2012/12/08/#ubuntu-arm.txt

=== XorA is now known as XorA|gone
[00:20] <achiang> vanhoof: all went well, going through oem-config now
[00:23] <vanhoof> cool
[00:24] <vanhoof> i got through it no probs a few times
[00:24] <vanhoof> you should be all set
[00:25] <vanhoof> achiang: screen /dev/ttyACM0 115200 ... once you're done
[00:25]  * vanhoof is in heaven
[00:26] <achiang> what's that, debug console?
[00:26] <vanhoof> yup
[00:26] <vanhoof> usb/serial
[00:26] <vanhoof> full console even
[00:27] <achiang> whee, i'm on R
[00:27] <achiang> machine name is 'mfer'
[00:27] <achiang> ;)
[00:27] <vanhoof> lol
[00:27] <achiang> what's new and shiny for me to play with?
[00:28] <vanhoof> plymouth, usb-serial, an actual panel, oem-config? :)
[00:28] <achiang> ah, the answer is "nothing" because i hit the button1 stuck issue
[00:28] <achiang> feh
[00:29] <vanhoof> you're landscape?
[00:29] <achiang> yeah
[00:29]  * vanhoof must have small fingers
[00:29] <vanhoof> i dont hit that often :)
[00:30] <vanhoof> achiang: if you get bored play w/ sound
[00:31] <vanhoof> i had some troubling noises today playing radio w/ rb
[00:31] <vanhoof> not sure if its my own breakage
[00:33] <vanhoof> achiang: but now you have dailys to flash ;)
[00:33] <achiang> from here on out, it's dist-upgrade every day, baby
[02:14]  * sfeole reads scrollback
[02:16] <sfeole> achiang, vanhoof , infinity : Yea guys, I went on a wiki updating rampage today. sort of sloppy but got the job done. Everything should be up there, updating instructions / new features / some bug updates etc...
[02:16]  * sfeole goes back into his cave
[02:20] <sfeole> also i fixed the 2nd heading for installed 13.04 on the nexus7, looks like my draft never saved.
[02:21] <sfeole> it now says "Manually Installing... blah blah"
=== albert is now known as Guest32322
=== Guest32322 is now known as Phryq
=== Phryq is now known as Hyde
=== Hyde is now known as Phryq
=== doko_ is now known as doko
[10:37] <wookey_> infinity: I see you fixed tcl crossing. What did you do about multiarching the tclconfig.sh path?
[10:38] <infinity> wookey_: I didn't touch tcl, just tk.
[10:38] <wookey_> Or did you just fix tcl bulding itself, not anything building against it?
[10:38] <infinity> wookey_: Though, tk builds against tcl, so maybe someone already fixed the thing you're talking about.
[10:38] <gurgalof> aww, I want to run 13.04 on the nexus, but I need Bluetooth...
[10:39] <wookey_> OK. That would be good if true - I've beenwondering what to do about it for months :-)
[10:39] <wookey_> I see you had a go at apr too - is there a bug for that? It was an abomination when we looked at it a couple of years back
[10:39] <wookey_> somebody tried to add cross-support but did it in a completely broken way
[10:40] <infinity> wookey_: I'm holding off on filing bugs (or committing directly) to apr/apr-util/apache2 until I'm happy with my "solutions" (and I use that term loosely).
[10:40] <infinity> wookey_: apr and apache2 work in Ubuntu now (apr-util needs some twiddling with libraries it links to), but neither is pretty, so I'll likely revisit.
[10:40] <wookey_> I have a really long mail for the apache people that we weren't allowed to send a couple of yrears back, asking what tey thought they were _trying_ to do
[10:41] <wookey_> so we could work out how to fix it properly
[10:41] <infinity> http://launchpadlibrarian.net/124947530/apr_1.4.6-3_1.4.6-3ubuntu1.diff.gz
[10:41] <infinity> ^-- The ugly is just a bunch of autoconf cache preseeding.
[10:41] <infinity> But there's got to be a less hideous way.
