hexdump01 has joined #linux-cros-arm
hexdump0815 has quit [Ping timeout: 480 seconds]
mps has joined #linux-cros-arm
<
mps>
jenneron[m]: I'm here
<
mps>
I would like if I can find just kernel in pmos repo and install it to test, not full pmos installation
<
mps>
lets see will it boot
<
jenneron[m]>
but I think it depends on package which is not in alpine
<
jenneron[m]>
(nevermind, it's probably a make dependency)
<
mps>
it boots and show display
<
mps>
display flicker with X
<
mps>
jenneron[m]: somewhat better performance with panforst than mainline kernel but still practically unusable
<
jenneron[m]>
weird
<
jenneron[m]>
it works fine on peach-pit
<
jenneron[m]>
mps: maybe try wayland
<
jenneron[m]>
and send me xorg logs and dmesg
<
mps>
jenneron[m]: nvm, it is good also without gpu
<
mps>
I mean usable
<
jenneron[m]>
well, you definitely can block probing of "panfrost" module in /etc/modprobe.d and it should be fine
<
mps>
I did that for long
<
jenneron[m]>
but it's interesting that it doesn't work for you, it works on peach-pit fine, so i can't reproduce it
<
jenneron[m]>
i would like to see logs
<
jenneron[m]>
and it is also interesting whether it is different in wayland or not
<
mps>
I don't like wayland
<
jenneron[m]>
for xorg some tearing is a usual behavior, but if you say unusable it's something else
<
jenneron[m]>
you can try to revert these changes one by one
<
mps>
I already rebuilt xorg with TearFree git commit for aarch64
<
mps>
another note, panfrost is slow also on samsung gru-kevin chromebook but there is usable
<
jenneron[m]>
it is slow on xorg in general
<
jenneron[m]>
in xorg i get 300-400 fps in glmark2, in wayland i get 750
<
jenneron[m]>
similar situation on rk3288
<
jenneron[m]>
but i'm not sure whether it is xorg on weak gpus problem or just a panfrost problem
<
mps>
anyway I don't use these devices often so not important much for me
<
mps>
xorg works fine on other machines for me
<
mps>
other gpu drivers
<
jenneron[m]>
if you ever decide to debug it, try that 5.18.5 tag and try to revert those patches one by one
<
jenneron[m]>
i think i will send a config blocking panfrost probe for peach-pi into pmaport for now
<
mps>
thanks for advice, but I prefer to follow mainline if possible at all
<
jenneron[m]>
yeah, it would be nice to solve this properly
<
jenneron[m]>
unfortunately i don't have knowledge for this
<
mps>
hmm, maybe could try this patch on mainline
<
jenneron[m]>
<mps> "thanks for advice, but I..." <- an option would be to move pmOS kernel to alpine and name it as "linux-exynos5"
<
mps>
uhm, these issue is assigned to me long ago
<
jenneron[m]>
but it may be easy to solve, i just haven't tried
<
mps>
alpine policy is to use patches which fixes serious bugs or security
<
mps>
so we don't like to add patches especially to kernels
<
jenneron[m]>
the same about packaging forks?
<
jenneron[m]>
in case of kernel
<
jenneron[m]>
i see
<
mps>
we are trying to follow upstream as much as possible
<
jenneron[m]>
well, for now it is impossible to have fully working sound and GPU with upstream
<
jenneron[m]>
the question is whether ones need it
<
mps>
because this I build locally different kernels for my machines
<
jenneron[m]>
i see
<
mps>
sound works fine on peach-pi
<
alpernebbi>
do you remember what errno you get from cros-ec?
<
jenneron[m]>
only outputs i guess
<
jenneron[m]>
exynos5-mainline also makes working mics and jack detection
<
mps>
not sure, didn't used mic for long
<
jenneron[m]>
alpernebbi: sorry, i don't remember
<
alpernebbi>
ok, no worries