**** BEGIN LOGGING AT Mon Dec 19 02:59:58 2011
Dec 19 03:29:03 <twb> lilstevie: does your load average always exceed 1?
Dec 19 03:29:10 <twb> Mind does, it seems weird
Dec 19 03:29:26 <twb> There's nothing in the process tree to account for it
Dec 19 03:31:38 <lilstevie> no
Dec 19 03:31:41 <lilstevie> mine does not
Dec 19 03:31:58 <lilstevie> mine rarely exceeds 0.5
Dec 19 03:38:02 <twb> Grmph
Dec 19 03:38:17 <twb> Min is usually 1.02 to 1.08, I suspect the 1 is that stupid resize2fs or something
Dec 19 03:39:23 <lilstevie> lolwut
Dec 19 03:39:51 <lilstevie> you know that program quits out as soon as it has finished its job right
Dec 19 03:40:07 <twb> Then why do I get printks about kinteractiveup hanging?
Dec 19 03:41:36 <twb> { dmesg | grep -2 kinteractiveup; ps u 61; }| pastebinit ==> http://paste.debian.net/149581/
Dec 19 03:47:23 <lilstevie> I get them too, but what makes you think that is resize2fs causing them
Dec 19 03:47:37 <lilstevie> fwiw I even get them when running a stock asus android rom
Dec 19 03:50:27 <twb> lilstevie: uh, because kinteractiveup is the kernel side of resize2fs
Dec 19 03:50:40 <twb> At least that's what google indicated, I haven't RTFS
Dec 19 03:51:14 <twb> If you're getting them from android then I'll RTFS now since I am likely wrong
Dec 19 03:51:16 <lilstevie> ok, but that has been a problem since long before resize2fs invoked
Dec 19 03:52:11 <twb> OK, then ignore me :-)
Dec 19 03:52:20 <twb> I'd still like to know where the load is from, tho...
Dec 19 03:54:09 <lilstevie> also I googled kinteractiveup to see where you got this information from and nothing of the sort showed up
Dec 19 03:55:05 <twb> I may have asked it leading questions like googling for "resize2fs AND kinteractiveup" :P
Dec 19 03:55:23 <twb> I mean it's an interactive fs resize, it needs kernel help, it seemed reasonable to me... :-(
Dec 19 03:56:13 <lilstevie> and the only results on google are you mentioning resize2fs and kinteractiveup in this room
Dec 19 03:56:56 <twb> Grmph
Dec 19 03:57:24 <twb> https://duckduckgo.com/lite?q=kinteractiveup%20resize2fs has a few
Dec 19 03:57:31 * lilstevie starts searching the kernel source tree for kinteractiveup
Dec 19 03:58:25 <lilstevie> twb: none of those really mention the two together, well except the first one and I would say that is unrelated :)
Dec 19 03:58:44 <twb> Mea culpa
Dec 19 03:58:49 <lilstevie> and infact I probably can kill the kinteractiveup thing quite easily
Dec 19 03:58:54 <lilstevie> it is the cpu govener
Dec 19 03:59:16 <twb> Oh!
Dec 19 03:59:22 <lilstevie> the symbol "kinteractiveup" belongs to /drivers/cpufreq/cpufreq_interactive.c
Dec 19 03:59:34 <lilstevie> a simple grep of the kernel source gave that :)
Dec 19 03:59:51 <lilstevie> but see I have it marked as low priority so I never looked
Dec 19 03:59:55 <twb> Can't find it in the 3.2 source, which is the only one I have handy
Dec 19 04:00:00 <lilstevie> now I know what it is though, dead
Dec 19 04:00:09 <twb> Yay
Dec 19 04:00:13 <lilstevie> well that is funny, cause the source I am using is based from the 3.2
Dec 19 04:00:13 <twb> I accidentally helped
Dec 19 04:00:27 <twb> I'm probably Doing It WrongTM
Dec 19 04:00:53 <twb> I don't even have a drivers/cpufreq/cpufreq_interactive.c ...
Dec 19 04:00:59 <lilstevie> lol
Dec 19 04:01:05 <twb> Does git blame say where it's from?
