From: tip-bot for Kirill Tkhai <tipbot@zytor.com>
To: linux-tip-commits@vger.kernel.org
Cc: ktkhai@parallels.com, peterz@infradead.org,
linux-kernel@vger.kernel.org, tglx@linutronix.de, hpa@zytor.com,
oleg@redhat.com, mingo@kernel.org, torvalds@linux-foundation.org
Subject: [tip:sched/core] sched/numa: Fix unsafe get_task_struct() in task_numa_assign()
Date: Tue, 28 Oct 2014 04:02:22 -0700 [thread overview]
Message-ID: <tip-1effd9f19324efb05fccc7421530e11a52db0278@git.kernel.org> (raw)
In-Reply-To: <1413962231.19914.130.camel@tkhai>
Commit-ID: 1effd9f19324efb05fccc7421530e11a52db0278
Gitweb: http://git.kernel.org/tip/1effd9f19324efb05fccc7421530e11a52db0278
Author: Kirill Tkhai <ktkhai@parallels.com>
AuthorDate: Wed, 22 Oct 2014 11:17:11 +0400
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 28 Oct 2014 10:46:02 +0100
sched/numa: Fix unsafe get_task_struct() in task_numa_assign()
Unlocked access to dst_rq->curr in task_numa_compare() is racy.
If curr task is exiting this may be a reason of use-after-free:
task_numa_compare() do_exit()
... current->flags |= PF_EXITING;
... release_task()
... ~~delayed_put_task_struct()~~
... schedule()
rcu_read_lock() ...
cur = ACCESS_ONCE(dst_rq->curr) ...
... rq->curr = next;
... context_switch()
... finish_task_switch()
... put_task_struct()
... __put_task_struct()
... free_task_struct()
task_numa_assign() ...
get_task_struct() ...
As noted by Oleg:
<<The lockless get_task_struct(tsk) is only safe if tsk == current
and didn't pass exit_notify(), or if this tsk was found on a rcu
protected list (say, for_each_process() or find_task_by_vpid()).
IOW, it is only safe if release_task() was not called before we
take rcu_read_lock(), in this case we can rely on the fact that
delayed_put_pid() can not drop the (potentially) last reference
until rcu_read_unlock().
And as Kirill pointed out task_numa_compare()->task_numa_assign()
path does get_task_struct(dst_rq->curr) and this is not safe. The
task_struct itself can't go away, but rcu_read_lock() can't save
us from the final put_task_struct() in finish_task_switch(); this
reference goes away without rcu gp>>
The patch provides simple check of PF_EXITING flag. If it's not set,
this guarantees that call_rcu() of delayed_put_task_struct() callback
hasn't happened yet, so we can safely do get_task_struct() in
task_numa_assign().
Locked dst_rq->lock protects from concurrency with the last schedule().
Reusing or unmapping of cur's memory may happen without it.
Suggested-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Kirill Tkhai <ktkhai@parallels.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/1413962231.19914.130.camel@tkhai
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
kernel/sched/fair.c | 14 ++++++++++++--
1 file changed, 12 insertions(+), 2 deletions(-)
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 0b069bf..fbc0b82 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -1164,9 +1164,19 @@ static void task_numa_compare(struct task_numa_env *env,
long moveimp = imp;
rcu_read_lock();
- cur = ACCESS_ONCE(dst_rq->curr);
- if (cur->pid == 0) /* idle */
+
+ raw_spin_lock_irq(&dst_rq->lock);
+ cur = dst_rq->curr;
+ /*
+ * No need to move the exiting task, and this ensures that ->curr
+ * wasn't reaped and thus get_task_struct() in task_numa_assign()
+ * is safe under RCU read lock.
+ * Note that rcu_read_lock() itself can't protect from the final
+ * put_task_struct() after the last schedule().
