robogoat has quit [Read error: Connection reset by peer]
robogoat has joined #linux-rockchip
<tony_>
I want to talk about LVDS resolution adjust.
GriefNorth has quit [Ping timeout: 240 seconds]
<tony_>
the laptop lcd use LVDS too.
<tony_>
why the laptop can adjust resolution ?
<tony_>
how to do it ?
naobsd has joined #linux-rockchip
naobsd has quit [Quit: naobsd]
naobsd has joined #linux-rockchip
naobsd has quit [Quit: naobsd]
Bludot has quit [Quit: Connection closed for inactivity]
em|biketron has quit [Read error: Connection reset by peer]
ganbold has quit [Ping timeout: 255 seconds]
visitor-14521 has joined #linux-rockchip
visitor-14521 is now known as visitor-14524
<visitor-14524>
Hmmm... Current state: Recovery with stock kernel image -- works. Recovery with a self-compiled kernel image -- display backlight switches on, then nothing. Are there any "usual suspects" to check why there isn't even the boot logo to see?
<amstan>
visitor-14524: are you talking about the chromebooks... or..?
<visitor-14524>
RK3188 based Android tablet, pretty sure the name "Odys Noon Pro" isn't telling anything, says to run a RK3188_NOON_PRO_Android4.4.2 image.
<amstan>
idk as much then
<amstan>
did you compile the stock kernel tree?
<amstan>
and did you use a similar compiler(as in.. not the newest gcc 5.1)?
<visitor-14524>
Using gcc-4.8, and kernel is said to be compiled with gcc-4.4. Sadly not the stock kernel sources as of yet, but tried with github "linux-rockchip" from vcarious sources.
nighty^ has joined #linux-rockchip
<visitor-14524>
So okay, I could try with downgrading the toolchain to exactly the same gcc version as the stock kernel advertises, and maybe switching off as many bells and whistles to eliminate possible causes of errors... But is it possible to have printk messages logged on the screen?
<tony_>
> it possible to have printk messages logged on the screen?
<tony_>
visitor-14524, have not a serial ?
<visitor-14524>
Uncertain. Maybe something on the inside of the device, but at the moment I'm (still) a bit loathe to disassemble that device.
<tony_>
amstan, why I can not adjust resolution on rk3188 kernel ?
<tony_>
amstan, would you give some advices ?
<rperier>
hi all
<visitor-14524>
Hello.
<visitor-14524>
Sadly, in the (g)olden days of Linux printk messages were dumped on the screen for those interested in looking at them. Hmm. Perhaps I need to switch _off_ the boot logo for that.
<tony_>
visitor-14524, why do you want to do that ?
<sjoerd>
visitor-14524: that can only happne if the kernle managed to initliaze the display controller and everything related to it
<tony_>
visitor-14524, serial will be better than screen.
<visitor-14524>
Serial would be better, sure, but I'm not that sure about my skills in hardware hacking or being able to locate the needed pins or even just soldering spots on the board...
<visitor-14524>
So yes, I'm trying to exhaust other options first before I get desperate :)
<visitor-14524>
Another idea would be to enable ttyUSB0 logging, but perhaps that requires even more to initialize than what would be needed to get the display up and running.
<sjoerd>
visitor-14524: not more, but a different batch of things
<sjoerd>
figuring out how to get serial out is really the best option
<sjoerd>
other you do really need the kernel sources/dtb to figure out how the hardware is put toghether
<visitor-14524>
dmesg output from the stock kernel booting up did offer some good clues. As well as the dump of the gpio pin allocation. But best I'll pare the configuration down to the bare essentials first and try to get some meaningful debug output.
<visitor-14524>
The fact that the backlight turns on means that at least _something_ seems to get working. If I'm not completely mistaken and it is the bootloader or such that did it.
