ssvb has quit [Ping timeout: 264 seconds]
ssvb has joined #linux-exynos
dlan has quit [Ping timeout: 255 seconds]
dlan has joined #linux-exynos
dlan has joined #linux-exynos
dlan has quit [Ping timeout: 256 seconds]
dlan has joined #linux-exynos
dlan has joined #linux-exynos
dlezcano has quit [Ping timeout: 265 seconds]
dlezcano has joined #linux-exynos
SteveCapper has quit [Changing host]
SteveCapper has joined #linux-exynos
vinipetrucci has joined #linux-exynos
vinipetrucci is now known as joaocashs
joaocashs has left #linux-exynos [#linux-exynos]
joaocashs has joined #linux-exynos
joaocashs has left #linux-exynos [#linux-exynos]
vinipetrucci has joined #linux-exynos
afaerber has quit [Quit: Verlassend]
ssvb has quit [Ping timeout: 276 seconds]
afaerber has joined #linux-exynos
ssvb has joined #linux-exynos
<vinipetrucci> hello guys, still not able to boot mainline kernel on my Chromebook 2 (Peach Pi) XE503C32.
<vinipetrucci> after hitting "ctrl-u" to boot from SD card at the dev screen, i get a black screen and cannot access the u-boot prompt.
<vinipetrucci> i've compiled/used the kernel from: git://git.kernel.org/pub/scm/linux/kernel/git/khilman/linux.git
<vinipetrucci> any thoughts?
<sjoerd> check if you've set the boot priorities on the sd card partitions right and that you've put u-boot in the right one
<javier__> vinipetrucci: which branch from kevin's tree did you build? (I assume is wip/exynos/v3.19-integ-orig)
<vinipetrucci> think i used master
<vinipetrucci> yea, master
<vinipetrucci> but before booting the kernel, i would need to get to the prompt right
<javier__> vinipetrucci: not necessarily, kevin's master is 3.19-rc4 and the patch that enables the needed kconfig options for display on pi only made it to 3.19-rc5
<vinipetrucci> i wrote nv_uboot-peach-pi.kpart into /dev/mmcblk1p4
<javier__> vinipetrucci: so at least you need that patch, there was a fix for the exynos dp phy that also made it to 3.19-rc5 iirc
<vinipetrucci> then: sudo cgpt add -i 4 -P 15 -S 1 -T 1 /dev/mmcblk1
<vinipetrucci> and: sudo cgpt add -i 2 -P 0 -S 0 -T 0 /dev/mmcblk1
<vinipetrucci> javier__: cool, thanks! i will rebuild with that branch now then
<javier__> vinipetrucci: ok
<vinipetrucci> i have enabled usb boot and non signed boot from cromiumos
<vinipetrucci> Wizzup: great, thx
<Wizzup> although it's not very detailed - just about the actual kernel building
<vinipetrucci> Wizzup: i think i got the kernel compilation right, just need to use an updated one, as javier__ suggested
<Wizzup> ack
<vinipetrucci> i think there is something with the nv_uboo thing
<sjoerd> yeah nv-uboot should give you the uboot output
<vinipetrucci> right
<Wizzup> I always have to press a button to actually get the u-boot prompt though
<vinipetrucci> any button?
<sjoerd> Wizzup: the prompt sure, but it does give output i assume ? (mainline uboot does for sure)
<javier__> if vinipetrucci didn't set a delay and nv-uboot is able to load a (wrong) kernel then he should not see u-boot prompt
<javier__> sjoerd: nv-uboot does not in that condition iirc
<Wizzup> sjoerd: yeah, it does, although fast
<sjoerd> oh
<Wizzup> and I should really switch to mainline u-boot :/
<Wizzup> It's just that I recall compiling u-boot as a huge pain with chromeos
<Wizzup> maybe it's better now...
<Wizzup> would be nice to have some guide on it :)
<Wizzup> http://linux-exynos.org/wiki/ARM_Chromebook/nv_u-boot ouch, this also still states simplefb is mandatory for mainline
<Wizzup> ack, I will try and document that soonish
<Wizzup> And provide binaries...
<vinipetrucci> javier__: does your linux-3.16-rc4-peach-pit branch already include the patches you mentioned? do you recommend trying yours first before testing kevin's "v3.19-integ-orig"?