Dec 19 04:01:11 <twb> I mean I just pulled thirty seconds ago
Dec 19 04:01:21 <twb> v3.2-rc6-38-g390f998
Dec 19 04:05:59 <lilstevie> that does not sound like an asus pull just fyi
Dec 19 04:06:38 <lilstevie> but it is an androidification patched into the tree
Dec 19 04:06:57 <twb> Stupid google
Dec 19 04:16:33 <lilstevie> heh
Dec 19 04:16:56 <twb> Well, at least I'm not trying to run Debian/kFreeBSD on this thing
Dec 19 04:18:41 <lilstevie> lol
Dec 19 07:05:52 <asiekierka> hello
Dec 19 07:07:55 <asiekierka> lilstevie - i'm using one of your ubuntu ARM images for the asus eee pad transformer... how do I get GLES up and running
Dec 19 07:08:03 <asiekierka> or is it already up and running and i'm just doing something wrong
Dec 19 07:08:34 <lilstevie> asiekierka: you don't :)
Dec 19 07:08:45 <lilstevie> it is not a supported configuration at this point in time
Dec 19 07:09:18 <lilstevie> well, you can; but you need to know how to compile u-boot
Dec 19 07:09:57 <asiekierka> well to be honest the current dualboot situation i use is linux normal, android recovery
Dec 19 07:10:21 <lilstevie> yes well you won't have that
Dec 19 07:10:56 <asiekierka> another bootloader? is that really needed, is there no kernel with the drivers or even just the drivers themselves?
Dec 19 07:11:04 <asiekierka> does the bootloader unlock something or is there some magic going on i don't know about
Dec 19 07:13:00 <asiekierka> if that's too complex to explain then just say it's magic
Dec 19 07:14:43 <lilstevie> no its quite simple really
Dec 19 07:14:51 <lilstevie> the kernel does not work properly on the asus bootloader
Dec 19 07:15:16 <asiekierka> :o
Dec 19 07:15:19 <asiekierka> that's new
Dec 19 07:15:33 <lilstevie> nono the 2.6.38 kernel
Dec 19 07:15:40 <asiekierka> i mean, this information
Dec 19 07:15:45 <asiekierka> the fact the kernel doesn't work properly is new
Dec 19 07:15:51 <lilstevie> oh :p I have said it a few times
Dec 19 07:16:04 <lilstevie> the muromec ported kernel is based off working with u-boot
Dec 19 07:16:08 <asiekierka> i just found it
Dec 19 07:16:19 <lilstevie> and it has a hell of a time working on the asus bootloader
Dec 19 07:16:28 <asiekierka> i found an uboot binary by muromec
Dec 19 07:16:35 <asiekierka> as i'm scared of bricking my phone to death
Dec 19 07:16:38 <asiekierka> err, tablet
Dec 19 07:16:48 <lilstevie> clocking is all whacked, the emmc driver doesn't work around the asus custom layout
Dec 19 07:16:49 <asiekierka> i'm scared of bricking my tablet forever, is it safe enough?
Dec 19 07:16:54 <lilstevie> also you cannot brick :)
Dec 19 07:17:01 <lilstevie> APX will always be there to save your ass
Dec 19 07:17:11 <lilstevie> but the muromec binary will not work for you :)
Dec 19 07:17:19 <asiekierka> oh. so i have to compile it myself.
Dec 19 07:17:25 <asiekierka> any instructions or do i have to use magic
Dec 19 07:17:30 <asiekierka> and figure it out myself
Dec 19 07:17:34 <lilstevie> I customised the bootloader a little more to be a bit more normal
Dec 19 07:17:50 <lilstevie> grab it from my git
Dec 19 07:17:53 <asiekierka> alright
Dec 19 07:18:01 <asiekierka> as i have to go soon, when i grab the uboot, what's the next step
Dec 19 07:18:03 <lilstevie> but you flash u-boot you say goodbye to booting android
Dec 19 07:18:08 <asiekierka> ...dammit.
Dec 19 07:18:24 <lilstevie> you *can* boot 2.6.38 with a microsd
Dec 19 07:18:31 <Pratik_> Hello All, I am not able to install flash player on my OMAP device running ubuntu 10.10
Dec 19 07:18:32 <lilstevie> just YMMV
Dec 19 07:18:52 <asiekierka> lilstevie well i probably use ubuntu more than android by now so...
