stikonas has quit [Remote host closed the connection]
field^Zzz1 has quit [Ping timeout: 265 seconds]
kevery has joined #linux-rockchip
mniip has quit [Ping timeout: 600 seconds]
mniip has joined #linux-rockchip
vagrantc has quit [Quit: rebooting...]
<JPEW>
Jin^eLD: tlwoerner: Ah, I think the confusion is that he replied to me off list saying it worked, and I told him to reply on-list for posterity :)
<JPEW>
Looks like it was just a reply to the latest message in the thread, which is pretty easy when using the gmail interface :)
vagrantc has joined #linux-rockchip
<tuxd3v>
vagrantc, inode, Jin^eLD: its something that uboot does, for sure, when kernel tries to start if partition were not umounted in a clean way, it run fsck
<tuxd3v>
I don't know who requested fsck, because I don't
<tuxd3v>
but yet it tries to run fsck in mmcblk0p2... which doesn't make any sense since mmcblk0p2 is the uboot.itb ???? :D
<vagrantc>
what aarguments are passed to the kernel?
<vagrantc>
or in /etc/fstab?
<vagrantc>
also, device numbers are not predictible ... better to use PARTUUID or UUID
<vagrantc>
what does the kernel actually receieve, since there are a number of empty variables there
<tuxd3v>
who is running fsck.ext4 on mmcblk0p2( uboot.itb )
<vagrantc>
probably doing exactly what it was told ... but most likely an initrd or part of the init process
<tuxd3v>
this is behaviour that I could determine until now
<vagrantc>
grep 'Kernel command line:' from the boot log
<vagrantc>
that's the first place i'd look
<tuxd3v>
if partitions as fscked before( clean ), uboot prints nothing, but it also doesn't boot the kernel, I think
<tuxd3v>
if i push the power cable, letting partitions not unmounted in a clean way, then it on startup tries to fscked the mmcblk0p2, that doesn't exist
<tuxd3v>
the sbc bootlog?
<tuxd3v>
let-me mount the sdcard and check :)
<tuxd3v>
thanks for the hint :)
<tuxd3v>
the sdcard contains a empty /var/log :(
<vagrantc>
uboot wouldn't fsck
<vagrantc>
tuxd3v: or the serial console output
<vagrantc>
uboot isn't going to fail to boot a kernel if the filesystem is clean ... that's not a ... linear order of events
<tuxd3v>
if fylesystem not clean it runs fsck
<tuxd3v>
but on the uboot.itb "pseudo-partition"
<tuxd3v>
he call it mmcblk0p2
<vagrantc>
but the kernel has already booted if you're running fsck
<tuxd3v>
not sure
<vagrantc>
i'm quite sure
<tuxd3v>
because later he shows a message saying that it fails to find 'fsck.ext4' file not found
<vagrantc>
something is telling your operating system to run fsck on that partition
<tuxd3v>
but hearlier on the process he made another fsck.ext4 in a correct way
<vagrantc>
but your kernel is already loaded and running, because fsck doesn't run on baremetal
<tuxd3v>
what I understand is that he tries to fscked all the partitions that shows in gpt
<vagrantc>
so i would check your /etc/fstab and the boot arguments passed to the kernel
<tuxd3v>
because I stupidly created 3 fake partitions he tries to clean them..
<vagrantc>
i am doubtful that it's just randomly checking partitions ... what OS is this?
<tuxd3v>
heheh
<tuxd3v>
its my own
<tuxd3v>
:)
<tuxd3v>
but it works in another sbcs correctly
<vagrantc>
ok, so it could be doing anything.
<vagrantc>
you might want to have a word with your operating system developer :P
<tuxd3v>
hehehe
<tuxd3v>
it could be
<tuxd3v>
for this case that the rootfs is playing tricks with me
<tuxd3v>
and I because already created dozens of them in the same way, are trusted it too much
<tuxd3v>
trusted -> trusting
<vagrantc>
it definitely seems wrong for it to run fsck on arbitrary filesystems ...
<vagrantc>
tuxd3v: is your OS *based* on any other more well-known OS ?
