Impose a limit on the number of watches that a user can hold so that they
can't use this mechanism to fill up all the available memory.
This is done by putting a counter in user_struct that's incremented when a
watch is allocated and decreased when it is released. If the number
exceeds the RLIMIT_NOFILE limit, the watch is rejected with EAGAIN.
This can be tested by the following means:
(1) Create a watch queue and attach it to fd 5 in the program given - in
this case, bash:
keyctl watch_session /tmp/nlog /tmp/gclog 5 bash
(2) In the shell, set the maximum number of files to, say, 99:
ulimit -n 99
(3) Add 200 keyrings:
for ((i=0; i<200; i++)); do keyctl newring a$i @s || break; done
(4) Try to watch all of the keyrings:
for ((i=0; i<200; i++)); do echo $i; keyctl watch_add 5 %:a$i || break; done
This should fail when the number of watches belonging to the user hits
99.
(5) Remove all the keyrings and all of those watches should go away:
for ((i=0; i<200; i++)); do keyctl unlink %:a$i; done
(6) Kill off the watch queue by exiting the shell spawned by
watch_session.
Fixes: c73be61cede5 ("pipe: Add general notification queue support")
Reported-by: Linus Torvalds <[email protected]>
Signed-off-by: David Howells <[email protected]>
---
include/linux/sched/user.h | 3 +++
kernel/watch_queue.c | 8 ++++++++
2 files changed, 11 insertions(+)
diff --git a/include/linux/sched/user.h b/include/linux/sched/user.h
index 917d88edb7b9..a8ec3b6093fc 100644
--- a/include/linux/sched/user.h
+++ b/include/linux/sched/user.h
@@ -36,6 +36,9 @@ struct user_struct {
defined(CONFIG_NET) || defined(CONFIG_IO_URING)
atomic_long_t locked_vm;
#endif
+#ifdef CONFIG_WATCH_QUEUE
+ atomic_t nr_watches; /* The number of watches this user currently has */
+#endif
/* Miscellaneous per-user rate limit */
struct ratelimit_state ratelimit;
diff --git a/kernel/watch_queue.c b/kernel/watch_queue.c
index f74020f6bd9d..0ef8f65bd2d7 100644
--- a/kernel/watch_queue.c
+++ b/kernel/watch_queue.c
@@ -393,6 +393,7 @@ static void free_watch(struct rcu_head *rcu)
struct watch *watch = container_of(rcu, struct watch, rcu);
put_watch_queue(rcu_access_pointer(watch->queue));
+ atomic_dec(&watch->cred->user->nr_watches);
put_cred(watch->cred);
}
@@ -452,6 +453,13 @@ int add_watch_to_object(struct watch *watch, struct watch_list *wlist)
watch->cred = get_current_cred();
rcu_assign_pointer(watch->watch_list, wlist);
+ if (atomic_inc_return(&watch->cred->user->nr_watches) >
+ task_rlimit(current, RLIMIT_NOFILE)) {
+ atomic_dec(&watch->cred->user->nr_watches);
+ put_cred(watch->cred);
+ return -EAGAIN;
+ }
+
spin_lock_bh(&wqueue->lock);
kref_get(&wqueue->usage);
kref_get(&watch->usage);
On Fri, Jul 24, 2020 at 11:17:26PM +0100, David Howells wrote:
> Impose a limit on the number of watches that a user can hold so that they
> can't use this mechanism to fill up all the available memory.
>
> This is done by putting a counter in user_struct that's incremented when a
> watch is allocated and decreased when it is released. If the number
> exceeds the RLIMIT_NOFILE limit, the watch is rejected with EAGAIN.
>
> This can be tested by the following means:
>
> (1) Create a watch queue and attach it to fd 5 in the program given - in
> this case, bash:
>
> keyctl watch_session /tmp/nlog /tmp/gclog 5 bash
>
> (2) In the shell, set the maximum number of files to, say, 99:
>
> ulimit -n 99
>
> (3) Add 200 keyrings:
>
> for ((i=0; i<200; i++)); do keyctl newring a$i @s || break; done
>
> (4) Try to watch all of the keyrings:
>
> for ((i=0; i<200; i++)); do echo $i; keyctl watch_add 5 %:a$i || break; done
>
> This should fail when the number of watches belonging to the user hits
> 99.
>
> (5) Remove all the keyrings and all of those watches should go away:
>
> for ((i=0; i<200; i++)); do keyctl unlink %:a$i; done
>
> (6) Kill off the watch queue by exiting the shell spawned by
> watch_session.
>
> Fixes: c73be61cede5 ("pipe: Add general notification queue support")
> Reported-by: Linus Torvalds <[email protected]>
> Signed-off-by: David Howells <[email protected]>
> ---
>
> include/linux/sched/user.h | 3 +++
> kernel/watch_queue.c | 8 ++++++++
> 2 files changed, 11 insertions(+)
>
> diff --git a/include/linux/sched/user.h b/include/linux/sched/user.h
> index 917d88edb7b9..a8ec3b6093fc 100644
> --- a/include/linux/sched/user.h
> +++ b/include/linux/sched/user.h
> @@ -36,6 +36,9 @@ struct user_struct {
> defined(CONFIG_NET) || defined(CONFIG_IO_URING)
> atomic_long_t locked_vm;
> #endif
> +#ifdef CONFIG_WATCH_QUEUE
> + atomic_t nr_watches; /* The number of watches this user currently has */
> +#endif
>
> /* Miscellaneous per-user rate limit */
> struct ratelimit_state ratelimit;
> diff --git a/kernel/watch_queue.c b/kernel/watch_queue.c
> index f74020f6bd9d..0ef8f65bd2d7 100644
> --- a/kernel/watch_queue.c
> +++ b/kernel/watch_queue.c
> @@ -393,6 +393,7 @@ static void free_watch(struct rcu_head *rcu)
> struct watch *watch = container_of(rcu, struct watch, rcu);
>
> put_watch_queue(rcu_access_pointer(watch->queue));
> + atomic_dec(&watch->cred->user->nr_watches);
> put_cred(watch->cred);
> }
>
> @@ -452,6 +453,13 @@ int add_watch_to_object(struct watch *watch, struct watch_list *wlist)
> watch->cred = get_current_cred();
> rcu_assign_pointer(watch->watch_list, wlist);
>
> + if (atomic_inc_return(&watch->cred->user->nr_watches) >
> + task_rlimit(current, RLIMIT_NOFILE)) {
> + atomic_dec(&watch->cred->user->nr_watches);
> + put_cred(watch->cred);
> + return -EAGAIN;
> + }
> +
> spin_lock_bh(&wqueue->lock);
> kref_get(&wqueue->usage);
> kref_get(&watch->usage);
>
>
Reviewed-by: Jarkko Sakkinen <[email protected]>
David, BTW, would it be possible to push keyrings to lore.kernel.org?
I don't have an archive for keyrings, which means that I cannot push
this forward.
/Jarkko
Hi Linus,
Could you consider taking this patch as a bugfix since the problem exists
already in upstream code?
David
David Howells <[email protected]> wrote:
> Could you consider taking this patch as a bugfix since the problem exists
> already in upstream code?
Alternatively, I can include it in a set with the mount notifications.
David