Re: [PATCH 2/2] apparmor: fix suspicious RCU usage warning inpolicy.c/policy.h

From: Paul E. McKenney
Date: Mon Sep 30 2013 - 23:53:54 EST

On Sun, Sep 29, 2013 at 08:39:22AM -0700, John Johansen wrote:
> The recent 3.12 pull request for apparmor was missing a couple rcu _protected
> access modifiers. Resulting in the follow suspicious RCU usage

Assuming the lock you called out is the right one (I have no idea!), this
looks good to me!

So why don't we need to worry that RCU read-side critical sections might
have modified the ->base.count field that aa_put_profile() references?
Because the RCU callback function is guaranteed to see the effect of any
RCU read-side critical sections that started before the corresponding
call_rcu() invocation. This of course assumes that you made the
structure inaccessible to readers before that same call_rcu() function.
(You did do this, didn't you? If not, you have very big problems over
and above the ->base.count field!)

Thanx, Paul

> [ 29.804534] [ INFO: suspicious RCU usage. ]
> [ 29.804539] 3.11.0+ #5 Not tainted
> [ 29.804541] -------------------------------
> [ 29.804545] security/apparmor/include/policy.h:363 suspicious rcu_dereference_check() usage!
> [ 29.804548]
> [ 29.804548] other info that might help us debug this:
> [ 29.804548]
> [ 29.804553]
> [ 29.804553] rcu_scheduler_active = 1, debug_locks = 1
> [ 29.804558] 2 locks held by apparmor_parser/1268:
> [ 29.804560] #0: (sb_writers#9){.+.+.+}, at: [<ffffffff81120a4c>] file_start_write+0x27/0x29
> [ 29.804576] #1: (&ns->lock){+.+.+.}, at: [<ffffffff811f5d88>] aa_replace_profiles+0x166/0x57c
> [ 29.804589]
> [ 29.804589] stack backtrace:
> [ 29.804595] CPU: 0 PID: 1268 Comm: apparmor_parser Not tainted 3.11.0+ #5
> [ 29.804599] Hardware name: ASUSTeK Computer Inc. UL50VT /UL50VT , BIOS 217 03/01/2010
> [ 29.804602] 0000000000000000 ffff8800b95a1d90 ffffffff8144eb9b ffff8800b94db540
> [ 29.804611] ffff8800b95a1dc0 ffffffff81087439 ffff880138cc3a18 ffff880138cc3a18
> [ 29.804619] ffff8800b9464a90 ffff880138cc3a38 ffff8800b95a1df0 ffffffff811f5084
> [ 29.804628] Call Trace:
> [ 29.804636] [<ffffffff8144eb9b>] dump_stack+0x4e/0x82
> [ 29.804642] [<ffffffff81087439>] lockdep_rcu_suspicious+0xfc/0x105
> [ 29.804649] [<ffffffff811f5084>] __aa_update_replacedby+0x53/0x7f
> [ 29.804655] [<ffffffff811f5408>] __replace_profile+0x11f/0x1ed
> [ 29.804661] [<ffffffff811f6032>] aa_replace_profiles+0x410/0x57c
> [ 29.804668] [<ffffffff811f16d4>] profile_replace+0x35/0x4c
> [ 29.804674] [<ffffffff81120fa3>] vfs_write+0xad/0x113
> [ 29.804680] [<ffffffff81121609>] SyS_write+0x44/0x7a
> [ 29.804687] [<ffffffff8145bfd2>] system_call_fastpath+0x16/0x1b
> [ 29.804691]
> [ 29.804694] ===============================
> [ 29.804697] [ INFO: suspicious RCU usage. ]
> [ 29.804700] 3.11.0+ #5 Not tainted
> [ 29.804703] -------------------------------
> [ 29.804706] security/apparmor/policy.c:566 suspicious rcu_dereference_check() usage!
> [ 29.804709]
> [ 29.804709] other info that might help us debug this:
> [ 29.804709]