[10:42] <infinity> As for apache2 itself, it's all fixed upstream, I just had to backport patches.
[10:42] <wookey_> Ah cool
[10:42] <infinity> (And fix debian/rules to be less retarded)
[10:43] <infinity> So, I'll probably commit the apache2 stuff to Debian (or file a bug and make sf do it), but the apr/apr-utils things need some thought, I think.
[10:43] <infinity> And, really, apache is useless without apr, so...
[12:04] <raster> mooo
[12:11]  * Tassadar hides the gnomes
[12:12] <raster> i'm not here to gorge on gnomes today
[12:22] <raster> yargh
[12:22] <raster> what is the magic to get sound out of this little beastie
[12:24] <Tassadar> You mean n7?
[12:28] <raster> yeah
[12:28] <raster> i have gotten it to hiss at me
[12:28] <raster> and thats about it
[12:28] <raster> yes - i know - suspend and resume
[12:28] <raster> doesnt do squat
[12:31] <raster> i've been swizzling everythng randomly in alsamixer for a while to try see what low-level knob makes it work
=== yofel_ is now known as yofel
[13:22] <marvin24> nvidia just published tegra boot process description: ftp://download.nvidia.com/tegra-public-appnotes/tegra-boot-flow.html
[13:22] <marvin24> (for those who didn't got it yet)
[13:23] <raster> nv is getting better at this
[13:24] <RaYmAn> sadly, it's not much use on most 'real' devices (given they are locked mostly down), but it's a nice move either way
[13:26] <raster> not much nv can do there
[13:26] <raster> thats oems
[13:26] <RaYmAn> indeed
[13:26]  * RaYmAn looks at google
[13:26] <raster> i have a working nexus7
[13:26] <raster> minus sound
[13:26] <raster> i cant divine how to make it peep
[13:28] <RaYmAn> raster: it's just a pity e.g. google makes the n7 unlockable and then locks down the bootloader still. :/
[13:28] <raster> well to some extent thats to avoid bricking
[13:29] <raster> oems are paranoid about having hoards of users send in bricked devices under warranty
[13:29] <RaYmAn> tegra is unbrickable if you have access to APX mode (e.g. tegrarcm)
[13:29] <marvin24> well, there is still the rom bootloader, which cannot be bricked by nature
[13:30] <marvin24> on the other hand, who cares if the locked bl can load an unlocked one?
[13:30] <raster> if a user cant boot it to a state of a ui withotu some cruptic pc tools
[13:30] <raster> then its a brick
[13:30] <raster> form an oem pov
[13:30] <raster> fastboot is not an answer in their eyes
[13:30] <marvin24> write a simple tools then
[13:30] <raster> it has to be unbrickable without tools
[13:30] <raster> ie withotu needint to plug it into a pc and "use tools"
[13:31] <raster> same with plugging in special mucro-sd cards with rescue images etc.
[13:31] <RaYmAn> but how is that different from now? users could easily perceive it as bricked even if they have fastboot.
[13:31] <RaYmAn> and that is possible now
[13:32] <raster> dont they stil lock the bootloader down
[13:32] <raster> but the rest u can play with as u like
[13:32] <lilstevie> flash a non working kernel to both boot and recovery
[13:32] <raster> so long as u use the bootloader to boot it
[13:32] <raster> ala the ubuntu nexus77 images
[13:32] <RaYmAn> lilstevie: isn't that fixed on n7?
[13:32] <lilstevie> and it will be in the state you just called a brick
[13:32] <lilstevie> RaYmAn, yeah, n7 also uses scratch rather than msc for bootloader messages
[13:33] <RaYmAn> hm
[13:33] <lilstevie> RaYmAn, but I am referring to "<raster> fastboot is not an answer in their eyes"
[13:33] <RaYmAn> did you test if that works on tf201?
