the user-settable attributes on a mount.  The caller must hold
- * sb->s_umount for writing.
- */
 static void set_mount_attributes(struct mount *mnt, unsigned int mnt_flags)
 {
 	mnt_flags |= mnt->mnt.mnt_flags & ~MNT_USER_SETTABLE_MASK;
@@ -2568,13 +2564,17 @@ static int do_reconfigure_mnt(struct path *path, unsigned int mnt_flags)
 	if (!can_change_locked_flags(mnt, mnt_flags))
 		return -EPERM;
 
-	down_write(&sb->s_umount);
+	/*
+	 * We're only checking whether the superblock is read-only not changing
+	 * it, so only take down_read(&sb->s_umount).
+	 */
+	down_read(&sb->s_umount);
 	lock_mount_hash();
 	ret = change_mount_ro_state(mnt, mnt_flags);
 	if (ret == 0)
 		set_mount_attributes(mnt, mnt_flags);
 	unlock_mount_hash();
-	up_write(&sb->s_umount);
+	up_read(&sb->s_umount);
 
 	mnt_warn_timestamp_expiry(path, &mnt->mnt);
 
Signature · 1 line
2.30.0
#4 lore raw

PATCH v5 4/42 fs: split out functions to hold writers

In reply to Christian Brauner

When a mount is marked read-only we set MNT_WRITE_HOLD on it if there aren't currently any active writers. Split this logic out into simple helpers that we can use in follow-up patches.

---
/* v2 */
patch not present

/* v3 */
patch not present

/* v4 */
patch introduced

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/namespace.c | 24 ++++++++++++++++++------
 1 file changed, 18 insertions(+), 6 deletions(-)
fs/namespace.c +18 -6
index 5177bf6d810e..4aab396e5f21 100644
@@ -469,10 +469,8 @@ void mnt_drop_write_file(struct file *file)
 }
 EXPORT_SYMBOL(mnt_drop_write_file);
 
-static int mnt_make_readonly(struct mount *mnt)
+static inline int mnt_hold_writers(struct mount *mnt)
 {
-	int ret = 0;
-
 	mnt->mnt.mnt_flags |= MNT_WRITE_HOLD;
 	/*
 	 * After storing MNT_WRITE_HOLD, we'll read the counters. This store
@@ -497,15 +495,29 @@ static int mnt_make_readonly(struct mount *mnt)
 	 * we're counting up here.
 	 */
 	if (mnt_get_writers(mnt) > 0)
-		ret = -EBUSY;
-	else
-		mnt->mnt.mnt_flags |= MNT_READONLY;
+		return -EBUSY;
+
+	return 0;
+}
+
+static inline void mnt_unhold_writers(struct mount *mnt)
+{
 	/*
 	 * MNT_READONLY must become visible before ~MNT_WRITE_HOLD, so writers
 	 * that become unheld will see MNT_READONLY.
 	 */
 	smp_wmb();
 	mnt->mnt.mnt_flags &= ~MNT_WRITE_HOLD;
+}
+
+static int mnt_make_readonly(struct mount *mnt)
+{
+	int ret;
+
+	ret = mnt_hold_writers(mnt);
+	if (!ret)
+		mnt->mnt.mnt_flags |= MNT_READONLY;
+	mnt_unhold_writers(mnt);
 	return ret;
 }
 
Signature · 1 line
2.30.0
#5 lore raw

PATCH v5 5/42 fs: add attr_flags_to_mnt_flags helper

In reply to Christian Brauner

Add a simple helper to translate uapi MOUNT_ATTR_* flags to MNT_* flags which we will use in follow-up patches too.

---
/* v2 */
patch not present

/* v3 */
patch not present

/* v4 */
patch introduced

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

- Christoph Hellwig <[email protected]>:
- Align "\" in defines further in.

---
 fs/namespace.c | 40 ++++++++++++++++++++++++----------------
 1 file changed, 24 insertions(+), 16 deletions(-)
fs/namespace.c +24 -16
index 4aab396e5f21..a1cfcab217e1 100644
@@ -3453,6 +3453,28 @@ SYSCALL_DEFINE5(mount, char __user *, dev_name, char __user *, dir_name,
 	return ret;
 }
 
+#define FSMOUNT_VALID_FLAGS \
+	(MOUNT_ATTR_RDONLY | MOUNT_ATTR_NOSUID | MOUNT_ATTR_NODEV | \
+	 MOUNT_ATTR_NOEXEC | MOUNT_ATTR__ATIME | MOUNT_ATTR_NODIRATIME)
+
+static unsigned int attr_flags_to_mnt_flags(u64 attr_flags)
+{
+	unsigned int mnt_flags = 0;
+
+	if (attr_flags & MOUNT_ATTR_RDONLY)
+		mnt_flags |= MNT_READONLY;
+	if (attr_flags & MOUNT_ATTR_NOSUID)
+		mnt_flags |= MNT_NOSUID;
+	if (attr_flags & MOUNT_ATTR_NODEV)
+		mnt_flags |= MNT_NODEV;
+	if (attr_flags & MOUNT_ATTR_NOEXEC)
+		mnt_flags |= MNT_NOEXEC;
+	if (attr_flags & MOUNT_ATTR_NODIRATIME)
+		mnt_flags |= MNT_NODIRATIME;
+
+	return mnt_flags;
+}
+
 /*
  * Create a kernel mount representation for a new, prepared superblock
  * (specified by fs_fd) and attach to an open_tree-like file descriptor.
@@ -3475,24 +3497,10 @@ SYSCALL_DEFINE3(fsmount, int, fs_fd, unsigned int, flags,
 	if ((flags & ~(FSMOUNT_CLOEXEC)) != 0)
 		return -EINVAL;
 
-	if (attr_flags & ~(MOUNT_ATTR_RDONLY |
-			   MOUNT_ATTR_NOSUID |
-			   MOUNT_ATTR_NODEV |
-			   MOUNT_ATTR_NOEXEC |
-			   MOUNT_ATTR__ATIME |
-			   MOUNT_ATTR_NODIRATIME))
+	if (attr_flags & ~FSMOUNT_VALID_FLAGS)
 		return -EINVAL;
 
-	if (attr_flags & MOUNT_ATTR_RDONLY)
-		mnt_flags |= MNT_READONLY;
-	if (attr_flags & MOUNT_ATTR_NOSUID)
-		mnt_flags |= MNT_NOSUID;
-	if (attr_flags & MOUNT_ATTR_NODEV)
-		mnt_flags |= MNT_NODEV;
-	if (attr_flags & MOUNT_ATTR_NOEXEC)
-		mnt_flags |= MNT_NOEXEC;
-	if (attr_flags & MOUNT_ATTR_NODIRATIME)
-		mnt_flags |= MNT_NODIRATIME;
+	mnt_flags = attr_flags_to_mnt_flags(attr_flags);
 
 	switch (attr_flags & MOUNT_ATTR__ATIME) {
 	case MOUNT_ATTR_STRICTATIME:
Signature · 1 line
2.30.0
#6 lore raw

PATCH v5 6/42 fs: add mount_setattr()

In reply to Christian Brauner

This implements the missing mount_setattr() syscall. While the new mount api allows to change the properties of a superblock there is currently no way to change the properties of a mount or a mount tree using file descriptors which the new mount api is based on. In addition the old mount api has the restriction that mount options cannot be applied recursively. This hasn't changed since changing mount options on a per-mount basis was implemented in [1] and has been a frequent request not just for convenience but also for security reasons. The legacy mount syscall is unable to accommodate this behavior without introducing a whole new set of flags because MS_REC | MS_REMOUNT | MS_BIND | MS_RDONLY | MS_NOEXEC | [...] only apply the mount option to the topmost mount. Changing MS_REC to apply to the whole mount tree would mean introducing a significant uapi change and would likely cause significant regressions.

The new mount_setattr() syscall allows to recursively clear and set mount options in one shot. Multiple calls to change mount options requesting the same changes are idempotent:

int mount_setattr(int dfd, const char *path, unsigned flags,

                  struct mount_attr *uattr, size_t usize);

Flags to modify path resolution behavior are specified in the @flags argument. Currently, AT_EMPTY_PATH, AT_RECURSIVE, AT_SYMLINK_NOFOLLOW, and AT_NO_AUTOMOUNT are supported. If useful, additional lookup flags to restrict path resolution as introduced with openat2() might be supported in the future.

The mount_setattr() syscall can be expected to grow over time and is designed with extensibility in mind. It follows the extensible syscall pattern we have used with other syscalls such as openat2(), clone3(), sched_{set,get}attr(), and others.
The set of mount options is passed in the uapi struct mount_attr which currently has the following layout:

struct mount_attr {
	__u64 attr_set;
	__u64 attr_clr;
	__u64 propagation;
};

The @attr_set and @attr_clr members are used to clear and set mount options. This way a user can e.g. request that a set of flags is to be raised such as turning mounts readonly by raising MOUNT_ATTR_RDONLY in @attr_set while at the same time requesting that another set of flags is to be lowered such as removing noexec from a mount tree by specifying MOUNT_ATTR_NOEXEC in @attr_clr.

Note, since the MOUNT_ATTR_<atime> values are an enum starting from 0, not a bitmap, users wanting to transition to a different atime setting cannot simply specify the atime setting in @attr_set, but must also specify MOUNT_ATTR__ATIME in the @attr_clr field. So we ensure that MOUNT_ATTR__ATIME can't be partially set in @attr_clr and that @attr_set can't have any atime bits set if MOUNT_ATTR__ATIME isn't set in @attr_clr.

The @propagation field lets callers specify the propagation type of a mount tree. Propagation is a single property that has four different settings and as such is not really a flag argument but an enum. Specifically, it would be unclear what setting and clearing propagation settings in combination would amount to. The legacy mount() syscall thus forbids the combination of multiple propagation settings too. The goal is to keep the semantics of mount propagation somewhat simple as they are overly complex as it is.

[1]: commit 2e4b7fcd9260 ("[PATCH] r/o bind mounts: honor mount writer counts at remount")

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Split into multiple helpers.

/* v3 */
- kernel test robot <[email protected]>:
- Fix unknown __u64 type by including linux/types.h in linux/mount.h.

/* v4 */
- Christoph Hellwig <[email protected]>:
- Make sure lines wrap at 80 chars.
- Move struct mount_kattr out of the internal.h header and completely into fs/namespace.c as it's not used outside of that file.
- Add missing space between ( and { when initializing mount_kattr.
- Split flag validation and calculation into separate preparatory patch.
- Simplify flag validation in build_mount_kattr() to avoid upper_32_bits() and lower_32_bits() calls. This will also lead to better code generation.
- Remove new propagation enums and simply use the old flags.
- Strictly adhere to 80 char limit.
- Restructure the time setting code in build_mount_kattr().

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

- Christoph Hellwig <[email protected]>:
- Simplify mount propagation flag checking.
- Wrap overly long line.

---
 arch/alpha/kernel/syscalls/syscall.tbl      |   1 +
 arch/arm/tools/syscall.tbl                  |   1 +
 arch/arm64/include/asm/unistd32.h           |   2 +
 arch/ia64/kernel/syscalls/syscall.tbl       |   1 +
 arch/m68k/kernel/syscalls/syscall.tbl       |   1 +
 arch/microblaze/kernel/syscalls/syscall.tbl |   1 +
 arch/mips/kernel/syscalls/syscall_n32.tbl   |   1 +
 arch/mips/kernel/syscalls/syscall_n64.tbl   |   1 +
 arch/mips/kernel/syscalls/syscall_o32.tbl   |   1 +
 arch/parisc/kernel/syscalls/syscall.tbl     |   1 +
 arch/powerpc/kernel/syscalls/syscall.tbl    |   1 +
 arch/s390/kernel/syscalls/syscall.tbl       |   1 +
 arch/sh/kernel/syscalls/syscall.tbl         |   1 +
 arch/sparc/kernel/syscalls/syscall.tbl      |   1 +
 arch/x86/entry/syscalls/syscall_32.tbl      |   1 +
 arch/x86/entry/syscalls/syscall_64.tbl      |   1 +
 arch/xtensa/kernel/syscalls/syscall.tbl     |   1 +
 fs/namespace.c                              | 260 ++++++++++++++++++++
 include/linux/syscalls.h                    |   4 +
 include/uapi/asm-generic/unistd.h           |   4 +-
 include/uapi/linux/mount.h                  |  14 ++
 tools/include/uapi/asm-generic/unistd.h     |   4 +-
 22 files changed, 302 insertions(+), 2 deletions(-)
arch/alpha/kernel/syscalls/syscall.tbl +1 -0
index a6617067dbe6..02f0244e005c 100644
@@ -481,3 +481,4 @@
 549	common	faccessat2			sys_faccessat2
 550	common	process_madvise			sys_process_madvise
 551	common	epoll_pwait2			sys_epoll_pwait2
+552	common	mount_setattr			sys_mount_setattr
arch/arm/tools/syscall.tbl +1 -0
index 20e1170e2e0a..dcc1191291a2 100644
@@ -455,3 +455,4 @@
 439	common	faccessat2			sys_faccessat2
 440	common	process_madvise			sys_process_madvise
 441	common	epoll_pwait2			sys_epoll_pwait2
+442	common	mount_setattr			sys_mount_setattr
arch/arm64/include/asm/unistd32.h +2 -0
index cccfbbefbf95..3d874f624056 100644
@@ -891,6 +891,8 @@ __SYSCALL(__NR_faccessat2, sys_faccessat2)
 __SYSCALL(__NR_process_madvise, sys_process_madvise)
 #define __NR_epoll_pwait2 441
 __SYSCALL(__NR_epoll_pwait2, compat_sys_epoll_pwait2)
+#define __NR_mount_setattr 442
+__SYSCALL(__NR_mount_setattr, sys_mount_setattr)
 
 /*
  * Please add new compat syscalls above this comment and update
arch/ia64/kernel/syscalls/syscall.tbl +1 -0
index bfc00f2bd437..d89231166e19 100644
@@ -362,3 +362,4 @@
 439	common	faccessat2			sys_faccessat2
 440	common	process_madvise			sys_process_madvise
 441	common	epoll_pwait2			sys_epoll_pwait2
+442	common	mount_setattr			sys_mount_setattr
arch/m68k/kernel/syscalls/syscall.tbl +1 -0
index 7fe4e45c864c..72bde6707dd3 100644
@@ -441,3 +441,4 @@
 439	common	faccessat2			sys_faccessat2
 440	common	process_madvise			sys_process_madvise
 441	common	epoll_pwait2			sys_epoll_pwait2
+442	common	mount_setattr			sys_mount_setattr
arch/microblaze/kernel/syscalls/syscall.tbl +1 -0
index a522adf194ab..d603a5ec9338 100644
@@ -447,3 +447,4 @@
 439	common	faccessat2			sys_faccessat2
 440	common	process_madvise			sys_process_madvise
 441	common	epoll_pwait2			sys_epoll_pwait2
+442	common	mount_setattr			sys_mount_setattr
arch/mips/kernel/syscalls/syscall_n32.tbl +1 -0
index 0f03ad223f33..8fd8c1790941 100644
@@ -380,3 +380,4 @@
 439	n32	faccessat2			sys_faccessat2
 440	n32	process_madvise			sys_process_madvise
 441	n32	epoll_pwait2			compat_sys_epoll_pwait2
+442	n32	mount_setattr			sys_mount_setattr
arch/mips/kernel/syscalls/syscall_n64.tbl +1 -0
index 91649690b52f..169f21438065 100644
@@ -356,3 +356,4 @@
 439	n64	faccessat2			sys_faccessat2
 440	n64	process_madvise			sys_process_madvise
 441	n64	epoll_pwait2			sys_epoll_pwait2
+442	n64	mount_setattr			sys_mount_setattr
arch/mips/kernel/syscalls/syscall_o32.tbl +1 -0
index 4bad0c40aed6..090d29ca80ff 100644
@@ -429,3 +429,4 @@
 439	o32	faccessat2			sys_faccessat2
 440	o32	process_madvise			sys_process_madvise
 441	o32	epoll_pwait2			sys_epoll_pwait2		compat_sys_epoll_pwait2
+442	o32	mount_setattr			sys_mount_setattr
arch/parisc/kernel/syscalls/syscall.tbl +1 -0
index 6bcc31966b44..271a92519683 100644
@@ -439,3 +439,4 @@
 439	common	faccessat2			sys_faccessat2
 440	common	process_madvise			sys_process_madvise
 441	common	epoll_pwait2			sys_epoll_pwait2		compat_sys_epoll_pwait2
+442	common	mount_setattr			sys_mount_setattr
arch/powerpc/kernel/syscalls/syscall.tbl +1 -0
index f744eb5cba88..72e5aa67ab8a 100644
@@ -531,3 +531,4 @@
 439	common	faccessat2			sys_faccessat2
 440	common	process_madvise			sys_process_madvise
 441	common	epoll_pwait2			sys_epoll_pwait2		compat_sys_epoll_pwait2
+442	common	mount_setattr			sys_mount_setattr
arch/s390/kernel/syscalls/syscall.tbl +1 -0
index d443423495e5..3abef2144dac 100644
@@ -444,3 +444,4 @@
 439  common	faccessat2		sys_faccessat2			sys_faccessat2
 440  common	process_madvise		sys_process_madvise		sys_process_madvise
 441  common	epoll_pwait2		sys_epoll_pwait2		compat_sys_epoll_pwait2
+442  common	mount_setattr		sys_mount_setattr		sys_mount_setattr
arch/sh/kernel/syscalls/syscall.tbl +1 -0
index 9df40ac0ebc0..d08eebad6b7f 100644
@@ -444,3 +444,4 @@
 439	common	faccessat2			sys_faccessat2
 440	common	process_madvise			sys_process_madvise
 441	common	epoll_pwait2			sys_epoll_pwait2
+442	common	mount_setattr			sys_mount_setattr
arch/sparc/kernel/syscalls/syscall.tbl +1 -0
index 40d8c7cd8298..84403a99039c 100644
@@ -487,3 +487,4 @@
 439	common	faccessat2			sys_faccessat2
 440	common	process_madvise			sys_process_madvise
 441	common	epoll_pwait2			sys_epoll_pwait2		compat_sys_epoll_pwait2
+442	common	mount_setattr			sys_mount_setattr
arch/x86/entry/syscalls/syscall_32.tbl +1 -0
index 874aeacde2dd..a1c9f496fca6 100644
@@ -446,3 +446,4 @@
 439	i386	faccessat2		sys_faccessat2
 440	i386	process_madvise		sys_process_madvise
 441	i386	epoll_pwait2		sys_epoll_pwait2		compat_sys_epoll_pwait2
+442	i386	mount_setattr		sys_mount_setattr
arch/x86/entry/syscalls/syscall_64.tbl +1 -0
index 78672124d28b..7bf01cbe582f 100644
@@ -363,6 +363,7 @@
 439	common	faccessat2		sys_faccessat2
 440	common	process_madvise		sys_process_madvise
 441	common	epoll_pwait2		sys_epoll_pwait2
+442	common	mount_setattr		sys_mount_setattr
 
 #
 # Due to a historical design error, certain syscalls are numbered differently
arch/xtensa/kernel/syscalls/syscall.tbl +1 -0
index 46116a28eeed..365a9b849224 100644
@@ -412,3 +412,4 @@
 439	common	faccessat2			sys_faccessat2
 440	common	process_madvise			sys_process_madvise
 441	common	epoll_pwait2			sys_epoll_pwait2
+442	common	mount_setattr			sys_mount_setattr
fs/namespace.c +260 -0

278 lines not on this page · raw message

include/linux/syscalls.h +4 -0
index 7688bc983de5..cd7b5c817ba2 100644
@@ -68,6 +68,7 @@ union bpf_attr;
 struct io_uring_params;
 struct clone_args;
 struct open_how;
+struct mount_attr;
 
 #include <linux/types.h>
 #include <linux/aio_abi.h>
@@ -1028,6 +1029,9 @@ asmlinkage long sys_open_tree(int dfd, const char __user *path, unsigned flags);
 asmlinkage long sys_move_mount(int from_dfd, const char __user *from_path,
 			       int to_dfd, const char __user *to_path,
 			       unsigned int ms_flags);
+asmlinkage long sys_mount_setattr(int dfd, const char __user *path,
+				  unsigned int flags,
+				  struct mount_attr __user *uattr, size_t usize);
 asmlinkage long sys_fsopen(const char __user *fs_name, unsigned int flags);
 asmlinkage long sys_fsconfig(int fs_fd, unsigned int cmd, const char __user *key,
 			     const void __user *value, int aux);
include/uapi/asm-generic/unistd.h +3 -1
index 728752917785..ce58cff99b66 100644
@@ -861,9 +861,11 @@ __SYSCALL(__NR_faccessat2, sys_faccessat2)
 __SYSCALL(__NR_process_madvise, sys_process_madvise)
 #define __NR_epoll_pwait2 441
 __SC_COMP(__NR_epoll_pwait2, sys_epoll_pwait2, compat_sys_epoll_pwait2)
+#define __NR_mount_setattr 442
+__SYSCALL(__NR_mount_setattr, sys_mount_setattr)
 
 #undef __NR_syscalls
-#define __NR_syscalls 442
+#define __NR_syscalls 443
 
 /*
  * 32 bit systems traditionally used different
include/uapi/linux/mount.h +14 -0
index dd8306ea336c..2255624e91c8 100644
@@ -1,6 +1,8 @@
 #ifndef _UAPI_LINUX_MOUNT_H
 #define _UAPI_LINUX_MOUNT_H
 
+#include <linux/types.h>
+
 /*
  * These are the fs-independent mount-flags: up to 32 flags are supported
  *
@@ -118,4 +120,16 @@ enum fsconfig_command {
 #define MOUNT_ATTR_STRICTATIME	0x00000020 /* - Always perform atime updates */
 #define MOUNT_ATTR_NODIRATIME	0x00000080 /* Do not update directory access times */
 
+/*
+ * mount_setattr()
+ */
+struct mount_attr {
+	__u64 attr_set;
+	__u64 attr_clr;
+	__u64 propagation;
+};
+
+/* List of all mount_attr versions. */
+#define MOUNT_ATTR_SIZE_VER0	24 /* sizeof first published struct */
+
 #endif /* _UAPI_LINUX_MOUNT_H */
tools/include/uapi/asm-generic/unistd.h +3 -1
index 728752917785..ce58cff99b66 100644
@@ -861,9 +861,11 @@ __SYSCALL(__NR_faccessat2, sys_faccessat2)
 __SYSCALL(__NR_process_madvise, sys_process_madvise)
 #define __NR_epoll_pwait2 441
 __SC_COMP(__NR_epoll_pwait2, sys_epoll_pwait2, compat_sys_epoll_pwait2)
+#define __NR_mount_setattr 442
+__SYSCALL(__NR_mount_setattr, sys_mount_setattr)
 
 #undef __NR_syscalls
-#define __NR_syscalls 442
+#define __NR_syscalls 443
 
 /*
  * 32 bit systems traditionally used different
Signature · 1 line
2.30.0
#7 lore raw

PATCH v5 7/42 tests: add mount_setattr() selftests

In reply to Christian Brauner

Add a range of selftests for the new mount_setattr() syscall to verify that it works as expected. This tests that:
- no invalid flags can be specified
- changing properties of a single mount works and leaves other mounts in the mount tree unchanged
- changing a mount tre to read-only when one of the mounts has writers fails and leaves the whole mount tree unchanged
- changing mount properties from multiple threads works
- changing atime settings works
- changing mount propagation works
- changing the mount options of a mount tree where the individual mounts in the tree have different mount options only changes the flags that were requested to change
- changing mount options from another mount namespace fails
- changing mount options from another user namespace fails

[==========] Running 9 tests from 2 test cases.
[ RUN ] mount_setattr.invalid_attributes
[ OK ] mount_setattr.invalid_attributes
[ RUN ] mount_setattr.basic
[ OK ] mount_setattr.basic
[ RUN ] mount_setattr.basic_recursive
[ OK ] mount_setattr.basic_recursive
[ RUN ] mount_setattr.mount_has_writers
[ OK ] mount_setattr.mount_has_writers
[ RUN ] mount_setattr.mixed_mount_options
[ OK ] mount_setattr.mixed_mount_options
[ RUN ] mount_setattr.time_changes
[ OK ] mount_setattr.time_changes
[ RUN ] mount_setattr.multi_threaded
[ OK ] mount_setattr.multi_threaded
[ RUN ] mount_setattr.wrong_user_namespace
[ OK ] mount_setattr.wrong_user_namespace
[ RUN ] mount_setattr.wrong_mount_namespace
[ OK ] mount_setattr.wrong_mount_namespace
[==========] 9 / 9 tests passed.
[ PASSED ]

---
/* v2 */
unchanged

/* v3 */
unchanged

/* v4 */
unchanged

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 tools/testing/selftests/Makefile              |   1 +
 .../selftests/mount_setattr/.gitignore        |   1 +
 .../testing/selftests/mount_setattr/Makefile  |   7 +
 tools/testing/selftests/mount_setattr/config  |   1 +
 .../mount_setattr/mount_setattr_test.c        | 941 ++++++++++++++++++
 5 files changed, 951 insertions(+)
 create mode 100644 tools/testing/selftests/mount_setattr/.gitignore
 create mode 100644 tools/testing/selftests/mount_setattr/Makefile
 create mode 100644 tools/testing/selftests/mount_setattr/config
 create mode 100644 tools/testing/selftests/mount_setattr/mount_setattr_test.c
tools/testing/selftests/Makefile +1 -0
index afbab4aeef3c..ca19bbcd4bc5 100644
@@ -33,6 +33,7 @@ TARGETS += memfd
 TARGETS += memory-hotplug
 TARGETS += mincore
 TARGETS += mount
+TARGETS += mount_setattr
 TARGETS += mqueue
 TARGETS += net
 TARGETS += net/forwarding
tools/testing/selftests/mount_setattr/.gitignore new file +1 -0
new file mode 100644
index 000000000000..5f74d8488472
@@ -0,0 +1 @@
+mount_setattr_test
tools/testing/selftests/mount_setattr/Makefile new file +7 -0
new file mode 100644
index 000000000000..2250f7dcb81e
@@ -0,0 +1,7 @@
+# SPDX-License-Identifier: GPL-2.0
+# Makefile for mount selftests.
+CFLAGS = -g -I../../../../usr/include/ -Wall -O2 -pthread
+
+TEST_GEN_FILES += mount_setattr_test
+
+include ../lib.mk
tools/testing/selftests/mount_setattr/config new file +1 -0
new file mode 100644
index 000000000000..416bd53ce982
@@ -0,0 +1 @@
+CONFIG_USER_NS=y
tools/testing/selftests/mount_setattr/mount_setattr_test.c new file +941 -0

941 lines not on this page · raw message

Signature · 1 line
2.30.0
#8 lore raw

PATCH v5 8/42 fs: add id translation helpers

In reply to Christian Brauner

Add simple helpers to make it easy to map kuids into and from idmapped mounts. We provide simple wrappers that filesystems can use to e.g. initialize inodes similar to i_{uid,gid}_read() and i_{uid,gid}_write(). Accessing an inode through an idmapped mount will require the inode to be mapped according to the mount's user namespace. If the fsids are used to compare against inodes or to initialize inodes they are required to be shifted from the mount's user namespace. Passing the initial user namespace to these helpers makes them a nop and so any non-idmapped paths will not be impacted.

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Get rid of the ifdefs and the config option that hid idmapped mounts.

/* v3 */
unchanged

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 include/linux/fs.h | 47 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 47 insertions(+)
include/linux/fs.h +47 -0
index fd47deea7c17..1da4c21fb588 100644
@@ -39,6 +39,7 @@
 #include <linux/fs_types.h>
 #include <linux/build_bug.h>
 #include <linux/stddef.h>
+#include <linux/cred.h>
 
 #include <asm/byteorder.h>
 #include <uapi/linux/fs.h>
@@ -1572,6 +1573,52 @@ static inline void i_gid_write(struct inode *inode, gid_t gid)
 	inode->i_gid = make_kgid(inode->i_sb->s_user_ns, gid);
 }
 
+static inline kuid_t kuid_into_mnt(struct user_namespace *mnt_userns,
+				   kuid_t kuid)
+{
+	return make_kuid(mnt_userns, __kuid_val(kuid));
+}
+
+static inline kgid_t kgid_into_mnt(struct user_namespace *mnt_userns,
+				   kgid_t kgid)
+{
+	return make_kgid(mnt_userns, __kgid_val(kgid));
+}
+
+static inline kuid_t i_uid_into_mnt(struct user_namespace *mnt_userns,
+				    const struct inode *inode)
+{
+	return kuid_into_mnt(mnt_userns, inode->i_uid);
+}
+
+static inline kgid_t i_gid_into_mnt(struct user_namespace *mnt_userns,
+				    const struct inode *inode)
+{
+	return kgid_into_mnt(mnt_userns, inode->i_gid);
+}
+
+static inline kuid_t kuid_from_mnt(struct user_namespace *mnt_userns,
+				   kuid_t kuid)
+{
+	return KUIDT_INIT(from_kuid(mnt_userns, kuid));
+}
+
+static inline kgid_t kgid_from_mnt(struct user_namespace *mnt_userns,
+				   kgid_t kgid)
+{
+	return KGIDT_INIT(from_kgid(mnt_userns, kgid));
+}
+
+static inline kuid_t fsuid_into_mnt(struct user_namespace *mnt_userns)
+{
+	return kuid_from_mnt(mnt_userns, current_fsuid());
+}
+
+static inline kgid_t fsgid_into_mnt(struct user_namespace *mnt_userns)
+{
+	return kgid_from_mnt(mnt_userns, current_fsgid());
+}
+
 extern struct timespec64 current_time(struct inode *inode);
 
 /*
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
#9 lore raw

PATCH v5 9/42 mount: attach mappings to mounts

In reply to Christian Brauner

In order to support per-mount idmappings vfsmounts will be marked with user namespaces. The idmapping associated with that 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.

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 we can allow the user namespace an already idmapped mounts has been marked with to be replaced with another one. The permission checking would then take into account whether the caller is privileged in the user namespace the mount is currently marked with and that is about to be replaced with another one. For now, we will enforce in later patches that once a mount has been idmapped it can't be remapped. This keeps permission checking and life-cycle management simple, especially since users can always create a new mount with a different idmapping anyway.

The idea to attach user namespaces to vfsmounts has been floated around in various forms at Linux Plumbers in ~2018 with the original idea tracing back to a discussion during a conference in St. Petersburg between Christoph, Tycho, and myself.

---
/* v2 */
patch introduced
- Christoph Hellwig <[email protected]>:
- Split internal implementation into separate patch and move syscall implementation later.

/* v3 */
- David Howells <[email protected]>:
- Remove MNT_IDMAPPED flag. We can simply check the pointer and use smp_load_acquire() in later patches.

- Tycho Andersen <[email protected]>:
- Use READ_ONCE() in mnt_user_ns().

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

- Christoph Hellwig <[email protected]>:
- Drop the READ_ONCE() from this patch. At this point in the series we don't allowing changing the vfsmount's userns. The infra to do that is only introduced as almost the last patch in the series and there we immediately use smp_load_acquire() and smp_store_release().

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/namespace.c        | 9 +++++++++
 include/linux/fs.h    | 1 +
 include/linux/mount.h | 6 ++++++
 3 files changed, 16 insertions(+)
fs/namespace.c +9 -0
index 6efae2681bcd..ceb2943f8458 100644
@@ -228,6 +228,7 @@ static struct mount *alloc_vfsmnt(const char *name)
 		INIT_HLIST_NODE(&mnt->mnt_mp_list);
 		INIT_LIST_HEAD(&mnt->mnt_umounting);
 		INIT_HLIST_HEAD(&mnt->mnt_stuck_children);
+		mnt->mnt.mnt_userns = &init_user_ns;
 	}
 	return mnt;
 
@@ -567,6 +568,11 @@ int sb_prepare_remount_readonly(struct super_block *sb)
 
 static void free_vfsmnt(struct mount *mnt)
 {
+	struct user_namespace *mnt_userns;
+
+	mnt_userns = mnt_user_ns(&mnt->mnt);
+	if (mnt_userns != &init_user_ns)
+		put_user_ns(mnt_userns);
 	kfree_const(mnt->mnt_devname);
 #ifdef CONFIG_SMP
 	free_percpu(mnt->mnt_pcp);
@@ -1075,6 +1081,9 @@ static struct mount *clone_mnt(struct mount *old, struct dentry *root,
 	mnt->mnt.mnt_flags &= ~(MNT_WRITE_HOLD|MNT_MARKED|MNT_INTERNAL);
 
 	atomic_inc(&sb->s_active);
+	mnt->mnt.mnt_userns = mnt_user_ns(&old->mnt);
+	if (mnt->mnt.mnt_userns != &init_user_ns)
+		mnt->mnt.mnt_userns = get_user_ns(mnt->mnt.mnt_userns);
 	mnt->mnt.mnt_sb = sb;
 	mnt->mnt.mnt_root = dget(root);
 	mnt->mnt_mountpoint = mnt->mnt.mnt_root;
include/linux/fs.h +1 -0
index 1da4c21fb588..0e2b8d235dca 100644
@@ -2278,6 +2278,7 @@ struct file_system_type {
 #define FS_HAS_SUBTYPE		4
 #define FS_USERNS_MOUNT		8	/* Can be mounted by userns root */
 #define FS_DISALLOW_NOTIFY_PERM	16	/* Disable fanotify permission events */
+#define FS_ALLOW_IDMAP         32      /* FS has been updated to handle vfs idmappings. */
 #define FS_THP_SUPPORT		8192	/* Remove once all fs converted */
 #define FS_RENAME_DOES_D_MOVE	32768	/* FS will handle d_move() during rename() internally. */
 	int (*init_fs_context)(struct fs_context *);
include/linux/mount.h +6 -0
index aaf343b38671..52de25e08319 100644
@@ -72,8 +72,14 @@ struct vfsmount {
 	struct dentry *mnt_root;	/* root of the mounted tree */
 	struct super_block *mnt_sb;	/* pointer to superblock */
 	int mnt_flags;
+	struct user_namespace *mnt_userns;
 } __randomize_layout;
 
+static inline struct user_namespace *mnt_user_ns(const struct vfsmount *mnt)
+{
+	return mnt->mnt_userns;
+}
+
 struct file; /* forward dec */
 struct path;
 
Signature · 1 line
2.30.0
#11 lore raw

PATCH v5 10/42 capability: handle idmapped mounts

In reply to Christian Brauner

In order to determine whether a caller holds privilege over a given inode the capability framework exposes the two helpers privileged_wrt_inode_uidgid() and capable_wrt_inode_uidgid(). The former verifies that the inode has a mapping in the caller's user namespace and the latter additionally verifies that the caller has the requested capability in their current user namespace.
If the inode is accessed through an idmapped mount we simply need to map it according to the mount's user namespace. Afterwards the checks are identical to non-idmapped inodes. If the initial user namespace is passed all operations are a nop so non-idmapped mounts will not see a change in behavior and will also not see any performance impact.

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Don't pollute the vfs with additional helpers simply extend the existing helpers with an additional argument and switch all callers.

/* v3 */
unchanged

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/attr.c                  |  8 ++++----
 fs/exec.c                  |  2 +-
 fs/inode.c                 |  2 +-
 fs/namei.c                 | 13 ++++++++-----
 fs/overlayfs/super.c       |  2 +-
 fs/posix_acl.c             |  2 +-
 fs/xfs/xfs_ioctl.c         |  2 +-
 include/linux/capability.h |  7 +++++--
 kernel/capability.c        | 14 +++++++++-----
 security/commoncap.c       |  5 +++--
 10 files changed, 34 insertions(+), 23 deletions(-)
fs/attr.c +4 -4
index b4bbdbd4c8ca..d270f640a192 100644
@@ -23,7 +23,7 @@ static bool chown_ok(const struct inode *inode, kuid_t uid)
 	if (uid_eq(current_fsuid(), inode->i_uid) &&
 	    uid_eq(uid, inode->i_uid))
 		return true;
-	if (capable_wrt_inode_uidgid(inode, CAP_CHOWN))
+	if (capable_wrt_inode_uidgid(&init_user_ns, inode, CAP_CHOWN))
 		return true;
 	if (uid_eq(inode->i_uid, INVALID_UID) &&
 	    ns_capable(inode->i_sb->s_user_ns, CAP_CHOWN))
@@ -36,7 +36,7 @@ static bool chgrp_ok(const struct inode *inode, kgid_t gid)
 	if (uid_eq(current_fsuid(), inode->i_uid) &&
 	    (in_group_p(gid) || gid_eq(gid, inode->i_gid)))
 		return true;
-	if (capable_wrt_inode_uidgid(inode, CAP_CHOWN))
+	if (capable_wrt_inode_uidgid(&init_user_ns, inode, CAP_CHOWN))
 		return true;
 	if (gid_eq(inode->i_gid, INVALID_GID) &&
 	    ns_capable(inode->i_sb->s_user_ns, CAP_CHOWN))
@@ -92,7 +92,7 @@ int setattr_prepare(struct dentry *dentry, struct iattr *attr)
 		/* Also check the setgid bit! */
 		if (!in_group_p((ia_valid & ATTR_GID) ? attr->ia_gid :
 				inode->i_gid) &&
-		    !capable_wrt_inode_uidgid(inode, CAP_FSETID))
+		    !capable_wrt_inode_uidgid(&init_user_ns, inode, CAP_FSETID))
 			attr->ia_mode &= ~S_ISGID;
 	}
 
@@ -193,7 +193,7 @@ void setattr_copy(struct inode *inode, const struct iattr *attr)
 		umode_t mode = attr->ia_mode;
 
 		if (!in_group_p(inode->i_gid) &&
-		    !capable_wrt_inode_uidgid(inode, CAP_FSETID))
+		    !capable_wrt_inode_uidgid(&init_user_ns, inode, CAP_FSETID))
 			mode &= ~S_ISGID;
 		inode->i_mode = mode;
 	}
fs/exec.c +1 -1
index 5d4d52039105..cea064a2c473 100644
@@ -1411,7 +1411,7 @@ void would_dump(struct linux_binprm *bprm, struct file *file)
 		/* Ensure mm->user_ns contains the executable */
 		user_ns = old = bprm->mm->user_ns;
 		while ((user_ns != &init_user_ns) &&
-		       !privileged_wrt_inode_uidgid(user_ns, inode))
+		       !privileged_wrt_inode_uidgid(user_ns, &init_user_ns, inode))
 			user_ns = user_ns->parent;
 
 		if (old != user_ns) {
fs/inode.c +1 -1
index 6442d97d9a4a..e1a58950f5e6 100644
@@ -2146,7 +2146,7 @@ void inode_init_owner(struct inode *inode, const struct inode *dir,
 			mode |= S_ISGID;
 		else if ((mode & (S_ISGID | S_IXGRP)) == (S_ISGID | S_IXGRP) &&
 			 !in_group_p(inode->i_gid) &&
-			 !capable_wrt_inode_uidgid(dir, CAP_FSETID))
+			 !capable_wrt_inode_uidgid(&init_user_ns, dir, CAP_FSETID))
 			mode &= ~S_ISGID;
 	} else
 		inode->i_gid = current_fsgid();
fs/namei.c +8 -5
index 78443a85480a..fd4724bce4f5 100644
@@ -357,10 +357,11 @@ int generic_permission(struct inode *inode, int mask)
 	if (S_ISDIR(inode->i_mode)) {
 		/* DACs are overridable for directories */
 		if (!(mask & MAY_WRITE))
-			if (capable_wrt_inode_uidgid(inode,
+			if (capable_wrt_inode_uidgid(&init_user_ns, inode,
 						     CAP_DAC_READ_SEARCH))
 				return 0;
-		if (capable_wrt_inode_uidgid(inode, CAP_DAC_OVERRIDE))
+		if (capable_wrt_inode_uidgid(&init_user_ns, inode,
+					     CAP_DAC_OVERRIDE))
 			return 0;
 		return -EACCES;
 	}
@@ -370,7 +371,8 @@ int generic_permission(struct inode *inode, int mask)
 	 */
 	mask &= MAY_READ | MAY_WRITE | MAY_EXEC;
 	if (mask == MAY_READ)
-		if (capable_wrt_inode_uidgid(inode, CAP_DAC_READ_SEARCH))
+		if (capable_wrt_inode_uidgid(&init_user_ns, inode,
+					     CAP_DAC_READ_SEARCH))
 			return 0;
 	/*
 	 * Read/write DACs are always overridable.
@@ -378,7 +380,8 @@ int generic_permission(struct inode *inode, int mask)
 	 * at least one exec bit set.
 	 */
 	if (!(mask & MAY_EXEC) || (inode->i_mode & S_IXUGO))
-		if (capable_wrt_inode_uidgid(inode, CAP_DAC_OVERRIDE))
+		if (capable_wrt_inode_uidgid(&init_user_ns, inode,
+					     CAP_DAC_OVERRIDE))
 			return 0;
 
 	return -EACCES;
@@ -2659,7 +2662,7 @@ int __check_sticky(struct inode *dir, struct inode *inode)
 		return 0;
 	if (uid_eq(dir->i_uid, fsuid))
 		return 0;
-	return !capable_wrt_inode_uidgid(inode, CAP_FOWNER);
+	return !capable_wrt_inode_uidgid(&init_user_ns, inode, CAP_FOWNER);
 }
 EXPORT_SYMBOL(__check_sticky);
 
fs/overlayfs/super.c +1 -1
index 2bd570cbe8a4..88d877787770 100644
@@ -1017,7 +1017,7 @@ ovl_posix_acl_xattr_set(const struct xattr_handler *handler,
 	if (unlikely(inode->i_mode & S_ISGID) &&
 	    handler->flags == ACL_TYPE_ACCESS &&
 	    !in_group_p(inode->i_gid) &&
-	    !capable_wrt_inode_uidgid(inode, CAP_FSETID)) {
+	    !capable_wrt_inode_uidgid(&init_user_ns, inode, CAP_FSETID)) {
 		struct iattr iattr = { .ia_valid = ATTR_KILL_SGID };
 
 		err = ovl_setattr(dentry, &iattr);
fs/posix_acl.c +1 -1
index 95882b3f5f62..4ca6d53c6f0a 100644
@@ -656,7 +656,7 @@ int posix_acl_update_mode(struct inode *inode, umode_t *mode_p,
 	if (error == 0)
 		*acl = NULL;
 	if (!in_group_p(inode->i_gid) &&
-	    !capable_wrt_inode_uidgid(inode, CAP_FSETID))
+	    !capable_wrt_inode_uidgid(&init_user_ns, inode, CAP_FSETID))
 		mode &= ~S_ISGID;
 	*mode_p = mode;
 	return 0;
fs/xfs/xfs_ioctl.c +1 -1
index 3fbd98f61ea5..97bd29fc8c43 100644
@@ -1502,7 +1502,7 @@ xfs_ioctl_setattr(
 	 */
 
 	if ((VFS_I(ip)->i_mode & (S_ISUID|S_ISGID)) &&
-	    !capable_wrt_inode_uidgid(VFS_I(ip), CAP_FSETID))
+	    !capable_wrt_inode_uidgid(&init_user_ns, VFS_I(ip), CAP_FSETID))
 		VFS_I(ip)->i_mode &= ~(S_ISUID|S_ISGID);
 
 	/* Change the ownerships and register project quota modifications */
include/linux/capability.h +5 -2
index b2f698915c0f..53d5acfd23c4 100644
@@ -247,8 +247,11 @@ static inline bool ns_capable_setid(struct user_namespace *ns, int cap)
 	return true;
 }
 #endif /* CONFIG_MULTIUSER */
-extern bool privileged_wrt_inode_uidgid(struct user_namespace *ns, const struct inode *inode);
-extern bool capable_wrt_inode_uidgid(const struct inode *inode, int cap);
+extern bool privileged_wrt_inode_uidgid(struct user_namespace *ns,
+					struct user_namespace *mnt_userns,
+					const struct inode *inode);
+extern bool capable_wrt_inode_uidgid(struct user_namespace *mnt_userns,
+				     const struct inode *inode, int cap);
 extern bool file_ns_capable(const struct file *file, struct user_namespace *ns, int cap);
 extern bool ptracer_capable(struct task_struct *tsk, struct user_namespace *ns);
 static inline bool perfmon_capable(void)
kernel/capability.c +9 -5
index de7eac903a2a..46a361dde042 100644
@@ -484,10 +484,12 @@ EXPORT_SYMBOL(file_ns_capable);
  *
  * Return true if the inode uid and gid are within the namespace.
  */
-bool privileged_wrt_inode_uidgid(struct user_namespace *ns, const struct inode *inode)
+bool privileged_wrt_inode_uidgid(struct user_namespace *ns,
+				 struct user_namespace *mnt_userns,
+				 const struct inode *inode)
 {
-	return kuid_has_mapping(ns, inode->i_uid) &&
-		kgid_has_mapping(ns, inode->i_gid);
+	return kuid_has_mapping(ns, i_uid_into_mnt(mnt_userns, inode)) &&
+	       kgid_has_mapping(ns, i_gid_into_mnt(mnt_userns, inode));
 }
 
 /**
@@ -499,11 +501,13 @@ bool privileged_wrt_inode_uidgid(struct user_namespace *ns, const struct inode *
  * its own user namespace and that the given inode's uid and gid are
  * mapped into the current user namespace.
  */
-bool capable_wrt_inode_uidgid(const struct inode *inode, int cap)
+bool capable_wrt_inode_uidgid(struct user_namespace *mnt_userns,
+			      const struct inode *inode, int cap)
 {
 	struct user_namespace *ns = current_user_ns();
 
-	return ns_capable(ns, cap) && privileged_wrt_inode_uidgid(ns, inode);
+	return ns_capable(ns, cap) &&
+	       privileged_wrt_inode_uidgid(ns, mnt_userns, inode);
 }
 EXPORT_SYMBOL(capable_wrt_inode_uidgid);
 
security/commoncap.c +3 -2
index bacc1111d871..88ee345f7bfa 100644
@@ -489,7 +489,7 @@ int cap_convert_nscap(struct dentry *dentry, const void **ivalue, size_t size)
 		return -EINVAL;
 	if (!validheader(size, cap))
 		return -EINVAL;
-	if (!capable_wrt_inode_uidgid(inode, CAP_SETFCAP))
+	if (!capable_wrt_inode_uidgid(&init_user_ns, inode, CAP_SETFCAP))
 		return -EPERM;
 	if (size == XATTR_CAPS_SZ_2)
 		if (ns_capable(inode->i_sb->s_user_ns, CAP_SETFCAP))
@@ -956,7 +956,8 @@ int cap_inode_removexattr(struct dentry *dentry, const char *name)
 		struct inode *inode = d_backing_inode(dentry);
 		if (!inode)
 			return -EINVAL;
-		if (!capable_wrt_inode_uidgid(inode, CAP_SETFCAP))
+		if (!capable_wrt_inode_uidgid(&init_user_ns, inode,
+					      CAP_SETFCAP))
 			return -EPERM;
 		return 0;
 	}
Signature · 1 line
2.30.0
Reviewed-by on v5 10/42
#12 lore raw

Re: [PATCH v5 10/42] capability: handle idmapped mounts

Quoted 11 lines · Christian Brauner

In order to determine whether a caller holds privilege over a given inode the capability framework exposes the two helpers privileged_wrt_inode_uidgid() and capable_wrt_inode_uidgid(). The former verifies that the inode has a mapping in the caller's user namespace and the latter additionally verifies that the caller has the requested capability in their current user namespace.
If the inode is accessed through an idmapped mount we simply need to map it according to the mount's user namespace. Afterwards the checks are identical to non-idmapped inodes. If the initial user namespace is passed all operations are a nop so non-idmapped mounts will not see a change in behavior and will also not see any performance impact.