<javier__> vinipetrucci: sorry, I have no idea what the linux-3.16-rc4-peach-pit branch is tbh
<vinipetrucci> James_T: git clone git://git.collabora.co.uk/git/user/javier/linux.git -b linux-3.16-rc4-peach-pit
<vinipetrucci> ops, javier__ git clone git://git.collabora.co.uk/git/user/javier/linux.git -b linux-3.16-rc4-peach-pit
<Wizzup> vinipetrucci: peach pit != peach-pi btw
<javier__> vinipetrucci: oh, that branch does not exist anymore
<Wizzup> argh, vinipetrucci, it seems like my instructions are outdated - hang on
<vinipetrucci> Wizzup: oh right, good catch
<vinipetrucci> javier__: i see
<Wizzup> You want one of the later wip/ branches
<javier__> vinipetrucci, Wizzup: I would prefer if the instructions mentioned a mainline kernel + additional patches needed to have all the features working instead one of my random branches :)
<Wizzup> I got those to boot on my chromebook
<Wizzup> javier__: yep, that's why they are under a user: page :)
<vinipetrucci> Wizzup: peach pi or pit ?
<Wizzup> pi and pit are different devices
<Wizzup> You could try this: http://cgit.collabora.com/git/user/javier/linux.git/log/?h=wip/exynos/dp-integ
<vinipetrucci> Wizzup: yea, but what was the one you have booted
<Wizzup> I got that to boot on snow and peach-pi
<vinipetrucci> Wizzup: i see
<Wizzup> iirc at least ;)
<Wizzup> definitely on snow at least, but I think it should also boot on pi
<javier__> that branch should give you display working on snow, pit and pi for sure
<vinipetrucci> cool
<Wizzup> vinipetrucci: if you like I can make you a wiki account so you can help out :)
<vinipetrucci> Wizzup: sure, how can i best help? describing the steps once i have it working? ;-)
<Wizzup> that, and perhaps, finding/fixing/marking outdated pages, I have some trouble getting free time, lots of work
<vinipetrucci> i want to get all 8 cores working to perform some research on het-aware scheduling... i was considering getting one of the ODROID-XU3 since it says HMP already works on it
<Wizzup> HMP also works on this with the right options
<Wizzup> this = peach-pi
<vinipetrucci> Wizzup: lots of work here too, but i can help out on my spare time with docs etc
<vinipetrucci> cool
<javier__> vinipetrucci: to get all cores you just need to disable CONFIG_BL_SWITCHER
<javier__> but keep in mind that from the scheduler/software pov is really a SMP system with 8 cores even though at the hw level are not the same
<vinipetrucci> javier__: sure
<vinipetrucci> javier__: but to disable i need to modify/recompile the kernel and boot it... would have similar issues with nv_boot thing right?
<javier__> vinipetrucci: yes, you need to modify recompile, there is a runtime way to change it by writing to a sysfs file but when I tried the system hang
<javier__> vinipetrucci: sorry, I didn't get what nv_uboot issue you are referring too
<vinipetrucci> javier__: do you recall what command did you use to change it at runtime?
<javier__> vinipetrucci: echo 0 > /sys/kernel/bL_switcher/active
<vinipetrucci> javier__: i meant the black screen at dev boot screen, but as you suggested that could be fixed using the right kernel with updated patches. i will give it a try
<javier__> vinipetrucci: yes, that's because you didn't enable the display options
<vinipetrucci> javier__: i can successfully run the echo command you gave me, the system didnt freeze, but when i read the "active" file again, it still has "1" as value... and /proc/cpuinfo still shows 4 cores...
<vinipetrucci> javier__: ok, that makes sense
<javier__> vinipetrucci: I haven't tested it in a while but yeah seems to be working now
<javier__> # echo 1 > /sys/kernel/bL_switcher/active
<javier__> # grep -c processor /proc/cpuinfo
<javier__> 4
<javier__> # echo 0 > /sys/kernel/bL_switcher/active
<javier__> # grep -c processor /proc/cpuinfo
<javier__> 8
<vinipetrucci> javier__: i see, cool. prob because i tried on older kernel version 3.8.11 that comes with crubuntu
<javier__> I'm running a 3.19-rc4 right now
* javier__ => lunch, bbl
vinipetrucci has quit [Ping timeout: 245 seconds]
<libv> nice, just plugged in a hdmi to vga convertor into the chromebook and it just worked
<libv> now for 1024x768
<si1v3r> xrandr?
<si1v3r> xrandr --output VGA-1 --mode 1024x768 :?