Dec 19 07:18:57 <asiekierka> also do i need a different kernel
Dec 19 07:19:02 <asiekierka> i used a hacked one from the topic with the hacked touchpad support
Dec 19 07:19:09 <asiekierka> that was a few months ago
Dec 19 07:19:12 <lilstevie> oh wow, no need to use that one :p
Dec 19 07:19:22 <asiekierka> before there was any kind of official touchpad support
Dec 19 07:19:23 <lilstevie> the newer kit includes that kind of stuff
Dec 19 07:19:26 <asiekierka> oh good
Dec 19 07:19:33 <asiekierka> so i grab everything from your git, correct?
Dec 19 07:19:34 <lilstevie> and wifi working with network manager
Dec 19 07:19:48 <lilstevie> I would say, wait a few days :)
Dec 19 07:19:53 <asiekierka> why so?
Dec 19 07:19:56 <asiekierka> are you doing something special
Dec 19 07:20:07 <lilstevie> enough people have been asking about u-boot so I am going to enable it in the next release
Dec 19 07:20:09 <asiekierka> okay
Dec 19 07:20:20 <asiekierka> i will wait a few days
Dec 19 07:20:21 <asiekierka> thanks, bye
Dec 19 07:20:33 <lilstevie> np
Dec 19 07:20:49 <Pratik_> I am not able to install flash player on my OMAP device running ubuntu 10.10, Could you help me?
Dec 19 07:21:58 <lilstevie> flash is not supported on an arm target
Dec 19 07:23:26 <twb> Because adobe are foss-hostile douchebags
Dec 19 07:23:38 <twb> You don't need it anyway, just tell youtube to use HTML5
Dec 19 07:24:32 <twb> I guess gnu thingoflash might work on arm, inasmuch as it works anywhere...
Dec 19 07:24:52 <micahg> Pratik_: try gnash or lightspark
Dec 19 07:25:28 <twb> Yeah, gnash was what I was thinking of
Dec 19 07:25:38 <Pratik_> ok will it as adobe flash player 10?
Dec 19 07:25:47 * micahg plans on SRUing gnash when they make another stable release
Dec 19 07:26:14 <Pratik_> thank you
Dec 19 07:26:59 <micahg> Pratik_: well, the non-armel one in natty does :-/, idk about the one in maverick
Dec 19 07:27:58 <Pratik_> ok
Dec 19 07:28:03 <Pratik_> i will try this
Dec 19 07:29:36 <Pratik_> actually I want to take camera input flash.
Dec 19 07:31:16 <twb> WTF is "camera input flash" ?
Dec 19 07:31:40 <Pratik_> web cam input to running flash in web browser
Dec 19 07:32:00 <Pratik_> to stream it to rtmp server like red5 and wowza
Dec 19 07:32:38 <twb> That's horrible
Dec 19 07:33:10 <lilstevie> lol
Dec 19 07:34:37 <twb> Which web monkey thought it was a good idea to give the RPC sandbox access to raw hardware
Dec 19 07:35:08 <twb> (Don't argue; flash is just as much an RPC as PostScript.)
Dec 19 09:57:52 <doko> lool, ogra_, janimo: do we know all the packages, where neon code is built explicitly?
Dec 19 09:58:26 <janimo> doko, you mean where there's neon assembly code activated based on hwcaps at runtime?
Dec 19 09:58:36 <janimo> I am not sure we have a list
Dec 19 10:00:08 <doko> janimo, either this or where a neon variant is built explicitly
Dec 19 10:00:41 <janimo> doko, I hope we do not have any hardcoded neon builds, if we do it is a bug we overlooked
Dec 19 10:01:15 <janimo> most cases should be optional runtime toggles and I know qt has that of the main packages we touched often
Dec 19 10:01:43 <janimo> also libav
Dec 19 10:02:02 <doko> janimo, looks like we need to identify all these to get them rebuilt after a toolchain fix
Dec 19 10:02:16 <janimo> doko, is there a neon related bug?
Dec 19 10:03:07 <doko> yes
Dec 19 11:10:14 <ogra_> doko, i only remember pixman ... probably cairo too
Dec 19 11:11:32 <ogra_> doko, we definitely *shouldnt* have any packages that are neon only
Dec 19 12:03:45 <ogra_> janimo, https://bugs.launchpad.net/gcc-linaro/+bug/903951
Dec 19 12:03:46 <ubot2> Launchpad bug 903951 in linaro-android "The kernel on staging-panda does not boot with the 11.12 released toolchain " [Critical,Confirmed]
Dec 19 12:04:28 <janimo> ogra_, hmm, sounds like our ac100 issue?