This adds a bunch of pointless > 80 char lines, that would be nice to fix up.

Otherwise looks good:

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers

#13 lore raw

PATCH v5 11/42 namei: make permission helpers idmapped mount aware

In reply to Christian Brauner

The two helpers inode_permission() and generic_permission() are used by the vfs to perform basic permission checking by verifying that the caller is privileged over an inode. In order to handle idmapped mounts we extend the two helpers with an additional user namespace argument. On idmapped mounts the two helpers will make sure to map the inode according to the mount's user namespace and then peform identical permission checks to inode_permission() and generic_permission(). If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Don't pollute the vfs with additional helpers simply extend the existing helpers with an additional argument and switch all callers.

/* v3 */
unchanged

/* v4 */
- "Serge E. Hallyn" <[email protected]>:
- Add proper documentation for all the changed permission checking helpers and adjust terminology to avoid any potential confusion.
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

- Christoph Hellwig <[email protected]>:
- Change commit message to reflect the fact that no new permission helpers are introduced but only the existing ones changed.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/attr.c                          |   3 +-
 fs/btrfs/inode.c                   |   2 +-
 fs/btrfs/ioctl.c                   |  10 +--
 fs/ceph/inode.c                    |   2 +-
 fs/cifs/cifsfs.c                   |   2 +-
 fs/configfs/symlink.c              |   2 +-
 fs/ecryptfs/inode.c                |   2 +-
 fs/exec.c                          |   2 +-
 fs/fuse/dir.c                      |   4 +-
 fs/gfs2/inode.c                    |   2 +-
 fs/hostfs/hostfs_kern.c            |   2 +-
 fs/init.c                          |   9 ++-
 fs/kernfs/inode.c                  |   2 +-
 fs/libfs.c                         |   7 +-
 fs/namei.c                         | 121 ++++++++++++++++++++---------
 fs/nfs/dir.c                       |   2 +-
 fs/nfsd/nfsfh.c                    |   2 +-
 fs/nfsd/vfs.c                      |   4 +-
 fs/nilfs2/inode.c                  |   2 +-
 fs/notify/fanotify/fanotify_user.c |   2 +-
 fs/notify/inotify/inotify_user.c   |   2 +-
 fs/ocfs2/file.c                    |   2 +-
 fs/ocfs2/refcounttree.c            |   4 +-
 fs/open.c                          |  10 +--
 fs/orangefs/inode.c                |   2 +-
 fs/overlayfs/file.c                |   2 +-
 fs/overlayfs/inode.c               |   4 +-
 fs/overlayfs/util.c                |   2 +-
 fs/posix_acl.c                     |  17 ++--
 fs/proc/base.c                     |   4 +-
 fs/proc/fd.c                       |   2 +-
 fs/reiserfs/xattr.c                |   2 +-
 fs/remap_range.c                   |   2 +-
 fs/udf/file.c                      |   2 +-
 fs/verity/enable.c                 |   2 +-
 fs/xattr.c                         |   2 +-
 include/linux/fs.h                 |   4 +-
 include/linux/posix_acl.h          |   4 +-
 ipc/mqueue.c                       |   2 +-
 kernel/bpf/inode.c                 |   4 +-
 kernel/cgroup/cgroup.c             |   2 +-
 kernel/sys.c                       |   2 +-
 mm/madvise.c                       |   2 +-
 mm/memcontrol.c                    |   2 +-
 mm/mincore.c                       |   2 +-
 net/unix/af_unix.c                 |   2 +-
 46 files changed, 168 insertions(+), 103 deletions(-)
fs/attr.c +2 -1
index d270f640a192..c9e29e589cec 100644
@@ -244,7 +244,8 @@ int notify_change(struct dentry * dentry, struct iattr * attr, struct inode **de
 			return -EPERM;
 
 		if (!inode_owner_or_capable(inode)) {
-			error = inode_permission(inode, MAY_WRITE);
+			error = inode_permission(&init_user_ns, inode,
+						 MAY_WRITE);
 			if (error)
 				return error;
 		}
fs/btrfs/inode.c +1 -1
index 070716650df8..f834ba3b16b7 100644
@@ -9863,7 +9863,7 @@ static int btrfs_permission(struct inode *inode, int mask)
 		if (BTRFS_I(inode)->flags & BTRFS_INODE_READONLY)
 			return -EACCES;
 	}
-	return generic_permission(inode, mask);
+	return generic_permission(&init_user_ns, inode, mask);
 }
 
 static int btrfs_tmpfile(struct inode *dir, struct dentry *dentry, umode_t mode)
fs/btrfs/ioctl.c +5 -5
index dde49a791f3e..7b47af3d8d5a 100644
@@ -922,7 +922,7 @@ static int btrfs_may_delete(struct inode *dir, struct dentry *victim, int isdir)
 	BUG_ON(d_inode(victim->d_parent) != dir);
 	audit_inode_child(dir, victim, AUDIT_TYPE_CHILD_DELETE);
 
-	error = inode_permission(dir, MAY_WRITE | MAY_EXEC);
+	error = inode_permission(&init_user_ns, dir, MAY_WRITE | MAY_EXEC);
 	if (error)
 		return error;
 	if (IS_APPEND(dir))
@@ -951,7 +951,7 @@ static inline int btrfs_may_create(struct inode *dir, struct dentry *child)
 		return -EEXIST;
 	if (IS_DEADDIR(dir))
 		return -ENOENT;
-	return inode_permission(dir, MAY_WRITE | MAY_EXEC);
+	return inode_permission(&init_user_ns, dir, MAY_WRITE | MAY_EXEC);
 }
 
 /*
@@ -2538,7 +2538,7 @@ static int btrfs_search_path_in_tree_user(struct inode *inode,
 				ret = PTR_ERR(temp_inode);
 				goto out_put;
 			}
-			ret = inode_permission(temp_inode, MAY_READ | MAY_EXEC);
+			ret = inode_permission(&init_user_ns, temp_inode, MAY_READ | MAY_EXEC);
 			iput(temp_inode);
 			if (ret) {
 				ret = -EACCES;
@@ -3068,7 +3068,7 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file,
 		if (root == dest)
 			goto out_dput;
 
-		err = inode_permission(inode, MAY_WRITE | MAY_EXEC);
+		err = inode_permission(&init_user_ns, inode, MAY_WRITE | MAY_EXEC);
 		if (err)
 			goto out_dput;
 	}
@@ -3139,7 +3139,7 @@ static int btrfs_ioctl_defrag(struct file *file, void __user *argp)
 		 * running and allows defrag on files open in read-only mode.
 		 */
 		if (!capable(CAP_SYS_ADMIN) &&
-		    inode_permission(inode, MAY_WRITE)) {
+		    inode_permission(&init_user_ns, inode, MAY_WRITE)) {
 			ret = -EPERM;
 			goto out;
 		}
fs/ceph/inode.c +1 -1
index adc8fc3c5d85..e8a15ee09bc1 100644
@@ -2331,7 +2331,7 @@ int ceph_permission(struct inode *inode, int mask)
 	err = ceph_do_getattr(inode, CEPH_CAP_AUTH_SHARED, false);
 
 	if (!err)
-		err = generic_permission(inode, mask);
+		err = generic_permission(&init_user_ns, inode, mask);
 	return err;
 }
 
fs/cifs/cifsfs.c +1 -1
index ce0d0037fd0a..ce14e6f8adb6 100644
@@ -320,7 +320,7 @@ static int cifs_permission(struct inode *inode, int mask)
 		on the client (above and beyond ACL on servers) for
 		servers which do not support setting and viewing mode bits,
 		so allowing client to check permissions is useful */
-		return generic_permission(inode, mask);
+		return generic_permission(&init_user_ns, inode, mask);
 }
 
 static struct kmem_cache *cifs_inode_cachep;
fs/configfs/symlink.c +1 -1
index cb61467478ca..0b592c55f38e 100644
@@ -197,7 +197,7 @@ int configfs_symlink(struct inode *dir, struct dentry *dentry, const char *symna
 	if (dentry->d_inode || d_unhashed(dentry))
 		ret = -EEXIST;
 	else
-		ret = inode_permission(dir, MAY_WRITE | MAY_EXEC);
+		ret = inode_permission(&init_user_ns, dir, MAY_WRITE | MAY_EXEC);
 	if (!ret)
 		ret = type->ct_item_ops->allow_link(parent_item, target_item);
 	if (!ret) {
fs/ecryptfs/inode.c +1 -1
index e23752d9a79f..9b1ae410983c 100644
@@ -864,7 +864,7 @@ int ecryptfs_truncate(struct dentry *dentry, loff_t new_length)
 static int
 ecryptfs_permission(struct inode *inode, int mask)
 {
-	return inode_permission(ecryptfs_inode_to_lower(inode), mask);
+	return inode_permission(&init_user_ns, ecryptfs_inode_to_lower(inode), mask);
 }
 
 /**
fs/exec.c +1 -1
index cea064a2c473..049a8d2669db 100644
@@ -1404,7 +1404,7 @@ EXPORT_SYMBOL(begin_new_exec);
 void would_dump(struct linux_binprm *bprm, struct file *file)
 {
 	struct inode *inode = file_inode(file);
-	if (inode_permission(inode, MAY_READ) < 0) {
+	if (inode_permission(&init_user_ns, inode, MAY_READ) < 0) {
 		struct user_namespace *old, *user_ns;
 		bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP;
 
fs/fuse/dir.c +2 -2
index 78f9f209078c..19a46f445d4c 100644
@@ -1280,7 +1280,7 @@ static int fuse_permission(struct inode *inode, int mask)
 	}
 
 	if (fc->default_permissions) {
-		err = generic_permission(inode, mask);
+		err = generic_permission(&init_user_ns, inode, mask);
 
 		/* If permission is denied, try to refresh file
 		   attributes.  This is also needed, because the root
@@ -1288,7 +1288,7 @@ static int fuse_permission(struct inode *inode, int mask)
 		if (err == -EACCES && !refreshed) {
 			err = fuse_perm_getattr(inode, mask);
 			if (!err)
-				err = generic_permission(inode, mask);
+				err = generic_permission(&init_user_ns, inode, mask);
 		}
 
 		/* Note: the opposite of the above test does not
fs/gfs2/inode.c +1 -1
index c1b77e8d6b1c..5b2ff0c74b67 100644
@@ -1852,7 +1852,7 @@ int gfs2_permission(struct inode *inode, int mask)
 	if ((mask & MAY_WRITE) && IS_IMMUTABLE(inode))
 		error = -EPERM;
 	else
-		error = generic_permission(inode, mask);
+		error = generic_permission(&init_user_ns, inode, mask);
 	if (gfs2_holder_initialized(&i_gh))
 		gfs2_glock_dq_uninit(&i_gh);
 
fs/hostfs/hostfs_kern.c +1 -1
index aea35459d390..b841a05a2b8c 100644
@@ -779,7 +779,7 @@ static int hostfs_permission(struct inode *ino, int desired)
 		err = access_file(name, r, w, x);
 	__putname(name);
 	if (!err)
-		err = generic_permission(ino, desired);
+		err = generic_permission(&init_user_ns, ino, desired);
 	return err;
 }
 
fs/init.c +6 -3
index e9c320a48cf1..2b4842f4802b 100644
@@ -49,7 +49,8 @@ int __init init_chdir(const char *filename)
 	error = kern_path(filename, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &path);
 	if (error)
 		return error;
-	error = inode_permission(path.dentry->d_inode, MAY_EXEC | MAY_CHDIR);
+	error = inode_permission(&init_user_ns, path.dentry->d_inode,
+				 MAY_EXEC | MAY_CHDIR);
 	if (!error)
 		set_fs_pwd(current->fs, &path);
 	path_put(&path);
@@ -64,7 +65,8 @@ int __init init_chroot(const char *filename)
 	error = kern_path(filename, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &path);
 	if (error)
 		return error;
-	error = inode_permission(path.dentry->d_inode, MAY_EXEC | MAY_CHDIR);
+	error = inode_permission(&init_user_ns, path.dentry->d_inode,
+				 MAY_EXEC | MAY_CHDIR);
 	if (error)
 		goto dput_and_out;
 	error = -EPERM;
@@ -118,7 +120,8 @@ int __init init_eaccess(const char *filename)
 	error = kern_path(filename, LOOKUP_FOLLOW, &path);
 	if (error)
 		return error;
-	error = inode_permission(d_inode(path.dentry), MAY_ACCESS);
+	error = inode_permission(&init_user_ns, d_inode(path.dentry),
+				 MAY_ACCESS);
 	path_put(&path);
 	return error;
 }
fs/kernfs/inode.c +1 -1
index fc2469a20fed..ff5598cc1de0 100644
@@ -285,7 +285,7 @@ int kernfs_iop_permission(struct inode *inode, int mask)
 	kernfs_refresh_inode(kn, inode);
 	mutex_unlock(&kernfs_mutex);
 
-	return generic_permission(inode, mask);
+	return generic_permission(&init_user_ns, inode, mask);
 }
 
 int kernfs_xattr_get(struct kernfs_node *kn, const char *name,
fs/libfs.c +6 -1
index d1c3bade9f30..f8b3c02b4f0f 100644
@@ -1318,9 +1318,14 @@ static ssize_t empty_dir_listxattr(struct dentry *dentry, char *list, size_t siz
 	return -EOPNOTSUPP;
 }
 
+static int empty_dir_permission(struct inode *inode, int mask)
+{
+	return generic_permission(&init_user_ns, inode, mask);
+}
+
 static const struct inode_operations empty_dir_inode_operations = {
 	.lookup		= empty_dir_lookup,
-	.permission	= generic_permission,
+	.permission	= empty_dir_permission,
 	.setattr	= empty_dir_setattr,
 	.getattr	= empty_dir_getattr,
 	.listxattr	= empty_dir_listxattr,
fs/namei.c +84 -37

300 lines not on this page · raw message

fs/nfs/dir.c +1 -1
index ef827ae193d2..727e01a84503 100644
@@ -2987,7 +2987,7 @@ int nfs_permission(struct inode *inode, int mask)
 
 	res = nfs_revalidate_inode(NFS_SERVER(inode), inode);
 	if (res == 0)
-		res = generic_permission(inode, mask);
+		res = generic_permission(&init_user_ns, inode, mask);
 	goto out;
 }
 EXPORT_SYMBOL_GPL(nfs_permission);
fs/nfsd/nfsfh.c +1 -1
index 66f2ef67792a..bbb271d9e1b5 100644
@@ -40,7 +40,7 @@ static int nfsd_acceptable(void *expv, struct dentry *dentry)
 		/* make sure parents give x permission to user */
 		int err;
 		parent = dget_parent(tdentry);
-		err = inode_permission(d_inode(parent), MAY_EXEC);
+		err = inode_permission(&init_user_ns, d_inode(parent), MAY_EXEC);
 		if (err < 0) {
 			dput(parent);
 			break;
fs/nfsd/vfs.c +2 -2
index 04937e51de56..b3ccb2997ea0 100644
@@ -2391,13 +2391,13 @@ nfsd_permission(struct svc_rqst *rqstp, struct svc_export *exp,
 		return 0;
 
 	/* This assumes  NFSD_MAY_{READ,WRITE,EXEC} == MAY_{READ,WRITE,EXEC} */
-	err = inode_permission(inode, acc & (MAY_READ|MAY_WRITE|MAY_EXEC));
+	err = inode_permission(&init_user_ns, inode, acc & (MAY_READ|MAY_WRITE|MAY_EXEC));
 
 	/* Allow read access to binaries even when mode 111 */
 	if (err == -EACCES && S_ISREG(inode->i_mode) &&
 	     (acc == (NFSD_MAY_READ | NFSD_MAY_OWNER_OVERRIDE) ||
 	      acc == (NFSD_MAY_READ | NFSD_MAY_READ_IF_EXEC)))
-		err = inode_permission(inode, MAY_EXEC);
+		err = inode_permission(&init_user_ns, inode, MAY_EXEC);
 
 	return err? nfserrno(err) : 0;
 }
fs/nilfs2/inode.c +1 -1
index 745d371d6fea..b6517220cad5 100644
@@ -851,7 +851,7 @@ int nilfs_permission(struct inode *inode, int mask)
 	    root->cno != NILFS_CPTREE_CURRENT_CNO)
 		return -EROFS; /* snapshot is not writable */
 
-	return generic_permission(inode, mask);
+	return generic_permission(&init_user_ns, inode, mask);
 }
 
 int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh)
fs/notify/fanotify/fanotify_user.c +1 -1
index dcab112e1f00..4ca13fb33c7c 100644
@@ -702,7 +702,7 @@ static int fanotify_find_path(int dfd, const char __user *filename,
 	}
 
 	/* you can only watch an inode if you have read permissions on it */
-	ret = inode_permission(path->dentry->d_inode, MAY_READ);
+	ret = inode_permission(&init_user_ns, path->dentry->d_inode, MAY_READ);
 	if (ret) {
 		path_put(path);
 		goto out;
fs/notify/inotify/inotify_user.c +1 -1
index 59c177011a0f..18e7024207a9 100644
@@ -352,7 +352,7 @@ static int inotify_find_inode(const char __user *dirname, struct path *path,
 	if (error)
 		return error;
 	/* you can only watch an inode if you have read permissions on it */
-	error = inode_permission(path->dentry->d_inode, MAY_READ);
+	error = inode_permission(&init_user_ns, path->dentry->d_inode, MAY_READ);
 	if (error) {
 		path_put(path);
 		return error;
fs/ocfs2/file.c +1 -1
index 85979e2214b3..0c75619adf54 100644
@@ -1355,7 +1355,7 @@ int ocfs2_permission(struct inode *inode, int mask)
 		dump_stack();
 	}
 
-	ret = generic_permission(inode, mask);
+	ret = generic_permission(&init_user_ns, inode, mask);
 
 	ocfs2_inode_unlock_tracker(inode, 0, &oh, had_lock);
 out:
fs/ocfs2/refcounttree.c +2 -2
index 3b397fa9c9e8..c26937824be1 100644
@@ -4346,7 +4346,7 @@ static inline int ocfs2_may_create(struct inode *dir, struct dentry *child)
 		return -EEXIST;
 	if (IS_DEADDIR(dir))
 		return -ENOENT;
-	return inode_permission(dir, MAY_WRITE | MAY_EXEC);
+	return inode_permission(&init_user_ns, dir, MAY_WRITE | MAY_EXEC);
 }
 
 /**
@@ -4400,7 +4400,7 @@ static int ocfs2_vfs_reflink(struct dentry *old_dentry, struct inode *dir,
 	 * file.
 	 */
 	if (!preserve) {
-		error = inode_permission(inode, MAY_READ);
+		error = inode_permission(&init_user_ns, inode, MAY_READ);
 		if (error)
 			return error;
 	}
fs/open.c +5 -5
index 1e06e443a565..4089ed2c51ec 100644
@@ -83,7 +83,7 @@ long vfs_truncate(const struct path *path, loff_t length)
 	if (error)
 		goto out;
 
-	error = inode_permission(inode, MAY_WRITE);
+	error = inode_permission(&init_user_ns, inode, MAY_WRITE);
 	if (error)
 		goto mnt_drop_write_and_out;
 
@@ -436,7 +436,7 @@ static long do_faccessat(int dfd, const char __user *filename, int mode, int fla
 			goto out_path_release;
 	}
 
-	res = inode_permission(inode, mode | MAY_ACCESS);
+	res = inode_permission(&init_user_ns, inode, mode | MAY_ACCESS);
 	/* SuS v2 requires we report a read only fs too */
 	if (res || !(mode & S_IWOTH) || special_file(inode->i_mode))
 		goto out_path_release;
@@ -492,7 +492,7 @@ SYSCALL_DEFINE1(chdir, const char __user *, filename)
 	if (error)
 		goto out;
 
-	error = inode_permission(path.dentry->d_inode, MAY_EXEC | MAY_CHDIR);
+	error = inode_permission(&init_user_ns, path.dentry->d_inode, MAY_EXEC | MAY_CHDIR);
 	if (error)
 		goto dput_and_out;
 
@@ -521,7 +521,7 @@ SYSCALL_DEFINE1(fchdir, unsigned int, fd)
 	if (!d_can_lookup(f.file->f_path.dentry))
 		goto out_putf;
 
-	error = inode_permission(file_inode(f.file), MAY_EXEC | MAY_CHDIR);
+	error = inode_permission(&init_user_ns, file_inode(f.file), MAY_EXEC | MAY_CHDIR);
 	if (!error)
 		set_fs_pwd(current->fs, &f.file->f_path);
 out_putf:
@@ -540,7 +540,7 @@ SYSCALL_DEFINE1(chroot, const char __user *, filename)
 	if (error)
 		goto out;
 
-	error = inode_permission(path.dentry->d_inode, MAY_EXEC | MAY_CHDIR);
+	error = inode_permission(&init_user_ns, path.dentry->d_inode, MAY_EXEC | MAY_CHDIR);
 	if (error)
 		goto dput_and_out;
 
fs/orangefs/inode.c +1 -1
index 48f0547d4850..4c790cc8042d 100644
@@ -933,7 +933,7 @@ int orangefs_permission(struct inode *inode, int mask)
 	if (ret < 0)
 		return ret;
 
-	return generic_permission(inode, mask);
+	return generic_permission(&init_user_ns, inode, mask);
 }
 
 int orangefs_update_time(struct inode *inode, struct timespec64 *time, int flags)
fs/overlayfs/file.c +1 -1
index bd9dd38347ae..b2948e7b3210 100644
@@ -50,7 +50,7 @@ static struct file *ovl_open_realfile(const struct file *file,
 		acc_mode |= MAY_APPEND;
 
 	old_cred = ovl_override_creds(inode->i_sb);
-	err = inode_permission(realinode, MAY_OPEN | acc_mode);
+	err = inode_permission(&init_user_ns, realinode, MAY_OPEN | acc_mode);
 	if (err) {
 		realfile = ERR_PTR(err);
 	} else {
fs/overlayfs/inode.c +2 -2
index d739e14c6814..c101ebbb7a77 100644
@@ -294,7 +294,7 @@ int ovl_permission(struct inode *inode, int mask)
 	 * Check overlay inode with the creds of task and underlying inode
 	 * with creds of mounter
 	 */
-	err = generic_permission(inode, mask);
+	err = generic_permission(&init_user_ns, inode, mask);
 	if (err)
 		return err;
 
@@ -305,7 +305,7 @@ int ovl_permission(struct inode *inode, int mask)
 		/* Make sure mounter can read file for copy up later */
 		mask |= MAY_READ;
 	}
-	err = inode_permission(realinode, mask);
+	err = inode_permission(&init_user_ns, realinode, mask);
 	revert_creds(old_cred);
 
 	return err;
fs/overlayfs/util.c +1 -1
index 6569031af3cd..de5c2047a0e9 100644
@@ -479,7 +479,7 @@ struct file *ovl_path_open(struct path *path, int flags)
 		BUG();
 	}
 
-	err = inode_permission(inode, acc_mode | MAY_OPEN);
+	err = inode_permission(&init_user_ns, inode, acc_mode | MAY_OPEN);
 	if (err)
 		return ERR_PTR(err);
 
fs/posix_acl.c +12 -5
index 4ca6d53c6f0a..5d9fe2fb2953 100644
@@ -345,10 +345,13 @@ EXPORT_SYMBOL(posix_acl_from_mode);
  * by the acl. Returns -E... otherwise.
  */
 int
-posix_acl_permission(struct inode *inode, const struct posix_acl *acl, int want)
+posix_acl_permission(struct user_namespace *mnt_userns, struct inode *inode,
+		     const struct posix_acl *acl, int want)
 {
 	const struct posix_acl_entry *pa, *pe, *mask_obj;
 	int found = 0;
+	kuid_t uid;
+	kgid_t gid;
 
 	want &= MAY_READ | MAY_WRITE | MAY_EXEC;
 
@@ -356,22 +359,26 @@ posix_acl_permission(struct inode *inode, const struct posix_acl *acl, int want)
                 switch(pa->e_tag) {
                         case ACL_USER_OBJ:
 				/* (May have been checked already) */
-				if (uid_eq(inode->i_uid, current_fsuid()))
+				uid = i_uid_into_mnt(mnt_userns, inode);
+				if (uid_eq(uid, current_fsuid()))
                                         goto check_perm;
                                 break;
                         case ACL_USER:
-				if (uid_eq(pa->e_uid, current_fsuid()))
+				uid = kuid_into_mnt(mnt_userns, pa->e_uid);
+				if (uid_eq(uid, current_fsuid()))
                                         goto mask;
 				break;
                         case ACL_GROUP_OBJ:
-                                if (in_group_p(inode->i_gid)) {
+				gid = i_gid_into_mnt(mnt_userns, inode);
+				if (in_group_p(gid)) {
 					found = 1;
 					if ((pa->e_perm & want) == want)
 						goto mask;
                                 }
 				break;
                         case ACL_GROUP:
-				if (in_group_p(pa->e_gid)) {
+				gid = kgid_into_mnt(mnt_userns, pa->e_gid);
+				if (in_group_p(gid)) {
 					found = 1;
 					if ((pa->e_perm & want) == want)
 						goto mask;
fs/proc/base.c +2 -2
index b3422cda2a91..b4ec9293625e 100644
@@ -751,7 +751,7 @@ static int proc_pid_permission(struct inode *inode, int mask)
 
 		return -EPERM;
 	}
-	return generic_permission(inode, mask);
+	return generic_permission(&init_user_ns, inode, mask);
 }
 
 
@@ -3492,7 +3492,7 @@ static int proc_tid_comm_permission(struct inode *inode, int mask)
 		return 0;
 	}
 
-	return generic_permission(inode, mask);
+	return generic_permission(&init_user_ns, inode, mask);
 }
 
 static const struct inode_operations proc_tid_comm_inode_operations = {
fs/proc/fd.c +1 -1
index cb51763ed554..d6e76461e135 100644
@@ -281,7 +281,7 @@ int proc_fd_permission(struct inode *inode, int mask)
 	struct task_struct *p;
 	int rv;
 
-	rv = generic_permission(inode, mask);
+	rv = generic_permission(&init_user_ns, inode, mask);
 	if (rv == 0)
 		return rv;
 
fs/reiserfs/xattr.c +1 -1
index fe63a7c3e0da..ec440d1957a1 100644
@@ -957,7 +957,7 @@ int reiserfs_permission(struct inode *inode, int mask)
 	if (IS_PRIVATE(inode))
 		return 0;
 
-	return generic_permission(inode, mask);
+	return generic_permission(&init_user_ns, inode, mask);
 }
 
 static int xattr_hide_revalidate(struct dentry *dentry, unsigned int flags)
fs/remap_range.c +1 -1
index 77dba3a49e65..29a4a4dbfe12 100644
@@ -438,7 +438,7 @@ static bool allow_file_dedupe(struct file *file)
 		return true;
 	if (uid_eq(current_fsuid(), file_inode(file)->i_uid))
 		return true;
-	if (!inode_permission(file_inode(file), MAY_WRITE))
+	if (!inode_permission(&init_user_ns, file_inode(file), MAY_WRITE))
 		return true;
 	return false;
 }
fs/udf/file.c +1 -1
index ad8eefad27d7..928283925d68 100644
@@ -183,7 +183,7 @@ long udf_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	long old_block, new_block;
 	int result;
 
-	if (inode_permission(inode, MAY_READ) != 0) {
+	if (inode_permission(&init_user_ns, inode, MAY_READ) != 0) {
 		udf_debug("no permission to access inode %lu\n", inode->i_ino);
 		return -EPERM;
 	}
fs/verity/enable.c +1 -1
index f7e997a01ad0..6809cf8a99b7 100644
@@ -369,7 +369,7 @@ int fsverity_ioctl_enable(struct file *filp, const void __user *uarg)
 	 * has verity enabled, and to stabilize the data being hashed.
 	 */
 
-	err = inode_permission(inode, MAY_WRITE);
+	err = inode_permission(&init_user_ns, inode, MAY_WRITE);
 	if (err)
 		return err;
 
fs/xattr.c +1 -1
index fd57153b1f61..56151bd9e642 100644
@@ -131,7 +131,7 @@ xattr_permission(struct inode *inode, const char *name, int mask)
 			return -EPERM;
 	}
 
-	return inode_permission(inode, mask);
+	return inode_permission(&init_user_ns, inode, mask);
 }
 
 /*
include/linux/fs.h +2 -2
index 0e2b8d235dca..bf62306c22a8 100644
@@ -2805,8 +2805,8 @@ static inline int bmap(struct inode *inode,  sector_t *block)
 #endif
 
 extern int notify_change(struct dentry *, struct iattr *, struct inode **);
-extern int inode_permission(struct inode *, int);
-extern int generic_permission(struct inode *, int);
+extern int inode_permission(struct user_namespace *, struct inode *, int);
+extern int generic_permission(struct user_namespace *, struct inode *, int);
 extern int __check_sticky(struct inode *dir, struct inode *inode);
 
 static inline bool execute_ok(struct inode *inode)
include/linux/posix_acl.h +3 -1
index 90797f1b421d..8276baefed13 100644
@@ -15,6 +15,8 @@
 #include <linux/refcount.h>
 #include <uapi/linux/posix_acl.h>
 
+struct user_namespace;
+
 struct posix_acl_entry {
 	short			e_tag;
 	unsigned short		e_perm;
@@ -62,7 +64,7 @@ posix_acl_release(struct posix_acl *acl)
 extern void posix_acl_init(struct posix_acl *, int);
 extern struct posix_acl *posix_acl_alloc(int, gfp_t);
 extern int posix_acl_valid(struct user_namespace *, const struct posix_acl *);
-extern int posix_acl_permission(struct inode *, const struct posix_acl *, int);
+extern int posix_acl_permission(struct user_namespace *, struct inode *, const struct posix_acl *, int);
 extern struct posix_acl *posix_acl_from_mode(umode_t, gfp_t);
 extern int posix_acl_equiv_mode(const struct posix_acl *, umode_t *);
 extern int __posix_acl_create(struct posix_acl **, gfp_t, umode_t *);
ipc/mqueue.c +1 -1
index beff0cfcd1e8..693f01fe1216 100644
@@ -873,7 +873,7 @@ static int prepare_open(struct dentry *dentry, int oflag, int ro,
 	if ((oflag & O_ACCMODE) == (O_RDWR | O_WRONLY))
 		return -EINVAL;
 	acc = oflag2acc[oflag & O_ACCMODE];
-	return inode_permission(d_inode(dentry), acc);
+	return inode_permission(&init_user_ns, d_inode(dentry), acc);
 }
 
 static int do_mq_open(const char __user *u_name, int oflag, umode_t mode,
kernel/bpf/inode.c +2 -2
index dd4b7fd60ee7..f1c393e5d47d 100644
@@ -507,7 +507,7 @@ static void *bpf_obj_do_get(const char __user *pathname,
 		return ERR_PTR(ret);
 
 	inode = d_backing_inode(path.dentry);
-	ret = inode_permission(inode, ACC_MODE(flags));
+	ret = inode_permission(&init_user_ns, inode, ACC_MODE(flags));
 	if (ret)
 		goto out;
 
@@ -558,7 +558,7 @@ int bpf_obj_get_user(const char __user *pathname, int flags)
 static struct bpf_prog *__get_prog_inode(struct inode *inode, enum bpf_prog_type type)
 {
 	struct bpf_prog *prog;
-	int ret = inode_permission(inode, MAY_READ);
+	int ret = inode_permission(&init_user_ns, inode, MAY_READ);
 	if (ret)
 		return ERR_PTR(ret);
 
kernel/cgroup/cgroup.c +1 -1
index 613845769103..091ffb5d2939 100644
@@ -4670,7 +4670,7 @@ static int cgroup_may_write(const struct cgroup *cgrp, struct super_block *sb)
 	if (!inode)
 		return -ENOMEM;
 
-	ret = inode_permission(inode, MAY_WRITE);
+	ret = inode_permission(&init_user_ns, inode, MAY_WRITE);
 	iput(inode);
 	return ret;
 }
kernel/sys.c +1 -1
index 51f00fe20e4d..41d7681e1f7e 100644
@@ -1848,7 +1848,7 @@ static int prctl_set_mm_exe_file(struct mm_struct *mm, unsigned int fd)
 	if (!S_ISREG(inode->i_mode) || path_noexec(&exe.file->f_path))
 		goto exit;
 
-	err = inode_permission(inode, MAY_EXEC);
+	err = inode_permission(&init_user_ns, inode, MAY_EXEC);
 	if (err)
 		goto exit;
 
mm/madvise.c +1 -1
index 6a660858784b..bb44b1a4534b 100644
@@ -540,7 +540,7 @@ static inline bool can_do_pageout(struct vm_area_struct *vma)
 	 * opens a side channel.
 	 */
 	return inode_owner_or_capable(file_inode(vma->vm_file)) ||
-		inode_permission(file_inode(vma->vm_file), MAY_WRITE) == 0;
+		inode_permission(&init_user_ns, file_inode(vma->vm_file), MAY_WRITE) == 0;
 }
 
 static long madvise_pageout(struct vm_area_struct *vma,
mm/memcontrol.c +1 -1
index 605f671203ef..8db67495b049 100644
@@ -4899,7 +4899,7 @@ static ssize_t memcg_write_event_control(struct kernfs_open_file *of,
 
 	/* the process need read permission on control file */
 	/* AV: shouldn't we check that it's been opened for read instead? */
-	ret = inode_permission(file_inode(cfile.file), MAY_READ);
+	ret = inode_permission(&init_user_ns, file_inode(cfile.file), MAY_READ);
 	if (ret < 0)
 		goto out_put_cfile;
 
mm/mincore.c +1 -1
index 02db1a834021..d5a58e61eac6 100644
@@ -167,7 +167,7 @@ static inline bool can_do_mincore(struct vm_area_struct *vma)
 	 * mappings, which opens a side channel.
 	 */
 	return inode_owner_or_capable(file_inode(vma->vm_file)) ||
-		inode_permission(file_inode(vma->vm_file), MAY_WRITE) == 0;
+		inode_permission(&init_user_ns, file_inode(vma->vm_file), MAY_WRITE) == 0;
 }
 
 static const struct mm_walk_ops mincore_walk_ops = {
net/unix/af_unix.c +1 -1
index 41c3303c3357..f568526d4a02 100644
@@ -936,7 +936,7 @@ static struct sock *unix_find_other(struct net *net,
 		if (err)
 			goto fail;
 		inode = d_backing_inode(path.dentry);
-		err = inode_permission(inode, MAY_WRITE);
+		err = inode_permission(&init_user_ns, inode, MAY_WRITE);
 		if (err)
 			goto put_fail;
 
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
Reviewed-by on v5 11/42
#14 lore raw

Re: [PATCH v5 11/42] namei: make permission helpers idmapped mount aware

Quoted 9 lines · Christian Brauner

The two helpers inode_permission() and generic_permission() are used by the vfs to perform basic permission checking by verifying that the caller is privileged over an inode. In order to handle idmapped mounts we extend the two helpers with an additional user namespace argument. On idmapped mounts the two helpers will make sure to map the inode according to the mount's user namespace and then peform identical permission checks to inode_permission() and generic_permission(). If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

More long lines in here.

Otherwise looks good:

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers

#15 lore raw

PATCH v5 12/42 inode: make init and permission helpers idmapped mount aware

In reply to Christian Brauner

The inode_owner_or_capable() helper determines whether the caller is the owner of the inode or is capable with respect to that inode. Allow it to handle idmapped mounts. If the inode is accessed through an idmapped mount we first need to map it according to 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.

Similarly, we allow the inode_init_owner() helper to handle idmapped mounts. It initializes a new inode on idmapped mounts by mapping the fsuid and fsgid of the caller from the mount's user namespace. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Don't pollute the vfs with additional helpers simply extend the existing helpers with an additional argument and switch all callers.

/* v3 */
unchanged

/* v4 */
- Christoph Hellwig <[email protected]>:
- Change commit message to reflect the fact that no new permission helpers are introduced but only the existing ones changed.

- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.
- Note that inode_owner_or_capable() contains a comparison of the caller's fsuid to the inode's i_uid to determine whether the caller is the owner of the inode. Serge asked whether it is possible that the caller's fsuid is INVALID_UID. If so and i_uid_into_mnt() makes the inode's i_uid INVALID_UID too then the caller could potentially gain privilege over an inode it doesn't own. But this is not feasible since the callers fsuid should never be INVALID_UID unless we have a severe bug how we manage set*id() calls or cred management. Also, if it is possible that the caller's fsuid were to be INVALID_UID we might already have an issue in case a filesystem was mounted with an s_user_ns that doesn't contain a mapping for a inode that was created when the filesystem was mounted with a different user namespace.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/9p/acl.c                  |  2 +-
 fs/9p/vfs_inode.c            |  2 +-
 fs/attr.c                    |  6 +++---
 fs/bfs/dir.c                 |  2 +-
 fs/btrfs/inode.c             |  2 +-
 fs/btrfs/ioctl.c             | 10 +++++-----
 fs/btrfs/tests/btrfs-tests.c |  2 +-
 fs/crypto/policy.c           |  2 +-
 fs/efivarfs/file.c           |  2 +-
 fs/ext2/ialloc.c             |  2 +-
 fs/ext2/ioctl.c              |  6 +++---
 fs/ext4/ialloc.c             |  2 +-
 fs/ext4/ioctl.c              | 14 +++++++-------
 fs/f2fs/file.c               | 14 +++++++-------
 fs/f2fs/namei.c              |  2 +-
 fs/f2fs/xattr.c              |  2 +-
 fs/fcntl.c                   |  2 +-
 fs/gfs2/file.c               |  2 +-
 fs/hfsplus/inode.c           |  2 +-
 fs/hfsplus/ioctl.c           |  2 +-
 fs/hugetlbfs/inode.c         |  2 +-
 fs/inode.c                   | 35 ++++++++++++++++++++++++++---------
 fs/jfs/ioctl.c               |  2 +-
 fs/jfs/jfs_inode.c           |  2 +-
 fs/minix/bitmap.c            |  2 +-
 fs/namei.c                   |  4 ++--
 fs/nilfs2/inode.c            |  2 +-
 fs/nilfs2/ioctl.c            |  2 +-
 fs/ocfs2/dlmfs/dlmfs.c       |  4 ++--
 fs/ocfs2/ioctl.c             |  2 +-
 fs/ocfs2/namei.c             |  2 +-
 fs/omfs/inode.c              |  2 +-
 fs/overlayfs/dir.c           |  2 +-
 fs/overlayfs/file.c          |  4 ++--
 fs/overlayfs/super.c         |  2 +-
 fs/overlayfs/util.c          |  2 +-
 fs/posix_acl.c               |  2 +-
 fs/ramfs/inode.c             |  2 +-
 fs/reiserfs/ioctl.c          |  4 ++--
 fs/reiserfs/namei.c          |  2 +-
 fs/sysv/ialloc.c             |  2 +-
 fs/ubifs/dir.c               |  2 +-
 fs/ubifs/ioctl.c             |  2 +-
 fs/udf/ialloc.c              |  2 +-
 fs/ufs/ialloc.c              |  2 +-
 fs/xattr.c                   |  2 +-
 fs/xfs/xfs_ioctl.c           |  2 +-
 fs/zonefs/super.c            |  2 +-
 include/linux/fs.h           |  9 +++++----
 kernel/bpf/inode.c           |  2 +-
 mm/madvise.c                 |  2 +-
 mm/mincore.c                 |  2 +-
 mm/shmem.c                   |  2 +-
 security/selinux/hooks.c     |  4 ++--
 54 files changed, 108 insertions(+), 90 deletions(-)
fs/9p/acl.c +1 -1
index 6261719f6f2a..d77b28e8d57a 100644
@@ -258,7 +258,7 @@ static int v9fs_xattr_set_acl(const struct xattr_handler *handler,
 
 	if (S_ISLNK(inode->i_mode))
 		return -EOPNOTSUPP;
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EPERM;
 	if (value) {
 		/* update the cached acl value */
fs/9p/vfs_inode.c +1 -1
index 4a937fac1acb..f66eb3c12c8a 100644
@@ -251,7 +251,7 @@ int v9fs_init_inode(struct v9fs_session_info *v9ses,
 {
 	int err = 0;
 
-	inode_init_owner(inode, NULL, mode);
+	inode_init_owner(&init_user_ns,inode,  NULL, mode);
 	inode->i_blocks = 0;
 	inode->i_rdev = rdev;
 	inode->i_atime = inode->i_mtime = inode->i_ctime = current_time(inode);
fs/attr.c +3 -3
index c9e29e589cec..00ae0b000146 100644
@@ -87,7 +87,7 @@ int setattr_prepare(struct dentry *dentry, struct iattr *attr)
 
 	/* Make sure a caller can chmod. */
 	if (ia_valid & ATTR_MODE) {
-		if (!inode_owner_or_capable(inode))
+		if (!inode_owner_or_capable(&init_user_ns, inode))
 			return -EPERM;
 		/* Also check the setgid bit! */
 		if (!in_group_p((ia_valid & ATTR_GID) ? attr->ia_gid :
@@ -98,7 +98,7 @@ int setattr_prepare(struct dentry *dentry, struct iattr *attr)
 
 	/* Check for setting the inode time. */
 	if (ia_valid & (ATTR_MTIME_SET | ATTR_ATIME_SET | ATTR_TIMES_SET)) {
-		if (!inode_owner_or_capable(inode))
+		if (!inode_owner_or_capable(&init_user_ns, inode))
 			return -EPERM;
 	}
 
@@ -243,7 +243,7 @@ int notify_change(struct dentry * dentry, struct iattr * attr, struct inode **de
 		if (IS_IMMUTABLE(inode))
 			return -EPERM;
 
-		if (!inode_owner_or_capable(inode)) {
+		if (!inode_owner_or_capable(&init_user_ns, inode)) {
 			error = inode_permission(&init_user_ns, inode,
 						 MAY_WRITE);
 			if (error)
fs/bfs/dir.c +1 -1
index d8dfe3a0cb39..be1335a8d25b 100644
@@ -96,7 +96,7 @@ static int bfs_create(struct inode *dir, struct dentry *dentry, umode_t mode,
 	}
 	set_bit(ino, info->si_imap);
 	info->si_freei--;
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
 	inode->i_blocks = 0;
 	inode->i_op = &bfs_file_inops;
fs/btrfs/inode.c +1 -1
index f834ba3b16b7..550f89553b80 100644
@@ -6190,7 +6190,7 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 	if (ret != 0)
 		goto fail_unlock;
 