<tuxd3v>
yes it is devuan based
<tuxd3v>
rootfs
<vagrantc>
ok, which is pretty much debian ...
<vagrantc>
so, check your kernel boot args and /etc/fstab
<tuxd3v>
yes
<tuxd3v>
it works in all boards that I have it loaded
<vagrantc>
ok, i'm done. :P
<tuxd3v>
vagrantc, thanks for the help :)
<tuxd3v>
but still, I believe the bootloader process is playing tricks with me
<tuxd3v>
:D
<tuxd3v>
but its true, not even a keyboard is recognized
<tuxd3v>
hdmi output only when it gives me a error
<tuxd3v>
in 50 attemps I see something :)
<tuxd3v>
vagrantc, my filesystem has problems, related with my rootfs coockings :S
<tuxd3v>
shame on me :(
<wens>
hmm, my rk3399 boards don't seem to reboot properly
vagrantc has quit [Quit: leaving]
ganbold has quit [Remote host closed the connection]
ganbold has joined #linux-rockchip
ganbold has quit [Remote host closed the connection]
kaspter has quit [Ping timeout: 240 seconds]
kaspter has joined #linux-rockchip
<anarsoul>
wens: update your ATF
_whitelogger has joined #linux-rockchip
kaspter has quit [Ping timeout: 264 seconds]
camus1 has joined #linux-rockchip
camus1 is now known as kaspter
_whitelogger has joined #linux-rockchip
<wens>
it is up to date
<wens>
roc-rk3399-pc hangs at "reboot: Restarting system"
<anarsoul>
up to date as from git master or latest release?
<wens>
git master
<wens>
pcie mezzanine is installed and enabled, if that makes any difference
lkcl has quit [Read error: Connection reset by peer]
<Esmil>
mps: yeah, suspend has never been stable for me, but I haven't noticed it should be the emmc in particular
<Esmil>
also I haven't used it in a while
<Esmil>
i get occasional timeouts when initialising usb3 (dwc3) though. right now I'm running the last commit here to see if i can catch exactly where it happens: https://github.com/esmil/linux/commits/kevin
<mps>
for me suspend-to-ram works fine on external mmc and ssd disk over usb-c
<mps>
and resume ofc
<mps>
suspend-to-ram works also on emmc, but after resume and some random times it 'crashes', as I wrote in bug report on above link
field^Zzz1 has joined #linux-rockchip
<Esmil>
yeah, i'm running from emmc still so that maybe why suspend hasn't been stable for me
lkcl has joined #linux-rockchip
<Jin^eLD>
JPEW: ok, that makes more sense then :)
ganbold has joined #linux-rockchip
lkcl has quit [Read error: Connection reset by peer]
return0e_ has joined #linux-rockchip
return0e has quit [Ping timeout: 264 seconds]
matthias_bgg has quit [Ping timeout: 250 seconds]
<mps>
Esmil: built kernel with your .config (adding f2fs) and boot, but I don't have display working, it is blank
<mps>
rebooting from mmc and looking at log files I see it boots
lkcl has joined #linux-rockchip
<Esmil>
mps: interesting. i'm still using the bootloader it came with, did you build your own?
<Esmil>
im just thinking it might initialise something differently
kaspter has quit [Remote host closed the connection]
kaspter has joined #linux-rockchip
<mps>
I use vboot-utils to create kernel and flash it to boot partition
<mps>
tried to make u-boot but didn't had success
<Esmil>
yeah, i do the same
stikonas has quit [Remote host closed the connection]
matthias_bgg has joined #linux-rockchip
stikonas has joined #linux-rockchip
<mps>
ok, I will pick some config option from your config and apply it to mine to see will that help
<Esmil>
could you paste the .config you just built?
<Esmil>
no no, that's fine. i just meant you shouldn't paste it in this channel. that's all
<mps>
ah, ofc, it would be a lot of msgs here :)
<Esmil>
yeah, no changes there to graphics stuff. do you have console=tty1 in your /proc/cmdline ?
