<lvrp16>
at first I thought it was the SD card but probing the lines shows that it is reset correctly. using u-boot psci suspend/poweroff/resume works fine, only after Linux has loaded and the system reboots does this loop happen
<f_>
Guess it's on cottonwood, when booting everything from SD?
<lvrp16>
yeah I have not tried eMMC yet, I should try it
<f_>
Can't reproduce on eMMC
<f_>
Oh well..
<f_>
Am booting from SPI..so I should probably try booting from eMMC.
<f_>
lvrp16: BTW I wasn't able to get X11 or Wayland to work on alta with Fedora
<lvrp16>
f_: you're right, it doesn't happen with eMMC. Only MicroSD
<lvrp16>
hmm it looks like SD is borked on reboot. more testing needed, thanks for the tip.
<f_>
But I'm starting to believe that it's more of a fedora issue than a board issue..
<lvrp16>
f_: what happens?
<f_>
Blank screen
<f_>
Seems like it "works", but I can't see anything.
<lvrp16>
maybe missing an X11 config?
<lvrp16>
wayland should work though
<f_>
(I can blindly try to open a terminal in weston and blindly type `reboot`)
<lvrp16>
x11, I've always had to add configuration to the /usr/share/X11 dir
<lvrp16>
wayland just works
<f_>
X11 yes
<f_>
I have that config.
<lvrp16>
is it a conflict between CVBS and HDMI?
<f_>
But it seems like people were able to reproduce this issue on Pi4 so..
<lvrp16>
outputting on CVBS instead of HDMI?
<f_>
lvrp16: Not sure, monitor still gets a signal.
<narmstrong>
lvrp16: it's probably the same issue we have on the Odroid N2 & C4
<lvrp16>
narmstrong: sorry, I'm not familiar, what issue?
<narmstrong>
f_: yes but who wants to go in court for that?
kilobyte_ch has joined #linux-amlogic
<f_>
chewitt: Yes but they were shared with the expectation that you don't reshare, right?
<chewitt>
correct
<f_>
so...ugh..won't do...
<f_>
May as well just use prebuilts on GT1..
<f_>
Or reverse changes made by beelink (easy for the most part)
<f_>
(acs..can just parse, BL21..can just look at UART debug logs...BL301 is the most complicated)
<f_>
With some luck Beelink's tree would probably be like WeTek's..
<f_>
(WeTek's tree only has very minor changes for WP2)
<f_>
WeTek has the same issue too.
<f_>
Technically, by just building and pushing to that fip repo, without corresponding source, you're violating the license.
psydroid has joined #linux-amlogic
<f_>
for WeTek it's mostly fine..you can just mod p200 acs to change rank mode and see it boot fine
<f_>
(so forcing wetek to publicly share becomes pointless)
<f_>
chewitt: See what's the issue?
<f_>
I also don't understand what's so secret in e.g. GT1 sources..
<chewitt>
the issue is .. they're a box vendor and Amlogic makes them sign NDAs
<chewitt>
and even though almost everything is GPL labelled .. that scares people into not sharing
<f_>
Doesn't sound very legal lol
Daanct12 has quit [Quit: WeeChat 4.1.2]
buzzmarshall has joined #linux-amlogic
psydroid2 has joined #linux-amlogic
f11f12 has joined #linux-amlogic
<f_>
TIL that pyamlboot can boot things from USB even when 2 amlogic boards are connected
<narmstrong>
f_: so this is a big question, because libusb makes it very hard to distinguish 2 same devices… there’s no way to pass a device path, but in recent version you can pass the « usb path » I plan to do add support for this
<f_>
Is it a question?
<narmstrong>
I need it anyway for my test farm, today I simply limit the gitlab job concurrency which is quite a bad solution
<f_>
I mean it works
<f_>
I have lepotato + kii-pro connected and can boot both from USB without trouble.
rockosov has joined #linux-amlogic
f11f12 has quit [Quit: Leaving]
djrscally has quit [Ping timeout: 252 seconds]
<exkc>
time to creak s912 i guess
vagrantc has joined #linux-amlogic
mripard has quit [Ping timeout: 245 seconds]
JohnnyonFlame has joined #linux-amlogic
<f_>
exkc: SPL on S912?
jacobk has joined #linux-amlogic
<narmstrong>
S912 is basically gxl with a second cpu cluster, so I think the gxl bl2 would work but bl31 should be fixed to bring up the second cluster
<f_>
Pretty much..
<narmstrong>
It’s even dumber that you think, the second cpu cluster hasn’t a dedicated pll, it uses the fixed 1ghz system pll fixed divider…
<f_>
Is GT1 gxm?
<f_>
It seems like gxm uses gxl BL2..
<rockosov>
narmstrong: Hello Neil! I hope you're doing well. We have sent you a pull request to the pyamlboot repo with the optimus (old) and ADNL (new) flashing protocols support. Please take a look when you have some time :-) https://github.com/superna9999/pyamlboot/pull/20
<rockosov>
It doesn't support secureboot flows in the current implementation, but non-secureboot flashing are working pretty well on the macos, windows, linux systems
<narmstrong>
rockosov: hi y s I saw the pr and I started looking at it, thank you very much so far it’s an excellent work
<narmstrong>
I’ll do a more extensive review in the week
<rockosov>
Actually, macos has problems on the some x86 stations, and M1 macbooks doesn't work currently. But it's in the progress.
<rockosov>
narmstrong: Thank you!
<rockosov>
narmstrong: Additionally, in the very near future (before NY I hope), we plan to send a U-Boot patch series with ADNL on-device client implementation. The Optimus client is currently in progress
<rockosov>
It will support all custom Amlogic tools and pyamlboot
psydroid2 has quit [Remote host closed the connection]