+ */
+ if ((cur->flags & PF_EXITING) || is_idle_task(cur))
cur = NULL;
+ raw_spin_unlock_irq(&dst_rq->lock);
/*
* "imp" is the fault differential for the source task between the
next prev parent reply other threads:[~2014-10-28 11:03 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-10-22 7:17 [PATCH v4] sched/numa: fix unsafe get_task_struct() in task_numa_assign() Kirill Tkhai
2014-10-22 21:30 ` introduce task_rcu_dereference? Oleg Nesterov
2014-10-22 22:23 ` Oleg Nesterov
2014-10-23 18:15 ` Oleg Nesterov
2014-10-23 8:10 ` Kirill Tkhai
2014-10-23 18:18 ` Oleg Nesterov
2014-10-24 7:51 ` Kirill Tkhai
2014-10-27 19:53 ` [PATCH 0/3] introduce task_rcu_dereference() Oleg Nesterov
2014-10-27 19:54 ` [PATCH 1/3] probe_kernel_address() can use __probe_kernel_read() Oleg Nesterov
2014-10-27 19:54 ` [PATCH 2/3] introduce probe_slab_address() Oleg Nesterov
2014-10-27 19:21 ` Christoph Lameter
2014-10-28 5:44 ` Kirill Tkhai
2014-10-28 5:48 ` Kirill Tkhai
2014-10-28 15:01 ` Peter Zijlstra
2014-10-28 17:56 ` Kirill Tkhai
2014-10-28 18:00 ` Kirill Tkhai
2014-10-28 19:55 ` Oleg Nesterov
2014-10-28 20:12 ` Oleg Nesterov
2014-10-29 5:10 ` Kirill Tkhai
2014-10-27 19:54 ` [PATCH 3/3] introduce task_rcu_dereference() Oleg Nesterov
2014-10-28 6:22 ` Kirill Tkhai
2016-05-18 17:02 ` Peter Zijlstra
2016-05-18 18:23 ` Oleg Nesterov
2016-05-18 19:10 ` Peter Zijlstra
2016-05-18 19:57 ` Oleg Nesterov
2016-05-26 11:34 ` Peter Zijlstra
2016-06-03 10:49 ` [tip:sched/core] sched/fair: Use task_rcu_dereference() tip-bot for Oleg Nesterov
2016-06-03 10:48 ` [tip:sched/core] sched/api: Introduce task_rcu_dereference() and try_get_task_struct() tip-bot for Oleg Nesterov
2014-10-28 11:02 ` tip-bot for Kirill Tkhai [this message]
2014-11-08 3:48 ` [PATCH v4] sched/numa: fix unsafe get_task_struct() in task_numa_assign() Sasha Levin
2014-11-09 14:07 ` Kirill Tkhai
2014-11-10 10:03 ` Peter Zijlstra
2014-11-10 15:48 ` Sasha Levin
2014-11-10 16:01 ` Peter Zijlstra
2014-11-16 9:50 ` [tip:sched/urgent] sched/numa: Avoid selecting oneself as swap target tip-bot for Peter Zijlstra
2014-11-10 16:03 ` [PATCH v4] sched/numa: fix unsafe get_task_struct() in task_numa_assign() Peter Zijlstra
2014-11-10 16:09 ` Sasha Levin
2014-11-10 16:16 ` Peter Zijlstra
2014-11-10 16:10 ` Kirill Tkhai
2014-11-10 16:36 ` Kirill Tkhai
2014-11-10 16:44 ` Sasha Levin
2014-11-10 20:01 ` Kirill Tkhai
2014-11-12 9:49 ` Kirill Tkhai
2014-11-15 2:38 ` Sasha Levin
2014-11-18 17:30 ` Sasha Levin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=tip-1effd9f19324efb05fccc7421530e11a52db0278@git.kernel.org \
--to=tipbot@zytor.com \
--cc=hpa@zytor.com \
--cc=ktkhai@parallels.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-tip-commits@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=oleg@redhat.com \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).