linux-kernel - Re: [PATCH] OMAP: DSS: DPI: disable vt-switch on suspend/resume.

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

Message-ID: <54ED9DF8.8030105@ti.com>
Date:	Wed, 25 Feb 2015 12:03:36 +0200
From:	Tomi Valkeinen <tomi.valkeinen@...com>
To:	NeilBrown <neilb@...e.de>
CC:	<linux-omap@...r.kernel.org>, <linux-fbdev@...r.kernel.org>,
	<linux-kernel@...r.kernel.org>,
	GTA04 owners <gta04-owner@...delico.com>
Subject: Re: [PATCH] OMAP: DSS: DPI: disable vt-switch on suspend/resume.

On 25/02/15 11:37, NeilBrown wrote:
> 
> These devices do not need to return to non-graphic console
> for suspend, so disable that option.
> This means there is less work to do in the suspend/resume cycle,
> making it smoother and cheaper.
> 
> 
> Signed-off-by: NeilBrown <neil@...wn.name>
> 
> --
> Hi Tomi,
>  I wonder if you would consider this patch too.  It works for me and
> removes pointless vt switching.  I assume no-one would need or want that
> switching on suspend/resume but I guess I cannot be certain.
> 
> Maybe a module-parameter could be used if there was any uncertainty.

I was just yesterday picking patches from TI's kernel on top of
mainline, and there's a similar one for omapfb and for omapdrm. Here's
for omapfb:

http://git.ti.com/android-sdk/kernel-video/commit/5915d8744c927307987b12b14bc6aa37b3bac05c?format=patch

The patch in TI's kernel claims it's needed to make resume work. You're
saying it just optimizes suspend/resume. I hadn't picked this up
earlier, because I didn't understand why pm_set_vt_switch() would be
needed to make resume work. And never had time to study it, so I still
don't know what it's about...

Looking at the drivers/tty/vt/vt_ioctl.c, it sounds to me that we should
always do pm_set_vt_switch(0), as omapdss restores everything just fine
on resume. Although it makes me wonder how it works if there are two
display controllers, one needing the switch and the other not...

Your patch does the call in a bit odd place, so I'll be queuing the
patches for omapfb and omapdrm. Or actually, maybe the call could be
done in one place in omapdss driver, as you do, but in omap_dsshw_probe().

 Tomi

Download attachment "signature.asc" of type "application/pgp-signature" (820 bytes)