_whitelogger has joined #linux-exynos
afaerber has quit [Ping timeout: 246 seconds]
afaerber has joined #linux-exynos
chewitt has joined #linux-exynos
mszyprow has joined #linux-exynos
GrimKriegor has quit [Quit: ZNC - https://znc.in]
GrimKriegor has joined #linux-exynos
GrimKriegor has quit [Changing host]
GrimKriegor has joined #linux-exynos
dllud has quit [Remote host closed the connection]
wwilly has joined #linux-exynos
dllud has joined #linux-exynos
chewitt has quit [Quit: Adios!]
PabloPL has joined #linux-exynos
PabloPL has quit [Client Quit]
mps has joined #linux-exynos
_whitelogger has joined #linux-exynos
<mps> mainline kernel 5.8 doesn't boot on my peach pi samsung chromebook, while 5.7 and earlier works very fine
<mps> anyone here know what should be changed/tweaked, or does 5.9 works
PabloPL has joined #linux-exynos
<krzk> mps: try the latest stable v5.8
<mps> krzk: I tried 5.8.14 but without success
<mps> 5.8.16 works?
PabloPL has quit [Quit: PabloPL]
PabloPL has joined #linux-exynos
PabloPL has quit [Client Quit]
<wwilly> I'm curious, does a chromebook print to the screen the boot info? does it expose an uart serial console or something like this?
<mps> wwilly: no by default, but on some serial console can be added, or put u-boot in boot SPI flash
<wwilly> "can be added" like wiring yourself on the motherboard?
<mps> yes
<wwilly> ok, I'm surprised, what stuff you need have a working UART electronically speaking?
<mps> depends on model
<mps> I read about that for samsung one plus (rk3399) but didn't tried, it works fine even without serial console
<mps> (well, it doesn't work fine but it works)
mszyprow has quit [Ping timeout: 246 seconds]
mszyprow has joined #linux-exynos
PabloPL has joined #linux-exynos
<mps> even 5.9.1 doesn't boot on peach pi chromebook (exynos 5800)
PabloPL has quit [Quit: Konversation terminated!]
<mszyprow> wwilly: all exynos based chromebooks have debug uart available on some test points on the motherboard. all you need is to solder some uart->usb adapter
<mszyprow> wwilly: the hardest thing is to find the one for 1.8V
<mszyprow> wwilly: although this requirement can be hacked, I've used standard 3.3V ftdi and resistor-based voltage divider
<mszyprow> mps: strange, v5.9 works fine here on my peach-pi
<mszyprow> mps: which u-boot version do you use?
<mps> mszyprow: huh, can't remember now, 2020.04 maybe
<mps> but u-boot works, I can boot 5.7 quite fine
<mszyprow> hmm
<mszyprow> I'm surprised that so recent u-boot still boots on peach-pi
<mszyprow> last time I've checked there were issues
<mszyprow> and I had to use quite old one to boot it
<mps> machine is now with my son and he lives on other part in city
<mps> not now with me*
<mszyprow> I use u-boot 2016.01
<mps> I'll check tomorrow and tell exact version
<mszyprow> mps: well, I will not play with uboot now as I have only remote access to chormebooks and I won't risk
<mps> understand
<mszyprow> mps: here v5.9 boots fine from tftp (via usb ethernet)
<mszyprow> mps: just exynos_defconfig and arch/arm/boot/dts/exynos5800-peach-pi.dtb
<mszyprow> mps: let me check v5.8
<mps> I used .config from working 5.7 to build new kernels, tomorrow I will try with default_exynos config
<mszyprow> mps: I don't remember any serious issue with it...
<mszyprow> the only really broken was v5.6 release
<mps> mszyprow: yes, also had issues with it
<mszyprow> v5.6 had broken scatter list handling code merged to DRM one day before the release... :/
<mps> let me see if i can found my notes about u-boot in archive
<mszyprow> mps: just checked vanilla v5.8 and it also boots fine here
<mszyprow> mps: maybe there are different revisions of peach-pi... but I have no idea how to check that
<mszyprow> mps: peach-pi are hard to buy anyway
<mps> mszyprow: I bought it 4-5 years ago
LiquidAcid has joined #linux-exynos
<mszyprow> mps: on the other hand, there are lots of used/recycled peach-pit boards available on the market here in Poland
<mps> mszyprow: nice to hear, Poland is not far for me
<mps> but nowadays when travels are limited, hmmm
<mps> mszyprow: found build script for u-boot but didn't noted version in it. will look tomorrow when son come here at work
<mszyprow> mps: if you have some time, you may try to bisect it, maybe the patch pointed by bisect will give us some hints
<mps> mszyprow: sorry, I'm short on time, preparing kernel 5.9 for alpine linux and trying to fix crystal language and some other things for next stable release (I'm alpine linux developer)
<mps> (alpine eats all my free time)
<mps> and, iirc arnd told me that he also owns peach pi chromebook, maybe will help
<arnd> I haven't really powered it on in a long time
<mps> mszyprow: also, I'm not in hurry, 5.7 works fine
<mps> arnd: np, I hope I will found problem when I finish with jobs in current queue
<mps> s/found/find/
glLiquidAcid has joined #linux-exynos
<mps> mszyprow: hm, maybe I found problem, CONFIG_THUMB2_KERNEL=y and I know that it stopped to work with newer binutils and kernels
LiquidAcid has quit [Remote host closed the connection]
glLiquidAcid has quit [Remote host closed the connection]
LiquidAcid has joined #linux-exynos
<mszyprow> mps: well, i use old gcc/binutils
<mszyprow> mps: gcc 4.9.4
<mps> mszyprow: I use binutils-2.35.1 and gcc-10.2.0, (current versions in alpine edge testing version)
<mps> and I had to disable CONFIG_THUMB2_KERNEL for armv7 kernel when upgraded to 5.8
wwilly has quit [Quit: This computer has gone to sleep]
mszyprow has quit [Ping timeout: 272 seconds]