Dec 19 12:04:37 <ogra_> yep
Dec 19 12:04:38 * janimo checks
Dec 19 12:04:45 <ogra_> and we should see it on other boards soon too
Dec 19 12:04:58 <janimo> that woudl rule out old kernel tree bugs, as panda has a much newer kernel
Dec 19 12:05:20 <ogra_> gar, the nvidia driver just killed all font rendering
Dec 19 12:05:24 * ogra_ restarts X
Dec 19 12:06:04 <janimo> ogra_, the new tegra l4t driver?
Dec 19 12:08:13 <lilstevie> I hate it when that happens
Dec 19 12:08:29 <lilstevie> janimo: I would guess alpha :p it was very common for me
Dec 19 12:09:04 <janimo> I hope the beta is better :)
Dec 19 12:09:30 <ogra_> janimo, right, the one from precise
Dec 19 12:09:36 * ogra_ is on vacation, so its ac100 time :)
Dec 19 12:09:49 <lilstevie> heh ouch
Dec 19 12:09:58 <lilstevie> so beta doesn't fix that issue then I take it
Dec 19 12:10:26 <ogra_> well
Dec 19 12:10:32 <ogra_> its oneiric i test on
Dec 19 12:10:46 <ogra_> we'll see how it behaves on precise
Dec 19 12:10:54 <lilstevie> it happened quite frequently with L4T alpha and oneiric on the tf101
Dec 19 12:11:01 <lilstevie> maybe 2/3 times a day
Dec 19 12:11:31 <ogra_> seems for oneiric the alternatives are wrong
Dec 19 12:11:45 <ogra_> also console switching is still broken
Dec 19 12:11:56 <ogra_> and suspend doesnt bring up the backlight on resume
Dec 19 12:12:09 <ogra_> beyond that it works great#
Dec 19 12:12:29 <ogra_> hdmi plugging is immediately recognized and gets me full 1080p
Dec 19 12:12:41 <lilstevie> thats nice
Dec 19 12:12:49 <ogra_> the overall load of the system as well as ram usage has dropped significantly
Dec 19 12:12:56 <lilstevie> my hdmi plug wouldn't recognise with alpha
Dec 19 12:13:42 <ogra_> es2gears runs with about 2000frames
Dec 19 12:15:11 <ogra_> i think the font issue was flash ... i was testing that alongside
Dec 19 12:16:25 <lilstevie> lol nice
Dec 19 12:16:35 * ogra_ curses gnome-power-manager for not keeping the backlight setting
Dec 19 12:17:51 <ogra_> wow, even the dash comes up at a usable speed now
Dec 19 12:18:00 <lilstevie> heh nice
Dec 19 12:18:14 <ogra_> around 3sec ... vs 15 or so
Dec 19 12:18:23 <lilstevie> dash is a little slow on fbdev
Dec 19 12:18:27 <lilstevie> yeah
Dec 19 12:18:29 <ogra_> yep
Dec 19 12:18:50 <xranby> ogra_: are you running with the 256kb default stack on this setup as well?
Dec 19 12:18:52 <ogra_> i always blamed disk IO ... but that doesnt seem to be it
Dec 19 12:18:55 <xranby> for pthreads?
Dec 19 12:19:25 <ogra_> i seem to, yes
Dec 19 12:19:37 <marvin24> lilstevie: hdmi problem is know and fixed if janimo pulls again ;-)
Dec 19 12:20:22 <lilstevie> marvin24: not ac100
Dec 19 12:20:37 <marvin24> ah, sorry
Dec 19 12:20:41 <lilstevie> tf101
Dec 19 12:20:56 <marvin24> so tf101 has 2000 fps?
