liquidAcid has quit [Quit: Leaving]
<Wizzup> swabbles: yo
<Wizzup> did you want me to make a dd of the chrome partition?
<Wizzup> swabbles: I'll make a dd-copy of it
tfiga has quit [Read error: Connection reset by peer]
tfiga has joined #linux-exynos
tfiga has quit [Read error: Connection reset by peer]
tfiga has joined #linux-exynos
<Wizzup> swabbles: armsoc works
tfiga has quit [Quit: http://quassel-irc.org - Chat comfortably. Anywhere.]
<Wizzup> speakers work, headphones work
<Wizzup> I thin
<Wizzup> speakers not yet, need to figure out what darn switch it was :)
<Wizzup> wifi is very unstable atm
Wizzup has quit [Quit: Lost terminal]
Wizzup has joined #linux-exynos
afaerber has joined #linux-exynos
<Wizzup> swabbles: do you remember the name of the wifi firmware?
<Wizzup> anyway.
<Wizzup> I took the firmware from chromeos.
<Wizzup> for wifi.
<Wizzup> Now it works as expected.
<Wizzup> as in, it doesn't fail miserably
<Wizzup> so yeah.
<Wizzup> getting there.
<Wizzup> as far as I am concerned the important thing now is finding out why /dev/fb0 acts so weird
<Wizzup> but I know so little about the graphics stacks...
<Wizzup> ah...
<Wizzup> Frame buffer device information:
<Wizzup> Name :
<Wizzup> Size : 0
<Wizzup> Address : (nil)
<Wizzup> That WOULD explain it.
<swabbles> Wizzup: the last time the firmware for Marvell WiFi was updated was somewhere in 2013, I believe.
tfiga has joined #linux-exynos
<swabbles> Or at least, the binary that the Chromebook 2 is supposed to use.
<Wizzup> maybe a DT thing
<Wizzup> swabbles: the one from linux-firmware-20140603 oesn't work though
<Wizzup> it dies very rapidly
<Wizzup> swabbles: do you know if the FTD contains FB entries?
<swabbles> Only fimd and lcd, IIRC.
<swabbles> Or well, that would be close to what you are looking for.
<Wizzup> yes. I found that
<Wizzup> Looking at the other chromebook is so weird after this high res.
liquidAcid has joined #linux-exynos
<Wizzup> I guess it's all relative.
<Wizzup> lol
<swabbles> Wizzup: what kind of issues did you get with the linux-firmware blob?
<swabbles> Because it seems that I have associated with the AP, DHCP seems to work fine, but pinging doesn't (for instance).
<Wizzup> Yes. It's super slow and quickly halts.
<Wizzup> I got a dhcp lease, tried ssh
<Wizzup> and didn't get far.
<Wizzup> waited for minutes, too.
<Wizzup> Just copy the mrvl folder from chromeos /lib/firmware
<swabbles> You only need sd8897_uapsta.bin IIRC.
<Wizzup> or rather, I copied the entire /lib/firmware from cros to my /root/ and mv mrvl mrvl_old in my own lib/firmware and copied the mrvl folder from /root/chros_firmware.
<Wizzup> and then it works just as nice as the cros wifi.
<swabbles> Heh, it is complaining that s5p_mfc firmware is missing.
<Wizzup> yes
<Wizzup> which is weird
<Wizzup> chrometwo ~ # ls /root/firmware-chromeos/s5p-mfc-v8.fw /lib/firmware/s5p-mfc/s5p-mfc*
<Wizzup> /lib/firmware/s5p-mfc/s5p-mfc.fw /lib/firmware/s5p-mfc/s5p-mfc-v6-v2.fw /lib/firmware/s5p-mfc/s5p-mfc-v8.fw
<Wizzup> /lib/firmware/s5p-mfc/s5p-mfc-v6.fw /lib/firmware/s5p-mfc/s5p-mfc-v7.fw /root/firmware-chromeos/s5p-mfc-v8.fw
<swabbles> sd8897_uapsta.bin was added on the 4th of November 2013 btw.
<swabbles> Has never been updated since.
<Wizzup> the md5sums were different
<Wizzup> I checked
<swabbles> It's a good idea to get someone to update linux-firmware then :).
<Wizzup> if it is better in all cases, idk :)
<swabbles> I only copied sd8897_uapsta.bin, WiFi does work properly now :D.
<Wizzup> swabbles: if I get turbofb to work
<Wizzup> then we're all set
<Wizzup> driver-wise it'll be the same as the first chromebook.
<swabbles> Or any Exynos Chromebook, thus far.
<tfiga> swabbles: mfc firmware should be on top of /lib/firmware
<tfiga> without the subdirectory
<Wizzup> What tfiga says seems correct, see also my ls
<swabbles> This is weird.
<tfiga> the layout in linux-firmware is broken
<tfiga> but AFAIK it is going to be fixed
<swabbles> >driver_version = mwifiex 1.0 (15.68.4.p42)
<swabbles> >WLAN (SDIO) firmware version: 15.69.2.p11
<swabbles> Last one is @ linux-firmware.git.
<Wizzup> is that the cros one?
