<Tenkawa>
trying to see if disabling iommu will fix this video problem
<Tenkawa>
this darn thing is picky
amitk_ has joined #linux-exynos
ssk has joined #linux-exynos
LanDi has quit [Quit: fui !]
ssvb has joined #linux-exynos
ssk has quit [Quit: Leaving]
<javier__>
Tenkawa: what kernel version are you trying? 4.0-rc1?
<javier__>
Tenkawa: ok, I built v4.0-rc1 and booted on Snow, as expected I got http://paste.debian.net/plain/159370 due exynos drm not working with iommu enabled
<javier__>
sjoerd: you mean the patch to disable iommu? he picked on his v4.0-samsung-defconfig
<javier__>
I mentioned this morning that is -rc material since 4.0-rc1 boot is broken due that on most exynos5 boards
<sjoerd>
javier__: no the dts bits
<javier__>
*mentioned to kukjin
<javier__>
sjoerd: I don't think so yet I don't know if that is technically a fix since display were not working anyways before 4.0-rc1
<sjoerd>
right
<sjoerd>
samsung-fixes-2 had a bunch of dts patches but i didn't look at the specifics
<javier__>
sjoerd: yeah but display on Snow is not working even with today's linux-next...
<sjoerd>
fair enough
<sjoerd>
one day :)
<javier__>
haha
zombah has joined #linux-exynos
gautam__ has joined #linux-exynos
ard has joined #linux-exynos
opuk has quit [Changing host]
opuk has joined #linux-exynos
ssk has joined #linux-exynos
prahal has joined #linux-exynos
prahal has quit [Ping timeout: 252 seconds]
ssk has quit [Ping timeout: 245 seconds]
prahal has joined #linux-exynos
ssk has joined #linux-exynos
ssk has quit [Quit: Leaving]
afaerber has quit [Ping timeout: 264 seconds]
afaerber has joined #linux-exynos
si1v3r has quit [Ping timeout: 250 seconds]
Tenkawa_ has joined #linux-exynos
<Tenkawa_>
hi all
<Wizzup>
hi
<Tenkawa_>
well the good news... 4.0.0-rc1 boots up with almost everything functional on exynos5250-snow
<Tenkawa_>
problem.... the screen just turns white
<Tenkawa_>
changing drm settings and dts changes its behaviour however nothing is makig text come up
<Tenkawa_>
any thoughts?
<Tenkawa_>
theres so many drm settings/combinations
<Tenkawa_>
I thought simplefb should take care of it. is there something I'm missing to force fb0 to be console in my bootline?
Tenkawa_ has quit [Changing host]
Tenkawa_ has joined #linux-exynos
<afaerber>
Tenkawa, to use simplefb, you need an nv-uboot with simplefb support and you need to disable anything in your .dts that loads the exynos drm, including possibly dp-controller, hdmi, panel nodes
<Tenkawa_>
then I definitely had it running using dtb because my scripts still use mkimage and a dts
<Tenkawa_>
so I have to be missing something simple
<afaerber>
you sound confused. you need a .dtb either way
<Tenkawa_>
afaerber: yes however it is possible to "not" use nv-uboot
<Tenkawa_>
I dont want to add nv-uboot to the mix
<Tenkawa_>
It was running fine up until a certain point
<Tenkawa_>
and I lost my kernel
<afaerber>
Tenkawa_, if you don't, then you can't use simplefb. you'll need to get the real exynos drm working
<Tenkawa_>
afaerber: thats fine too... the conundrum is it activates and creates a white screen (everything else is functional)
<Tenkawa_>
so I need to determine which drm setup the dts needs and/or which drivers in the kernel drm stack to adjust
<afaerber>
welcome to the club :)
<Tenkawa_>
heheheheh
<Tenkawa_>
and drm and dts are both not my forte
<Tenkawa_>
at least the rest of the box works (kb, wifi, etc).. just cant see it
<Tenkawa_>
you know its going to be something monumentally simple in the end
<Tenkawa_>
heheheh
<afaerber>
probably ;)
<Tenkawa_>
it always is haahaaa
<Tenkawa_>
from what I can tell I think it is fimd related
<Tenkawa_>
compiling another test
<Tenkawa_>
got this part down to a science
Tenkawa_ has quit [Quit: reboot]
Tenkawa_ has joined #linux-exynos
Tenkawa_ has quit [Changing host]
Tenkawa_ has joined #linux-exynos
<Tenkawa_>
bah no dice on that try
<Tenkawa_>
still white background
<Tenkawa_>
going to see what syslog says
<Tenkawa_>
pwm-backlight probe keeps deferring
<Tenkawa_>
it appears to use FIMD in 3.8
<Tenkawa_>
fimd and displatpor
<Tenkawa_>
er displayport
gautam__ has quit [Quit: Leaving]
<Tenkawa_>
ok so fimd is definitely the display
<Tenkawa_>
think I may have found it... was burried
* Tenkawa_
hopes
<Tenkawa_>
FIMD was hard to get to select based on its dependencies
<Tenkawa_>
brb
Tenkawa_ has quit [Quit: lets try that again]
Tenkawa_ has joined #linux-exynos
Tenkawa_ has quit [Changing host]
Tenkawa_ has joined #linux-exynos
<Tenkawa_>
uggh
<Tenkawa_>
this thing just doesnt like me
<Tenkawa_>
just keeps requesting deferral
<javier__>
Tenkawa_: is a regression in the exynos drm driver, I don't think is something trivial
<Tenkawa_>
well crap
<Tenkawa_>
why was it working a few code changes ago?