Dec 19 12:21:13 <lilstevie> no, ogra is talking ac100
Dec 19 12:21:23 <lilstevie> the issues I am talking about are tf101
Dec 19 12:21:23 * marvin24 confused again
Dec 19 12:21:44 <janimo> marvin24, I will pull for next 3.0 upload, sure. Not sure ATM to work around the issue by forcing a build with 4.5 or wait till gcc is fixed - it appears to be raised by others too
Dec 19 12:21:47 <lilstevie> ogra said he was getting 2k fps on es2gears
Dec 19 12:21:54 <ogra_> right
Dec 19 12:22:07 <ogra_> i said that in #ac100 before
Dec 19 12:22:25 <ogra_> but since janimo isnt a resident there anymore i repeated it here :)
Dec 19 12:22:43 <marvin24> janimo: you can build with the recommended option
Dec 19 12:22:54 <janimo> yeah, I was unable to keep tabs on everything happening on that channel :)
Dec 19 12:23:19 <ogra_> heh
Dec 19 12:23:28 <lilstevie> heh'
Dec 19 12:23:38 <marvin24> ogra_: you said 460 on #ac100
Dec 19 12:23:45 * ogra_ wishes we would find a quirk for suspend/resume 
Dec 19 12:23:47 <marvin24> and this is the value I also get
Dec 19 12:23:52 <janimo> marvin24, right, more or less same amount of change as forcing 4.5. Just wondering it should do any toolchain workaround at all, or wait till it is fixed, it should not be more than a week or two since it appears serious
Dec 19 12:24:09 <ogra_> marvin24, 2000 frames .... 400 something FPS
Dec 19 12:24:09 <lilstevie> I'm only in #ac100 cause ogra recommended it for kernel stuff
Dec 19 12:24:23 <janimo> ogra_, suspend/resume broken only with the L4T driver right?
Dec 19 12:24:35 <janimo> are there no issues it was released for 2.6.36 only?
Dec 19 12:24:43 <ogra_> janimo, yep, backlight doesnt seem to come back
Dec 19 12:25:17 * ogra_ treis console switching again without having a broken flash running
Dec 19 12:26:41 * janimo contemplates building Libreoffice on the ac100
Dec 19 12:26:55 <janimo> panda would likely be faster though
Dec 19 12:31:00 <ogra_> well, that didnt go so well
Dec 19 12:31:32 <ogra_> i can switch to the console and also switch consoles then ... but as soon as i switch back to X everything hangs hard
Dec 19 12:32:14 * ogra_ will try precise later today ... that requires an armel resinatll though
Dec 19 12:32:18 <ogra_> *reinstall
Dec 19 12:32:47 <janimo> xranby, hi, I suppose the mmap() bug you reported is still present on the ac100. Can you check if it is ok with the 3.0 kernel you tried?
Dec 19 12:33:16 <janimo> I applied the supposed fix to the 2.6.38 branch for oneiric but I still got an error with the app built from your original .c file attached to the bug report
Dec 19 12:33:22 <ogra_> janimo, oh, btw did you check what mmap_min_addr is defaulting to in the ac100 build ?
Dec 19 12:33:27 <janimo> context for others https://bugs.launchpad.net/ubuntu/+source/linux/+bug/861296
Dec 19 12:33:28 <ubot2> Launchpad bug 861296 in linux-ac100 "mmap fails to allocate 2030Mb heap on ARM" [Undecided,Confirmed]
Dec 19 12:33:29 <ogra_> seems it was wrong in all our arm kernels
Dec 19 12:33:44 <xranby> janimo: i will check in a minute
Dec 19 12:33:47 <janimo> ogra_, I did not as it was mentioned in that bug ac100 is fine :)
Dec 19 12:33:52 <ogra_> (should be 32k but it was 64... only procps saved us from hard locking)
Dec 19 12:33:55 <janimo> but I could double check I guess
Dec 19 12:34:31 <janimo> ogra_, it is 32K in precise now
Dec 19 12:34:43 <ogra_> ogra@horus:~$ grep MMAP_MIN /boot/config-2.6.38-1001-ac100
Dec 19 12:34:43 <ogra_> CONFIG_DEFAULT_MMAP_MIN_ADDR=32768
Dec 19 12:34:47 <ogra_> yep, looks fine
Dec 19 12:34:47 <janimo> yep
Dec 19 12:34:56 <ogra_> thats oneiric :)
Dec 19 12:35:01 <ogra_> so it was always right
Dec 19 12:35:08 <janimo> so same in both cases as the configs are the same ,more or less
Dec 19 12:35:10 <ogra_> yay for community kernels *g*
Dec 19 12:35:17 <janimo> :)
Dec 19 12:35:24 <janimo> yay for sheer luck :)
Dec 19 12:35:29 <ogra_> heh
Dec 19 12:35:44 <ogra_> well, i created the vers first config after a lucid panda kernel
Dec 19 12:35:48 <ogra_> *very
Dec 19 12:35:58 <ogra_> i guess that value was always just carried over
Dec 19 12:36:00 <janimo> so panda was not affected either?
