All of lore.kernel.org
 help / color / mirror / Atom feed
From: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
To: Kieran Bingham <kieran.bingham@ideasonboard.com>
Cc: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>,
	linux-renesas-soc@vger.kernel.org
Subject: Re: [VSP-Tests PATCH 5/7] vsp-lib: Allow forcing pixel perfect comparisons
Date: Tue, 19 Feb 2019 18:35:37 +0200	[thread overview]
Message-ID: <20190219163537.GD3526@pendragon.ideasonboard.com> (raw)
In-Reply-To: <c69a16fc-98d5-8fe9-0c1d-7b06995b4ced@ideasonboard.com>

Hi Kieran,

On Sun, Feb 17, 2019 at 08:51:52PM +0000, Kieran Bingham wrote:
> On 17/02/2019 11:59, Laurent Pinchart wrote:
> > On Tue, Dec 04, 2018 at 03:51:44PM +0000, Kieran Bingham wrote:
> >> Provide a means for the tester to request pixel perfect matches on tests.
> >> This can be either through setting the environment variable VSP_PIXEL_PERFECT, or
> >> by passing either '-p' or '--pixel-perfect' on the test command line.
> > 
> > What's the use case for this ?
> 
> Some of the tests (i.e. scaling) provide 'fuzzy comparisons' to allow "a
> close match" to pass.
> 
> It can be desirable to 'turn that off' so that you can generate a set of
> frames (using the '-k' option) to diff the expected against the actual
> frames.

Ideally we should improve the reference frame generation tool to
generate pixel-perfect frames :-) In the meantime I agree this could be
useful, but I think we should then integrate this better by marking a
test run as reference, saving frames to a good location, and then
reusing them for the subsequent runs. It's a bit too manual in its
proposed form for an automated test suite :-)

> This helps while improving the system without making changes to the
> specific tests.
> 
> It could be called 'disable fuzzy matching' if you like?
> 
> >> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
> >> ---
> >>  scripts/vsp-lib.sh | 11 ++++++++++-
> >>  1 file changed, 10 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/scripts/vsp-lib.sh b/scripts/vsp-lib.sh
> >> index cf15a045ea8c..9140254c7459 100755
> >> --- a/scripts/vsp-lib.sh
> >> +++ b/scripts/vsp-lib.sh
> >> @@ -287,7 +287,11 @@ compare_frames() {
> >>  	params=${params//)/_}
> >>  	params=$pipe-$in_fmt-$out_fmt-$size$params
> >>  
> >> -	if [ x$__vsp_pixel_perfect != xtrue ] ; then
> >> +	# The system can hint when pixel-perfection is not supported,
> >> +	# however the user can override to force this requirement with
> >> +	# VSP_PIXEL_PERFECT=1 in the environment or by passing -p, or
> >> +	# --pixel-perfect on the commandline.
> >> +	if [ x$__vsp_pixel_perfect != xtrue -a x$VSP_PIXEL_PERFECT != x1 ] ; then
> >>  		method=fuzzy
> >>  	fi
> >>  
> >> @@ -1113,10 +1117,15 @@ case $1 in
> >>  		export VSP_KEEP_FRAMES=1
> >>  		shift
> >>  		;;
> >> +	-p|--pixel-perfect)
> >> +		export VSP_PIXEL_PERFECT=1
> >> +		shift
> >> +		;;
> >>  	-h|--help)
> >>  		echo "$(basename $0): VSP Test library"
> >>  		echo "  -x|--debug          enable shell debug"
> >>  		echo "  -k|--keep-frames    keep generated and captured frames"
> >> +		echo "  -p|--pixel-perfect  frames must match with pixel perfection"
> >>  		echo "  -h|--help           this help"
> >>  		exit
> >>  		shift

-- 
Regards,

Laurent Pinchart

  reply	other threads:[~2019-02-19 16:35 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-12-04 15:51 [VSP-Tests PATCH 0/7] Reset controls and unloved patches Kieran Bingham
2018-12-04 15:51 ` [VSP-Tests PATCH 1/7] tests: add pseudo platform test Kieran Bingham
2019-02-17  2:41   ` Laurent Pinchart
2019-02-17 20:41     ` Kieran Bingham
2018-12-04 15:51 ` [VSP-Tests PATCH 2/7] tests: Provide copy test to validate 1xN streams Kieran Bingham
2019-02-17  2:47   ` Laurent Pinchart
2019-02-17 20:43     ` Kieran Bingham
2018-12-04 15:51 ` [VSP-Tests PATCH 3/7] vsp-lib: Provide command line argument parsing Kieran Bingham
2019-02-17  2:48   ` Laurent Pinchart
2019-02-17 20:47     ` Kieran Bingham
2018-12-04 15:51 ` [VSP-Tests PATCH 4/7] vsp-lib: trivial: Fix spelling of Reference Kieran Bingham
2019-02-17  2:42   ` Laurent Pinchart
2019-02-17 20:48     ` Kieran Bingham
2018-12-04 15:51 ` [VSP-Tests PATCH 5/7] vsp-lib: Allow forcing pixel perfect comparisons Kieran Bingham
2019-02-17 11:59   ` Laurent Pinchart
2019-02-17 20:51     ` Kieran Bingham
2019-02-19 16:35       ` Laurent Pinchart [this message]
2018-12-04 15:51 ` [VSP-Tests PATCH 6/7] vsp-lib: Support late queuing of buffers in yavta Kieran Bingham
2019-02-17 12:00   ` Laurent Pinchart
2019-02-17 20:57     ` Kieran Bingham
2019-02-19 16:37       ` Laurent Pinchart
2018-12-04 15:51 ` [VSP-Tests PATCH 7/7] vsp-lib: Reset controls to defaults on each test run Kieran Bingham
2019-02-17 12:01   ` Laurent Pinchart
2019-02-17 21:00     ` Kieran Bingham

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=20190219163537.GD3526@pendragon.ideasonboard.com \
    --to=laurent.pinchart@ideasonboard.com \
    --cc=kieran.bingham+renesas@ideasonboard.com \
    --cc=kieran.bingham@ideasonboard.com \
    --cc=linux-renesas-soc@vger.kernel.org \
    /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.