[13:33] <lilstevie> RaYmAn, I did not
[13:33] <lilstevie> RaYmAn, it is in the shutdown kernel code
[13:33] <raster> lilstevie: i work for an oem
[13:34] <raster> if ti requires anything more than absolute basic knowledge to fix
[13:34] <lilstevie> raster, still doesn't change the fact, even unlocking a device will cause the definition you just called "brick"
[13:34] <raster> (press a reset button for example) it's undesired and everything is done to avoid it
[13:34] <raster> problem is a reset button is also aloiability in normal use
[13:35] <raster> lilstevie: yes - its a risk google has pushed asus to do
[13:35] <lilstevie> raster, that asus also offer on their entire tegra 3 range
[13:35] <raster> prbably so google can use the n7 itself for hacking as well as allow others to do custom android roms and otherwise spread android around in general
[13:36] <RaYmAn> raster: tbh, if oems wanted to get around that, it *should* be possible to hook up boot strap pins in such a way that it can boot a special (signed) image incl. bootloader from sdcard. Then a win app that writes the sdcard
[13:37] <RaYmAn> also, they could just charge to fix bricks that could be fixed with e.g. nvflash if it was available
[13:38] <raster> RaYmAn: you expect them to figure out all for a tiny percentage of their customerbase
[13:38] <raster> when they can just continue with the old solution they have used for many years
[13:41] <RaYmAn> nah, I've lost faith in them bothering to do such stuff long ago
[13:41] <raster> so have most
[13:41] <raster> :)
[13:41] <RaYmAn> though, it would probably cut down on software repair costs =P
[13:42] <RaYmAn> anyways, your argument was that e.g. nvflash was too complex for normal users (I disagree) - there's plenty of ways they could make it simpler
[13:43] <raster> u are free to sit in meetings with some taiwanese, korea, etc. ppl and make your argument
[13:43] <raster> :)
[13:44] <RaYmAn> I don't expect an invitation any time soon
[13:44] <RaYmAn> ;)
[13:44] <raster> i've been there, done that... many times over the years :)
[13:45] <raster> and i get the bonus of the "listen to me" badge onmy chest
[13:45] <raster> even then its insanely hard
[13:45] <lilstevie> although they may have to soon if the sony tool thing catches on
[13:45] <raster> the things that make themj move.. is if someone else moves
[13:46] <raster> and if it means market share gain, profits or somethnig
[13:46] <raster> that gets attention
[13:46] <RaYmAn> yeah, that's a pretty decent way of doing it - nice and semi-userfriendly tool that installs drivers and reflashes devices. It would be fairly easy for e.g. asus or google to do something similar with a custom nvflash for n7
[13:51] <RaYmAn> raster: incidentally, it's amusing they spend so many resources on locking down things. You'd think it would be less resources to just make it open :)
[13:51] <gurgalof_mobil> i flashed raring using the tool in ubuntu,  but i only get a 8GB partition on my 32GB nexus7
[13:53] <raster> RaYmAn: the locking down has already been done and it was a product requirement from the product specs team to begin with
[13:53] <raster> or something like that
[13:55] <gurgalof_mobil> and i have a black bar under the notification bar in raring
[13:57] <RaYmAn> raster: it's a choice e.g. google makes. Of course it has already been done, but it's still a choice they made
[13:57] <RaYmAn> and then you have stuff like the arm chromebook which is completely open/unlockable.
[13:58] <raster> a choice made by a product specs dept that doesnt knowq nor care about hackers/development
[13:58] <raster> as they are not
[13:58] <raster> they expect the o9nly development happens inside this engineering dept they have
[13:58] <raster> and nowhere else
[13:58] <RaYmAn> I somehow doubt google is that naive
[13:58] <raster> people wanting to hack on their PRODUCTs is a bizarre foreign concept to them
[13:58] <raster> raster: did i not mention oems above?
[13:58] <raster> not google?
[13:58] <raster> :)
[13:59] <RaYmAn> google is essentially an oem too :)
[13:59] <raster> well since they bougt moto - sure
[14:00] <raster> tho within large companies its not so simple
[14:00] <raster> its a lot of hops from left to right hand
[14:00] <raster> :)
[14:02] <gurgalof_mobil> is everyone sleeping?
