UbuntuIRC / 2012 /01 /11 /#ubuntu-arm.txt
niansa
Initial commit
4aa5fce
=== Jack87|Away is now known as Jack87
=== Crisco is now known as Criscolo
=== zyga_ is now known as zyga
[10:14] <ogra_> lilstevie, around ?
=== Jack87 is now known as Jack87|Away
[10:48] <brendand> hey, does the hard-float image work on the pandaboard?
[10:53] <ogra_> sure
[10:54] <brendand> does it make any difference?
[10:54] <ogra_> it fells faster, i didnt measure anything myself
[10:54] <ogra_> so thats a totally subjective impression
[10:54] <brendand> yeah, that might be an interesting experiment
[12:21] <lilstevie> ogra_: kinda, redecorating bedroom
=== jkridner__ is now known as jkridner
=== Jack87|Away is now known as Jack87
[13:30] <lilstevie> ogra_: if you are about I am here now :)
[13:31] <ogra_> lilstevie, i was wondering if you have anything newer for the transformer than 2.6.36
[13:31] * ogra_ is just playing with an ubuntu installed transformer and the kernel seems to have some issues
[13:31] <ogra_> though the config is a bit odd anyway
[13:32] <lilstevie> the config is a bit off yeah, been working on some things though
[13:32] <lilstevie> I will have something newer soon
[13:33] <ogra_> but you are on the same source ?
[13:33] <ogra_> k
[13:35] <lilstevie> well I am working on bringing it over to mainline
[13:36] <lilstevie> there is the 2.6.38 CrOS kernel if you are using u-boot though
[13:38] <ogra_> not sure they use u-boot on this device
[13:38] <ogra_> its not mine, i just got it to look at some performance issues
[13:38] <lilstevie> ah,
[13:39] <lilstevie> yeah probably stock bootloader then
[13:39] <ogra_> (and noticed a bunch of udev errors on boot)
[13:39] <lilstevie> hm
[13:39] <ogra_> looks like it doesnt have dnotify enabled and other minor things
[13:39] <lilstevie> can you grab the config from /proc/config.gz and see what the generation date is?
[13:43] <ogra_> # Linux kernel version: 2.6.36.4
[13:43] <ogra_> # Tue Jan 3 09:26:31 2012
[13:43] <ogra_> but i think they recompiled the kernel already
[13:46] <lilstevie> wow
[13:46] <lilstevie> jan 3
[13:46] <lilstevie> :p
[13:46] <lilstevie> thats not my kernel
[13:46] <lilstevie> :p
[13:47] <lilstevie> but it is built off the latest probably
[13:47] <lilstevie> there are some odd things missing from the config, I am aware of that, and I have been adding them as I have been made aware of them
[13:47] <brendand> ogra_ - any chance the hf images might not work on older pandaboards?
[13:47] <lilstevie> although, can't say I have seen any udev errors
[13:47] <ogra_> define older
[13:47] <brendand> i have an A1 and it isn't going anywhere
[13:48] <ogra_> we tested onm EA1 which is even older
[13:48] <ogra_> *on
[13:48] <brendand> okay
[13:48] <infinity> Define "not going anywhere"?
[13:49] <brendand> maybe it's just the latest image then
[13:49] <infinity> Usually that's a bad SD, or a a badly-seated card.
[13:49] <brendand> hmm
[13:49] <GrueMaster> brendand: I have run it on every HW revision board I have (EA1-A3 & 4460).
[13:49] <brendand> well, the light pattern seems a bit weird. not what i usually see
[13:49] <GrueMaster> And I have today's daily image running now on 4460.
[13:50] <GrueMaster> Precise kernel doesn't blink the led's like maverick-oneiric kernels did.
[13:51] <brendand> when i switch my monitor to the dvi input, it just switches back again. that usually means nothing being output on the dvi
[13:52] <GrueMaster> brendand: That is a different issue (and one I haven't tested).
[13:53] <GrueMaster> Try the HDMI port.
[13:53] <brendand> jeez
[13:54] <brendand> all of a sudden it's the HDMI port that needs to be used. great
[13:54] <GrueMaster> Erm, that has always been the case.
[13:54] <GrueMaster> At least that is how I have tested since Maverick.
[13:56] <brendand> oh well, no worries
[14:05] <kerute> hello
[14:07] <kerute> lilstevie_: sorry to bother you, if i want to try the 2.6.38 in OLiFe, what should I do ? i tried to flash using the 'update cros kernel' or by replacing the 2.6.36 with tne .38 but i either end up in initramfs shell telling me there is no mmcblk0p8 or stuck on 'asus' boot screen
[14:08] <lilstevie> kerute: at this point, solve the reason why it wont read asus/nvidias bastardized gpt implimentation
[14:09] <lilstevie> or use u-boot
[14:09] <kerute> ok so u-boot i have to compile it from your repo (u-boot-tegra) and flash it using OLiFe ?
[14:10] <lilstevie> yes
[14:10] <kerute> ok thanks a lot
[14:11] <kerute> for the first point i'm not sure i have enought knowledge :)
=== Jack87 is now known as Jack87|Away
[15:08] <ppisati> would kexec be a nice addition to our kernels?
[15:09] * infinity glances at #ubuntu-kernel
[15:11] <ppisati> well, if you want it i need a bug to be opened against our P kernels
[15:11] <ppisati> teh patchset is quite big, but it works
[15:11] <ppisati> so i won't backport it to O
[15:15] <infinity> ppisati: I didn't say I wanted it.
[15:16] <infinity> ppisati: Do the x86 Ubuntu kernels have kexec?
[15:16] <ppisati> yep
[15:16] <ppisati> it was broken on arm
[15:16] <ppisati> but a series of patches make it work on arm now
[15:16] <infinity> Ahh. Well, I'm a fan of our ARM kernels mirroring the x86 ones.
[15:16] <ppisati> yep
[15:16] <infinity> Just as an element of least surprise.
[15:16] <ppisati> the problem is that all the stuff is in 3.3
[15:17] <ppisati> and the backport for omap-only is about...
[15:18] <ppisati> ~17 ptches ATM
[15:18] <ppisati> the entire patches with all the archs is about...
[15:18] <ppisati> 70 patches? more or less
[15:19] <ppisati> sorry, with all the arm chips
[15:19] <ppisati> and i want to apply it to master
[15:19] <ppisati> so i'll percolate down to all the topic branches
[15:19] <ppisati> it'll
[15:19] <ppisati> (can't type today...)
[15:28] <GrueMaster> iirc, kexec was attempted multiple times on arm. seems to come up every cycle. I think there is an old blueprint on it.
[15:30] <GrueMaster> ppisati: do a google search: arm kexec site:launchpad.net
[15:30] <GrueMaster> lots of entries.
[15:39] <ppisati> GrueMaster: cool
[15:39] <ppisati> there's still an issue
[15:39] <ppisati> but it's workable
[15:40] <ppisati> aka nic is dead after kexec
[15:46] <GrueMaster> How is this considered "workable"?
[15:48] <ppisati> in the sense that it's an isolated fix
[15:48] <ppisati> some stuff is not initialised properly in smsc911x
[15:48] <ppisati> and when it boots from u-boot is ok
[15:48] <ppisati> else it craps out
=== Quintasan_ is now known as Quintasan
[19:34] <calculus> I have ubuntu 11.10 on a pandaboard, did apt-get dist-upgrade, connected a tv to the hdmi port, and video works, but I am trying to get the audio to come out on the tv as well... any ideas?
[19:34] <calculus> is there something in alsamixer that needs to be turned on/unmuted?