<juri_>
the machine is running all free software, armbian with the RYF approved wificard from thinkpenguin, running the open firmware for it. in the other usb port is a usb hub, with my 3d printer (rambo board based lulzbot) plugged into it.
rellla has joined #linux-sunxi
<mru>
do you have logs from before those rcu messages started?
<mru>
that's probably where things first went wrong
jbrown has quit [Ping timeout: 246 seconds]
<juri_>
it aparently died in something related to printing.. or at least, it took pronsole with it. pronsole was unresponsive (and the printer was jammed) way before the oops.
<mru>
that call stack should be very well tested indeed
<KotCzarny>
usb 1-1: ath: firmware panic! exccause
<KotCzarny>
i wonder if it has anything to do with it
<juri_>
I wouldn't think so, but.. after all, it seems to have recovered.. and i am logging in via wireless.
<KotCzarny>
usbprinter reconnected few times too
<juri_>
yeah, that's power cycles.
<juri_>
that's a long period of time. this exact same print was successful during one of those previous runs.
<KotCzarny>
i would bet on some usb driver misbehaviour, but i'm not an expert
<mru>
usb is always a likely suspect
<mru>
that said, the cdc-acm driver is probably not too terrible
tnovotny has quit [Remote host closed the connection]
ldevulder_ has joined #linux-sunxi
<juri_>
mru: since you seem to be the expert of the moment, is there anything else you'd like to know, before i reboot it.. or anything i should do before running this print again, to make sure all is healthy, and the next report is higher quality?
<KotCzarny>
can you run the top ? is there some hung process on the list?
<juri_>
there was, but a signal 15 killed it.
jonasbits has quit [Ping timeout: 268 seconds]
<juri_>
the terminal it's in still has not returned to me, however.
<mru>
it won't return
<juri_>
oh. the process is still running.
jbrown has joined #linux-sunxi
<mru>
the system is all messed up
<mru>
needs to be rebooted
ldevulder has quit [Ping timeout: 276 seconds]
<mru>
is this the first time something like this happens?
<juri_>
yes.
<KotCzarny>
does rootfs still respond?
<KotCzarny>
ie. swap works etc
<mru>
I'd just sync and pull the power at this point
<juri_>
this system has not been in heavy use, but it has completed many prints.
<juri_>
yes, rootfs is still readable.
<juri_>
0 swap was in use.
<KotCzarny>
there is always a chance for: 1/ power fluke that flipped some bits, 2/ radiation fluke that flipped some bits, 3/ silicone fluke that flipped some bits, 4/ bug in kernel
<juri_>
I have no radiation sources out of the ordinary, and am in a building with 3 floors above me. :)
<KotCzarny>
does that print job run as root?
<mru>
it's almost certainly _some_ kind of memory corruption
<juri_>
KotCzarny: no, i run it as a user.
<juri_>
additionally, the room is 73 degrees farenheight, with 38% humidity.
<KotCzarny>
btw. if you live on the first floor or in basement, radioactive gases like radon like to creep near the floor
<KotCzarny>
but that's mountain area hazard
<juri_>
third floor.
<mru>
radon is a regional thing, not necessarily linked to mountains
<juri_>
new building, and in a corner where one of the walls is made of concrete.
<KotCzarny>
mru: but often related to mining operations
<KotCzarny>
juri: also remember that russian government never announces radioactive events immediately
<KotCzarny>
:>
<mru>
radon is produced by radioactive decay of uranium and other heavy elements
<KotCzarny>
which also applies to .us gov
<juri_>
KotCzarny: I'm in berlin, at the point that the line that used to be the wall meets the ringbahn. :)
<KotCzarny>
anyway, i would worry if it repeats
<KotCzarny>
right now it looks like random event
<mru>
random bit flips happen
<juri_>
I'm just trying to make sure i report this correctly. :)
<KotCzarny>
if you find a way to reproduce it, sure, at this moment even if one would like to debug it, it's impossible
<KotCzarny>
still betting on usb though
<juri_>
I do not have a CPU or memory heatsink on the machine. i am using it with the original case.
<juri_>
rebooting.
<mru>
do a full power cycle, just to be sure
rellla has joined #linux-sunxi
<juri_>
powered by a 5V 3A power supply.
<juri_>
(power cycled)
<mru>
should be plenty
<KotCzarny>
heatsink isnt bad idea
<mru>
shouldn't be needed though
<KotCzarny>
they are cheap but shave few degrees
<juri_>
I'll let you know if it reproduces.
souther has quit [Ping timeout: 276 seconds]
souther has joined #linux-sunxi
* KotCzarny
imagines reproducing bugs. ugh
<mru>
you need both a male and a female for that to happen
<fALSO>
;-)
<KotCzarny>
kernel might have all kinds
<juri_>
indeed. ;)
selfbg has quit [Quit: selfbg]
xqdzn has joined #linux-sunxi
Perlovka has quit [Quit: Perlovka]
<montjoie>
anyone with an sun7i-a20-olinuxino-lime2 using uboot ethernet ?
Perlovka has joined #linux-sunxi
<montjoie>
seems that the gigabit is unstable
Perlovka has quit [Client Quit]
Perlovka has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
Perlovka has quit [Quit: Perlovka]
Perlovka has joined #linux-sunxi
reinforce has quit [Quit: Leaving.]
<libv>
there's the tx delay thing in newer lime2 versions with the different phy?
<libv>
but i have not gone and tried to use ethernet from uboot directly
xqdzn has quit [Remote host closed the connection]
<montjoie>
the tx delay thing ? not sure to have understood you
jonasbits has joined #linux-sunxi
tllim has joined #linux-sunxi
reinforce has joined #linux-sunxi
aloo_shu has joined #linux-sunxi
cnxsoft has quit [Remote host closed the connection]