<mps>
here it is: cros_secure console=tty0 console=ttyS2,115200n8 earlyprintk=ttyS2,115200n8 init=/sbin/init root=PARTUUID=4......... rootwait rw noinitrd loglevel=7
<mps>
current cmdline, I mine
<Esmil>
yeah, maybe try just console=tty1. i mean it's just weird that my screen lights up fine, but yours doesn't with almost identical .configs
<mps>
and this is my vboot cmdline file: console=tty0 console=ttyS2,115200n8 earlyprintk=ttyS2,115200n8 init=/sbin/init
<mps>
but, right now I have to go to $day_job, maybe will try this night, hope will not be tired too much
indy has quit [Ping timeout: 265 seconds]
<mps>
btw, two or three weeks ago I noticed that when disabling hdmi for rockchip it looks more stable but then sound doesn't work
sb35 has joined #linux-rockchip
indy has joined #linux-rockchip
lkcl has quit [Read error: Connection reset by peer]
kevery has quit [Ping timeout: 256 seconds]
lkcl has joined #linux-rockchip
matthias_bgg has quit [Quit: Leaving]
matthias_bgg has joined #linux-rockchip
lkcl has quit [Read error: Connection reset by peer]
PoueT has joined #linux-rockchip
chewitt has joined #linux-rockchip
jelly has quit [Quit: ""]
jelly has joined #linux-rockchip
MoeIcenowy has quit [*.net *.split]
tuxd3v has quit [*.net *.split]
j45 has quit [*.net *.split]
niceplace has quit [*.net *.split]
MoeIcenowy has joined #linux-rockchip
niceplace has joined #linux-rockchip
j45 has joined #linux-rockchip
tuxd3v has joined #linux-rockchip
<repk>
So I manage to boot u-boot SPL but it seems it does not load TPL on nanopi-m4. It reboot at "Trying to boot from MMC1"
<repk>
I have dd u-boot.itb at sector 16384 though
vicencb has joined #linux-rockchip
<repk>
s/TPL/bl31/
<tuxd3v>
repk, does you have a rootfs?
<repk>
no
<repk>
tuxd3v: But I was expected to see BL31 log and u-boot boot menu
vagrantc has joined #linux-rockchip
<tuxd3v>
about that, I don't know
<tuxd3v>
what I know is that MMC1 is the sdcard i think
<tuxd3v>
and the rootfs needs a fstab with mmcblk1p?
<tuxd3v>
not mmcblk0p?
<repk>
tuxd3v: oh well I am way before that stage
<repk>
I am at bl31 loading
<tuxd3v>
how are you booting?
<tuxd3v>
"Trying to boot from MMC1" is a commun message
<tuxd3v>
it means it will search mmcblk0p? for 'extlinux.conf' or 'boot.scr'
<tuxd3v>
what is your uboot scheme partition
<tuxd3v>
use a paste web tool :)
<repk>
tuxd3v: IMHO "Trying to boot from MMC1" in SPL does not mean that. It will look for u-boot.itb at sector 0x4000 to start bl31 and then bl33 (i.e. u-boot tpl).
<repk>
So I don't have a partition scheme so far
<repk>
only iidbloader.img at sector 64 and uboot.itb at sector 16384
<repk>
this should be enough to start at least arm-trusted-firmware bl31 and u-boot tpl
<repk>
but it fails at SPL
<tuxd3v>
well idbloader.img + uboot.itv is what I have plus bl31
<tuxd3v>
but I don't have bl33
<repk>
well bl33 is u-boot.img
vagrantc has quit [Quit: leaving]
cristian__c has joined #linux-rockchip
<repk>
hm that is bl31_entry()'s call to bl31 @ 0x40000 that is failing
cristian_c has quit [Ping timeout: 256 seconds]
<tuxd3v>
does any one managed to get sound via Jack or HDMI?
<repk>
Ok I have UART now (mainline ATF has 115200 baudrate configured instead of 1500000). And I have the callstack, that is an assert on bl_params version from u-boot. Maybe that changed recently
<repk>
Ok got u-boot working
<repk>
for the record ATF need to be compiled with "ERROR_DEPRECATED=0", as u-boot spl seems to still use "old" bl31_param structure