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 	inode_set_bytes(inode, 0);
 
 	inode->i_mtime = current_time(inode);
fs/btrfs/ioctl.c +5 -5
index 7b47af3d8d5a..950b4f1b9092 100644
@@ -213,7 +213,7 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
 	const char *comp = NULL;
 	u32 binode_flags;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EPERM;
 
 	if (btrfs_root_readonly(root))
@@ -429,7 +429,7 @@ static int btrfs_ioctl_fssetxattr(struct file *file, void __user *arg)
 	unsigned old_i_flags;
 	int ret = 0;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EPERM;
 
 	if (btrfs_root_readonly(root))
@@ -1862,7 +1862,7 @@ static noinline int __btrfs_ioctl_snap_create(struct file *file,
 			btrfs_info(BTRFS_I(file_inode(file))->root->fs_info,
 				   "Snapshot src from another FS");
 			ret = -EXDEV;
-		} else if (!inode_owner_or_capable(src_inode)) {
+		} else if (!inode_owner_or_capable(&init_user_ns, src_inode)) {
 			/*
 			 * Subvolume creation is not restricted, but snapshots
 			 * are limited to own subvolumes only
@@ -1982,7 +1982,7 @@ static noinline int btrfs_ioctl_subvol_setflags(struct file *file,
 	u64 flags;
 	int ret = 0;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EPERM;
 
 	ret = mnt_want_write_file(file);
@@ -4451,7 +4451,7 @@ static long _btrfs_ioctl_set_received_subvol(struct file *file,
 	int ret = 0;
 	int received_uuid_changed;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EPERM;
 
 	ret = mnt_want_write_file(file);
fs/btrfs/tests/btrfs-tests.c +1 -1
index 6bd97bd4cb37..3a4099a2bf05 100644
@@ -62,7 +62,7 @@ struct inode *btrfs_new_test_inode(void)
 	BTRFS_I(inode)->location.type = BTRFS_INODE_ITEM_KEY;
 	BTRFS_I(inode)->location.objectid = BTRFS_FIRST_FREE_OBJECTID;
 	BTRFS_I(inode)->location.offset = 0;
-	inode_init_owner(inode, NULL, S_IFREG);
+	inode_init_owner(&init_user_ns, inode, NULL, S_IFREG);
 
 	return inode;
 }
fs/crypto/policy.c +1 -1
index a51cef6bd27f..ed3d623724cd 100644
@@ -465,7 +465,7 @@ int fscrypt_ioctl_set_policy(struct file *filp, const void __user *arg)
 		return -EFAULT;
 	policy.version = version;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EACCES;
 
 	ret = mnt_want_write_file(filp);
fs/efivarfs/file.c +1 -1
index feaa5e182b7b..e6bc0302643b 100644
@@ -137,7 +137,7 @@ efivarfs_ioc_setxflags(struct file *file, void __user *arg)
 	unsigned int oldflags = efivarfs_getflags(inode);
 	int error;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EACCES;
 
 	if (copy_from_user(&flags, arg, sizeof(flags)))
fs/ext2/ialloc.c +1 -1
index 432c3febea6d..df14e750e9fe 100644
@@ -551,7 +551,7 @@ struct inode *ext2_new_inode(struct inode *dir, umode_t mode,
 		inode->i_uid = current_fsuid();
 		inode->i_gid = dir->i_gid;
 	} else
-		inode_init_owner(inode, dir, mode);
+		inode_init_owner(&init_user_ns, inode, dir, mode);
 
 	inode->i_ino = ino;
 	inode->i_blocks = 0;
fs/ext2/ioctl.c +3 -3
index 32a8d10b579d..b399cbb7022d 100644
@@ -39,7 +39,7 @@ long ext2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 		if (ret)
 			return ret;
 
-		if (!inode_owner_or_capable(inode)) {
+		if (!inode_owner_or_capable(&init_user_ns, inode)) {
 			ret = -EACCES;
 			goto setflags_out;
 		}
@@ -84,7 +84,7 @@ long ext2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	case EXT2_IOC_SETVERSION: {
 		__u32 generation;
 
-		if (!inode_owner_or_capable(inode))
+		if (!inode_owner_or_capable(&init_user_ns, inode))
 			return -EPERM;
 		ret = mnt_want_write_file(filp);
 		if (ret)
@@ -117,7 +117,7 @@ long ext2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 		if (!test_opt(inode->i_sb, RESERVATION) ||!S_ISREG(inode->i_mode))
 			return -ENOTTY;
 
-		if (!inode_owner_or_capable(inode))
+		if (!inode_owner_or_capable(&init_user_ns, inode))
 			return -EACCES;
 
 		if (get_user(rsv_window_size, (int __user *)arg))
fs/ext4/ialloc.c +1 -1
index b215c564bc31..00c1ec6eee16 100644
@@ -972,7 +972,7 @@ struct inode *__ext4_new_inode(handle_t *handle, struct inode *dir,
 		inode->i_uid = current_fsuid();
 		inode->i_gid = dir->i_gid;
 	} else
-		inode_init_owner(inode, dir, mode);
+		inode_init_owner(&init_user_ns, inode, dir, mode);
 
 	if (ext4_has_feature_project(sb) &&
 	    ext4_test_inode_flag(dir, EXT4_INODE_PROJINHERIT))
fs/ext4/ioctl.c +7 -7
index 524e13432447..42335ae5dd94 100644
@@ -139,7 +139,7 @@ static long swap_inode_boot_loader(struct super_block *sb,
 	}
 
 	if (IS_RDONLY(inode) || IS_APPEND(inode) || IS_IMMUTABLE(inode) ||
-	    !inode_owner_or_capable(inode) || !capable(CAP_SYS_ADMIN)) {
+	    !inode_owner_or_capable(&init_user_ns, inode) || !capable(CAP_SYS_ADMIN)) {
 		err = -EPERM;
 		goto journal_err_out;
 	}
@@ -829,7 +829,7 @@ static long __ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	case FS_IOC_SETFLAGS: {
 		int err;
 
-		if (!inode_owner_or_capable(inode))
+		if (!inode_owner_or_capable(&init_user_ns, inode))
 			return -EACCES;
 
 		if (get_user(flags, (int __user *) arg))
@@ -871,7 +871,7 @@ static long __ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 		__u32 generation;
 		int err;
 
-		if (!inode_owner_or_capable(inode))
+		if (!inode_owner_or_capable(&init_user_ns, inode))
 			return -EPERM;
 
 		if (ext4_has_metadata_csum(inode->i_sb)) {
@@ -1010,7 +1010,7 @@ static long __ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	case EXT4_IOC_MIGRATE:
 	{
 		int err;
-		if (!inode_owner_or_capable(inode))
+		if (!inode_owner_or_capable(&init_user_ns, inode))
 			return -EACCES;
 
 		err = mnt_want_write_file(filp);
@@ -1032,7 +1032,7 @@ static long __ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	case EXT4_IOC_ALLOC_DA_BLKS:
 	{
 		int err;
-		if (!inode_owner_or_capable(inode))
+		if (!inode_owner_or_capable(&init_user_ns, inode))
 			return -EACCES;
 
 		err = mnt_want_write_file(filp);
@@ -1214,7 +1214,7 @@ static long __ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 
 	case EXT4_IOC_CLEAR_ES_CACHE:
 	{
-		if (!inode_owner_or_capable(inode))
+		if (!inode_owner_or_capable(&init_user_ns, inode))
 			return -EACCES;
 		ext4_clear_inode_es(inode);
 		return 0;
@@ -1260,7 +1260,7 @@ static long __ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 			return -EFAULT;
 
 		/* Make sure caller has proper permission */
-		if (!inode_owner_or_capable(inode))
+		if (!inode_owner_or_capable(&init_user_ns, inode))
 			return -EACCES;
 
 		if (fa.fsx_xflags & ~EXT4_SUPPORTED_FS_XFLAGS)
fs/f2fs/file.c +7 -7
index f585545277d7..5fc0ff28b5dd 100644
@@ -1961,7 +1961,7 @@ static int f2fs_ioc_setflags(struct file *filp, unsigned long arg)
 	u32 iflags;
 	int ret;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EACCES;
 
 	if (get_user(fsflags, (int __user *)arg))
@@ -2008,7 +2008,7 @@ static int f2fs_ioc_start_atomic_write(struct file *filp)
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
 	int ret;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EACCES;
 
 	if (!S_ISREG(inode->i_mode))
@@ -2075,7 +2075,7 @@ static int f2fs_ioc_commit_atomic_write(struct file *filp)
 	struct inode *inode = file_inode(filp);
 	int ret;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EACCES;
 
 	ret = mnt_want_write_file(filp);
@@ -2117,7 +2117,7 @@ static int f2fs_ioc_start_volatile_write(struct file *filp)
 	struct inode *inode = file_inode(filp);
 	int ret;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EACCES;
 
 	if (!S_ISREG(inode->i_mode))
@@ -2152,7 +2152,7 @@ static int f2fs_ioc_release_volatile_write(struct file *filp)
 	struct inode *inode = file_inode(filp);
 	int ret;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EACCES;
 
 	ret = mnt_want_write_file(filp);
@@ -2181,7 +2181,7 @@ static int f2fs_ioc_abort_volatile_write(struct file *filp)
 	struct inode *inode = file_inode(filp);
 	int ret;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EACCES;
 
 	ret = mnt_want_write_file(filp);
@@ -3158,7 +3158,7 @@ static int f2fs_ioc_fssetxattr(struct file *filp, unsigned long arg)
 		return -EFAULT;
 
 	/* Make sure caller has proper permission */
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EACCES;
 
 	if (fa.fsx_xflags & ~F2FS_SUPPORTED_XFLAGS)
fs/f2fs/namei.c +1 -1
index 6edb1ab579a1..ad98926bacac 100644
@@ -46,7 +46,7 @@ static struct inode *f2fs_new_inode(struct inode *dir, umode_t mode)
 
 	nid_free = true;
 
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 
 	inode->i_ino = ino;
 	inode->i_blocks = 0;
fs/f2fs/xattr.c +1 -1
index 65afcc3cc68a..d772bf13a814 100644
@@ -114,7 +114,7 @@ static int f2fs_xattr_advise_set(const struct xattr_handler *handler,
 	unsigned char old_advise = F2FS_I(inode)->i_advise;
 	unsigned char new_advise;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EPERM;
 	if (value == NULL)
 		return -EINVAL;
fs/fcntl.c +1 -1
index 05b36b28f2e8..74d99731fd43 100644
@@ -46,7 +46,7 @@ static int setfl(int fd, struct file * filp, unsigned long arg)
 
 	/* O_NOATIME can only be set by the owner or superuser */
 	if ((arg & O_NOATIME) && !(filp->f_flags & O_NOATIME))
-		if (!inode_owner_or_capable(inode))
+		if (!inode_owner_or_capable(&init_user_ns, inode))
 			return -EPERM;
 
 	/* required for strict SunOS emulation */
fs/gfs2/file.c +1 -1
index b39b339feddc..1d994bdfffaa 100644
@@ -238,7 +238,7 @@ static int do_gfs2_set_flags(struct file *filp, u32 reqflags, u32 mask,
 		goto out;
 
 	error = -EACCES;
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		goto out;
 
 	error = 0;
fs/hfsplus/inode.c +1 -1
index e3da9e96b835..21357046b039 100644
@@ -376,7 +376,7 @@ struct inode *hfsplus_new_inode(struct super_block *sb, struct inode *dir,
 		return NULL;
 
 	inode->i_ino = sbi->next_cnid++;
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 	set_nlink(inode, 1);
 	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
 
fs/hfsplus/ioctl.c +1 -1
index ce15b9496b77..3edb1926d127 100644
@@ -91,7 +91,7 @@ static int hfsplus_ioctl_setflags(struct file *file, int __user *user_flags)
 	if (err)
 		goto out;
 
-	if (!inode_owner_or_capable(inode)) {
+	if (!inode_owner_or_capable(&init_user_ns, inode)) {
 		err = -EACCES;
 		goto out_drop_write;
 	}
fs/hugetlbfs/inode.c +1 -1
index b5c109703daa..6737929e443c 100644
@@ -836,7 +836,7 @@ static struct inode *hugetlbfs_get_inode(struct super_block *sb,
 		struct hugetlbfs_inode_info *info = HUGETLBFS_I(inode);
 
 		inode->i_ino = get_next_ino();
-		inode_init_owner(inode, dir, mode);
+		inode_init_owner(&init_user_ns, inode, dir, mode);
 		lockdep_set_class(&inode->i_mapping->i_mmap_rwsem,
 				&hugetlbfs_i_mmap_rwsem_key);
 		inode->i_mapping->a_ops = &hugetlbfs_aops;
fs/inode.c +26 -9
index e1a58950f5e6..a9ac97a27784 100644
@@ -2130,14 +2130,21 @@ EXPORT_SYMBOL(init_special_inode);
 
 /**
  * inode_init_owner - Init uid,gid,mode for new inode according to posix standards
+ * @mnt_userns:	User namespace of the mount the inode was created from
  * @inode: New inode
  * @dir: Directory inode
  * @mode: mode of the new inode
+ *
+ * If the inode has been created through an idmapped mount the user namespace of
+ * the vfsmount must be passed through @mnt_userns. This function will then take
+ * care to map the inode according to @mnt_userns before checking permissions
+ * and initializing i_uid and i_gid. On non-idmapped mounts or if permission
+ * checking is to be performed on the raw inode simply passs init_user_ns.
  */
-void inode_init_owner(struct inode *inode, const struct inode *dir,
-			umode_t mode)
+void inode_init_owner(struct user_namespace *mnt_userns, struct inode *inode,
+		      const struct inode *dir, umode_t mode)
 {
-	inode->i_uid = current_fsuid();
+	inode->i_uid = fsuid_into_mnt(mnt_userns);
 	if (dir && dir->i_mode & S_ISGID) {
 		inode->i_gid = dir->i_gid;
 
@@ -2145,31 +2152,41 @@ void inode_init_owner(struct inode *inode, const struct inode *dir,
 		if (S_ISDIR(mode))
 			mode |= S_ISGID;
 		else if ((mode & (S_ISGID | S_IXGRP)) == (S_ISGID | S_IXGRP) &&
-			 !in_group_p(inode->i_gid) &&
-			 !capable_wrt_inode_uidgid(&init_user_ns, dir, CAP_FSETID))
+			 !in_group_p(i_gid_into_mnt(mnt_userns, dir)) &&
+			 !capable_wrt_inode_uidgid(mnt_userns, dir, CAP_FSETID))
 			mode &= ~S_ISGID;
 	} else
-		inode->i_gid = current_fsgid();
+		inode->i_gid = fsgid_into_mnt(mnt_userns);
 	inode->i_mode = mode;
 }
 EXPORT_SYMBOL(inode_init_owner);
 
 /**
  * inode_owner_or_capable - check current task permissions to inode
+ * @mnt_userns:	user namespace of the mount the inode was found from
  * @inode: inode being checked
  *
  * Return true if current either has CAP_FOWNER in a namespace with the
  * inode owner uid mapped, or owns the file.
+ *
+ * If the inode has been found through an idmapped mount the user namespace of
+ * the vfsmount must be passed through @mnt_userns. This function will then take
+ * care to map the inode according to @mnt_userns before checking permissions.
+ * On non-idmapped mounts or if permission checking is to be performed on the
+ * raw inode simply passs init_user_ns.
  */
-bool inode_owner_or_capable(const struct inode *inode)
+bool inode_owner_or_capable(struct user_namespace *mnt_userns,
+			    const struct inode *inode)
 {
+	kuid_t i_uid;
 	struct user_namespace *ns;
 
-	if (uid_eq(current_fsuid(), inode->i_uid))
+	i_uid = i_uid_into_mnt(mnt_userns, inode);
+	if (uid_eq(current_fsuid(), i_uid))
 		return true;
 
 	ns = current_user_ns();
-	if (kuid_has_mapping(ns, inode->i_uid) && ns_capable(ns, CAP_FOWNER))
+	if (kuid_has_mapping(ns, i_uid) && ns_capable(ns, CAP_FOWNER))
 		return true;
 	return false;
 }
fs/jfs/ioctl.c +1 -1
index 10ee0ecca1a8..2581d4db58ff 100644
@@ -76,7 +76,7 @@ long jfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 		if (err)
 			return err;
 
-		if (!inode_owner_or_capable(inode)) {
+		if (!inode_owner_or_capable(&init_user_ns, inode)) {
 			err = -EACCES;
 			goto setflags_out;
 		}
fs/jfs/jfs_inode.c +1 -1
index 4cef170630db..59379089e939 100644
@@ -64,7 +64,7 @@ struct inode *ialloc(struct inode *parent, umode_t mode)
 		goto fail_put;
 	}
 
-	inode_init_owner(inode, parent, mode);
+	inode_init_owner(&init_user_ns, inode, parent, mode);
 	/*
 	 * New inodes need to save sane values on disk when
 	 * uid & gid mount options are used
fs/minix/bitmap.c +1 -1
index f4e5e5181a14..9115948c624e 100644
@@ -252,7 +252,7 @@ struct inode *minix_new_inode(const struct inode *dir, umode_t mode, int *error)
 		iput(inode);
 		return NULL;
 	}
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 	inode->i_ino = j;
 	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
 	inode->i_blocks = 0;
fs/namei.c +2 -2
index 630a3b5be249..cd124e18cec5 100644
@@ -1086,7 +1086,7 @@ int may_linkat(struct path *link)
 	/* Source inode owner (or CAP_FOWNER) can hardlink all they like,
 	 * otherwise, it must be a safe source.
 	 */
-	if (safe_hardlink_source(inode) || inode_owner_or_capable(inode))
+	if (safe_hardlink_source(inode) || inode_owner_or_capable(&init_user_ns, inode))
 		return 0;
 
 	audit_log_path_denied(AUDIT_ANOM_LINK, "linkat");
@@ -2938,7 +2938,7 @@ static int may_open(const struct path *path, int acc_mode, int flag)
 	}
 
 	/* O_NOATIME can only be set by the owner or superuser */
-	if (flag & O_NOATIME && !inode_owner_or_capable(inode))
+	if (flag & O_NOATIME && !inode_owner_or_capable(&init_user_ns, inode))
 		return -EPERM;
 
 	return 0;
fs/nilfs2/inode.c +1 -1
index b6517220cad5..11225a659736 100644
@@ -348,7 +348,7 @@ struct inode *nilfs_new_inode(struct inode *dir, umode_t mode)
 	/* reference count of i_bh inherits from nilfs_mdt_read_block() */
 
 	atomic64_inc(&root->inodes_count);
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 	inode->i_ino = ino;
 	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
 
fs/nilfs2/ioctl.c +1 -1
index 07d26f61f22a..b053b40315bf 100644
@@ -132,7 +132,7 @@ static int nilfs_ioctl_setflags(struct inode *inode, struct file *filp,
 	unsigned int flags, oldflags;
 	int ret;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EACCES;
 
 	if (get_user(flags, (int __user *)argp))
fs/ocfs2/dlmfs/dlmfs.c +2 -2
index 583820ec63e2..37c7d03a6284 100644
@@ -329,7 +329,7 @@ static struct inode *dlmfs_get_root_inode(struct super_block *sb)
 
 	if (inode) {
 		inode->i_ino = get_next_ino();
-		inode_init_owner(inode, NULL, mode);
+		inode_init_owner(&init_user_ns, inode, NULL, mode);
 		inode->i_atime = inode->i_mtime = inode->i_ctime = current_time(inode);
 		inc_nlink(inode);
 
@@ -352,7 +352,7 @@ static struct inode *dlmfs_get_inode(struct inode *parent,
 		return NULL;
 
 	inode->i_ino = get_next_ino();
-	inode_init_owner(inode, parent, mode);
+	inode_init_owner(&init_user_ns, inode, parent, mode);
 	inode->i_atime = inode->i_mtime = inode->i_ctime = current_time(inode);
 
 	ip = DLMFS_I(inode);
fs/ocfs2/ioctl.c +1 -1
index 89984172fc4a..50c9b30ee9f6 100644
@@ -96,7 +96,7 @@ static int ocfs2_set_inode_attr(struct inode *inode, unsigned flags,
 	}
 
 	status = -EACCES;
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		goto bail_unlock;
 
 	if (!S_ISDIR(inode->i_mode))
fs/ocfs2/namei.c +1 -1
index 2a237ab00453..908b79e1082b 100644
@@ -198,7 +198,7 @@ static struct inode *ocfs2_get_init_inode(struct inode *dir, umode_t mode)
 	 * callers. */
 	if (S_ISDIR(mode))
 		set_nlink(inode, 2);
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 	status = dquot_initialize(inode);
 	if (status)
 		return ERR_PTR(status);
fs/omfs/inode.c +1 -1
index ce93ccca8639..2a0e83236c01 100644
@@ -48,7 +48,7 @@ struct inode *omfs_new_inode(struct inode *dir, umode_t mode)
 		goto fail;
 
 	inode->i_ino = new_block;
-	inode_init_owner(inode, NULL, mode);
+	inode_init_owner(&init_user_ns, inode, NULL, mode);
 	inode->i_mapping->a_ops = &omfs_aops;
 
 	inode->i_atime = inode->i_mtime = inode->i_ctime = current_time(inode);
fs/overlayfs/dir.c +1 -1
index 28a075b5f5b2..98a23353b19a 100644
@@ -636,7 +636,7 @@ static int ovl_create_object(struct dentry *dentry, int mode, dev_t rdev,
 	inode->i_state |= I_CREATING;
 	spin_unlock(&inode->i_lock);
 
-	inode_init_owner(inode, dentry->d_parent->d_inode, mode);
+	inode_init_owner(&init_user_ns, inode, dentry->d_parent->d_inode, mode);
 	attr.mode = inode->i_mode;
 
 	err = ovl_create_or_link(dentry, inode, &attr, false);
fs/overlayfs/file.c +2 -2
index b2948e7b3210..7d8b84c715b3 100644
@@ -54,7 +54,7 @@ static struct file *ovl_open_realfile(const struct file *file,
 	if (err) {
 		realfile = ERR_PTR(err);
 	} else {
-		if (!inode_owner_or_capable(realinode))
+		if (!inode_owner_or_capable(&init_user_ns, realinode))
 			flags &= ~O_NOATIME;
 
 		realfile = open_with_fake_path(&file->f_path, flags, realinode,
@@ -520,7 +520,7 @@ static long ovl_ioctl_set_flags(struct file *file, unsigned int cmd,
 	long ret;
 	struct inode *inode = file_inode(file);
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EACCES;
 
 	ret = mnt_want_write_file(file);
fs/overlayfs/super.c +1 -1
index 88d877787770..3e925deaa19a 100644
@@ -1005,7 +1005,7 @@ ovl_posix_acl_xattr_set(const struct xattr_handler *handler,
 		goto out_acl_release;
 	}
 	err = -EPERM;
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		goto out_acl_release;
 
 	posix_acl_release(acl);
fs/overlayfs/util.c +1 -1
index de5c2047a0e9..06013b7b1e87 100644
@@ -484,7 +484,7 @@ struct file *ovl_path_open(struct path *path, int flags)
 		return ERR_PTR(err);
 
 	/* O_NOATIME is an optimization, don't fail if not permitted */
-	if (inode_owner_or_capable(inode))
+	if (inode_owner_or_capable(&init_user_ns, inode))
 		flags |= O_NOATIME;
 
 	return dentry_open(path, flags, current_cred());
fs/posix_acl.c +1 -1
index 5d9fe2fb2953..9ce8214bfdac 100644
@@ -874,7 +874,7 @@ set_posix_acl(struct inode *inode, int type, struct posix_acl *acl)
 
 	if (type == ACL_TYPE_DEFAULT && !S_ISDIR(inode->i_mode))
 		return acl ? -EACCES : 0;
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EPERM;
 
 	if (acl) {
fs/ramfs/inode.c +1 -1
index ee179a81b3da..3fd4326f36b5 100644
@@ -67,7 +67,7 @@ struct inode *ramfs_get_inode(struct super_block *sb,
 
 	if (inode) {
 		inode->i_ino = get_next_ino();
-		inode_init_owner(inode, dir, mode);
+		inode_init_owner(&init_user_ns, inode, dir, mode);
 		inode->i_mapping->a_ops = &ramfs_aops;
 		mapping_set_gfp_mask(inode->i_mapping, GFP_HIGHUSER);
 		mapping_set_unevictable(inode->i_mapping);
fs/reiserfs/ioctl.c +2 -2
index adb21bea3d60..4f1cbd930179 100644
@@ -59,7 +59,7 @@ long reiserfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 			if (err)
 				break;
 
-			if (!inode_owner_or_capable(inode)) {
+			if (!inode_owner_or_capable(&init_user_ns, inode)) {
 				err = -EPERM;
 				goto setflags_out;
 			}
@@ -101,7 +101,7 @@ long reiserfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 		err = put_user(inode->i_generation, (int __user *)arg);
 		break;
 	case REISERFS_IOC_SETVERSION:
-		if (!inode_owner_or_capable(inode)) {
+		if (!inode_owner_or_capable(&init_user_ns, inode)) {
 			err = -EPERM;
 			break;
 		}
fs/reiserfs/namei.c +1 -1
index 1594687582f0..a67a7d371725 100644
@@ -615,7 +615,7 @@ static int new_inode_init(struct inode *inode, struct inode *dir, umode_t mode)
 	 * the quota init calls have to know who to charge the quota to, so
 	 * we have to set uid and gid here
 	 */
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 	return dquot_initialize(inode);
 }
 
fs/sysv/ialloc.c +1 -1
index 6c9801986af6..50df794a3c1f 100644
@@ -163,7 +163,7 @@ struct inode * sysv_new_inode(const struct inode * dir, umode_t mode)
 	*sbi->s_sb_fic_count = cpu_to_fs16(sbi, count);
 	fs16_add(sbi, sbi->s_sb_total_free_inodes, -1);
 	dirty_sb(sb);
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 	inode->i_ino = fs16_to_cpu(sbi, ino);
 	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
 	inode->i_blocks = 0;
fs/ubifs/dir.c +1 -1
index 9a6b8660425a..694e7714545b 100644
@@ -94,7 +94,7 @@ struct inode *ubifs_new_inode(struct ubifs_info *c, struct inode *dir,
 	 */
 	inode->i_flags |= S_NOCMTIME;
 
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 	inode->i_mtime = inode->i_atime = inode->i_ctime =
 			 current_time(inode);
 	inode->i_mapping->nrpages = 0;
fs/ubifs/ioctl.c +1 -1
index 4363d85a3fd4..2326d5122beb 100644
@@ -155,7 +155,7 @@ long ubifs_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 		if (IS_RDONLY(inode))
 			return -EROFS;
 
-		if (!inode_owner_or_capable(inode))
+		if (!inode_owner_or_capable(&init_user_ns, inode))
 			return -EACCES;
 
 		if (get_user(flags, (int __user *) arg))
fs/udf/ialloc.c +1 -1
index 84ed23edebfd..2ecf0e87660e 100644
@@ -103,7 +103,7 @@ struct inode *udf_new_inode(struct inode *dir, umode_t mode)
 		mutex_unlock(&sbi->s_alloc_mutex);
 	}
 
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 	if (UDF_QUERY_FLAG(sb, UDF_FLAG_UID_SET))
 		inode->i_uid = sbi->s_uid;
 	if (UDF_QUERY_FLAG(sb, UDF_FLAG_GID_SET))
fs/ufs/ialloc.c +1 -1
index 969fd60436d3..7e3e08c0166f 100644
@@ -289,7 +289,7 @@ struct inode *ufs_new_inode(struct inode *dir, umode_t mode)
 	ufs_mark_sb_dirty(sb);
 
 	inode->i_ino = cg * uspi->s_ipg + bit;
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 	inode->i_blocks = 0;
 	inode->i_generation = 0;
 	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
fs/xattr.c +1 -1
index 56151bd9e642..7773ed7e0562 100644
@@ -127,7 +127,7 @@ xattr_permission(struct inode *inode, const char *name, int mask)
 		if (!S_ISREG(inode->i_mode) && !S_ISDIR(inode->i_mode))
 			return (mask & MAY_WRITE) ? -EPERM : -ENODATA;
 		if (S_ISDIR(inode->i_mode) && (inode->i_mode & S_ISVTX) &&
-		    (mask & MAY_WRITE) && !inode_owner_or_capable(inode))
+		    (mask & MAY_WRITE) && !inode_owner_or_capable(&init_user_ns, inode))
 			return -EPERM;
 	}
 
fs/xfs/xfs_ioctl.c +1 -1
index 97bd29fc8c43..218e80afc859 100644
@@ -1300,7 +1300,7 @@ xfs_ioctl_setattr_get_trans(
 	 * The user ID of the calling process must be equal to the file owner
 	 * ID, except in cases where the CAP_FSETID capability is applicable.
 	 */
-	if (!inode_owner_or_capable(VFS_I(ip))) {
+	if (!inode_owner_or_capable(&init_user_ns, VFS_I(ip))) {
 		error = -EPERM;
 		goto out_cancel;
 	}
fs/zonefs/super.c +1 -1
index bec47f2d074b..569525ee8f69 100644
@@ -1223,7 +1223,7 @@ static void zonefs_init_dir_inode(struct inode *parent, struct inode *inode,
 	struct super_block *sb = parent->i_sb;
 
 	inode->i_ino = blkdev_nr_zones(sb->s_bdev->bd_disk) + type + 1;
-	inode_init_owner(inode, parent, S_IFDIR | 0555);
+	inode_init_owner(&init_user_ns, inode, parent, S_IFDIR | 0555);
 	inode->i_op = &zonefs_dir_inode_operations;
 	inode->i_fop = &simple_dir_operations;
 	set_nlink(inode, 2);
include/linux/fs.h +5 -4
index bf62306c22a8..430434317f9e 100644
@@ -1761,8 +1761,8 @@ static inline bool sb_start_intwrite_trylock(struct super_block *sb)
 	return __sb_start_write_trylock(sb, SB_FREEZE_FS);
 }
 
-
-extern bool inode_owner_or_capable(const struct inode *inode);
+extern bool inode_owner_or_capable(struct user_namespace *mnt_userns,
+				   const struct inode *inode);
 
 /*
  * VFS helper functions..
@@ -1804,8 +1804,9 @@ extern long compat_ptr_ioctl(struct file *file, unsigned int cmd,
 /*
  * VFS file helper functions.
  */
-extern void inode_init_owner(struct inode *inode, const struct inode *dir,
-			umode_t mode);
+extern void inode_init_owner(struct user_namespace *mnt_userns,
+			     struct inode *inode, const struct inode *dir,
+			     umode_t mode);
 extern bool may_open_dev(const struct path *path);
 
 /*
kernel/bpf/inode.c +1 -1
index f1c393e5d47d..dd6f124c12c7 100644
@@ -122,7 +122,7 @@ static struct inode *bpf_get_inode(struct super_block *sb,
 	inode->i_mtime = inode->i_atime;
 	inode->i_ctime = inode->i_atime;
 
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 
 	return inode;
 }
mm/madvise.c +1 -1
index bb44b1a4534b..6830ebe3fc5c 100644
@@ -539,7 +539,7 @@ static inline bool can_do_pageout(struct vm_area_struct *vma)
 	 * otherwise we'd be including shared non-exclusive mappings, which
 	 * opens a side channel.
 	 */
-	return inode_owner_or_capable(file_inode(vma->vm_file)) ||
+	return inode_owner_or_capable(&init_user_ns, file_inode(vma->vm_file)) ||
 		inode_permission(&init_user_ns, file_inode(vma->vm_file), MAY_WRITE) == 0;
 }
 
mm/mincore.c +1 -1
index d5a58e61eac6..ad2dfb7a4500 100644
@@ -166,7 +166,7 @@ static inline bool can_do_mincore(struct vm_area_struct *vma)
 	 * for writing; otherwise we'd be including shared non-exclusive
 	 * mappings, which opens a side channel.
 	 */
-	return inode_owner_or_capable(file_inode(vma->vm_file)) ||
+	return inode_owner_or_capable(&init_user_ns, file_inode(vma->vm_file)) ||
 		inode_permission(&init_user_ns, file_inode(vma->vm_file), MAY_WRITE) == 0;
 }
 
mm/shmem.c +1 -1
index 7c6b6d8f6c39..1c68c9edba5e 100644
@@ -2303,7 +2303,7 @@ static struct inode *shmem_get_inode(struct super_block *sb, const struct inode
 	inode = new_inode(sb);
 	if (inode) {
 		inode->i_ino = ino;
-		inode_init_owner(inode, dir, mode);
+		inode_init_owner(&init_user_ns, inode, dir, mode);
 		inode->i_blocks = 0;
 		inode->i_atime = inode->i_mtime = inode->i_ctime = current_time(inode);
 		inode->i_generation = prandom_u32();
security/selinux/hooks.c +2 -2
index 644b17ec9e63..9d6d3da2caf2 100644
@@ -3140,13 +3140,13 @@ static int selinux_inode_setxattr(struct dentry *dentry, const char *name,
 	}
 
 	if (!selinux_initialized(&selinux_state))
-		return (inode_owner_or_capable(inode) ? 0 : -EPERM);
+		return (inode_owner_or_capable(&init_user_ns, inode) ? 0 : -EPERM);
 
 	sbsec = inode->i_sb->s_security;
 	if (!(sbsec->flags & SBLABEL_MNT))
 		return -EOPNOTSUPP;
 
-	if (!inode_owner_or_capable(inode))
+	if (!inode_owner_or_capable(&init_user_ns, inode))
 		return -EPERM;
 
 	ad.type = LSM_AUDIT_DATA_DENTRY;
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
Reviewed-by on v5 12/42
#16 lore raw

Re: [PATCH v5 12/42] inode: make init and permission helpers idmapped mount aware

include/linux/fs.h static inline bool sb_start_intwrite_trylock(struct super_block *sb) · line 1765

Quoted 8 lines
include/linux/fs.h +2 -2
@@ -1761,8 +1761,8 @@ static inline bool sb_start_intwrite_trylock(struct super_block *sb)
 	return __sb_start_write_trylock(sb, SB_FREEZE_FS);
 }
 
-
-extern bool inode_owner_or_capable(const struct inode *inode);
+extern bool inode_owner_or_capable(struct user_namespace *mnt_userns,
+				   const struct inode *inode);

While you're at it - it would be nice to drop the superflous externs in the headers.

Otherwise looks good:

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers

#17 lore raw

PATCH v5 13/42 attr: handle idmapped mounts

In reply to Christian Brauner

When file attributes are changed filesystems almost exclusively 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 it is mapped according to 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.

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Don't pollute the vfs with additional helpers simply extend the existing helpers with an additional argument and switch all callers.

/* v3 */
- kernel test robot <[email protected]>:
- Adapt __setattr_copy() in fs/f2fs when CONFIG_F2FS_FS_POSIX_ACL is selected.

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 arch/powerpc/platforms/cell/spufs/inode.c |   2 +-
 drivers/base/devtmpfs.c                   |   4 +-
 fs/9p/vfs_inode.c                         |   4 +-
 fs/9p/vfs_inode_dotl.c                    |   4 +-
 fs/adfs/inode.c                           |   2 +-
 fs/affs/inode.c                           |   4 +-
 fs/attr.c                                 | 117 ++++++++++++++++------
 fs/btrfs/inode.c                          |   4 +-
 fs/cachefiles/interface.c                 |   4 +-
 fs/ceph/inode.c                           |   2 +-
 fs/cifs/inode.c                           |   8 +-
 fs/ecryptfs/inode.c                       |   6 +-
 fs/exfat/file.c                           |   4 +-
 fs/ext2/inode.c                           |   4 +-
 fs/ext4/inode.c                           |   4 +-
 fs/f2fs/file.c                            |  10 +-
 fs/fat/file.c                             |   4 +-
 fs/fuse/dir.c                             |   2 +-
 fs/gfs2/inode.c                           |   4 +-
 fs/hfs/inode.c                            |   4 +-
 fs/hfsplus/inode.c                        |   4 +-
 fs/hostfs/hostfs_kern.c                   |   4 +-
 fs/hpfs/inode.c                           |   4 +-
 fs/hugetlbfs/inode.c                      |   4 +-
 fs/inode.c                                |   2 +-
 fs/jffs2/fs.c                             |   2 +-
 fs/jfs/file.c                             |   4 +-
 fs/kernfs/inode.c                         |   4 +-
 fs/libfs.c                                |   4 +-
 fs/minix/file.c                           |   4 +-
 fs/nfsd/nfsproc.c                         |   2 +-
 fs/nfsd/vfs.c                             |   4 +-
 fs/nilfs2/inode.c                         |   4 +-
 fs/ntfs/inode.c                           |   2 +-
 fs/ocfs2/dlmfs/dlmfs.c                    |   4 +-
 fs/ocfs2/file.c                           |   4 +-
 fs/omfs/file.c                            |   4 +-
 fs/open.c                                 |   6 +-
 fs/orangefs/inode.c                       |   4 +-
 fs/overlayfs/copy_up.c                    |   8 +-
 fs/overlayfs/dir.c                        |   2 +-
 fs/overlayfs/inode.c                      |   4 +-
 fs/overlayfs/super.c                      |   2 +-
 fs/proc/base.c                            |   4 +-
 fs/proc/generic.c                         |   4 +-
 fs/proc/proc_sysctl.c                     |   4 +-
 fs/ramfs/file-nommu.c                     |   4 +-
 fs/reiserfs/inode.c                       |   4 +-
 fs/sysv/file.c                            |   4 +-
 fs/ubifs/file.c                           |   2 +-
 fs/udf/file.c                             |   4 +-
 fs/ufs/inode.c                            |   4 +-
 fs/utimes.c                               |   2 +-
 fs/xfs/xfs_iops.c                         |   2 +-
 fs/zonefs/super.c                         |   4 +-
 include/linux/fs.h                        |   8 +-
 mm/shmem.c                                |   4 +-
 57 files changed, 200 insertions(+), 137 deletions(-)
arch/powerpc/platforms/cell/spufs/inode.c +1 -1
index 25390569e24c..3de526eb2275 100644
@@ -98,7 +98,7 @@ spufs_setattr(struct dentry *dentry, struct iattr *attr)
 	if ((attr->ia_valid & ATTR_SIZE) &&
 	    (attr->ia_size != inode->i_size))
 		return -EINVAL;
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
drivers/base/devtmpfs.c +2 -2
index eac184e6d657..2e0c3cdb4184 100644
@@ -221,7 +221,7 @@ static int handle_create(const char *nodename, umode_t mode, kuid_t uid,
 		newattrs.ia_gid = gid;
 		newattrs.ia_valid = ATTR_MODE|ATTR_UID|ATTR_GID;
 		inode_lock(d_inode(dentry));
-		notify_change(dentry, &newattrs, NULL);
+		notify_change(&init_user_ns, dentry, &newattrs, NULL);
 		inode_unlock(d_inode(dentry));
 
 		/* mark as kernel-created inode */
@@ -328,7 +328,7 @@ static int handle_remove(const char *nodename, struct device *dev)
 			newattrs.ia_valid =
 				ATTR_UID|ATTR_GID|ATTR_MODE;
 			inode_lock(d_inode(dentry));
-			notify_change(dentry, &newattrs, NULL);
+			notify_change(&init_user_ns, dentry, &newattrs, NULL);
 			inode_unlock(d_inode(dentry));
 			err = vfs_unlink(d_inode(parent.dentry), dentry, NULL);
 			if (!err || err == -ENOENT)
fs/9p/vfs_inode.c +2 -2
index f66eb3c12c8a..9c3ff6e9ab82 100644
@@ -1062,7 +1062,7 @@ static int v9fs_vfs_setattr(struct dentry *dentry, struct iattr *iattr)
 	struct p9_wstat wstat;
 
 	p9_debug(P9_DEBUG_VFS, "\n");
-	retval = setattr_prepare(dentry, iattr);
+	retval = setattr_prepare(&init_user_ns, dentry, iattr);
 	if (retval)
 		return retval;
 
@@ -1118,7 +1118,7 @@ static int v9fs_vfs_setattr(struct dentry *dentry, struct iattr *iattr)
 
 	v9fs_invalidate_inode_attr(d_inode(dentry));
 
-	setattr_copy(d_inode(dentry), iattr);
+	setattr_copy(&init_user_ns, d_inode(dentry), iattr);
 	mark_inode_dirty(d_inode(dentry));
 	return 0;
 }
fs/9p/vfs_inode_dotl.c +2 -2
index 823c2eb5f1bf..302553101fcb 100644
@@ -549,7 +549,7 @@ int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr)
 
 	p9_debug(P9_DEBUG_VFS, "\n");
 
-	retval = setattr_prepare(dentry, iattr);
+	retval = setattr_prepare(&init_user_ns, dentry, iattr);
 	if (retval)
 		return retval;
 
@@ -590,7 +590,7 @@ int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr)
 		truncate_setsize(inode, iattr->ia_size);
 
 	v9fs_invalidate_inode_attr(inode);
-	setattr_copy(inode, iattr);
+	setattr_copy(&init_user_ns, inode, iattr);
 	mark_inode_dirty(inode);
 	if (iattr->ia_valid & ATTR_MODE) {
 		/* We also want to update ACL when we update mode bits */
fs/adfs/inode.c +1 -1
index 32620f4a7623..278dcee6ae22 100644
@@ -299,7 +299,7 @@ adfs_notify_change(struct dentry *dentry, struct iattr *attr)
 	unsigned int ia_valid = attr->ia_valid;
 	int error;
 	
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 
 	/*
 	 * we can't change the UID or GID of any file -
fs/affs/inode.c +2 -2
index 044412110b52..767e5bdfb703 100644
@@ -223,7 +223,7 @@ affs_notify_change(struct dentry *dentry, struct iattr *attr)
 
 	pr_debug("notify_change(%lu,0x%x)\n", inode->i_ino, attr->ia_valid);
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		goto out;
 
@@ -249,7 +249,7 @@ affs_notify_change(struct dentry *dentry, struct iattr *attr)
 		affs_truncate(inode);
 	}
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 
 	if (attr->ia_valid & ATTR_MODE)
fs/attr.c +88 -29

224 lines not on this page · raw message

fs/btrfs/inode.c +2 -2
index 550f89553b80..19d432a453b1 100644
@@ -5054,7 +5054,7 @@ static int btrfs_setattr(struct dentry *dentry, struct iattr *attr)
 	if (btrfs_root_readonly(root))
 		return -EROFS;
 
-	err = setattr_prepare(dentry, attr);
+	err = setattr_prepare(&init_user_ns, dentry, attr);
 	if (err)
 		return err;
 
@@ -5065,7 +5065,7 @@ static int btrfs_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (attr->ia_valid) {
-		setattr_copy(inode, attr);
+		setattr_copy(&init_user_ns, inode, attr);
 		inode_inc_iversion(inode);
 		err = btrfs_dirty_inode(inode);
 
fs/cachefiles/interface.c +2 -2
index 4cea5fbf695e..5efa6a3702c0 100644
@@ -470,14 +470,14 @@ static int cachefiles_attr_changed(struct fscache_object *_object)
 		_debug("discard tail %llx", oi_size);
 		newattrs.ia_valid = ATTR_SIZE;
 		newattrs.ia_size = oi_size & PAGE_MASK;
-		ret = notify_change(object->backer, &newattrs, NULL);
+		ret = notify_change(&init_user_ns, object->backer, &newattrs, NULL);
 		if (ret < 0)
 			goto truncate_failed;
 	}
 
 	newattrs.ia_valid = ATTR_SIZE;
 	newattrs.ia_size = ni_size;
-	ret = notify_change(object->backer, &newattrs, NULL);
+	ret = notify_change(&init_user_ns, object->backer, &newattrs, NULL);
 
 truncate_failed:
 	inode_unlock(d_inode(object->backer));
fs/ceph/inode.c +1 -1
index e8a15ee09bc1..285d3baca27e 100644
@@ -2247,7 +2247,7 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr)
 	if (ceph_snap(inode) != CEPH_NOSNAP)
 		return -EROFS;
 
-	err = setattr_prepare(dentry, attr);
+	err = setattr_prepare(&init_user_ns, dentry, attr);
 	if (err != 0)
 		return err;
 
fs/cifs/inode.c +4 -4
index a83b3a8ffaac..27554f71f744 100644
@@ -2610,7 +2610,7 @@ cifs_setattr_unix(struct dentry *direntry, struct iattr *attrs)
 	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_PERM)
 		attrs->ia_valid |= ATTR_FORCE;
 
-	rc = setattr_prepare(direntry, attrs);
+	rc = setattr_prepare(&init_user_ns, direntry, attrs);
 	if (rc < 0)
 		goto out;
 
@@ -2715,7 +2715,7 @@ cifs_setattr_unix(struct dentry *direntry, struct iattr *attrs)
 	    attrs->ia_size != i_size_read(inode))
 		truncate_setsize(inode, attrs->ia_size);
 
-	setattr_copy(inode, attrs);
+	setattr_copy(&init_user_ns, inode, attrs);
 	mark_inode_dirty(inode);
 
 	/* force revalidate when any of these times are set since some
@@ -2757,7 +2757,7 @@ cifs_setattr_nounix(struct dentry *direntry, struct iattr *attrs)
 	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_PERM)
 		attrs->ia_valid |= ATTR_FORCE;
 
-	rc = setattr_prepare(direntry, attrs);
+	rc = setattr_prepare(&init_user_ns, direntry, attrs);
 	if (rc < 0) {
 		free_xid(xid);
 		return rc;
@@ -2913,7 +2913,7 @@ cifs_setattr_nounix(struct dentry *direntry, struct iattr *attrs)
 	    attrs->ia_size != i_size_read(inode))
 		truncate_setsize(inode, attrs->ia_size);
 
-	setattr_copy(inode, attrs);
+	setattr_copy(&init_user_ns, inode, attrs);
 	mark_inode_dirty(inode);
 
 cifs_setattr_exit:
fs/ecryptfs/inode.c +3 -3
index 9b1ae410983c..2454aef02a68 100644
@@ -855,7 +855,7 @@ int ecryptfs_truncate(struct dentry *dentry, loff_t new_length)
 		struct dentry *lower_dentry = ecryptfs_dentry_to_lower(dentry);
 
 		inode_lock(d_inode(lower_dentry));
-		rc = notify_change(lower_dentry, &lower_ia, NULL);
+		rc = notify_change(&init_user_ns, lower_dentry, &lower_ia, NULL);
 		inode_unlock(d_inode(lower_dentry));
 	}
 	return rc;
@@ -933,7 +933,7 @@ static int ecryptfs_setattr(struct dentry *dentry, struct iattr *ia)
 	}
 	mutex_unlock(&crypt_stat->cs_mutex);
 
-	rc = setattr_prepare(dentry, ia);
+	rc = setattr_prepare(&init_user_ns, dentry, ia);
 	if (rc)
 		goto out;
 	if (ia->ia_valid & ATTR_SIZE) {
@@ -959,7 +959,7 @@ static int ecryptfs_setattr(struct dentry *dentry, struct iattr *ia)
 		lower_ia.ia_valid &= ~ATTR_MODE;
 
 	inode_lock(d_inode(lower_dentry));
-	rc = notify_change(lower_dentry, &lower_ia, NULL);
+	rc = notify_change(&init_user_ns, lower_dentry, &lower_ia, NULL);
 	inode_unlock(d_inode(lower_dentry));
 out:
 	fsstack_copy_attr_all(inode, lower_inode);
fs/exfat/file.c +2 -2
index a92478eabfa4..ace35aa8e64b 100644
@@ -305,7 +305,7 @@ int exfat_setattr(struct dentry *dentry, struct iattr *attr)
 				ATTR_TIMES_SET);
 	}
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	attr->ia_valid = ia_valid;
 	if (error)
 		goto out;
@@ -340,7 +340,7 @@ int exfat_setattr(struct dentry *dentry, struct iattr *attr)
 		up_write(&EXFAT_I(inode)->truncate_lock);
 	}
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	exfat_truncate_atime(&inode->i_atime);
 	mark_inode_dirty(inode);
 
fs/ext2/inode.c +2 -2
index 78c417d3c898..06c0cf28c1a0 100644
@@ -1669,7 +1669,7 @@ int ext2_setattr(struct dentry *dentry, struct iattr *iattr)
 	struct inode *inode = d_inode(dentry);
 	int error;
 
