[09:40] <alkisg> Hi, I'm getting screen corruption with https://launchpad.net/ubuntu/+source/linux-raspi2, can I do something about it, e.g. change some configuration value?
[10:31] <ppisati> alkisg: kernel version? and how do you get that?
[10:31] <ppisati> alkisg: plain framebuffer? or X?
[10:31] <ppisati> alkisg: beucase i just connected my raspi2 to my lcd
[10:31] <alkisg> ppisati:  4.2.0-1014.21           
[10:31] <ppisati> alkisg: and with the 13.19 kernel, i don't see any corruption
[10:31] <alkisg> I tried with fbdev and fbturbo
[10:31] <ppisati> alkisg: oh ok, let me try that
[10:31] <alkisg> Thanks!
[10:31] <ppisati> alkisg: ok, first le me test the new kernel
[10:32] <ppisati> alkisg: then i'll ask you about the rest
[10:32] <alkisg> Cool, thank you
[11:03] <ppisati> alkisg: are you using snappy? or another image?
[11:03] <ppisati> alkisg: cat /proc/cmdline
[11:03] <alkisg> ppisati: I'm using debootstrap in 14.04
[11:03] <alkisg> I manually installed the raspi2 kernel (actually I copied it to my ppa/trusty)
[11:04] <alkisg> $ cat /proc/cmdline
[11:04] <alkisg> dma.dmachans=0x7f35 bcm2708_fb.fbwidth=1920 bcm2708_fb.fbheight=1080 bcm2709.boardrev=0xa01041 bcm2709.serial=0x32c4a54a smsc95xx.macaddr=B8:27:EB:C4:A5:4A bcm2708_fb.fbswap=1 bcm2709.disk_led_gpio=47
[11:04] <alkisg> bcm2709.disk_led_active_low=0 sdhci-bcm2708.emmc_clock_freq=250000000 vc_mem.mem_base=0x3dc00000 vc_mem.mem_size=0x3f000000  dwc_otg.lpm_enable=0 console=ttyAMA0,115200 kgdboc=ttyAMA0,115200 console=tty1 init=/sbin/init-ltsp nbdroot=10.161.254.11:/opt/ltsp/armhf root=/dev/nbd0 elevator=deadline rootwait
[11:05] <alkisg> ppisati: did you try with the 4.2.0-1014.21      version, and it worked without corruption?
[11:06] <alkisg> By using the rpi2 kernel of fo0bar's ppa, I don't get screen corruption
[11:06] <ppisati> alkisg: yes, the framebuffer looks good here
[11:06] <alkisg> Do you think the version of the firmware would play a role?
[11:06] <alkisg> Which one are you using, the latest one from git?
[11:07] <ppisati> alkisg: https://launchpad.net/~p-pisati/+archive/ubuntu/embedded/+packages
[11:07] <ppisati> alkisg: raspberrypi2-firmware - 4.1.10-ba7a8fb-1
[11:08] <alkisg> Ah cool, I'll try with those
[11:09] <alkisg> ppisati: I need to leave for now, but I'll work on it tomorrow + for some more days, and I'll ping you with the results. Do you have some blog etc somewhere with instructions about your ppa?
[11:09] <alkisg> (meaning, which of those packages I should install, and which ones aren't necessary....)
[11:09] <ppisati> alkisg: i have a script that creates ubuntu armhf images
[11:09] <ppisati> alkisg: or i can point you
[11:09] <ppisati> alkisg: to a premade image
[11:09] <alkisg> I'm trying to support rpi2 in LTSP (Linux Terminal Server Project)
[11:10] <alkisg> Sure, please point me to whatever script/info you have...
[11:10] <alkisg> Thanks!
[11:10] <ppisati> alkisg: prebuilt images - http://people.canonical.com/~ppisati/ubuntu_embedded/
[11:10] <ppisati> alkisg: for the script...
[11:10] <ppisati> alkisg: https://github.com/piso77/ubuntu-embedded
[11:10] <ppisati> alkisg: sudo ./make_img.sh -b raspi2 -d 15.10
[11:11] <alkisg> Thank you! :)
[11:15] <ogra_> ppisati, do you plan to push the firmware package to the archive ? 
[11:15] <ppisati> ogra_: no plan ATM
[11:15] <ogra_> k
[11:15] <ppisati> ogra_: do you ned it?
[11:15] <ppisati> *need
[11:15] <ogra_> not urgently, no 
[11:17] <ogra_> hmm, i just noticed their VideoCore "hardfp" build is armv6 only ... 
[11:18] <ogra_> i wonder if that would actually work with our kernel (if the version matched)
[11:18] <alkisg> ppisati: what would be the proper path for ltsp users now? Currently to use your ppa, and in the future, e.g. in 16.04, those packages with the same names will be in the ubuntu archives?
[11:20] <ppisati> alkisg: except for the linux-firmware package, all the other packages are in the archive now
[11:20] <alkisg> The infrastructure in LTSP is already there, we just need the PPA, kernel variant, and extra package names for the rpi2. If we put those in ltsp-build-client.conf, the rpi2 will be supported by LTSP.
[11:21] <alkisg> ppisati: is the fbturbo driver needed?
[11:22] <ppisati> alkisg: never used, so i can't tell you
[11:22]  * alkisg tries to check if anything from https://launchpad.net/~fo0bar/+archive/ubuntu/rpi2/+packages is needed...
[11:22] <alkisg> I did some x11perf benchmarks with/without fbturbo, the difference was from small to non-existing...
[11:23] <alkisg> OK thanks a lot, I'll try with your firmware version to see if I still get corruption or not.
[11:23]  * alkisg waves
[11:23] <ppisati> alkisg: i'm going out for lunch soon, i'll do some X tests in the afternoon
[17:09] <caribou> ogasawara: apw: have you planned to take a cab upon your arrival in Austin ? I'll be there a bit before you
[17:09] <bjf> caribou, there's several of us arriving about that time
[17:09] <apw> caribou, i am planning on meeting up with whoever is still there
[17:09] <caribou> bjf: yeah, I was looking for your nic
[17:09] <ogasawara> caribou: yep, shall we try and find you?
[17:09] <caribou> ok, I'll wait for you then
[17:10] <apw> caribou, are you landing in the US for the first time in austin
[17:10] <caribou> ogasawara: apw bjf looks like there is 4 of 5 of us arriving arount 16:30
[17:10] <caribou> apw: nope, was there for the last sprint
[17:10] <bjf> caribou, i think there's enough that it should be easy to find someone to share a cab with
[17:11] <caribou> bjf: ok, I'll hang around until I see some of you
[17:11] <caribou> apw: I can hardly remember : the cab area is upstair or downstair ?
[17:11] <apw> caribou, sorry no i mean are you passing through imigration in austin
[17:12] <apw> caribou, i am so i will be no doubt slower than average
[17:12] <caribou> apw: no, in ATL
[17:12] <apw> ok
[17:12] <caribou> ok, I'll look around for any of you. Have a safe trim
[17:12] <caribou> s/trim/stip
[17:15] <arges> caribou: cabs are downstairs
[17:15] <bjf> cabs are at the end of the "cabs" signs
[17:15] <arges> : )
[17:15] <apw> heheh
[17:16] <caribou> arges: k. I remember meeting people there last time
[19:34] <noethics> hey, i'm wondering if there is some inotify-esque interface that instead of pooling resources can just match for any fs event with a certain dir prefix. or possibly every fs event and then i can filter myself