ChanServ changed the topic of #linux-rockchip to: Rockchip development discussion | Wiki at http://linux-rockchip.info | Logs at http://irclog.whitequark.org/linux-rockchip | ML at http://groups.google.com/group/linux-rockchip
nighty has joined #linux-rockchip
cnxsoft has joined #linux-rockchip
zhecka_ has quit [Read error: Connection reset by peer]
zhecka_ has joined #linux-rockchip
<cnxsoft> Does anybody know about the status of RK3036 Kylin board? No development activity for about a year. It must have been canceled, and maybe the processor too?
cnxsoft has quit [Remote host closed the connection]
cnxsoft has joined #linux-rockchip
stdint has quit [Ping timeout: 256 seconds]
stdint has joined #linux-rockchip
ganbold has quit [Quit: This computer has gone to sleep]
k_ has quit [Quit: Page closed]
cnxsoft has quit [Read error: Connection reset by peer]
cnxsoft has joined #linux-rockchip
wens has joined #linux-rockchip
wens has quit [Ping timeout: 264 seconds]
stdint has quit [Ping timeout: 240 seconds]
wzyy2 has joined #linux-rockchip
stdint has joined #linux-rockchip
ganbold has joined #linux-rockchip
wens has joined #linux-rockchip
ganbold has quit [Quit: This computer has gone to sleep]
ganbold has joined #linux-rockchip
wens has quit [Ping timeout: 240 seconds]
wens has joined #linux-rockchip
ganbold has quit [Quit: This computer has gone to sleep]
ganbold has joined #linux-rockchip
wens has quit [Ping timeout: 256 seconds]
ganbold has quit [Quit: This computer has gone to sleep]
ganbold has joined #linux-rockchip
ganbold has quit [Quit: This computer has gone to sleep]
Myy has joined #linux-rockchip
ganbold has joined #linux-rockchip
Omegamoon 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
ganbold has quit [Client Quit]
ganbold has joined #linux-rockchip
wens has joined #linux-rockchip
ganbold has quit [Quit: This computer has gone to sleep]
wens has quit [Remote host closed the connection]
wens has joined #linux-rockchip
maz has quit [Ping timeout: 245 seconds]
wens has quit [Ping timeout: 256 seconds]
wens has joined #linux-rockchip
matthias_bgg has joined #linux-rockchip
wens has quit [Ping timeout: 245 seconds]
mrueg has quit [Ping timeout: 255 seconds]
maz has joined #linux-rockchip
wens has joined #linux-rockchip
nighty has quit [Quit: Disappears in a puff of smoke]
mrueg has joined #linux-rockchip
<Myy> Okay, it seems that the Mali driver ignores half of the provided options when compiling it Out-of-Tree
<Myy> I'm starting to think that CONFIG_XXX is ignored for any OOT module... It just affects the Makefile but the macros definitions are not available...
<phh> That's likely indeed
<phh> Since it needs to generate the config.h
<phh> Now that you mention that I'm suprised it works for me
<Myy> Isn't that because you already have a config.h in the mali folder, for example ?
nighty has joined #linux-rockchip
mrjay has joined #linux-rockchip
mrjay has quit [Ping timeout: 260 seconds]
libv_ has joined #linux-rockchip
libv has quit [Ping timeout: 264 seconds]
cnxsoft has quit [Quit: cnxsoft]
JohnDoe_71Rus has joined #linux-rockchip
mijk has joined #linux-rockchip
<mijk> I tried to compile a kernel for my rk3126 tablet but after flashing it it just does a boot loop. I'm not sure if I messed it up or if I need to flash the NAND
levd has quit [Ping timeout: 240 seconds]
libv_ is now known as libv
mrjay has joined #linux-rockchip
nighty has quit [Remote host closed the connection]
nighty has joined #linux-rockchip
<Myy> Without serial interfaces, it's really a guess work.
<mijk> I have a kernel.dts
paulk-collins has joined #linux-rockchip
<Myy> Well, that's a good start. However, in order to understand what's going on during the boot loop, you'll need to have some kind of feedback.