-	error = setattr_prepare(dentry, iattr);
+	error = setattr_prepare(&init_user_ns, dentry, iattr);
 	if (error)
 		return error;
 
@@ -1689,7 +1689,7 @@ int ext2_setattr(struct dentry *dentry, struct iattr *iattr)
 		if (error)
 			return error;
 	}
-	setattr_copy(inode, iattr);
+	setattr_copy(&init_user_ns, inode, iattr);
 	if (iattr->ia_valid & ATTR_MODE)
 		error = posix_acl_chmod(inode, inode->i_mode);
 	mark_inode_dirty(inode);
fs/ext4/inode.c +2 -2
index 27946882d4ce..644510b5e06c 100644
@@ -5333,7 +5333,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 				  ATTR_GID | ATTR_TIMES_SET))))
 		return -EPERM;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
@@ -5508,7 +5508,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (!error) {
-		setattr_copy(inode, attr);
+		setattr_copy(&init_user_ns, inode, attr);
 		mark_inode_dirty(inode);
 	}
 
fs/f2fs/file.c +6 -4
index 5fc0ff28b5dd..90d7b89176de 100644
@@ -831,7 +831,8 @@ int f2fs_getattr(const struct path *path, struct kstat *stat,
 }
 
 #ifdef CONFIG_F2FS_FS_POSIX_ACL
-static void __setattr_copy(struct inode *inode, const struct iattr *attr)
+static void __setattr_copy(struct user_namespace *mnt_userns, struct inode *inode,
+			   const struct iattr *attr)
 {
 	unsigned int ia_valid = attr->ia_valid;
 
@@ -847,8 +848,9 @@ static void __setattr_copy(struct inode *inode, const struct iattr *attr)
 		inode->i_ctime = attr->ia_ctime;
 	if (ia_valid & ATTR_MODE) {
 		umode_t mode = attr->ia_mode;
+		kgid_t kgid = i_gid_into_mnt(mnt_userns, inode);
 
-		if (!in_group_p(inode->i_gid) && !capable(CAP_FSETID))
+		if (!in_group_p(kgid) && !capable(CAP_FSETID))
 			mode &= ~S_ISGID;
 		set_acl_inode(inode, mode);
 	}
@@ -869,7 +871,7 @@ int f2fs_setattr(struct dentry *dentry, struct iattr *attr)
 		!f2fs_is_compress_backend_ready(inode))
 		return -EOPNOTSUPP;
 
-	err = setattr_prepare(dentry, attr);
+	err = setattr_prepare(&init_user_ns, dentry, attr);
 	if (err)
 		return err;
 
@@ -945,7 +947,7 @@ int f2fs_setattr(struct dentry *dentry, struct iattr *attr)
 		spin_unlock(&F2FS_I(inode)->i_size_lock);
 	}
 
-	__setattr_copy(inode, attr);
+	__setattr_copy(&init_user_ns, inode, attr);
 
 	if (attr->ia_valid & ATTR_MODE) {
 		err = posix_acl_chmod(inode, f2fs_get_inode_mode(inode));
fs/fat/file.c +2 -2
index f9ee27cf4d7c..805b501467e9 100644
@@ -480,7 +480,7 @@ int fat_setattr(struct dentry *dentry, struct iattr *attr)
 			attr->ia_valid &= ~TIMES_SET_FLAGS;
 	}
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	attr->ia_valid = ia_valid;
 	if (error) {
 		if (sbi->options.quiet)
@@ -550,7 +550,7 @@ int fat_setattr(struct dentry *dentry, struct iattr *attr)
 		fat_truncate_time(inode, &attr->ia_mtime, S_MTIME);
 	attr->ia_valid &= ~(ATTR_ATIME|ATTR_CTIME|ATTR_MTIME);
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 out:
 	return error;
fs/fuse/dir.c +1 -1
index 19a46f445d4c..eb648ed19312 100644
@@ -1610,7 +1610,7 @@ int fuse_do_setattr(struct dentry *dentry, struct iattr *attr,
 	if (!fc->default_permissions)
 		attr->ia_valid |= ATTR_FORCE;
 
-	err = setattr_prepare(dentry, attr);
+	err = setattr_prepare(&init_user_ns, dentry, attr);
 	if (err)
 		return err;
 
fs/gfs2/inode.c +2 -2
index 5b2ff0c74b67..59c25181d108 100644
@@ -1861,7 +1861,7 @@ int gfs2_permission(struct inode *inode, int mask)
 
 static int __gfs2_setattr_simple(struct inode *inode, struct iattr *attr)
 {
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
@@ -1982,7 +1982,7 @@ static int gfs2_setattr(struct dentry *dentry, struct iattr *attr)
 	if (IS_IMMUTABLE(inode) || IS_APPEND(inode))
 		goto error;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		goto error;
 
fs/hfs/inode.c +2 -2
index f35a37c65e5f..c646218b72bf 100644
@@ -608,7 +608,7 @@ int hfs_inode_setattr(struct dentry *dentry, struct iattr * attr)
 	struct hfs_sb_info *hsb = HFS_SB(inode->i_sb);
 	int error;
 
-	error = setattr_prepare(dentry, attr); /* basic permission checks */
+	error = setattr_prepare(&init_user_ns, dentry, attr); /* basic permission checks */
 	if (error)
 		return error;
 
@@ -647,7 +647,7 @@ int hfs_inode_setattr(struct dentry *dentry, struct iattr * attr)
 						  current_time(inode);
 	}
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
fs/hfsplus/inode.c +2 -2
index 21357046b039..ffa137f8234e 100644
@@ -246,7 +246,7 @@ static int hfsplus_setattr(struct dentry *dentry, struct iattr *attr)
 	struct inode *inode = d_inode(dentry);
 	int error;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
@@ -264,7 +264,7 @@ static int hfsplus_setattr(struct dentry *dentry, struct iattr *attr)
 		inode->i_mtime = inode->i_ctime = current_time(inode);
 	}
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 
 	return 0;
fs/hostfs/hostfs_kern.c +2 -2
index b841a05a2b8c..6970e29a5287 100644
@@ -792,7 +792,7 @@ static int hostfs_setattr(struct dentry *dentry, struct iattr *attr)
 
 	int fd = HOSTFS_I(inode)->fd;
 
-	err = setattr_prepare(dentry, attr);
+	err = setattr_prepare(&init_user_ns, dentry, attr);
 	if (err)
 		return err;
 
@@ -849,7 +849,7 @@ static int hostfs_setattr(struct dentry *dentry, struct iattr *attr)
 	    attr->ia_size != i_size_read(inode))
 		truncate_setsize(inode, attr->ia_size);
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
fs/hpfs/inode.c +2 -2
index eb8b4baf0f2e..8ba2152a78ba 100644
@@ -274,7 +274,7 @@ int hpfs_setattr(struct dentry *dentry, struct iattr *attr)
 	if ((attr->ia_valid & ATTR_SIZE) && attr->ia_size > inode->i_size)
 		goto out_unlock;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		goto out_unlock;
 
@@ -288,7 +288,7 @@ int hpfs_setattr(struct dentry *dentry, struct iattr *attr)
 		hpfs_truncate(inode);
 	}
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 
 	hpfs_write_inode(inode);
 
fs/hugetlbfs/inode.c +2 -2
index 6737929e443c..327e572b4e00 100644
@@ -761,7 +761,7 @@ static int hugetlbfs_setattr(struct dentry *dentry, struct iattr *attr)
 
 	BUG_ON(!inode);
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
@@ -780,7 +780,7 @@ static int hugetlbfs_setattr(struct dentry *dentry, struct iattr *attr)
 			return error;
 	}
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
fs/inode.c +1 -1
index a9ac97a27784..49b512592dcd 100644
@@ -1912,7 +1912,7 @@ static int __remove_privs(struct dentry *dentry, int kill)
 	 * Note we call this on write, so notify_change will not
 	 * encounter any conflicting delegations:
 	 */
-	return notify_change(dentry, &newattrs, NULL);
+	return notify_change(&init_user_ns, dentry, &newattrs, NULL);
 }
 
 /*
fs/jffs2/fs.c +1 -1
index 78858f6e9583..67993808f4da 100644
@@ -195,7 +195,7 @@ int jffs2_setattr(struct dentry *dentry, struct iattr *iattr)
 	struct inode *inode = d_inode(dentry);
 	int rc;
 
-	rc = setattr_prepare(dentry, iattr);
+	rc = setattr_prepare(&init_user_ns, dentry, iattr);
 	if (rc)
 		return rc;
 
fs/jfs/file.c +2 -2
index 930d2701f206..ff49876e9c9b 100644
@@ -90,7 +90,7 @@ int jfs_setattr(struct dentry *dentry, struct iattr *iattr)
 	struct inode *inode = d_inode(dentry);
 	int rc;
 
-	rc = setattr_prepare(dentry, iattr);
+	rc = setattr_prepare(&init_user_ns, dentry, iattr);
 	if (rc)
 		return rc;
 
@@ -118,7 +118,7 @@ int jfs_setattr(struct dentry *dentry, struct iattr *iattr)
 		jfs_truncate(inode);
 	}
 
-	setattr_copy(inode, iattr);
+	setattr_copy(&init_user_ns, inode, iattr);
 	mark_inode_dirty(inode);
 
 	if (iattr->ia_valid & ATTR_MODE)
fs/kernfs/inode.c +2 -2
index ff5598cc1de0..86bd4c593b78 100644
@@ -122,7 +122,7 @@ int kernfs_iop_setattr(struct dentry *dentry, struct iattr *iattr)
 		return -EINVAL;
 
 	mutex_lock(&kernfs_mutex);
-	error = setattr_prepare(dentry, iattr);
+	error = setattr_prepare(&init_user_ns, dentry, iattr);
 	if (error)
 		goto out;
 
@@ -131,7 +131,7 @@ int kernfs_iop_setattr(struct dentry *dentry, struct iattr *iattr)
 		goto out;
 
 	/* this ignores size changes */
-	setattr_copy(inode, iattr);
+	setattr_copy(&init_user_ns, inode, iattr);
 
 out:
 	mutex_unlock(&kernfs_mutex);
fs/libfs.c +2 -2
index f8b3c02b4f0f..a73fe109403c 100644
@@ -497,13 +497,13 @@ int simple_setattr(struct dentry *dentry, struct iattr *iattr)
 	struct inode *inode = d_inode(dentry);
 	int error;
 
-	error = setattr_prepare(dentry, iattr);
+	error = setattr_prepare(&init_user_ns, dentry, iattr);
 	if (error)
 		return error;
 
 	if (iattr->ia_valid & ATTR_SIZE)
 		truncate_setsize(inode, iattr->ia_size);
-	setattr_copy(inode, iattr);
+	setattr_copy(&init_user_ns, inode, iattr);
 	mark_inode_dirty(inode);
 	return 0;
 }
fs/minix/file.c +2 -2
index c50b0a20fcd9..f07acd268577 100644
@@ -27,7 +27,7 @@ static int minix_setattr(struct dentry *dentry, struct iattr *attr)
 	struct inode *inode = d_inode(dentry);
 	int error;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
@@ -41,7 +41,7 @@ static int minix_setattr(struct dentry *dentry, struct iattr *attr)
 		minix_truncate(inode);
 	}
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
fs/nfsd/nfsproc.c +1 -1
index 9473d048efec..0ea0554d20d1 100644
@@ -90,7 +90,7 @@ nfsd_proc_setattr(struct svc_rqst *rqstp)
 		if (delta < 0)
 			delta = -delta;
 		if (delta < MAX_TOUCH_TIME_ERROR &&
-		    setattr_prepare(fhp->fh_dentry, iap) != 0) {
+		    setattr_prepare(&init_user_ns, fhp->fh_dentry, iap) != 0) {
 			/*
 			 * Turn off ATTR_[AM]TIME_SET but leave ATTR_[AM]TIME.
 			 * This will cause notify_change to set these times
fs/nfsd/vfs.c +2 -2
index b3ccb2997ea0..5db88d703b63 100644
@@ -448,7 +448,7 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap,
 			.ia_size	= iap->ia_size,
 		};
 
-		host_err = notify_change(dentry, &size_attr, NULL);
+		host_err = notify_change(&init_user_ns, dentry, &size_attr, NULL);
 		if (host_err)
 			goto out_unlock;
 		iap->ia_valid &= ~ATTR_SIZE;
@@ -463,7 +463,7 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap,
 	}
 
 	iap->ia_valid |= ATTR_CTIME;
-	host_err = notify_change(dentry, iap, NULL);
+	host_err = notify_change(&init_user_ns, dentry, iap, NULL);
 
 out_unlock:
 	fh_unlock(fhp);
fs/nilfs2/inode.c +2 -2
index 11225a659736..8aad3c48092a 100644
@@ -812,7 +812,7 @@ int nilfs_setattr(struct dentry *dentry, struct iattr *iattr)
 	struct super_block *sb = inode->i_sb;
 	int err;
 
-	err = setattr_prepare(dentry, iattr);
+	err = setattr_prepare(&init_user_ns, dentry, iattr);
 	if (err)
 		return err;
 
@@ -827,7 +827,7 @@ int nilfs_setattr(struct dentry *dentry, struct iattr *iattr)
 		nilfs_truncate(inode);
 	}
 
-	setattr_copy(inode, iattr);
+	setattr_copy(&init_user_ns, inode, iattr);
 	mark_inode_dirty(inode);
 
 	if (iattr->ia_valid & ATTR_MODE) {
fs/ntfs/inode.c +1 -1
index f7e4cbc26eaf..38f4cf1d4497 100644
@@ -2866,7 +2866,7 @@ int ntfs_setattr(struct dentry *dentry, struct iattr *attr)
 	int err;
 	unsigned int ia_valid = attr->ia_valid;
 
-	err = setattr_prepare(dentry, attr);
+	err = setattr_prepare(&init_user_ns, dentry, attr);
 	if (err)
 		goto out;
 	/* We do not support NTFS ACLs yet. */
fs/ocfs2/dlmfs/dlmfs.c +2 -2
index 37c7d03a6284..9fa66cd1f622 100644
@@ -196,11 +196,11 @@ static int dlmfs_file_setattr(struct dentry *dentry, struct iattr *attr)
 	struct inode *inode = d_inode(dentry);
 
 	attr->ia_valid &= ~ATTR_SIZE;
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
fs/ocfs2/file.c +2 -2
index 0c75619adf54..cabf355b148f 100644
@@ -1142,7 +1142,7 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr)
 	if (!(attr->ia_valid & OCFS2_VALID_ATTRS))
 		return 0;
 
-	status = setattr_prepare(dentry, attr);
+	status = setattr_prepare(&init_user_ns, dentry, attr);
 	if (status)
 		return status;
 
@@ -1263,7 +1263,7 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr)
 		}
 	}
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 
 	status = ocfs2_mark_inode_dirty(handle, inode, bh);
fs/omfs/file.c +2 -2
index 2c7b70ee1388..729339cd7902 100644
@@ -348,7 +348,7 @@ static int omfs_setattr(struct dentry *dentry, struct iattr *attr)
 	struct inode *inode = d_inode(dentry);
 	int error;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
@@ -361,7 +361,7 @@ static int omfs_setattr(struct dentry *dentry, struct iattr *attr)
 		omfs_truncate(inode);
 	}
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
fs/open.c +3 -3
index 4089ed2c51ec..bae00ee63af8 100644
@@ -61,7 +61,7 @@ int do_truncate(struct dentry *dentry, loff_t length, unsigned int time_attrs,
 
 	inode_lock(dentry->d_inode);
 	/* Note any delegations or leases have already been broken: */
-	ret = notify_change(dentry, &newattrs, NULL);
+	ret = notify_change(&init_user_ns, dentry, &newattrs, NULL);
 	inode_unlock(dentry->d_inode);
 	return ret;
 }
@@ -580,7 +580,7 @@ int chmod_common(const struct path *path, umode_t mode)
 		goto out_unlock;
 	newattrs.ia_mode = (mode & S_IALLUGO) | (inode->i_mode & ~S_IALLUGO);
 	newattrs.ia_valid = ATTR_MODE | ATTR_CTIME;
-	error = notify_change(path->dentry, &newattrs, &delegated_inode);
+	error = notify_change(&init_user_ns, path->dentry, &newattrs, &delegated_inode);
 out_unlock:
 	inode_unlock(inode);
 	if (delegated_inode) {
@@ -671,7 +671,7 @@ int chown_common(const struct path *path, uid_t user, gid_t group)
 	inode_lock(inode);
 	error = security_path_chown(path, uid, gid);
 	if (!error)
-		error = notify_change(path->dentry, &newattrs, &delegated_inode);
+		error = notify_change(&init_user_ns, path->dentry, &newattrs, &delegated_inode);
 	inode_unlock(inode);
 	if (delegated_inode) {
 		error = break_deleg_wait(&delegated_inode);
fs/orangefs/inode.c +2 -2
index 4c790cc8042d..8ac9491ceb9a 100644
@@ -855,7 +855,7 @@ int __orangefs_setattr(struct inode *inode, struct iattr *iattr)
 		ORANGEFS_I(inode)->attr_uid = current_fsuid();
 		ORANGEFS_I(inode)->attr_gid = current_fsgid();
 	}
-	setattr_copy(inode, iattr);
+	setattr_copy(&init_user_ns, inode, iattr);
 	spin_unlock(&inode->i_lock);
 	mark_inode_dirty(inode);
 
@@ -876,7 +876,7 @@ int orangefs_setattr(struct dentry *dentry, struct iattr *iattr)
 	int ret;
 	gossip_debug(GOSSIP_INODE_DEBUG, "__orangefs_setattr: called on %pd\n",
 	    dentry);
-	ret = setattr_prepare(dentry, iattr);
+	ret = setattr_prepare(&init_user_ns, dentry, iattr);
 	if (ret)
 	        goto out;
 	ret = __orangefs_setattr(d_inode(dentry), iattr);
fs/overlayfs/copy_up.c +4 -4
index e5b616c93e11..3e9957ae19fa 100644
@@ -235,7 +235,7 @@ static int ovl_set_size(struct dentry *upperdentry, struct kstat *stat)
 		.ia_size = stat->size,
 	};
 
-	return notify_change(upperdentry, &attr, NULL);
+	return notify_change(&init_user_ns, upperdentry, &attr, NULL);
 }
 
 static int ovl_set_timestamps(struct dentry *upperdentry, struct kstat *stat)
@@ -247,7 +247,7 @@ static int ovl_set_timestamps(struct dentry *upperdentry, struct kstat *stat)
 		.ia_mtime = stat->mtime,
 	};
 
-	return notify_change(upperdentry, &attr, NULL);
+	return notify_change(&init_user_ns, upperdentry, &attr, NULL);
 }
 
 int ovl_set_attr(struct dentry *upperdentry, struct kstat *stat)
@@ -259,7 +259,7 @@ int ovl_set_attr(struct dentry *upperdentry, struct kstat *stat)
 			.ia_valid = ATTR_MODE,
 			.ia_mode = stat->mode,
 		};
-		err = notify_change(upperdentry, &attr, NULL);
+		err = notify_change(&init_user_ns, upperdentry, &attr, NULL);
 	}
 	if (!err) {
 		struct iattr attr = {
@@ -267,7 +267,7 @@ int ovl_set_attr(struct dentry *upperdentry, struct kstat *stat)
 			.ia_uid = stat->uid,
 			.ia_gid = stat->gid,
 		};
-		err = notify_change(upperdentry, &attr, NULL);
+		err = notify_change(&init_user_ns, upperdentry, &attr, NULL);
 	}
 	if (!err)
 		ovl_set_timestamps(upperdentry, stat);
fs/overlayfs/dir.c +1 -1
index 98a23353b19a..29840820a46c 100644
@@ -508,7 +508,7 @@ static int ovl_create_over_whiteout(struct dentry *dentry, struct inode *inode,
 			.ia_mode = cattr->mode,
 		};
 		inode_lock(newdentry->d_inode);
-		err = notify_change(newdentry, &attr, NULL);
+		err = notify_change(&init_user_ns, newdentry, &attr, NULL);
 		inode_unlock(newdentry->d_inode);
 		if (err)
 			goto out_cleanup;
fs/overlayfs/inode.c +2 -2
index c101ebbb7a77..5aa66881dbd7 100644
@@ -21,7 +21,7 @@ int ovl_setattr(struct dentry *dentry, struct iattr *attr)
 	struct dentry *upperdentry;
 	const struct cred *old_cred;
 
-	err = setattr_prepare(dentry, attr);
+	err = setattr_prepare(&init_user_ns, dentry, attr);
 	if (err)
 		return err;
 
@@ -79,7 +79,7 @@ int ovl_setattr(struct dentry *dentry, struct iattr *attr)
 
 		inode_lock(upperdentry->d_inode);
 		old_cred = ovl_override_creds(dentry->d_sb);
-		err = notify_change(upperdentry, attr, NULL);
+		err = notify_change(&init_user_ns, upperdentry, attr, NULL);
 		revert_creds(old_cred);
 		if (!err)
 			ovl_copyattr(upperdentry->d_inode, dentry->d_inode);
fs/overlayfs/super.c +1 -1
index 3e925deaa19a..39b2e9aa0e5b 100644
@@ -804,7 +804,7 @@ static struct dentry *ovl_workdir_create(struct ovl_fs *ofs,
 
 		/* Clear any inherited mode bits */
 		inode_lock(work->d_inode);
-		err = notify_change(work, &attr, NULL);
+		err = notify_change(&init_user_ns, work, &attr, NULL);
 		inode_unlock(work->d_inode);
 		if (err)
 			goto out_dput;
fs/proc/base.c +2 -2
index b4ec9293625e..bb4e63a3684f 100644
@@ -693,11 +693,11 @@ int proc_setattr(struct dentry *dentry, struct iattr *attr)
 	if (attr->ia_valid & ATTR_MODE)
 		return -EPERM;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
fs/proc/generic.c +2 -2
index 6c0a05f55d6b..6d4fabab8aa7 100644
@@ -121,11 +121,11 @@ static int proc_notify_change(struct dentry *dentry, struct iattr *iattr)
 	struct proc_dir_entry *de = PDE(inode);
 	int error;
 
-	error = setattr_prepare(dentry, iattr);
+	error = setattr_prepare(&init_user_ns, dentry, iattr);
 	if (error)
 		return error;
 
-	setattr_copy(inode, iattr);
+	setattr_copy(&init_user_ns, inode, iattr);
 	mark_inode_dirty(inode);
 
 	proc_set_user(de, inode->i_uid, inode->i_gid);
fs/proc/proc_sysctl.c +2 -2
index 317899222d7f..ec67dbc1f705 100644
@@ -821,11 +821,11 @@ static int proc_sys_setattr(struct dentry *dentry, struct iattr *attr)
 	if (attr->ia_valid & (ATTR_MODE | ATTR_UID | ATTR_GID))
 		return -EPERM;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
fs/ramfs/file-nommu.c +2 -2
index 355523f4a4bf..f0358fe410d3 100644
@@ -165,7 +165,7 @@ static int ramfs_nommu_setattr(struct dentry *dentry, struct iattr *ia)
 	int ret = 0;
 
 	/* POSIX UID/GID verification for setting inode attributes */
-	ret = setattr_prepare(dentry, ia);
+	ret = setattr_prepare(&init_user_ns, dentry, ia);
 	if (ret)
 		return ret;
 
@@ -185,7 +185,7 @@ static int ramfs_nommu_setattr(struct dentry *dentry, struct iattr *ia)
 		}
 	}
 
-	setattr_copy(inode, ia);
+	setattr_copy(&init_user_ns, inode, ia);
  out:
 	ia->ia_valid = old_ia_valid;
 	return ret;
fs/reiserfs/inode.c +2 -2
index c76d563dec0e..944f2b487cf8 100644
@@ -3288,7 +3288,7 @@ int reiserfs_setattr(struct dentry *dentry, struct iattr *attr)
 	unsigned int ia_valid;
 	int error;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
@@ -3413,7 +3413,7 @@ int reiserfs_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (!error) {
-		setattr_copy(inode, attr);
+		setattr_copy(&init_user_ns, inode, attr);
 		mark_inode_dirty(inode);
 	}
 
fs/sysv/file.c +2 -2
index 45fc79a18594..ca7e216b7b9e 100644
@@ -34,7 +34,7 @@ static int sysv_setattr(struct dentry *dentry, struct iattr *attr)
 	struct inode *inode = d_inode(dentry);
 	int error;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
@@ -47,7 +47,7 @@ static int sysv_setattr(struct dentry *dentry, struct iattr *attr)
 		sysv_truncate(inode);
 	}
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
fs/ubifs/file.c +1 -1
index 2bc7780d2963..76ef392b1e41 100644
@@ -1265,7 +1265,7 @@ int ubifs_setattr(struct dentry *dentry, struct iattr *attr)
 
 	dbg_gen("ino %lu, mode %#x, ia_valid %#x",
 		inode->i_ino, inode->i_mode, attr->ia_valid);
-	err = setattr_prepare(dentry, attr);
+	err = setattr_prepare(&init_user_ns, dentry, attr);
 	if (err)
 		return err;
 
fs/udf/file.c +2 -2
index 928283925d68..0d5ade491b1a 100644
@@ -259,7 +259,7 @@ static int udf_setattr(struct dentry *dentry, struct iattr *attr)
 	struct super_block *sb = inode->i_sb;
 	int error;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
@@ -282,7 +282,7 @@ static int udf_setattr(struct dentry *dentry, struct iattr *attr)
 	if (attr->ia_valid & ATTR_MODE)
 		udf_update_extra_perms(inode, attr->ia_mode);
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
fs/ufs/inode.c +2 -2
index c843ec858cf7..6b51f3b20143 100644
@@ -1217,7 +1217,7 @@ int ufs_setattr(struct dentry *dentry, struct iattr *attr)
 	unsigned int ia_valid = attr->ia_valid;
 	int error;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
@@ -1227,7 +1227,7 @@ int ufs_setattr(struct dentry *dentry, struct iattr *attr)
 			return error;
 	}
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	mark_inode_dirty(inode);
 	return 0;
 }
fs/utimes.c +1 -1
index fd3cc4226224..1a4130bee157 100644
@@ -62,7 +62,7 @@ int vfs_utimes(const struct path *path, struct timespec64 *times)
 	}
 retry_deleg:
 	inode_lock(inode);
-	error = notify_change(path->dentry, &newattrs, &delegated_inode);
+	error = notify_change(&init_user_ns, path->dentry, &newattrs, &delegated_inode);
 	inode_unlock(inode);
 	if (delegated_inode) {
 		error = break_deleg_wait(&delegated_inode);
fs/xfs/xfs_iops.c +1 -1
index 67c8dc9de8aa..08a478d25122 100644
@@ -637,7 +637,7 @@ xfs_vn_change_ok(
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return -EIO;
 
-	return setattr_prepare(dentry, iattr);
+	return setattr_prepare(&init_user_ns, dentry, iattr);
 }
 
 /*
fs/zonefs/super.c +2 -2
index 569525ee8f69..8a1f69677784 100644
@@ -488,7 +488,7 @@ static int zonefs_inode_setattr(struct dentry *dentry, struct iattr *iattr)
 	if (unlikely(IS_IMMUTABLE(inode)))
 		return -EPERM;
 
-	ret = setattr_prepare(dentry, iattr);
+	ret = setattr_prepare(&init_user_ns, dentry, iattr);
 	if (ret)
 		return ret;
 
@@ -516,7 +516,7 @@ static int zonefs_inode_setattr(struct dentry *dentry, struct iattr *iattr)
 			return ret;
 	}
 
-	setattr_copy(inode, iattr);
+	setattr_copy(&init_user_ns, inode, iattr);
 
 	return 0;
 }
include/linux/fs.h +5 -3
index 430434317f9e..ef993857682b 100644
@@ -2805,7 +2805,8 @@ static inline int bmap(struct inode *inode,  sector_t *block)
 }
 #endif
 
-extern int notify_change(struct dentry *, struct iattr *, struct inode **);
+extern int notify_change(struct user_namespace *, struct dentry *,
+			 struct iattr *, struct inode **);
 extern int inode_permission(struct user_namespace *, struct inode *, int);
 extern int generic_permission(struct user_namespace *, struct inode *, int);
 extern int __check_sticky(struct inode *dir, struct inode *inode);
@@ -3260,9 +3261,10 @@ extern int buffer_migrate_page_norefs(struct address_space *,
 #define buffer_migrate_page_norefs NULL
 #endif
 
-extern int setattr_prepare(struct dentry *, struct iattr *);
+extern int setattr_prepare(struct user_namespace *, struct dentry *, struct iattr *);
 extern int inode_newsize_ok(const struct inode *, loff_t offset);
-extern void setattr_copy(struct inode *inode, const struct iattr *attr);
+extern void setattr_copy(struct user_namespace *, struct inode *inode,
+			 const struct iattr *attr);
 
 extern int file_update_time(struct file *file);
 
mm/shmem.c +2 -2
index 1c68c9edba5e..1cb451e131ec 100644
@@ -1087,7 +1087,7 @@ static int shmem_setattr(struct dentry *dentry, struct iattr *attr)
 	struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb);
 	int error;
 
-	error = setattr_prepare(dentry, attr);
+	error = setattr_prepare(&init_user_ns, dentry, attr);
 	if (error)
 		return error;
 
@@ -1141,7 +1141,7 @@ static int shmem_setattr(struct dentry *dentry, struct iattr *attr)
 		}
 	}
 
-	setattr_copy(inode, attr);
+	setattr_copy(&init_user_ns, inode, attr);
 	if (attr->ia_valid & ATTR_MODE)
 		error = posix_acl_chmod(inode, inode->i_mode);
 	return error;
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
#19 lore raw

PATCH v5 14/42 acl: handle idmapped mounts

In reply to Christian Brauner

The posix acl permission checking helpers determine whether a caller is privileged over an inode according to the acls associated with the inode. Add helpers that make it possible to handle acls on idmapped mounts.

The vfs and the filesystems targeted by this first iteration make use of posix_acl_fix_xattr_from_user() and posix_acl_fix_xattr_to_user() to translate basic posix access and default permissions such as the ACL_USER and ACL_GROUP type according to the initial user namespace (or the superblock's user namespace) to and from the caller's current user namespace. Adapt these two helpers to handle idmapped mounts whereby we either shift from or into the mount's user namespace depending on in which direction we're translating.
Similarly, cap_convert_nscap() is used by the vfs to translate user namespace and non-user namespace aware filesystem capabilities from the superblock's user namespace to the caller's user namespace. Enable it to handle idmapped mounts by accounting for the mount's user namespace.

In addition the fileystems targeted in the first iteration of this patch series make use of the posix_acl_chmod() and, posix_acl_update_mode() helpers. Both helpers perform permission checks on the target inode. Let them handle idmapped mounts. These two helpers are called when posix acls are set by the respective filesystems to handle this case we extend the ->set() method to take an additional user namespace argument to pass the mount's user namespace down.

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Don't pollute the vfs with additional helpers simply extend the existing helpers with an additional argument and switch all callers.

/* v3 */
unchanged

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 Documentation/filesystems/locking.rst |  6 +--
 Documentation/filesystems/porting.rst |  2 +
 fs/9p/acl.c                           |  4 +-
 fs/9p/xattr.c                         |  1 +
 fs/afs/xattr.c                        |  2 +
 fs/btrfs/acl.c                        |  2 +-
 fs/btrfs/inode.c                      |  2 +-
 fs/btrfs/xattr.c                      |  2 +
 fs/ceph/acl.c                         |  2 +-
 fs/ceph/inode.c                       |  2 +-
 fs/ceph/xattr.c                       |  1 +
 fs/cifs/xattr.c                       |  1 +
 fs/ecryptfs/inode.c                   |  1 +
 fs/ext2/acl.c                         |  2 +-
 fs/ext2/inode.c                       |  2 +-
 fs/ext2/xattr_security.c              |  1 +
 fs/ext2/xattr_trusted.c               |  1 +
 fs/ext2/xattr_user.c                  |  1 +
 fs/ext4/acl.c                         |  2 +-
 fs/ext4/inode.c                       |  2 +-
 fs/ext4/xattr_hurd.c                  |  1 +
 fs/ext4/xattr_security.c              |  1 +
 fs/ext4/xattr_trusted.c               |  1 +
 fs/ext4/xattr_user.c                  |  1 +
 fs/f2fs/acl.c                         |  2 +-
 fs/f2fs/file.c                        |  2 +-
 fs/f2fs/xattr.c                       |  2 +
 fs/fuse/xattr.c                       |  2 +
 fs/gfs2/acl.c                         |  2 +-
 fs/gfs2/inode.c                       |  2 +-
 fs/gfs2/xattr.c                       |  1 +
 fs/hfs/attr.c                         |  1 +
 fs/hfsplus/xattr.c                    |  1 +
 fs/hfsplus/xattr_security.c           |  1 +
 fs/hfsplus/xattr_trusted.c            |  1 +
 fs/hfsplus/xattr_user.c               |  1 +
 fs/jffs2/acl.c                        |  2 +-
 fs/jffs2/fs.c                         |  2 +-
 fs/jffs2/security.c                   |  1 +
 fs/jffs2/xattr_trusted.c              |  1 +
 fs/jffs2/xattr_user.c                 |  1 +
 fs/jfs/acl.c                          |  2 +-
 fs/jfs/file.c                         |  2 +-
 fs/jfs/xattr.c                        |  2 +
 fs/kernfs/inode.c                     |  2 +
 fs/nfs/nfs4proc.c                     |  3 ++
 fs/nfsd/nfs2acl.c                     |  4 +-
 fs/nfsd/nfs3acl.c                     |  4 +-
 fs/nfsd/nfs4acl.c                     |  4 +-
 fs/ocfs2/acl.c                        |  2 +-
 fs/ocfs2/xattr.c                      |  3 ++
 fs/orangefs/acl.c                     |  2 +-
 fs/orangefs/inode.c                   |  2 +-
 fs/orangefs/xattr.c                   |  1 +
 fs/overlayfs/super.c                  |  3 ++
 fs/posix_acl.c                        | 77 +++++++++++++++++++--------
 fs/reiserfs/xattr_acl.c               |  4 +-
 fs/reiserfs/xattr_security.c          |  3 +-
 fs/reiserfs/xattr_trusted.c           |  3 +-
 fs/reiserfs/xattr_user.c              |  3 +-
 fs/ubifs/xattr.c                      |  1 +
 fs/xattr.c                            | 10 ++--
 fs/xfs/xfs_acl.c                      |  2 +-
 fs/xfs/xfs_iops.c                     |  2 +-
 fs/xfs/xfs_xattr.c                    |  3 +-
 include/linux/capability.h            |  4 +-
 include/linux/posix_acl.h             |  9 ++--
 include/linux/posix_acl_xattr.h       | 12 +++--
 include/linux/xattr.h                 |  6 +--
 mm/shmem.c                            |  3 +-
 net/socket.c                          |  1 +
 security/commoncap.c                  | 47 +++++++++++++---
 72 files changed, 210 insertions(+), 84 deletions(-)
Documentation/filesystems/locking.rst +3 -3
index c0f2c7586531..daf132a038ad 100644
@@ -126,9 +126,9 @@ prototypes::
 	int (*get)(const struct xattr_handler *handler, struct dentry *dentry,
 		   struct inode *inode, const char *name, void *buffer,
 		   size_t size);
-	int (*set)(const struct xattr_handler *handler, struct dentry *dentry,
-		   struct inode *inode, const char *name, const void *buffer,
-		   size_t size, int flags);
+	int (*set)(const struct xattr_handler *handler, struct user_namespace *mnt_userns,
+                   struct dentry *dentry, struct inode *inode, const char *name,
+                   const void *buffer, size_t size, int flags);
 
 locking rules:
 	all may block
Documentation/filesystems/porting.rst +2 -0
index 867036aa90b8..de1dcec3b5b8 100644
@@ -717,6 +717,8 @@ be removed.  Switch while you still can; the old one won't stay.
 **mandatory**
 
 ->setxattr() and xattr_handler.set() get dentry and inode passed separately.
+The xattr_handler.set() gets passed the user namespace of the mount the inode
+is seen from so filesystems can idmap the i_uid and i_gid accordingly.
 dentry might be yet to be attached to inode, so do _not_ use its ->d_inode
 in the instances.  Rationale: !@#!@# security_d_instantiate() needs to be
 called before we attach dentry to inode and !@#!@##!@$!$#!@#$!@$!@$ smack
fs/9p/acl.c +3 -1
index d77b28e8d57a..1c14f18a6ec9 100644
@@ -239,6 +239,7 @@ static int v9fs_xattr_get_acl(const struct xattr_handler *handler,
 }
 
 static int v9fs_xattr_set_acl(const struct xattr_handler *handler,
+			      struct user_namespace *mnt_userns,
 			      struct dentry *dentry, struct inode *inode,
 			      const char *name, const void *value,
 			      size_t size, int flags)
@@ -279,7 +280,8 @@ static int v9fs_xattr_set_acl(const struct xattr_handler *handler,
 			struct iattr iattr = { 0 };
 			struct posix_acl *old_acl = acl;
 
-			retval = posix_acl_update_mode(inode, &iattr.ia_mode, &acl);
+			retval = posix_acl_update_mode(mnt_userns, inode,
+						       &iattr.ia_mode, &acl);
 			if (retval)
 				goto err_out;
 			if (!acl) {
fs/9p/xattr.c +1 -0
index 87217dd0433e..ee331845e2c7 100644
@@ -157,6 +157,7 @@ static int v9fs_xattr_handler_get(const struct xattr_handler *handler,
 }
 
 static int v9fs_xattr_handler_set(const struct xattr_handler *handler,
+				  struct user_namespace *mnt_userns,
 				  struct dentry *dentry, struct inode *inode,
 				  const char *name, const void *value,
 				  size_t size, int flags)
fs/afs/xattr.c +2 -0
index 95c573dcda11..c629caae5002 100644
@@ -120,6 +120,7 @@ static const struct afs_operation_ops afs_store_acl_operation = {
  * Set a file's AFS3 ACL.
  */
 static int afs_xattr_set_acl(const struct xattr_handler *handler,
+			     struct user_namespace *mnt_userns,
                              struct dentry *dentry,
                              struct inode *inode, const char *name,
                              const void *buffer, size_t size, int flags)
@@ -248,6 +249,7 @@ static const struct afs_operation_ops yfs_store_opaque_acl2_operation = {
  * Set a file's YFS ACL.
  */
 static int afs_xattr_set_yfs(const struct xattr_handler *handler,
+			     struct user_namespace *mnt_userns,
                              struct dentry *dentry,
                              struct inode *inode, const char *name,
                              const void *buffer, size_t size, int flags)
fs/btrfs/acl.c +1 -1
index a0af1b952c4d..b5a683e895c6 100644
@@ -113,7 +113,7 @@ int btrfs_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 	umode_t old_mode = inode->i_mode;
 
 	if (type == ACL_TYPE_ACCESS && acl) {
-		ret = posix_acl_update_mode(inode, &inode->i_mode, &acl);
+		ret = posix_acl_update_mode(&init_user_ns, inode, &inode->i_mode, &acl);
 		if (ret)
 			return ret;
 	}
fs/btrfs/inode.c +1 -1
index 19d432a453b1..97bbf80ee12d 100644
@@ -5070,7 +5070,7 @@ static int btrfs_setattr(struct dentry *dentry, struct iattr *attr)
 		err = btrfs_dirty_inode(inode);
 
 		if (!err && attr->ia_valid & ATTR_MODE)
-			err = posix_acl_chmod(inode, inode->i_mode);
+			err = posix_acl_chmod(&init_user_ns, inode, inode->i_mode);
 	}
 
 	return err;
fs/btrfs/xattr.c +2 -0
index af6246f36a9e..b025102e435f 100644
@@ -362,6 +362,7 @@ static int btrfs_xattr_handler_get(const struct xattr_handler *handler,
 }
 
 static int btrfs_xattr_handler_set(const struct xattr_handler *handler,
+				   struct user_namespace *mnt_userns,
 				   struct dentry *unused, struct inode *inode,
 				   const char *name, const void *buffer,
 				   size_t size, int flags)
@@ -371,6 +372,7 @@ static int btrfs_xattr_handler_set(const struct xattr_handler *handler,
 }
 
 static int btrfs_xattr_handler_set_prop(const struct xattr_handler *handler,
+					struct user_namespace *mnt_userns,
 					struct dentry *unused, struct inode *inode,
 					const char *name, const void *value,
 					size_t size, int flags)
fs/ceph/acl.c +1 -1
index e0465741c591..bceab4c01585 100644
@@ -100,7 +100,7 @@ int ceph_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 	case ACL_TYPE_ACCESS:
 		name = XATTR_NAME_POSIX_ACL_ACCESS;
 		if (acl) {
-			ret = posix_acl_update_mode(inode, &new_mode, &acl);
+			ret = posix_acl_update_mode(&init_user_ns, inode, &new_mode, &acl);
 			if (ret)
 				goto out;
 		}
fs/ceph/inode.c +1 -1
index 285d3baca27e..145e26a4ddbb 100644
@@ -2262,7 +2262,7 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr)
 	err = __ceph_setattr(inode, attr);
 
 	if (err >= 0 && (attr->ia_valid & ATTR_MODE))
-		err = posix_acl_chmod(inode, attr->ia_mode);
+		err = posix_acl_chmod(&init_user_ns, inode, attr->ia_mode);
 
 	return err;
 }
fs/ceph/xattr.c +1 -0
index 24997982de01..02f59bcb4f27 100644
@@ -1238,6 +1238,7 @@ static int ceph_get_xattr_handler(const struct xattr_handler *handler,
 }
 
 static int ceph_set_xattr_handler(const struct xattr_handler *handler,
+				  struct user_namespace *mnt_userns,
 				  struct dentry *unused, struct inode *inode,
 				  const char *name, const void *value,
 				  size_t size, int flags)
fs/cifs/xattr.c +1 -0
index 6b658a1172ef..41a611e76bb7 100644
@@ -101,6 +101,7 @@ static int cifs_creation_time_set(unsigned int xid, struct cifs_tcon *pTcon,
 }
 
 static int cifs_xattr_set(const struct xattr_handler *handler,
+			  struct user_namespace *mnt_userns,
 			  struct dentry *dentry, struct inode *inode,
 			  const char *name, const void *value,
 			  size_t size, int flags)
fs/ecryptfs/inode.c +1 -0
index 2454aef02a68..fa794c98c395 100644
@@ -1131,6 +1131,7 @@ static int ecryptfs_xattr_get(const struct xattr_handler *handler,
 }
 
 static int ecryptfs_xattr_set(const struct xattr_handler *handler,
+			      struct user_namespace *mnt_userns,
 			      struct dentry *dentry, struct inode *inode,
 			      const char *name, const void *value, size_t size,
 			      int flags)
fs/ext2/acl.c +1 -1
index cf4c77f8dd08..826987b23ccb 100644
@@ -223,7 +223,7 @@ ext2_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 	umode_t mode = inode->i_mode;
 
 	if (type == ACL_TYPE_ACCESS && acl) {
-		error = posix_acl_update_mode(inode, &mode, &acl);
+		error = posix_acl_update_mode(&init_user_ns, inode, &mode, &acl);
 		if (error)
 			return error;
 		update_mode = 1;
fs/ext2/inode.c +1 -1
index 06c0cf28c1a0..9de813635d8d 100644
@@ -1691,7 +1691,7 @@ int ext2_setattr(struct dentry *dentry, struct iattr *iattr)
 	}
 	setattr_copy(&init_user_ns, inode, iattr);
 	if (iattr->ia_valid & ATTR_MODE)
-		error = posix_acl_chmod(inode, inode->i_mode);
+		error = posix_acl_chmod(&init_user_ns, inode, inode->i_mode);
 	mark_inode_dirty(inode);
 
 	return error;
fs/ext2/xattr_security.c +1 -0
index 9a682e440acb..ebade1f52451 100644
@@ -19,6 +19,7 @@ ext2_xattr_security_get(const struct xattr_handler *handler,
 
 static int
 ext2_xattr_security_set(const struct xattr_handler *handler,
+			struct user_namespace *mnt_userns,
 			struct dentry *unused, struct inode *inode,
 			const char *name, const void *value,
 			size_t size, int flags)
fs/ext2/xattr_trusted.c +1 -0
index 49add1107850..18a87d5dd1ab 100644
@@ -26,6 +26,7 @@ ext2_xattr_trusted_get(const struct xattr_handler *handler,
 
 static int
 ext2_xattr_trusted_set(const struct xattr_handler *handler,
+		       struct user_namespace *mnt_userns,
 		       struct dentry *unused, struct inode *inode,
 		       const char *name, const void *value,
 		       size_t size, int flags)
fs/ext2/xattr_user.c +1 -0
index c243a3b4d69d..58092449f8ff 100644
@@ -30,6 +30,7 @@ ext2_xattr_user_get(const struct xattr_handler *handler,
 
 static int
 ext2_xattr_user_set(const struct xattr_handler *handler,
+		    struct user_namespace *mnt_userns,
 		    struct dentry *unused, struct inode *inode,
 		    const char *name, const void *value,
 		    size_t size, int flags)
fs/ext4/acl.c +1 -1
index 68aaed48315f..4aad060010d8 100644
@@ -245,7 +245,7 @@ ext4_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 	ext4_fc_start_update(inode);
 
