14:15 <wwilly> krzk, hi, do you actually run your https://git.kernel.org/pub/scm/linux/kernel/git/krzk/linux.git/ for-next on a odroid-xu3 or 4??

14:17 <wwilly> directly after boot, just logged from uart https://pastebin.com/yvsWHY69

14:17 <wwilly> 5.6.0-rc1-13504-g0c09ac8e27a0, pulled from 2 hours ago

14:30 <krzk> wwilly: Hi, I do not have a working XU3 anymore... and I did not have XU4, so answer is not

14:30 <krzk> but it is running on HC1 always

14:31 <wwilly> do they have same network chara?

14:31 <krzk> wwilly: latest boot for my repo is: https://krzk.eu/#/builders/21/builds/1801

14:32 <wwilly> I'm not sure how to read this page

14:32 <krzk> wwilly: it's pretty standard Buildbot :)

14:33 <krzk> wwilly: just find step "Boot: odroidhc1" and read serial

14:33 <krzk> there is no such warning you have

14:33 <krzk> let me take a look at later ones

14:33 <wwilly> ok

14:34 <krzk> wwilly: and here is dump of warnings from dmesg:

14:34 <krzk> https://krzk.eu/#/builders/21/builds/1801/steps/18/logs/stdio

14:34 <krzk> so answer is no - I did not see warning of yours

14:35 <wwilly> ok, it's when I try to ssh on it

14:35 <wwilly> so, boot, root login in uart no prob, + simple ping no prob + ssh make a fault

14:36 <krzk> Your error comes from network watchdog for XU3 network card, my boards do not have it

14:36 <wwilly> ok

14:36 <wwilly> is it possible to deactivate the watchdog?

14:36 <krzk> I can try however XU, I think it also had this adapter

14:38 <krzk> wwilly: regular watchdog yes, but this one I don't know

14:39 <wwilly> shit...

14:39 <PabloPL> krzk: Hi, do You have maybe espresso7420 or other 7420 board connected to Your build farm ?

14:40 <krzk> PabloPL: nope, never received it from Samsung (and I think it is not official and not easy to get)

14:41 <wwilly> krzk, do you want a mail on the mailing list to have a trace of the error I got?

14:41 <PabloPL> krzk: Ok, i was asking because i'm getting some strange error (cpu stall on all cores) on mainline on 7420 (Galaxy s6 edge)

14:41 <wwilly> sadly, hard for me to trace that down, and also no time currently, I have a paper deadline really soon

14:43 <krzk> wwilly: public bug reports are always welcomed :) maybe someone will take a look. If it is reproducible always, try bisecting. If it points to recent commit enabling SCHED_MC, then it's not really conclusive... it would require enabling this option on earlier versions.

14:43 <krzk> wwilly: Understand

14:43 <wwilly> krzk, I remove sched_mc & energy model too to check it's not because of this already, the same

14:43 <krzk> snawrocki: Did you see something like this? (mentioned by wwilly)

14:44 <wwilly> krzk, in fact, I tested your branch because of mc and em :)

14:48 <wwilly> krzk, snawrocki after a little while more: https://pastebin.com/3YDv6xzP

14:48 <wwilly> related to the same issue I guess

14:50 <wwilly> both are without SCHED_MC and ENERGY_MODEL

14:51 <krzk> also on XU3? Without SCHED_MC and ENERGY MODEL?

14:51 <wwilly> on xu3 yes

14:51 <wwilly> without SCHED_MC and ENERGY MODEL yes

14:53 <krzk> wwilly: XU booted fine on latest linux-next (and XU also comes with smsc95xx)

14:55 <wwilly> krzk, booted and accessed via ssh?

14:57 <wwilly> because this thing happen when I try ssh on it

15:00 <krzk> wwilly: yes, all tests are run through SSH so it's included

15:00 <krzk> Also this is NFS root boot so network has to work reliably - otherwise it would not come up

15:11 <snawrocki> krzk: I haven't touched the kernel for some time, but Marek is running tests periodically and he didn't mention anything about something like this recently

15:12 <snawrocki> I'm building the kernel, wil see in a while if I can reproduce that bug

15:23 <snawrocki> krzk for-next branch works well with exynos_defconfig on XU3, I tried UART and SSH

15:23 <snawrocki> 0c09ac8e27a03 (HEAD, krzk/for-next) Merge branch 'next/dt' into for-next

15:27 <wwilly> uhm ok ...

15:27 <wwilly> so it's my problem only...

15:27 <wwilly> shit again

15:33 <snawrocki> wwilly: is network working after that warning?

15:33 <wwilly> randomly

15:34 <wwilly> sometime, for a few seconds the ping is lost

15:34 <wwilly> during this lost pings, ssh isn't working, but uart yes

15:35 <wwilly> uart responding normally

15:37 <snawrocki> I see, I'm afraid the best option might be to try to bisect that

15:38 <snawrocki> what's the newest kernel that works for you?

15:38 <wwilly> I was working nicely with 5.3.11

15:39 <snawrocki> Ok

15:39 <wwilly> I back patched dts stuff to this version the one I wanted to try from last krzk sources

15:40 <wwilly> so will continue with this one for my paper deadline, after the deadline I will check which kernel stable release becomes buggy

15:40 <wwilly> I will start playing next week snawrocki krzk

17:07 <PabloPL> krzk: You know what is status of https://patchwork.kernel.org/cover/10646929/ ? (why it's not in mainline)

17:09 <PabloPL> krzk: looking for fix for my booting issue on 7420 (getting cpu stall after "clocksource: Switched to clocksource arch_sys_counter" - will check if it would fix this)