<!-- Some styling for better description lists --><style type='text/css'>dt { font-weight: bold;float: left;display:inline;margin-right: 1em} dd { display:block; margin-left: 2em}</style>

   ***: andrey_utkin has quit IRC (Ping timeout: 248 seconds)
   hverkuil: <u>tfiga</u>: ping
   grohne: do I understand correctly, that after setting some V4L2_CID_* and receiving a successful 0, one should check back the value one tried to set to see what actually got set and that value may differ from the requested value?
   tfiga: <u>hverkuil</u>: pong
   hverkuil: <u>tfiga</u>: were you planning to make an update to the codec API documentation soon? And if so, how soon? :-)
   tfiga: <u>hverkuil</u>: yes, this week, at least for decoder
   hverkuil: <u>tfiga</u>: ok, I'm about to post the first RFC of the vicodec driver. Obviously the idea is to make that a reference driver for stateful codecs (and later for stateless codecs as well)
   grohne: also there is V4L2_CID_EXPOSURE and V4L2_CID_EXPOSURE_ABSOLUTE. the latter is documented quite well, but everyone seems to use the former
   <br> can someone relate these?
   hverkuil: <u>tfiga</u>: I will probably wait with making more changes until you've posted a new version of the codec API docs.
   tfiga: <u>hverkuil</u>: okay, I'll keep it in mind when planning things
   <br> that's some useful work, thanks :)
   hverkuil: <u>grohne</u>: controls may be modified, yes. Typically this happens for integers so the value is mapped into the range.
   <br> Attempting to set an invalid menu control value will always return an error.
   grohne: <u>hverkuil</u>: thank you
   hverkuil: You usually are not able to map the HW exposure to some absolute unit due to lack of HW info, so that's the reason exposure_absolute is rarely used.
   ttomov: <u>hverkuil</u>: Hi. I'm running the v4l2-compliance on the new QComm CAMSS driver. I got a few failures and I'm starting to look at them.
   <br> <u>hverkuil</u>: I'm seeing one which I'm confused about. It is "fail: v4l2-test-subdevs.cpp(482): s_sel.which != which". I think that s_sel.which will always be 0 and if testSubDevSelection() is called with parameter which = 1, it will always fail.
   hverkuil: <u>ttomov</u>: ah, a v4l2-compliance bug. I only ever tested this with vimc and that driver doesn't support selection.
   ***: griffinp has quit IRC (Quit: ZNC - http://znc.in)
   ttomov: <u>hverkuil</u>: ok, thanks for info. I'll continue looking at the rest of the failures.
   hverkuil: It's missing s_sel.which = which; s_sel.pad = pad; lines right after s_sel = sel;
   ttomov: I see
   hverkuil: Try that and see if that fixes the test.
   ttomov: ok
   <br> <u>hverkuil</u>: I put sel.which = which; before s_sel = sel; and it fixes it
   hverkuil: <u>ttomov</u>: fix pushed.
   ttomov: <u>hverkuil</u>: fix received :) thanks
   grohne: what does one usually do when some V4L2_CID_ has only allows certain values not representable by the usual min/max/def/step configuration? is there a way to enumerate all working values from userspace?
   <br> why did you leave VIDIOC_TRY_EXT_CTRLS implementation-defined? "It is up to the driver whether wrong values are automatically adjusted to a valid value or if an error is returned."
   <br> all drivers but media/platform/vsp1/vsp1_hgt.c choose the "adjust" approach. only that one chooses the "error" approach
   ***: benjiG has left
   <br> paulk-leonov has quit IRC (Ping timeout: 245 seconds)