<Myy> Since the problem can be anything, from the kernel not having the filesystems functionnalities built-in and requiring an initrd, to some boot device change, to some specific drivers not compiled in this kernel, ...
<Myy> If you can have a UART interface, you might be able to view U-Boot and kernel messages.
<mijk> well, there's the microusb cable
<Myy> The OTG cable... I don't know if you can configure the kernel to redirect the log output to USB.
<phh> (or you can remux usb into serial)
<Myy> Is there an easy way to do this ?
<Myy> Does it requires soldering ?( ̄~ ̄;)
<mrjay> Myy: it requires usb serial adapter, microusb cable some skills to put this all together
<Myy> Interesting. However, can you use such techniques to capture the serial output of a tablet, without additional software configuration ?
<mrjay> Myy: no, software configuration is required
<phh> the patch isn't in mainline, though it has been posted
<Myy> Is there some tutorial / FAQ about how to configure a kernel to do so. I mean, worst case, would it be possible to encapsulate kernel messages in USB packets, send them to the OTG and record these USB packets with wireshark on the other side (PC) ?
<phh> well that would be using usb as usb slave/gadget
<phh> that's not what we're refering to
<Myy> Alright
<phh> but yes that's fairly easy to do, you might need to change your dts to change your usb port to slave instead of host, and then in make xxxconfig, enable gadget, using serial per default, then console=ttyGS0
<Myy> Ooooh. I'll need to test this !
<mrjay> Myy: what tablet are you using and what kernel are you compiling(vendor/mainline)?
<Myy> Well, actually, it's mijk which is having some problem with kernel flashing. I also had a RK tablet on which I tried to flash kernels but I wasn't able to do anything on it. That's why I'm interested by these tips.
<Myy> "<mijk> I tried to compile a kernel for my rk3126 tablet but after flashing it it just does a boot loop. I'm not sure if I messed it up or if I need to flash the NAND"
<mrjay> Myy: ok
<Myy> I guess that if he can recompile the kernel and use the gadget trick, he might have more informations about what's happening during the kernel boot process.
<phh> I doubt it, it is likely he has problems before even that
<Myy> So, what's the common advice to get U-boot messages ?
* phh has never played with a bootloader yet
marty__ has joined #linux-rockchip
<Myy> Ow (´・ω・` )
<Myy> I guess that in such cases, if you really determined to get serial output, you're good for some JTAG pins soldering ?
<marty__> Hey, is someone around here who knows how to build android for the rk3288 ? I've synced the source but I am kind of lost, since I don't see any kernel config for the device nor any instructions how to create the proper update.img for the bootloder :/
<mrjay> Myy: i think serial console is the only option
<phh> narmstrong: you created system.img/boot.img/.... ?
<phh> Myy: all rockchip systems I know have uart pins
<phh> (even highly integrated like gpd xd)
<Myy> phh, mrjay, I see. So if you have a USB<->UART dongle, it should be fairly simple to get console output on any RKXXX system ?
<phh> Myy: yup
<Myy> Nice !
<marty__> I was able to build the boot/misc/recovery/system.img but it seems like for the mkupdate script I need the kernel aswell ?
<phh> marty__: use the "ota" build argument
<phh> this will use boot.img for the kernel
<marty__> phh: so instead of "lunch rk3288_box-userdebug; make" I do "make ota" ?
<phh> no
<phh> you do the standard make
<phh> it is after
<marty__> Oh I see
<phh> I think it is mkimage.sh script
<marty__> phh: But Where am I supposed to find the update.img after running mkscript ?
<phh> can you link me to which repo you're using?
