diff options
| author | Miklos Szeredi <mszeredi@redhat.com> | 2023-10-25 16:02:03 +0200 |
|---|---|---|
| committer | Christian Brauner <brauner@kernel.org> | 2023-12-14 11:49:17 +0100 |
| commit | b4c2bea8ceaa50cd42a8f73667389d801a3ecf2d (patch) | |
| tree | 840f95317cfa1468b784232fa0e97bd5918adea9 /include/linux/syscalls.h | |
| parent | 68385d77c05b401f748acecc0e0a8eff10489334 (diff) | |
add listmount(2) syscall
Add way to query the children of a particular mount. This is a more
flexible way to iterate the mount tree than having to parse
/proc/self/mountinfo.
Lookup the mount by the new 64bit mount ID. If a mount needs to be
queried based on path, then statx(2) can be used to first query the
mount ID belonging to the path.
Return an array of new (64bit) mount ID's. Without privileges only
mounts are listed which are reachable from the task's root.
Folded into this patch are several later improvements. Keeping them
separate would make the history pointlessly confusing:
* Recursive listing of mounts is the default now (cf. [1]).
* Remove explicit LISTMOUNT_UNREACHABLE flag (cf. [1]) and fail if mount
is unreachable from current root. This also makes permission checking
consistent with statmount() (cf. [3]).
* Start listing mounts in unique mount ID order (cf. [2]) to allow
continuing listmount() from a midpoint.
* Allow to continue listmount(). The @request_mask parameter is renamed
and to @param to be usable by both statmount() and listmount().
If @param is set to a mount id then listmount() will continue listing
mounts from that id on. This allows listing mounts in multiple
listmount invocations without having to resize the buffer. If @param
is zero then the listing starts from the beginning (cf. [4]).
* Don't return EOVERFLOW, instead return the buffer size which allows to
detect a full buffer as well (cf. [4]).
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Link: https://lore.kernel.org/r/20231025140205.3586473-6-mszeredi@redhat.com
Reviewed-by: Ian Kent <raven@themaw.net>
Link: https://lore.kernel.org/r/20231128160337.29094-2-mszeredi@redhat.com [1] (folded)
Link: https://lore.kernel.org/r/20231128160337.29094-3-mszeredi@redhat.com [2] (folded)
Link: https://lore.kernel.org/r/20231128160337.29094-4-mszeredi@redhat.com [3] (folded)
Link: https://lore.kernel.org/r/20231128160337.29094-5-mszeredi@redhat.com [4] (folded)
[Christian Brauner <brauner@kernel.org>: various smaller fixes]
Signed-off-by: Christian Brauner <brauner@kernel.org>
Diffstat (limited to 'include/linux/syscalls.h')
| -rw-r--r-- | include/linux/syscalls.h | 3 |
1 files changed, 3 insertions, 0 deletions
diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index 530ca9adf5f1..2d6d3e76e3f7 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -412,6 +412,9 @@ asmlinkage long sys_fstatfs64(unsigned int fd, size_t sz, asmlinkage long sys_statmount(const struct mnt_id_req __user *req, struct statmount __user *buf, size_t bufsize, unsigned int flags); +asmlinkage long sys_listmount(const struct mnt_id_req __user *req, + u64 __user *buf, size_t bufsize, + unsigned int flags); asmlinkage long sys_truncate(const char __user *path, long length); asmlinkage long sys_ftruncate(unsigned int fd, unsigned long length); #if BITS_PER_LONG == 32 |