<Tenkawa_>
I was really hoping to test 4.0-rc on here
<Tenkawa_>
hold that thought.. need to reboot
Tenkawa_ has quit [Quit: leaving]
Tenkawa_ has joined #linux-exynos
Tenkawa_ has quit [Changing host]
Tenkawa_ has joined #linux-exynos
<Tenkawa_>
guess I'll be waiting for a hopeful fix
<javier__>
Tenkawa: it never really worked with mainline, what you were probably using was a 3.19 + the posed patches to add display support for snow
<Tenkawa_>
OH
<Tenkawa_>
those patches avail for 4 yet?
<Tenkawa_>
(memory isnt what it use to be)
<Tenkawa_>
javier__: you're probably right
<Tenkawa_>
I seem to recall something like that
<Tenkawa_>
I'm a dork
afaerber_ has joined #linux-exynos
<Tenkawa_>
javier__: thank you for being patient with my misremembering
<Tenkawa_>
I'd even use 3.10 with those patches if they are still out there
<Tenkawa_>
I just lost my copy
afaerber has quit [Ping timeout: 244 seconds]
<javier__>
Tenkawa_: no worries, everything already landed in linux-next but is not working. Padovan is looking at it so it should be working soon
afaerber_ has quit [Quit: Verlassend]
<Tenkawa_>
ahh
<Tenkawa_>
linux-next is what I'm tracking so thats good for me
<Tenkawa_>
refresh my memory.. whats the best/appropriate way to get my git up to date with those changes without pulling the whole thing again
* tyler-baker
will test his chromebook 1 on mainline
<mkatic>
the patch adds some kds handling to dma-buf
<daniels>
mkatic: you'll also need that if you want to avoid tearing
<daniels>
mkatic: since that branch was in the middle of atomic rework, i went for the alternate approach, of clocking the mali high enough that tearing wasn't visible :\
<mkatic>
ok, i'll look into your tree, thank you
<mkatic>
tenkawa: i'm running 3.19.0-next.2015022 on my snow
<mkatic>
i have no problems with the display
<Tenkawa_>
mkatic: yeah it broke recently
<Tenkawa_>
it boots however lcd panel wont init
<Tenkawa_>
flashes and does nothing
<Tenkawa_>
keyboard/wifi/etc all work fine
<Tenkawa_>
just cant see a darn thing
<mkatic>
i did try 3.19.0-next.20150227 and that one got me a black display
<Tenkawa_>
yep
<Tenkawa_>
thats the prob
<Tenkawa_>
22 you said is working though?
<Tenkawa_>
I'll pull it for now then
<Tenkawa_>
I got to get away from 3.18
<mkatic>
yes, i only have a minor change to get wifi going
<Tenkawa_>
is that the next from kernel.org btw?
<mkatic>
runs quite well for me, much more stable than chromeos 3.8 kernel
<Tenkawa_>
yeah
<mkatic>
yes
<Tenkawa_>
cool thanks
<Tenkawa_>
pulling now
<Tenkawa_>
or at least about to
<mkatic>
forgot to mention it, shutdown and suspend are broken
<Tenkawa_>
yeah
<Tenkawa_>
I found that out the hard way
<Tenkawa_>
and thank you....
<Tenkawa_>
now I have to remember how to get git to go to a specific rev
<mkatic>
also, wizzupo pointed out that one should revert 68ff2a00dbf590dc9483c20f85fd1f1faa054604 to get wifi going on linux-next 20150222
<mkatic>
daniels: How come chromebook mali platform layer doesn't come included with arm mali tarballs?