<visitor-14524>
Proprietary Rockchip bootloader, V1.03, in that case.... Yes, stone-age - that device never received an update, not for the lack of trying... :(
<sjoerd>
visitor-14524: driving the backlight is quite seperate from driving a display subsystem though
<visitor-14524>
I know... but it tells me that it got that far during booting before things get haywire. If, for example, USB would be initialized earlier, it could be that I can use ttyUSB as the debug console, for example.
<visitor-14524>
Heh, if I get crazy enough, I'd implement something like switching the backlight on and off in a pattern, much like a BIOS beep code on a borked PC... :)
<visitor-14524>
That actually might be worth a look. Wonder why it never made it into the kernel... now to learn Morse code :)
<rperier>
"will blink out the panic message in morse code using the keyboard LEDs" <--- seriosuly ? :D
<sjoerd>
I just remembered it because i found it funny :)
<sjoerd>
it's ancient
<sjoerd>
back in the day when i think ps/2 keyboard were still qutie common on PCs
<rperier>
it is, indeed...
<visitor-14524>
But it is an idea which could have quite some merit. Found other suggestions to use the PC speaker and a smartphone with an app to transmit the debug data via sound.
<rperier>
ah ah people are crazy, that's funny and useful... but that's "morse code" :D
<visitor-14524>
Wonder why it never made it into the mainline kernel? Or a pared down version with something akin to BIOS beep codes. It's definitely something one would use when facing a borked embedded system with no access to any serial port or such.
<sjoerd>
visitor-14524: probably because tit ends to be simpler to get serial out
<sjoerd>
and a lot less cumbersome to work with
naobsd has joined #linux-rockchip
naobsd has quit [Client Quit]
visitor-14524_ has joined #linux-rockchip
visitor-14524 has quit [Ping timeout: 246 seconds]
visitor-14524_ is now known as visitor-14524
ssvb has joined #linux-rockchip
<karlp>
sjoerd: my washing machine does some audio squealing for diagnostics.. it's paired with a phone app that you hold near the machine that listens to it.
<sjoerd>
karlp: sure but the amount of information you get out of that is rather limited isn't it ;)
* karlp
shrugs. it's just like any old modem.
<karlp>
more than blinking morse code for manual operators, that's jsut a garbage.
<sjoerd>
Sure but what we're talking about here is potentially quite low-level diagnostics
<sjoerd>
blinking a led tends to be a lot more trivial then outputing audio
<visitor-14524>
Just to think of it, BIOS beeps are even more limited in the amount of information they contain, yet they're a good start to find what you're looking for.
maz_ has joined #linux-rockchip
tony_ has quit [Quit: Leaving]
VargaD has quit [Ping timeout: 246 seconds]
rperier has quit [Ping timeout: 246 seconds]
sunilmohan_w has quit [Ping timeout: 244 seconds]
sunilmohan_w has joined #linux-rockchip
rperier has joined #linux-rockchip
dianders has quit [Ping timeout: 246 seconds]
dianders has joined #linux-rockchip
VargaD has joined #linux-rockchip
Danukeru has quit [Ping timeout: 246 seconds]
Danukeru has joined #linux-rockchip
naobsd has joined #linux-rockchip
ganbold has joined #linux-rockchip
GTRsdk has quit [Remote host closed the connection]
ssvb has quit [Ping timeout: 252 seconds]
rperier has quit [Changing host]
rperier has joined #linux-rockchip
ganbold has quit [Quit: This computer has gone to sleep]
ganbold has joined #linux-rockchip
ganbold has quit [Client Quit]
ganbold has joined #linux-rockchip
dowu has joined #linux-rockchip
<dowu>
Hi, all.
<dowu>
I am totally new on rockchip but with a little bit knowledge of linux.
<dowu>
I installed rkflashtool properly. and have an error message, "rkflashtool: info: rkflashtool v5.2; rkflashtool: fatal: cannot open device"
<dowu>
what is wrong?
<karlp>
that's the sort of thing you'd get with no perms for the device.
