/srv/irclogs.ubuntu.com/2011/08/18/#ubuntu-arm.txt

prpplaguecooloney: hey buddy!03:41
cooloneyprpplague: hey, man03:41
prpplaguecooloney: what's cooking today?03:41
cooloneyprpplague: heh, acutally i don't have much rice today03:42
prpplaguecooloney: hehe03:42
prpplaguecooloney: guess you'll have to use beer instead03:43
cooloneyprpplague: good idea. need buy beer then.03:43
cooloneyprpplague: will you goto plumbers conf?03:43
prpplaguehehe03:43
prpplaguecooloney: sorry no, too many things to work on03:44
cooloneyprpplague: ic03:44
cooloneyprpplague: i think you will go to Prague for ELC-E03:44
prpplaguecooloney: OH?03:44
prpplaguecooloney: you'll be able to make it03:44
prpplague?03:44
cooloneyprpplague: oh, i won't, i will go to Orlando for our UDS03:45
prpplaguecooloney: yea i figured as much03:45
prpplaguecooloney: we'll have to meet up in the spring03:46
cooloneyprpplague: right, might be during ELC in SFO.03:46
prpplaguecooloney: sounds like a plan03:47
prpplaguehehe03:47
=== doko__ is now known as doko
ogra_steev, thats on purpose, ubuntu-core is the bare minimum to *run* a machine, it isnt configured at all, thats up to you (and it is expected that you know what you are doing)07:47
ogra_grr, mono killed all builds07:49
ogra_s/builds/image builds/07:49
=== arun__ is now known as arun
mahmohok, so I was able to install to SD/MMC on the panda where I was seeing usb-stick problems during install yesterday (new build too though)12:25
mahmohcmagina: ogra_ ^12:25
cmaginamahmoh: good to hear12:26
ogra_yes, d-i was fixed12:26
mahmohI'm going to try again to the usb-stick just to ensure it's not a build change that fixed it, if it's still broke is it worth entering a bug against a particular usb-stick  install ?12:26
ogra_(see oneiric-changes ;) )12:26
mahmohogra_: so the problem may not have been the stick after all?12:26
ogra_well, d-i was using an old kernel12:27
ogra_thats also the reason why you get questions about the kernel package usually12:27
mahmohit was failing at kernel install time but installing an old kernel shouldn't fail, no?12:27
ogra_heh, depends on the bugs in that kernel ;)12:28
mahmohhere's another one for you, the net install prompts for a host mirror and pre-populates with "mirror" vs. ports.ubuntu.com, that's a bug too right?12:29
mahmohogra_: ^?12:30
ogra_yes12:30
ogra_sounds like a preseed file issue12:30
mahmohif so, which package should that bug get assigned to?12:30
mahmohyeah12:30
mahmohogra_: ^^12:34
ogra_just go for debian-installer (and check if there arent bugs already, might not be arch specific)12:34
mahmohok, so is there nayone to ping about this or just wait?   bug 82871812:46
ubot2Launchpad bug 828718 in debian-installer "arm net install prepopulates archive with - mirror - vs. - ports.ubuntu.com" [Undecided,New] https://launchpad.net/bugs/82871812:46
mahmohogra_: ^ ;)12:49
rsalvetijanimo: seems the pvr crash is somehow related with the new xorg =\13:07
rsalvetifirst impression is that the fbdev is being loaded 2 times, and at the second time it gets loaded it crashes with a corrupted double-linked list13:08
rsalvetiinvestigating13:08
rsalvetieven with the linaro kernel I'm getting it, so it's not related with the kernel13:08
janimorsalveti, ok, for me it worked with a Xorg of last week, so maybe it will not be too hard to find the issue13:19
rsalvetijanimo: oh, good to know13:19
rsalvetimaybe I broke something :P13:19
janimorsalveti, it seems to me that setting setenv bootdelay in boot.scr on panda has no effect, as the default delay of 3 is used before actually starting the boot sequence?13:20
janimoso reading boot.scr happens after the countdown13:21
rsalvetijanimo: yeah, guess this is happening before boot.scr13:22
rsalvetiso something to change at the source code13:22
janimook.13:22
rsalvetiat the board specific header I guess13:22
janimoor have it read boot.scr but not run anything from it, just load env values13:22
janimootherwise I am not sure how the autoboot was supposed to work if not settable from a non-default script13:23
rsalvetijanimo: yeah, makes sense13:37
rsalvetijanimo: seems u-boot in general is only setting autoboot at compile time, for all boards13:40
rsalvetibut what you want kind of make sense13:41
rsalvetiwould work more like what we have at grub13:41
rsalvetithat the user can set up how many seconds before autoboot13:41
rsalvetimaybe something to improve at u-boot later on13:42
rsalvetijcrigby should know it better, maybe this was discussed already13:42
rsalvetiguess jcrigby should be up in a few minutes ;-)13:42
jcrigbyjanimo, rsalveti: I have noticed too that doing stuff in the boot script (either boot.scr or uEnv.txt) is not optimal since it happens "late".13:45
jcrigbyfwiw a patch hit the u-boot list this week that sets default boot delay to 3 seconds instead of 1013:46
rsalvetijanimo: ok, the segfault I'm getting is only happening with using the latest libs package as I removed the 99-pvr.conf file13:46
