linux-kernel - Re: [PATCH] media: i2c: improve suspend/resume switch performance for GT9769 VCM driver

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: <Zs7W9PDY2ny5JI8w@kekkonen.localdomain>
Date: Wed, 28 Aug 2024 07:51:16 +0000
From: Sakari Ailus <sakari.ailus@...ux.intel.com>
To: Zhi Mao <zhi.mao@...iatek.com>
Cc: Mauro Carvalho Chehab <mchehab@...nel.org>,
	Matthias Brugger <matthias.bgg@...il.com>,
	AngeloGioacchino Del Regno <angelogioacchino.delregno@...labora.com>,
	linux-media@...r.kernel.org, linux-kernel@...r.kernel.org,
	linux-arm-kernel@...ts.infradead.org,
	linux-mediatek@...ts.infradead.org, shengnan.wang@...iatek.com,
	yaya.chang@...iatek.com, teddy.chen@...iatek.com,
	yunkec@...omium.org, 10572168@...com
Subject: Re: [PATCH] media: i2c: improve suspend/resume switch performance
 for GT9769 VCM driver

Hi Zhi,

Thanks for the patch.

On Sat, Aug 17, 2024 at 03:34:02PM +0800, Zhi Mao wrote:
> GT9769 VCM power-on default setting is PD=0,
> so it is not necessary to set again in dw9768_init function,
> and it also has no requirement of setting PD=1
> before power-off in dw9768_release function.
> For GT9769 VCM, PD mode control will add extra time
> when switching between suspend and resume.
> e.g. chrome camera AP can switch between video and photo mode,
> the behavior corresponding to VCM is suspend and resume,
> it will cause camera preview is not smooth.

If this is the problem, wouldn't either of these two be a better option:

- keep the file handle open in the user space, to avoid powering off the
  VCM or

- add autosuspend support to the driver.

I also wouldn't differentiate driver behaviour between the chips. If the
hardware default really is different (is it, this is rare for
register-compatible parts), then the driver needs to reprogram it (at least
on the one with a different default).

-- 
Kind regards,

Sakari Ailus