<dowu>
what is perms?
<visitor-14524>
Try again as root or with the appropiate udev entry.
<pjr>
or device not in "recovery mode"?
<dowu>
root has the same error message too.
<pjr>
ls -lah /dev/ttyUSB*
<visitor-14524>
and lsusb, please.
<dowu>
my env. is vmwareplayer and ubuntu
<pjr>
as root: lsusb -vv
<dowu>
as "root lsusb -vv" too long messages...
<dowu>
what should I say?
lerc has quit [Quit: No Ping reply in 180 seconds.]
<visitor-14524>
Simply lsusb. Do you see your device in the list? Should be something starting with 2207:
<dowu>
Yes, the list has rockchip rk3188
<visitor-14524>
And what exactly does it say for VendorID:ProductID?
<dowu>
idVender 0x2207, idProduct 0x0010
<visitor-14524>
Definitely not USB flash mode, then. that would be 2207:310b (I think), but definitely not :0010
lerc has joined #linux-rockchip
<dowu>
Actually I am a newbie on this board. I do not know who made this board. someone told me this board has 8G flash
<dowu>
Can I change this mode into USB flash mode?
<dowu>
I connected target b/d through USB(OTG) into host PC(Windows 10) and vmwareplayer.
<visitor-14524>
Depends on your device. If it's Android, try "adb reboot bootloader"; "rkflashtool b" would get you out of it again.
<visitor-14524>
But YMMV. Danger zone ahead, don't try if you don't have a backup and all that.
<dowu>
Yes, the target b/d has android. But I want install linux instead of Android.
<dowu>
as the result of "adb reboot bootloader" error: device not found
<dowu>
not at all of toggle switches.
ganbold has quit [Quit: This computer has gone to sleep]
ganbold has joined #linux-rockchip
<visitor-14524>
Just FYI: "Rockchip" and "RK3188" are just the name of the SoC being built in, the hardware surrounding it - starting with display, camera, whatever - does vary wildly. You need to be VERY sure if the image you obtained is one geared for your device.
<dowu>
yes, on right top area of vmware player, an icon is triggered to "connected" which is
<dowu>
I will make phone call to my coworker to get more info of Android and some others
<dowu>
Thank you for your lots of inform.
<dowu>
I will be better step by step...
<visitor-14524>
np. I'm more or less stuck with a self-compiled kernel (for a RK3188 tablet) as well. Just getting the SoC name right in the kernel config is far by not enough.
<visitor-14524>
*RK3188-based tablet, to be more precise.
ssvb has joined #linux-rockchip
visitor-14524 has quit [Quit: It's dead, Jim.]
cnxsoft has quit [Quit: cnxsoft]
chrs_ has quit [Ping timeout: 252 seconds]
<GekkePrutser>
Anyone here who uses the Chromebook flip C100 with the Arch Linux image?
<GekkePrutser>
I just wonder if it's any good (does suspend work, rotation, touch screen etc)
<GekkePrutser>
I just got my flip yesterday and I've been playing with Crouton which works OK, but I'm not a fan of all the Google tracking
<maz_>
GekkePrutser: I use a c201 with the kernel extracted from the Arch image, and it works really nicely.
<GekkePrutser>
Just the kernel alone
<GekkePrutser>
?
<GekkePrutser>
But I'll give it a try, I just don't want to mess up the flash
<GekkePrutser>
(Or brick it, in other words)
<GekkePrutser>
But I suppose I could install it on SD first
<maz_>
GekkePrutser: yes, just the kernel. I use Debian otherwise, so no need for the Arch install.
<GekkePrutser>
Ahh I didn't realise you could do that!
<GekkePrutser>
Do you even get GPU acceleration for video, and suspend?
<maz_>
GekkePrutser: I boot from SD, the chromeos install is untouched.
<GekkePrutser>
Ok I'll try that first so!
