linux-kernel - Re: [PATCH v2 4/5] platform/x86: intel-vbtn: support KEY_ROTATE_LOCK

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 for Android: free password hash cracker in your pocket

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

Date:   Fri, 10 Nov 2017 03:20:19 +0100
From:   Stefan Brüns <stefan.bruens@...h-aachen.de>
To:     Darren Hart <dvhart@...radead.org>
CC:     Bastien Nocera <hadess@...ess.net>,
        <platform-driver-x86@...r.kernel.org>,
        <linux-input@...r.kernel.org>,
        AceLan Kao <acelan.kao@...onical.com>,
        "Andy Shevchenko" <andy@...radead.org>,
        <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2 4/5] platform/x86: intel-vbtn: support KEY_ROTATE_LOCK_TOGGLE

On Friday, November 10, 2017 2:54:22 AM CET Darren Hart wrote:
> On Fri, Nov 10, 2017 at 01:15:09AM +0100, Stefan Brüns wrote:
> > On Friday, November 10, 2017 12:30:46 AM CET Bastien Nocera wrote:
> > > On Thu, 2017-11-09 at 23:44 +0100, Stefan Brüns wrote:
> > > > The Rotate Lock button event is emitted on the XPS 12 (BIOS A8, but
> > > > not
> > > > on BIOS A2).
> > > > 
> > > > Signed-off-by: Stefan Brüns <stefan.bruens@...h-aachen.de>
> > > > ---
> > > > 
> > > > Changes in v2:
> > > > - Emit KEY_ROTATE_LOCK_TOGGLE instead of KEY_ROTATE_DISPLAY
> > > > - Use separate up/down events
> > > > 
> > > >  drivers/platform/x86/intel-vbtn.c | 2 ++
> > > >  1 file changed, 2 insertions(+)
> > > > 
> > > > diff --git a/drivers/platform/x86/intel-vbtn.c
> > > > b/drivers/platform/x86/intel-vbtn.c index e3f6375af85c..a484bcc6393b
> > > > 100644
> > > > --- a/drivers/platform/x86/intel-vbtn.c
> > > > +++ b/drivers/platform/x86/intel-vbtn.c
> > > > @@ -42,6 +42,8 @@ static const struct key_entry intel_vbtn_keymap[] =
> > > > {
> > > > 
> > > >  	{ KE_IGNORE, 0xC5, { KEY_VOLUMEUP } },		/* volume-up key release 
*/
> > > >  	{ KE_KEY, 0xC6, { KEY_VOLUMEDOWN } },		/* volume-down key press 
*/
> > > >  	{ KE_IGNORE, 0xC7, { KEY_VOLUMEDOWN } },	/* volume-down key 
release
> > 
> > */
> > 
> > > > +	{ KE_KEY,    0xC8, { KEY_ROTATE_LOCK_TOGGLE } },	/* rotate-lock key
> > > > press */ +	{ KE_KEY,    0xC9, { KEY_ROTATE_LOCK_TOGGLE } },	/*
> > > > rotate-lock key release */
> > > 
> > > How are those events sent? When pressing and releasing the key, do you
> > > receive 0xC8 followed by 0xC9? Or do you receive 0xC8 when pressing and
> > > releasing the first time, and 0xC9 when pressing and releasing a second
> > > time?
> > > 
> > > If the former, then it's not going to work. The release is supposed to
> > > be ignored, as you send the event with sparse_keymap_report_event().
> > > 
> > > If the latter, and there's an actual state, does it disable a device
> > > on-board, or activate an LED? If so, then it would need to be a switch,
> > > not a key.
> > 
> > Do you think I don't test the patches before sending? Let me tell you, it
> > *does* work.
> > 
> > You could also read the cover letter, where you find more details, putting
> > the patches in relation to each other.
> 
> A point of process. If there is context that is needed to explain the
> patch, it belongs in the patch, not just in the cover letter. The cover
> letter is effectively lost once the patches are merged.
> 
> > Just in case its not yet clear:
> > The codes are emitted when pressing a button. It is a button, not a
> > switch.
> > There is no state handled in hardware. On press (as noted by the code
> > comment), event code 0xc8 is emitted. On release, event code 0xc9 is
> > emitted.
> This sounds like the "former" scenario Bastien described, for which I
> understand the use case to be:
> 
> User presses and releases the rotate lock button to prevent the
> accelerometer for triggering screen rotation.
> 
> User presses and releases the rotate lock button to allow the accelerometer
> to trigger screen rotation.
> 
> Is that correct?
> 
> If so, why do we need to emit two KEY_ROTATE_LOCK_TOGGLE events each time
> instead of just the press event like the volume buttons?

Volume buttons *should* send separate press/release events, to allow software 
autorepeat.

For the rotate lock button, I see no reason *not* to report the actual state 
instead of doing an autorelease.

Kind regards,

Stefan

-- 
Stefan Brüns  /  Bergstraße 21  /  52062 Aachen
home: +49 241 53809034     mobile: +49 151 50412019
Download attachment "signature.asc" of type "application/pgp-signature" (196 bytes)