21:35Lyude: oh hey - looks like I wasn't wrong about type C hotplugging after all
21:36imirkin: that it's a pain?
21:36Lyude: currently reviewing some patches for out of band HPD events from type C ports, apparently the HPD pins are no longer guaranteed to be on the GPU
21:36Lyude: imirkin: no that there were extra typec events we weren't handling :P
21:36imirkin: that's just an implementation detail of said pain...
21:36Lyude: i suppose
21:37karolherbst: Lyude: wait what?
21:38Lyude: karolherbst: yeah - I'm reviewing a bunch of patches from hans for supporting this universally across drivers
21:38karolherbst: uhhh.. annoying
21:38Lyude: i mean we're about to have it supported soon :), I'm just glad for that
21:51Lyude: looks like we'll need to hook up a callback or two for nouveau but that's it
21:52Lyude: or, oh. no, it needs more then that :\
21:52Lyude: at least we know these are a thing that exist now though
21:53Lyude: this will likely need some kind of info from nvidia though, skeggsb_ if you want to come up with something for that person from lenovo to look into ^ :)
22:48karolherbst: so I'd really like to get https://gitlab.freedesktop.org/mesa/drm/-/merge_requests/188 in, should I just merge or does anybody wants to review it before I'll just do that?
22:49karolherbst: I wouldn't mind that much if that wouldn't add a new API call
22:50imirkin: karolherbst: i don't think this matters, but
22:50imirkin: nouveau_pushbuf_new starts out with something similar
22:50imirkin: where it also sets req.nr_push = 0;
22:51imirkin: this isn't strictly required, since the thing should get zero-initialized
22:51imirkin: but i'd encourage you to do the same thing there too
22:51imirkin: and add a similar comment that this is a no-op submission
22:51karolherbst: mhh yeah, makes sense
23:12imirkin: karolherbst: also i dunno about sticking "bool" into public API's
23:13karolherbst: yeah.... not sure either, but since C99 those are actually well defined.. soo
23:13imirkin: not all is C99 in the world
23:13imirkin: that's my point
23:13karolherbst: not even the nouveau ddx?
23:14imirkin: i assume it is
23:14imirkin: but drm_nouveau is meant to be more general
23:15karolherbst: amdgpu seems to use bool
23:15karolherbst: inside amdgpu/amdgpu.h
23:15imirkin: fine i guess
23:16karolherbst: I mean.. if something still defaults to c89 they kind of deserve the pain or so