<Tenkawa_>
ok.. i might have to ask how to do that later
<tyler-baker>
arrg yeah my snow's panel isn't working on next-20150303
<Tenkawa_>
tyler-baker: yup
<Tenkawa_>
]
<daniels>
mkatic: the platform layer is meant to be supplied by whoever supplies the, well, platform
<daniels>
mkatic: since it handles things like the exact register and clock details for power-on
<daniels>
mkatic: over time i'd expect that to using devicetree to describe as much as possible, but pre-DT that just wasn't a thing
<mkatic>
makes sense, thank you
<Tenkawa_>
tyler-baker: I'm going to try backing up all the dts/dtb defs and migrate those to the newer source where its broke
<mkatic>
i assume that most up to date version of the chromebook layer is the one the tree you posted here?
<tyler-baker>
Tenkawa_, I just reported DRM exynos failures on next
<Tenkawa_>
and see what that breaks in the process
<Tenkawa_>
cool
<Tenkawa_>
thank you
<tyler-baker>
Tenkawa_, let me try to add the IOMMU option back
<Tenkawa_>
iommu makes it really go weird
<Tenkawa_>
for me anyway
<tyler-baker>
well that was the case until today
<Tenkawa_>
ahhhh
<tyler-baker>
now we are seeing DRM related failures on the latest next on exynos instead of the IOMMU issue
<tyler-baker>
Tenkawa_, have you found a kernel version that works 3.19+?
<Tenkawa_>
tyler-baker: nope
<tyler-baker>
ok
<tyler-baker>
let me help run this down
<Tenkawa_>
I'll be very happy to help test
<Tenkawa_>
I was really enjoying the newer kernel on here until I had to rebuild the box (f2fs glitch)
<Tenkawa_>
ok... lets see if I can checkout the right tag
<Tenkawa_>
darnit
<Tenkawa_>
only saved the 4 one.. got to go dload again
<Tenkawa_>
isnt there a gitweb fo kernel.org?
<Tenkawa_>
nm
<Tenkawa_>
I was being goofy
<Tenkawa_>
which node worked? 22nd right?
<Tenkawa_>
scrollback is stuck
<mkatic>
20150222, correct
<Tenkawa_>
excellent8
<mkatic>
make sure you disable IOMMU
<Tenkawa_>
going to set it up to build here in a sec
<Tenkawa_>
fortunately I have my usb wifi too just in case
<mkatic>
just revert commit 68ff2a00dbf590dc9483c20f85fd1f1faa054604
<mkatic>
and enable userspace fw loader helper
<mkatic>
hm
<mkatic>
actually, i'm not sure if revert that commit is necessary if you enable the userspace fw loader helper
<mkatic>
better ask wizzup
<Tenkawa_>
ok
<Tenkawa_>
got my ralink just in case
<Tenkawa_>
that i know works
<mkatic>
gotta go, talk to you tomorrow
mkatic has quit [Quit: Leaving]
<Tenkawa_>
ok.. hopefully I'll bbiaf... might have to compile one module in for wifi if all goes well...
Tenkawa_ has quit [Quit: leaving]
Zotan has quit [Remote host closed the connection]
Zotan has joined #linux-exynos
<javier__>
tyler-baker: is so frustrating how fragile the exynos drm driver is :(
<javier__>
I think kevin is right and we should just disable until it becomes stable...
<daniels>
atomic \o/
<tyler-baker>
javier__, I agree, it's becoming difficult to for us to bisect some of these issue on next
<daniels>
it still has a load of issues, but atomic very neatly solves a hell of a lot of the things it got so very badly wrong, including fb/vblank refcounting
<javier__>
daniels: yeah but all the atomic work is meant for 4.1 right? while we need to fix the current issues in 4.0-rc1 and -next regardless of the atomic work
<daniels>
javier__: yeah agreed, i meant in general rather than right immediately now
<javier__>
daniels: totally agree
Tenkawa_ has joined #linux-exynos
Tenkawa_ has quit [Changing host]
Tenkawa_ has joined #linux-exynos
<Tenkawa_>
yay!!
<Tenkawa_>
back on 3.19
<Tenkawa_>
and internal wifi
<tyler-baker>
woot
<Tenkawa_>
thanks all
<Tenkawa_>
that 0322 version worked great
<Tenkawa_>
I'll start trying to determine why its broke newer tomorrow
<Tenkawa_>
oh thats so much better
<Tenkawa_>
time to go home.. I should be back later