Dec 19 12:36:01 <lilstevie> ac100 has an awesome community :p
Dec 19 12:36:04 <ogra_> it was
Dec 19 12:36:12 <ogra_> but the breakage occured in natty for panda iirc
Dec 19 12:36:20 <ogra_> lilstevie, definitely
Dec 19 12:36:23 <xranby> janimo: i have to activate some swap in order to run that test on my ac100
Dec 19 12:36:31 <ogra_> i wish we had more such communities for other devices
Dec 19 12:36:39 <janimo> xranby, so it only triggers with swap?
Dec 19 12:36:54 <janimo> can it be reproduces w/o swap but other alloc size?
Dec 19 12:37:09 * ogra_ is really impressed how the nvidia driver seems to have sped up the world on this device
Dec 19 12:37:10 <lilstevie> ogra_: heh I wish the tf101 had as awesome of a community
Dec 19 12:37:17 <janimo> xranby, anyway you know better. The 3.0 kernel has the two changes applied that are supposed to fix it so I hope all is well
Dec 19 12:37:18 <xranby> janimo: no but you cant pass the test without 4Gb of memory
Dec 19 12:37:40 <xranby> so you need to add swap since the ram on the ac100 are too limited
Dec 19 12:37:51 <lilstevie> as it stands the tf101 community is, well me Lo
Dec 19 12:37:57 <janimo> lilstevie, if it only was as easy to boot and rewrite as the ac100 :) I hear it is locked down in newer models
Dec 19 12:38:17 <lilstevie> janimo: that is correct
Dec 19 12:38:21 <janimo> xranby, ah, had no idea. Will add more swap when testing my oneiric backport then
Dec 19 12:38:45 <lilstevie> well not so much newer model
Dec 19 12:38:53 <lilstevie> the only difference is the production run
Dec 19 12:39:01 <xranby> janimo: the test passed on the 3.0 ac100 kernel running armhf using 3gb of swap
Dec 19 12:39:10 <xranby> \o/
Dec 19 12:43:05 <janimo> xranby, thanks, great
Dec 19 12:43:22 <janimo> I'll more confidently test for the SRU kernel then :)
Dec 19 12:44:35 * ogra_ wonders why he has a "zaurus" module loaded
**** BEGIN LOGGING AT Mon Dec 19 12:51:54 2011
Dec 19 12:52:07 <ogra_> hmm, and shouldnt i see some tegrafb stuff in dmesg ?
Dec 19 12:52:07 <ogra_> i suspect thats the bit breaking console switching
Dec 19 12:55:35 <ogra_> hmm, tegrafb is set on my kernel
Dec 19 13:05:59 <ogra_> hmm, the kernel thats shipped in L4T has the backlight lcd driver enabled
Dec 19 13:06:08 <ogra_> i wonder if it relies on this for resume
Dec 19 13:11:34 <ogra_> hmm, we have LCD compiled in ... while they have it as module
Dec 19 13:24:57 * ogra_ finds it really intresting that hsi system stays below 300MB using the binary driver ... with the same set of apps open i usually got it using all RAm and at least 100-200M swap
Dec 19 13:26:07 <ogra_> oh, i so wish we had an armhf version :(
Dec 19 13:27:13 <lilstevie> :(
Dec 19 13:27:57 <ogra_> it would fly !
Dec 19 13:28:33 <lilstevie> yeah
Dec 19 13:28:43 <lilstevie> hf really does fly comparitivly
Dec 19 13:28:51 <ogra_> right
Dec 19 13:29:01 <ogra_> and el flies with the binary driver
Dec 19 13:29:08 <lilstevie> heh
Dec 19 13:30:07 <lilstevie> I wish I had a stable kernel
Dec 19 13:30:09 <lilstevie> for accel
Dec 19 13:34:39 <ogra_> hmpf
Dec 19 13:35:04 * ogra_ has a hanging totem that even doesnt let the ps output finish
Dec 19 13:35:14 <ogra_> killing doesnt help either
Dec 19 13:35:23 * ogra_ reboots
Dec 19 13:35:54 <ogra_> hmm, or not ...
