[01:55] Well, I've got a near-chroot method working. [01:55] *chroot method near-working... [01:55] Now I've figured out how the Ubuntu for Android guys did it... haha [01:58] Any ARM Xorg guys on right now? [02:26] ogra_: so my 'ssh hangs hard when I try to connect to my nexus7' bug is still here, even after repeated reinstalls === zz_chihchun is now known as chihchun === vhadil is now known as naruto === naruto is now known as Guest19242 [07:39] hi [07:39] i'm trying to cross compile a kernel on my desktop for a panda board but i get an dependency error [07:40] but the dependencys are met and i still get that error [07:48] wingze: what error? [07:53] good morning [07:56] good morning [07:57] dpkg-checkbuilddeps: Unmet build dependencies: libelf-dev libnewt-dev binutils-dev libdw-dev libgtk2.0-dev | gtk2-devel [07:57] i have all that installed but still get the error [08:08] this is crazy [08:12] ok got it working it was a build dependency error it needed gcc-arm-linux-gnueabi === yofel_ is now known as yofel === chihchun is now known as zz_chihchun [09:50] I have a headless target which plymouth is not installed, and I'm missing some of the logs from initramfs, even when I have console= console= in the kernel commandline. Is it possible to get this log out on the console, or do I need plymouth even on a headless system? [10:58] sveinse: plymouth has text-mode plugins & can run fine on headless and output text. [10:59] sveinse: i dunno how to do this without plymouth. [11:19] is there any reasonably priced 10" tablet that can run Ubuntu well [11:19] bonus points for Flash support in any way, shape or form (for livestreams that don't HTML5) [11:47] xnox: Hmm. plymouth is installed after all (dep'd by mountall). But for some reason nothing appears in the console during initramfs. I have another device (another kernel) also with the console= console= bootcmd construct, but here it does work. [11:48] Kernel config perhaps. Don't remember seeing any config related to that (since the initial printk of the kernel is output properly on the console) [12:21] asiekierka: flash is a no go with ubuntu-arm [12:22] asiekierka: for 10" tablets I don't really know of one that runs it "well", especially at a reasonable price, it runs fine on the tf201 but that isn't reasonably priced, and while fine I wouldn't call it "well" === chihchun is now known as zz_chihchun [13:13] lilstevie: do you have a device tree for tf201? === rsalveti_ is now known as rsalveti [13:40] janimo, hmm, your HDA trick doesnt work if i build the HDA driver as module [13:40] :( [13:40] well we build it in I suggest [13:40] most our hw stuff is built-in anyway [13:40] as we know it is there [13:40] well, i'm *extremely* short on bootimg space [13:40] ah I remember [13:41] how much more can fit there? [13:41] so modular would indeed be preferred if we add any stuff [13:41] -1M [13:41] :P [13:41] ohm that should fit fine then for a while :) [13:41] the boot is slow enough that it justifies to have plymouth in the initrd [13:42] which i cant do atm [13:42] I did not look closely at what we have modular or built-in [13:42] everything is builtin [13:42] I think the majority of settings are as in the original android kernel [13:42] well, nearly at least [13:42] vmlinuz is 4.5M [13:43] (ac100 is 1.2 or so) [13:43] I say when we hit that limit, we take out non-essential hw stuff (maybe all the USB peripherals) [13:43] you dont get it ? we are about 1M above [13:43] and high-level networking and fss [13:43] ah, aboce? I thought 1M to go [13:43] it currently only works due to some evil hacks and diversions i add [13:43] that - 1M confused me [13:44] well, you asked how much more :) [13:44] well I thought you said dash 1M not minus 1M [13:44] * ogra_ is neadly done with a set of quietening patches for the boot [13:44] *nearly [13:45] well anything not strictly tegraSoC hw could be modular IMHO [13:45] theoretically [13:45] sadly practically that will fall over in many places [13:45] well FSs that are not ext4, TCP, and other bits Asus/Google did not modify should be fine I think [13:46] Iknow battery and some other bits don't work as modules [13:46] all the non std drivers that are woven into android stuff usually require to be builtin [13:46] right, but the big space consumers are network,fs and the rest of the code [13:46] battery, wlan, bluetooth, i bet the accelerometer and all other sensors too [13:46] so if we can make those modular it should fix the bootimg size issue [13:46] right, but those are all relatively small footprint i bet [13:47] well, wlan would really help [13:47] we could drop all network bits to modules that way (protocols etc) [13:47] wlan itself wont gain us much, but firewall, protocols etc [14:15] ogra_, do we have a nexus meeting today? [14:16] yup [14:18] GRRR ! [14:18] i just had the boot quiet ... [14:19] enabling HDA gets me another message on the screen [14:19] why cant people use pr_info for info messages instead of pr_err [14:19] grmbl [14:20] ogra_, is onboard not working when the gksu dialog after apport pops up a known issue? [14:20] yes, gksu is supposed to go asap [14:20] I seem to have seen lots of onboard related bugs filed but am not sure which are still relevznt [14:20] what replaces gksu? [14:20] pkexec, since about 4 releases now [14:21] I keep getting an apport dialog on each session start, but cannot send the report due to this [14:21] * janimo never heard of pkexec [14:21] yet you use it all the time :) [14:21] so by asap you mean a couple weeks or a couple more cycles? :) [14:22] its part of policykit/consolekit [14:22] i mean this cycle [14:22] there are two apps left that need porting, then we can unseed gksu [14:22] one is update-notifier (which is the apport bit you see) and the other is xdiagnose [14:22] I don't know what makes up the 'January-2013 Linux Plumbing stack' [14:23] where does that come from ? [14:23] I can only assume it is different from 'December 2012 Linux Plumbing stack' and 'February 2013 Linux Plumbing stack' [14:24] oh, you mean wrt pkexec :) [14:24] wrt the whole moving thing which keeps getting in the way of actually writing features [14:24] half-joking [14:25] yay, completely quiet boot ! [14:25] everything from kernel up until the login manager [14:25] if we now just could have plymouth in initrd [14:26] ogra_, what happened to the no initrd at all attempt? [14:26] i got stuck at the console-setup race [14:26] and was told bootspped isnt a prio anymore [14:26] *speed [14:27] so for now i just care about it having bling :) [14:27] marvin24_: nope [14:28] marvin24_: probably should work on that [14:31] janimo: Who's on armadaxp SRU duty right now? You or Ike? [14:31] Ike I guess [14:31] Kay. We have a rushed revert in master that needs a quick rebase for derivatives. [14:32] I was told I'd be doing non-SRU stuff last December, so my last armadaxp thing was the B0 enablement in Precise [14:32] but if I misunderstood I can be convinced otherwise [14:32] infinity, or if there's an emergency I can of course try handling it [14:33] what happened to NCommander ? [14:33] doesnt he do any server kernel work anymore ? [14:33] ogra_, not on armadaxp at least [14:34] janimo: I'll ping Ike. [14:34] janimo, btw, compiling HDA in or as module doesnt get me any sound working (suspend/resume still does) [14:34] are you sure your HDA stuff isnt just a red herring ? [14:34] ogra_, very strange. I got that repeatedly after enabling it [14:34] well I am not 100% sure, it just worked in my testing [14:35] ogra do you have the volume turned on at boot? [14:35] no, it is muted [14:35] but unmuting doesnt get me any sound [14:35] that sucks [14:36] * ogra_ tries another reboot [14:36] could it be i accidentally set some alsa configs which made it work? I'll dump my alsa state and paste it [14:36] now that rotation works wutomatically, we should probably consider to just boot in portrait [14:37] ogra_, alsa store gives me this: http://paste.ubuntu.com/1569919/ [14:38] ogra_, I agree with portrait booting. needs to drop the fbcon rotate cmdline arg too [14:38] it also doesnt seem to make any difference if HDA is card 0 or 1 [14:38] ogra_, so reboot and still no sound? [14:38] janimo, and i think yoiu did a hack to the touch driver to switch to 1280x800 from 800x1280 [14:39] i guess we need to drop that too [14:39] janimo, right, but sound after resume [14:39] ogra_, right, I tested with that dropped this week [14:39] will need the matrices shuffled in the rotate scripts [14:39] great, telly me once it entered the kernel and i'll adjust the cmdline [14:39] but better that way so portrait is the 'normal' one [14:39] and we drop that needless delta vs the android kernel [14:39] right [14:40] and getting plymouth inoto landscape seems impossible anyway [14:40] ogra_, btw the cmdline could be altered separately no? Isn't that only affecting the fbcon (so bootup messages) ? [14:40] separately from ? [14:40] from the rotate script and kernel stuff [14:40] the cmdline has two parts, first is hardcoded in the bootloader ... [14:41] second is from abootimg [14:41] the abootimg part can be altered [14:41] right, abootimg fbcon rotate=1 [14:41] yeah [14:41] which is only about display in console mode [14:41] i'll drop that from the builds as soon as the kernel fix is there [14:41] the kernel bit is only about touch input [15:00] janimo, http://paste.ubuntu.com/1569980/ [15:01] thats the diff to my automatically stored asound.state file [15:01] (yours is +) [15:01] seems you played with all the controls :) [15:02] i wonder if you probably stil have sound working if you now disable HDA in your kernel [15:49] ogra_, can you try this .deb? http://people.canonical.com/~jani/linux-image-3.1.10-8-nexus7_3.1.10-8.21_armhf.deb [15:49] sure [15:49] I freshly rebuilt with the HDA and the axes inversion changes [15:49] it works for me if I set the volume up (it is again off at boot) [15:51] bah, the noise is back at boot :P [15:52] nope [15:52] no sound [15:53] janimo, how exactly do you test btw ? [15:53] * ogra_ just relies on the plop plop from the volume control applet [15:53] (when adjusting with the buttons) [15:54] ogra_, canberra-gtk-play --file=/usr/share/sounds/gnome/default/alerts/glass.ogg [15:54] geez, complex [15:55] ergh [15:55] well, I have no media to play, and this is a simple thing to test via ssh [15:55] and indeed the touchscreen is totally screwed now [15:55] with export DISPLAY=:0 of course [15:55] well, i just press vol-up/down [15:55] makes an event sound if audio works [15:58] * ogra_ reboots back to a sane kernel that likes my touching :P [16:01] ogra_, so no sound with this deb? [16:01] janimo, right [16:01] well, after resume i had it [16:01] as with my kernel [16:01] well, lets have a meeting :) [16:02] ogra_, well with this kernel I get sound without having to do suspend [16:03] ogra_, http://paste.ubuntu.com/1570120/ the matrices changed around to have rotation work with this deb [16:03] to be uploaded once we have the kernel change [16:34] xnox, Bus 001 Device 067: ID 04e8:6866 Samsung Electronics Co., Ltd [16:34] Galaxy Nexus [16:37] janimo: thanks. [16:39] anyone else with a nexus 7 willing to test the deb I linked above to see if it fixes the no sound on boot problem? [16:40] yeah, why not, tell me what to do [16:42] just install and reboot I guess) [16:42] right, like we discussed in the meeting :) [16:43] Tassadar, yes, install and reboot but make sure you have ssh enabled or a keyboard or mouse [16:43] as the touchscreen will be messed up :) [16:43] via ssh then export DISPLAY=:0 [16:43] and canberra-gtk-play --file=/usr/share/sounds/gnome/default/alerts/glass.ogg [16:43] or any other sound playing app you wish [16:44] okay, I'll just update my installation first [16:44] for me this works on boot without having to issue a suspend [16:44] but you may need to turn the volume on [16:56] hmm, it does not work for me [16:57] Tassadar, does it work after a suspend cycle (as it did before) [16:57] yes [16:57] Anyone got some initfs advice for me? I'm trying to get Ubuntu to run natively on the Toshiba Thrive. [16:58] just to make sure it is right kernel, my /proc/version looks like this: [16:58] Linux version 3.1.10-8-nexus7 (jani@parapapio) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-19ubuntu1) ) #21 SMP PREEMPT Fri Jan 25 16:55:55 EET 2013 [16:58] so I think it is the right one [16:58] yeah, it is, the touch screen is messed up [17:00] I was able to get into a chroot Ubuntu, but the X server never displayed anything. I know it had access to the framebuffer because there was garbage on the display before I started it. [17:00] it doesn't need any modules or cmdline params during boot or something like that, does it? You said it is built in the kernel, right? [17:04] Tassadar, yes it is the right kernel - especially since the touchscreen handling is broken right [17:04] Tassadar, indeed no extra boot params [17:04] Tassadar, thanks for testing [17:04] I am building anothe deb now without the messed up touch now [17:05] theres one hda-codec message if that's any help, "hda-codec: No codec parser is available" [17:07] Tassadar, I have that too in dmesg === XenGi is now known as XenGi_ [17:19] Would I need to set a custom cmdline when I make the boot image on mine? If so, what? [17:24] Tassadar, btw I put a new deb at the same URL, this time without the messed up screen [17:25] did you test with the canberra command line ? [17:25] ogra_, did you test with that canberra line btw? [17:26] that works for me but playing an mp3 in rhythmbox is silent [17:27] or rather it plays crack. I think I'll soon do a test-reinstall :) [17:29] well, we dont ship any mp3 codecs [17:30] rebooting to your latest kernel now [17:32] ogra_, well I installed mpg123 and rhythmbox installed its own [17:32] my alsa settings are likely broken now [17:32] not sure how the gtk player just works [17:32] anyway, no change with the kernel from the ML [17:32] volume up? [17:32] Hmmm... Sound works perfectly, but I'm not getting anything to display. [17:32] and yes, i used the canberra methid too [17:33] *method [17:33] You and I are in completely opposite sides of the spectrum... haha [17:33] volume is indeed up [17:33] AmEv, does sound not work for you with the regular kernel unless suspending first? [17:34] I'm actually on a Tegra 2 device in a Android-to-Ubuntuchroot. [17:34] I have absolutely zero speaker issues. [17:34] My problem is that I can't get anything to display. [17:35] what do you want to display ? [17:35] android doesnt ship an X server [17:35] janimo: no, no sound until suspend/resume [17:35] right, same here [17:35] I've got Ubuntu on an SD card. [17:35] but touchscreen works) [17:35] Chroot. [17:36] AmEv, but you are aware that ubuntu apps use X ? [17:36] Yes. [17:36] Tassadar, mine is a 16G model if that is any help but I doubt it has to do with the model [17:36] mine too [17:36] I'm trying to start X, but nothing's displaying. [17:36] android doesnt ship an X server, you cant display anything [17:37] janimo, i really think HDA is a red herring ... [17:37] janimo, but i also think you must have something that makes it work :) [17:37] ogra_, could well be [17:37] yes, so I should not just wipe and reinstall [17:37] I may have a gem here somewhere [17:37] :) [17:37] we should consult diwic again [17:38] I've built an X-compatible kernel. I've been able to get Zygote, Surfaceflinger, Media, and DRM stopped, then chroot into Ubuntu, and start X. [17:38] My final goal is to get a working native, zero-Android solution. [17:38] Kernel source is https://github.com/Thrive-Hackers/tostab3-gnu-linux-kernel [17:39] ok, indeed it now works with the stock kernel [17:39] sigh [17:39] my nexus just got fixed and I don't know how [17:40] In the meantime, I'm trying to esssentially shut down Android and shut down Ubuntu. [17:40] both bohdi and plasma have sound working without suspend, but I don't see jeff here right now :/ [17:40] *start up Ubuntu [17:41] * janimo wonders whether bodhi and plasma are applications or people [17:41] distributions) [17:42] bohdi is based on debian and plasma active on MER (it is "kde on tablet" project), but both use ubuntu's kernel [17:43] Still, if someone does want to try remoting in, I'm willing to. === W1N9Zr250 is now known as W1N9Zr0 [17:47] Almost in... === doko_ is now known as doko [17:54] Hmmm... WiFi module isn't loaded. [17:57] Okay, this is new... X is segfaulting. [18:00] Wall of text coming in... [18:01] root@localhost:/# startx [18:01] startx [18:01] xauth: error in locking authority file /data/.Xauthority [18:01] xauth: error in locking authority file /data/.Xauthority [18:01] X.Org X Server 1.11.3 [18:01] Release Date: 2011-12-16 [18:01] X Protocol Version 11, Revision 0 [18:01] Build Operating System: Linux 2.6.38-1209-omap4 armv7l Ubuntu [18:01] Current Operating System: Linux localhost 2.6.39.4-gnu-v0.01-g49f5e55-dirty #6 SMP PREEMPT Thu Jan 24 18:24:20 MST 2013 armv7l [18:01] Build Date: 05 April 2012 02:47:41AM [18:01] xorg-server 2:1.11.4-0ubuntu10 (For technical support please see http://www.ubuntu.com/support) [18:01] Current version of pixman: 0.24.4 [18:01] Before reporting problems, check http://wiki.x.org [18:01] to make sure that you have the latest version. [18:01] Markers: (--) probed, (**) from config file, (==) default setting, [18:01] (++) from command line, (!!) notice, (II) informational, [18:01] (WW) warning, (EE) error, (NI) not implemented, (??) unknown. [18:01] (==) Log file: "/var/log/Xorg.0.log", Time: Fri Jan 25 09:59:34 2013 [18:01] (==) Using system config directory "/usr/share/X11/xorg.conf.d" [18:01] Backtrace: [18:01] Segmentation fault at address 0x4 [18:01] Caught signal 11 (Segmentation fault). Server aborting [18:01] Please consult the The X.Org Foundation support [18:01] at http://wiki.x.org [18:01] for help. [18:01] Please also check the log file at "/var/log/Xorg.0.log" for additional information. [18:01] ddxSigGiveUp: Closing log [18:01] Server terminated with error (1). Closing log file. [18:01] xinit: giving up [18:01] xinit: unable to connect to X server: No such file or directory [18:01] xinit: server error [18:01] xauth: error in locking authority file /data/.Xauthority [18:01] root@localhost:/# [18:08] Ouehy..... I'm on 2.6.39..... [18:13] Might that be what's happening? [18:20] Well, anyone know how to set up a initfs properly, and the proper cmdline? [18:23] I should probably mention that I was using http://blog.rot13.org/2012/08/x11-running-on-nook-color-without-android-stack.html as a basis for a lot of what I was doing... [18:33] ogra_: hmm, any idea why modem-manager in raring takes over /dev/ttyACM0 when I connect my nexus7 to my raring laptop? [18:34] nope, thats a qerustion for cyphermox i think, we noticed other modemmanager issues as well, might be the same bug, not sure [18:34] *question even [18:37] ah, it apparently opens the port several times and takes a long time to probe it, but eventually lets go [18:37] yeah, it sends AT commands [18:37] seems it does that for the serial gadget connection too [18:37] which can confuse getty a bit [18:38] slangasek, btw, make sure to upgrade to the latest, there were a ton of fixes this week [18:38] this was installed yesterdy [18:38] well, we didnt have images for the whole week :) [18:39] we did yesterday [18:39] oh, i missed that, only saw that todays failed [18:39] perfect then :) [18:50] * ogra_ goes for dinner [18:55] Yep. My audio device is tegra8903. Perfect sound in Ubuntu.