| Message ID | 20241128-work-pidfs-v1-0-80f267639d98@kernel.org (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
On Thu, Nov 28, 2024 at 1:34 PM Christian Brauner <brauner@kernel.org> wrote: > > Hey, > > This reworks the inode number allocation for pidfs in order to support > file handles properly. > > Recently we received a patchset that aims to enable file handle encoding > and decoding via name_to_handle_at(2) and open_by_handle_at(2). > > A crucical step in the patch series is how to go from inode number to > struct pid without leaking information into unprivileged contexts. The > issue is that in order to find a struct pid the pid number in the > initial pid namespace must be encoded into the file handle via > name_to_handle_at(2). This can be used by containers using a separate > pid namespace to learn what the pid number of a given process in the > initial pid namespace is. While this is a weak information leak it could > be used in various exploits and in general is an ugly wart in the > design. > > To solve this problem a new way is needed to lookup a struct pid based > on the inode number allocated for that struct pid. The other part is to > remove the custom inode number allocation on 32bit systems that is also > an ugly wart that should go away. > > So, a new scheme is used that I was discusssing with Tejun some time > back. A cyclic ida is used for the lower 32 bits and a the high 32 bits > are used for the generation number. This gives a 64 bit inode number > that is unique on both 32 bit and 64 bit. The lower 32 bit number is > recycled slowly and can be used to lookup struct pids. > > So after applying the pidfs file handle series at > https://lore.kernel.org/r/20241101135452.19359-1-erin.shepherd@e43.eu on > top of the patches here we should be able to simplify encoding and > decoding to something like: > > diff --git a/fs/pidfs.c b/fs/pidfs.c > index e71294d3d607..a38b833a2d38 100644 > --- a/fs/pidfs.c > +++ b/fs/pidfs.c > @@ -78,7 +78,7 @@ void pidfs_remove_pid(struct pid *pid) > } > > /* Find a struct pid based on the inode number. */ > -static __maybe_unused struct pid *pidfs_ino_get_pid(u64 ino) > +static struct pid *pidfs_ino_get_pid(u64 ino) > { > ino_t pid_ino = pidfs_ino(ino); > u32 gen = pidfs_gen(ino); > @@ -475,49 +475,37 @@ static const struct dentry_operations pidfs_dentry_operations = { > .d_prune = stashed_dentry_prune, > }; > > -#define PIDFD_FID_LEN 3 > - > -struct pidfd_fid { > - u64 ino; > - s32 pid; > -} __packed; > - > -static int pidfs_encode_fh(struct inode *inode, u32 *fh, int *max_len, > +static int pidfs_encode_fh(struct inode *inode, __u32 *fh, int *max_len, > struct inode *parent) > { > struct pid *pid = inode->i_private; > - struct pidfd_fid *fid = (struct pidfd_fid *)fh; > > - if (*max_len < PIDFD_FID_LEN) { > - *max_len = PIDFD_FID_LEN; > + if (*max_len < 2) { > + *max_len = 2; > return FILEID_INVALID; > } > > - fid->ino = pid->ino; > - fid->pid = pid_nr(pid); > - *max_len = PIDFD_FID_LEN; > + *max_len = 2; > + *(u64 *)fh = pid->ino; > return FILEID_INO64_GEN; Semantic remark: /* * 64 bit inode number, 32 bit generation number. */ FILEID_INO64_GEN = 0x81, filesystems are free to abuse the constants and return whatever id they want (e.g. shmem_encode_fh()), but if you want to play by the rules, this would be either: /* * 64 bit unique kernfs id */ FILEID_KERNFS = 0xfe, or: /* * 32bit inode number, 32 bit generation number. */ FILEID_INO32_GEN = 1, which is at least sometimes correct. or define: /* * 64 bit inode number. */ FILEID_INO64 = 0x80, Thanks, Amir.