> [ 29.804714]
> [ 29.804714] rcu_scheduler_active = 1, debug_locks = 1
> [ 29.804718] 2 locks held by apparmor_parser/1268:
> [ 29.804721] #0: (sb_writers#9){.+.+.+}, at: [<ffffffff81120a4c>] file_start_write+0x27/0x29
> [ 29.804733] #1: (&ns->lock){+.+.+.}, at: [<ffffffff811f5d88>] aa_replace_profiles+0x166/0x57c
> [ 29.804744]
> [ 29.804744] stack backtrace:
> [ 29.804750] CPU: 0 PID: 1268 Comm: apparmor_parser Not tainted 3.11.0+ #5
> [ 29.804753] Hardware name: ASUSTeK Computer Inc. UL50VT /UL50VT , BIOS 217 03/01/2010
> [ 29.804756] 0000000000000000 ffff8800b95a1d80 ffffffff8144eb9b ffff8800b94db540
> [ 29.804764] ffff8800b95a1db0 ffffffff81087439 ffff8800b95b02b0 0000000000000000
> [ 29.804772] ffff8800b9efba08 ffff880138cc3a38 ffff8800b95a1dd0 ffffffff811f4f94
> [ 29.804779] Call Trace:
> [ 29.804786] [<ffffffff8144eb9b>] dump_stack+0x4e/0x82
> [ 29.804791] [<ffffffff81087439>] lockdep_rcu_suspicious+0xfc/0x105
> [ 29.804798] [<ffffffff811f4f94>] aa_free_replacedby_kref+0x4d/0x62
> [ 29.804804] [<ffffffff811f4f47>] ? aa_put_namespace+0x17/0x17
> [ 29.804810] [<ffffffff811f4f0b>] kref_put+0x36/0x40
> [ 29.804816] [<ffffffff811f5423>] __replace_profile+0x13a/0x1ed
> [ 29.804822] [<ffffffff811f6032>] aa_replace_profiles+0x410/0x57c
> [ 29.804829] [<ffffffff811f16d4>] profile_replace+0x35/0x4c
> [ 29.804835] [<ffffffff81120fa3>] vfs_write+0xad/0x113
> [ 29.804840] [<ffffffff81121609>] SyS_write+0x44/0x7a
> [ 29.804847] [<ffffffff8145bfd2>] system_call_fastpath+0x16/0x1b
>
> Reported-by: miles.lane@xxxxxxxxx
> CC: paulmck@xxxxxxxxxxxxxxxxxx
> Signed-off-by: John Johansen <john.johansen@xxxxxxxxxxxxx>
> ---
> security/apparmor/include/policy.h | 4 +++-
> security/apparmor/policy.c | 3 ++-
> 2 files changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/security/apparmor/include/policy.h b/security/apparmor/include/policy.h
> index f2d4b63..c28b0f2 100644
> --- a/security/apparmor/include/policy.h
> +++ b/security/apparmor/include/policy.h
> @@ -360,7 +360,9 @@ static inline void aa_put_replacedby(struct aa_replacedby *p)
> static inline void __aa_update_replacedby(struct aa_profile *orig,
> struct aa_profile *new)
> {
> - struct aa_profile *tmp = rcu_dereference(orig->replacedby->profile);
> + struct aa_profile *tmp;
> + tmp = rcu_dereference_protected(orig->replacedby->profile,
> + mutex_is_locked(&orig->ns->lock));
> rcu_assign_pointer(orig->replacedby->profile, aa_get_profile(new));
> orig->flags |= PFLAG_INVALID;
> aa_put_profile(tmp);
> diff --git a/security/apparmor/policy.c b/security/apparmor/policy.c
> index 6172509..345bec0 100644
> --- a/security/apparmor/policy.c
> +++ b/security/apparmor/policy.c
> @@ -563,7 +563,8 @@ void __init aa_free_root_ns(void)
> static void free_replacedby(struct aa_replacedby *r)
> {
> if (r) {
> - aa_put_profile(rcu_dereference(r->profile));
> + /* r->profile will not be updated any more as r is dead */
> + aa_put_profile(rcu_dereference_protected(r->profile, true));
> kzfree(r);
> }
> }
> --
> 1.8.3.2
>

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/