From: Hans Verkuil <hverkuil@xs4all.nl>
To: Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
Philipp Zabel <p.zabel@pengutronix.de>
Cc: linux-media@vger.kernel.org,
Mauro Carvalho Chehab <m.chehab@samsung.com>,
Hans Verkuil <hans.verkuil@cisco.com>,
kernel@pengutronix.de,
Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Subject: Re: [RFC v2] [media] v4l2: add V4L2 pixel format array and helper functions
Date: Thu, 28 Aug 2014 18:40:53 +0200 [thread overview]
Message-ID: <53FF5B95.4030705@xs4all.nl> (raw)
In-Reply-To: <2088388.O2EqQOIWv7@avalon>
On 08/28/2014 06:25 PM, Laurent Pinchart wrote:
> Hi Philipp,
>
> On Thursday 28 August 2014 18:09:35 Philipp Zabel wrote:
>> Am Donnerstag, den 28.08.2014, 14:24 +0200 schrieb Laurent Pinchart:
>>>> A driver could then do the following:
>>>>
>>>> static struct v4l2_pixfmt_info driver_formats[] = {
>>>>
>>>> { .pixelformat = V4L2_PIX_FMT_YUYV },
>>>> { .pixelformat = V4L2_PIX_FMT_YUV420 },
>>>>
>>>> };
>>>>
>>>> int driver_probe(...)
>>>> {
>>>>
>>>> ...
>>>> v4l2_init_pixfmt_array(driver_formats,
>>>>
>>>> ARRAY_SIZE(driver_formats));
>>>>
>>>> ...
>>>>
>>>> }
>>>
>>> Good question. This option consumes more memory, and prevents the driver-
>>> specific format info arrays to be const, which bothers me a bit.
>>
>> Also, this wouldn't help drivers that don't want to take these
>> additional steps, which probably includes a lot of camera drivers with
>> only a few formats.
>>
>>> On the other hand it allows drivers to override some of the default
>>> values for odd cases.
>>
>> Hm, but those cases don't have to use the v4l2_pixfmt_info at all.
>>
>>> I won't nack this approach, but I'm wondering whether a better
>>> solution wouldn't be possible. Hans, Mauro, Guennadi, any opinion ?
>>
>> We could keep the global v4l2_pixfmt_info array sorted by fourcc value
>> and do a binary search (would have to be kept in mind when adding new
>> formats)
>
> I like that option, provided we can ensure that the array is sorted. This can
> get a bit tricky, and Hans might wear his "don't over-optimize" hat :-)
Well, for small sets of data (which this is) a binary search may well be
slower than a simple search. So yes, you should do some performance tests
before going with the more complex option.
By placing the commonly used pixel formats at the beginning of the list I
suspect a simple search is the fastest lookup method, and very easy to
implement as well.
Regards,
Hans
>
>> or build a hash table (more complicated code, consumes memory).
>
next prev parent reply other threads:[~2014-08-28 16:41 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-08-26 9:00 [RFC v2] [media] v4l2: add V4L2 pixel format array and helper functions Philipp Zabel
2014-08-26 10:01 ` Laurent Pinchart
2014-08-27 9:30 ` Philipp Zabel
2014-08-28 12:24 ` Laurent Pinchart
2014-08-28 16:09 ` Philipp Zabel
2014-08-28 16:25 ` Laurent Pinchart
2014-08-28 16:40 ` Hans Verkuil [this message]
2014-08-28 17:18 ` Mauro Carvalho Chehab
2014-08-28 17:32 ` Hans Verkuil
2014-08-28 17:41 ` Hans Verkuil
2014-09-02 10:00 ` Philipp Zabel
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=53FF5B95.4030705@xs4all.nl \
--to=hverkuil@xs4all.nl \
--cc=g.liakhovetski@gmx.de \
--cc=hans.verkuil@cisco.com \
--cc=kernel@pengutronix.de \
--cc=laurent.pinchart@ideasonboard.com \
--cc=linux-media@vger.kernel.org \
--cc=m.chehab@samsung.com \
--cc=p.zabel@pengutronix.de \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).