prahal has joined #linux-exynos
prahal has quit [Ping timeout: 240 seconds]
prahal has joined #linux-exynos
prahal has quit [Client Quit]
indy has quit [Ping timeout: 264 seconds]
indy has joined #linux-exynos
liquidAcid has joined #linux-exynos
afaerber_ is now known as afaerber
spring has joined #linux-exynos
spring is now known as Guest81043
Guest81043 is now known as springchromebook
<springchromebook> hola, looking for any tips getting a mainline kernel to drive the displayport interface on a hp chromebook 11
<springchromebook> kernel (3.19.7) boots but with a black screen
<springchromebook> if I use the .config from the chromebook, system locks up on boot, exynos_defconfig boots with no video
<springchromebook> I found some reference to a bug with the DRM IOMMU (sorry, no link handy), enabling that does indeed reproduce a lockup at boot with 3.19.7 and the exynos_defconfig, but disabling that option with the chromeos .config does not prevent it from locking up at boot
<Wizzup> hiya
<Wizzup> start with linux v4.1-rc2
<springchromebook> ok, I did try 4.0 also
<Wizzup> from v4.1-rc2 display worked out of the box for me on snow
<Wizzup> and the hw is quite similar I think.
<springchromebook> I think the bug I mentioned was for 4.0, which is why i tried 3.19.7
<springchromebook> Wizzup: is that with the simple fb
<Wizzup> no.
<Wizzup> I disabled simplefb
<springchromebook> nice, ok i will try that kernel
<Wizzup> actually, almost everythibf works in v4.1-rc2
<Wizzup> apart from sound
<springchromebook> bummer, but that doesnt bother me too much
<liquidAcid> Wizzup, what's the issue with sound?
<Wizzup> Do you get sound on snow?
<Wizzup> On peach-pi I get the 'audio too fast and random hickups
<Wizzup> '
<Wizzup> on snow I don't think I managed to make it output anything
<Wizzup> javier said it had something to do with the bootloader (or booting, I don't recall), but he also wasn't exactly sure what was up atm
<springchromebook> ah, just to be clear, I dont need any of the mali blobs, right
<Wizzup> no, you don't.
<Wizzup> mali is only 3d accel
<Wizzup> and I mean that: only
<springchromebook> thats what I thought, but both the Archlinux and chrome kernels that did work have the mali blobs
<liquidAcid> Wizzup, k, i thought it might be related to the dma issues that prahal pointed out recently: https://github.com/prahal/linux/commit/5e60cfc9fa5101a346e29ef5f944fbbad300c72d
<Wizzup> liquidAcid: well, that *may* explain it for peach actually
<Wizzup> Although I do not know the details - the audio was way too fast
<Wizzup> (guessing what changing it to 'burst' would do)
prahal has joined #linux-exynos
<Wizzup> I guess I can try later -- not today though
<springchromebook> Wizzup: ok, thanks for the tip. will give this a shot.
<Wizzup> let me know - and please update the wiki if it does work ;)
<springchromebook> should I use the exynos_defconfig or the chromeos one?
<Wizzup> ehm
<Wizzup> exynos_defconfig
<Wizzup> mainline has no chromeos config I presume?
<Wizzup> or does it?
<springchromebook> there is some options for the chrome ec driver iirc
<Wizzup> just exynos_defconfig
<Wizzup> I only add a few things (dm_crypt compiled in instead of module) and disable simplefb
<Wizzup> but I don't think you actually need to disable simplefb anymore
prahal has quit [Ping timeout: 240 seconds]
prahal has joined #linux-exynos
prahal has quit [Quit: prahal]
prahal has joined #linux-exynos
prahal has quit [Quit: prahal]
liquidAcid has quit [Quit: Leaving]
<springchromebook> Wizzup: I didnt disable simplefb, locks up on boot - will try disabling it
<springchromebook> nope, still locks up... :(, hmm
<springchromebook> I guess I
<springchromebook> (oops) will try the chromeos config
<springchromebook> interestingly 4.1-rc2 has a new driver for the displayport-lvds chip that is used on this device... sounds plausible that it might be useful
<springchromebook> bah, this locks up too
<springchromebook> ok, I am giving up on this and buying an x86 system i guess
<springchromebook> thanks for help, sorry it didnt work out
<Wizzup> you give up quite fast :)
<Wizzup> we've been at this for over a year
<Wizzup> there's other people with a spring machine too
<Wizzup> there are*
<Wizzup> have some patience :)
springchromebook has quit [Ping timeout: 246 seconds]
<Wizzup> guess not
afaerber_ has joined #linux-exynos
afaerber has quit [Ping timeout: 276 seconds]
* si1v3r checks ebay for a cheap spring
<si1v3r> I think there's a perception issue. "Linux was hacked to run on x" is not the same as "linux supports x".