<Wizzup> si1v3r: no xrandr on fbturbo/fb
<Wizzup> and armsoc didn't work for me at least
<Wizzup> nor did xf86-video-modesetting
<si1v3r> Oh, lame.
afaerber has quit [Quit: Verlassend]
roadki11 has joined #linux-exynos
roadki11 has quit [Client Quit]
<libv> it's actually quite worse than that
<libv> it's almost as if i never existed and people hadn't learned that structured display driver development was even possible
afaerber has joined #linux-exynos
amitk has quit [Quit: leaving]
<libv> but it seems javier__ also is unable to make mainline work with snow
<libv> or at least, work properly
<Wizzup> I have a mainline snow
<Wizzup> I'll be back in a bit and try to document it
<si1v3r> Wizzup: I think I did what you did
<si1v3r> I wrote it down too ^
<daniels> for snow, wip/exynos/drm-bridge-3.19-rc5 is much more up to date than dp-integ
<daniels> and has a lot fewer patches on top of mainline
<libv> daniels: do you have any idea to what extent hdmi works on that?
* libv is really considering hacking the 3.8 kernel drm driver directly as that currently seems the least painful and most promising to get 1024x768
<Wizzup> daniels: ah, I guess I should try that later then
<daniels> libv: see
<daniels> libv: see javier__ on linux-samsung-soc; works fine for me on pi with a very small patch
<libv> hrm, apparently all that is pretty dependent on modetables, and there is no 75MHz dotclock, or 1024x768 modetable for exynos5, only for 5420 it seems :(
<libv> that's going to be an insane amount of work for 11d, in which i actually have to write the demo still
* libv is going to have to hope that 1280x720 is supported by those 4y old projectors in the new FOSDEM building
<Wizzup> maybe a stupid suggestion, but perhaps you could use some hw to downscale it
<Wizzup> Also, I think I definitely got armsoc to work on 3.4 or 3.8 kernels
<Wizzup> It was just slow, but it worked. And I could do some modesetting with it (not too reliably)
<Wizzup> Just didn't get it to work on mainline
<libv> Wizzup: good point
<libv> that actually exists for a reasonable amount of money
<Wizzup> Alt. perhaps you can ask the fosdem people to point a cam. on the screen... ;-)
<si1v3r> A university would have a projector that will scan documents, usually.
<Wizzup> But yeah, it's annoying that this doesn't work nicely yet
<libv> and for the same price for which i bought my original converter box in feb 2012 after the lima driver was shown on a 7" screen to about 120 people :)
<si1v3r> Somekind of camera projector rube goldberg overhead projector.
<libv> which has no scaling
<Wizzup> I also used a USB 2.0 external display (chinese Displaylink copy) and that worked ok, but somehow I figure that's not ok for your usecase
<libv> for demoing the current state of affairs of the tamil driver, that's indeed not the best option :)
<libv> ordered
<libv> this thing has some buttons so the resolution can be chosen.
<libv> smashing idea Wizzup
<libv> plus, it seems a lot smaller than my current hdmi->vga device
<Wizzup> What device did you order exactly? (Now you've piqued my interest)
<libv> i was certain i'd only hit those cheap in-cable things or the one i already have or something similar
<libv> also, this would've really saved kusma's ass last fosdem :)
<Wizzup> I presume it also downscales? (and not just upscales?)
<libv> Wizzup: wait and see :)
<Wizzup> alright
<libv> if not, there's a good chance that those projectors in the K building do 720p
<libv> they were brand new in 2012, so they should
<Wizzup> ok
<Wizzup> The VGA to HDMI Converter Ligawo 6526624 allows you to easily connect a VGA analog device such as a netbook or laptop with VGA output to a projector or TV set with an HDMI connection. The Ligawo VGA to HDMI converter not working 1:1, he always scales the incoming signal via VGA to 720p or 1080p HDMI. The switch from 720p to 1080p is made directly to the converter.
<Wizzup> hm, perhaps this is not the same one
<Wizzup> do you want hdmi out though?
<Wizzup> ah that is a different device.
<libv> the one i just ordered has two buttons allowing you to select the resolution after scaling
<Wizzup> yeah, I tried to find the same on amazon.co.uk
<Wizzup> my german is not too bad, but still ;)
vinipetrucci has joined #linux-exynos
vinipetrucci has quit [Ping timeout: 245 seconds]
vinipetrucci has joined #linux-exynos
vinipetrucci has quit [Ping timeout: 245 seconds]
vinipetrucci has joined #linux-exynos
vinipetrucci has quit [Ping timeout: 264 seconds]
vinipetrucci has joined #linux-exynos
vinipetrucci has quit [Ping timeout: 265 seconds]
vinipetrucci has joined #linux-exynos