rsalvetijanimo: by trying to make xorg to find about the hw and driver without the need of a config file13:46
rsalvetiseems it works fine for natty, but for oneiric it doesn't13:46
rsalvetiwill check that later, but if you enable that ppa it'll probably work fine, if you use the xorg config file for pvr13:47
rsalvetijust tested and it worked with the lt kernel, let me reboot at the ubuntu one now13:47
rsalvetijcrigby: great, that helps already :-)13:47
rsalvetijcrigby: seems that trying to load the boot.scr to at least set up the env vars would be useful in some way13:48
rsalvetibut don't know if it'd break some other stuff :-)13:48
=== Guest67514 is now known as lag
janimorsalveti, ah indeed, I still needed pvr.conf , the linaro ppa libs (the newer ones) did not work without it13:55
=== lag is now known as Guest3769
rsalvetijanimo: yup, seems to work with the ubuntu kernel13:58
rsalvetiwith the fixes I pointed at the email13:58
rsalvetijanimo: so, if you want to test it now, just grab the lt-ti kernel and the dkms and the libs packages available at the overlay13:59
janimorsalveti, great hopefully ppisati will upload them in the next round13:59
rsalvetiall available for oneiric13:59
rsalvetithe dkms should be in shape to push to ubuntu already13:59
janimorsalveti, I already have it running with the lt kernel, but needed the ubuntu one for easier install, and to also be able to run perf13:59
rsalvetibut the lib package still needs some love13:59
rsalvetijanimo: oh yeah, true14:00
rsalvetican give you the packages later if you need14:00
rsalvetimeeting time14:00
janimook14:00
=== zyga is now known as zyga-afk
=== davidm` is now known as davidm
austeregrim-ghi'm surprised there isn't more action for ubuntu on arm devices17:54
infinityThere's a fair bit.17:54
austeregrim-ghI'm interested17:54
austeregrim-ghI've got a toshiba thrive, I've got some interest in trying to put ubuntu on it...17:55
austeregrim-ghbut the ubuntu pages lack alot of basic info... for arm.17:55
infinityWell, every new device requires some hacking and fiddling to make go.17:56
infinityAnd we don't tend to focus much on tablets, except for hobby hacking, since Ubuntu isn't (yet) a particularly useful tablet OS.17:56
austeregrim-ghwe've got an unlocked boot loader, and have rooted the android os on it... just looking to go one step further17:56
austeregrim-ghi agree ubuntu isn't ready for tablets yet... but there's a usb port on it, that I can pop a keyboard into17:57
austeregrim-ghim sure as tablets increase in userbase, ubuntu will eventually follow into ... but I'll admit, I'm new to it... but I understand alot. :-P17:59
infinityIf it's based on the nvidia tegra (I'm guessing it might be, but I can't find proper tech specs), the work we've done for the Toshiba ac100 netbooks might directly translate over.18:01
austeregrim-ghtegra218:01
infinityYeah, the ac100 is tegra2 as well.18:02
infinityAnd likely very similar under the hood, though I'd make no promises.18:02
austeregrim-ghthats what I was thinking.. but the ac100's page (via ubuntu's wiki) is old and hasn't been updated in a while18:02
infinityWe'll refresh some of those docs once our ac100 images are shipping (which is soon).18:02
infinityIf we can turn the ac100 stuff into a more generic toshiba-tegra image or something, that would be shiny.  But I've never actually even touched a Thrive, so...18:03
austeregrim-ghwould there be a posibility of getting a test image?18:04
infinityThat would require our tests images to be useful.  Give it a day or two, though, ogra's hard at work on that.18:05
austeregrim-ghno worries, it was more of a put my toes in the water to see what fish bite..18:05
austeregrim-ghalso if theres anything that I can get from the thrive to assist, information wise. I can probably get it.. whatever linux reports from android.18:08
austeregrim-ghfor general comparison, about hardware18:08
infinityCurrently, I think we have too much on our plate right now, unless a community member steps up (you?) to learn how it all works and make it go. :P18:09
infinityBut I'm perfectly willing to change my position on that when I'm less busy. ;)18:10
infinityI wouldn't be at all shocked to discover that the thrive is just the ac100 without a keyboard, though.  The ac100 always felt more like a proof-of-concept product than a finished deal.18:10
austeregrim-ghI've been on the forefront of rooting the thrive in android, working with the tabletroms.com folks...18:11
austeregrim-ghso I would see me taking to ubuntu (or linux in general) as my next plausible step18:11
infinity*nod*18:13
infinityWell, I'd recommend talking to ogra about his ac100 stuff as a start.18:13
infinityBut, like I said, giving him a few days to deliver it might help. ;)18:13
austeregrim-ghunderstood18:13
austeregrim-ghbtw, Thrive model = AT105 (does not surprise me if your theory is at all correct)18:16
=== Amaranthus is now known as Amaranth

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!