<swabbles> cros appears to be 15.68.4.p42
<Wizzup> speakers work
afaerber has quit [Quit: Verlassend]
afaerber has joined #linux-exynos
liquidAcid has quit [Quit: Leaving]
<Wizzup> swabbles: okay
<Wizzup> going to nuke the ssd
<Wizzup> :D
<Wizzup> btw
<Wizzup> why do you suggest gdisk instead of cgpt?
<Wizzup> as in there
<Wizzup> It seems a lot less verbose
<Wizzup> compared to all the <enter> stuff
<Wizzup> I guess it's personal preference :)
<WarheadsSE> Wizzup: I use fbturbo atm, just not with mali.
<Wizzup> WarheadsSE: fbturbo is great, I use it too for my chromebook
<Wizzup> comparatively great
leming has joined #linux-exynos
<WarheadsSE> Yeah, just the ASM speed ups help
<Wizzup> It doesn't work on the new chromebook C32 atm though
<Wizzup> well, mostly armsoc is super slow
<WarheadsSE> Oh? Going to load mine up on a 16G C10 uSD later
<WarheadsSE> Yeah, not using armsoc on anything
<Wizzup> WarheadsSE: fbset -i shows size: 0 ; addr: nil for fb0
<WarheadsSE> I've never gotten a speedup on that
<Wizzup> and writing to it doesn't produce anything sensible
<WarheadsSE> yeah, thats a problem
<Wizzup> that needs to be fixed
<Wizzup> btw, did you say the arch arm people were also interested?
<Wizzup> somewhere tonight or tomorrow evening I think we're ready to let the world know linux runs "okayish" on the devices
* WarheadsSE cough
<WarheadsSE> I am Arch Linux ARM core dev.
<WarheadsSE> Myself and leming do ~ 90% of the work
<WarheadsSE> leming doing the larger portion
<Wizzup> ah
<Wizzup> :p
<WarheadsSE> But we're scheming for device support tonight/tomorrow
<WarheadsSE> Got to go mow .. whee
<Wizzup> okay
<leming> suppose i'll see if this kernel will boot
<swabbles> Wizzup: because fdisk-like utils allow you to use sizes in human-friendly formats mostly.
<Wizzup> I quite like the bpaste way, but sure
* Wizzup dislikes creating initramfses
<Wizzup> so
<Wizzup> the kernel doesn't compile for me if I *disable* optimise for size
* Wizzup cries
<Wizzup> swabbles: btw, I think the btrfs compression saves me about 3.5GB
<Wizzup> at 6GB in use
<Wizzup> swabbles: 'The device you inserted does not contain chromeos' is what it tells me if I want to boot from ssd
<Wizzup> lolwut
<Wizzup> ah, nvm
<Wizzup> I need to sign the kernel obviously.
<Wizzup> (and make the kpart)
* Wizzup slaps himself
<Wizzup> meh, black screen with this initramfs one
<Wizzup> btw
<Wizzup> the instructions look really good I think
<Wizzup> At least, I'm happy
<Wizzup> maybe we should actually link them somewhere now though
<leming> could condense the drive setup using just cgpt.. far less commands for the same end result
<Wizzup> leming: as in http://bpaste.net/raw/139643/ ?
<Wizzup> (thats for snow btw)
<leming> hah, yeah, those are gentooized versions of our installation process for snow :P
<Wizzup> I can make a wiki account for you
<leming> sure
<leming> can just use those few cgpt commands instead of all the gdisk stuff plus cgpt
<leming> btw, noticed the uboot script got pulled, is that not necessary?
<swabbles> Meh, I would have used cgpt if it used human-friendly size formatting.
<swabbles> leming: well, it doesn't seem to work for me.
<swabbles> @ script.
<swabbles> Neither does nv u-boot seem to work atm.
<leming> yeah, it's not easy to parse cgpt sizes with your brain
<leming> not having much luck booting up here.. backtraces are jamming up the screen and can't scroll up :/
<swabbles> About drm/rootfs?
<Wizzup> probably can't find rootfs
<leming> yeah, drm_crtc.c stuff
<Wizzup> cmdline force on?
<leming> currently set to use bootloader args if available, suppose i could force what i want through the kernel
<leming> probably the only way to do it without better control via a chainloaded uboot
<Wizzup> That's what we found so far
<leming> anything special needed other than root and console? anything for video?
<swabbles> console=tty1 earlyprinkt debug should be fine video-wise.
<leming> easy enough
<leming> you guys poked at the 3.10 and 3.14 branches they have? or are the needed bits only in 3.8?
<Wizzup> I didn't see a dts in 3.14 at least
<leming> ah
<leming> yep, forcing command line did the trick, booted up nicely now
<leming> how's the experience been with that mwifiex_sdio wifi?
<leming> for me, it's practically unusable
<swabbles> leming: with the sd8879_uapsta.bin file from ChromeOS it works fine.
<swabbles> With the firmware blob in linux-firmware not so much.
<leming> ah
<swabbles> The irony is that the firmware blob in linux-firmware is supposedly newer (according to the versions).
<swabbles> :P
<leming> i don't think i've seen a working mwifiex_sdio implementation without some other version of the firmware, now that i think about it
<leming> indeed.. replaced with the blob from cromeos and everything is swell, thanks