11:42 <bryanodonoghue> seeing a fair bit of ghosting on arch linux x1e with wayland on gnome and hyprland - looks like some kind of buffer switching problem per plane

11:43 <bryanodonoghue> OpenGL ES profile version string: OpenGL ES 3.2 Mesa 25.0.1-arch1.2

11:43 <bryanodonoghue> is there a missing mesa update ?

12:02 <SpieringsAE> I'm on the same mesa, not noticing anything weird on hyprland

12:03 <SpieringsAE> I do have that config tweak, don't know if it is still necessary

12:03 <SpieringsAE> render {

12:03 <SpieringsAE> explicit_sync = 0

12:04 <SpieringsAE> explicit_sync_kms = 0

12:04 <SpieringsAE> >

12:04 <SpieringsAE> }

12:04 <SpieringsAE> I believe that mr got accepted to fix it but don't know if it is already in the current hyprland version

12:06 <bryanodonoghue> hmm ok appreciated let me give that a go

12:09 <SpieringsAE> yeah this one I believe: https://github.com/hyprwm/Hyprland/pull/9398

12:09 <SpieringsAE> yeah thats 11 days after the last release so its not in yet

12:21 <SpieringsAE> bryanodonoghue: did it fix it for you on hyprland?

12:25 <anthony25> yep, I had glitches because hyprland on the last release didn't check if explicit sync was supported by the driver

12:25 <anthony25> and msm doesn't expose explicit sync

12:26 <bryanodonoghue> hmm it looks better now I think

12:26 <anthony25> so either you can disable explicit_sync in your hyprland config, or build Hyprland with this patch, or expose the explicit sync attribute in msm

12:27 <anthony25> (I wrote this patch for Hyprland)

12:28 <anthony25> msm supports explicit sync, but there was a bug in an older version of mesa that has been fixed since, so it has been disabled in the driver

12:32 <SpieringsAE> bryanodonoghue: it was night and day in some applications, halloy was barely usable without this fix, it flickered constantly

12:32 <SpieringsAE> for me at least

12:42 <anthony25> same, I use keepassxc in a scratchpad, and the pop-ups were flickering a lot

12:45 <anthony25> but this patch disable explicit sync if the driver doesn't support it, so "render.explicit_sync = 0" and this patch do the same thing

13:40 <robclark> another option would be to cherry-pick https://patchwork.freedesktop.org/series/145438/ on the kernel side, fwiw

13:40 <robclark> (that is in msm-next for v6.15)

13:46 <SpieringsAE> nice!

13:51 <bryanodonoghue> seems to do the same fix for me

13:51 <bryanodonoghue> I wonder if this will help with the ghosting i saw in gnome

14:05 <robclark> hmm, I hadn't seen any similar issues, or any issues, in gnome-shell wayland (f41).. but not quite sure what mean by ghosting?

14:09 <anthony25> robclark: ha you re-enabled explicit sync?

14:09 <anthony25> Nice!

14:10 <robclark> yeah, well the mesa fix was over a year ago.. so I decided enough time had elapsed.. if you are using an ancient mesa and bleeding edge kernel, that is a you problem ;-)

14:30 <anthony25> haha, I agree :p

14:30 <anthony25> and the distribution can always revert this if needed

14:32 <robclark> is this the ath12k issue that folks have been complaining about.. every now and then my laptop dies, and it seems to start with this:

14:32 <robclark> https://www.irccloud.com/pastebin/whI47a8q/

14:34 <JensGlathe[m]> one of them

14:34 <JensGlathe[m]> apparently its a bunch

14:36 <jhovold> robclark: looks like this one: https://bugzilla.kernel.org/show_bug.cgi?id=219796

14:37 <robclark> ahh, could be.. I tend to notice it when I show up in the office.. which unsurprisingly has a different wifi network from home ;-)

14:37 <jhovold> steev, JensGlathe[m]: how many msdu_done warnings do you see per day roughly?

14:37 <JensGlathe[m]> Between 0 and 20

14:37 <JensGlathe[m]> Usually maybe 5

14:40 <jhovold> perhaps you can help me test a fix over the weekend then, will post in a bit

15:15 <bryanodonoghue> robclark

15:15 <bryanodonoghue> https://youtube.com/watch?v=CPxt869L7WI&si=TOmy3soAoupJlkLB

15:15 <bryanodonoghue> only happened once

15:33 <robclark> huh, weird.. but does look like something is out of sync

15:52 <craftyguy> jhovold: no ath11k issues at all with your rc7 :)

15:54 <craftyguy> feel free to add my t-b

15:55 <jhovold> craftyguy: thanks again for testing, could you reply to the fix with your t-b?

16:03 <craftyguy> ohhh ya I see now that you cc'd me on it, sorry!

16:33 <jhovold> craftyguy: thanks!

17:06 <JensGlathe[m]> Re WCN7850 crashes, NULL pointer derefs, and the like

17:06 <JensGlathe[m]> I sort of accidentally pulled the whole branch for [this patch](https://lore.kernel.org/all/20250226113118.7626-1-quic_lingbok@quicinc.com/), which were about 100 ath related commits, apparently most of them not in ath-next. Built it, tested it here, published as 6.14-rc7-2, and it seems actually the most stable fix for the WCN7850 mess. Reports on the Ubuntu Concept X1E discourse seem to concur.

17:12 <anthony25> Does it fix the bluetooth disconnections?

17:13 <JensGlathe[m]> never had any... didn't pay attention to it

18:23 <steev> jhovold: enough that i was willing to simply comment it out

18:23 <steev> but it's far greater than 5

20:19 <JensGlathe[m]> Fans of the Thinkbook are trying to go supersonic