[11:03] <mchehab> mkrufky: yeah, it is just kAPI. It could break media-build, but, if it does, we fix it later [11:04] <mchehab> one of the reasons why we were using cflags were due to the old building system. With the new one, it is easy to fix things like that - and we do it already with include/media/i2c [11:04] <mchehab> and other similar stuff [11:05] <mchehab> backport patches that touch header files could need changes, but I don't think there are any touching at the dvb header files [11:07] <mchehab> sailus: with regards to that patch series that document mc-centric stuff, I opted to not resend it for now, as I've no idea what changes we agreed. I recall that you were supposed to send us something [11:08] <mchehab> I'd appreciate if you could do it early-2018, if possible, in time for merging it to 4.16 [11:35] <sailus> mchehab: Something? :-) [11:36] <sailus> I can revisit the patchset, but right now I don't remember what this might have exactly been. :-o [11:36] <mchehab> I don't remember the details, but I guess you would either be sending a proposal or writing a meeting notes [11:36] <mchehab> (or both?) [11:38] <mchehab> the main topic where focused on what name we should use for non-subdev, non-mc device nodes created by V4L2 core [11:42] <mchehab> e. g. a definition that includes /dev/{video,vbi,radio,swradio,v4l-touch}", but will exclude /dev/v4l-subdev [11:43] <mchehab> on a separate topic, I'm starting to review the cio2 patchset [11:43] <mchehab> patch 2 generates 4 new doc warnings: [11:43] <mchehab> /devel/v4l/patchwork/Documentation/output/videodev2.h.rst:6: WARNING: undefined label: v4l2-pix-fmt-ipu3-sbggr10 (if the link has no caption the label must precede a section header) [11:43] <mchehab> /devel/v4l/patchwork/Documentation/output/videodev2.h.rst:6: WARNING: undefined label: v4l2-pix-fmt-ipu3-sgbrg10 (if the link has no caption the label must precede a section header) [11:43] <mchehab> /devel/v4l/patchwork/Documentation/output/videodev2.h.rst:6: WARNING: undefined label: v4l2-pix-fmt-ipu3-sgrbg10 (if the link has no caption the label must precede a section header) [11:43] <mchehab> /devel/v4l/patchwork/Documentation/output/videodev2.h.rst:6: WARNING: undefined label: v4l2-pix-fmt-ipu3-srggb10 (if the link has no caption the label must precede a section header) [11:45] <mchehab> probably, videodev2.h.rst.exceptions need to be patched [11:49] <sailus> Right, now I vaguely remember this. We didn't discuss all that many topics in the meeting indeed. I'll send notes on that. Thanks for the reminder. [11:49] <sailus> Ouch, I'll check the documentation patches. [11:52] <mchehab> argh, a macro named arr_size() with a different meaning than ARR_SIZE() is risky... [11:52] <mchehab> someone might send a patch replacing it some day [11:52] <mchehab> also, such macros should be in upper case [11:53] <mchehab> this doesn't block merge, but it would be good if you could send a patch replacing it by a better name [11:53] <mchehab> (patch 3) [11:53] <mchehab> +#define arr_size(a) ((a)->end - (a)->begin + 1) [11:54] <mchehab> btw, that entire code is weird: cio2_fbpt_rearrange() [11:54] <mchehab> what are it rearranging? buffers? why? [11:57] <sailus> It's rearranging buffer pointers that are accessed by the hardware. [11:57] <sailus> The hardware starts from the beginning when it is woken up after sleep. [11:57] <sailus> But the V4L2 buffers don't work that way, the hardware must continue processing from the next buffer, not first. [11:58] <sailus> The documentation fix is trivial. [11:58] <mchehab> this happens just when woken up after sleep? [11:59] <sailus> When the hardware is restarted essentially. [11:59] <sailus> The streaming state is maintained over sleep. [11:59] <sailus> So in this case the array is rearranged. [11:59] <sailus> Would you be fine with pulling this, and adding the fixes on top, or change the existing patches? [12:00] <mchehab> if those are the only issues, you can send separate patches [12:00] <sailus> Ack. Let me know if there's something else. [12:01] <mchehab> yeah, that's it [12:01] <mchehab> I guess I was expecting more patches :-) [12:02] <sailus> :-) [12:02] <mchehab> patches applied [12:03] <mchehab> I'll merge right now the patches that fix those warnings: [12:03] <mchehab> ./include/media/v4l2-async.h:80: warning: Function parameter or member 'match.fwnode' not described in 'v4l2_async_subdev' [12:03] <mchehab> I missed some dmxdev warnings due to that! [12:04] <mchehab> as it is producing 20 warnings [12:04] <sailus> Obrigado! [12:04] <sailus> I'll send the two patches addressing the issues above. [12:04] <mchehab> OK [13:39] *** ChanServ sets mode: +v mchehab` [13:43] <mchehab> ok, we're now warning-free again when building the documentation :-) [13:43] <mchehab> and the DVB kAPI documentation is again in sync with the implementation [13:44] <mchehab> we still need to make the V4L2 API in sync, but that's a hard work :-) [13:44] <mchehab> s/API/kAPI/ [14:35] <sailus> That's very nice indeed. [14:36] <sailus> How about the meeting notes on Prague meeting btw.? [14:36] <sailus> I remember there was discussion and a bunch of comments but I haven't seen any notes publicly. [14:36] <sailus> It'd be good to have them out, even if they're not perfect. [14:42] <mchehab> heh... didn't have time to work on it yet :-( [14:43] <mchehab> need to consolidate all comments into a new version [14:43] <mchehab> and submit the draft to linux-media [14:43] <mchehab> for a final review [14:45] <sailus> That'd be great. [14:45] <sailus> It's easier to refer to meeting notes when they exist. ;-) [15:00] <mchehab> :-) [15:00] <mchehab> well, someone needs time to look at the replies :-) [15:01] <mchehab> I won't be sad if you volunteer yourself :-D