<maz_>
I don't care much for the GPU. actually, I explicitely want it disabled.
<GekkePrutser>
Ahh ok
<maz_>
suspend/resime works just fine.
<maz_>
resume*
<GekkePrutser>
Nice!! I never got that working on my AC100
<GekkePrutser>
I work a lot on the console as well but because I will be travelling with the flip I'll probably use it to watch video too
<maz_>
I just have a small script that configures the wakeup source at boot time.
<GekkePrutser>
Nice, I'll have to get started on this.. I also prefer Debian to Arch (just because I'm more familiar with it)
chrs_ has joined #linux-rockchip
dowu has quit [Ping timeout: 246 seconds]
<maz_>
I just moved the sdcard I used on my previous chromebook with a different kernel and modules, fixed the odd issues (name of the network interface, wakeup sources...), and got everything going in about 30 minutes.
Bludot has joined #linux-rockchip
<GekkePrutser>
Nice!!
<GekkePrutser>
I never had a chromebook before but I'm sure I can get it going (eventually :) )
<GekkePrutser>
The AC100 was abit challenge also
<GekkePrutser>
"a big" I mean
tonokip9 has joined #linux-rockchip
JohnDoe_71Rus has joined #linux-rockchip
<mrutland>
Are there any docs on booting rk3368 (with a custom kernel)? I have a Tronsmart Orion R68 Meta, and it looks like the built-in U-Boot doesn't respond to anything by default. I'm somewhat hesitant to reflash things, though I suspect that's necessary.
<mmind00>
the hex values denoting the flash locations should already be visible in the commandline of the upstream kernel
<mmind00>
there should be a "recovery button" somewhere on the device, which will bring uboot into the current rockchip-specific flash-mode
<mrutland>
mmind00: Cheers! So far (yesterday) I wasn't abel to get rkflashtool to identify that the board was present, so I suspect I'm missing something more fundamental w.r.t. how to put the board in the appropriate mode
<mrutland>
mmind00: I wasn't able to spot one, so far, just power and reset
<mrutland>
Many thanks for the pointers above! I'll have another look to see if I missed a header
<mmind00>
mrutland: there seem to be two buttons on the board ... one is obviously the powerbutton and the other one probably the recovery one
<mmind00>
mrutland: procedure normally is having the recovery button pressed when powering on
<mrutland>
Ah; I assumed because it said 'reset' that it wasn't the same thing. I'll give that a shot when I get back to the board in a couple of hours or so :)
<mmind00>
mrutland: hmm could also be a reset button
<mrutland>
mmind00: I haven't tested it so far, so I'll let you know when I find out
<mrutland>
mmind00: I think so, 0x330 was certainly a portion of it.
kido has quit [Ping timeout: 272 seconds]
<mrutland>
That said I was playing fun and gaems trying to used VM, so my problems may be of my own creation
<mrutland>
*trying to use USB in a VM
GriefNorth has joined #linux-rockchip
<mmind00>
mrutland: hehe ... anyway, they're quite creative hiding the recovery stuff ... on one rk3288 box you have to poke through the headphone jack to reach the button behind it :-)
<karlp>
mmind00: interesting, I wonder where they got that jack with a button built in?
pjr has joined #linux-rockchip
pjr has quit [Ping timeout: 248 seconds]
<mmind00>
karlp: nope, not built in ... the button was just positioned directly behind the "hollow" jack
pjr has quit [Ping timeout: 248 seconds]
pjr has joined #linux-rockchip
<karlp>
neat trick.
cristian_c has joined #linux-rockchip
<mmind00>
mrutland: http://www.tronsmart.com/?p=3521 ... zip including the Updating_instruction at the bottom of the list ... TL;DR: the small button is the recovery button
<mrutland>
mmind00: Ah, sweet. I'll give that a go. Thanks for digging around!
