[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20201118233841.GS1437@paulmck-ThinkPad-P72>
Date: Wed, 18 Nov 2020 15:38:41 -0800
From: "Paul E. McKenney" <paulmck@...nel.org>
To: Marco Elver <elver@...gle.com>
Cc: Steven Rostedt <rostedt@...dmis.org>,
Anders Roxell <anders.roxell@...aro.org>,
Andrew Morton <akpm@...ux-foundation.org>,
Alexander Potapenko <glider@...gle.com>,
Dmitry Vyukov <dvyukov@...gle.com>,
Jann Horn <jannh@...gle.com>,
Mark Rutland <mark.rutland@....com>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
Linux-MM <linux-mm@...ck.org>,
kasan-dev <kasan-dev@...glegroups.com>, rcu@...r.kernel.org,
Peter Zijlstra <peterz@...radead.org>,
Tejun Heo <tj@...nel.org>,
Lai Jiangshan <jiangshanlai@...il.com>
Subject: Re: [PATCH] kfence: Avoid stalling work queue task without
allocations
On Wed, Nov 18, 2020 at 11:56:21PM +0100, Marco Elver wrote:
> On Tue, Nov 17, 2020 at 10:29AM -0800, Paul E. McKenney wrote:
> [...]
> > But it would be good to get the kcompactd() people to look at this (not
> > immediately seeing who they are in MAINTAINERS). Perhaps preemption is
> > disabled somehow and I am failing to see it.
> >
> > Failing that, maybe someone knows of a way to check for overly long
> > timeout handlers.
>
> I think I figured out one piece of the puzzle. Bisection keeps pointing
> me at some -rcu merge commit, which kept throwing me off. Nor did it
> help that reproduction is a bit flaky. However, I think there are 2
> independent problems, but the manifestation of 1 problem triggers the
> 2nd problem:
>
> 1. problem: slowed forward progress (workqueue lockup / RCU stall reports)
>
> 2. problem: DEADLOCK which causes complete system lockup
>
> | ...
> | CPU0
> | ----
> | lock(rcu_node_0);
> | <Interrupt>
> | lock(rcu_node_0);
> |
> | *** DEADLOCK ***
> |
> | 1 lock held by event_benchmark/105:
> | #0: ffffbb6e0b804458 (rcu_node_0){?.-.}-{2:2}, at: print_other_cpu_stall kernel/rcu/tree_stall.h:493 [inline]
> | #0: ffffbb6e0b804458 (rcu_node_0){?.-.}-{2:2}, at: check_cpu_stall kernel/rcu/tree_stall.h:652 [inline]
> | #0: ffffbb6e0b804458 (rcu_node_0){?.-.}-{2:2}, at: rcu_pending kernel/rcu/tree.c:3752 [inline]
> | #0: ffffbb6e0b804458 (rcu_node_0){?.-.}-{2:2}, at: rcu_sched_clock_irq+0x428/0xd40 kernel/rcu/tree.c:2581
> | ...
>
> Problem 2 can with reasonable confidence (5 trials) be fixed by reverting:
>
> rcu: Don't invoke try_invoke_on_locked_down_task() with irqs disabled
>
> At which point the system always boots to user space -- albeit with a
> bunch of warnings still (attached). The supposed "good" version doesn't
> end up with all those warnings deterministically, so I couldn't say if
> the warnings are expected due to recent changes or not (Arm64 QEMU
> emulation, 1 CPU, and lots of debugging tools on).
>
> Does any of that make sense?
Marco, it makes all too much sense! :-/
Does the patch below help?
Thanx, Paul
------------------------------------------------------------------------
commit 444ef3bbd0f243b912fdfd51f326704f8ee872bf
Author: Peter Zijlstra <peterz@...radead.org>
Date: Sat Aug 29 10:22:24 2020 -0700
sched/core: Allow try_invoke_on_locked_down_task() with irqs disabled
The try_invoke_on_locked_down_task() function currently requires
that interrupts be enabled, but it is called with interrupts
disabled from rcu_print_task_stall(), resulting in an "IRQs not
enabled as expected" diagnostic. This commit therefore updates
try_invoke_on_locked_down_task() to use raw_spin_lock_irqsave() instead
of raw_spin_lock_irq(), thus allowing use from either context.
Link: https://lore.kernel.org/lkml/000000000000903d5805ab908fc4@google.com/
Link: https://lore.kernel.org/lkml/20200928075729.GC2611@hirez.programming.kicks-ass.net/
Reported-by: syzbot+cb3b69ae80afd6535b0e@...kaller.appspotmail.com
Signed-off-by: Peter Zijlstra <peterz@...radead.org>
Signed-off-by: Paul E. McKenney <paulmck@...nel.org>
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index e172f2d..09ef5cf 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -2984,7 +2984,7 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags)
/**
* try_invoke_on_locked_down_task - Invoke a function on task in fixed state
- * @p: Process for which the function is to be invoked.
+ * @p: Process for which the function is to be invoked, can be @current.
* @func: Function to invoke.
* @arg: Argument to function.
*
@@ -3002,12 +3002,11 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags)
*/
bool try_invoke_on_locked_down_task(struct task_struct *p, bool (*func)(struct task_struct *t, void *arg), void *arg)
{
- bool ret = false;
struct rq_flags rf;
+ bool ret = false;
struct rq *rq;
- lockdep_assert_irqs_enabled();
- raw_spin_lock_irq(&p->pi_lock);
+ raw_spin_lock_irqsave(&p->pi_lock, rf.flags);
if (p->on_rq) {
rq = __task_rq_lock(p, &rf);
if (task_rq(p) == rq)
@@ -3024,7 +3023,7 @@ bool try_invoke_on_locked_down_task(struct task_struct *p, bool (*func)(struct t
ret = func(p, arg);
}
}
- raw_spin_unlock_irq(&p->pi_lock);
+ raw_spin_unlock_irqrestore(&p->pi_lock, rf.flags);
return ret;
}
Powered by blists - more mailing lists