From: "Rafael J. Wysocki" <rjw@rjwysocki.net>
To: "Claude. Yen" <Claude.Yen@mediatek.com>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>,
Daniel Lezcano <daniel.lezcano@linaro.org>,
Frederic Weisbecker <fweisbec@gmail.com>,
Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@kernel.org>,
Matthias Brugger <matthias.bgg@gmail.com>,
Bjorn Andersson <bjorn.andersson@linaro.org>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
Stephen Boyd <swboyd@chromium.org>,
Douglas Anderson <dianders@chromium.org>,
Linux PM <linux-pm@vger.kernel.org>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Linux ARM <linux-arm-kernel@lists.infradead.org>,
"moderated list:ARM/Mediatek SoC..."
<linux-mediatek@lists.infradead.org>,
wsd_upstream <wsd_upstream@mediatek.com>
Subject: Re: [PATCH] PM: s2idle: Introduce syscore callbacks in s2idle flow
Date: Thu, 10 Sep 2020 14:55:06 +0200 [thread overview]
Message-ID: <1955367.r5QahNuf3v@kreacher> (raw)
In-Reply-To: <1599099247.4435.4.camel@mtksdccf07>
On Thursday, September 3, 2020 4:14:07 AM CEST Claude. Yen wrote:
> On Tue, 2020-09-01 at 13:57 +0200, Rafael J. Wysocki wrote:
> > On Tue, Sep 1, 2020 at 9:05 AM Claude Yen <claude.yen@mediatek.com> wrote:
> > >
> > > This series based on 5.9-rc1
> > > This patch makes s2idle call existing syscore callbacks. Currently,
> > > when s2idle is selected as system suspend method, callbacks hooked
> > > by register_syscore_ops() will not be triggered. This may induce
> > > unexpected results.
> >
> > They are not executed by design.
> >
> > > For example, sched_clock_suspend() was added to s2idle flow in
> > > commit 3f2552f7e9c5 ("timers/sched_clock: Prevent generic sched_clock
> > > wrap caused by tick_freeze()") to fix clock wrap problem. However,
> > > sched_clock_suspend() is originally registered in syscore callback.
> >
> > I'm not sure why this matters here.
>
> If functions in syscore callbacks are needed in s2idle, explicit
> migration is needed like commit 3f2552f7e9c5 ("timers/sched_clock:
> Prevent generic sched_clock wrap caused by tick_freeze()").
> Thus, I am wondering if such effort could be saved.
Yes, it could.
You can define platform ops for s2idle and invoke what's needed from there.
> > > With this patch, if another syscore callback is needed in s2idle,
> > > additional migration effort could be saved.
> >
> > s2idle cannot execute syscore callbacks, because it doesn' take
> > non-boot CPUs offline and it won't do that.
> >
> > Thanks!
>
> Yes, the current design of syscore callback needs non-boot CPUs offline.
> Considering the following case: in s2idle flow, there is a status that
> only one CPU is alive and other CPUs have enter deepest idle state.
> This situation is similar to getting non-boot CPUs offline, though all
> CPUs are online from kernel's perspective.
It is only similar AFAICS.
You don't migrate interrupts during s2idle, for example.
> Reply from Stephen mentioned that if an operation is needed in both
> S2R and s2idle, CPU_PM notifier can be utilized.
> In my opinion, CPU_PM notifier is particularly for CPU entering idle
> state. In contrast, syscore callback is for system going low power
> state. There exists semantic difference between these two callbacks.
Fair enough.
> Could the current design of syscore callback be re-designed as
> system-wide suspend callback?
No, it couldn't.
> Proposed suspend flow in this patch:
>
> Freeze tasks
> |
> V
> Device suspend callbacks
> |
> |-------------s2idle----------
> | |
> V |
> Disable nonboot CPUs Is this CPU last core to enter idle?
> | |
> V |-------------
> syscore callbacks | |
> | No Yes
> V | |
> platform suspend V V
> enter idle syscore callback
> |
> V
> enter idle
>
The primary problem with this is that on some architectures (x86 at least)
the syscore things cannot be run during the s2idle flow.
Also there is a way to invoke them through the platform ops as I said.
Thanks!
prev parent reply other threads:[~2020-09-10 21:43 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-09-01 7:04 [PATCH] PM: s2idle: Introduce syscore callbacks in s2idle flow Claude Yen
2020-09-01 11:57 ` Rafael J. Wysocki
[not found] ` <1599099247.4435.4.camel@mtksdccf07>
2020-09-10 12:55 ` Rafael J. Wysocki [this message]
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=1955367.r5QahNuf3v@kreacher \
--to=rjw@rjwysocki.net \
--cc=Claude.Yen@mediatek.com \
--cc=bjorn.andersson@linaro.org \
--cc=daniel.lezcano@linaro.org \
--cc=dianders@chromium.org \
--cc=fweisbec@gmail.com \
--cc=gregkh@linuxfoundation.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mediatek@lists.infradead.org \
--cc=linux-pm@vger.kernel.org \
--cc=matthias.bgg@gmail.com \
--cc=mingo@kernel.org \
--cc=rafael@kernel.org \
--cc=swboyd@chromium.org \
--cc=tglx@linutronix.de \
--cc=wsd_upstream@mediatek.com \
/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).