 	if ((type == ACL_TYPE_ACCESS) && acl) {
-		error = posix_acl_update_mode(inode, &mode, &acl);
+		error = posix_acl_update_mode(&init_user_ns, inode, &mode, &acl);
 		if (error)
 			goto out_stop;
 		if (mode != inode->i_mode)
fs/ext4/inode.c +1 -1
index 644510b5e06c..55990b74b9e4 100644
@@ -5520,7 +5520,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 		ext4_orphan_del(NULL, inode);
 
 	if (!error && (ia_valid & ATTR_MODE))
-		rc = posix_acl_chmod(inode, inode->i_mode);
+		rc = posix_acl_chmod(&init_user_ns, inode, inode->i_mode);
 
 err_out:
 	if  (error)
fs/ext4/xattr_hurd.c +1 -0
index 8cfa74a56361..c78df5790377 100644
@@ -32,6 +32,7 @@ ext4_xattr_hurd_get(const struct xattr_handler *handler,
 
 static int
 ext4_xattr_hurd_set(const struct xattr_handler *handler,
+		    struct user_namespace *mnt_userns,
 		    struct dentry *unused, struct inode *inode,
 		    const char *name, const void *value,
 		    size_t size, int flags)
fs/ext4/xattr_security.c +1 -0
index 197a9d8a15ef..8213f66f7b2d 100644
@@ -23,6 +23,7 @@ ext4_xattr_security_get(const struct xattr_handler *handler,
 
 static int
 ext4_xattr_security_set(const struct xattr_handler *handler,
+			struct user_namespace *mnt_userns,
 			struct dentry *unused, struct inode *inode,
 			const char *name, const void *value,
 			size_t size, int flags)
fs/ext4/xattr_trusted.c +1 -0
index e9389e5d75c3..7c21ffb26d25 100644
@@ -30,6 +30,7 @@ ext4_xattr_trusted_get(const struct xattr_handler *handler,
 
 static int
 ext4_xattr_trusted_set(const struct xattr_handler *handler,
+		       struct user_namespace *mnt_userns,
 		       struct dentry *unused, struct inode *inode,
 		       const char *name, const void *value,
 		       size_t size, int flags)
fs/ext4/xattr_user.c +1 -0
index d4546184b34b..2fe7ff0a479c 100644
@@ -31,6 +31,7 @@ ext4_xattr_user_get(const struct xattr_handler *handler,
 
 static int
 ext4_xattr_user_set(const struct xattr_handler *handler,
+		    struct user_namespace *mnt_userns,
 		    struct dentry *unused, struct inode *inode,
 		    const char *name, const void *value,
 		    size_t size, int flags)
fs/f2fs/acl.c +1 -1
index 1e5e9b1136ee..144e434834b3 100644
@@ -213,7 +213,7 @@ static int __f2fs_set_acl(struct inode *inode, int type,
 	case ACL_TYPE_ACCESS:
 		name_index = F2FS_XATTR_INDEX_POSIX_ACL_ACCESS;
 		if (acl && !ipage) {
-			error = posix_acl_update_mode(inode, &mode, &acl);
+			error = posix_acl_update_mode(&init_user_ns, inode, &mode, &acl);
 			if (error)
 				return error;
 			set_acl_inode(inode, mode);
fs/f2fs/file.c +1 -1
index 90d7b89176de..ba18b7e28d51 100644
@@ -950,7 +950,7 @@ int f2fs_setattr(struct dentry *dentry, struct iattr *attr)
 	__setattr_copy(&init_user_ns, inode, attr);
 
 	if (attr->ia_valid & ATTR_MODE) {
-		err = posix_acl_chmod(inode, f2fs_get_inode_mode(inode));
+		err = posix_acl_chmod(&init_user_ns, inode, f2fs_get_inode_mode(inode));
 		if (err || is_inode_flag_set(inode, FI_ACL_MODE)) {
 			inode->i_mode = F2FS_I(inode)->i_acl_mode;
 			clear_inode_flag(inode, FI_ACL_MODE);
fs/f2fs/xattr.c +2 -0
index d772bf13a814..10081bf74324 100644
@@ -64,6 +64,7 @@ static int f2fs_xattr_generic_get(const struct xattr_handler *handler,
 }
 
 static int f2fs_xattr_generic_set(const struct xattr_handler *handler,
+		struct user_namespace *mnt_userns,
 		struct dentry *unused, struct inode *inode,
 		const char *name, const void *value,
 		size_t size, int flags)
@@ -107,6 +108,7 @@ static int f2fs_xattr_advise_get(const struct xattr_handler *handler,
 }
 
 static int f2fs_xattr_advise_set(const struct xattr_handler *handler,
+		struct user_namespace *mnt_userns,
 		struct dentry *unused, struct inode *inode,
 		const char *name, const void *value,
 		size_t size, int flags)
fs/fuse/xattr.c +2 -0
index cdea18de94f7..1a7d7ace54e1 100644
@@ -188,6 +188,7 @@ static int fuse_xattr_get(const struct xattr_handler *handler,
 }
 
 static int fuse_xattr_set(const struct xattr_handler *handler,
+			  struct user_namespace *mnt_userns,
 			  struct dentry *dentry, struct inode *inode,
 			  const char *name, const void *value, size_t size,
 			  int flags)
@@ -214,6 +215,7 @@ static int no_xattr_get(const struct xattr_handler *handler,
 }
 
 static int no_xattr_set(const struct xattr_handler *handler,
+			struct user_namespace *mnt_userns,
 			struct dentry *dentry, struct inode *nodee,
 			const char *name, const void *value,
 			size_t size, int flags)
fs/gfs2/acl.c +1 -1
index 2e939f5fe751..ce88ef29eef0 100644
@@ -130,7 +130,7 @@ int gfs2_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 
 	mode = inode->i_mode;
 	if (type == ACL_TYPE_ACCESS && acl) {
-		ret = posix_acl_update_mode(inode, &mode, &acl);
+		ret = posix_acl_update_mode(&init_user_ns, inode, &mode, &acl);
 		if (ret)
 			goto unlock;
 	}
fs/gfs2/inode.c +1 -1
index 59c25181d108..aef2f7587392 100644
@@ -1993,7 +1993,7 @@ static int gfs2_setattr(struct dentry *dentry, struct iattr *attr)
 	else {
 		error = gfs2_setattr_simple(inode, attr);
 		if (!error && attr->ia_valid & ATTR_MODE)
-			error = posix_acl_chmod(inode, inode->i_mode);
+			error = posix_acl_chmod(&init_user_ns, inode, inode->i_mode);
 	}
 
 error:
fs/gfs2/xattr.c +1 -0
index 9d7667bc4292..13969a813410 100644
@@ -1214,6 +1214,7 @@ int __gfs2_xattr_set(struct inode *inode, const char *name,
 }
 
 static int gfs2_xattr_set(const struct xattr_handler *handler,
+			  struct user_namespace *mnt_userns,
 			  struct dentry *unused, struct inode *inode,
 			  const char *name, const void *value,
 			  size_t size, int flags)
fs/hfs/attr.c +1 -0
index 74fa62643136..2bd54efaf416 100644
@@ -121,6 +121,7 @@ static int hfs_xattr_get(const struct xattr_handler *handler,
 }
 
 static int hfs_xattr_set(const struct xattr_handler *handler,
+			 struct user_namespace *mnt_userns,
 			 struct dentry *unused, struct inode *inode,
 			 const char *name, const void *value, size_t size,
 			 int flags)
fs/hfsplus/xattr.c +1 -0
index bb0b27d88e50..4d169c5a2673 100644
@@ -858,6 +858,7 @@ static int hfsplus_osx_getxattr(const struct xattr_handler *handler,
 }
 
 static int hfsplus_osx_setxattr(const struct xattr_handler *handler,
+				struct user_namespace *mnt_userns,
 				struct dentry *unused, struct inode *inode,
 				const char *name, const void *buffer,
 				size_t size, int flags)
fs/hfsplus/xattr_security.c +1 -0
index cfbe6a3bfb1e..c1c7a16cbf21 100644
@@ -23,6 +23,7 @@ static int hfsplus_security_getxattr(const struct xattr_handler *handler,
 }
 
 static int hfsplus_security_setxattr(const struct xattr_handler *handler,
+				     struct user_namespace *mnt_userns,
 				     struct dentry *unused, struct inode *inode,
 				     const char *name, const void *buffer,
 				     size_t size, int flags)
fs/hfsplus/xattr_trusted.c +1 -0
index fbad91e1dada..e150372ec564 100644
@@ -22,6 +22,7 @@ static int hfsplus_trusted_getxattr(const struct xattr_handler *handler,
 }
 
 static int hfsplus_trusted_setxattr(const struct xattr_handler *handler,
+				    struct user_namespace *mnt_userns,
 				    struct dentry *unused, struct inode *inode,
 				    const char *name, const void *buffer,
 				    size_t size, int flags)
fs/hfsplus/xattr_user.c +1 -0
index 74d19faf255e..a6b60b153916 100644
@@ -22,6 +22,7 @@ static int hfsplus_user_getxattr(const struct xattr_handler *handler,
 }
 
 static int hfsplus_user_setxattr(const struct xattr_handler *handler,
+				 struct user_namespace *mnt_userns,
 				 struct dentry *unused, struct inode *inode,
 				 const char *name, const void *buffer,
 				 size_t size, int flags)
fs/jffs2/acl.c +1 -1
index 093ffbd82395..cf07a2fdf8bf 100644
@@ -236,7 +236,7 @@ int jffs2_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 		if (acl) {
 			umode_t mode;
 
-			rc = posix_acl_update_mode(inode, &mode, &acl);
+			rc = posix_acl_update_mode(&init_user_ns, inode, &mode, &acl);
 			if (rc)
 				return rc;
 			if (inode->i_mode != mode) {
fs/jffs2/fs.c +1 -1
index 67993808f4da..ee9f51bab4c6 100644
@@ -201,7 +201,7 @@ int jffs2_setattr(struct dentry *dentry, struct iattr *iattr)
 
 	rc = jffs2_do_setattr(inode, iattr);
 	if (!rc && (iattr->ia_valid & ATTR_MODE))
-		rc = posix_acl_chmod(inode, inode->i_mode);
+		rc = posix_acl_chmod(&init_user_ns, inode, inode->i_mode);
 
 	return rc;
 }
fs/jffs2/security.c +1 -0
index c2332e30f218..aef5522551db 100644
@@ -57,6 +57,7 @@ static int jffs2_security_getxattr(const struct xattr_handler *handler,
 }
 
 static int jffs2_security_setxattr(const struct xattr_handler *handler,
+				   struct user_namespace *mnt_userns,
 				   struct dentry *unused, struct inode *inode,
 				   const char *name, const void *buffer,
 				   size_t size, int flags)
fs/jffs2/xattr_trusted.c +1 -0
index 5d6030826c52..cc3f24883e7d 100644
@@ -25,6 +25,7 @@ static int jffs2_trusted_getxattr(const struct xattr_handler *handler,
 }
 
 static int jffs2_trusted_setxattr(const struct xattr_handler *handler,
+				  struct user_namespace *mnt_userns,
 				  struct dentry *unused, struct inode *inode,
 				  const char *name, const void *buffer,
 				  size_t size, int flags)
fs/jffs2/xattr_user.c +1 -0
index 9d027b4abcf9..fb945977c013 100644
@@ -25,6 +25,7 @@ static int jffs2_user_getxattr(const struct xattr_handler *handler,
 }
 
 static int jffs2_user_setxattr(const struct xattr_handler *handler,
+			       struct user_namespace *mnt_userns,
 			       struct dentry *unused, struct inode *inode,
 			       const char *name, const void *buffer,
 			       size_t size, int flags)
fs/jfs/acl.c +1 -1
index 92cc0ac2d1fc..cf79a34bfada 100644
@@ -101,7 +101,7 @@ int jfs_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 	tid = txBegin(inode->i_sb, 0);
 	mutex_lock(&JFS_IP(inode)->commit_mutex);
 	if (type == ACL_TYPE_ACCESS && acl) {
-		rc = posix_acl_update_mode(inode, &mode, &acl);
+		rc = posix_acl_update_mode(&init_user_ns, inode, &mode, &acl);
 		if (rc)
 			goto end_tx;
 		if (mode != inode->i_mode)
fs/jfs/file.c +1 -1
index ff49876e9c9b..61c3b0c1fbf6 100644
@@ -122,7 +122,7 @@ int jfs_setattr(struct dentry *dentry, struct iattr *iattr)
 	mark_inode_dirty(inode);
 
 	if (iattr->ia_valid & ATTR_MODE)
-		rc = posix_acl_chmod(inode, inode->i_mode);
+		rc = posix_acl_chmod(&init_user_ns, inode, inode->i_mode);
 	return rc;
 }
 
fs/jfs/xattr.c +2 -0
index db41e7803163..f9273f6901c8 100644
@@ -932,6 +932,7 @@ static int jfs_xattr_get(const struct xattr_handler *handler,
 }
 
 static int jfs_xattr_set(const struct xattr_handler *handler,
+			 struct user_namespace *mnt_userns,
 			 struct dentry *unused, struct inode *inode,
 			 const char *name, const void *value,
 			 size_t size, int flags)
@@ -950,6 +951,7 @@ static int jfs_xattr_get_os2(const struct xattr_handler *handler,
 }
 
 static int jfs_xattr_set_os2(const struct xattr_handler *handler,
+			     struct user_namespace *mnt_userns,
 			     struct dentry *unused, struct inode *inode,
 			     const char *name, const void *value,
 			     size_t size, int flags)
fs/kernfs/inode.c +2 -0
index 86bd4c593b78..7e44052b42e1 100644
@@ -319,6 +319,7 @@ static int kernfs_vfs_xattr_get(const struct xattr_handler *handler,
 }
 
 static int kernfs_vfs_xattr_set(const struct xattr_handler *handler,
+				struct user_namespace *mnt_userns,
 				struct dentry *unused, struct inode *inode,
 				const char *suffix, const void *value,
 				size_t size, int flags)
@@ -385,6 +386,7 @@ static int kernfs_vfs_user_xattr_rm(struct kernfs_node *kn,
 }
 
 static int kernfs_vfs_user_xattr_set(const struct xattr_handler *handler,
+				     struct user_namespace *mnt_userns,
 				     struct dentry *unused, struct inode *inode,
 				     const char *suffix, const void *value,
 				     size_t size, int flags)
fs/nfs/nfs4proc.c +3 -0
index 0ce04e0e5d82..98ac98bb999c 100644
@@ -7497,6 +7497,7 @@ nfs4_release_lockowner(struct nfs_server *server, struct nfs4_lock_state *lsp)
 #define XATTR_NAME_NFSV4_ACL "system.nfs4_acl"
 
 static int nfs4_xattr_set_nfs4_acl(const struct xattr_handler *handler,
+				   struct user_namespace *mnt_userns,
 				   struct dentry *unused, struct inode *inode,
 				   const char *key, const void *buf,
 				   size_t buflen, int flags)
@@ -7519,6 +7520,7 @@ static bool nfs4_xattr_list_nfs4_acl(struct dentry *dentry)
 #ifdef CONFIG_NFS_V4_SECURITY_LABEL
 
 static int nfs4_xattr_set_nfs4_label(const struct xattr_handler *handler,
+				     struct user_namespace *mnt_userns,
 				     struct dentry *unused, struct inode *inode,
 				     const char *key, const void *buf,
 				     size_t buflen, int flags)
@@ -7569,6 +7571,7 @@ nfs4_listxattr_nfs4_label(struct inode *inode, char *list, size_t list_len)
 
 #ifdef CONFIG_NFS_V4_2
 static int nfs4_xattr_set_nfs4_user(const struct xattr_handler *handler,
+				    struct user_namespace *mnt_userns,
 				    struct dentry *unused, struct inode *inode,
 				    const char *key, const void *buf,
 				    size_t buflen, int flags)
fs/nfsd/nfs2acl.c +2 -2
index b0f66604532a..6808645cb6b0 100644
@@ -113,10 +113,10 @@ static __be32 nfsacld_proc_setacl(struct svc_rqst *rqstp)
 
 	fh_lock(fh);
 
-	error = set_posix_acl(inode, ACL_TYPE_ACCESS, argp->acl_access);
+	error = set_posix_acl(&init_user_ns, inode, ACL_TYPE_ACCESS, argp->acl_access);
 	if (error)
 		goto out_drop_lock;
-	error = set_posix_acl(inode, ACL_TYPE_DEFAULT, argp->acl_default);
+	error = set_posix_acl(&init_user_ns, inode, ACL_TYPE_DEFAULT, argp->acl_default);
 	if (error)
 		goto out_drop_lock;
 
fs/nfsd/nfs3acl.c +2 -2
index 7c30876a31a1..c38a818ca041 100644
@@ -103,10 +103,10 @@ static __be32 nfsd3_proc_setacl(struct svc_rqst *rqstp)
 
 	fh_lock(fh);
 
-	error = set_posix_acl(inode, ACL_TYPE_ACCESS, argp->acl_access);
+	error = set_posix_acl(&init_user_ns, inode, ACL_TYPE_ACCESS, argp->acl_access);
 	if (error)
 		goto out_drop_lock;
-	error = set_posix_acl(inode, ACL_TYPE_DEFAULT, argp->acl_default);
+	error = set_posix_acl(&init_user_ns, inode, ACL_TYPE_DEFAULT, argp->acl_default);
 
 out_drop_lock:
 	fh_unlock(fh);
fs/nfsd/nfs4acl.c +2 -2
index 71292a0d6f09..0b35bc490aee 100644
@@ -781,12 +781,12 @@ nfsd4_set_nfs4_acl(struct svc_rqst *rqstp, struct svc_fh *fhp,
 
 	fh_lock(fhp);
 
-	host_error = set_posix_acl(inode, ACL_TYPE_ACCESS, pacl);
+	host_error = set_posix_acl(&init_user_ns, inode, ACL_TYPE_ACCESS, pacl);
 	if (host_error < 0)
 		goto out_drop_lock;
 
 	if (S_ISDIR(inode->i_mode)) {
-		host_error = set_posix_acl(inode, ACL_TYPE_DEFAULT, dpacl);
+		host_error = set_posix_acl(&init_user_ns, inode, ACL_TYPE_DEFAULT, dpacl);
 	}
 
 out_drop_lock:
fs/ocfs2/acl.c +1 -1
index 7b07f5df3a29..7e64dbe93251 100644
@@ -274,7 +274,7 @@ int ocfs2_iop_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 	if (type == ACL_TYPE_ACCESS && acl) {
 		umode_t mode;
 
-		status = posix_acl_update_mode(inode, &mode, &acl);
+		status = posix_acl_update_mode(&init_user_ns, inode, &mode, &acl);
 		if (status)
 			goto unlock;
 
fs/ocfs2/xattr.c +3 -0
index 9ccd19d8f7b1..36ae47a4aef6 100644
@@ -7249,6 +7249,7 @@ static int ocfs2_xattr_security_get(const struct xattr_handler *handler,
 }
 
 static int ocfs2_xattr_security_set(const struct xattr_handler *handler,
+				    struct user_namespace *mnt_userns,
 				    struct dentry *unused, struct inode *inode,
 				    const char *name, const void *value,
 				    size_t size, int flags)
@@ -7321,6 +7322,7 @@ static int ocfs2_xattr_trusted_get(const struct xattr_handler *handler,
 }
 
 static int ocfs2_xattr_trusted_set(const struct xattr_handler *handler,
+				   struct user_namespace *mnt_userns,
 				   struct dentry *unused, struct inode *inode,
 				   const char *name, const void *value,
 				   size_t size, int flags)
@@ -7351,6 +7353,7 @@ static int ocfs2_xattr_user_get(const struct xattr_handler *handler,
 }
 
 static int ocfs2_xattr_user_set(const struct xattr_handler *handler,
+				struct user_namespace *mnt_userns,
 				struct dentry *unused, struct inode *inode,
 				const char *name, const void *value,
 				size_t size, int flags)
fs/orangefs/acl.c +1 -1
index a25e6c890975..ba55d61906c2 100644
@@ -132,7 +132,7 @@ int orangefs_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 		 * and "mode" to the new desired value. It is up to
 		 * us to propagate the new mode back to the server...
 		 */
-		error = posix_acl_update_mode(inode, &iattr.ia_mode, &acl);
+		error = posix_acl_update_mode(&init_user_ns, inode, &iattr.ia_mode, &acl);
 		if (error) {
 			gossip_err("%s: posix_acl_update_mode err: %d\n",
 				   __func__,
fs/orangefs/inode.c +1 -1
index 8ac9491ceb9a..563fe9ab8eb2 100644
@@ -861,7 +861,7 @@ int __orangefs_setattr(struct inode *inode, struct iattr *iattr)
 
 	if (iattr->ia_valid & ATTR_MODE)
 		/* change mod on a file that has ACLs */
-		ret = posix_acl_chmod(inode, inode->i_mode);
+		ret = posix_acl_chmod(&init_user_ns, inode, inode->i_mode);
 
 	ret = 0;
 out:
fs/orangefs/xattr.c +1 -0
index bdc285aea360..9a5b757fbd2f 100644
@@ -526,6 +526,7 @@ ssize_t orangefs_listxattr(struct dentry *dentry, char *buffer, size_t size)
 }
 
 static int orangefs_xattr_set_default(const struct xattr_handler *handler,
+				      struct user_namespace *mnt_userns,
 				      struct dentry *unused,
 				      struct inode *inode,
 				      const char *name,
fs/overlayfs/super.c +3 -0
index 39b2e9aa0e5b..e24c995c5fd4 100644
@@ -980,6 +980,7 @@ ovl_posix_acl_xattr_get(const struct xattr_handler *handler,
 
 static int __maybe_unused
 ovl_posix_acl_xattr_set(const struct xattr_handler *handler,
+			struct user_namespace *mnt_userns,
 			struct dentry *dentry, struct inode *inode,
 			const char *name, const void *value,
 			size_t size, int flags)
@@ -1044,6 +1045,7 @@ static int ovl_own_xattr_get(const struct xattr_handler *handler,
 }
 
 static int ovl_own_xattr_set(const struct xattr_handler *handler,
+			     struct user_namespace *mnt_userns,
 			     struct dentry *dentry, struct inode *inode,
 			     const char *name, const void *value,
 			     size_t size, int flags)
@@ -1059,6 +1061,7 @@ static int ovl_other_xattr_get(const struct xattr_handler *handler,
 }
 
 static int ovl_other_xattr_set(const struct xattr_handler *handler,
+			       struct user_namespace *mnt_userns,
 			       struct dentry *dentry, struct inode *inode,
 			       const char *name, const void *value,
 			       size_t size, int flags)
fs/posix_acl.c +55 -22

164 lines not on this page · raw message

fs/reiserfs/xattr_acl.c +2 -2
index ccd40df6eb45..b8f397134c17 100644
@@ -40,7 +40,7 @@ reiserfs_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 	reiserfs_write_unlock(inode->i_sb);
 	if (error == 0) {
 		if (type == ACL_TYPE_ACCESS && acl) {
-			error = posix_acl_update_mode(inode, &mode, &acl);
+			error = posix_acl_update_mode(&init_user_ns, inode, &mode, &acl);
 			if (error)
 				goto unlock;
 			update_mode = 1;
@@ -399,5 +399,5 @@ int reiserfs_acl_chmod(struct inode *inode)
 	    !reiserfs_posixacl(inode->i_sb))
 		return 0;
 
-	return posix_acl_chmod(inode, inode->i_mode);
+	return posix_acl_chmod(&init_user_ns, inode, inode->i_mode);
 }
fs/reiserfs/xattr_security.c +2 -1
index 20be9a0e5870..3411ce2bfbd0 100644
@@ -21,7 +21,8 @@ security_get(const struct xattr_handler *handler, struct dentry *unused,
 }
 
 static int
-security_set(const struct xattr_handler *handler, struct dentry *unused,
+security_set(const struct xattr_handler *handler, struct user_namespace *mnt_userns,
+	     struct dentry *unused,
 	     struct inode *inode, const char *name, const void *buffer,
 	     size_t size, int flags)
 {
fs/reiserfs/xattr_trusted.c +2 -1
index 5ed48da3d02b..d338a24f26e7 100644
@@ -20,7 +20,8 @@ trusted_get(const struct xattr_handler *handler, struct dentry *unused,
 }
 
 static int
-trusted_set(const struct xattr_handler *handler, struct dentry *unused,
+trusted_set(const struct xattr_handler *handler, struct user_namespace *mnt_userns,
+	    struct dentry *unused,
 	    struct inode *inode, const char *name, const void *buffer,
 	    size_t size, int flags)
 {
fs/reiserfs/xattr_user.c +2 -1
index a573ca45bacc..65d9cd10a5ea 100644
@@ -18,7 +18,8 @@ user_get(const struct xattr_handler *handler, struct dentry *unused,
 }
 
 static int
-user_set(const struct xattr_handler *handler, struct dentry *unused,
+user_set(const struct xattr_handler *handler, struct user_namespace *mnt_userns,
+	 struct dentry *unused,
 	 struct inode *inode, const char *name, const void *buffer,
 	 size_t size, int flags)
 {
fs/ubifs/xattr.c +1 -0
index a0b9b349efe6..8f4135c22ab6 100644
@@ -681,6 +681,7 @@ static int xattr_get(const struct xattr_handler *handler,
 }
 
 static int xattr_set(const struct xattr_handler *handler,
+			   struct user_namespace *mnt_userns,
 			   struct dentry *dentry, struct inode *inode,
 			   const char *name, const void *value,
 			   size_t size, int flags)
fs/xattr.c +5 -5
index 7773ed7e0562..dc6068c6b7e5 100644
@@ -174,7 +174,7 @@ __vfs_setxattr(struct dentry *dentry, struct inode *inode, const char *name,
 		return -EOPNOTSUPP;
 	if (size == 0)
 		value = "";  /* empty EA, do not remove */
-	return handler->set(handler, dentry, inode, name, value, size, flags);
+	return handler->set(handler, &init_user_ns, dentry, inode, name, value, size, flags);
 }
 EXPORT_SYMBOL(__vfs_setxattr);
 
@@ -280,7 +280,7 @@ vfs_setxattr(struct dentry *dentry, const char *name, const void *value,
 	int error;
 
 	if (size && strcmp(name, XATTR_NAME_CAPS) == 0) {
-		error = cap_convert_nscap(dentry, &value, size);
+		error = cap_convert_nscap(&init_user_ns, dentry, &value, size);
 		if (error < 0)
 			return error;
 		size = error;
@@ -449,7 +449,7 @@ __vfs_removexattr(struct dentry *dentry, const char *name)
 		return PTR_ERR(handler);
 	if (!handler->set)
 		return -EOPNOTSUPP;
-	return handler->set(handler, dentry, inode, name, NULL, 0, XATTR_REPLACE);
+	return handler->set(handler, &init_user_ns, dentry, inode, name, NULL, 0, XATTR_REPLACE);
 }
 EXPORT_SYMBOL(__vfs_removexattr);
 
@@ -547,7 +547,7 @@ setxattr(struct dentry *d, const char __user *name, const void __user *value,
 		}
 		if ((strcmp(kname, XATTR_NAME_POSIX_ACL_ACCESS) == 0) ||
 		    (strcmp(kname, XATTR_NAME_POSIX_ACL_DEFAULT) == 0))
-			posix_acl_fix_xattr_from_user(kvalue, size);
+			posix_acl_fix_xattr_from_user(&init_user_ns, kvalue, size);
 	}
 
 	error = vfs_setxattr(d, kname, kvalue, size, flags);
@@ -641,7 +641,7 @@ getxattr(struct dentry *d, const char __user *name, void __user *value,
 	if (error > 0) {
 		if ((strcmp(kname, XATTR_NAME_POSIX_ACL_ACCESS) == 0) ||
 		    (strcmp(kname, XATTR_NAME_POSIX_ACL_DEFAULT) == 0))
-			posix_acl_fix_xattr_to_user(kvalue, error);
+			posix_acl_fix_xattr_to_user(&init_user_ns, kvalue, error);
 		if (size && copy_to_user(value, kvalue, error))
 			error = -EFAULT;
 	} else if (error == -ERANGE && size >= XATTR_SIZE_MAX) {
fs/xfs/xfs_acl.c +1 -1
index 779cb73b3d00..0927247d656f 100644
@@ -252,7 +252,7 @@ xfs_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 		return error;
 
 	if (type == ACL_TYPE_ACCESS) {
-		error = posix_acl_update_mode(inode, &mode, &acl);
+		error = posix_acl_update_mode(&init_user_ns, inode, &mode, &acl);
 		if (error)
 			return error;
 		set_mode = true;
fs/xfs/xfs_iops.c +1 -1
index 08a478d25122..26d22edef741 100644
@@ -807,7 +807,7 @@ xfs_setattr_nonsize(
 	 * 	     Posix ACL code seems to care about this issue either.
 	 */
 	if (mask & ATTR_MODE) {
-		error = posix_acl_chmod(inode, inode->i_mode);
+		error = posix_acl_chmod(&init_user_ns, inode, inode->i_mode);
 		if (error)
 			return error;
 	}
fs/xfs/xfs_xattr.c +2 -1
index bca48b308c02..4ea08dd47826 100644
@@ -38,7 +38,8 @@ xfs_xattr_get(const struct xattr_handler *handler, struct dentry *unused,
 }
 
 static int
-xfs_xattr_set(const struct xattr_handler *handler, struct dentry *unused,
+xfs_xattr_set(const struct xattr_handler *handler, struct user_namespace *mnt_userns,
+		struct dentry *unused,
 		struct inode *inode, const char *name, const void *value,
 		size_t size, int flags)
 {
include/linux/capability.h +3 -1
index 53d5acfd23c4..575436b977bc 100644
@@ -273,6 +273,8 @@ static inline bool checkpoint_restore_ns_capable(struct user_namespace *ns)
 /* audit system wants to get cap info from files as well */
 extern int get_vfs_caps_from_disk(const struct dentry *dentry, struct cpu_vfs_cap_data *cpu_caps);
 
-extern int cap_convert_nscap(struct dentry *dentry, const void **ivalue, size_t size);
+extern int cap_convert_nscap(struct user_namespace *mnt_userns,
+			     struct dentry *dentry, const void **ivalue,
+			     size_t size);
 
 #endif /* !_LINUX_CAPABILITY_H */
include/linux/posix_acl.h +5 -4
index 8276baefed13..50e266fdaaef 100644
@@ -71,13 +71,13 @@ extern int __posix_acl_create(struct posix_acl **, gfp_t, umode_t *);
 extern int __posix_acl_chmod(struct posix_acl **, gfp_t, umode_t);
 
 extern struct posix_acl *get_posix_acl(struct inode *, int);
-extern int set_posix_acl(struct inode *, int, struct posix_acl *);
+extern int set_posix_acl(struct user_namespace *, struct inode *, int, struct posix_acl *);
 
 #ifdef CONFIG_FS_POSIX_ACL
-extern int posix_acl_chmod(struct inode *, umode_t);
+extern int posix_acl_chmod(struct user_namespace *, struct inode *, umode_t);
 extern int posix_acl_create(struct inode *, umode_t *, struct posix_acl **,
 		struct posix_acl **);
-extern int posix_acl_update_mode(struct inode *, umode_t *, struct posix_acl **);
+extern int posix_acl_update_mode(struct user_namespace *, struct inode *, umode_t *, struct posix_acl **);
 
 extern int simple_set_acl(struct inode *, struct posix_acl *, int);
 extern int simple_acl_create(struct inode *, struct inode *);
@@ -94,7 +94,8 @@ static inline void cache_no_acl(struct inode *inode)
 	inode->i_default_acl = NULL;
 }
 #else
-static inline int posix_acl_chmod(struct inode *inode, umode_t mode)
+static inline int posix_acl_chmod(struct user_namespace *mnt_userns,
+				  struct inode *inode, umode_t mode)
 {
 	return 0;
 }
include/linux/posix_acl_xattr.h +8 -4
index 2387709991b5..060e8d203181 100644
@@ -33,13 +33,17 @@ posix_acl_xattr_count(size_t size)
 }
 
 #ifdef CONFIG_FS_POSIX_ACL
-void posix_acl_fix_xattr_from_user(void *value, size_t size);
-void posix_acl_fix_xattr_to_user(void *value, size_t size);
+void posix_acl_fix_xattr_from_user(struct user_namespace *mnt_userns,
+				   void *value, size_t size);
+void posix_acl_fix_xattr_to_user(struct user_namespace *mnt_userns,
+				 void *value, size_t size);
 #else
-static inline void posix_acl_fix_xattr_from_user(void *value, size_t size)
+static inline void posix_acl_fix_xattr_from_user(struct user_namespace *mnt_userns,
+						 void *value, size_t size)
 {
 }
-static inline void posix_acl_fix_xattr_to_user(void *value, size_t size)
+static inline void posix_acl_fix_xattr_to_user(struct user_namespace *mnt_userns,
+					       void *value, size_t size)
 {
 }
 #endif
include/linux/xattr.h +3 -3
index 10b4dc2709f0..d3c7ad273392 100644
@@ -34,9 +34,9 @@ struct xattr_handler {
 	int (*get)(const struct xattr_handler *, struct dentry *dentry,
 		   struct inode *inode, const char *name, void *buffer,
 		   size_t size);
-	int (*set)(const struct xattr_handler *, struct dentry *dentry,
-		   struct inode *inode, const char *name, const void *buffer,
-		   size_t size, int flags);
+	int (*set)(const struct xattr_handler *, struct user_namespace *mnt_userns,
+		   struct dentry *dentry, struct inode *inode, const char *name,
+		   const void *buffer, size_t size, int flags);
 };
 
 const char *xattr_full_name(const struct xattr_handler *, const char *);
mm/shmem.c +2 -1
index 1cb451e131ec..23b8e9c15a42 100644
@@ -1143,7 +1143,7 @@ static int shmem_setattr(struct dentry *dentry, struct iattr *attr)
 
 	setattr_copy(&init_user_ns, inode, attr);
 	if (attr->ia_valid & ATTR_MODE)
-		error = posix_acl_chmod(inode, inode->i_mode);
+		error = posix_acl_chmod(&init_user_ns, inode, inode->i_mode);
 	return error;
 }
 
@@ -3273,6 +3273,7 @@ static int shmem_xattr_handler_get(const struct xattr_handler *handler,
 }
 
 static int shmem_xattr_handler_set(const struct xattr_handler *handler,
+				   struct user_namespace *mnt_userns,
 				   struct dentry *unused, struct inode *inode,
 				   const char *name, const void *value,
 				   size_t size, int flags)
net/socket.c +1 -0
index 33e8b6c4e1d3..c76703c6f480 100644
@@ -334,6 +334,7 @@ static const struct xattr_handler sockfs_xattr_handler = {
 };
 
 static int sockfs_security_xattr_set(const struct xattr_handler *handler,
+				     struct user_namespace *mnt_userns,
 				     struct dentry *dentry, struct inode *inode,
 				     const char *suffix, const void *value,
 				     size_t size, int flags)
security/commoncap.c +40 -7

89 lines not on this page · raw message

Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
#21 lore raw

PATCH v5 15/42 fs: add file_user_ns() helper

In reply to Christian Brauner

Add a simple helper to retrieve the user namespace associated with the vfsmount of a file. Christoph correctly points out that this makes codepaths (e.g. ioctls) way easier to follow that would otherwise dereference via mnt_user_ns(file->f_path.mnt).

In order to make file_user_ns() static inline we'd need to include mount.h in either file.h or fs.h which seems undesirable so let's simply not force file_user_ns() to be inline.

---
/* v2 */
patch not present

/* v3 */
patch not present

/* v4 */
patch not present

/* v5 */
patch introduced
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/namei.c         | 6 ++++++
 include/linux/fs.h | 1 +
 2 files changed, 7 insertions(+)
fs/namei.c +6 -0
index cd124e18cec5..d8dee449e92a 100644
@@ -259,6 +259,12 @@ void putname(struct filename *name)
 		__putname(name);
 }
 
+struct user_namespace *file_user_ns(struct file *file)
+{
+	return mnt_user_ns(file->f_path.mnt);
+}
+EXPORT_SYMBOL(file_user_ns);
+
 /**
  * check_acl - perform ACL permission checking
  * @mnt_userns:	user namespace of the mount the inode was found from
include/linux/fs.h +1 -0
index ef993857682b..89e41a821bad 100644
@@ -2585,6 +2585,7 @@ static inline struct file *file_clone_open(struct file *file)
 	return dentry_open(&file->f_path, file->f_flags, file->f_cred);
 }
 extern int filp_close(struct file *, fl_owner_t id);
+extern struct user_namespace *file_user_ns(struct file *file);
 
 extern struct filename *getname_flags(const char __user *, int, int *);
 extern struct filename *getname(const char __user *);
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
Reviewed-by on v5 15/42
#22 lore raw

Re: [PATCH v5 15/42] fs: add file_user_ns() helper

Quoted 8 lines · Christian Brauner

Add a simple helper to retrieve the user namespace associated with the vfsmount of a file. Christoph correctly points out that this makes codepaths (e.g. ioctls) way easier to follow that would otherwise dereference via mnt_user_ns(file->f_path.mnt).

In order to make file_user_ns() static inline we'd need to include mount.h in either file.h or fs.h which seems undesirable so let's simply not force file_user_ns() to be inline.

I'd be tempted to just make this an inline.

Otherwise this looks ok:

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers

reviewer
#23 lore raw

Re: [PATCH v5 15/42] fs: add file_user_ns() helper

In reply to Christian Brauner
Quoted 8 lines · Christian Brauner

Add a simple helper to retrieve the user namespace associated with the vfsmount of a file. Christoph correctly points out that this makes codepaths (e.g. ioctls) way easier to follow that would otherwise dereference via mnt_user_ns(file->f_path.mnt).

In order to make file_user_ns() static inline we'd need to include mount.h in either file.h or fs.h which seems undesirable so let's simply not force file_user_ns() to be inline.

[...]

fs/namei.c void putname(struct filename *name) · line 265

Quoted 4 lines
fs/namei.c +4 -0
+struct user_namespace *file_user_ns(struct file *file)
+{
+       return mnt_user_ns(file->f_path.mnt);
+}

That name is confusing to me, because when I think of "the userns of a file", it's file->f_cred->user_ns. There are a bunch of places that look at that, as you can see from grepping for "f_cred->user_ns".

If you really want this to be a separate helper, can you maybe give it a clearer name? file_mnt_user_ns(), or something like that, idk.
_______________________________________________
Containers mailing list [email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers

#24 lore raw

Re: [PATCH v5 15/42] fs: add file_user_ns() helper

Quoted 20 lines · Jann Horn
Quoted 8 lines · Christian Brauner

Add a simple helper to retrieve the user namespace associated with the vfsmount of a file. Christoph correctly points out that this makes codepaths (e.g. ioctls) way easier to follow that would otherwise dereference via mnt_user_ns(file->f_path.mnt).

In order to make file_user_ns() static inline we'd need to include mount.h in either file.h or fs.h which seems undesirable so let's simply not force file_user_ns() to be inline.

[...]

fs/namei.c void putname(struct filename *name) · line 265

Quoted 4 lines
fs/namei.c +4 -0
+struct user_namespace *file_user_ns(struct file *file)
+{
+       return mnt_user_ns(file->f_path.mnt);
+}

That name is confusing to me, because when I think of "the userns of a file", it's file->f_cred->user_ns. There are a bunch of places that look at that, as you can see from grepping for "f_cred->user_ns".

Quoted 2 lines

If you really want this to be a separate helper, can you maybe give it a clearer name? file_mnt_user_ns(), or something like that, idk.

Done.
_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers

#25 lore raw

PATCH v5 16/42 xattr: handle idmapped mounts

In reply to Christian Brauner

From: Tycho Andersen <[email protected]>

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.

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Don't pollute the vfs with additional helpers simply extend the existing helpers with an additional argument and switch all callers.

/* v3 */
unchanged

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

- Christoph Hellwig <[email protected]>:
- Use new file_userns_helper().

