13:25 <aholler_> I've give 3.14 (linux-linaro-lsk-v3.14-mx6-drm) a try, and galcore failed: http://dpaste.com/1QNJ9DA

13:42 <jnettlet> aholler, setup CONFIG_DMA_CMA=y in your .config and make sure the allocation is over 128M We generally set it to 256

13:44 <aholler> thanks, will try it later

14:17 <aholler> I did a make oldconfig to not loose my configuration, therefor that wasn't set.

14:20 <aholler> but besides that it booted

14:22 <aholler> using my old dtb for the wandboard and having added the rfkill-driver

14:27 <aholler> cec as module fails (did that because Im playing at that front):

14:27 <aholler> drivers/built-in.o: In function `hdmi_set_dvi_mode':

14:27 <aholler> :(.text+0x8ff70): undefined reference to `hdmi_cec_start_device'

14:27 <aholler> :(.text+0x8ff7c): undefined reference to `hdmi_cec_stop_device'

14:27 <aholler> make: *** [vmlinux] Fehler 1

14:29 <aholler> likely a missing export

14:30 <aholler> ah, no, hdmi build and cec as module doesn't work

14:30 <aholler> +in

14:40 <aholler> [ 80.971373] imx-hdmi-audio imx-hdmi-audio: failed to config slave dma

14:40 <aholler> [ 80.971390] imx-hdmi-audio imx-hdmi-audio: ASoC: imx-hdmi-audio hw params failed: -22

14:41 <aholler> but video (xbmc-imx6 fb) plays

15:10 <aholler> hmm, dma_cma was turned on in my old config too. don't know why it disappeared

15:22 <dv_> aholler: what defconfig are you using ?

15:22 <dv_> imx_v7_cbi_hb_defconfig ?

15:24 <aholler> none, oldconfig. can't remember on what is was based ;)

15:24 <aholler> s/is/it/

15:25 <aholler> anyway, I now first play with cec before giving 3.14 another try. cec doesn't seem to work with 3.10 and 3.14 (using a patched libcec and modified kernel driver) so there seems to be some work there necessary

15:27 <aholler> libcec from imx-lib is able to turn the tv off, so it isn't a problem of the hw. Just buggy driver :/

15:27 <aholler> should be able to find that ;)

15:30 <dv_> try the config I mentioned

15:30 <dv_> this is the one I use

15:30 <dv_> works perfectly

15:31 <dv_> oh wait

15:31 <dv_> thats the one for the cubox-i , and you are using a wandboard. nevermind :)

15:31 <dv_> btw, I got a mail from somebody who wants to add wandboard support to the 3.14 kernel. jnettlet, heard anything about that?

15:33 <aholler> several changes for the dt are necessary to use a wandboard.

15:34 <dv_> sure

15:35 <dv_> I sent a mail to check for updates. I suppose you'd like to work with that person on it.

15:42 <aholler> I already thought about sending patches myself. but on the other side I'm unsure if I really want to support wandboard :/

15:42 <aholler> they don't seem to care for anything

15:47 <vpeter> aholler: I think here https://github.com/linux4kix/linux-linaro-stable-mx6 wandboard is supported.

15:48 <vpeter> at least dts's are there.

15:48 <aholler> yes, but that's the old stuff which doesn't work the c1

15:49 <vpeter> then nevermind

15:49 <aholler> you need to split the dt as done in mainline. and rfkill-wand is necessary to use wifi and bt

15:51 <aholler> e.g. the old wifi-power-pin is now used for cec

15:51 <aholler> (with c1)

15:52 <vpeter> I think this is too much for me. I only add led and gpio-ir for my board (tbs matrix btw).

15:53 <aholler> it isn't that much. it's just cumbersome ;)

15:55 <aholler> but thanks to dt, all changes have just be done in the dt. it was much more difficult when hw-specific stuff was embedded deeply in several vendor-drivers

15:58 <jnettlet> dv_, haven't heard a peep

15:58 <dv_> aholler: lets see if I get a response

16:01 <aholler> maybe just cherry-picking the patch from mainline works as a base. haven't tried it

16:02 <dv_> but I find it strange that the wandboard support is lacking. whenever I see imx6 mentioned, I often see their name close to it

16:04 <aholler> the board configuration fitted my needs best. and it is cheap

16:04 <aholler> but edid is broken :(

16:04 <dv_> edid is often broken

16:04 <dv_> makes you wonder how useful it really is

16:05 <aholler> yes, the beagle learned it very late too

16:05 <jnettlet> aholler, much of my device-tree is backported for 3.16.y

16:05 <dv_> but overall I'd definitely base my work on the 3.14 kernel

16:06 <dv_> considering the many improvements over 3.10.17 it has

16:09 <jnettlet> definitely a lot more stable.

16:11 <jnettlet> The in kernel NEON and aes integration makes ssh so much better. I can actually saturate 100Mbps with scp

16:11 <dv_> yay

16:12 <dv_> hopefully more people learn of this kernel and consider switching to it for their boards

16:14 <jnettlet> Like FSL :-)

17:13 <vpeter> Is it possible to get dtb file name used from userspace inside running linux?

17:20 <jnettlet> vpeter, what are you trying to do?

17:23 <vpeter> I have base linux system which covers multiple imx6 platforms. The only change is which dtb is loaded from u-boot. That's why I would like to know the file or some other identification when system is running.

17:25 <jnettlet> vpeter, if this is within the kernel there are helper functions cpu_is_imx6q

17:25 <jnettlet> etc

17:26 <jnettlet> if it is from userspace you can check sysfs

17:27 <jnettlet> /sys/devices/soc0/soc_id

17:30 <aholler> cat /proc/device-tree/compatible

17:31 <vpeter> aholler: Don't have this one. Maybe some kernel config missing?

17:31 <vpeter> jnettlet: I know this one. But this gives same data.

17:32 <aholler> vpeter: look at drivers->device.tree

17:32 <aholler> there is an option for dt in proc

17:33 <vpeter> will do. Thank you.

17:46 <vpeter> Anyone having hummingboard? What returns cat /sys/devices/soc0/machine

21:44 <jqz> Hello, anyone have E9 minipc from here:http://en.embedsky.com/product_info.php?cateid=169&id=169 and playing with kernel 3.14? Is there a patches I could find that has mxc video drivers in it?