uvos | so http://uvos.xyz/maserati/ is back now | 00:44 |
---|---|---|
uvos | recoverd from backups after the server suffered harddrive failure | 00:45 |
uvos | freemangordon: and i can confirm that mapphone-pannel is used by the vendor kernel | 00:45 |
Wizzup | freemangordon: ping | 12:50 |
Wizzup | freemangordon: hmm | 13:01 |
Wizzup | freemangordon: https://wizzup.org/dirlist/maemo-leste/lima/new-traces/lima-bug-report.trace | 13:02 |
Wizzup | this is fallback, rendered on lima, with egl swap buffer behaviour set | 13:02 |
Wizzup | freemangordon: something I noticed, if I replay this on my laptop it is fine/smooth | 13:02 |
Wizzup | freemangordon: but if I make the replay window fullscreened, it also has problems in apitrace replay | 13:02 |
Wizzup | freemangordon: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5800 | 13:21 |
rafael2k | I think I'll download mobian with the same kernel I'm using and copy over the audio setup | 13:33 |
rafael2k | it seems maemo audio samplerate is twice fast as it should in pulseaudio | 13:34 |
rafael2k | using alsa directly works perfect | 13:34 |
rafael2k | btw, battery in pp after full discchange still shows the message that calibration is needed... it does not bother me much, but it is there | 13:35 |
rafael2k | I'm using X with noAccell, and apart of the sound configuration I need to fix, 5.15.10 is rock-solid on the pp, no issues at all, modem works, wifi works, screen is fine, leds are fine and so on | 13:42 |
Wizzup | rafael2k: hm, you still need no accel? | 13:58 |
Wizzup | rafael2k: I am not sure how calibration works in pinephone, it might be different from n900 / droid4 | 14:01 |
buZz | oh i had that issue of 'double samplerate through pulse' aswell , but weirdly not from every application | 14:09 |
buZz | and wasnt on droid4 but some other debian derivate (iirc raspbian) | 14:10 |
buZz | s/droid4/leste/ | 14:10 |
freemangordon | Wizzup: great | 16:41 |
freemangordon | (bug report) | 16:42 |
Wizzup | freemangordon: any comments? | 17:38 |
freemangordon | no, not really, lets wait for lima guys to comment | 17:42 |
freemangordon | I assume they have PP, otherwise I am not sure what image you can provide | 17:42 |
Wizzup | freemangordon: they asked for pinephone image | 18:14 |
Wizzup | I will work on getting them one from -devel | 18:14 |
Wizzup | hoping we can do it in the CI | 18:14 |
uvos | Wizzup: you package arrived :) | 18:17 |
uvos | installing leste on d3 is quite painfull atm, since for it to work with current images i would have to backport the kernel changes to 5.10, i ended up ddin over the rootfs of my primary bionic | 18:17 |
uvos | also the backlight is really unhappy on my device | 18:17 |
uvos | i flickers like crazy on the mainline kernel | 18:17 |
uvos | will have to figure that one out | 18:17 |
uvos | d3 is really impressivly thin for a slider | 18:18 |
uvos | also there are some bugs in the solana clown boot branch, but ill straigten those out | 18:20 |
uvos | looks like downgrading bionic works btw, so i think we can have d3/bionic use the same root method etc | 18:21 |
Wizzup | uvos: yeah let's use the opportunity to do it right from the get-go | 18:52 |
Wizzup | uvos: right @ backlight | 18:52 |
Wizzup | ah check @ downgrading | 18:52 |
uvos | Wizzup: right | 19:24 |
freemangordon | Wizzup: sorry for the delay re non-CMA buffers, hopefully tomorrow will start work on it | 19:35 |
freemangordon | Wizzup: btw, do you know if I can 'subscribe' to the lime issue? | 19:37 |
uvos | aw crap | 19:42 |
uvos | they blew a fuse for android 4.1 | 19:42 |
uvos | (the first bionic i got that never had 4.1 installed but came with 4.0.4 i can downgrade the other one dosent acept the old cdt.bin) | 19:43 |
uvos | :( | 19:43 |
freemangordon | Wizzup: hmm, maybe we shall try PP with 20.x, I am starting to think there is some regression in mesa 21.x | 20:11 |
Wizzup | freemangordon: I think that's not very productive imho | 20:13 |
Wizzup | fwiw I am on mesa 21.2.5 | 20:13 |
Wizzup | with amdgpu/laptop | 20:14 |
freemangordon | what is not productive? | 20:14 |
freemangordon | switching versions back and forth? | 20:14 |
Wizzup | because I don't think the problem is there and had other problems like crashes and shader issues | 20:14 |
Wizzup | 21.2.5 is a big improvement over what we had before | 20:14 |
freemangordon | Wizzup: the point is that for me on intel mesa 20.x replay is fine, for guys on #lima on intel 21.x it is not | 20:15 |
freemangordon | but well, you lead :) | 20:17 |
freemangordon | Wizzup: one thing though - could you try to replay on PP befire upgrading mesa? | 20:18 |
freemangordon | *before | 20:18 |
freemangordon | if you dont; want to waste time, I can dd new image later on, just LMK | 20:19 |
Wizzup | freemangordon: ok but I am on 21.x and it looks ok | 20:27 |
Wizzup | freemangordon: you mean replay on old pp image? | 20:28 |
freemangordon | yes | 20:32 |
freemangordon | Wizzup: the point is - if it replays fine on intel 20.x, but not on intel 21.x, there is a slim chance it will replay fine on lima 20.x | 20:34 |
Wizzup | yeah, but I am more inclined to blame other things | 20:36 |
freemangordon | ok | 20:36 |
uvos | what are the mkfs.ext4 parameters to make the old kernels able to read the fs | 20:37 |
uvos | again? | 20:37 |
parazyd | mkfs -O ^64bit,^metadata_csum -t ext4 | 21:32 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!