---
 fs/cachefiles/xattr.c                 |  16 ++--
 fs/ecryptfs/crypto.c                  |   4 +-
 fs/ecryptfs/inode.c                   |   4 +-
 fs/ecryptfs/mmap.c                    |   4 +-
 fs/nfsd/vfs.c                         |  12 +--
 fs/overlayfs/copy_up.c                |  12 +--
 fs/overlayfs/dir.c                    |   2 +-
 fs/overlayfs/inode.c                  |   8 +-
 fs/overlayfs/overlayfs.h              |   6 +-
 fs/overlayfs/super.c                  |   4 +-
 fs/xattr.c                            | 120 +++++++++++++++-----------
 include/linux/xattr.h                 |  24 ++++--
 security/apparmor/domain.c            |   4 +-
 security/commoncap.c                  |   6 +-
 security/integrity/evm/evm_crypto.c   |  11 +--
 security/integrity/evm/evm_main.c     |   4 +-
 security/integrity/ima/ima_appraise.c |   8 +-
 security/selinux/hooks.c              |   2 +-
 security/smack/smack_lsm.c            |   4 +-
 19 files changed, 143 insertions(+), 112 deletions(-)
fs/cachefiles/xattr.c +8 -8
index 72e42438f3d7..3b6a3f1610f4 100644
@@ -39,7 +39,7 @@ int cachefiles_check_object_type(struct cachefiles_object *object)
 	_enter("%p{%s}", object, type);
 
 	/* attempt to install a type label directly */
-	ret = vfs_setxattr(dentry, cachefiles_xattr_cache, type, 2,
+	ret = vfs_setxattr(&init_user_ns, dentry, cachefiles_xattr_cache, type, 2,
 			   XATTR_CREATE);
 	if (ret == 0) {
 		_debug("SET"); /* we succeeded */
@@ -54,7 +54,7 @@ int cachefiles_check_object_type(struct cachefiles_object *object)
 	}
 
 	/* read the current type label */
-	ret = vfs_getxattr(dentry, cachefiles_xattr_cache, xtype, 3);
+	ret = vfs_getxattr(&init_user_ns, dentry, cachefiles_xattr_cache, xtype, 3);
 	if (ret < 0) {
 		if (ret == -ERANGE)
 			goto bad_type_length;
@@ -110,7 +110,7 @@ int cachefiles_set_object_xattr(struct cachefiles_object *object,
 	_debug("SET #%u", auxdata->len);
 
 	clear_bit(FSCACHE_COOKIE_AUX_UPDATED, &object->fscache.cookie->flags);
-	ret = vfs_setxattr(dentry, cachefiles_xattr_cache,
+	ret = vfs_setxattr(&init_user_ns, dentry, cachefiles_xattr_cache,
 			   &auxdata->type, auxdata->len,
 			   XATTR_CREATE);
 	if (ret < 0 && ret != -ENOMEM)
@@ -140,7 +140,7 @@ int cachefiles_update_object_xattr(struct cachefiles_object *object,
 	_debug("SET #%u", auxdata->len);
 
 	clear_bit(FSCACHE_COOKIE_AUX_UPDATED, &object->fscache.cookie->flags);
-	ret = vfs_setxattr(dentry, cachefiles_xattr_cache,
+	ret = vfs_setxattr(&init_user_ns, dentry, cachefiles_xattr_cache,
 			   &auxdata->type, auxdata->len,
 			   XATTR_REPLACE);
 	if (ret < 0 && ret != -ENOMEM)
@@ -171,7 +171,7 @@ int cachefiles_check_auxdata(struct cachefiles_object *object)
 	if (!auxbuf)
 		return -ENOMEM;
 
-	xlen = vfs_getxattr(dentry, cachefiles_xattr_cache,
+	xlen = vfs_getxattr(&init_user_ns, dentry, cachefiles_xattr_cache,
 			    &auxbuf->type, 512 + 1);
 	ret = -ESTALE;
 	if (xlen < 1 ||
@@ -213,7 +213,7 @@ int cachefiles_check_object_xattr(struct cachefiles_object *object,
 	}
 
 	/* read the current type label */
-	ret = vfs_getxattr(dentry, cachefiles_xattr_cache,
+	ret = vfs_getxattr(&init_user_ns, dentry, cachefiles_xattr_cache,
 			   &auxbuf->type, 512 + 1);
 	if (ret < 0) {
 		if (ret == -ENODATA)
@@ -270,7 +270,7 @@ int cachefiles_check_object_xattr(struct cachefiles_object *object,
 		}
 
 		/* update the current label */
-		ret = vfs_setxattr(dentry, cachefiles_xattr_cache,
+		ret = vfs_setxattr(&init_user_ns, dentry, cachefiles_xattr_cache,
 				   &auxdata->type, auxdata->len,
 				   XATTR_REPLACE);
 		if (ret < 0) {
@@ -309,7 +309,7 @@ int cachefiles_remove_object_xattr(struct cachefiles_cache *cache,
 {
 	int ret;
 
-	ret = vfs_removexattr(dentry, cachefiles_xattr_cache);
+	ret = vfs_removexattr(&init_user_ns, dentry, cachefiles_xattr_cache);
 	if (ret < 0) {
 		if (ret == -ENOENT || ret == -ENODATA)
 			ret = 0;
fs/ecryptfs/crypto.c +2 -2
index 0681540c48d9..943e523f4c9d 100644
@@ -1110,8 +1110,8 @@ ecryptfs_write_metadata_to_xattr(struct dentry *ecryptfs_dentry,
 	}
 
 	inode_lock(lower_inode);
-	rc = __vfs_setxattr(lower_dentry, lower_inode, ECRYPTFS_XATTR_NAME,
-			    page_virt, size, 0);
+	rc = __vfs_setxattr(&init_user_ns, lower_dentry, lower_inode,
+			    ECRYPTFS_XATTR_NAME, page_virt, size, 0);
 	if (!rc && ecryptfs_inode)
 		fsstack_copy_attr_all(ecryptfs_inode, lower_inode);
 	inode_unlock(lower_inode);
fs/ecryptfs/inode.c +2 -2
index fa794c98c395..7de9fda25dad 100644
@@ -1022,7 +1022,7 @@ ecryptfs_setxattr(struct dentry *dentry, struct inode *inode,
 		rc = -EOPNOTSUPP;
 		goto out;
 	}
-	rc = vfs_setxattr(lower_dentry, name, value, size, flags);
+	rc = vfs_setxattr(&init_user_ns, lower_dentry, name, value, size, flags);
 	if (!rc && inode)
 		fsstack_copy_attr_all(inode, d_inode(lower_dentry));
 out:
@@ -1087,7 +1087,7 @@ static int ecryptfs_removexattr(struct dentry *dentry, struct inode *inode,
 		goto out;
 	}
 	inode_lock(lower_inode);
-	rc = __vfs_removexattr(lower_dentry, name);
+	rc = __vfs_removexattr(&init_user_ns, lower_dentry, name);
 	inode_unlock(lower_inode);
 out:
 	return rc;
fs/ecryptfs/mmap.c +2 -2
index 019572c6b39a..2f333a40ff4d 100644
@@ -426,8 +426,8 @@ static int ecryptfs_write_inode_size_to_xattr(struct inode *ecryptfs_inode)
 	if (size < 0)
 		size = 8;
 	put_unaligned_be64(i_size_read(ecryptfs_inode), xattr_virt);
-	rc = __vfs_setxattr(lower_dentry, lower_inode, ECRYPTFS_XATTR_NAME,
-			    xattr_virt, size, 0);
+	rc = __vfs_setxattr(&init_user_ns, lower_dentry, lower_inode,
+			    ECRYPTFS_XATTR_NAME, xattr_virt, size, 0);
 	inode_unlock(lower_inode);
 	if (rc)
 		printk(KERN_ERR "Error whilst attempting to write inode size "
fs/nfsd/vfs.c +6 -6
index 5db88d703b63..494362c2e4fc 100644
@@ -499,7 +499,7 @@ int nfsd4_is_junction(struct dentry *dentry)
 		return 0;
 	if (!(inode->i_mode & S_ISVTX))
 		return 0;
-	if (vfs_getxattr(dentry, NFSD_JUNCTION_XATTR_NAME, NULL, 0) <= 0)
+	if (vfs_getxattr(&init_user_ns, dentry, NFSD_JUNCTION_XATTR_NAME, NULL, 0) <= 0)
 		return 0;
 	return 1;
 }
@@ -2149,7 +2149,7 @@ nfsd_getxattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char *name,
 
 	inode_lock_shared(inode);
 
-	len = vfs_getxattr(dentry, name, NULL, 0);
+	len = vfs_getxattr(&init_user_ns, dentry, name, NULL, 0);
 
 	/*
 	 * Zero-length attribute, just return.
@@ -2176,7 +2176,7 @@ nfsd_getxattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char *name,
 		goto out;
 	}
 
-	len = vfs_getxattr(dentry, name, buf, len);
+	len = vfs_getxattr(&init_user_ns, dentry, name, buf, len);
 	if (len <= 0) {
 		kvfree(buf);
 		buf = NULL;
@@ -2283,7 +2283,7 @@ nfsd_removexattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char *name)
 
 	fh_lock(fhp);
 
-	ret = __vfs_removexattr_locked(fhp->fh_dentry, name, NULL);
+	ret = __vfs_removexattr_locked(&init_user_ns, fhp->fh_dentry, name, NULL);
 
 	fh_unlock(fhp);
 	fh_drop_write(fhp);
@@ -2307,8 +2307,8 @@ nfsd_setxattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char *name,
 		return nfserrno(ret);
 	fh_lock(fhp);
 
-	ret = __vfs_setxattr_locked(fhp->fh_dentry, name, buf, len, flags,
-				    NULL);
+	ret = __vfs_setxattr_locked(&init_user_ns, fhp->fh_dentry, name, buf,
+				    len, flags, NULL);
 
 	fh_unlock(fhp);
 	fh_drop_write(fhp);
fs/overlayfs/copy_up.c +6 -6
index 3e9957ae19fa..d17e0d4040b6 100644
@@ -85,9 +85,9 @@ int ovl_copy_xattr(struct super_block *sb, struct dentry *old,
 		if (ovl_is_private_xattr(sb, name))
 			continue;
 retry:
-		size = vfs_getxattr(old, name, value, value_size);
+		size = vfs_getxattr(&init_user_ns, old, name, value, value_size);
 		if (size == -ERANGE)
-			size = vfs_getxattr(old, name, NULL, 0);
+			size = vfs_getxattr(&init_user_ns, old, name, NULL, 0);
 
 		if (size < 0) {
 			error = size;
@@ -114,7 +114,7 @@ int ovl_copy_xattr(struct super_block *sb, struct dentry *old,
 			error = 0;
 			continue; /* Discard */
 		}
-		error = vfs_setxattr(new, name, value, size, 0);
+		error = vfs_setxattr(&init_user_ns, new, name, value, size, 0);
 		if (error) {
 			if (error != -EOPNOTSUPP || ovl_must_copy_xattr(name))
 				break;
@@ -795,7 +795,7 @@ static ssize_t ovl_getxattr(struct dentry *dentry, char *name, char **value)
 	ssize_t res;
 	char *buf;
 
-	res = vfs_getxattr(dentry, name, NULL, 0);
+	res = vfs_getxattr(&init_user_ns, dentry, name, NULL, 0);
 	if (res == -ENODATA || res == -EOPNOTSUPP)
 		res = 0;
 
@@ -804,7 +804,7 @@ static ssize_t ovl_getxattr(struct dentry *dentry, char *name, char **value)
 		if (!buf)
 			return -ENOMEM;
 
-		res = vfs_getxattr(dentry, name, buf, res);
+		res = vfs_getxattr(&init_user_ns, dentry, name, buf, res);
 		if (res < 0)
 			kfree(buf);
 		else
@@ -846,7 +846,7 @@ static int ovl_copy_up_meta_inode_data(struct ovl_copy_up_ctx *c)
 	 * don't want that to happen for normal copy-up operation.
 	 */
 	if (capability) {
-		err = vfs_setxattr(upperpath.dentry, XATTR_NAME_CAPS,
+		err = vfs_setxattr(&init_user_ns, upperpath.dentry, XATTR_NAME_CAPS,
 				   capability, cap_size, 0);
 		if (err)
 			goto out_free;
fs/overlayfs/dir.c +1 -1
index 29840820a46c..d75c96cb18c3 100644
@@ -449,7 +449,7 @@ static int ovl_set_upper_acl(struct dentry *upperdentry, const char *name,
 	if (err < 0)
 		goto out_free;
 
-	err = vfs_setxattr(upperdentry, name, buffer, size, XATTR_CREATE);
+	err = vfs_setxattr(&init_user_ns, upperdentry, name, buffer, size, XATTR_CREATE);
 out_free:
 	kfree(buffer);
 	return err;
fs/overlayfs/inode.c +4 -4
index 5aa66881dbd7..fc6252ccc6d7 100644
@@ -352,7 +352,7 @@ int ovl_xattr_set(struct dentry *dentry, struct inode *inode, const char *name,
 		goto out;
 
 	if (!value && !upperdentry) {
-		err = vfs_getxattr(realdentry, name, NULL, 0);
+		err = vfs_getxattr(&init_user_ns, realdentry, name, NULL, 0);
 		if (err < 0)
 			goto out_drop_write;
 	}
@@ -367,10 +367,10 @@ int ovl_xattr_set(struct dentry *dentry, struct inode *inode, const char *name,
 
 	old_cred = ovl_override_creds(dentry->d_sb);
 	if (value)
-		err = vfs_setxattr(realdentry, name, value, size, flags);
+		err = vfs_setxattr(&init_user_ns, realdentry, name, value, size, flags);
 	else {
 		WARN_ON(flags != XATTR_REPLACE);
-		err = vfs_removexattr(realdentry, name);
+		err = vfs_removexattr(&init_user_ns, realdentry, name);
 	}
 	revert_creds(old_cred);
 
@@ -392,7 +392,7 @@ int ovl_xattr_get(struct dentry *dentry, struct inode *inode, const char *name,
 		ovl_i_dentry_upper(inode) ?: ovl_dentry_lower(dentry);
 
 	old_cred = ovl_override_creds(dentry->d_sb);
-	res = vfs_getxattr(realdentry, name, value, size);
+	res = vfs_getxattr(&init_user_ns, realdentry, name, value, size);
 	revert_creds(old_cred);
 	return res;
 }
fs/overlayfs/overlayfs.h +3 -3
index b487e48c7fd4..0002834f664a 100644
@@ -186,7 +186,7 @@ static inline ssize_t ovl_do_getxattr(struct ovl_fs *ofs, struct dentry *dentry,
 				      size_t size)
 {
 	const char *name = ovl_xattr(ofs, ox);
-	return vfs_getxattr(dentry, name, value, size);
+	return vfs_getxattr(&init_user_ns, dentry, name, value, size);
 }
 
 static inline int ovl_do_setxattr(struct ovl_fs *ofs, struct dentry *dentry,
@@ -194,7 +194,7 @@ static inline int ovl_do_setxattr(struct ovl_fs *ofs, struct dentry *dentry,
 				  size_t size)
 {
 	const char *name = ovl_xattr(ofs, ox);
-	int err = vfs_setxattr(dentry, name, value, size, 0);
+	int err = vfs_setxattr(&init_user_ns, dentry, name, value, size, 0);
 	pr_debug("setxattr(%pd2, \"%s\", \"%*pE\", %zu, 0) = %i\n",
 		 dentry, name, min((int)size, 48), value, size, err);
 	return err;
@@ -204,7 +204,7 @@ static inline int ovl_do_removexattr(struct ovl_fs *ofs, struct dentry *dentry,
 				     enum ovl_xattr ox)
 {
 	const char *name = ovl_xattr(ofs, ox);
-	int err = vfs_removexattr(dentry, name);
+	int err = vfs_removexattr(&init_user_ns, dentry, name);
 	pr_debug("removexattr(%pd2, \"%s\") = %i\n", dentry, name, err);
 	return err;
 }
fs/overlayfs/super.c +2 -2
index e24c995c5fd4..e414dcd307b3 100644
@@ -794,11 +794,11 @@ static struct dentry *ovl_workdir_create(struct ovl_fs *ofs,
 		 * allowed as upper are limited to "normal" ones, where checking
 		 * for the above two errors is sufficient.
 		 */
-		err = vfs_removexattr(work, XATTR_NAME_POSIX_ACL_DEFAULT);
+		err = vfs_removexattr(&init_user_ns, work, XATTR_NAME_POSIX_ACL_DEFAULT);
 		if (err && err != -ENODATA && err != -EOPNOTSUPP)
 			goto out_dput;
 
-		err = vfs_removexattr(work, XATTR_NAME_POSIX_ACL_ACCESS);
+		err = vfs_removexattr(&init_user_ns, work, XATTR_NAME_POSIX_ACL_ACCESS);
 		if (err && err != -ENODATA && err != -EOPNOTSUPP)
 			goto out_dput;
 
fs/xattr.c +72 -48

356 lines not on this page · raw message

include/linux/xattr.h +15 -9
index d3c7ad273392..26d490512db7 100644
@@ -16,6 +16,7 @@
 #include <linux/types.h>
 #include <linux/spinlock.h>
 #include <linux/mm.h>
+#include <linux/user_namespace.h>
 #include <uapi/linux/xattr.h>
 
 struct inode;
@@ -48,18 +49,23 @@ struct xattr {
 };
 
 ssize_t __vfs_getxattr(struct dentry *, struct inode *, const char *, void *, size_t);
-ssize_t vfs_getxattr(struct dentry *, const char *, void *, size_t);
+ssize_t vfs_getxattr(struct user_namespace *, struct dentry *, const char *, void *, size_t);
 ssize_t vfs_listxattr(struct dentry *d, char *list, size_t size);
-int __vfs_setxattr(struct dentry *, struct inode *, const char *, const void *, size_t, int);
-int __vfs_setxattr_noperm(struct dentry *, const char *, const void *, size_t, int);
-int __vfs_setxattr_locked(struct dentry *, const char *, const void *, size_t, int, struct inode **);
-int vfs_setxattr(struct dentry *, const char *, const void *, size_t, int);
-int __vfs_removexattr(struct dentry *, const char *);
-int __vfs_removexattr_locked(struct dentry *, const char *, struct inode **);
-int vfs_removexattr(struct dentry *, const char *);
+int __vfs_setxattr(struct user_namespace *, struct dentry *, struct inode *,
+		   const char *, const void *, size_t, int);
+int __vfs_setxattr_noperm(struct user_namespace *, struct dentry *,
+			  const char *, const void *, size_t, int);
+int __vfs_setxattr_locked(struct user_namespace *, struct dentry *,
+			  const char *, const void *, size_t, int,
+			  struct inode **);
+int vfs_setxattr(struct user_namespace *, struct dentry *, const char *, const void *, size_t, int);
+int __vfs_removexattr(struct user_namespace *, struct dentry *, const char *);
+int __vfs_removexattr_locked(struct user_namespace *, struct dentry *, const char *, struct inode **);
+int vfs_removexattr(struct user_namespace *, struct dentry *, const char *);
 
 ssize_t generic_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size);
-ssize_t vfs_getxattr_alloc(struct dentry *dentry, const char *name,
+ssize_t vfs_getxattr_alloc(struct user_namespace *mnt_userns,
+			   struct dentry *dentry, const char *name,
 			   char **xattr_value, size_t size, gfp_t flags);
 
 int xattr_supported_namespace(struct inode *inode, const char *prefix);
security/apparmor/domain.c +2 -2
index f919ebd042fd..16f184bc48de 100644
@@ -324,8 +324,8 @@ static int aa_xattrs_match(const struct linux_binprm *bprm,
 	d = bprm->file->f_path.dentry;
 
 	for (i = 0; i < profile->xattr_count; i++) {
-		size = vfs_getxattr_alloc(d, profile->xattrs[i], &value,
-					  value_size, GFP_KERNEL);
+		size = vfs_getxattr_alloc(&init_user_ns, d, profile->xattrs[i],
+					  &value, value_size, GFP_KERNEL);
 		if (size >= 0) {
 			u32 perm;
 
security/commoncap.c +3 -3
index 9a0b97a099af..b6f89ae5c88a 100644
@@ -313,7 +313,7 @@ int cap_inode_killpriv(struct dentry *dentry)
 {
 	int error;
 
-	error = __vfs_removexattr(dentry, XATTR_NAME_CAPS);
+	error = __vfs_removexattr(&init_user_ns, dentry, XATTR_NAME_CAPS);
 	if (error == -EOPNOTSUPP)
 		error = 0;
 	return error;
@@ -386,8 +386,8 @@ int cap_inode_getsecurity(struct inode *inode, const char *name, void **buffer,
 		return -EINVAL;
 
 	size = sizeof(struct vfs_ns_cap_data);
-	ret = (int) vfs_getxattr_alloc(dentry, XATTR_NAME_CAPS,
-				 &tmpbuf, size, GFP_NOFS);
+	ret = (int)vfs_getxattr_alloc(&init_user_ns, dentry, XATTR_NAME_CAPS,
+				      &tmpbuf, size, GFP_NOFS);
 	dput(dentry);
 
 	if (ret < 0)
security/integrity/evm/evm_crypto.c +6 -5
index 168c3b78ac47..f720f78cbbb1 100644
@@ -222,7 +222,7 @@ static int evm_calc_hmac_or_hash(struct dentry *dentry,
 				ima_present = true;
 			continue;
 		}
-		size = vfs_getxattr_alloc(dentry, xattr->name,
+		size = vfs_getxattr_alloc(&init_user_ns, dentry, xattr->name,
 					  &xattr_value, xattr_size, GFP_NOFS);
 		if (size == -ENOMEM) {
 			error = -ENOMEM;
@@ -275,8 +275,8 @@ static int evm_is_immutable(struct dentry *dentry, struct inode *inode)
 		return 1;
 
 	/* Do this the hard way */
-	rc = vfs_getxattr_alloc(dentry, XATTR_NAME_EVM, (char **)&xattr_data, 0,
-				GFP_NOFS);
+	rc = vfs_getxattr_alloc(&init_user_ns, dentry, XATTR_NAME_EVM,
+				(char **)&xattr_data, 0, GFP_NOFS);
 	if (rc <= 0) {
 		if (rc == -ENODATA)
 			return 0;
@@ -319,11 +319,12 @@ int evm_update_evmxattr(struct dentry *dentry, const char *xattr_name,
 			   xattr_value_len, &data);
 	if (rc == 0) {
 		data.hdr.xattr.sha1.type = EVM_XATTR_HMAC;
-		rc = __vfs_setxattr_noperm(dentry, XATTR_NAME_EVM,
+		rc = __vfs_setxattr_noperm(&init_user_ns, dentry,
+					   XATTR_NAME_EVM,
 					   &data.hdr.xattr.data[1],
 					   SHA1_DIGEST_SIZE + 1, 0);
 	} else if (rc == -ENODATA && (inode->i_opflags & IOP_XATTR)) {
-		rc = __vfs_removexattr(dentry, XATTR_NAME_EVM);
+		rc = __vfs_removexattr(&init_user_ns, dentry, XATTR_NAME_EVM);
 	}
 	return rc;
 }
security/integrity/evm/evm_main.c +2 -2
index 76d19146d74b..0de367aaa2d3 100644
@@ -146,8 +146,8 @@ static enum integrity_status evm_verify_hmac(struct dentry *dentry,
 	/* if status is not PASS, try to check again - against -ENOMEM */
 
 	/* first need to know the sig type */
-	rc = vfs_getxattr_alloc(dentry, XATTR_NAME_EVM, (char **)&xattr_data, 0,
-				GFP_NOFS);
+	rc = vfs_getxattr_alloc(&init_user_ns, dentry, XATTR_NAME_EVM,
+				(char **)&xattr_data, 0, GFP_NOFS);
 	if (rc <= 0) {
 		evm_status = INTEGRITY_FAIL;
 		if (rc == -ENODATA) {
security/integrity/ima/ima_appraise.c +4 -4
index 8361941ee0a1..70b643c41c6b 100644
@@ -94,7 +94,7 @@ static int ima_fix_xattr(struct dentry *dentry,
 		iint->ima_hash->xattr.ng.type = IMA_XATTR_DIGEST_NG;
 		iint->ima_hash->xattr.ng.algo = algo;
 	}
-	rc = __vfs_setxattr_noperm(dentry, XATTR_NAME_IMA,
+	rc = __vfs_setxattr_noperm(&init_user_ns, dentry, XATTR_NAME_IMA,
 				   &iint->ima_hash->xattr.data[offset],
 				   (sizeof(iint->ima_hash->xattr) - offset) +
 				   iint->ima_hash->length, 0);
@@ -215,8 +215,8 @@ int ima_read_xattr(struct dentry *dentry,
 {
 	ssize_t ret;
 
-	ret = vfs_getxattr_alloc(dentry, XATTR_NAME_IMA, (char **)xattr_value,
-				 0, GFP_NOFS);
+	ret = vfs_getxattr_alloc(&init_user_ns, dentry, XATTR_NAME_IMA,
+				 (char **)xattr_value, 0, GFP_NOFS);
 	if (ret == -EOPNOTSUPP)
 		ret = 0;
 	return ret;
@@ -520,7 +520,7 @@ void ima_inode_post_setattr(struct dentry *dentry)
 
 	action = ima_must_appraise(inode, MAY_ACCESS, POST_SETATTR);
 	if (!action)
-		__vfs_removexattr(dentry, XATTR_NAME_IMA);
+		__vfs_removexattr(&init_user_ns, dentry, XATTR_NAME_IMA);
 	iint = integrity_iint_find(inode);
 	if (iint) {
 		set_bit(IMA_CHANGE_ATTR, &iint->atomic_flags);
security/selinux/hooks.c +1 -1
index 9d6d3da2caf2..98756f53a929 100644
@@ -6526,7 +6526,7 @@ static int selinux_inode_notifysecctx(struct inode *inode, void *ctx, u32 ctxlen
  */
 static int selinux_inode_setsecctx(struct dentry *dentry, void *ctx, u32 ctxlen)
 {
-	return __vfs_setxattr_noperm(dentry, XATTR_NAME_SELINUX, ctx, ctxlen, 0);
+	return __vfs_setxattr_noperm(&init_user_ns, dentry, XATTR_NAME_SELINUX, ctx, ctxlen, 0);
 }
 
 static int selinux_inode_getsecctx(struct inode *inode, void **ctx, u32 *ctxlen)
security/smack/smack_lsm.c +2 -2
index f69c3dd9a0c6..8b0b90c6dda2 100644
@@ -3425,7 +3425,7 @@ static void smack_d_instantiate(struct dentry *opt_dentry, struct inode *inode)
 			 */
 			if (isp->smk_flags & SMK_INODE_CHANGED) {
 				isp->smk_flags &= ~SMK_INODE_CHANGED;
-				rc = __vfs_setxattr(dp, inode,
+				rc = __vfs_setxattr(&init_user_ns, dp, inode,
 					XATTR_NAME_SMACKTRANSMUTE,
 					TRANS_TRUE, TRANS_TRUE_SIZE,
 					0);
@@ -4602,7 +4602,7 @@ static int smack_inode_notifysecctx(struct inode *inode, void *ctx, u32 ctxlen)
 
 static int smack_inode_setsecctx(struct dentry *dentry, void *ctx, u32 ctxlen)
 {
-	return __vfs_setxattr_noperm(dentry, XATTR_NAME_SMACK, ctx, ctxlen, 0);
+	return __vfs_setxattr_noperm(&init_user_ns, dentry, XATTR_NAME_SMACK, ctx, ctxlen, 0);
 }
 
 static int smack_inode_getsecctx(struct inode *inode, void **ctx, u32 *ctxlen)
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
#27 lore raw

PATCH v5 17/42 commoncap: handle idmapped mounts

In reply to Christian Brauner

When interacting with user namespace and non-user namespace aware filesystem capabilities the vfs will perform various security checks to determine whether or not the filesystem capabilities can be used by the caller (e.g. during exec), or even whether they need to be removed. The main infrastructure for this resides in the capability codepaths but they are called through the LSM security infrastructure even though they are not technically an LSM or optional. This extends the existing security hooks security_inode_removexattr(), security_inode_killpriv(), security_inode_getsecurity() to pass down the mount's user namespace and makes them aware of idmapped mounts.

In order to actually get filesystem capabilities from disk the capability infrastructure exposes the get_vfs_caps_from_disk() helper. For user namespace aware filesystem capabilities a root uid is stored alongside the capabilities.

In order to determine whether the caller can make use of the filesystem capability or whether it needs to be ignored it is translated according to the superblock's user namespace. If it can be translated to uid 0 according to that id mapping the caller can use the filesystem capabilities stored on disk. If we are accessing the inode that holds the filesystem capabilities through an idmapped mount we need to map the root uid according to the mount's user namespace.

Afterwards the checks are identical to non-idmapped mounts. Reading filesystem caps from disk enforces that the root uid associated with the filesystem capability must have a mapping in the superblock's user namespace and that the caller is either in the same user namespace or is a descendant of the superblock's user namespace. For filesystems that are mountable inside user namespace the caller can just mount the filesystem and won't usually need to idmap it. If they do want to idmap it they can create an idmapped mount and mark it with a user namespace they created and which is thus a descendant of s_user_ns. For filesystems that are not mountable inside user namespaces the descendant rule is trivially true because the s_user_ns will be the initial user namespace.

If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Don't pollute the vfs with additional helpers simply extend the existing helpers with an additional argument and switch all callers.

/* v3 */
- Paul Moore <[email protected]>:
- All audit logging is currently based on init_user_ns and will remain so until other work unrelated to this patchset has landed so always pass &init_user_ns in audit.

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

- Christoph Hellwig <[email protected]>:
- Use new file_userns_helper().

---
 fs/attr.c                     |  2 +-
 fs/xattr.c                    | 14 ++++----
 include/linux/capability.h    |  4 ++-
 include/linux/lsm_hook_defs.h | 15 +++++----
 include/linux/lsm_hooks.h     |  1 +
 include/linux/security.h      | 46 ++++++++++++++++---------
 kernel/auditsc.c              |  5 +--
 security/commoncap.c          | 63 ++++++++++++++++++++++++++---------
 security/security.c           | 25 ++++++++------
 security/selinux/hooks.c      | 20 ++++++-----
 security/smack/smack_lsm.c    | 14 ++++----
 11 files changed, 137 insertions(+), 72 deletions(-)
fs/attr.c +1 -1
index 939402dea6f7..3ee0b1b0f207 100644
@@ -143,7 +143,7 @@ int setattr_prepare(struct user_namespace *mnt_userns, struct dentry *dentry,
 	if (ia_valid & ATTR_KILL_PRIV) {
 		int error;
 
-		error = security_inode_killpriv(dentry);
+		error = security_inode_killpriv(mnt_userns, dentry);
 		if (error)
 			return error;
 	}
fs/xattr.c +7 -7
index 1c7d1d4e5f32..03afd803e50b 100644
@@ -259,7 +259,7 @@ __vfs_setxattr_locked(struct user_namespace *mnt_userns, struct dentry *dentry,
 	if (error)
 		return error;
 
-	error = security_inode_setxattr(dentry, name, value, size, flags);
+	error = security_inode_setxattr(mnt_userns, dentry, name, value, size, flags);
 	if (error)
 		goto out;
 
@@ -309,18 +309,18 @@ vfs_setxattr(struct user_namespace *mnt_userns, struct dentry *dentry,
 EXPORT_SYMBOL_GPL(vfs_setxattr);
 
 static ssize_t
-xattr_getsecurity(struct inode *inode, const char *name, void *value,
-			size_t size)
+xattr_getsecurity(struct user_namespace *mnt_userns, struct inode *inode,
+		  const char *name, void *value, size_t size)
 {
 	void *buffer = NULL;
 	ssize_t len;
 
 	if (!value || !size) {
-		len = security_inode_getsecurity(inode, name, &buffer, false);
+		len = security_inode_getsecurity(mnt_userns, inode, name, &buffer, false);
 		goto out_noalloc;
 	}
 
-	len = security_inode_getsecurity(inode, name, &buffer, true);
+	len = security_inode_getsecurity(mnt_userns, inode, name, &buffer, true);
 	if (len < 0)
 		return len;
 	if (size < len) {
@@ -410,7 +410,7 @@ vfs_getxattr(struct user_namespace *mnt_userns, struct dentry *dentry,
 	if (!strncmp(name, XATTR_SECURITY_PREFIX,
 				XATTR_SECURITY_PREFIX_LEN)) {
 		const char *suffix = name + XATTR_SECURITY_PREFIX_LEN;
-		int ret = xattr_getsecurity(inode, suffix, value, size);
+		int ret = xattr_getsecurity(mnt_userns, inode, suffix, value, size);
 		/*
 		 * Only overwrite the return value if a security module
 		 * is actually active.
@@ -479,7 +479,7 @@ __vfs_removexattr_locked(struct user_namespace *mnt_userns, struct dentry *dentr
 	if (error)
 		return error;
 
-	error = security_inode_removexattr(dentry, name);
+	error = security_inode_removexattr(mnt_userns, dentry, name);
 	if (error)
 		goto out;
 
include/linux/capability.h +3 -1
index 575436b977bc..95421b941ffc 100644
@@ -271,7 +271,9 @@ static inline bool checkpoint_restore_ns_capable(struct user_namespace *ns)
 }
 
 /* audit system wants to get cap info from files as well */
-extern int get_vfs_caps_from_disk(const struct dentry *dentry, struct cpu_vfs_cap_data *cpu_caps);
+extern int get_vfs_caps_from_disk(struct user_namespace *mnt_userns,
+				  const struct dentry *dentry,
+				  struct cpu_vfs_cap_data *cpu_caps);
 
 extern int cap_convert_nscap(struct user_namespace *mnt_userns,
 			     struct dentry *dentry, const void **ivalue,
include/linux/lsm_hook_defs.h +9 -6
index 7aaa753b8608..df4cdad6681e 100644
@@ -133,17 +133,20 @@ LSM_HOOK(int, 0, inode_follow_link, struct dentry *dentry, struct inode *inode,
 LSM_HOOK(int, 0, inode_permission, struct inode *inode, int mask)
 LSM_HOOK(int, 0, inode_setattr, struct dentry *dentry, struct iattr *attr)
 LSM_HOOK(int, 0, inode_getattr, const struct path *path)
-LSM_HOOK(int, 0, inode_setxattr, struct dentry *dentry, const char *name,
-	 const void *value, size_t size, int flags)
+LSM_HOOK(int, 0, inode_setxattr, struct user_namespace *mnt_userns,
+	 struct dentry *dentry, const char *name, const void *value,
+	 size_t size, int flags)
 LSM_HOOK(void, LSM_RET_VOID, inode_post_setxattr, struct dentry *dentry,
 	 const char *name, const void *value, size_t size, int flags)
 LSM_HOOK(int, 0, inode_getxattr, struct dentry *dentry, const char *name)
 LSM_HOOK(int, 0, inode_listxattr, struct dentry *dentry)
-LSM_HOOK(int, 0, inode_removexattr, struct dentry *dentry, const char *name)
+LSM_HOOK(int, 0, inode_removexattr, struct user_namespace *mnt_userns,
+	 struct dentry *dentry, const char *name)
 LSM_HOOK(int, 0, inode_need_killpriv, struct dentry *dentry)
-LSM_HOOK(int, 0, inode_killpriv, struct dentry *dentry)
-LSM_HOOK(int, -EOPNOTSUPP, inode_getsecurity, struct inode *inode,
-	 const char *name, void **buffer, bool alloc)
+LSM_HOOK(int, 0, inode_killpriv, struct user_namespace *mnt_userns,
+	 struct dentry *dentry)
+LSM_HOOK(int, -EOPNOTSUPP, inode_getsecurity, struct user_namespace *mnt_userns,
+	 struct inode *inode, const char *name, void **buffer, bool alloc)
 LSM_HOOK(int, -EOPNOTSUPP, inode_setsecurity, struct inode *inode,
 	 const char *name, const void *value, size_t size, int flags)
 LSM_HOOK(int, 0, inode_listsecurity, struct inode *inode, char *buffer,
include/linux/lsm_hooks.h +1 -0
index a19adef1f088..98a5e263aabb 100644
@@ -444,6 +444,7 @@
  * @inode_killpriv:
  *	The setuid bit is being removed.  Remove similar security labels.
  *	Called with the dentry->d_inode->i_mutex held.
+ *	@mnt_userns: user namespace of the mount
  *	@dentry is the dentry being changed.
  *	Return 0 on success.  If error is returned, then the operation
  *	causing setuid bit removal is failed.
include/linux/security.h +30 -16

88 lines not on this page · raw message

kernel/auditsc.c +3 -2
index ce8c9e2279ba..fdfdd71405ff 100644
@@ -1930,7 +1930,7 @@ static inline int audit_copy_fcaps(struct audit_names *name,
 	if (!dentry)
 		return 0;
 
-	rc = get_vfs_caps_from_disk(dentry, &caps);
+	rc = get_vfs_caps_from_disk(&init_user_ns, dentry, &caps);
 	if (rc)
 		return rc;
 
@@ -2481,7 +2481,8 @@ int __audit_log_bprm_fcaps(struct linux_binprm *bprm,
 	ax->d.next = context->aux;
 	context->aux = (void *)ax;
 
-	get_vfs_caps_from_disk(bprm->file->f_path.dentry, &vcaps);
+	get_vfs_caps_from_disk(&init_user_ns,
+			       bprm->file->f_path.dentry, &vcaps);
 
 	ax->fcap.permitted = vcaps.permitted;
 	ax->fcap.inheritable = vcaps.inheritable;
security/commoncap.c +48 -15

140 lines not on this page · raw message

security/security.c +15 -10
index 7b09cfbae94f..698a9f17bad7 100644
@@ -1280,7 +1280,8 @@ int security_inode_getattr(const struct path *path)
 	return call_int_hook(inode_getattr, 0, path);
 }
 
-int security_inode_setxattr(struct dentry *dentry, const char *name,
+int security_inode_setxattr(struct user_namespace *mnt_userns,
+			    struct dentry *dentry, const char *name,
 			    const void *value, size_t size, int flags)
 {
 	int ret;
@@ -1291,8 +1292,8 @@ int security_inode_setxattr(struct dentry *dentry, const char *name,
 	 * SELinux and Smack integrate the cap call,
 	 * so assume that all LSMs supplying this call do so.
 	 */
-	ret = call_int_hook(inode_setxattr, 1, dentry, name, value, size,
-				flags);
+	ret = call_int_hook(inode_setxattr, 1, mnt_userns, dentry, name, value,
+			    size, flags);
 
 	if (ret == 1)
 		ret = cap_inode_setxattr(dentry, name, value, size, flags);
@@ -1327,7 +1328,8 @@ int security_inode_listxattr(struct dentry *dentry)
 	return call_int_hook(inode_listxattr, 0, dentry);
 }
 
-int security_inode_removexattr(struct dentry *dentry, const char *name)
+int security_inode_removexattr(struct user_namespace *mnt_userns,
+			       struct dentry *dentry, const char *name)
 {
 	int ret;
 
@@ -1337,9 +1339,9 @@ int security_inode_removexattr(struct dentry *dentry, const char *name)
 	 * SELinux and Smack integrate the cap call,
 	 * so assume that all LSMs supplying this call do so.
 	 */
-	ret = call_int_hook(inode_removexattr, 1, dentry, name);
+	ret = call_int_hook(inode_removexattr, 1, mnt_userns, dentry, name);
 	if (ret == 1)
-		ret = cap_inode_removexattr(dentry, name);
+		ret = cap_inode_removexattr(mnt_userns, dentry, name);
 	if (ret)
 		return ret;
 	ret = ima_inode_removexattr(dentry, name);
@@ -1353,12 +1355,15 @@ int security_inode_need_killpriv(struct dentry *dentry)
 	return call_int_hook(inode_need_killpriv, 0, dentry);
 }
 
-int security_inode_killpriv(struct dentry *dentry)
+int security_inode_killpriv(struct user_namespace *mnt_userns,
+			    struct dentry *dentry)
 {
-	return call_int_hook(inode_killpriv, 0, dentry);
+	return call_int_hook(inode_killpriv, 0, mnt_userns, dentry);
 }
 
-int security_inode_getsecurity(struct inode *inode, const char *name, void **buffer, bool alloc)
+int security_inode_getsecurity(struct user_namespace *mnt_userns,
+			       struct inode *inode, const char *name,
+			       void **buffer, bool alloc)
 {
 	struct security_hook_list *hp;
 	int rc;
@@ -1369,7 +1374,7 @@ int security_inode_getsecurity(struct inode *inode, const char *name, void **buf
 	 * Only one module will provide an attribute with a given name.
 	 */
 	hlist_for_each_entry(hp, &security_hook_heads.inode_getsecurity, list) {
-		rc = hp->hook.inode_getsecurity(inode, name, buffer, alloc);
+		rc = hp->hook.inode_getsecurity(mnt_userns, inode, name, buffer, alloc);
 		if (rc != LSM_RET_DEFAULT(inode_getsecurity))
 			return rc;
 	}
security/selinux/hooks.c +12 -8
index 98756f53a929..6408d763b652 100644
@@ -3119,7 +3119,8 @@ static bool has_cap_mac_admin(bool audit)
 	return true;
 }
 
-static int selinux_inode_setxattr(struct dentry *dentry, const char *name,
+static int selinux_inode_setxattr(struct user_namespace *mnt_userns,
+				  struct dentry *dentry, const char *name,
 				  const void *value, size_t size, int flags)
 {
 	struct inode *inode = d_backing_inode(dentry);
@@ -3140,13 +3141,13 @@ static int selinux_inode_setxattr(struct dentry *dentry, const char *name,
 	}
 
 	if (!selinux_initialized(&selinux_state))
-		return (inode_owner_or_capable(&init_user_ns, inode) ? 0 : -EPERM);
+		return (inode_owner_or_capable(mnt_userns, inode) ? 0 : -EPERM);
 
 	sbsec = inode->i_sb->s_security;
 	if (!(sbsec->flags & SBLABEL_MNT))
 		return -EOPNOTSUPP;
 
-	if (!inode_owner_or_capable(&init_user_ns, inode))
+	if (!inode_owner_or_capable(mnt_userns, inode))
 		return -EPERM;
 
 	ad.type = LSM_AUDIT_DATA_DENTRY;
@@ -3267,10 +3268,11 @@ static int selinux_inode_listxattr(struct dentry *dentry)
 	return dentry_has_perm(cred, dentry, FILE__GETATTR);
 }
 
-static int selinux_inode_removexattr(struct dentry *dentry, const char *name)
+static int selinux_inode_removexattr(struct user_namespace *mnt_userns,
+				     struct dentry *dentry, const char *name)
 {
 	if (strcmp(name, XATTR_NAME_SELINUX)) {
-		int rc = cap_inode_removexattr(dentry, name);
+		int rc = cap_inode_removexattr(mnt_userns, dentry, name);
 		if (rc)
 			return rc;
 
@@ -3336,7 +3338,9 @@ static int selinux_path_notify(const struct path *path, u64 mask,
  *
  * Permission check is handled by selinux_inode_getxattr hook.
  */
-static int selinux_inode_getsecurity(struct inode *inode, const char *name, void **buffer, bool alloc)
+static int selinux_inode_getsecurity(struct user_namespace *mnt_userns,
+				     struct inode *inode, const char *name,
+				     void **buffer, bool alloc)
 {
 	u32 size;
 	int error;
@@ -6532,8 +6536,8 @@ static int selinux_inode_setsecctx(struct dentry *dentry, void *ctx, u32 ctxlen)
 static int selinux_inode_getsecctx(struct inode *inode, void **ctx, u32 *ctxlen)
 {
 	int len = 0;
-	len = selinux_inode_getsecurity(inode, XATTR_SELINUX_SUFFIX,
-						ctx, true);
+	len = selinux_inode_getsecurity(&init_user_ns, inode,
+					XATTR_SELINUX_SUFFIX, ctx, true);
 	if (len < 0)
 		return len;
 	*ctxlen = len;
security/smack/smack_lsm.c +8 -6
index 8b0b90c6dda2..e2ea446defcc 100644
@@ -1240,7 +1240,8 @@ static int smack_inode_getattr(const struct path *path)
  *
  * Returns 0 if access is permitted, an error code otherwise
  */
-static int smack_inode_setxattr(struct dentry *dentry, const char *name,
+static int smack_inode_setxattr(struct user_namespace *mnt_userns,
+				struct dentry *dentry, const char *name,
 				const void *value, size_t size, int flags)
 {
 	struct smk_audit_info ad;
@@ -1362,7 +1363,8 @@ static int smack_inode_getxattr(struct dentry *dentry, const char *name)
  *
  * Returns 0 if access is permitted, an error code otherwise
  */
-static int smack_inode_removexattr(struct dentry *dentry, const char *name)
+static int smack_inode_removexattr(struct user_namespace *mnt_userns,
+				   struct dentry *dentry, const char *name)
 {
 	struct inode_smack *isp;
 	struct smk_audit_info ad;
@@ -1377,7 +1379,7 @@ static int smack_inode_removexattr(struct dentry *dentry, const char *name)
 		if (!smack_privileged(CAP_MAC_ADMIN))
 			rc = -EPERM;
 	} else
-		rc = cap_inode_removexattr(dentry, name);
+		rc = cap_inode_removexattr(mnt_userns, dentry, name);
 
 	if (rc != 0)
 		return rc;
@@ -1420,9 +1422,9 @@ static int smack_inode_removexattr(struct dentry *dentry, const char *name)
  *
  * Returns the size of the attribute or an error code
  */
-static int smack_inode_getsecurity(struct inode *inode,
-				   const char *name, void **buffer,
-				   bool alloc)
+static int smack_inode_getsecurity(struct user_namespace *mnt_userns,
+				   struct inode *inode, const char *name,
+				   void **buffer, bool alloc)
 {
 	struct socket_smack *ssp;
 	struct socket *sock;
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
#29 lore raw

PATCH v5 18/42 stat: handle idmapped mounts

In reply to Christian Brauner

The generic_fillattr() helper fills in the basic attributes associated with an inode. Enable it to handle idmapped mounts. If the inode is accessed through an idmapped mount we need to map it according to the mount's user namespace before we store the uid and gid.
If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Don't pollute the vfs with additional helpers simply extend the existing helpers with an additional argument and switch all callers.