<marty__> phh: I'm trying to compile android 5, I've used repo on the manifest in : http://git.us.linux-rockchip.org/cgit/manifest/
<phh> (it's a bit sad they haven't release android 6)
<marty__> Yeah :(
<phh> (I mean they have it internally, or even publicly for OEMs)
<marty__> As of right now I would be really happy to even get android 5 working^^
<phh> right so you do mkimage.sh ota, this will create various stuff in rockdev/image-rk3288_box
<phh> then you can di boot boot.img; di system system.img; etc
<phh> if you want to make a full-blown upgrade.img, I doubt this will work out of the box, and you'll have more luck with google
<phh> the tool to do that is rkimagemacker
<phh> maker
<marty__> oh okay
<marty__> But do you know how to create the ressource.img from the kernel ?
mrjay has quit [Quit: Page closed]
<marty__> 'create boot.img with kernel... error: could not load secondstage 'kernel/resource.img''
<phh> mmm yes
<marty__> or can I ignore that ?
<phh> no don't ignore it
<marty__> I was kind of confused towards building the kernel because it doesn't seem like there is an rk3288_defconfig for it
<phh> make rk3288_box.img
<phh> in kernel folder
<phh> it's rockchip_android_defconfig I think?
<phh> or rockchip_defconfig?
<phh> yes rockchip_defconfig
<marty__> do I have to do anything besides rockchip_defconfig ?
<marty__> Because it's only printing out like 4 lines ....
<phh> well I'd do something like make -j8 rockchip_defconfig zImage modules dtbs rk3288-box.img
<Myy> if you're cross-compiling, you might want to double check the ARCH and CROSS_COMPILE environment variables.
<marty__> I'm running on linux/ubunut, but it seems like it's compiling now, shouldn't take too long
<phh> Myy: iirc rockchip's changed kernel's makefile for that
<phh> Myy: make rockchip_defconfig wouldn't work on x86 ;)
<phh> I mean except for sofia sdk which never existed
<Myy> Ah, nice !
<marty__> but it works on x64 ?
<phh> marty__: yes don't worry
<Myy> I often forget to set it when cross-compiling mainline kernels, which leads to : "This is a nice config you got there ! How about I nuke all these ARM options things. Now, tell me, what CPU are you using ?"
<phh> yup :P
<Myy> I almost forgot : 4.10.0-rc6 seems to have integrated some of the MMC patches
<Myy> I have not tested it yet but you might be able to boot it this time.
<phh> ah nice
<marty__> Compiling takes longer than before, guess thats a good sign ^^
<Myy> Should be !
<marty__> it failes on "modules" ERROR: "mali_platform_device_register" [drivers/gpu/arm/mali400/mali/mali.ko] undefined!
<Myy> If that's a RK3288, you could try disabling the Mali400 drivers as they won't be needed. Only the Midgard ones are on such platforms. (Mali-T760)
<phh> I think you can just ignore that message then
<marty__> Yeah it is an RK3288 with an Mali-t760
<phh> l i
<marty__> Yeah but it killed the whole process ^^
<phh> mmm really?
<phh> ok well then disable it in make menuconfig
<phh> but in my memory missing symbols in module was merely a warning
<marty__> Where is it in the menuconfig ? :o
<marty__> nvm
<marty__> I found it
<marty__> How do I apply the custom .config file it created ?
<phh> there is nothing to do
<phh> when you do make it will just use .config
<marty__> Now it fails on mv: Call of stat for 'include/generated/utsrelease.h.tmp' not possible : File or Directory not found
<Myy> Is it the only error message ?
<Myy> Isn't there something like "No space left on device" or something ?
<marty__> # # configuration written to .config # scripts/kconfig/conf --silentoldconfig Kconfig scripts/kconfig/conf --silentoldconfig Kconfig scripts/kconfig/conf --silentoldconfig Kconfig *** Error during update of the configuration. /home/ubuntu/Android/jenkins/workspace/Vidroid-rk3288/kernel/scripts/kconfig/Makefile:36: die Regel für Ziel „silentoldconfig“ scheiterte make[3]: *** [silentoldconfig] Fehler 1 /home/ubuntu/Andro
<marty__> Whoops there is some german in there
<marty__> It starts off with : *** Error during update of the configuration.
<marty__> and then make[3]: *** [silentoldconfig] Error 1
<Myy> Yeah, seems like something went wrong but what exactly... try running the same make command but append V=1 after make. Like make V=1 -j8 ...
