linux-kernel - Re: [PATCH v2 1/9] media: v4l2-subdev: Add a pad variant of .query_dv

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]

Date: Wed, 28 Feb 2024 16:34:49 +0100
From: Paweł Anikiel <panikiel@...gle.com>
To: Hans Verkuil <hverkuil-cisco@...all.nl>
Cc: airlied@...il.com, akpm@...ux-foundation.org, conor+dt@...nel.org, 
	daniel@...ll.ch, dinguyen@...nel.org, krzysztof.kozlowski+dt@...aro.org, 
	maarten.lankhorst@...ux.intel.com, mchehab@...nel.org, mripard@...nel.org, 
	robh+dt@...nel.org, tzimmermann@...e.de, devicetree@...r.kernel.org, 
	dri-devel@...ts.freedesktop.org, linux-kernel@...r.kernel.org, 
	linux-media@...r.kernel.org, chromeos-krk-upstreaming@...gle.com, 
	ribalda@...omium.org
Subject: Re: [PATCH v2 1/9] media: v4l2-subdev: Add a pad variant of .query_dv_timings()

On Wed, Feb 28, 2024 at 12:25 PM Hans Verkuil <hverkuil-cisco@...all.nl> wrote:
>
> Hi Paweł,
>
> On 21/02/2024 17:02, Paweł Anikiel wrote:
> > Currently, .query_dv_timings() is defined as a video callback without
> > a pad argument. This is a problem if the subdevice can have different
> > dv timings for each pad (e.g. a DisplayPort receiver with multiple
> > virtual channels).
> >
> > To solve this, add a pad variant of this callback which includes
> > the pad number as an argument.
>
> So now we have two query_dv_timings ops: one for video ops, and one
> for pad ops. That's not very maintainable. I would suggest switching
> all current users of the video op over to the pad op.

I agree it would be better if there was only one. However, I have some concerns:
1. Isn't there a problem with backwards compatibility? For example, an
out-of-tree driver is likely to use this callback, which would break.
I'm asking because I'm not familiar with how such API changes are
handled.
2. If I do switch all current users to the pad op, I can't test those
changes. Isn't that a problem?
3. Would I need to get ACK from all the driver maintainers?