linux-kernel - Re: [PATCH v2] ACPI / PM: Propagate KEY

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20190329123244.GA20104@amd>
Date:   Fri, 29 Mar 2019 13:32:44 +0100
From:   Pavel Machek <pavel@....cz>
To:     Hans de Goede <hdegoede@...hat.com>
Cc:     "Chen, Hu" <hu1.chen@...el.com>,
        "Rafael J. Wysocki" <rjw@...ysocki.net>,
        Len Brown <lenb@...nel.org>, linux-acpi@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] ACPI / PM: Propagate KEY_POWER to user space when
 resume

On Fri 2019-03-29 12:38:39, Hans de Goede wrote:
> Hi,
> 
> On 3/29/19 11:25 AM, Pavel Machek wrote:
> >>I run Android on x86 PC (it's a NUC). Everytime I press the power button
> >>to wake the system, it suspends right away. After some debug, I find
> >>that Android wants to see KEY_POWER at resume. Otherwise, its
> >>opportunistic suspend will kick in shortly.
> >>
> >>However, other OS such as Ubuntu doesn't like KEY_POWER at resume. So
> >>add a knob "/sys/module/button/parameters/key_power_at_resume" for users
> >>to select.
> >>
> >>Signed-off-by: Chen, Hu <hu1.chen@...el.com>
> >
> >NAK.
> >
> >Fix android, lets not break kernel.
> 
> It is not that simple, as I explained in my other reply to this
> patch, we alreayd have inconsistent behavior here inside the kernel.
> 
> When KEY_POWER is handled by the gpio-keys driver it does explicitly
> send a KET_POWER press event when the system is woken up through the
> power-button.

Which may be okay.

> Arguably that is more consistent, e.g. some systems can also be woken
> up through a home-button press and in that case we do want the KEY_HOMEPAGE
> to be propagated to userspace after the wakeup so that we not only wake
> but also switch to the homescreen (whatever that might be).

Which may also be okay.

> Note I'm not saying that I'm happy with any of this, but simply NACK-ing
> this patch is IMHO not the answer.

Well, to add a knob
"/sys/module/button/parameters/key_power_at_resume" is really not
acceptable. Android does not know that it needs to set it, so it will
not set it, and the problem remains. Plus, we get a great mess in
future.

But yes, we might want userland to know why the system woke up. And it
would be good if everything also worked with wake-on-lan. Hmm.

It is also possible that PC's ACPI power button should generate
something else than KEY_POWER. Power button on my USB keyboard
generates that, and that is really quite different button the one on
the box.

I'm not saying I know what the solution is. But we should have one
solution, not a knob to select between different solutions.

Best regards,
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Download attachment "signature.asc" of type "application/pgp-signature" (182 bytes)