/* v3 */
unchanged

/* v4 */
unchanged

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/9p/vfs_inode.c      |  4 ++--
 fs/9p/vfs_inode_dotl.c |  4 ++--
 fs/afs/inode.c         |  2 +-
 fs/btrfs/inode.c       |  2 +-
 fs/ceph/inode.c        |  2 +-
 fs/cifs/inode.c        |  2 +-
 fs/coda/inode.c        |  2 +-
 fs/ecryptfs/inode.c    |  4 ++--
 fs/erofs/inode.c       |  2 +-
 fs/exfat/file.c        |  2 +-
 fs/ext2/inode.c        |  2 +-
 fs/ext4/inode.c        |  2 +-
 fs/f2fs/file.c         |  2 +-
 fs/fat/file.c          |  2 +-
 fs/fuse/dir.c          |  2 +-
 fs/gfs2/inode.c        |  2 +-
 fs/hfsplus/inode.c     |  2 +-
 fs/kernfs/inode.c      |  2 +-
 fs/libfs.c             |  4 ++--
 fs/minix/inode.c       |  2 +-
 fs/nfs/inode.c         |  2 +-
 fs/nfs/namespace.c     |  2 +-
 fs/ocfs2/file.c        |  2 +-
 fs/orangefs/inode.c    |  2 +-
 fs/proc/base.c         |  4 ++--
 fs/proc/generic.c      |  2 +-
 fs/proc/proc_net.c     |  2 +-
 fs/proc/proc_sysctl.c  |  2 +-
 fs/proc/root.c         |  2 +-
 fs/stat.c              | 20 ++++++++++++++------
 fs/sysv/itree.c        |  2 +-
 fs/ubifs/dir.c         |  2 +-
 fs/udf/symlink.c       |  2 +-
 fs/vboxsf/utils.c      |  2 +-
 include/linux/fs.h     |  2 +-
 mm/shmem.c             |  2 +-
 36 files changed, 54 insertions(+), 46 deletions(-)
fs/9p/vfs_inode.c +2 -2
index 9c3ff6e9ab82..c21b146c8d91 100644
@@ -1027,7 +1027,7 @@ v9fs_vfs_getattr(const struct path *path, struct kstat *stat,
 	p9_debug(P9_DEBUG_VFS, "dentry: %p\n", dentry);
 	v9ses = v9fs_dentry2v9ses(dentry);
 	if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) {
-		generic_fillattr(d_inode(dentry), stat);
+		generic_fillattr(&init_user_ns, d_inode(dentry), stat);
 		return 0;
 	}
 	fid = v9fs_fid_lookup(dentry);
@@ -1040,7 +1040,7 @@ v9fs_vfs_getattr(const struct path *path, struct kstat *stat,
 		return PTR_ERR(st);
 
 	v9fs_stat2inode(st, d_inode(dentry), dentry->d_sb, 0);
-	generic_fillattr(d_inode(dentry), stat);
+	generic_fillattr(&init_user_ns, d_inode(dentry), stat);
 
 	p9stat_free(st);
 	kfree(st);
fs/9p/vfs_inode_dotl.c +2 -2
index 302553101fcb..984f28315d2a 100644
@@ -468,7 +468,7 @@ v9fs_vfs_getattr_dotl(const struct path *path, struct kstat *stat,
 	p9_debug(P9_DEBUG_VFS, "dentry: %p\n", dentry);
 	v9ses = v9fs_dentry2v9ses(dentry);
 	if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) {
-		generic_fillattr(d_inode(dentry), stat);
+		generic_fillattr(&init_user_ns, d_inode(dentry), stat);
 		return 0;
 	}
 	fid = v9fs_fid_lookup(dentry);
@@ -485,7 +485,7 @@ v9fs_vfs_getattr_dotl(const struct path *path, struct kstat *stat,
 		return PTR_ERR(st);
 
 	v9fs_stat2inode_dotl(st, d_inode(dentry), 0);
-	generic_fillattr(d_inode(dentry), stat);
+	generic_fillattr(&init_user_ns, d_inode(dentry), stat);
 	/* Change block size to what the server returned */
 	stat->blksize = st->st_blksize;
 
fs/afs/inode.c +1 -1
index b0d7b892090d..795ee5cb3817 100644
@@ -745,7 +745,7 @@ int afs_getattr(const struct path *path, struct kstat *stat,
 
 	do {
 		read_seqbegin_or_lock(&vnode->cb_lock, &seq);
-		generic_fillattr(inode, stat);
+		generic_fillattr(&init_user_ns, inode, stat);
 		if (test_bit(AFS_VNODE_SILLY_DELETED, &vnode->flags) &&
 		    stat->nlink > 0)
 			stat->nlink -= 1;
fs/btrfs/inode.c +1 -1
index 97bbf80ee12d..c8045c9d51d9 100644
@@ -8841,7 +8841,7 @@ static int btrfs_getattr(const struct path *path, struct kstat *stat,
 				  STATX_ATTR_IMMUTABLE |
 				  STATX_ATTR_NODUMP);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	stat->dev = BTRFS_I(inode)->root->anon_dev;
 
 	spin_lock(&BTRFS_I(inode)->lock);
fs/ceph/inode.c +1 -1
index 145e26a4ddbb..179a2bb88538 100644
@@ -2385,7 +2385,7 @@ int ceph_getattr(const struct path *path, struct kstat *stat,
 			return err;
 	}
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	stat->ino = ceph_present_inode(inode);
 
 	/*
fs/cifs/inode.c +1 -1
index 27554f71f744..374abce7efaf 100644
@@ -2408,7 +2408,7 @@ int cifs_getattr(const struct path *path, struct kstat *stat,
 			return rc;
 	}
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	stat->blksize = cifs_sb->ctx->bsize;
 	stat->ino = CIFS_I(inode)->uniqueid;
 
fs/coda/inode.c +1 -1
index b1c70e2b9b1e..4d113e191cb8 100644
@@ -256,7 +256,7 @@ int coda_getattr(const struct path *path, struct kstat *stat,
 {
 	int err = coda_revalidate_inode(d_inode(path->dentry));
 	if (!err)
-		generic_fillattr(d_inode(path->dentry), stat);
+		generic_fillattr(&init_user_ns, d_inode(path->dentry), stat);
 	return err;
 }
 
fs/ecryptfs/inode.c +2 -2
index 7de9fda25dad..0a218d7c3023 100644
@@ -975,7 +975,7 @@ static int ecryptfs_getattr_link(const struct path *path, struct kstat *stat,
 
 	mount_crypt_stat = &ecryptfs_superblock_to_private(
 						dentry->d_sb)->mount_crypt_stat;
-	generic_fillattr(d_inode(dentry), stat);
+	generic_fillattr(&init_user_ns, d_inode(dentry), stat);
 	if (mount_crypt_stat->flags & ECRYPTFS_GLOBAL_ENCRYPT_FILENAMES) {
 		char *target;
 		size_t targetsiz;
@@ -1003,7 +1003,7 @@ static int ecryptfs_getattr(const struct path *path, struct kstat *stat,
 	if (!rc) {
 		fsstack_copy_attr_all(d_inode(dentry),
 				      ecryptfs_inode_to_lower(d_inode(dentry)));
-		generic_fillattr(d_inode(dentry), stat);
+		generic_fillattr(&init_user_ns, d_inode(dentry), stat);
 		stat->blocks = lower_stat.blocks;
 	}
 	return rc;
fs/erofs/inode.c +1 -1
index 3e21c0e8adae..083818063ac6 100644
@@ -343,7 +343,7 @@ int erofs_getattr(const struct path *path, struct kstat *stat,
 	stat->attributes_mask |= (STATX_ATTR_COMPRESSED |
 				  STATX_ATTR_IMMUTABLE);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	return 0;
 }
 
fs/exfat/file.c +1 -1
index ace35aa8e64b..e9705b3295d3 100644
@@ -273,7 +273,7 @@ int exfat_getattr(const struct path *path, struct kstat *stat,
 	struct inode *inode = d_backing_inode(path->dentry);
 	struct exfat_inode_info *ei = EXFAT_I(inode);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	exfat_truncate_atime(&stat->atime);
 	stat->result_mask |= STATX_BTIME;
 	stat->btime.tv_sec = ei->i_crtime.tv_sec;
fs/ext2/inode.c +1 -1
index 9de813635d8d..3d8acafca8ce 100644
@@ -1660,7 +1660,7 @@ int ext2_getattr(const struct path *path, struct kstat *stat,
 			STATX_ATTR_IMMUTABLE |
 			STATX_ATTR_NODUMP);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	return 0;
 }
 
fs/ext4/inode.c +1 -1
index 55990b74b9e4..69caa7b939d3 100644
@@ -5567,7 +5567,7 @@ int ext4_getattr(const struct path *path, struct kstat *stat,
 				  STATX_ATTR_NODUMP |
 				  STATX_ATTR_VERITY);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	return 0;
 }
 
fs/f2fs/file.c +1 -1
index ba18b7e28d51..5020c740f068 100644
@@ -820,7 +820,7 @@ int f2fs_getattr(const struct path *path, struct kstat *stat,
 				  STATX_ATTR_NODUMP |
 				  STATX_ATTR_VERITY);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 
 	/* we need to show initial sectors used for inline_data/dentries */
 	if ((S_ISREG(inode->i_mode) && f2fs_has_inline_data(inode)) ||
fs/fat/file.c +1 -1
index 805b501467e9..f7e04f533d31 100644
@@ -398,7 +398,7 @@ int fat_getattr(const struct path *path, struct kstat *stat,
 		u32 request_mask, unsigned int flags)
 {
 	struct inode *inode = d_inode(path->dentry);
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	stat->blksize = MSDOS_SB(inode->i_sb)->cluster_size;
 
 	if (MSDOS_SB(inode->i_sb)->options.nfs == FAT_NFS_NOSTALE_RO) {
fs/fuse/dir.c +1 -1
index eb648ed19312..f2f252e3da7f 100644
@@ -1087,7 +1087,7 @@ static int fuse_update_get_attr(struct inode *inode, struct file *file,
 		forget_all_cached_acls(inode);
 		err = fuse_do_getattr(inode, stat, file);
 	} else if (stat) {
-		generic_fillattr(inode, stat);
+		generic_fillattr(&init_user_ns, inode, stat);
 		stat->mode = fi->orig_i_mode;
 		stat->ino = fi->orig_ino;
 	}
fs/gfs2/inode.c +1 -1
index aef2f7587392..9b773b2b9d32 100644
@@ -2049,7 +2049,7 @@ static int gfs2_getattr(const struct path *path, struct kstat *stat,
 				  STATX_ATTR_IMMUTABLE |
 				  STATX_ATTR_NODUMP);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 
 	if (gfs2_holder_initialized(&gh))
 		gfs2_glock_dq_uninit(&gh);
fs/hfsplus/inode.c +1 -1
index ffa137f8234e..642e067d8fe8 100644
@@ -286,7 +286,7 @@ int hfsplus_getattr(const struct path *path, struct kstat *stat,
 	stat->attributes_mask |= STATX_ATTR_APPEND | STATX_ATTR_IMMUTABLE |
 				 STATX_ATTR_NODUMP;
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	return 0;
 }
 
fs/kernfs/inode.c +1 -1
index 7e44052b42e1..032d3d7546d8 100644
@@ -193,7 +193,7 @@ int kernfs_iop_getattr(const struct path *path, struct kstat *stat,
 	kernfs_refresh_inode(kn, inode);
 	mutex_unlock(&kernfs_mutex);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	return 0;
 }
 
fs/libfs.c +2 -2
index a73fe109403c..508e9ea8e6f3 100644
@@ -31,7 +31,7 @@ int simple_getattr(const struct path *path, struct kstat *stat,
 		   u32 request_mask, unsigned int query_flags)
 {
 	struct inode *inode = d_inode(path->dentry);
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	stat->blocks = inode->i_mapping->nrpages << (PAGE_SHIFT - 9);
 	return 0;
 }
@@ -1304,7 +1304,7 @@ static int empty_dir_getattr(const struct path *path, struct kstat *stat,
 			     u32 request_mask, unsigned int query_flags)
 {
 	struct inode *inode = d_inode(path->dentry);
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	return 0;
 }
 
fs/minix/inode.c +1 -1
index 34f546404aa1..91c81d2fc90d 100644
@@ -658,7 +658,7 @@ int minix_getattr(const struct path *path, struct kstat *stat,
 	struct super_block *sb = path->dentry->d_sb;
 	struct inode *inode = d_inode(path->dentry);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	if (INODE_VERSION(inode) == MINIX_V1)
 		stat->blocks = (BLOCK_SIZE / 512) * V1_minix_blocks(stat->size, sb);
 	else
fs/nfs/inode.c +1 -1
index 522aa10a1a3e..cab123ec1664 100644
@@ -857,7 +857,7 @@ int nfs_getattr(const struct path *path, struct kstat *stat,
 	/* Only return attributes that were revalidated. */
 	stat->result_mask &= request_mask;
 out_no_update:
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	stat->ino = nfs_compat_user_ino64(NFS_FILEID(inode));
 	if (S_ISDIR(inode->i_mode))
 		stat->blksize = NFS_SERVER(inode)->dtsize;
fs/nfs/namespace.c +1 -1
index 2bcbe38afe2e..55fc711e368b 100644
@@ -213,7 +213,7 @@ nfs_namespace_getattr(const struct path *path, struct kstat *stat,
 {
 	if (NFS_FH(d_inode(path->dentry))->size != 0)
 		return nfs_getattr(path, stat, request_mask, query_flags);
-	generic_fillattr(d_inode(path->dentry), stat);
+	generic_fillattr(&init_user_ns, d_inode(path->dentry), stat);
 	return 0;
 }
 
fs/ocfs2/file.c +1 -1
index cabf355b148f..a070d4c9b6ed 100644
@@ -1313,7 +1313,7 @@ int ocfs2_getattr(const struct path *path, struct kstat *stat,
 		goto bail;
 	}
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	/*
 	 * If there is inline data in the inode, the inode will normally not
 	 * have data blocks allocated (it may have an external xattr block).
fs/orangefs/inode.c +1 -1
index 563fe9ab8eb2..b94032f77e61 100644
@@ -903,7 +903,7 @@ int orangefs_getattr(const struct path *path, struct kstat *stat,
 	ret = orangefs_inode_getattr(inode,
 	    request_mask & STATX_SIZE ? ORANGEFS_GETATTR_SIZE : 0);
 	if (ret == 0) {
-		generic_fillattr(inode, stat);
+		generic_fillattr(&init_user_ns, inode, stat);
 
 		/* override block size reported to stat */
 		if (!(request_mask & STATX_SIZE))
fs/proc/base.c +2 -2
index bb4e63a3684f..d45aa68c1f17 100644
@@ -1934,7 +1934,7 @@ int pid_getattr(const struct path *path, struct kstat *stat,
 	struct proc_fs_info *fs_info = proc_sb_info(inode->i_sb);
 	struct task_struct *task;
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 
 	stat->uid = GLOBAL_ROOT_UID;
 	stat->gid = GLOBAL_ROOT_GID;
@@ -3803,7 +3803,7 @@ static int proc_task_getattr(const struct path *path, struct kstat *stat,
 {
 	struct inode *inode = d_inode(path->dentry);
 	struct task_struct *p = get_proc_task(inode);
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 
 	if (p) {
 		stat->nlink += get_nr_threads(p);
fs/proc/generic.c +1 -1
index 6d4fabab8aa7..0db96a761149 100644
@@ -145,7 +145,7 @@ static int proc_getattr(const struct path *path, struct kstat *stat,
 		}
 	}
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	return 0;
 }
 
fs/proc/proc_net.c +1 -1
index 18601042af99..4aef49ccf571 100644
@@ -297,7 +297,7 @@ static int proc_tgid_net_getattr(const struct path *path, struct kstat *stat,
 
 	net = get_proc_task_net(inode);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 
 	if (net != NULL) {
 		stat->nlink = net->proc_net->nlink;
fs/proc/proc_sysctl.c +1 -1
index ec67dbc1f705..87c828348140 100644
@@ -840,7 +840,7 @@ static int proc_sys_getattr(const struct path *path, struct kstat *stat,
 	if (IS_ERR(head))
 		return PTR_ERR(head);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	if (table)
 		stat->mode = (stat->mode & S_IFMT) | table->mode;
 
fs/proc/root.c +1 -1
index 5e444d4f9717..244e4b6f15ef 100644
@@ -311,7 +311,7 @@ void __init proc_root_init(void)
 static int proc_root_getattr(const struct path *path, struct kstat *stat,
 			     u32 request_mask, unsigned int query_flags)
 {
-	generic_fillattr(d_inode(path->dentry), stat);
+	generic_fillattr(&init_user_ns, d_inode(path->dentry), stat);
 	stat->nlink = proc_root.nlink + nr_processes();
 	return 0;
 }
fs/stat.c +14 -6
index dacecdda2e79..8c58c37ef1ab 100644
@@ -26,21 +26,29 @@
 
 /**
  * generic_fillattr - Fill in the basic attributes from the inode struct
- * @inode: Inode to use as the source
- * @stat: Where to fill in the attributes
+ * @mnt_userns:	user namespace of the mount the inode was found from
+ * @inode:	Inode to use as the source
+ * @stat:	Where to fill in the attributes
  *
  * Fill in the basic attributes in the kstat structure from data that's to be
  * found on the VFS inode structure.  This is the default if no getattr inode
  * operation is supplied.
+ *
+ * If the inode has been found through an idmapped mount the user namespace of
+ * the vfsmount must be passed through @mnt_userns. This function will then take
+ * care to map the inode according to @mnt_userns before filling in the uid and
+ * gid filds. On non-idmapped mounts or if permission checking is to be
+ * performed on the raw inode simply passs init_user_ns.
  */
-void generic_fillattr(struct inode *inode, struct kstat *stat)
+void generic_fillattr(struct user_namespace *mnt_userns, struct inode *inode,
+		      struct kstat *stat)
 {
 	stat->dev = inode->i_sb->s_dev;
 	stat->ino = inode->i_ino;
 	stat->mode = inode->i_mode;
 	stat->nlink = inode->i_nlink;
-	stat->uid = inode->i_uid;
-	stat->gid = inode->i_gid;
+	stat->uid = i_uid_into_mnt(mnt_userns, inode);
+	stat->gid = i_gid_into_mnt(mnt_userns, inode);
 	stat->rdev = inode->i_rdev;
 	stat->size = i_size_read(inode);
 	stat->atime = inode->i_atime;
@@ -87,7 +95,7 @@ int vfs_getattr_nosec(const struct path *path, struct kstat *stat,
 		return inode->i_op->getattr(path, stat, request_mask,
 					    query_flags);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(mnt_user_ns(path->mnt), inode, stat);
 	return 0;
 }
 EXPORT_SYMBOL(vfs_getattr_nosec);
fs/sysv/itree.c +1 -1
index bcb67b0cabe7..83cffab6955f 100644
@@ -445,7 +445,7 @@ int sysv_getattr(const struct path *path, struct kstat *stat,
 		 u32 request_mask, unsigned int flags)
 {
 	struct super_block *s = path->dentry->d_sb;
-	generic_fillattr(d_inode(path->dentry), stat);
+	generic_fillattr(&init_user_ns, d_inode(path->dentry), stat);
 	stat->blocks = (s->s_blocksize / 512) * sysv_nblocks(s, stat->size);
 	stat->blksize = s->s_blocksize;
 	return 0;
fs/ubifs/dir.c +1 -1
index 694e7714545b..a8881ed61620 100644
@@ -1589,7 +1589,7 @@ int ubifs_getattr(const struct path *path, struct kstat *stat,
 				STATX_ATTR_ENCRYPTED |
 				STATX_ATTR_IMMUTABLE);
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	stat->blksize = UBIFS_BLOCK_SIZE;
 	stat->size = ui->ui_size;
 
fs/udf/symlink.c +1 -1
index c973db239604..54a44d1f023c 100644
@@ -159,7 +159,7 @@ static int udf_symlink_getattr(const struct path *path, struct kstat *stat,
 	struct inode *inode = d_backing_inode(dentry);
 	struct page *page;
 
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 	page = read_mapping_page(inode->i_mapping, 0, NULL);
 	if (IS_ERR(page))
 		return PTR_ERR(page);
fs/vboxsf/utils.c +1 -1
index 018057546067..d2cd1c99f48e 100644
@@ -233,7 +233,7 @@ int vboxsf_getattr(const struct path *path, struct kstat *kstat,
 	if (err)
 		return err;
 
-	generic_fillattr(d_inode(dentry), kstat);
+	generic_fillattr(&init_user_ns, d_inode(dentry), kstat);
 	return 0;
 }
 
include/linux/fs.h +1 -1
index 89e41a821bad..e4876e373636 100644
@@ -3141,7 +3141,7 @@ extern int __page_symlink(struct inode *inode, const char *symname, int len,
 extern int page_symlink(struct inode *inode, const char *symname, int len);
 extern const struct inode_operations page_symlink_inode_operations;
 extern void kfree_link(void *);
-extern void generic_fillattr(struct inode *, struct kstat *);
+extern void generic_fillattr(struct user_namespace *, struct inode *, struct kstat *);
 extern int vfs_getattr_nosec(const struct path *, struct kstat *, u32, unsigned int);
 extern int vfs_getattr(const struct path *, struct kstat *, u32, unsigned int);
 void __inode_add_bytes(struct inode *inode, loff_t bytes);
mm/shmem.c +1 -1
index 23b8e9c15a42..339d5530d3a9 100644
@@ -1072,7 +1072,7 @@ static int shmem_getattr(const struct path *path, struct kstat *stat,
 		shmem_recalc_inode(inode);
 		spin_unlock_irq(&info->lock);
 	}
-	generic_fillattr(inode, stat);
+	generic_fillattr(&init_user_ns, inode, stat);
 
 	if (is_huge_enabled(sb_info))
 		stat->blksize = HPAGE_PMD_SIZE;
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
Reviewed-by on v5 18/42
#30 lore raw

Re: [PATCH v5 18/42] stat: handle idmapped mounts

Quoted 6 lines · Christian Brauner

The generic_fillattr() helper fills in the basic attributes associated with an inode. Enable it to handle idmapped mounts. If the inode is accessed through an idmapped mount we need to map it according to the mount's user namespace before we store the uid and gid.
If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

Looks good,

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers

#31 lore raw

PATCH v5 19/42 namei: handle idmapped mounts in may_*() helpers

In reply to Christian Brauner

The may_follow_link(), may_linkat(), may_lookup(), may_open(), may_o_create(), may_create_in_sticky(), may_delete(), and may_create() helpers determine whether the caller is privileged enough to perform the associated operations. Let them handle idmapped mounts by mapping the inode or fsids according to the mount's user namespace. Afterwards the checks are identical to non-idmapped inodes. The patch takes care to retrieve the mount's user namespace right before performing permission checks and passing it down into the fileystem so the user namespace can't change in between by someone idmapping a mount that is currently not idmapped. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
unchanged

/* v3 */
unchanged

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/btrfs/ioctl.c   |   5 +-
 fs/init.c          |   2 +-
 fs/inode.c         |   2 +-
 fs/internal.h      |   2 +-
 fs/namei.c         | 147 ++++++++++++++++++++++++++++-----------------
 fs/xattr.c         |   2 +-
 include/linux/fs.h |  14 +++--
 7 files changed, 108 insertions(+), 66 deletions(-)
fs/btrfs/ioctl.c +3 -2
index 950b4f1b9092..05c33b8bd75b 100644
@@ -927,8 +927,9 @@ static int btrfs_may_delete(struct inode *dir, struct dentry *victim, int isdir)
 		return error;
 	if (IS_APPEND(dir))
 		return -EPERM;
-	if (check_sticky(dir, d_inode(victim)) || IS_APPEND(d_inode(victim)) ||
-	    IS_IMMUTABLE(d_inode(victim)) || IS_SWAPFILE(d_inode(victim)))
+	if (check_sticky(&init_user_ns, dir, d_inode(victim)) ||
+	    IS_APPEND(d_inode(victim)) || IS_IMMUTABLE(d_inode(victim)) ||
+	    IS_SWAPFILE(d_inode(victim)))
 		return -EPERM;
 	if (isdir) {
 		if (!d_is_dir(victim))
fs/init.c +1 -1
index 2b4842f4802b..dd3d698afcd8 100644
@@ -184,7 +184,7 @@ int __init init_link(const char *oldname, const char *newname)
 	error = -EXDEV;
 	if (old_path.mnt != new_path.mnt)
 		goto out_dput;
-	error = may_linkat(&old_path);
+	error = may_linkat(&init_user_ns, &old_path);
 	if (unlikely(error))
 		goto out_dput;
 	error = security_path_link(old_path.dentry, &new_path, new_dentry);
fs/inode.c +1 -1
index 49b512592dcd..46116ef44c9f 100644
@@ -1796,7 +1796,7 @@ bool atime_needs_update(const struct path *path, struct inode *inode)
 	/* Atime updates will likely cause i_uid and i_gid to be written
 	 * back improprely if their true value is unknown to the vfs.
 	 */
-	if (HAS_UNMAPPED_ID(inode))
+	if (HAS_UNMAPPED_ID(mnt_user_ns(mnt), inode))
 		return false;
 
 	if (IS_NOATIME(inode))
fs/internal.h +1 -1
index 77c50befbfbe..6c8a4eddc7e6 100644
@@ -73,7 +73,7 @@ extern int vfs_path_lookup(struct dentry *, struct vfsmount *,
 			   const char *, unsigned int, struct path *);
 long do_rmdir(int dfd, struct filename *name);
 long do_unlinkat(int dfd, struct filename *name);
-int may_linkat(struct path *link);
+int may_linkat(struct user_namespace *mnt_userns, struct path *link);
 int do_renameat2(int olddfd, struct filename *oldname, int newdfd,
 		 struct filename *newname, unsigned int flags);
 
fs/namei.c +92 -55

437 lines not on this page · raw message

fs/xattr.c +1 -1
index 03afd803e50b..f0bfe207ebab 100644
@@ -98,7 +98,7 @@ xattr_permission(struct user_namespace *mnt_userns, struct inode *inode,
 		 * to be writen back improperly if their true value is
 		 * unknown to the vfs.
 		 */
-		if (HAS_UNMAPPED_ID(inode))
+		if (HAS_UNMAPPED_ID(mnt_userns, inode))
 			return -EPERM;
 	}
 
include/linux/fs.h +9 -5
index e4876e373636..e7c802478730 100644
@@ -2083,9 +2083,11 @@ static inline bool sb_rdonly(const struct super_block *sb) { return sb->s_flags
 #define IS_WHITEOUT(inode)	(S_ISCHR(inode->i_mode) && \
 				 (inode)->i_rdev == WHITEOUT_DEV)
 
-static inline bool HAS_UNMAPPED_ID(struct inode *inode)
+static inline bool HAS_UNMAPPED_ID(struct user_namespace *mnt_userns,
+				   struct inode *inode)
 {
-	return !uid_valid(inode->i_uid) || !gid_valid(inode->i_gid);
+	return !uid_valid(i_uid_into_mnt(mnt_userns, inode)) ||
+	       !gid_valid(i_gid_into_mnt(mnt_userns, inode));
 }
 
 static inline enum rw_hint file_write_hint(struct file *file)
@@ -2810,7 +2812,8 @@ extern int notify_change(struct user_namespace *, struct dentry *,
 			 struct iattr *, struct inode **);
 extern int inode_permission(struct user_namespace *, struct inode *, int);
 extern int generic_permission(struct user_namespace *, struct inode *, int);
-extern int __check_sticky(struct inode *dir, struct inode *inode);
+extern int __check_sticky(struct user_namespace *mnt_userns, struct inode *dir,
+			  struct inode *inode);
 
 static inline bool execute_ok(struct inode *inode)
 {
@@ -3429,12 +3432,13 @@ static inline bool is_sxid(umode_t mode)
 	return (mode & S_ISUID) || ((mode & S_ISGID) && (mode & S_IXGRP));
 }
 
-static inline int check_sticky(struct inode *dir, struct inode *inode)
+static inline int check_sticky(struct user_namespace *mnt_userns,
+			       struct inode *dir, struct inode *inode)
 {
 	if (!(dir->i_mode & S_ISVTX))
 		return 0;
 
-	return __check_sticky(dir, inode);
+	return __check_sticky(mnt_userns, dir, inode);
 }
 
 static inline void inode_has_no_xattr(struct inode *inode)
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
Reviewed-by on v5 19/42
#32 lore raw

Re: [PATCH v5 19/42] namei: handle idmapped mounts in may_*() helpers

Quoted 11 lines · Christian Brauner

The may_follow_link(), may_linkat(), may_lookup(), may_open(), may_o_create(), may_create_in_sticky(), may_delete(), and may_create() helpers determine whether the caller is privileged enough to perform the associated operations. Let them handle idmapped mounts by mapping the inode or fsids according to the mount's user namespace. Afterwards the checks are identical to non-idmapped inodes. The patch takes care to retrieve the mount's user namespace right before performing permission checks and passing it down into the fileystem so the user namespace can't change in between by someone idmapping a mount that is currently not idmapped. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

Looks good,

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers

#33 lore raw

PATCH v5 20/42 namei: introduce struct renamedata

In reply to Christian Brauner

In order to handle idmapped mounts we will extend the vfs rename helper to take two new arguments in follow up patches. Since this operations already takes a bunch of arguments add a simple struct renamedata (based on struct nameidata) and make the current helper use it before we extend it.

---
/* v2 */
unchanged

/* v3 */
unchanged

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

- Christoph Hellwig <[email protected]>:
- Declare variable on separate lines instead of chaining them and wrapping with weird indentation.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/cachefiles/namei.c    |  9 +++++++--
 fs/ecryptfs/inode.c      | 10 +++++++---
 fs/namei.c               | 21 +++++++++++++++------
 fs/nfsd/vfs.c            |  8 +++++++-
 fs/overlayfs/overlayfs.h |  9 ++++++++-
 include/linux/fs.h       | 12 +++++++++++-
 6 files changed, 55 insertions(+), 14 deletions(-)
fs/cachefiles/namei.c +7 -2
index ecc8ecbbfa5a..7b987de0babe 100644
@@ -412,9 +412,14 @@ static int cachefiles_bury_object(struct cachefiles_cache *cache,
 	if (ret < 0) {
 		cachefiles_io_error(cache, "Rename security error %d", ret);
 	} else {
+		struct renamedata rd = {
+			.old_dir	= d_inode(dir),
+			.old_dentry	= rep,
+			.new_dir	= d_inode(cache->graveyard),
+			.new_dentry	= grave,
+		};
 		trace_cachefiles_rename(object, rep, grave, why);
-		ret = vfs_rename(d_inode(dir), rep,
-				 d_inode(cache->graveyard), grave, NULL, 0);
+		ret = vfs_rename(&rd);
 		if (ret != 0 && ret != -ENOMEM)
 			cachefiles_io_error(cache,
 					    "Rename failed with error %d", ret);
fs/ecryptfs/inode.c +7 -3
index 0a218d7c3023..0697480ee2eb 100644
@@ -590,6 +590,7 @@ ecryptfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	struct dentry *lower_new_dir_dentry;
 	struct dentry *trap;
 	struct inode *target_inode;
+	struct renamedata rd = {};
 
 	if (flags)
 		return -EINVAL;
@@ -619,9 +620,12 @@ ecryptfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		rc = -ENOTEMPTY;
 		goto out_lock;
 	}
-	rc = vfs_rename(d_inode(lower_old_dir_dentry), lower_old_dentry,
-			d_inode(lower_new_dir_dentry), lower_new_dentry,
-			NULL, 0);
+
+	rd.old_dir	= d_inode(lower_old_dir_dentry);
+	rd.old_dentry	= lower_old_dentry;
+	rd.new_dir	= d_inode(lower_new_dir_dentry);
+	rd.new_dentry	= lower_new_dentry;
+	rc = vfs_rename(&rd);
 	if (rc)
 		goto out_lock;
 	if (target_inode)
fs/namei.c +15 -6
index 05f1db8e7ed0..d145dceeaeb0 100644
@@ -4315,12 +4315,15 @@ SYSCALL_DEFINE2(link, const char __user *, oldname, const char __user *, newname
  *	   ->i_mutex on parents, which works but leads to some truly excessive
  *	   locking].
  */
-int vfs_rename(struct inode *old_dir, struct dentry *old_dentry,
-	       struct inode *new_dir, struct dentry *new_dentry,
-	       struct inode **delegated_inode, unsigned int flags)
+int vfs_rename(struct renamedata *rd)
 {
 	int error;
 	struct user_namespace *mnt_userns = &init_user_ns;
+	struct inode *old_dir = rd->old_dir, *new_dir = rd->new_dir;
+	struct dentry *old_dentry = rd->old_dentry;
+	struct dentry *new_dentry = rd->new_dentry;
+	struct inode **delegated_inode = rd->delegated_inode;
+	unsigned int flags = rd->flags;
 	bool is_dir = d_is_dir(old_dentry);
 	struct inode *source = old_dentry->d_inode;
 	struct inode *target = new_dentry->d_inode;
@@ -4444,6 +4447,7 @@ EXPORT_SYMBOL(vfs_rename);
 int do_renameat2(int olddfd, struct filename *from, int newdfd,
 		 struct filename *to, unsigned int flags)
 {
+	struct renamedata rd;
 	struct dentry *old_dentry, *new_dentry;
 	struct dentry *trap;
 	struct path old_path, new_path;
@@ -4547,9 +4551,14 @@ int do_renameat2(int olddfd, struct filename *from, int newdfd,
 				     &new_path, new_dentry, flags);
 	if (error)
 		goto exit5;
-	error = vfs_rename(old_path.dentry->d_inode, old_dentry,
-			   new_path.dentry->d_inode, new_dentry,
-			   &delegated_inode, flags);
+
+	rd.old_dir	   = old_path.dentry->d_inode;
+	rd.old_dentry	   = old_dentry;
+	rd.new_dir	   = new_path.dentry->d_inode;
+	rd.new_dentry	   = new_dentry;
+	rd.delegated_inode = &delegated_inode;
+	rd.flags	   = flags;
+	error = vfs_rename(&rd);
 exit5:
 	dput(new_dentry);
 exit4:
fs/nfsd/vfs.c +7 -1
index 494362c2e4fc..045f49a4352d 100644
@@ -1797,7 +1797,13 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen,
 		close_cached = true;
 		goto out_dput_old;
 	} else {
-		host_err = vfs_rename(fdir, odentry, tdir, ndentry, NULL, 0);
+		struct renamedata rd = {
+			.old_dir	= fdir,
+			.old_dentry	= odentry,
+			.new_dir	= tdir,
+			.new_dentry	= ndentry,
+		};
+		host_err = vfs_rename(&rd);
 		if (!host_err) {
 			host_err = commit_metadata(tfhp);
 			if (!host_err)
fs/overlayfs/overlayfs.h +8 -1
index 0002834f664a..426899681df7 100644
@@ -214,9 +214,16 @@ static inline int ovl_do_rename(struct inode *olddir, struct dentry *olddentry,
 				unsigned int flags)
 {
 	int err;
+	struct renamedata rd = {
+		.old_dir 	= olddir,
+		.old_dentry 	= olddentry,
+		.new_dir 	= newdir,
+		.new_dentry 	= newdentry,
+		.flags 		= flags,
+	};
 
 	pr_debug("rename(%pd2, %pd2, 0x%x)\n", olddentry, newdentry, flags);
-	err = vfs_rename(olddir, olddentry, newdir, newdentry, NULL, flags);
+	err = vfs_rename(&rd);
 	if (err) {
 		pr_debug("...rename(%pd2, %pd2, ...) = %i\n",
 			 olddentry, newdentry, err);
include/linux/fs.h +11 -1
index e7c802478730..e0e27e4dd817 100644
@@ -1774,7 +1774,17 @@ extern int vfs_symlink(struct inode *, struct dentry *, const char *);
 extern int vfs_link(struct dentry *, struct inode *, struct dentry *, struct inode **);
 extern int vfs_rmdir(struct inode *, struct dentry *);
 extern int vfs_unlink(struct inode *, struct dentry *, struct inode **);
-extern int vfs_rename(struct inode *, struct dentry *, struct inode *, struct dentry *, struct inode **, unsigned int);
+
+struct renamedata {
+	struct inode *old_dir;
+	struct dentry *old_dentry;
+	struct inode *new_dir;
+	struct dentry *new_dentry;
+	struct inode **delegated_inode;
+	unsigned int flags;
+} __randomize_layout;
+
+extern int vfs_rename(struct renamedata *);
 
 static inline int vfs_whiteout(struct inode *dir, struct dentry *dentry)
 {
Signature · 1 line
2.30.0
#35 lore raw

PATCH v5 21/42 namei: prepare for idmapped mounts

In reply to Christian Brauner

The various vfs_*() helpers are called by filesystems or by the vfs itself to perform core operations create, link, mkdir, mknod, rename, rmdir, tmpfile and unlink. Enable them to handle idmapped mounts. If the inode is accessed through an idmapped mount it is mapped according to the mount's user namespace. Afterwards the checks and operations 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.

---
/* v2 */
- Christoph Hellwig <[email protected]>:
- Don't pollute the vfs with additional helpers simply extend the existing helpers with an additional argument and switch all callers.

/* v3 */
unchanged

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 drivers/base/devtmpfs.c  |   8 +-
 fs/cachefiles/namei.c    |  10 +-
 fs/ecryptfs/inode.c      |  30 +++---
 fs/init.c                |  12 +--
 fs/namei.c               | 224 +++++++++++++++++++++++++++++++--------
 fs/nfsd/nfs4recover.c    |   6 +-
 fs/nfsd/vfs.c            |  19 ++--
 fs/overlayfs/dir.c       |   4 +-
 fs/overlayfs/overlayfs.h |  20 ++--
 include/linux/fs.h       |  25 +++--
 ipc/mqueue.c             |   2 +-
 net/unix/af_unix.c       |   2 +-
 12 files changed, 259 insertions(+), 103 deletions(-)
drivers/base/devtmpfs.c +4 -4
index 2e0c3cdb4184..fd4e86c58111 100644
@@ -162,7 +162,7 @@ static int dev_mkdir(const char *name, umode_t mode)
 	if (IS_ERR(dentry))
 		return PTR_ERR(dentry);
 
-	err = vfs_mkdir(d_inode(path.dentry), dentry, mode);
+	err = vfs_mkdir(&init_user_ns, d_inode(path.dentry), dentry, mode);
 	if (!err)
 		/* mark as kernel-created inode */
 		d_inode(dentry)->i_private = &thread;
@@ -212,7 +212,7 @@ static int handle_create(const char *nodename, umode_t mode, kuid_t uid,
 	if (IS_ERR(dentry))
 		return PTR_ERR(dentry);
 
-	err = vfs_mknod(d_inode(path.dentry), dentry, mode, dev->devt);
+	err = vfs_mknod(&init_user_ns, d_inode(path.dentry), dentry, mode, dev->devt);
 	if (!err) {
 		struct iattr newattrs;
 
@@ -242,7 +242,7 @@ static int dev_rmdir(const char *name)
 		return PTR_ERR(dentry);
 	if (d_really_is_positive(dentry)) {
 		if (d_inode(dentry)->i_private == &thread)
-			err = vfs_rmdir(d_inode(parent.dentry), dentry);
+			err = vfs_rmdir(&init_user_ns, d_inode(parent.dentry), dentry);
 		else
 			err = -EPERM;
 	} else {
@@ -330,7 +330,7 @@ static int handle_remove(const char *nodename, struct device *dev)
 			inode_lock(d_inode(dentry));
 			notify_change(&init_user_ns, dentry, &newattrs, NULL);
 			inode_unlock(d_inode(dentry));
-			err = vfs_unlink(d_inode(parent.dentry), dentry, NULL);
+			err = vfs_unlink(&init_user_ns, d_inode(parent.dentry), dentry, NULL);
 			if (!err || err == -ENOENT)
 				deleted = 1;
 		}
fs/cachefiles/namei.c +6 -4
index 7b987de0babe..78e1deed3af0 100644
@@ -311,7 +311,7 @@ static int cachefiles_bury_object(struct cachefiles_cache *cache,
 			cachefiles_io_error(cache, "Unlink security error");
 		} else {
 			trace_cachefiles_unlink(object, rep, why);
-			ret = vfs_unlink(d_inode(dir), rep, NULL);
+			ret = vfs_unlink(&init_user_ns, d_inode(dir), rep, NULL);
 
 			if (preemptive)
 				cachefiles_mark_object_buried(cache, rep, why);
@@ -413,8 +413,10 @@ static int cachefiles_bury_object(struct cachefiles_cache *cache,
 		cachefiles_io_error(cache, "Rename security error %d", ret);
 	} else {
 		struct renamedata rd = {
+			.old_mnt_userns	= &init_user_ns,
 			.old_dir	= d_inode(dir),
 			.old_dentry	= rep,
+			.new_mnt_userns	= &init_user_ns,
 			.new_dir	= d_inode(cache->graveyard),
 			.new_dentry	= grave,
 		};
@@ -566,7 +568,7 @@ int cachefiles_walk_to_object(struct cachefiles_object *parent,
 			if (ret < 0)
 				goto create_error;
 			start = jiffies;
-			ret = vfs_mkdir(d_inode(dir), next, 0);
+			ret = vfs_mkdir(&init_user_ns, d_inode(dir), next, 0);
 			cachefiles_hist(cachefiles_mkdir_histogram, start);
 			if (!key)
 				trace_cachefiles_mkdir(object, next, ret);
@@ -602,7 +604,7 @@ int cachefiles_walk_to_object(struct cachefiles_object *parent,
 			if (ret < 0)
 				goto create_error;
 			start = jiffies;
-			ret = vfs_create(d_inode(dir), next, S_IFREG, true);
+			ret = vfs_create(&init_user_ns, d_inode(dir), next, S_IFREG, true);
 			cachefiles_hist(cachefiles_create_histogram, start);
 			trace_cachefiles_create(object, next, ret);
 			if (ret < 0)
@@ -796,7 +798,7 @@ struct dentry *cachefiles_get_directory(struct cachefiles_cache *cache,
 		ret = security_path_mkdir(&path, subdir, 0700);
 		if (ret < 0)
 			goto mkdir_error;
-		ret = vfs_mkdir(d_inode(dir), subdir, 0700);
+		ret = vfs_mkdir(&init_user_ns, d_inode(dir), subdir, 0700);
 		if (ret < 0)
 			goto mkdir_error;
 
fs/ecryptfs/inode.c +17 -13
index 0697480ee2eb..0c97a1c37e5d 100644
@@ -141,7 +141,7 @@ static int ecryptfs_do_unlink(struct inode *dir, struct dentry *dentry,
 	else if (d_unhashed(lower_dentry))
 		rc = -EINVAL;
 	else
-		rc = vfs_unlink(lower_dir_inode, lower_dentry, NULL);
+		rc = vfs_unlink(&init_user_ns, lower_dir_inode, lower_dentry, NULL);
 	if (rc) {
 		printk(KERN_ERR "Error in vfs_unlink; rc = [%d]\n", rc);
 		goto out_unlock;
@@ -180,7 +180,8 @@ ecryptfs_do_create(struct inode *directory_inode,
 
 	lower_dentry = ecryptfs_dentry_to_lower(ecryptfs_dentry);
 	lower_dir_dentry = lock_parent(lower_dentry);
-	rc = vfs_create(d_inode(lower_dir_dentry), lower_dentry, mode, true);
+	rc = vfs_create(&init_user_ns, d_inode(lower_dir_dentry), lower_dentry,
+			mode, true);
 	if (rc) {
 		printk(KERN_ERR "%s: Failure to create dentry in lower fs; "
 		       "rc = [%d]\n", __func__, rc);
@@ -190,7 +191,7 @@ ecryptfs_do_create(struct inode *directory_inode,
 	inode = __ecryptfs_get_inode(d_inode(lower_dentry),
 				     directory_inode->i_sb);
 	if (IS_ERR(inode)) {
-		vfs_unlink(d_inode(lower_dir_dentry), lower_dentry, NULL);
+		vfs_unlink(&init_user_ns, d_inode(lower_dir_dentry), lower_dentry, NULL);
 		goto out_lock;
 	}
 	fsstack_copy_attr_times(directory_inode, d_inode(lower_dir_dentry));
@@ -436,8 +437,8 @@ static int ecryptfs_link(struct dentry *old_dentry, struct inode *dir,
 	dget(lower_old_dentry);
 	dget(lower_new_dentry);
 	lower_dir_dentry = lock_parent(lower_new_dentry);
-	rc = vfs_link(lower_old_dentry, d_inode(lower_dir_dentry),
-		      lower_new_dentry, NULL);
+	rc = vfs_link(lower_old_dentry, &init_user_ns,
+		      d_inode(lower_dir_dentry), lower_new_dentry, NULL);
 	if (rc || d_really_is_negative(lower_new_dentry))
 		goto out_lock;
 	rc = ecryptfs_interpose(lower_new_dentry, new_dentry, dir->i_sb);
@@ -481,7 +482,7 @@ static int ecryptfs_symlink(struct inode *dir, struct dentry *dentry,
 						  strlen(symname));
 	if (rc)
 		goto out_lock;
-	rc = vfs_symlink(d_inode(lower_dir_dentry), lower_dentry,
+	rc = vfs_symlink(&init_user_ns, d_inode(lower_dir_dentry), lower_dentry,
 			 encoded_symname);
 	kfree(encoded_symname);
 	if (rc || d_really_is_negative(lower_dentry))
@@ -507,7 +508,7 @@ static int ecryptfs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode
 
 	lower_dentry = ecryptfs_dentry_to_lower(dentry);
 	lower_dir_dentry = lock_parent(lower_dentry);
-	rc = vfs_mkdir(d_inode(lower_dir_dentry), lower_dentry, mode);
+	rc = vfs_mkdir(&init_user_ns, d_inode(lower_dir_dentry), lower_dentry, mode);
 	if (rc || d_really_is_negative(lower_dentry))
 		goto out;
 	rc = ecryptfs_interpose(lower_dentry, dentry, dir->i_sb);
@@ -541,7 +542,7 @@ static int ecryptfs_rmdir(struct inode *dir, struct dentry *dentry)
 	else if (d_unhashed(lower_dentry))
 		rc = -EINVAL;
 	else
-		rc = vfs_rmdir(lower_dir_inode, lower_dentry);
+		rc = vfs_rmdir(&init_user_ns, lower_dir_inode, lower_dentry);
 	if (!rc) {
 		clear_nlink(d_inode(dentry));
 		fsstack_copy_attr_times(dir, lower_dir_inode);
@@ -563,7 +564,8 @@ ecryptfs_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, dev_t dev
 
 	lower_dentry = ecryptfs_dentry_to_lower(dentry);
 	lower_dir_dentry = lock_parent(lower_dentry);
-	rc = vfs_mknod(d_inode(lower_dir_dentry), lower_dentry, mode, dev);
+	rc = vfs_mknod(&init_user_ns, d_inode(lower_dir_dentry), lower_dentry,
+		       mode, dev);
 	if (rc || d_really_is_negative(lower_dentry))
 		goto out;
 	rc = ecryptfs_interpose(lower_dentry, dentry, dir->i_sb);
@@ -621,10 +623,12 @@ ecryptfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		goto out_lock;
 	}
 
-	rd.old_dir	= d_inode(lower_old_dir_dentry);
-	rd.old_dentry	= lower_old_dentry;
-	rd.new_dir	= d_inode(lower_new_dir_dentry);
-	rd.new_dentry	= lower_new_dentry;
+	rd.old_mnt_userns	= &init_user_ns;
+	rd.old_dir		= d_inode(lower_old_dir_dentry);
+	rd.old_dentry		= lower_old_dentry;
+	rd.new_mnt_userns	= &init_user_ns;
+	rd.new_dir		= d_inode(lower_new_dir_dentry);
+	rd.new_dentry		= lower_new_dentry;
 	rc = vfs_rename(&rd);
 	if (rc)
 		goto out_lock;
fs/init.c +6 -6
index dd3d698afcd8..06d8e52ce18e 100644
@@ -160,8 +160,8 @@ int __init init_mknod(const char *filename, umode_t mode, unsigned int dev)
 		mode &= ~current_umask();
 	error = security_path_mknod(&path, dentry, mode, dev);
 	if (!error)
-		error = vfs_mknod(path.dentry->d_inode, dentry, mode,
-				  new_decode_dev(dev));
+		error = vfs_mknod(&init_user_ns, path.dentry->d_inode, dentry,
+				  mode, new_decode_dev(dev));
 	done_path_create(&path, dentry);
 	return error;
 }
@@ -190,8 +190,8 @@ int __init init_link(const char *oldname, const char *newname)
 	error = security_path_link(old_path.dentry, &new_path, new_dentry);
 	if (error)
 		goto out_dput;
-	error = vfs_link(old_path.dentry, new_path.dentry->d_inode, new_dentry,
-			 NULL);
+	error = vfs_link(old_path.dentry, &init_user_ns,
+			 new_path.dentry->d_inode, new_dentry, NULL);
 out_dput:
 	done_path_create(&new_path, new_dentry);
 out:
@@ -210,7 +210,7 @@ int __init init_symlink(const char *oldname, const char *newname)
 		return PTR_ERR(dentry);
 	error = security_path_symlink(&path, dentry, oldname);
 	if (!error)
-		error = vfs_symlink(path.dentry->d_inode, dentry, oldname);
+		error = vfs_symlink(&init_user_ns, path.dentry->d_inode, dentry, oldname);
 	done_path_create(&path, dentry);
 	return error;
 }
@@ -233,7 +233,7 @@ int __init init_mkdir(const char *pathname, umode_t mode)
 		mode &= ~current_umask();
 	error = security_path_mkdir(&path, dentry, mode);
 	if (!error)
-		error = vfs_mkdir(path.dentry->d_inode, dentry, mode);
+		error = vfs_mkdir(&init_user_ns, path.dentry->d_inode, dentry, mode);
 	done_path_create(&path, dentry);
 	return error;
 }
fs/namei.c +182 -42

431 lines not on this page · raw message

fs/nfsd/nfs4recover.c +3 -3
index 186fa2c2c6ba..891395c6c7d3 100644
@@ -233,7 +233,7 @@ nfsd4_create_clid_dir(struct nfs4_client *clp)
 		 * as well be forgiving and just succeed silently.
 		 */
 		goto out_put;
-	status = vfs_mkdir(d_inode(dir), dentry, S_IRWXU);
+	status = vfs_mkdir(&init_user_ns, d_inode(dir), dentry, S_IRWXU);
 out_put:
 	dput(dentry);
 out_unlock:
@@ -353,7 +353,7 @@ nfsd4_unlink_clid_dir(char *name, int namlen, struct nfsd_net *nn)
 	status = -ENOENT;
 	if (d_really_is_negative(dentry))
 		goto out;
-	status = vfs_rmdir(d_inode(dir), dentry);
+	status = vfs_rmdir(&init_user_ns, d_inode(dir), dentry);
 out:
 	dput(dentry);
 out_unlock:
@@ -443,7 +443,7 @@ purge_old(struct dentry *parent, struct dentry *child, struct nfsd_net *nn)
 	if (nfs4_has_reclaimed_state(name, nn))
 		goto out_free;
 
