Re: [PATCH v3 3/6] drm/modes: Allow to specify rotation and reflection on the commandline

From: "Noralf Trønnes" <noralf@tronnes.org>
To: Maxime Ripard <maxime.ripard@bootlin.com>
Cc: eben@raspberrypi.org, David Airlie <airlied@linux.ie>,
	Maarten Lankhorst <maarten.lankhorst@linux.intel.com>,
	dri-devel@lists.freedesktop.org,
	Paul Kocialkowski <paul.kocialkowski@bootlin.com>,
	Sean Paul <seanpaul@chromium.org>,
	Thomas Petazzoni <thomas.petazzoni@bootlin.com>,
	Daniel Vetter <daniel.vetter@intel.com>,
	linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH v3 3/6] drm/modes: Allow to specify rotation and reflection on the commandline
Date: Thu, 13 Jun 2019 16:12:05 +0200	[thread overview]
Message-ID: <4169c6d9-5a64-8d05-9496-263cd9eda485@tronnes.org> (raw)
In-Reply-To: <20190613125024.46yiiy6zrrqojajy@flea>

Den 13.06.2019 14.50, skrev Maxime Ripard:
> Hi,
> 
> On Wed, Jun 12, 2019 at 03:21:19PM +0200, Noralf Trønnes wrote:
>>>> The only way I see for that to happen, is to set
>>>> ->panel_orientation. And to repeat myself, imo that makes
>>>> 'orientation' a better name for this video= option.
>>>
>>> orientation and rotation are two different things to me. The
>>> orientation of a panel for example is absolute, while the rotation is
>>> a transformation. In this particular case, I think that both the
>>> orientation and the rotation should be taken into account, with the
>>> orientation being the default state, and the hardware / panel will
>>> tell us that, while the rotation would be a transformation from that
>>> default to whatever the user wants.
>>>
>>> More importantly, the orientation is a property of the hardware (ie,
>>> how the display has been assembled), while the rotation is a software
>>> construct.
>>>
>>> And if the property being used to expose that is the rotation, I guess
>>> it would make sense to just use the same name and remain consistent.
>>
>> Ok, I see. Based on this, I would assume that rotation would be relative
>> to the orientation, but I see that in this patch rotation doesn't take
>> orintation into account, it just overwrites it.
> 
> Yeah, that's a good point. I've updated the next version to add the
> rotation on the command line to the orientation.
> 
>> I don't how userspace deals with rotation on top of orientation.
> 
> The orientation is exposed through the property, and the result is
> available through the plane's rotation, so I guess that it's enough?
> 

I was just wondering if Xserver/wayland applied rotation on top of
orientation or not. Maybe it would make sense to treat internal drm
clients the same. I understand that the DRM rotation property doesn't
apply on top orientation, just wondering how mutter/wayland (whoever is
in charge) does this. Not important to me, so feel free to disregard,
I'm just wondering.

Noralf.

> Maxime
> 
> --
> Maxime Ripard, Bootlin
> Embedded Linux and Kernel engineering
> https://bootlin.com
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel