All of lore.kernel.org
 help / color / mirror / Atom feed
From: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
To: Sakari Ailus <sakari.ailus@iki.fi>
Cc: linux-media@vger.kernel.org, Antoine Reversat <a.reversat@gmail.com>
Subject: Re: [PATCH] omap3isp: Use monotonic timestamps for statistics buffers
Date: Fri, 14 Sep 2012 23:49:27 +0200	[thread overview]
Message-ID: <1853693.vMPZbuiXtB@avalon> (raw)
In-Reply-To: <20120913210139.GL6834@valkosipuli.retiisi.org.uk>

Hi Sakari,

On Friday 14 September 2012 00:01:39 Sakari Ailus wrote:
> On Thu, Sep 13, 2012 at 09:53:23PM +0200, Laurent Pinchart wrote:
> > V4L2 buffers use the monotonic clock, while statistics buffers use wall
> > time. This makes it difficult to correlate video frames and statistics.
> > 
> > Switch statistics buffers to the monotonic clock to fix this.
> > 
> > Reported-by: Antoine Reversat <a.reversat@gmail.com>
> > Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
> > ---
> > 
> >  drivers/media/platform/omap3isp/ispstat.c |    6 +++++-
> >  1 files changed, 5 insertions(+), 1 deletions(-)
> > 
> > diff --git a/drivers/media/platform/omap3isp/ispstat.c
> > b/drivers/media/platform/omap3isp/ispstat.c index b8640be..52263cc 100644
> > --- a/drivers/media/platform/omap3isp/ispstat.c
> > +++ b/drivers/media/platform/omap3isp/ispstat.c
> > @@ -253,10 +253,14 @@ isp_stat_buf_find_oldest_or_empty(struct ispstat
> > *stat)> 
> >  static int isp_stat_buf_queue(struct ispstat *stat)
> >  {
> > +	struct timespec ts;
> > +
> >  	if (!stat->active_buf)
> >  		return STAT_NO_BUF;
> > 
> > -	do_gettimeofday(&stat->active_buf->ts);
> > +	ktime_get_ts(&ts);
> > +	stat->active_buf->ts.tv_sec = ts.tv_sec;
> > +	stat->active_buf->ts.tv_usec = ts.tv_nsec / NSEC_PER_USEC;
> > 
> >  	stat->active_buf->buf_size = stat->buf_size;
> >  	if (isp_stat_buf_check_magic(stat, stat->active_buf)) {
> 
> I didn't think wall clock timestamps were used for statistics. This change
> is sure going to affect anyone using them --- which likely equates to no-one
> since I can hardly see use for wall clock timestams in such use.
> 
> How about using struct timespec instead?

Sounds good. I've tested it on ARM, x86-32 and x86-64, and both structures 
have the same size. I'll send a v2.

-- 
Regards,

Laurent Pinchart


      reply	other threads:[~2012-09-14 21:48 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-09-13 19:53 [PATCH] omap3isp: Use monotonic timestamps for statistics buffers Laurent Pinchart
2012-09-13 21:01 ` Sakari Ailus
2012-09-14 21:49   ` Laurent Pinchart [this message]

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=1853693.vMPZbuiXtB@avalon \
    --to=laurent.pinchart@ideasonboard.com \
    --cc=a.reversat@gmail.com \
    --cc=linux-media@vger.kernel.org \
    --cc=sakari.ailus@iki.fi \
    /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.