[05:22] <mani_s> kbingham, ping [07:38] *** tth__ has left [09:44] <svarbanov> mchehab: can you (or hverkuil ) delegate venus patches to me, so that I can change their state in patchwork? [10:50] <bbrezillon> Kwiboo: great! I can definitely wait a few days. [10:50] <bbrezillon> and no, no plans for HEVC support yet [10:50] <bbrezillon> the next thing on our list is VP9 [12:03] <ribalda> random question. Any ideas if there is any plan to move the AMD encoder drivers to v4l2 m2m? [12:05] <ribalda> I am thinking about the VCEs : https://github.com/obsproject/obs-amd-encoder/wiki/hardware-support [12:08] <hverkuil> ribalda: no plans AFAIK. [12:09] <ribalda> hverkuil: :( and is anyone of AMD interested in linux-media? I have never met anyone from AMD at media summits. But I have never been outside Europe [12:14] <hverkuil> Not a clue. [12:22] <ribalda> Converting a mesa vaapi driver into v4l2 m2m is something doable in a reasonable amount of time? [12:28] <hverkuil> Not a clue! :-) [12:28] <hverkuil> Sorry, I know absolutely nothing about this. [12:43] <OnVar> I did more work on the capture card drive and now I've got 0 failures and warnings. :) [12:45] <OnVar> I did a bit of work on updating the tw9910 driver, but hit a snag. I want to be able to tell the driver how I want the chip to be configured, but I imagine that would mean editing the platform data structure. [12:45] <OnVar> Wouldn't that potentially break some drivers if I add more fields to the structure? [12:47] <hverkuil> OnVar: should be fine: new fields would be 0 unless explicitly set by the bridge driver. So as long as '0' means 'keep the old behavior' it should have no effect on existing bridge drivers. [12:49] <OnVar> Alright, I wasn't sure about making the assumption that the structure would be zeroed out. Thanks. [13:21] <doublef> Hi all, I'm intending on proposing a couple of patches working around an issue with the GEO GC6500 camera, and possibly others. Patches at https://github.com/digitalloggers/geocam-bpp/tree/master/camera-bpp/5 should apply to the media master branch. It would be nice to have your preliminary feedback before posting them to linux-media (issue previously discussed at https://linuxtv.org/irc/irclogger_log/v4l?date=2019-08-22,Thu&sel=87#l83). [13:21] <doublef> pinchartl: your input especially welcome:) ^ [13:38] <ezequielg> doublef: hello! I think you are fine posting those patches on the mailing list. Discussion can happen directly there. [13:39] <ezequielg> Discussing things in the mailing list is actually encouraged, as it makes the work transparent to others. [13:40] <pinchartl> doublef: please CC me when posting the patches [13:43] <doublef> ezequielg: pinchartl: thanks, will prettify and post them. [13:49] <kbingham> kos_tom, I'm afraid we haven't looked at the DRM side for GMSL yet. There's always too much overlap between DRM/V4L2 on video :D [13:50] <kbingham> mani_s, Hola, how can I help [13:55] <kbingham> mani_s, Aha - I see you're asking over in #96boards about where the enable pin is on the max9286. Do you mean something like pin 47 (!PWDN)? (Page 19 in my datasheet) [14:22] <mani_s> kbingham, just playing w/ your gmsl patches and fixed one issue which stopped me from using max9286 on db410c :D [14:22] <mani_s> wondering if I can propose a patch for that or you can integrate it in your next version! [14:24] <kbingham> mani_s, Certainly! [14:25] <kbingham> I think you could either send your patch in reply to the latest max9286 series or just post it, clearly marked that it depends on the MAX9286 driver. [14:26] <mani_s> kbingham, thanks! will do. it is a one line fix btw! [14:27] <mani_s> kbingham, one more question: what kind of changes are required to convert a v4l2 platform driver to use vc? I'm currently looking at qcom camss [14:27] <kbingham> Excellent - should be easy then :) - Please cc neg : "Niklas Söderlund <niklas.soderlund@ragnatech.se>", as he has also worked on this driver recently. [14:40] <kbingham> mani_s, I think VC is becoming a big topic that's had a few rounds of development and discussions already. It may need some further dedicated development resourcing too. I believe jmondi is our current resident expert on VC and owns the latest published series. [14:42] <mani_s> kbingham, okay, then I'll wait for it to become stable then ;-) [14:44] <jmondi> kbingham: my understanding is that we're back at square zero on VC [14:44] <kbingham> jmondi, Excellent. Just like I said, you're the resident expert :D [14:44] <pinchartl> kbingham: yes, we've never reached an agreement on how to move forward :-( [19:39] <KitsuWhooa> Apologies for that random link, I did something silly [23:08] <ric96> Logitech Brio Led Controls not showing up: [23:08] <ric96> Using kernel 5.2.17 with fedora 30 [23:08] <ric96> The followinging modes for led contron do not show upled1_mode (menu) : min=0 max=3 default=0 value=3 [23:08] <ric96> led1_frequency (int) : min=0 max=255 step=1 default=0 value=0 [23:08] <ric96> any idea on this?