-	status = vfs_rmdir(d_inode(parent), child);
+	status = vfs_rmdir(&init_user_ns, d_inode(parent), child);
 	if (status)
 		printk("failed to remove client recovery directory %pd\n",
 				child);
fs/nfsd/vfs.c +11 -8
index 045f49a4352d..402b5f301884 100644
@@ -1254,12 +1254,12 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	host_err = 0;
 	switch (type) {
 	case S_IFREG:
-		host_err = vfs_create(dirp, dchild, iap->ia_mode, true);
+		host_err = vfs_create(&init_user_ns, dirp, dchild, iap->ia_mode, true);
 		if (!host_err)
 			nfsd_check_ignore_resizing(iap);
 		break;
 	case S_IFDIR:
-		host_err = vfs_mkdir(dirp, dchild, iap->ia_mode);
+		host_err = vfs_mkdir(&init_user_ns, dirp, dchild, iap->ia_mode);
 		if (!host_err && unlikely(d_unhashed(dchild))) {
 			struct dentry *d;
 			d = lookup_one_len(dchild->d_name.name,
@@ -1287,7 +1287,8 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	case S_IFBLK:
 	case S_IFIFO:
 	case S_IFSOCK:
-		host_err = vfs_mknod(dirp, dchild, iap->ia_mode, rdev);
+		host_err = vfs_mknod(&init_user_ns, dirp, dchild,
+				     iap->ia_mode, rdev);
 		break;
 	default:
 		printk(KERN_WARNING "nfsd: bad file type %o in nfsd_create\n",
@@ -1485,7 +1486,7 @@ do_nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (!IS_POSIXACL(dirp))
 		iap->ia_mode &= ~current_umask();
 
-	host_err = vfs_create(dirp, dchild, iap->ia_mode, true);
+	host_err = vfs_create(&init_user_ns, dirp, dchild, iap->ia_mode, true);
 	if (host_err < 0) {
 		fh_drop_write(fhp);
 		goto out_nfserr;
@@ -1609,7 +1610,7 @@ nfsd_symlink(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (IS_ERR(dnew))
 		goto out_nfserr;
 
-	host_err = vfs_symlink(d_inode(dentry), dnew, path);
+	host_err = vfs_symlink(&init_user_ns, d_inode(dentry), dnew, path);
 	err = nfserrno(host_err);
 	if (!err)
 		err = nfserrno(commit_metadata(fhp));
@@ -1677,7 +1678,7 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 	err = nfserr_noent;
 	if (d_really_is_negative(dold))
 		goto out_dput;
-	host_err = vfs_link(dold, dirp, dnew, NULL);
+	host_err = vfs_link(dold, &init_user_ns, dirp, dnew, NULL);
 	if (!host_err) {
 		err = nfserrno(commit_metadata(ffhp));
 		if (!err)
@@ -1798,8 +1799,10 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen,
 		goto out_dput_old;
 	} else {
 		struct renamedata rd = {
+			.old_mnt_userns	= &init_user_ns,
 			.old_dir	= fdir,
 			.old_dentry	= odentry,
+			.new_mnt_userns	= &init_user_ns,
 			.new_dir	= tdir,
 			.new_dentry	= ndentry,
 		};
@@ -1890,9 +1893,9 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 	if (type != S_IFDIR) {
 		if (rdentry->d_sb->s_export_op->flags & EXPORT_OP_CLOSE_BEFORE_UNLINK)
 			nfsd_close_cached_files(rdentry);
-		host_err = vfs_unlink(dirp, rdentry, NULL);
+		host_err = vfs_unlink(&init_user_ns, dirp, rdentry, NULL);
 	} else {
-		host_err = vfs_rmdir(dirp, rdentry);
+		host_err = vfs_rmdir(&init_user_ns, dirp, rdentry);
 	}
 
 	if (!host_err)
fs/overlayfs/dir.c +2 -2
index d75c96cb18c3..6904cc2ed7bb 100644
@@ -821,9 +821,9 @@ static int ovl_remove_upper(struct dentry *dentry, bool is_dir,
 		goto out_dput_upper;
 
 	if (is_dir)
-		err = vfs_rmdir(dir, upper);
+		err = vfs_rmdir(&init_user_ns, dir, upper);
 	else
-		err = vfs_unlink(dir, upper, NULL);
+		err = vfs_unlink(&init_user_ns, dir, upper, NULL);
 	ovl_dir_modified(dentry->d_parent, ovl_type_origin(dentry));
 
 	/*
fs/overlayfs/overlayfs.h +11 -9
index 426899681df7..5e9eb46e741a 100644
@@ -123,7 +123,7 @@ static inline const char *ovl_xattr(struct ovl_fs *ofs, enum ovl_xattr ox)
 
 static inline int ovl_do_rmdir(struct inode *dir, struct dentry *dentry)
 {
-	int err = vfs_rmdir(dir, dentry);
+	int err = vfs_rmdir(&init_user_ns, dir, dentry);
 
 	pr_debug("rmdir(%pd2) = %i\n", dentry, err);
 	return err;
@@ -131,7 +131,7 @@ static inline int ovl_do_rmdir(struct inode *dir, struct dentry *dentry)
 
 static inline int ovl_do_unlink(struct inode *dir, struct dentry *dentry)
 {
-	int err = vfs_unlink(dir, dentry, NULL);
+	int err = vfs_unlink(&init_user_ns, dir, dentry, NULL);
 
 	pr_debug("unlink(%pd2) = %i\n", dentry, err);
 	return err;
@@ -140,7 +140,7 @@ static inline int ovl_do_unlink(struct inode *dir, struct dentry *dentry)
 static inline int ovl_do_link(struct dentry *old_dentry, struct inode *dir,
 			      struct dentry *new_dentry)
 {
-	int err = vfs_link(old_dentry, dir, new_dentry, NULL);
+	int err = vfs_link(old_dentry, &init_user_ns, dir, new_dentry, NULL);
 
 	pr_debug("link(%pd2, %pd2) = %i\n", old_dentry, new_dentry, err);
 	return err;
@@ -149,7 +149,7 @@ static inline int ovl_do_link(struct dentry *old_dentry, struct inode *dir,
 static inline int ovl_do_create(struct inode *dir, struct dentry *dentry,
 				umode_t mode)
 {
-	int err = vfs_create(dir, dentry, mode, true);
+	int err = vfs_create(&init_user_ns, dir, dentry, mode, true);
 
 	pr_debug("create(%pd2, 0%o) = %i\n", dentry, mode, err);
 	return err;
@@ -158,7 +158,7 @@ static inline int ovl_do_create(struct inode *dir, struct dentry *dentry,
 static inline int ovl_do_mkdir(struct inode *dir, struct dentry *dentry,
 			       umode_t mode)
 {
-	int err = vfs_mkdir(dir, dentry, mode);
+	int err = vfs_mkdir(&init_user_ns, dir, dentry, mode);
 	pr_debug("mkdir(%pd2, 0%o) = %i\n", dentry, mode, err);
 	return err;
 }
@@ -166,7 +166,7 @@ static inline int ovl_do_mkdir(struct inode *dir, struct dentry *dentry,
 static inline int ovl_do_mknod(struct inode *dir, struct dentry *dentry,
 			       umode_t mode, dev_t dev)
 {
-	int err = vfs_mknod(dir, dentry, mode, dev);
+	int err = vfs_mknod(&init_user_ns, dir, dentry, mode, dev);
 
 	pr_debug("mknod(%pd2, 0%o, 0%o) = %i\n", dentry, mode, dev, err);
 	return err;
@@ -175,7 +175,7 @@ static inline int ovl_do_mknod(struct inode *dir, struct dentry *dentry,
 static inline int ovl_do_symlink(struct inode *dir, struct dentry *dentry,
 				 const char *oldname)
 {
-	int err = vfs_symlink(dir, dentry, oldname);
+	int err = vfs_symlink(&init_user_ns, dir, dentry, oldname);
 
 	pr_debug("symlink(\"%s\", %pd2) = %i\n", oldname, dentry, err);
 	return err;
@@ -215,8 +215,10 @@ static inline int ovl_do_rename(struct inode *olddir, struct dentry *olddentry,
 {
 	int err;
 	struct renamedata rd = {
+		.old_mnt_userns	= &init_user_ns,
 		.old_dir 	= olddir,
 		.old_dentry 	= olddentry,
+		.new_mnt_userns	= &init_user_ns,
 		.new_dir 	= newdir,
 		.new_dentry 	= newdentry,
 		.flags 		= flags,
@@ -233,14 +235,14 @@ static inline int ovl_do_rename(struct inode *olddir, struct dentry *olddentry,
 
 static inline int ovl_do_whiteout(struct inode *dir, struct dentry *dentry)
 {
-	int err = vfs_whiteout(dir, dentry);
+	int err = vfs_whiteout(&init_user_ns, dir, dentry);
 	pr_debug("whiteout(%pd2) = %i\n", dentry, err);
 	return err;
 }
 
 static inline struct dentry *ovl_do_tmpfile(struct dentry *dentry, umode_t mode)
 {
-	struct dentry *ret = vfs_tmpfile(dentry, mode, 0);
+	struct dentry *ret = vfs_tmpfile(&init_user_ns, dentry, mode, 0);
 	int err = PTR_ERR_OR_ZERO(ret);
 
 	pr_debug("tmpfile(%pd2, 0%o) = %i\n", dentry, mode, err);
include/linux/fs.h +15 -10
index e0e27e4dd817..e74995558c31 100644
@@ -1767,17 +1767,20 @@ extern bool inode_owner_or_capable(struct user_namespace *mnt_userns,
 /*
  * VFS helper functions..
  */
-extern int vfs_create(struct inode *, struct dentry *, umode_t, bool);
-extern int vfs_mkdir(struct inode *, struct dentry *, umode_t);
-extern int vfs_mknod(struct inode *, struct dentry *, umode_t, dev_t);
-extern int vfs_symlink(struct inode *, struct dentry *, const char *);
-extern int vfs_link(struct dentry *, struct inode *, struct dentry *, struct inode **);
-extern int vfs_rmdir(struct inode *, struct dentry *);
-extern int vfs_unlink(struct inode *, struct dentry *, struct inode **);
+extern int vfs_create(struct user_namespace *, struct inode *, struct dentry *, umode_t, bool);
+extern int vfs_mkdir(struct user_namespace *, struct inode *, struct dentry *, umode_t);
+extern int vfs_mknod(struct user_namespace *, struct inode *, struct dentry *, umode_t, dev_t);
+extern int vfs_symlink(struct user_namespace *, struct inode *, struct dentry *, const char *);
+extern int vfs_link(struct dentry *, struct user_namespace *, struct inode *,
+		    struct dentry *, struct inode **);
+extern int vfs_rmdir(struct user_namespace *, struct inode *, struct dentry *);
+extern int vfs_unlink(struct user_namespace *, struct inode *, struct dentry *, struct inode **);
 
 struct renamedata {
+	struct user_namespace *old_mnt_userns;
 	struct inode *old_dir;
 	struct dentry *old_dentry;
+	struct user_namespace *new_mnt_userns;
 	struct inode *new_dir;
 	struct dentry *new_dentry;
 	struct inode **delegated_inode;
@@ -1786,12 +1789,14 @@ struct renamedata {
 
 extern int vfs_rename(struct renamedata *);
 
-static inline int vfs_whiteout(struct inode *dir, struct dentry *dentry)
+static inline int vfs_whiteout(struct user_namespace *mnt_userns,
+			       struct inode *dir, struct dentry *dentry)
 {
-	return vfs_mknod(dir, dentry, S_IFCHR | WHITEOUT_MODE, WHITEOUT_DEV);
+	return vfs_mknod(mnt_userns, dir, dentry, S_IFCHR | WHITEOUT_MODE, WHITEOUT_DEV);
 }
 
-extern struct dentry *vfs_tmpfile(struct dentry *dentry, umode_t mode,
+extern struct dentry *vfs_tmpfile(struct user_namespace *mnt_userns,
+				  struct dentry *dentry, umode_t mode,
 				  int open_flag);
 
 int vfs_mkobj(struct dentry *, umode_t,
ipc/mqueue.c +1 -1
index 693f01fe1216..8b8c300854db 100644
@@ -965,7 +965,7 @@ SYSCALL_DEFINE1(mq_unlink, const char __user *, u_name)
 		err = -ENOENT;
 	} else {
 		ihold(inode);
-		err = vfs_unlink(d_inode(dentry->d_parent), dentry, NULL);
+		err = vfs_unlink(&init_user_ns, d_inode(dentry->d_parent), dentry, NULL);
 	}
 	dput(dentry);
 
net/unix/af_unix.c +1 -1
index f568526d4a02..b4987805e5e5 100644
@@ -996,7 +996,7 @@ static int unix_mknod(const char *sun_path, umode_t mode, struct path *res)
 	 */
 	err = security_path_mknod(&path, dentry, mode, 0);
 	if (!err) {
-		err = vfs_mknod(d_inode(path.dentry), dentry, mode, 0);
+		err = vfs_mknod(&init_user_ns, d_inode(path.dentry), dentry, mode, 0);
 		if (!err) {
 			res->mnt = mntget(path.mnt);
 			res->dentry = dget(dentry);
Signature · 1 line
2.30.0
#37 lore raw

PATCH v5 22/42 open: handle idmapped mounts in do_truncate()

In reply to Christian Brauner

When truncating files the vfs will verify that the caller is privileged over the inode. Extend it to handle idmapped mounts. If the inode is accessed through an idmapped mount it is mapped according to the mount's user namespace. Afterwards the permissions 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.

---
/* v2 */
unchanged

/* v3 */
unchanged

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

- Christoph Hellwig <[email protected]>:
- Use new file_userns_helper().

---
 fs/coredump.c      | 14 ++++++++++----
 fs/inode.c         | 13 +++++++++----
 fs/namei.c         |  6 +++---
 fs/open.c          | 21 +++++++++++++--------
 include/linux/fs.h |  4 ++--
 5 files changed, 37 insertions(+), 21 deletions(-)
fs/coredump.c +10 -4
index a2f6ecc8e345..929d3f3bbe21 100644
@@ -703,6 +703,7 @@ void do_coredump(const kernel_siginfo_t *siginfo)
 			goto close_fail;
 		}
 	} else {
+		struct user_namespace *mnt_userns;
 		struct inode *inode;
 		int open_flags = O_CREAT | O_RDWR | O_NOFOLLOW |
 				 O_LARGEFILE | O_EXCL;
@@ -780,13 +781,14 @@ void do_coredump(const kernel_siginfo_t *siginfo)
 		 * a process dumps core while its cwd is e.g. on a vfat
 		 * filesystem.
 		 */
-		if (!uid_eq(inode->i_uid, current_fsuid()))
+		mnt_userns = file_user_ns(cprm.file);
+		if (!uid_eq(i_uid_into_mnt(mnt_userns, inode), current_fsuid()))
 			goto close_fail;
 		if ((inode->i_mode & 0677) != 0600)
 			goto close_fail;
 		if (!(cprm.file->f_mode & FMODE_CAN_WRITE))
 			goto close_fail;
-		if (do_truncate(cprm.file->f_path.dentry, 0, 0, cprm.file))
+		if (do_truncate(mnt_userns, cprm.file->f_path.dentry, 0, 0, cprm.file))
 			goto close_fail;
 	}
 
@@ -930,8 +932,12 @@ void dump_truncate(struct coredump_params *cprm)
 
 	if (file->f_op->llseek && file->f_op->llseek != no_llseek) {
 		offset = file->f_op->llseek(file, 0, SEEK_CUR);
-		if (i_size_read(file->f_mapping->host) < offset)
-			do_truncate(file->f_path.dentry, offset, 0, file);
+		if (i_size_read(file->f_mapping->host) < offset) {
+			struct user_namespace *mnt_userns;
+
+			mnt_userns = file_user_ns(file);
+			do_truncate(mnt_userns, file->f_path.dentry, offset, 0, file);
+		}
 	}
 }
 EXPORT_SYMBOL(dump_truncate);
fs/inode.c +9 -4
index 46116ef44c9f..aefcfd31a9c7 100644
@@ -1903,7 +1903,8 @@ int dentry_needs_remove_privs(struct dentry *dentry)
 	return mask;
 }
 
-static int __remove_privs(struct dentry *dentry, int kill)
+static int __remove_privs(struct user_namespace *mnt_userns, struct dentry *dentry,
+			  int kill)
 {
 	struct iattr newattrs;
 
@@ -1912,7 +1913,7 @@ static int __remove_privs(struct dentry *dentry, int kill)
 	 * Note we call this on write, so notify_change will not
 	 * encounter any conflicting delegations:
 	 */
-	return notify_change(&init_user_ns, dentry, &newattrs, NULL);
+	return notify_change(mnt_userns, dentry, &newattrs, NULL);
 }
 
 /*
@@ -1938,8 +1939,12 @@ int file_remove_privs(struct file *file)
 	kill = dentry_needs_remove_privs(dentry);
 	if (kill < 0)
 		return kill;
-	if (kill)
-		error = __remove_privs(dentry, kill);
+	if (kill) {
+		struct user_namespace *mnt_userns;
+
+		mnt_userns = file_user_ns(file);
+		error = __remove_privs(mnt_userns, dentry, kill);
+	}
 	if (!error)
 		inode_has_no_xattr(inode);
 
fs/namei.c +3 -3
index e085831c6b85..016d52fda656 100644
@@ -3013,9 +3013,9 @@ static int handle_truncate(struct file *filp)
 	if (!error)
 		error = security_path_truncate(path);
 	if (!error) {
-		error = do_truncate(path->dentry, 0,
-				    ATTR_MTIME|ATTR_CTIME|ATTR_OPEN,
-				    filp);
+		error = do_truncate(file_user_ns(filp),
+				    path->dentry, 0,
+				    ATTR_MTIME | ATTR_CTIME | ATTR_OPEN, filp);
 	}
 	put_write_access(inode);
 	return error;
fs/open.c +13 -8
index bae00ee63af8..a9f3a3b46ef1 100644
@@ -35,8 +35,8 @@
 
 #include "internal.h"
 
-int do_truncate(struct dentry *dentry, loff_t length, unsigned int time_attrs,
-	struct file *filp)
+int do_truncate(struct user_namespace *mnt_userns, struct dentry *dentry,
+		loff_t length, unsigned int time_attrs, struct file *filp)
 {
 	int ret;
 	struct iattr newattrs;
@@ -61,13 +61,14 @@ int do_truncate(struct dentry *dentry, loff_t length, unsigned int time_attrs,
 
 	inode_lock(dentry->d_inode);
 	/* Note any delegations or leases have already been broken: */
-	ret = notify_change(&init_user_ns, dentry, &newattrs, NULL);
+	ret = notify_change(mnt_userns, dentry, &newattrs, NULL);
 	inode_unlock(dentry->d_inode);
 	return ret;
 }
 
 long vfs_truncate(const struct path *path, loff_t length)
 {
+	struct user_namespace *mnt_userns;
 	struct inode *inode;
 	long error;
 
@@ -83,7 +84,8 @@ long vfs_truncate(const struct path *path, loff_t length)
 	if (error)
 		goto out;
 
-	error = inode_permission(&init_user_ns, inode, MAY_WRITE);
+	mnt_userns = mnt_user_ns(path->mnt);
+	error = inode_permission(mnt_userns, inode, MAY_WRITE);
 	if (error)
 		goto mnt_drop_write_and_out;
 
@@ -107,7 +109,7 @@ long vfs_truncate(const struct path *path, loff_t length)
 	if (!error)
 		error = security_path_truncate(path);
 	if (!error)
-		error = do_truncate(path->dentry, length, 0, NULL);
+		error = do_truncate(mnt_userns, path->dentry, length, 0, NULL);
 
 put_write_and_out:
 	put_write_access(inode);
@@ -186,13 +188,16 @@ long do_sys_ftruncate(unsigned int fd, loff_t length, int small)
 	/* Check IS_APPEND on real upper inode */
 	if (IS_APPEND(file_inode(f.file)))
 		goto out_putf;
-
 	sb_start_write(inode->i_sb);
 	error = locks_verify_truncate(inode, f.file, length);
 	if (!error)
 		error = security_path_truncate(&f.file->f_path);
-	if (!error)
-		error = do_truncate(dentry, length, ATTR_MTIME|ATTR_CTIME, f.file);
+	if (!error) {
+		struct user_namespace *mnt_userns;
+
+		mnt_userns = file_user_ns(f.file);
+		error = do_truncate(mnt_userns, dentry, length, ATTR_MTIME | ATTR_CTIME, f.file);
+	}
 	sb_end_write(inode->i_sb);
 out_putf:
 	fdput(f);
include/linux/fs.h +2 -2
index e74995558c31..168b8d3f5afb 100644
@@ -2584,8 +2584,8 @@ struct filename {
 static_assert(offsetof(struct filename, iname) % sizeof(long) == 0);
 
 extern long vfs_truncate(const struct path *, loff_t);
-extern int do_truncate(struct dentry *, loff_t start, unsigned int time_attrs,
-		       struct file *filp);
+extern int do_truncate(struct user_namespace *, struct dentry *, loff_t start,
+		       unsigned int time_attrs, struct file *filp);
 extern int vfs_fallocate(struct file *file, int mode, loff_t offset,
 			loff_t len);
 extern long do_sys_open(int dfd, const char __user *filename, int flags,
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
#38 lore raw

Re: [PATCH v5 22/42] open: handle idmapped mounts in do_truncate()

fs/coredump.c void dump_truncate(struct coredump_params *cprm) · line 940

Quoted 12 lines · Christian Brauner
fs/coredump.c +6 -2
@@ -930,8 +932,12 @@ void dump_truncate(struct coredump_params *cprm)
 
 	if (file->f_op->llseek && file->f_op->llseek != no_llseek) {
 		offset = file->f_op->llseek(file, 0, SEEK_CUR);
-		if (i_size_read(file->f_mapping->host) < offset)
-			do_truncate(file->f_path.dentry, offset, 0, file);
+		if (i_size_read(file->f_mapping->host) < offset) {
+			struct user_namespace *mnt_userns;
+
+			mnt_userns = file_user_ns(file);
+			do_truncate(mnt_userns, file->f_path.dentry, offset, 0, file);
+		}

I think we can skip the local variable here. In fact for all callers of do_truncate except vfs_truncate a little file_truncate helper that takes a struct file would help readability a lot.
_______________________________________________
Containers mailing list [email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers

#39 lore raw

PATCH v5 23/42 open: handle idmapped mounts

In reply to Christian Brauner

For core file operations such as changing directories or chrooting, determining file access, changing mode or ownership the vfs will verify that the caller is privileged over the inode. Extend the various helpers to handle idmapped mounts. If the inode is accessed through an idmapped mount it is mapped according to the mount's user namespace. Afterwards the permissions checks are identical to non-idmapped mounts. When changing file ownership we need to map the uid and gid from the mount's user namespace. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
unchanged

/* v3 */
- David Howells <[email protected]>:
- Remove mnt_idmapped() check after removing mnt_idmapped() helper in earlier patches.

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

- Christoph Hellwig <[email protected]>:
- Use new file_userns_helper().

---
 fs/open.c | 29 ++++++++++++++++++++++-------
 1 file changed, 22 insertions(+), 7 deletions(-)
fs/open.c +22 -7
index a9f3a3b46ef1..ac26f0a363a0 100644
@@ -401,6 +401,7 @@ static const struct cred *access_override_creds(void)
 
 static long do_faccessat(int dfd, const char __user *filename, int mode, int flags)
 {
+	struct user_namespace *mnt_userns;
 	struct path path;
 	struct inode *inode;
 	int res;
@@ -441,7 +442,8 @@ static long do_faccessat(int dfd, const char __user *filename, int mode, int fla
 			goto out_path_release;
 	}
 
-	res = inode_permission(&init_user_ns, inode, mode | MAY_ACCESS);
+	mnt_userns = mnt_user_ns(path.mnt);
+	res = inode_permission(mnt_userns, inode, mode | MAY_ACCESS);
 	/* SuS v2 requires we report a read only fs too */
 	if (res || !(mode & S_IWOTH) || special_file(inode->i_mode))
 		goto out_path_release;
@@ -489,6 +491,7 @@ SYSCALL_DEFINE2(access, const char __user *, filename, int, mode)
 
 SYSCALL_DEFINE1(chdir, const char __user *, filename)
 {
+	struct user_namespace *mnt_userns;
 	struct path path;
 	int error;
 	unsigned int lookup_flags = LOOKUP_FOLLOW | LOOKUP_DIRECTORY;
@@ -497,7 +500,8 @@ SYSCALL_DEFINE1(chdir, const char __user *, filename)
 	if (error)
 		goto out;
 
-	error = inode_permission(&init_user_ns, path.dentry->d_inode, MAY_EXEC | MAY_CHDIR);
+	mnt_userns = mnt_user_ns(path.mnt);
+	error = inode_permission(mnt_userns, path.dentry->d_inode, MAY_EXEC | MAY_CHDIR);
 	if (error)
 		goto dput_and_out;
 
@@ -515,6 +519,7 @@ SYSCALL_DEFINE1(chdir, const char __user *, filename)
 
 SYSCALL_DEFINE1(fchdir, unsigned int, fd)
 {
+	struct user_namespace *mnt_userns;
 	struct fd f = fdget_raw(fd);
 	int error;
 
@@ -526,7 +531,8 @@ SYSCALL_DEFINE1(fchdir, unsigned int, fd)
 	if (!d_can_lookup(f.file->f_path.dentry))
 		goto out_putf;
 
-	error = inode_permission(&init_user_ns, file_inode(f.file), MAY_EXEC | MAY_CHDIR);
+	mnt_userns = file_user_ns(f.file);
+	error = inode_permission(mnt_userns, file_inode(f.file), MAY_EXEC | MAY_CHDIR);
 	if (!error)
 		set_fs_pwd(current->fs, &f.file->f_path);
 out_putf:
@@ -537,6 +543,7 @@ SYSCALL_DEFINE1(fchdir, unsigned int, fd)
 
 SYSCALL_DEFINE1(chroot, const char __user *, filename)
 {
+	struct user_namespace *mnt_userns;
 	struct path path;
 	int error;
 	unsigned int lookup_flags = LOOKUP_FOLLOW | LOOKUP_DIRECTORY;
@@ -545,7 +552,8 @@ SYSCALL_DEFINE1(chroot, const char __user *, filename)
 	if (error)
 		goto out;
 
-	error = inode_permission(&init_user_ns, path.dentry->d_inode, MAY_EXEC | MAY_CHDIR);
+	mnt_userns = mnt_user_ns(path.mnt);
+	error = inode_permission(mnt_userns, path.dentry->d_inode, MAY_EXEC | MAY_CHDIR);
 	if (error)
 		goto dput_and_out;
 
@@ -570,6 +578,7 @@ SYSCALL_DEFINE1(chroot, const char __user *, filename)
 
 int chmod_common(const struct path *path, umode_t mode)
 {
+	struct user_namespace *mnt_userns;
 	struct inode *inode = path->dentry->d_inode;
 	struct inode *delegated_inode = NULL;
 	struct iattr newattrs;
@@ -585,7 +594,8 @@ int chmod_common(const struct path *path, umode_t mode)
 		goto out_unlock;
 	newattrs.ia_mode = (mode & S_IALLUGO) | (inode->i_mode & ~S_IALLUGO);
 	newattrs.ia_valid = ATTR_MODE | ATTR_CTIME;
-	error = notify_change(&init_user_ns, path->dentry, &newattrs, &delegated_inode);
+	mnt_userns = mnt_user_ns(path->mnt);
+	error = notify_change(mnt_userns, path->dentry, &newattrs, &delegated_inode);
 out_unlock:
 	inode_unlock(inode);
 	if (delegated_inode) {
@@ -646,6 +656,7 @@ SYSCALL_DEFINE2(chmod, const char __user *, filename, umode_t, mode)
 
 int chown_common(const struct path *path, uid_t user, gid_t group)
 {
+	struct user_namespace *mnt_userns;
 	struct inode *inode = path->dentry->d_inode;
 	struct inode *delegated_inode = NULL;
 	int error;
@@ -656,6 +667,10 @@ int chown_common(const struct path *path, uid_t user, gid_t group)
 	uid = make_kuid(current_user_ns(), user);
 	gid = make_kgid(current_user_ns(), group);
 
+	mnt_userns = mnt_user_ns(path->mnt);
+	uid = kuid_from_mnt(mnt_userns, uid);
+	gid = kgid_from_mnt(mnt_userns, gid);
+
 retry_deleg:
 	newattrs.ia_valid =  ATTR_CTIME;
 	if (user != (uid_t) -1) {
@@ -676,7 +691,7 @@ int chown_common(const struct path *path, uid_t user, gid_t group)
 	inode_lock(inode);
 	error = security_path_chown(path, uid, gid);
 	if (!error)
-		error = notify_change(&init_user_ns, path->dentry, &newattrs, &delegated_inode);
+		error = notify_change(mnt_userns, path->dentry, &newattrs, &delegated_inode);
 	inode_unlock(inode);
 	if (delegated_inode) {
 		error = break_deleg_wait(&delegated_inode);
@@ -1137,7 +1152,7 @@ struct file *filp_open(const char *filename, int flags, umode_t mode)
 {
 	struct filename *name = getname_kernel(filename);
 	struct file *file = ERR_CAST(name);
-	
+
 	if (!IS_ERR(name)) {
 		file = file_open_name(name, flags, mode);
 		putname(name);
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
#42 lore raw

PATCH v5 24/42 af_unix: handle idmapped mounts

In reply to Christian Brauner

When binding a non-abstract AF_UNIX socket it will gain a representation in the filesystem. Enable the socket infrastructure to handle idmapped mounts by passing down the user namespace of the mount the socket will be created from. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
unchanged

/* v3 */
unchanged

/* v4 */
unchanged

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 net/unix/af_unix.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/unix/af_unix.c +1 -1
index b4987805e5e5..4be33240e9cc 100644
@@ -996,7 +996,7 @@ static int unix_mknod(const char *sun_path, umode_t mode, struct path *res)
 	 */
 	err = security_path_mknod(&path, dentry, mode, 0);
 	if (!err) {
-		err = vfs_mknod(&init_user_ns, d_inode(path.dentry), dentry, mode, 0);
+		err = vfs_mknod(mnt_user_ns(path.mnt), d_inode(path.dentry), dentry, mode, 0);
 		if (!err) {
 			res->mnt = mntget(path.mnt);
 			res->dentry = dget(dentry);
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
#44 lore raw

PATCH v5 25/42 utimes: handle idmapped mounts

In reply to Christian Brauner

Enable the vfs_utimes() helper to handle idmapped mounts by passing down the mount's user namespace. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
unchanged

/* v3 */
unchanged

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/utimes.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
fs/utimes.c +3 -1
index 1a4130bee157..33fa753bab3a 100644
@@ -22,6 +22,7 @@ int vfs_utimes(const struct path *path, struct timespec64 *times)
 	struct iattr newattrs;
 	struct inode *inode = path->dentry->d_inode;
 	struct inode *delegated_inode = NULL;
+	struct user_namespace *mnt_userns;
 
 	if (times) {
 		if (!nsec_valid(times[0].tv_nsec) ||
@@ -61,8 +62,9 @@ int vfs_utimes(const struct path *path, struct timespec64 *times)
 		newattrs.ia_valid |= ATTR_TOUCH;
 	}
 retry_deleg:
+	mnt_userns = mnt_user_ns(path->mnt);
 	inode_lock(inode);
-	error = notify_change(&init_user_ns, path->dentry, &newattrs, &delegated_inode);
+	error = notify_change(mnt_userns, path->dentry, &newattrs, &delegated_inode);
 	inode_unlock(inode);
 	if (delegated_inode) {
 		error = break_deleg_wait(&delegated_inode);
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
Reviewed-by on v5 25/42
#45 lore raw

Re: [PATCH v5 25/42] utimes: handle idmapped mounts

Quoted 4 lines · Christian Brauner

Enable the vfs_utimes() helper to handle idmapped mounts by passing down the mount's user namespace. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

No real need for the local variable here. Same applies to a bunch of other patches as well.

Otherwise looks good:

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers

#46 lore raw

PATCH v5 26/42 fcntl: handle idmapped mounts

In reply to Christian Brauner

Enable the setfl() helper to handle idmapped mounts by passing down the mount's user namespace. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
patch introduced

/* v3 */
unchanged

/* v4 */
unchanged

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/fcntl.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
fs/fcntl.c +2 -1
index 74d99731fd43..58706031e603 100644
@@ -25,6 +25,7 @@
 #include <linux/user_namespace.h>
 #include <linux/memfd.h>
 #include <linux/compat.h>
+#include <linux/mount.h>
 
 #include <linux/poll.h>
 #include <asm/siginfo.h>
@@ -46,7 +47,7 @@ static int setfl(int fd, struct file * filp, unsigned long arg)
 
 	/* O_NOATIME can only be set by the owner or superuser */
 	if ((arg & O_NOATIME) && !(filp->f_flags & O_NOATIME))
-		if (!inode_owner_or_capable(&init_user_ns, inode))
+		if (!inode_owner_or_capable(mnt_user_ns(filp->f_path.mnt), inode))
 			return -EPERM;
 
 	/* required for strict SunOS emulation */
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
#48 lore raw

PATCH v5 27/42 notify: handle idmapped mounts

In reply to Christian Brauner

Enable notify implementations to handle idmapped mounts by passing down the mount's user namespace. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
patch introduced

/* v3 */
unchanged

/* v4 */
unchanged

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/notify/fanotify/fanotify_user.c | 2 +-
 fs/notify/inotify/inotify_user.c   | 3 ++-
 2 files changed, 3 insertions(+), 2 deletions(-)
fs/notify/fanotify/fanotify_user.c +1 -1
index 4ca13fb33c7c..c60b8cc1f94d 100644
@@ -702,7 +702,7 @@ static int fanotify_find_path(int dfd, const char __user *filename,
 	}
 
 	/* you can only watch an inode if you have read permissions on it */
-	ret = inode_permission(&init_user_ns, path->dentry->d_inode, MAY_READ);
+	ret = inode_permission(mnt_user_ns(path->mnt), path->dentry->d_inode, MAY_READ);
 	if (ret) {
 		path_put(path);
 		goto out;
fs/notify/inotify/inotify_user.c +2 -1
index 18e7024207a9..cecff22549d7 100644
@@ -31,6 +31,7 @@
 #include <linux/wait.h>
 #include <linux/memcontrol.h>
 #include <linux/security.h>
+#include <linux/mount.h>
 
 #include "inotify.h"
 #include "../fdinfo.h"
@@ -352,7 +353,7 @@ static int inotify_find_inode(const char __user *dirname, struct path *path,
 	if (error)
 		return error;
 	/* you can only watch an inode if you have read permissions on it */
-	error = inode_permission(&init_user_ns, path->dentry->d_inode, MAY_READ);
+	error = inode_permission(mnt_user_ns(path->mnt), path->dentry->d_inode, MAY_READ);
 	if (error) {
 		path_put(path);
 		return error;
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
#50 lore raw

PATCH v5 28/42 init: handle idmapped mounts

In reply to Christian Brauner

Enable the init helpers to handle idmapped mounts by passing down the mount's user namespace. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
patch introduced

/* v3 */
unchanged

/* v4 */
unchanged

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/init.c | 24 ++++++++++++++----------
 1 file changed, 14 insertions(+), 10 deletions(-)
fs/init.c +14 -10
index 06d8e52ce18e..1db2b392e4e6 100644
@@ -49,7 +49,7 @@ int __init init_chdir(const char *filename)
 	error = kern_path(filename, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &path);
 	if (error)
 		return error;
-	error = inode_permission(&init_user_ns, path.dentry->d_inode,
+	error = inode_permission(mnt_user_ns(path.mnt), path.dentry->d_inode,
 				 MAY_EXEC | MAY_CHDIR);
 	if (!error)
 		set_fs_pwd(current->fs, &path);
@@ -65,7 +65,7 @@ int __init init_chroot(const char *filename)
 	error = kern_path(filename, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &path);
 	if (error)
 		return error;
-	error = inode_permission(&init_user_ns, path.dentry->d_inode,
+	error = inode_permission(mnt_user_ns(path.mnt), path.dentry->d_inode,
 				 MAY_EXEC | MAY_CHDIR);
 	if (error)
 		goto dput_and_out;
@@ -120,7 +120,7 @@ int __init init_eaccess(const char *filename)
 	error = kern_path(filename, LOOKUP_FOLLOW, &path);
 	if (error)
 		return error;
-	error = inode_permission(&init_user_ns, d_inode(path.dentry),
+	error = inode_permission(mnt_user_ns(path.mnt), d_inode(path.dentry),
 				 MAY_ACCESS);
 	path_put(&path);
 	return error;
@@ -160,8 +160,8 @@ int __init init_mknod(const char *filename, umode_t mode, unsigned int dev)
 		mode &= ~current_umask();
 	error = security_path_mknod(&path, dentry, mode, dev);
 	if (!error)
-		error = vfs_mknod(&init_user_ns, path.dentry->d_inode, dentry,
-				  mode, new_decode_dev(dev));
+		error = vfs_mknod(mnt_user_ns(path.mnt), path.dentry->d_inode,
+				  dentry, mode, new_decode_dev(dev));
 	done_path_create(&path, dentry);
 	return error;
 }
@@ -170,6 +170,7 @@ int __init init_link(const char *oldname, const char *newname)
 {
 	struct dentry *new_dentry;
 	struct path old_path, new_path;
+	struct user_namespace *mnt_userns;
 	int error;
 
 	error = kern_path(oldname, 0, &old_path);
@@ -184,14 +185,15 @@ int __init init_link(const char *oldname, const char *newname)
 	error = -EXDEV;
 	if (old_path.mnt != new_path.mnt)
 		goto out_dput;
-	error = may_linkat(&init_user_ns, &old_path);
+	mnt_userns = mnt_user_ns(new_path.mnt);
+	error = may_linkat(mnt_userns, &old_path);
 	if (unlikely(error))
 		goto out_dput;
 	error = security_path_link(old_path.dentry, &new_path, new_dentry);
 	if (error)
 		goto out_dput;
-	error = vfs_link(old_path.dentry, &init_user_ns,
-			 new_path.dentry->d_inode, new_dentry, NULL);
+	error = vfs_link(old_path.dentry, mnt_userns, new_path.dentry->d_inode,
+			 new_dentry, NULL);
 out_dput:
 	done_path_create(&new_path, new_dentry);
 out:
@@ -210,7 +212,8 @@ int __init init_symlink(const char *oldname, const char *newname)
 		return PTR_ERR(dentry);
 	error = security_path_symlink(&path, dentry, oldname);
 	if (!error)
-		error = vfs_symlink(&init_user_ns, path.dentry->d_inode, dentry, oldname);
+		error = vfs_symlink(mnt_user_ns(path.mnt), path.dentry->d_inode,
+				    dentry, oldname);
 	done_path_create(&path, dentry);
 	return error;
 }
@@ -233,7 +236,8 @@ int __init init_mkdir(const char *pathname, umode_t mode)
 		mode &= ~current_umask();
 	error = security_path_mkdir(&path, dentry, mode);
 	if (!error)
-		error = vfs_mkdir(&init_user_ns, path.dentry->d_inode, dentry, mode);
+		error = vfs_mkdir(mnt_user_ns(path.mnt), path.dentry->d_inode,
+				  dentry, mode);
 	done_path_create(&path, dentry);
 	return error;
 }
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers
#51 lore raw

PATCH v5 29/42 ioctl: handle idmapped mounts

In reply to Christian Brauner

Enable generic ioctls to handle idmapped mounts by passing down the mount's user namespace. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
patch introduced

/* v3 */
unchanged

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

- Christoph Hellwig <[email protected]>:
- Use new file_userns_helper().

---
 fs/remap_range.c   | 7 +++++--
 fs/verity/enable.c | 2 +-
 2 files changed, 6 insertions(+), 3 deletions(-)
fs/remap_range.c +5 -2
index 29a4a4dbfe12..dc68394ec9a1 100644
@@ -432,13 +432,16 @@ EXPORT_SYMBOL(vfs_clone_file_range);
 /* Check whether we are allowed to dedupe the destination file */
 static bool allow_file_dedupe(struct file *file)
 {
+	struct user_namespace *mnt_userns = file_user_ns(file);
+	struct inode *inode = file_inode(file);
+
 	if (capable(CAP_SYS_ADMIN))
 		return true;
 	if (file->f_mode & FMODE_WRITE)
 		return true;
-	if (uid_eq(current_fsuid(), file_inode(file)->i_uid))
+	if (uid_eq(current_fsuid(), i_uid_into_mnt(mnt_userns, inode)))
 		return true;
-	if (!inode_permission(&init_user_ns, file_inode(file), MAY_WRITE))
+	if (!inode_permission(mnt_userns, inode, MAY_WRITE))
 		return true;
 	return false;
 }
fs/verity/enable.c +1 -1
index 6809cf8a99b7..9a221e368fa6 100644
@@ -369,7 +369,7 @@ int fsverity_ioctl_enable(struct file *filp, const void __user *uarg)
 	 * has verity enabled, and to stabilize the data being hashed.
 	 */
 
-	err = inode_permission(&init_user_ns, inode, MAY_WRITE);
+	err = inode_permission(file_user_ns(filp), inode, MAY_WRITE);
 	if (err)
 		return err;
 
Signature · 1 line
2.30.0
Reviewed-by on v5 29/42
#52 lore raw

Re: [PATCH v5 29/42] ioctl: handle idmapped mounts

Quoted 3 lines · Christian Brauner

Enable generic ioctls to handle idmapped mounts by passing down the mount's user namespace. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

Looks good, although file_permission would help here as well..

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers

#53 lore raw

PATCH v5 30/42 would_dump: handle idmapped mounts

In reply to Christian Brauner

When determining whether or not to create a coredump the vfs will verify that the caller is privileged over the inode. Make the would_dump() helper handle idmapped mounts by passing down the mount's user namespace of the exec file. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

---
/* v2 */
unchanged

/* v3 */
unchanged

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

- Christoph Hellwig <[email protected]>:
- Use new file_userns_helper().

---
 fs/exec.c  | 5 +++--
 fs/fcntl.c | 2 +-
 2 files changed, 4 insertions(+), 3 deletions(-)
fs/exec.c +3 -2
index 049a8d2669db..36a1b927a8ae 100644
@@ -1404,14 +1404,15 @@ EXPORT_SYMBOL(begin_new_exec);
 void would_dump(struct linux_binprm *bprm, struct file *file)
 {
 	struct inode *inode = file_inode(file);
-	if (inode_permission(&init_user_ns, inode, MAY_READ) < 0) {
+	struct user_namespace *mnt_userns = file_user_ns(file);
+	if (inode_permission(mnt_userns, inode, MAY_READ) < 0) {
 		struct user_namespace *old, *user_ns;
 		bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP;
 
 		/* Ensure mm->user_ns contains the executable */
 		user_ns = old = bprm->mm->user_ns;
 		while ((user_ns != &init_user_ns) &&
-		       !privileged_wrt_inode_uidgid(user_ns, &init_user_ns, inode))
+		       !privileged_wrt_inode_uidgid(user_ns, mnt_userns, inode))
 			user_ns = user_ns->parent;
 
 		if (old != user_ns) {
fs/fcntl.c +1 -1
index 58706031e603..ebbe831d5736 100644
@@ -47,7 +47,7 @@ static int setfl(int fd, struct file * filp, unsigned long arg)
 
 	/* O_NOATIME can only be set by the owner or superuser */
 	if ((arg & O_NOATIME) && !(filp->f_flags & O_NOATIME))
-		if (!inode_owner_or_capable(mnt_user_ns(filp->f_path.mnt), inode))
+		if (!inode_owner_or_capable(file_user_ns(filp), inode))
 			return -EPERM;
 
 	/* required for strict SunOS emulation */
Signature · 1 line
2.30.0
Reviewed-by on v5 30/42
#54 lore raw

Re: [PATCH v5 30/42] would_dump: handle idmapped mounts

Quoted 5 lines · Christian Brauner

When determining whether or not to create a coredump the vfs will verify that the caller is privileged over the inode. Make the would_dump() helper handle idmapped mounts by passing down the mount's user namespace of the exec file. If the initial user namespace is passed nothing changes so non-idmapped mounts will see identical behavior as before.

Looks good,

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers

#55 lore raw

PATCH v5 31/42 exec: handle idmapped mounts

In reply to Christian Brauner

When executing a setuid binary the kernel will verify in bprm_fill_uid() that the inode has a mapping in the caller's user namespace before setting the callers uid and gid. Let bprm_fill_uid() handle idmapped mounts. If the inode is accessed through an idmapped mount it is mapped according to 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.

---
/* v2 */
unchanged

/* v3 */
unchanged

/* v4 */
- Serge Hallyn <[email protected]>:
- Use "mnt_userns" to refer to a vfsmount's userns everywhere to make terminology consistent.

/* v5 */
base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837

---
 fs/exec.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
fs/exec.c +5 -2
index 36a1b927a8ae..61ae826d5a16 100644
@@ -1580,6 +1580,7 @@ static void check_unsafe_exec(struct linux_binprm *bprm)
 static void bprm_fill_uid(struct linux_binprm *bprm, struct file *file)
 {
 	/* Handle suid and sgid on files */
+	struct user_namespace *mnt_userns;
 	struct inode *inode;
 	unsigned int mode;
 	kuid_t uid;
@@ -1596,13 +1597,15 @@ static void bprm_fill_uid(struct linux_binprm *bprm, struct file *file)
 	if (!(mode & (S_ISUID|S_ISGID)))
 		return;
 
+	mnt_userns = mnt_user_ns(file->f_path.mnt);
+
 	/* Be careful if suid/sgid is set */
 	inode_lock(inode);
 
 	/* reload atomically mode/uid/gid now that lock held */
 	mode = inode->i_mode;
-	uid = inode->i_uid;
-	gid = inode->i_gid;
+	uid = i_uid_into_mnt(mnt_userns, inode);
+	gid = i_gid_into_mnt(mnt_userns, inode);
 	inode_unlock(inode);
 
 	/* We ignore suid/sgid if there are no mappings for them in the ns */
Signature · 6 lines
2.30.0

_______________________________________________
Containers mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/containers