[14:06] <smartboyhw> gurgalof_mobil, NO
[14:06] <gurgalof_mobil> :D
[14:07] <gurgalof_mobil> feels like no one notices me, but ill ask again later...
[14:07] <gurgalof_mobil> hmm i didnt even write a question, well well
[14:11] <ogra_> gurgalof_mobil,  the current ikmmge only uses a 6G partition, we're working on getting it to work for bigger devices too, the broken panel shadow is known and reported as a bug already
[14:12] <ogra_> *image
[14:12] <gurgalof_mobil> okay,  thanks
[14:13] <gurgalof_mobil> i'll revert to quantal for the moment then
[14:14] <ogra_> raster, do you use pulse in youir setup ? iirc fixing the sound was a combo of alsa UCM profiles and pulse profile changes
[14:14] <raster> yup
[14:14] <raster> pulse is there
[14:14] <gurgalof_mobil> oh and the slideshow in raring says 12.10
[14:15] <infinity> gurgalof_mobil: It usually does for the first few months.
[14:15] <raster> i use it on all y setups
[14:15] <raster> hell
[14:15] <raster> my mier eeven detects pulse and is happy
[14:15] <infinity> gurgalof_mobil: It'll say 13.04 before release. :P
[14:15] <ogra_> raster, also note that the master is muted by default ... only suspend/resume isnt enough
[14:15] <raster> yup
[14:15] <raster> i checked
[14:15] <raster> :)
[14:15] <gurgalof_mobil> :P hopefully
[14:15] <raster> pulse is unmuted
[14:15] <raster> rhythmobx doest play
[14:15] <ogra_> hmmm, then i dont know
[14:16] <raster> aplay doesnt play (bypassing pulse going right to alsa)
[14:16] <ogra_> it works flawless under unity for me
[14:16] <raster> weird
[14:16] <ogra_> boot, suspend/resume, hit vol-up omce (easier than unmuting) always gets me sound
[14:16] <raster> i didnt get to test audio until i had the rest settled
[14:17] <infinity> ogra_: That's an odd definition of "flawless". ;)
[14:17] <ogra_> heh
[14:17] <ogra_> well, "as described" :)
[14:18] <infinity> Thou shalt swing the mutilated chicken thrice above thine head, whilst chanting "ALSA, our lord, hear my prayers."
[14:18] <ogra_> *giggle*
[14:18] <raster> ogra_: yaaargh
[14:19] <raster> any info as to.. what this magic pulse + alsa magic was?
[14:19] <ogra_> ask diwic during the week
[14:19] <raster> ok
[14:19] <raster> either way
[14:19] <raster> suspend/resume is all happy now
[14:19] <ogra_> he is our sound guru and added the fixes
[14:19] <raster> it was wifi simply not being shut off before suspend
[14:19] <ogra_> ah
[14:19] <raster> u have to rfkill it or it gens wake interrupts
[14:19] <raster> :)
[14:20] <raster> my screen is all nicely portraited now
[14:20] <ogra_> yeah,, the pm-utils scripts care for that iirc
[14:20] <ogra_> portrait with working input ?
[14:20] <raster> i tried the 6.1-7 xorg thing and it fixes some of the funkiness of the touch when in portrait
[14:20] <raster> yeah
[14:20] <ogra_> oh, sweet
[14:20] <gurgalof_mobil> in quantal BT stays on while suspending as moving bt-mouse unsuspends it
[14:20] <raster> tho the pointer jumps around like its obsessed
[14:20] <raster> inptu works tho
[14:21] <ogra_> you have a pointer ?
