From: Mike Galbraith <umgwanakikbuti@gmail.com>
To: Jan Kiszka <jan.kiszka@siemens.com>
Cc: Steven Rostedt <rostedt@goodmis.org>,
Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
RT <linux-rt-users@vger.kernel.org>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH RT 3.18] irq_work: Provide a soft-irq based queue
Date: Mon, 20 Apr 2015 10:03:56 +0200 [thread overview]
Message-ID: <1429517036.3226.9.camel@gmail.com> (raw)
In-Reply-To: <552FE320.6050601@siemens.com>
On Thu, 2015-04-16 at 18:28 +0200, Jan Kiszka wrote:
> Instead of turning all irq_work requests into lazy ones on -rt, just
> move their execution from hard into soft-irq context.
>
> This resolves deadlocks of ftrace which will queue work from
> arbitrary
> contexts, including those that have locks held that are needed for
> raising a soft-irq.
Yup, trace-cmd record -> dead-box fully repeatable, and now fixed.
> Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
> ---
>
> Second try, looks much better so far. And it also removes my concerns
> regarding other potential cases besides ftrace.
>
> kernel/irq_work.c | 26 ++++++++++----------------
> 1 file changed, 10 insertions(+), 16 deletions(-)
>
> diff --git a/kernel/irq_work.c b/kernel/irq_work.c
> index 9dda38a..3f6ffcd 100644
> --- a/kernel/irq_work.c
> +++ b/kernel/irq_work.c
> @@ -85,12 +85,9 @@ bool irq_work_queue_on(struct irq_work *work, int
> cpu)
> raise_irqwork = llist_add(&work->llnode,
> &per_cpu(hirq_work_list, cpu));
> else
> - raise_irqwork = llist_add(&work->llnode,
> - &per_cpu(lazy_list, cpu));
> -#else
> +#endif
> raise_irqwork = llist_add(&work->llnode,
> &per_cpu(raised_list, cpu));
> -#endif
>
> if (raise_irqwork)
> arch_send_call_function_single_ipi(cpu);
> @@ -114,21 +111,20 @@ bool irq_work_queue(struct irq_work *work)
> if (work->flags & IRQ_WORK_HARD_IRQ) {
> if (llist_add(&work->llnode, this_cpu_ptr(&hirq_work_list)))
> arch_irq_work_raise();
> - } else {
> + } else
> +#endif
> + if (work->flags & IRQ_WORK_LAZY) {
> if (llist_add(&work->llnode, this_cpu_ptr(&lazy_list)) &&
> tick_nohz_tick_stopped())
> +#ifdef CONFIG_PREEMPT_RT_FULL
> raise_softirq(TIMER_SOFTIRQ);
> - }
> #else
> - if (work->flags & IRQ_WORK_LAZY) {
> - if (llist_add(&work->llnode, this_cpu_ptr(&lazy_list)) &&
> - tick_nohz_tick_stopped())
> arch_irq_work_raise();
> +#endif
> } else {
> if (llist_add(&work->llnode, this_cpu_ptr(&raised_list)))
> arch_irq_work_raise();
> }
> -#endif
>
> preempt_enable();
>
> @@ -202,6 +198,8 @@ void irq_work_run(void)
> {
> #ifdef CONFIG_PREEMPT_RT_FULL
> irq_work_run_list(this_cpu_ptr(&hirq_work_list));
> + if (!llist_empty(this_cpu_ptr(&raised_list)))
> + raise_softirq(TIMER_SOFTIRQ);
> #else
> irq_work_run_list(this_cpu_ptr(&raised_list));
> irq_work_run_list(this_cpu_ptr(&lazy_list));
> @@ -211,15 +209,11 @@ EXPORT_SYMBOL_GPL(irq_work_run);
>
> void irq_work_tick(void)
> {
> -#ifdef CONFIG_PREEMPT_RT_FULL
> - irq_work_run_list(this_cpu_ptr(&lazy_list));
> -#else
> - struct llist_head *raised = &__get_cpu_var(raised_list);
> + struct llist_head *raised = this_cpu_ptr(&raised_list);
>
> if (!llist_empty(raised) && !arch_irq_work_has_interrupt())
> irq_work_run_list(raised);
> - irq_work_run_list(&__get_cpu_var(lazy_list));
> -#endif
> + irq_work_run_list(this_cpu_ptr(&lazy_list));
> }
>
> /*
next prev parent reply other threads:[~2015-04-20 8:04 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-04-16 14:06 [PATCH RT 3.18] ring-buffer: Mark irq_work as HARD_IRQ to prevent deadlocks Jan Kiszka
2015-04-16 14:12 ` Steven Rostedt
2015-04-16 14:26 ` Sebastian Andrzej Siewior
2015-04-16 14:28 ` Jan Kiszka
2015-04-16 14:57 ` Sebastian Andrzej Siewior
2015-04-16 15:31 ` Jan Kiszka
2015-04-16 15:10 ` Steven Rostedt
2015-04-16 15:29 ` Jan Kiszka
2015-04-16 15:33 ` Sebastian Andrzej Siewior
2015-04-16 16:28 ` [PATCH RT 3.18] irq_work: Provide a soft-irq based queue Jan Kiszka
2015-04-20 8:03 ` Mike Galbraith [this message]
2015-04-23 6:11 ` Mike Galbraith
2015-04-23 6:29 ` Jan Kiszka
2015-04-23 6:58 ` Mike Galbraith
2015-04-23 7:14 ` Jan Kiszka
2015-04-23 6:50 ` Jan Kiszka
2015-04-23 7:01 ` Mike Galbraith
2015-04-23 7:12 ` Jan Kiszka
2015-04-23 7:19 ` Mike Galbraith
2015-04-23 21:00 ` Steven Rostedt
2015-04-24 6:54 ` Mike Galbraith
2015-04-24 9:00 ` Jan Kiszka
2015-04-24 9:59 ` Mike Galbraith
2015-04-25 7:20 ` Mike Galbraith
2015-04-25 7:26 ` Jan Kiszka
2015-05-18 19:52 ` Sebastian Andrzej Siewior
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=1429517036.3226.9.camel@gmail.com \
--to=umgwanakikbuti@gmail.com \
--cc=bigeasy@linutronix.de \
--cc=jan.kiszka@siemens.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rt-users@vger.kernel.org \
--cc=rostedt@goodmis.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).