diff --git a/fs/pidfs.c b/fs/pidfs.c index e71294d3d607..a38b833a2d38 100644 --- a/fs/pidfs.c +++ b/fs/pidfs.c @@ -78,7 +78,7 @@ void pidfs_remove_pid(struct pid *pid) } /* Find a struct pid based on the inode number. */ -static __maybe_unused struct pid *pidfs_ino_get_pid(u64 ino) +static struct pid *pidfs_ino_get_pid(u64 ino) { ino_t pid_ino = pidfs_ino(ino); u32 gen = pidfs_gen(ino); @@ -475,49 +475,37 @@ static const struct dentry_operations pidfs_dentry_operations = { .d_prune = stashed_dentry_prune, }; -#define PIDFD_FID_LEN 3 - -struct pidfd_fid { - u64 ino; - s32 pid; -} __packed; - -static int pidfs_encode_fh(struct inode *inode, u32 *fh, int *max_len, +static int pidfs_encode_fh(struct inode *inode, __u32 *fh, int *max_len, struct inode *parent) { struct pid *pid = inode->i_private; - struct pidfd_fid *fid = (struct pidfd_fid *)fh; - if (*max_len < PIDFD_FID_LEN) { - *max_len = PIDFD_FID_LEN; + if (*max_len < 2) { + *max_len = 2; return FILEID_INVALID; } - fid->ino = pid->ino; - fid->pid = pid_nr(pid); - *max_len = PIDFD_FID_LEN; + *max_len = 2; + *(u64 *)fh = pid->ino; return FILEID_INO64_GEN; } static struct dentry *pidfs_fh_to_dentry(struct super_block *sb, - struct fid *gen_fid, + struct fid *fid, int fh_len, int fh_type) { int ret; struct path path; - struct pidfd_fid *fid = (struct pidfd_fid *)gen_fid; struct pid *pid; + u64 pid_ino; - if (fh_type != FILEID_INO64_GEN || fh_len < PIDFD_FID_LEN) + if (fh_type != FILEID_INO64_GEN || fh_len < 2) return NULL; - scoped_guard(rcu) { - pid = find_pid_ns(fid->pid, &init_pid_ns); - if (!pid || pid->ino != fid->ino || pid_vnr(pid) == 0) - return NULL; - - pid = get_pid(pid); - } + pid_ino = *(u64 *)fid; + pid = pidfs_ino_get_pid(pid_ino); + if (!pid) + return NULL; ret = path_from_stashed(&pid->stashed, pidfs_mnt, pid, &path); if (ret < 0)
Hey, This reworks the inode number allocation for pidfs in order to support file handles properly. Recently we received a patchset that aims to enable file handle encoding and decoding via name_to_handle_at(2) and open_by_handle_at(2). A crucical step in the patch series is how to go from inode number to struct pid without leaking information into unprivileged contexts. The issue is that in order to find a struct pid the pid number in the initial pid namespace must be encoded into the file handle via name_to_handle_at(2). This can be used by containers using a separate pid namespace to learn what the pid number of a given process in the initial pid namespace is. While this is a weak information leak it could be used in various exploits and in general is an ugly wart in the design. To solve this problem a new way is needed to lookup a struct pid based on the inode number allocated for that struct pid. The other part is to remove the custom inode number allocation on 32bit systems that is also an ugly wart that should go away. So, a new scheme is used that I was discusssing with Tejun some time back. A cyclic ida is used for the lower 32 bits and a the high 32 bits are used for the generation number. This gives a 64 bit inode number that is unique on both 32 bit and 64 bit. The lower 32 bit number is recycled slowly and can be used to lookup struct pids. So after applying the pidfs file handle series at https://lore.kernel.org/r/20241101135452.19359-1-erin.shepherd@e43.eu on top of the patches here we should be able to simplify encoding and decoding to something like: Thanks! Christian --- Christian Brauner (2): pidfs: rework inode number allocation pidfs: remove 32bit inode number handling fs/pidfs.c | 138 +++++++++++++++++++++++++++++++++++--------------- include/linux/pidfs.h | 2 + kernel/pid.c | 11 ++-- 3 files changed, 103 insertions(+), 48 deletions(-) --- base-commit: b86545e02e8c22fb89218f29d381fa8e8b91d815 change-id: 20241128-work-pidfs-2bd42c7ea772