10:58 <lautriv_> does a vanilla or patched ( non-android) kernel work reasonable on a rockchip (RK3188) ?

11:15 <mmind00> lautriv_: depends on if you can live without graphics

11:15 <mmind00> lautriv_: aka only the rk3288 has graphics support currently

11:19 <lautriv_> mmind00, if i could live without graphics, the whole idea would be pointless.

11:19 <cristian_c> hello

11:20 <cristian_c> I've got an issue with my screen: I see two desktops in my screen (1 full desktop + 1/3 desktop below the first one)

11:22 <cristian_c> Issue happens also in boot process. I've found that using: echo 1 > /sys/class/graphics/fb0/blank (1 can be replaced by 2, 3 or whatever number different by zero, the second desktop becomes black and entire display freezes and it slowly obscurates itself (as in suspend mode)

11:23 <cristian_c> if I type echo 0 > /sys/class/graphics/fb0/blank, it enables itself again

11:23 <cristian_c> same effect with fb1, btw

11:24 <cristian_c> f I try with fb2, I can obscurate and reactivate external hdmi screen

11:24 <cristian_c> tw, external hdmi screen is not affected by double desktop issue

11:24 <cristian_c> Any ideas?

12:19 <rperier> usually when you submit a patches serie to a maintainer you are supposed to base this serie onto the maintainer's tree or linux-next ? or both ? (personally I do everything with next...)

12:21 <mmind00> rperier: while next works most of the time, maintainers normally expect patches to be based on their tree

12:21 <mmind00> rperier: there may be cases where some other tree also touches similar areas that the maintainer's tree (like big scale cleanups)

12:22 <rperier> no way to apply the patches serie for analogix... :/

12:22 <cristian_c> another question

12:22 <cristian_c> I've got rk903 wifi chip

12:22 <cristian_c> I've created /system/etc/firmware directory

12:23 <rperier> mmind00: I see, perhaps that I should try with the drm tree directly

12:23 <cristian_c> I've put firmware files into it

12:24 <cristian_c> but, again, I can't see wifi launcher start in dmesg, so wifi is not initialized

12:24 <rperier> I am just trying to boot a "core-image-sato" image (minimal image with x11) on my veyron, for that, I would like to test edp first

12:24 <cristian_c> Any ideas?

12:27 <rperier> I mean technically I might rebase it by hand but... there are 14 patches :D

17:12 <lautriv_> certain links on the wiki vanished, what is the most useable repo for a rk3188 based thingie where i have to gather details ?

17:15 <cristian_c> maybe github repositories were updatrd

17:15 <cristian_c> updated

17:15 <cristian_c> or maybe they was canceled

17:16 <lautriv_> github gives a 404

17:17 <cristian_c> lautriv_: what device do you own?

17:22 <lautriv_> cristian_c, that is actually not 100% defined, it's a TV-Box containing RK3188/rtl8101(eth)/RK1000(SPDIF)/RTL87xx(Wifi)/8G Flash/2G RAM/OTG/HDMI/2xUSB2.0/AV/µSD

17:23 <cristian_c> lautriv_: you are developing android official firmware?

17:24 <cristian_c> either are you a simple customer?

17:24 <cristian_c> :)

17:31 <lautriv_> cristian_c, the plan is to put a real OS on ;)

17:31 <cristian_c> lol

17:33 <cristian_c> lautriv_: you could use phjanderson github repository

17:34 <lautriv_> i found someone postulating this hardware is "CS918", not found in wiki.

17:42 <lautriv_> cristian_c, the only thing i found with that name is a hot-patch tool.

17:45 <lautriv_> ok, found something .... appears most people lost interrest 2 years ago, probably because all the propietary surrounders.

18:07 <cristian_c> lautriv_: ah, ok

18:08 <cristian_c> lautriv_: I think you should not encounter particular issues

18:09 <cristian_c> lautriv_: usually, you should only care particular wifi chip

18:11 <cristian_c> lautriv_: I can link an rk3188 prebuilt kernel, but you can build a new obe, if you want

18:59 <rperier> amstan: linux-veyron (chromeos kernel) works very fine if it is flashed but not if it is loaded via kexec

19:00 <rperier> via kexec the kernel hangs while initializing mali

19:01 <lautriv_> rperier, kexec is only for single-core, is this a single-core ?

19:13 <rperier> lautriv_: no kexec works for smp

19:14 <rperier> (I use kexec on all my rockchip-based devices, all of these are smp)

19:14 <c0d3z3r0> rperier: except for rk3188 which doesn't work at the moment

19:15 <rperier> woooo! my veyron is booting "core-image-sato" with the chromeos kernel and an x11 server :D (generated via meta-rockchip).

19:15 <rperier> next step is mainline with edp

19:20 <lautriv_> rperier, it doesn't matter how much cores are around but it matters if more than CPU0 is running on kexec -e.

19:47 <lautriv_> hm, while dealing with other arms i had somehow the impression rockchip is better supported by FOSS, appears i was mistaken.

19:54 <mmind00> lautriv_: it actually is quite well supported ... although the old socs are currently limited to tricle-down effects

19:56 <mmind00> lautriv_: currently most development happens on the rk3288, due to the numerous chromebooks based on it being released

20:02 <lautriv_> mmind00, not even a defconfig for uboot, no way to boot from external e.g.SD-Card, everywhere propietary blobs ain't good support.

20:04 <mmind00> lautriv_: of course if you're looking at the vendor kernel, you won't get happy ... mainline rk3288 uboot is in the works: http://git.denx.de/?p=u-boot/u-boot-dm.git;a=shortlog;h=refs/heads/rockchip-working

20:12 <lautriv_> mmind00, seriously, even wondermedia chips (NDA) have better support, probably the best to show this thingie my sledgehammer and put rockchip on the useless list.