From: Thomas Gleixner <tglx@linutronix.de>
To: Neeraj Upadhyay <neeraju@codeaurora.org>
Cc: josh@joshtriplett.org, peterz@infradead.org, mingo@kernel.org,
jiangshanlai@gmail.com, dzickus@redhat.com,
brendan.jackman@arm.com, malat@debian.org,
linux-kernel@vger.kernel.org, sramana@codeaurora.org,
linux-arm-msm@vger.kernel.org
Subject: Re: [PATCH] cpu/hotplug: Fix rollback during error-out in takedown_cpu()
Date: Wed, 5 Sep 2018 14:23:46 +0200 (CEST) [thread overview]
Message-ID: <alpine.DEB.2.21.1809051419580.1416@nanos.tec.linutronix.de> (raw)
In-Reply-To: <alpine.DEB.2.21.1809051244500.1416@nanos.tec.linutronix.de>
On Wed, 5 Sep 2018, Thomas Gleixner wrote:
> On Tue, 4 Sep 2018, Neeraj Upadhyay wrote:
> > ret = cpuhp_down_callbacks(cpu, st, target);
> > if (ret && st->state > CPUHP_TEARDOWN_CPU && st->state < prev_state) {
> > - cpuhp_reset_state(st, prev_state);
> > + /*
> > + * As st->last is not set, cpuhp_reset_state() increments
> > + * st->state, which results in CPUHP_AP_SMPBOOT_THREADS being
> > + * skipped during rollback. So, don't use it here.
> > + */
> > + st->rollback = true;
> > + st->target = prev_state;
> > + st->bringup = !st->bringup;
>
> No, this is just papering over the actual problem.
>
> The state inconsistency happens in take_cpu_down() when it returns with a
> failure from __cpu_disable() because that returns with state = TEARDOWN_CPU
> and st->state is then incremented in undo_cpu_down().
>
> That's the real issue and we need to analyze the whole cpu_down rollback
> logic first.
And looking closer this is a general issue. Just that the TEARDOWN state
makes it simple to observe. It's universaly broken, when the first teardown
callback fails because, st->state is only decremented _AFTER_ the callback
returns success, but undo_cpu_down() increments unconditionally.
Patch below.
Thanks,
tglx
----
--- a/kernel/cpu.c
+++ b/kernel/cpu.c
@@ -916,7 +916,8 @@ static int cpuhp_down_callbacks(unsigned
ret = cpuhp_invoke_callback(cpu, st->state, false, NULL, NULL);
if (ret) {
st->target = prev_state;
- undo_cpu_down(cpu, st);
+ if (st->state < prev_state)
+ undo_cpu_down(cpu, st);
break;
}
}
@@ -969,7 +970,7 @@ static int __ref _cpu_down(unsigned int
* to do the further cleanups.
*/
ret = cpuhp_down_callbacks(cpu, st, target);
- if (ret && st->state > CPUHP_TEARDOWN_CPU && st->state < prev_state) {
+ if (ret && st->state == CPUHP_TEARDOWN_CPU && st->state < prev_state) {
cpuhp_reset_state(st, prev_state);
__cpuhp_kick_ap(st);
}
next prev parent reply other threads:[~2018-09-05 12:23 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-09-04 6:33 [PATCH] cpu/hotplug: Fix rollback during error-out in takedown_cpu() Neeraj Upadhyay
2018-09-04 12:12 ` Mukesh Ojha
2018-09-05 11:33 ` Thomas Gleixner
2018-09-05 12:09 ` Mukesh Ojha
2018-09-05 13:14 ` Thomas Gleixner
2018-09-05 12:23 ` Thomas Gleixner [this message]
2018-09-05 12:51 ` Neeraj Upadhyay
2018-09-05 13:17 ` Thomas Gleixner
2018-09-06 2:56 ` Neeraj Upadhyay
2018-09-06 8:18 ` Thomas Gleixner
2018-09-06 9:00 ` Neeraj Upadhyay
2018-09-05 14:53 ` Sudeep Holla
2018-09-05 15:19 ` Geert Uytterhoeven
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=alpine.DEB.2.21.1809051419580.1416@nanos.tec.linutronix.de \
--to=tglx@linutronix.de \
--cc=brendan.jackman@arm.com \
--cc=dzickus@redhat.com \
--cc=jiangshanlai@gmail.com \
--cc=josh@joshtriplett.org \
--cc=linux-arm-msm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=malat@debian.org \
--cc=mingo@kernel.org \
--cc=neeraju@codeaurora.org \
--cc=peterz@infradead.org \
--cc=sramana@codeaurora.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).