<si1v3r> or even, vendor support is not the same as mainline support.
leowt has joined #linux-exynos
<leowt> today i started digging a little on the dts files from odroid, copied them and started removing things that nanopc-t1 doesnt have, or in other gpio
<leowt> javier__ i have found out that nanopc-t1 doesnt have the max regulator that odroid has
<leowt> i also found the schematic for the board, and there is no max regulator
<leowt> so, the componentes use the power references from 4412 directly?
<leowt> why is odroid using the max regulator?
<sjoerd> leowt: which of the various odroid boards
<leowt> it is on the odroid-common, so i suppose every one
<leowt> but i am using odroidx as comparsion
<leowt> sjoerd, yes, but nanopc-t1 dont
<leowt> thats what i am saying
<sjoerd> ah sorry i misunderstood you
<leowt> :)
<sjoerd> what pmic does the nano have?
<leowt> i am trying to find out just now
<leowt> i thought for instance that 4412 had an internal one
<leowt> i cant find one
<sjoerd> looks like it's a set of fixed regulators
<sjoerd> see page 11
* sjoerd has to go and make dinner, later :)
<leowt> sjoerd, exactly
<leowt> my priority is to enable hdmi, fb and mfc to play h264
<leowt> firstly i see that on that schematic hdmi is connected to NTS0104
<leowt> that has somekind of i2c controll
<leowt> being so
<leowt> by looking at exynos-common dtsi
<leowt> there is a ddc field on hdmi
<leowt> i suppose that is where i should enable that circuit
<leowt> is that so?
<leowt> btw, im i supposed to pass anything to de boot args so that hdmi can fire up?
<sjoerd> nope, shouldn't be needed
<leowt> sjoerd, i dont have fb, no hdmi messages on the dmesg
<sjoerd> that NST chip just seems to be a level shifter fwiw
<leowt> i saw i2c and thought that was somehow enabled by it
<leowt> but i see that it is only doing his job for hdmi scl and sda
<leowt> anyways
<leowt> am i supposed to declare anything of hdmi on device tree?
<leowt> theres hdmi_hpd that is already ok with the correct gpio
<sjoerd> same as the odroid have a i uess, just the supplies (although it might be fine with dummies) and the dcc lines
<sjoerd> note that for an x2 the dmesg output also seems rather limited wrt. hdmi
* sjoerd must say he hasn't tested that hdmi actually works on an x2 in ages
<leowt> dts is like this
<leowt> what should i put in ddc?
<leowt> btw, that hdmi_dcc, tbat e
<leowt> that means that on odroid board it is enabling some ic with i2c?
<leowt> i am now reading that ddc is somekind of hdmi i2c
<leowt> so, if it is in the wrong i2c address it makes hdmi to not work?
<sjoerd> yeah ddc is jut an i2c bus to talk to the monitor
<leowt> fixed it, but still now sign of hdmi
<sjoerd> did the driver load correctly and all that ?
<leowt> sjoerd what should be the kernel components to have hdmi fb?
<sjoerd> at least CONFIG_DRM_EXYNOS_HDMI and CONFIG_DRM_EXYNOS
<leowt> CONFIG_VIDEO_SAMSUNG_S5P_HDMI?
<sjoerd> that's the oldstyle one
<leowt> sjoerd is the one on 4.1-rc2
<leowt> no exynos_hdmi appears
<leowt> found it, sorry
<sjoerd> they should all be in exynos_defconfig
<leowt> what is the difference between exynos and s5p tv ones?
<leowt> sjoerd yey i have a hdmi signal
<sjoerd> leowt: the s5p tv one is a v4l2 output device
<sjoerd> while the exynos hdmi one is a DRM driver
<leowt> sjoerd, i have signal, how can i test properly for output? i have done "gst-launch-1.0 -v videotestsrc ! fbdevsink device="/dev/fb0""
<leowt> with no success
<leowt> it behaves like it is playing
<leowt> no error messages
<leowt> also tried "cat /dev/urandom > /dev/fb0" and fbi
<leowt> no success
<leowt> i suppose it has something to do with this: "[drm:hdmi_commit] *ERROR* failed to find hdmiphy conf"
<chrs_> fbi? where?
liquidAcid has joined #linux-exynos
liquidAcid has quit [Quit: Leaving]
leowt has quit [Quit: Textual IRC Client: www.textualapp.com]