ChanServ changed the topic of #linux-rockchip to: Rockchip development discussion | Wiki at http://linux-rockchip.info | Logs at http://irclog.whitequark.org/linux-rockchip | ML at http://groups.google.com/group/linux-rockchip
<amstan> steev: but why, i'm using mmind's exact kernel and he has it running, what else is different?
<steev> amstan: no, i meant, on 3.14
<amstan> steev: that's also weird, because i'm using the chromeos kernel
<steev> the logs of the 3.14 reflect what i see, the logs of 4.1 reflect... it almost working
nighty-_ has quit [Quit: Disappears in a puff of smoke]
<steev> hm, actually, 4.1 says it's card name is Rockchip, 3.14 says vgem
<amstan> hmm
RokQuarry has joined #linux-rockchip
<amstan> steev: and this doesn't look different? https://bpaste.net/show/c568abc7c046 (this is the 3.14 kernel compiled using the chromeos build system)
<steev> [ 53.227] (II) DriverName is [vgem]
<steev> that looks like what i was getting using mmind00's armsoc driver
<steev> hm
<steev> why does it open it twice
<steev> oh nevermind
<steev> look at likes 87-98 on your most recent paste, and then on the 4.1 paste
<steev> 3.14 has card0 (i'm assuming that is vgem) and card1 (maybe that's the one we want?)
<steev> looks like it
<steev> according to /sys/devices/display-subsystem.5/drm/card1/card1-eDP-1/modes
<steev> amstan: for giggles, i'm gonna build a small image, with vgem disabled
<amstan> there you go
<amstan> vgem is out
<amstan> looks better
<steev> so vgem being in, is what was fucking it up
<amstan> leming: so we made it go farther by disabling VGEM
<amstan> matches 4.1 now
<leming> kernel option?
<amstan> yep
<amstan> CONFIG_DRM_VGEM=n
<amstan> still doesn't work, but at least it's the same problem on all kernels
* amstan is leaving for food
* mmind00 is leaving for bed ;-)
<leming> you're making sure /dev/mali0 is writable by your user?
<steev> i wasn't
<steev> but then, i didn't expect it to work yet
<steev> also, i didn't realize i was missing it heh
levd has joined #linux-rockchip
levd1 has joined #linux-rockchip
levd has quit [Read error: Connection reset by peer]
<amstan> leming: i wasn't either, good idea
<amstan> works! \o/
<steev> amstan: \o/
<steev> so yay
<amstan> kde crashes though back to the login screen
<steev> amstan: so install ze mali-drivers and then the armsoc driver from mmind with vgem turned off, basically
<amstan> but... the loading screen is so smooooth
<amstan> EGL_VERSION 1.4 Midgard!
<amstan> leaving for board games
<amstan> cya
levd has joined #linux-rockchip
levd1 has quit [Ping timeout: 252 seconds]
levd has quit [Ping timeout: 248 seconds]
levd1 has joined #linux-rockchip
levd has joined #linux-rockchip
levd1 has quit [Ping timeout: 264 seconds]
markm has quit [Ping timeout: 250 seconds]
markm has joined #linux-rockchip
ganbold_ has quit [Quit: Leaving]
ganbold has joined #linux-rockchip
RokQuarry has quit []
levd1 has joined #linux-rockchip
levd has quit [Ping timeout: 252 seconds]
levd has joined #linux-rockchip
levd1 has quit [Ping timeout: 240 seconds]
levd1 has joined #linux-rockchip
levd has quit [Ping timeout: 256 seconds]
levd1 has quit [Ping timeout: 255 seconds]
markm has quit [Ping timeout: 246 seconds]
markm__ has joined #linux-rockchip
field^Mop has joined #linux-rockchip
GriefNorth has joined #linux-rockchip
<mmind00> amstan steev: congrats :-D
<amstan> mmind00: still don't have it working when i got home
<amstan> but i'm tired
<mmind00> do you have libGL.so.1 linking to mesa?
