All of lore.kernel.org
 help / color / mirror / Atom feed
From: Sakari Ailus <sakari.ailus@iki.fi>
To: linux-media@vger.kernel.org
Cc: laurent.pinchart@ideasonboard.com, javier@osg.samsung.com,
	mchehab@osg.samsung.com, hverkuil@xs4all.nl
Subject: [PATCH 00/19] Unrestricted media entity ID range support
Date: Tue, 27 Oct 2015 01:01:31 +0200	[thread overview]
Message-ID: <1445900510-1398-1-git-send-email-sakari.ailus@iki.fi> (raw)

Hi,

This is an update to my previous RFC patchset here:

<URL:http://www.spinics.net/lists/linux-media/msg93481.html>

In addition to the problems the patchset previously resolved and features
it implemented:

- unrestricted media entity ID support and
- API to manage entity enumerations

this set now solves one additional problem:

- unrestricted number of media entities.

The set also paves way to dynamic media entity registration but does not
implement it yet. In a completely dynamic system there's a lot more
mandatory error handling and difficult corner cases that one has to deal
with. There are cases where entity enumeration or graph walk may not fail
such as stopping streaming or power count calculation in link
enabling/disabling. This can be resolved by allocating entity enumerations
at a safe time and keeping them around until when they're needed and
releasing when they're no longer needed.

Keeping around entity enumerations together with dynamic entity
enumeration allocation brings up another problem: how do you ensure an
entity enumeration is large enough for all the entities that can be
encountered?

These additional problems are left for the future. This set does not
intend to address them.

The set has been tested with the omap3isp driver while the exynos4-is,
xilinx, vsp1 and omap4iss driver have been compile tested only. I'd
appreciate if others who have access to the hardware tested them.

What's changed is that the media entity enumerations are now dynamically
allocated if the number of entities in an enumeration exceeds a certain
constant value (which is now 64). The implication is that now error
handling is mandatory in entity enum initialisation (and thus graph walk
as well) as memory allocation may fail. Enumerations are allocated and
re-used in a few occasions in file handle open/close and pipeline
start/stop in order to guarantee successful graph walk, for instance.

The set also contains a fix for the omap4iss power management code. I kept
it in the set since another omap4iss driver patch depends on it.

Reviews would be very welcome.

-- 
Kind regards,
Sakari

             reply	other threads:[~2015-10-26 23:03 UTC|newest]

Thread overview: 41+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-10-26 23:01 Sakari Ailus [this message]
2015-10-26 23:01 ` [PATCH 01/19] media: Enforce single entity->pipe in a pipeline Sakari Ailus
2015-10-28  0:18   ` Mauro Carvalho Chehab
2015-10-26 23:01 ` [PATCH 02/19] media: Introduce internal index for media entities Sakari Ailus
2015-10-28  0:20   ` Mauro Carvalho Chehab
2015-10-26 23:01 ` [PATCH 03/19] media: Add an API to manage entity enumerations Sakari Ailus
2015-10-28  2:09   ` Mauro Carvalho Chehab
2015-11-03 22:21     ` Sakari Ailus
2015-10-26 23:01 ` [PATCH 04/19] media: Move struct media_entity_graph definition up Sakari Ailus
2015-10-28  0:36   ` Mauro Carvalho Chehab
2015-11-03 22:22     ` Sakari Ailus
2015-11-29 13:07       ` Sakari Ailus
2015-10-26 23:01 ` [PATCH 05/19] media: Move media graph state for streamon/off to the pipeline Sakari Ailus
2015-10-28  0:38   ` Mauro Carvalho Chehab
2015-11-03 22:23     ` Sakari Ailus
2015-10-26 23:01 ` [PATCH 06/19] media: Amend media graph walk API by init and cleanup functions Sakari Ailus
2015-10-28  2:09   ` Mauro Carvalho Chehab
2015-10-26 23:01 ` [PATCH 07/19] media: Use the new media_entity_graph_walk_start() Sakari Ailus
2015-10-28  0:43   ` Mauro Carvalho Chehab
2015-11-03 22:42     ` Sakari Ailus
2015-10-26 23:01 ` [PATCH 08/19] v4l: omap3isp: Use the new media_entity_graph_walk_start() interface Sakari Ailus
2015-10-26 23:01 ` [PATCH 09/19] v4l: exynos4-is: " Sakari Ailus
2015-10-26 23:01 ` [PATCH 10/19] v4l: xilinx: " Sakari Ailus
2015-10-26 23:01 ` [PATCH 11/19] v4l: vsp1: " Sakari Ailus
2015-10-26 23:01 ` [PATCH 12/19] media: Use entity enums in graph walk Sakari Ailus
2015-10-28  0:48   ` Mauro Carvalho Chehab
2015-10-26 23:01 ` [PATCH 13/19] media: Keep using the same graph walk object for a given pipeline Sakari Ailus
2015-10-28  0:51   ` Mauro Carvalho Chehab
2015-10-26 23:01 ` [PATCH 14/19] v4l: omap3isp: Use media entity enumeration API Sakari Ailus
2015-10-28  1:30   ` Mauro Carvalho Chehab
2015-11-03 22:44     ` Sakari Ailus
2015-10-26 23:01 ` [PATCH 15/19] v4l: vsp1: " Sakari Ailus
2015-10-28  1:33   ` Mauro Carvalho Chehab
2015-10-26 23:01 ` [PATCH 16/19] staging: omap4iss: Fix sub-device power management code Sakari Ailus
2015-10-28  1:59   ` Mauro Carvalho Chehab
2015-10-26 23:01 ` [PATCH 17/19] staging: v4l: omap4iss: Use media entity enum API Sakari Ailus
2015-10-28  2:01   ` Mauro Carvalho Chehab
2015-10-26 23:01 ` [PATCH 18/19] staging: v4l: omap4iss: Use the new media_entity_graph_walk_start() interface Sakari Ailus
2015-10-28  2:08   ` Mauro Carvalho Chehab
2015-10-26 23:01 ` [PATCH 19/19] media: Rename MEDIA_ENTITY_ENUM_MAX_ID as MEDIA_ENTITY_ENUM_STACK_ALLOC Sakari Ailus
2015-10-28  2:10   ` Mauro Carvalho Chehab

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1445900510-1398-1-git-send-email-sakari.ailus@iki.fi \
    --to=sakari.ailus@iki.fi \
    --cc=hverkuil@xs4all.nl \
    --cc=javier@osg.samsung.com \
    --cc=laurent.pinchart@ideasonboard.com \
    --cc=linux-media@vger.kernel.org \
    --cc=mchehab@osg.samsung.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.