<mmind00>
mrutland: glad to help :-)
<cristian_c>
hello
<cristian_c>
I've got a rk3066 device
<cristian_c>
backlight values are inverted in /sys/class/backlight/..../brightness file
<cristian_c>
but it looks good
<cristian_c>
issue is the following one: on the screen/display, I see one screen + 1/3 screen below the prevuous one
<cristian_c>
*previous one
<cristian_c>
so, I see two desktops on the display (second desktop is cut, btw)
<cristian_c>
How can I fix this issue?
<cristian_c>
Any ideas?
GriefNorth has quit [Ping timeout: 240 seconds]
JohnDoe1 has joined #linux-rockchip
JohnDoe_71Rus has quit [Ping timeout: 260 seconds]
markm has joined #linux-rockchip
<mmind00>
norris: you might've seen, I've picked up your regulator changes just now ... so only need to send the new board in future versions :-)
<norris>
mmind00: yes, just saw, but thanks for the notice!
<mmind00>
norris: btw. do you plan to send a next board-dts version in the near future?
<norris>
mmind00: you mean the jaq dts? yes, I was thinking of sending today
<mmind00>
norris: great ... then I'll hold off my pull request a bit :-)
<norris>
mmind00: oh, didn't know it was holding you up. sorry
<mmind00>
norris: no no
<cristian_c>
another question
<mmind00>
norris: not holding up ... as we have a rc8 this time around, I just planned to do a small pull req with some accumulated fixes
<mmind00>
norris: and when you plan to send this today, we can probably just include that board too
<cristian_c>
I've tried block2mtd
<norris>
mmind00: ok, I guess that makes sense
<cristian_c>
when I load block2mtd with a loop block device, /dev/mtdblock0 is created
<norris>
mmind00: btw, thanks for the rc8 tip; didn't read my LKML yet :)
<rperier>
don't compute something, why almost everything is included in RK3288 TRM (all IP) and why this is not the case for previous SoC... an great example is the PMU :/
ganbold has quit [Ping timeout: 255 seconds]
<cristian_c>
but if I try to mount mtdblock with jffs2 fs, I get an error message
<mmind00>
norris: :-D ... always glad to be of service
<cristian_c>
if I read last messages in dmesg, I read: jffs2: Too few erase blocks (0)
<rperier>
"For detailed information about PMU, please refer to RK30xx PMU.pdf" yeeeppeee ^^
<mmind00>
rperier: in the really old kernels, even the clock_data was binary only (without source), so I guess they considered it some sort of secret in the past
<cristian_c>
how can I fix this issue?
<cristian_c>
Any ideas?
<mmind00>
rperier: if you need something looked up there, just ask ... but I'm of course not sharing the document itself
<norris>
cristian_c: why are you asking that here? that looks like linux-mtd
<cristian_c>
I've set the proper erase size value, in block2mtd loading command
<rperier>
mmind00: yeah, I know, don't worry. Perhaps that you could extract some info about PMU for me. I was looking what is need to add support for rk3066-pmu (and if it does make sense on this soc)
<rperier>
that's not a real question I know :D
<mmind00>
rperier: :-P ... the pmu registers on the rk3306 contain a lot less stuff ... mainly the suspend stuff and the powerdomain settings
<rperier>
the problem is that I don't know what I am looking for because I don't have these informations. I will look at how is implemented and how it works on RK3288 , and I will back to you with detailed questions
<mmind00>
rperier: ok ... at the moment I'm also not sure if the rk3066/rk3188 can at least reach some shallow suspend mode without additional sram code
<mmind00>
rperier: for example there does not seem to exist a mechanism similar to the FAST_BOOT_ADDR register on the rk3288
<mmind00>
rperier: so I'm not even sure where it would resume from
CTXz-dev has joined #linux-rockchip
<CTXz-dev>
Hi
pizthewiz has joined #linux-rockchip
CTXz-dev has left #linux-rockchip ["Leaving"]
akaizen_ has quit [Remote host closed the connection]