07:18 pq: emersion, if predicted vblanks are within 1 ms, the CRTC updates get bundled - something like that.
07:19 pq: emersion, and always when enabling/disabling outputs. Well, should. Happens at least on start-up and vt-switch in, but not sure about enabling more outputs on the fly.
07:28 pq: pcercuei, I've never even attempted it myself, but I hear you'd set up the palette in "the LUT". Which LUT in modern days, I don't know. I suppose the one that legacy KMS API calls "the gamma LUT".
07:29 emersion: cool
07:53 sravn: airlied: drm_panel_{attach,detach} patch caused a conflict. I fixed it up and pushed a resolution. There was a lot of other files involved (ttm stuff etc), but things did build. hope it is OK
07:54 sravn: Well, sparc64 did not build, but thats un-related. Some NR_CPUS thingy
08:04 sravn: drm-tip is broken. My alpha build fails with a recursive dependency.
08:04 sravn: git bisect points to 4057263e3ce2 ("Revert "drm/i915: Don't select BROKEN"")
08:04 sravn: This is an old commit that is not in upstream
08:05 sravn: But only present in drm-intel/topic/core-for-CI
12:42 hanetzer: agd5f_: yeah, found it grepping logs. Any particular data I should give other than the usual?
17:07 pcercuei: it's pretty calm here
17:08 pcercuei: Quick question, how can I support paletted 8bpp dumb buffers in my driver (ingenic-drm)?
17:08 pcercuei: just have the palette as a plane property?
20:50 Vanfanel: emersion: I am trying to set a custom cursor via the atomic interface, and you gave me a clue: that I neet to set these properties of the cursor plane: CRTC_ID, CRTC_X/Y/W/H, SRC_X/Y/W/H.
20:51 Vanfanel: My question is: is it normal that CRTC_ID for the cursor plane is set to 0 by default?
20:52 sravn: pcercuei: Maybe https://lists.freedesktop.org/archives/dri-devel/2018-August/186693.html will answer your question
20:52 Vanfanel: I mean... the primary plane I use for display comes with a configured CRTC_ID; Is it expected that the cursor plane does not?
20:54 sravn: pcercuei: The linked post is from 2018, so a bit outdated. But I think this was where it started. atmel_hlcdc have support for gamma - maybe worth a look
21:09 emersion: Vanfanel: the initial configuration you get is the one that was active before your compositor tool over
21:10 emersion: if you're running your compositor from a tty, it'll be the configuration used by kmscon: 1 primary plane, no cursor plane
21:10 emersion: so yeah, that's expected
21:10 emersion: possible_crtcs can tell you which CRTCs a plane is compatible with
21:14 Vanfanel: emersion: yes, I have selected a cursor plane that's compatible with the CRTC in use. So.. I must manually connect the cursor plane to the CRTC...ok
21:14 emersion: yes
21:15 emersion: you should do the same for the primary plane too, just in case
22:17 linusw: Would someone be so kind to merge v5.8 final into drm-misc-next? I have a fixe there that my new patches need as baseline.
23:01 pcercuei: sravn: I'll take a look, thanks