[14:21] <ogra_> unity/compiz hides it if there isnt an actual mouse
[14:21] <ogra_> (that hides such issues)
[14:21] <raster> yeah
[14:22] <raster> i haveth one
[14:22] <raster> i can hide it if i want
[14:22] <raster> i just like seeing it
[14:22] <raster> its kind alike a nice debug thing
[14:22] <raster> :)
[14:22] <raster> i'm down to 180m of mem usage
[14:22] <ogra_> heh, yeah
[14:23] <raster> for some reaosn my mystery missing ram... comes back after some time
[14:23] <ogra_> i think there are some oddities with the kernel
[14:23] <raster> definitely
[14:23] <raster> like on boot - 420m ram gone
[14:23] <raster> i compile a bunch of stuff
[14:23] <raster> suspend/resume a few times
[14:23] <ogra_> it also shows some strange load values that i cant really confirm
[14:23] <raster> restart wm etc.
[14:23] <raster> and it goes DOWN.. to about 150-200
[14:24] <raster> oh yeah
[14:24] <raster> and DONT use powertop
[14:24] <gurgalof_mobil> WAT
[14:24] <raster> it somehow turns off the cores
[14:24] <raster> and makes u have only 1 core
[14:24] <ogra_> file a bug !
[14:24] <raster> everything nicely slows down to a crawl
[14:24] <raster> :)
[14:24] <ogra_> that should definitely be fixed
[14:24] <raster> i'll kick arjan next i get the chance
[14:24] <ogra_> great
[14:25] <raster> :)
[14:25] <raster> rightr now.. i dont know what it did
[14:25] <lilstevie> you should confirm that is powertop that did it
[14:25] <raster> i was compiling at the time
[14:25] <raster> lilstevie: certain - i was runing htop at the time
[14:25] <raster> and i rant it
[14:25] <lilstevie> I've come across that exact issue as a kernel bug on the tf201
[14:25] <raster> and then immediately everything jumpt to a single core
[14:25] <raster> it has never happend otherwise
[14:25] <raster> so far anyway
[14:26] <lilstevie> raster, lather rinse repeat
[14:26] <raster> busy lathering a lot of other things
[14:26] <raster> :)
[14:27] <lilstevie> I've never been able to trace the exact moment before, but I have never run powertop
[14:27] <raster> they were too suspiciously timed together
[14:27] <raster> i dont know if its pt's bug tho
[14:27] <raster> it may simply innocently do something that happens to trigger the kernel bug
[14:28] <raster> these things happen :) i'm rather used to them
[14:33] <lilstevie> someone else should probably try the above on a n7, cause if it is the case, it is n7 specific and doesn't affect T3 in general
[14:33]  * lilstevie just tested
[14:35] <raster> i only have the n7
[14:35] <raster> so...  :)
[14:37] <lilstevie> well then :p reproduce
[14:38] <raster> \would love to
[14:38] <raster> but i'm busy with.. other fun
[14:38] <raster> like nixing backlight artifacts
[15:22] <gurgalof> where do i find the 12.10 nexus7 image?
[15:24] <gurgalof> or can i resize the partiton to 32GB? instead of 6?
[15:25] <Tassadar> http://hwe.ubuntu.com/uds-r/nexus7/
[15:27] <krawi> hi all
[15:27] <krawi> can someone give me a hint for alsa playback on pandaboard with ubuntu 12.04 server?
[15:27] <krawi> in syslog: OMAP ABE Media: asoc: OMAP ABE Media no valid playback route from source to sink
[15:28] <krawi> in mplayer [AO_ALSA] alsa-lib: pcm_hw.c:1293:(snd_pcm_hw_open) open '/dev/snd/pcmC0D0p' failed (-22): Invalid argument
[15:28] <krawi> and nothing is played
[16:04] <krawi> one step foward
[16:04] <krawi> now it is: omap_hwmod: aess: _wait_target_disable failed
=== Snark_ is now known as Snark
[17:35] <krawi> got it
[18:30] <gurgalof> can i resize the partition on my nexus7 with gparted?
[18:39] <gurgalof> or is there some other way?
=== Ursinha is now known as Ursinha-afk
=== Ursinha-afk is now known as Ursinha
[19:14] <gurgalof> why do we have swap enabled in 13.04 on nexus7?
[19:16] <gurgalof> I think everyone is sleeping again...
=== awafaa_ is now known as awafaa
=== Quintasan_ is now known as Quintasan