<amstan> no, uninstalled mesa
<mmind00> amstan: that's probably it
<amstan> it worked earlier today though
<mmind00> amstan: strange, i.e. at least for me kde crashes when it's missing libGL
<amstan> but with the kernel compiled on my home computer(to remove vgem) i couldn't get it to work again
<mmind00> strange
<steev> amstan: was armsoc build before or after installing mali drivers
<steev> and, i haven't quite gotten that far yet.... i only just finished building an image i'm happy with
<amstan> i think it's from the arch repos
<steev> oh, i thought you built it yourself
<steev> from mmind00's stuff or
<amstan> it is, yes
<steev> ah
<steev> not sure, i'm not quite to that part yet
<steev> need to yank the stuff off my other sdcard first
Bludot has quit [Quit: Connection closed for inactivity]
Bludot has joined #linux-rockchip
field^Mop has quit [Ping timeout: 240 seconds]
hipboi_ has joined #linux-rockchip
hipboi has quit [Ping timeout: 264 seconds]
ganbold has quit [Quit: This computer has gone to sleep]
ganbold has joined #linux-rockchip
markm__ has quit [Ping timeout: 256 seconds]
markm__ has joined #linux-rockchip
Bludot has quit [Quit: Connection closed for inactivity]
cosm has quit [Ping timeout: 250 seconds]
cosm has joined #linux-rockchip
cosm has quit [Ping timeout: 255 seconds]
cosm has joined #linux-rockchip
markm__ has quit [Ping timeout: 256 seconds]
markm__ has joined #linux-rockchip
cristian__c has joined #linux-rockchip
cristian__c is now known as cristian_c
RokQuarry has joined #linux-rockchip
premoboss has joined #linux-rockchip
RokQuarry has quit [Ping timeout: 246 seconds]
RokQuarry has joined #linux-rockchip
cosm has quit [Ping timeout: 255 seconds]
Bludot has joined #linux-rockchip
cosm has joined #linux-rockchip
RokQuarry has quit [Ping timeout: 250 seconds]
premoboss has quit [Read error: No route to host]
markm__ has quit [Ping timeout: 256 seconds]
cosm has quit [Quit: Leaving]
RokQuarry has joined #linux-rockchip
premoboss has joined #linux-rockchip
field^Mop has joined #linux-rockchip
markm has joined #linux-rockchip
<steev> mmind00: maybe we should create the libGL.so symlinks in the update-mali-drivers script?
<steev> or... somehow in the debian package
<steev> amstan: so, i have it sorta working here - but when using the armsoc driver.... cannot use rotation...
<mmind00> steev: maybe ... the glx-alternative-(mesa/fglrx/nvidia) packages in Debian do something like this too ... which of course would collide if glshim was to provide libGL ... that is the question I'm still thinking about, how both can work in a correct way, respecting package policy and everything ;-)
<steev> yeah i'll take a look at them when i get a chance
<steev> one thing i've noticed.... performance is not as buttery smooth as it is with fbdev and xorg 1.17
<steev> and mesa
<steev> mmind00: libdrm-rockchip - is that needed?
<mmind00> steev: nope ... everything seems to use generic libdrm functions only
<steev> hm
<mmind00> steev: i.e. the libdrm-rockchip stuff only relates to the older drm in 3.14 it seems
<steev> well i'm using that :)
<mmind00> haha
<mmind00> steev: but the whole 3.14 also uses a lot of other stuff in the drm ... for example, if I remember correctly, the memory fencing between gpu/cpu is also something chromeos-specific which results in custom ioctls and from them custom libdrm-rockchip code
<steev> right
<steev> amstan: "rk3288-dmc ff610000.dmc: timeout before pausing cpus\nrk3288-dmc ff610000.dmc: failed to change dmc freq" - anything to worry about there?
Bludot has quit [Quit: Connection closed for inactivity]
<amstan> steev: i noticed that too
markm has quit [Ping timeout: 252 seconds]
cristian_c has quit [Quit: Bye]
RokQuarry has quit [Ping timeout: 246 seconds]
markm has joined #linux-rockchip
RokQuarry has joined #linux-rockchip
markm has quit [Ping timeout: 246 seconds]
markm has joined #linux-rockchip
cyrozap has left #linux-rockchip [#linux-rockchip]
<steev> /* This can happen if a irq/softirq delays us long enough. */
<steev> and the other one happens in rk3288_dmcclk_set_rate
<amstan> steev: so you got it working too?
<amstan> steev: because with the arch kernel with VGEM=n it doesn't work
<amstan> though, to be fair it also kernel panics once in a while
<steev> amstan: yes, but as I said I went back to fbdev and mesa because performance was kinda meh
<steev> Plus the whole screen rotation not working
<steev> I'm about to go wash my car but I can post up my kernel config
<steev> After that, I mean
GriefNorth has quit [Ping timeout: 255 seconds]
<amstan> steev: ok, i'm an idiot
<amstan> steev: see any problems with this? "sudo chmod +x /dev/mali0"
<amstan> i was typing that the whole time
<amstan> convinced that's what i would have to type
<karlp> only gave root the +x?
field^Mop has quit [Ping timeout: 255 seconds]
<mmind00> amstan: :-D
<amstan> karlp: looks like it :)
<amstan> what am i supposed to do with libgl?
<amstan> keep it the mesa one?
<amstan> chromium wants libgl apparently
<mmind00> amstan: yep, just symlink it to the mesa one
<mmind00> this will simply fall through to software rendering
markm has quit [Read error: Connection reset by peer]
<mmind00> amstan: and you might want to use something like "chromium-browser --use-gl=egl" ;-)
markm has joined #linux-rockchip
<steev> amstan: lol. Bruh just copy the udev rule already
<amstan> lol
<steev> Also... I've been monitoring CPU usage. When the screen is rotated, one of the cpus stays at a steady 25% for processing. Is there something similar in cros so I can monitor the CPU usage?
<steev> Also, for some reason, when I switched to using armsoc, my display became eDP-1
<amstan> mmind00: yay, works nicely
<amstan> steev: yeah, it's sad that we can't rotate
<amstan> have to get that going somehow
<amstan> yep, noticed that too
<steev> I'm gonna look into building the armsoc one from cros. I did it once but new sdcard
RokQuarry has quit [Ping timeout: 246 seconds]