[libcamera-devel] [PATCH v3 5/5] libcamera: delayed_controls: Fix off-by-one error in get()

[Kieran Bingham]

Hi Naush,

On 03/03/2021 11:24, Naushir Patuck wrote:
> Hi Kieran,
> 
> On Wed, 3 Mar 2021 at 11:00, Kieran Bingham
> <kieran.bingham at ideasonboard.com
> <mailto:kieran.bingham at ideasonboard.com>> wrote:
> 
>     Hi Naush,
> 
>     On 01/03/2021 13:31, Naushir Patuck wrote:
>     > There was an off-by-one error in DelayedControls::get() when picking
>     > controls from the queue to return back to the pipeline handler.
>     > This is only noticeable as small oscillations in brightness when
>     closely
>     > viewing frame while AGC is running. The old StaggeredCtrl did not show
>     > this error as the startup queuing mechanism has changed in
>     > DelayedControls.
>     >
>     > Fix this by indexing to the correct position in the queue.
> 
>     It seems this behaviour is currently expected by the test framework.
> 
>     Running the tests (whole suite with `ninja test`) identifies a
>     failure on:
> 
> 
>     > 44/61 libcamera / delayed_contols                                 
>           FAIL            0.12s   (exit status 255 or signal 127 SIGinvalid)
> 
> 
>     Running this test alone on each patch in the series:
> 
>     > git rebase -i libcamera.org/master <http://libcamera.org/master>
>     -x "ninja -C build && ./build/test/delayed_contols"
> 
>     stops at this patch.
> 
>     Could you also investigate the test, and update it accordingly in this
>     patch to make sure that it is correct and doesn't fail please?
> 
> 
> Sorry, my bad.  I had run the tests, and did not see a failure. 
> However, on closer
> inspection, the test was marked as SKIPPED, as I do not have the "vivid
> video"
> device driver available.
> 
> Let me see what's involved in getting this device available to use on our
> kernel, and I will update the tests to reflect the new state of the world.

Aha yes, we need vimc, vivid and vim2m to be able to run the tests
without hardware.

You can run that on the RPi, or on your host laptop with those modules
loaded.

Niklas' "lc-compliance" tool which I hope we'll see start to integrate
soon will do deeper per-pipeline tests on a running system.

For now the unit-tests are supposed to run 'without' hardware.

--
Kieran

> Regards,
> Naush
>  
> 
> 
>     I anticipate that this is just a case of needing to update the tests to
>     match, if we assume it is the test that is wrong in this instance, but
>     we should check that too, and make sure we're not breaking some other
>     assumptions elsewhere.
> 
>     --
>     Kieran
> 
> 
>     > Signed-off-by: Naushir Patuck <naush at raspberrypi.com
>     <mailto:naush at raspberrypi.com>>
>     > Reported-by: David Plowman <david.plowman at raspberrypi.com
>     <mailto:david.plowman at raspberrypi.com>>
>     > Fixes: 3d4b7b005911 ("libcamera: delayed_controls: Add helper for
>     controls that apply with a delay")
>     > Tested-by: David Plowman <david.plowman at raspberrypi.com
>     <mailto:david.plowman at raspberrypi.com>>
>     > Reviewed-by: Paul Elder <paul.elder at ideasonboard.com
>     <mailto:paul.elder at ideasonboard.com>>
>     > ---
>     >  src/libcamera/delayed_controls.cpp | 2 +-
>     >  1 file changed, 1 insertion(+), 1 deletion(-)
>     >
>     > diff --git a/src/libcamera/delayed_controls.cpp
>     b/src/libcamera/delayed_controls.cpp
>     > index 5a05741c0285..138761c9852e 100644
>     > --- a/src/libcamera/delayed_controls.cpp
>     > +++ b/src/libcamera/delayed_controls.cpp
>     > @@ -184,7 +184,7 @@ bool DelayedControls::push(const ControlList
>     &controls)
>     >   */
>     >  ControlList DelayedControls::get(uint32_t sequence)
>     >  {
>     > -     uint32_t adjustedSeq = sequence - firstSequence_ + 1;
>     > +     uint32_t adjustedSeq = sequence - firstSequence_;
>     >       unsigned int index = std::max<int>(0, adjustedSeq - maxDelay_);
>     > 
>     >       ControlList out(device_->controls());

-- 
Regards
--
Kieran