<Myy> Then paste the result on ghostbin.com or some other place.
wzyy2 has quit [Ping timeout: 252 seconds]
<marty__> I've cut out the log after the first error because it would be too long
<Myy> I'll take a look
<Myy> Seems like there's something wrong with the dts file
<Myy> Line 498
<phh> ok my bad
<phh> marty__: delete the "dtbs" in the build line
<marty__> didn't worked , I had to delete the rk31*.dts files for it in order to compile
matthias_bgg has quit [Ping timeout: 256 seconds]
<marty__> strange, well as of now, I've disabled the Mali 400/300/200 GPU driver and removed the dts files of the rk31* , it went through without error, I guess I just try it out now <.<
<Myy> Without a DTB file, however, this will be extremely difficult. DTB files are needed by the different embedded device drivers.
<Myy> Give it a try, though.
<marty__> doesn't it work anyways cuz I'm working on a rk32*
<marty__> It complained about the rk31* dts files :/
<Myy> Oh !
<Myy> Fine then
<marty__> Myy: I have all the img files now, do I have to flash them via upgrade tool ? or fastboot ?
<Myy> Good question. phh or others android tinkerers might be able to help you out on this question. I never tried it (don't have enough disk space).
<phh> marty__: well you can use fastboot, "standard way" would be upgrade_tool
<phh> di boot boot.img
<phh> di system system.img
<phh> rd
<marty__> di flashes the image ?
<phh> di stands for download image
<phh> first argument is partition name
<phh> second is file
<marty__> yeah but shouldnt it be upload ?
<marty__> Or is it "download to the device" mmm :)
<phh> usually for embedded devices we indeed call it download
<marty__> Makes sense then
<marty__> but upgrade_tool says "parameter is invalid please check"
<marty__> I've run upgrade_tool , selected the device
<marty__> and then : di boot boot.img
<marty__> boot.img is in the same folder as the upgrade_tool
<Myy> What's the size of the boot.img ?
<marty__> 8.2 MB
<Myy> Is the device in download mode ?
<marty__> argh
<marty__> nevermind
<marty__> sudo is the keyword >.<
<Myy> Ah... good old sudo make me a sandwich
<marty__> yeah ....
<marty__> Do I have to flash the pcba_* images as well ?
<Myy> Hmm, tutorials does not seem to flash those files
<Myy> I don't know if they're a good reference though
<marty__> It didn't worked x)
<marty__> I guess I try my luck tomorrow again, but thank you all for you help :)
<Myy> Alright then :)
akaizen has quit [Read error: Connection reset by peer]
akaizen has joined #linux-rockchip
LongChair has quit [Quit: LongChair]
LongWork is now known as LongChair
LongChair1 has joined #linux-rockchip
JohnDoe_71Rus has quit [Quit: KVIrc 4.9.2 Aria http://www.kvirc.net/]
Omegamoon has left #linux-rockchip [#linux-rockchip]
<mijk> it looks like I'm the only person on the planet with a rk3126 looking to install Linux on it
<mijk> the others threw them out in the trash
marty__ has quit [Ping timeout: 260 seconds]
paulk-collins has quit [Quit: Leaving]
<Myy> (or sold them)
<Myy> The new kernel seems to have solved the MMC problem but not the MMC spam problem, which is still annoying. But at least it works.
<mijk> I'm just hoping to do something with it
<mijk> my mother in law bought it for the kids and it turned out to be a lemon, huge waste of money, couldn't return it
<mijk> I'm hoping to get some use out of it at least
<Myy> I highly suggest getting an UART<->USB cable if you don't have any
<Myy> This will be extremely useful to get debugging messages from the machine when things go wrong
<mijk> I have one but it's full sized usb
<Myy> Well, as long as you can plug the UART cables on the tablet UART pins, it's okay.
<Myy> You'll have to split it open though.
<Myy> 4.10.0-rc6 kernel available for RK3288 MiQi devices https://github.com/Miouyouyou/MyyQi
Myy has quit [Quit: Leaving]