Dec 19 13:35:59 <ogra_> even reboot hangs
Dec 19 13:36:08 <ogra_> time for a hard reset
Dec 19 13:37:14 <lilstevie> lol
Dec 19 16:01:42 <lool> doko: I don't have a list, however I tried to have such NEON flavors only be built when the toolchain does *not* default to NEON, I did that in a couple of packages, a bit like the vfp pass on Debian armel and not Ubuntu armel by checking whether vfp is turned on or not
Dec 19 16:02:14 <lool> I think that's what we should generally do, so that if someone rebuilds Debian/Ubuntu sources with different toolchain defaults, he gets the best thing
Dec 19 16:24:36 <doko> lool, right, but do you still know which packages you did touch?
Dec 19 16:25:00 <doko> GrueMaster, kernel test ping
Dec 19 16:25:39 <GrueMaster> I'm here (but not necessarily awake).
Dec 19 16:26:22 <GrueMaster> What would you like tested?
Dec 19 16:27:23 <doko> GrueMaster, the linux-ti-omap, just copied into oneiric-proposed
Dec 19 16:28:21 <GrueMaster> Is there a specific test you need or just run my normal SRU test suite?
Dec 19 16:30:10 <doko> just your tests, and maybe the test case in bug 861296, although I already did that
Dec 19 16:30:11 <ubot2> Launchpad bug 861296 in linux-ac100 "mmap fails to allocate 2030Mb heap on ARM" [Undecided,Confirmed] https://launchpad.net/bugs/861296
Dec 19 16:32:38 <GrueMaster> One of my tests actually hits that, so I'll see pass/fail fairly easily (test takes ~1.5 hours to run though).
Dec 19 16:35:22 <GrueMaster> Starting job now.  Will post results in a few hours (have to reimage as part of the test, and other tests take time).
Dec 19 16:36:04 <janimo> GrueMaster, is omap4 armhf image in good shape? I need to set it up to build Libreoffice on the panda
Dec 19 16:37:10 <GrueMaster> I reimaged on Saturday and it seemed ok (netinstall).  Haven't checked the daily pre-installed.  I can fire one up here very quickly if you want.
Dec 19 16:43:13 <kinta> Hello,  what tablet do you think is working better with ubuntu? (i am planning to get new one)
Dec 19 16:47:28 <doko> janimo, armhf chroot should be enough
Dec 19 16:48:00 <janimo> doko, right but I need to test armhf more seriously so may do the full switch
Dec 19 16:48:45 <janimo> GrueMaster, no, thanks, just wanted to make sure there's no outstanding issue so I don't find out about it at the end of the ubiquity run :)
Dec 19 16:49:28 <GrueMaster> Well, I haven't looked at desktop in over a week (SRU automation).
Dec 19 16:49:52 <GrueMaster> Nor server.
Dec 19 16:49:52 <ogra_> youwill still have to manually run oem-config-remove after first login
Dec 19 16:50:06 <ogra_> there is still a dbus bug that prevents it from running
Dec 19 16:50:22 <GrueMaster> ogra_: That's desktop only, right?
Dec 19 16:50:24 * ogra_ will look into that right after vacation
Dec 19 16:50:30 <GrueMaster> I don't remember server being an issue.
Dec 19 16:50:34 <ogra_> i think thats server preinstalled too
Dec 19 16:50:42 <ogra_> but not sure
Dec 19 16:50:50 <ogra_> adam tested that one iirc
Dec 19 16:51:03 <ogra_> i only did a few different desktop tests here
Dec 19 16:51:14 <ogra_> and during vacation i will only do ac100
Dec 19 16:52:30 <GrueMaster> Well, the only image I can't test at this time is mx5.  I am almost finished with SRU automation (it runs a full install and test suite from jenkins now).
Dec 19 16:52:48 <GrueMaster> My next automation setup is server images.
Dec 19 16:54:09 <GrueMaster> Core is already mostly automated, just need to integrate it into a daily job.  Might also expand it to run in LXC.
Dec 19 17:37:52 <Francis_Albert> how are you ?
Dec 19 17:51:14 <RoyK> hi all. trying to install ubuntu 11.10 on a pandaboard fails - installer restarts somewhere 75% into the installation - anyone taht knows a fix for this?
Dec 19 17:53:36 <GrueMaster> RoyK: Yes, once the installer restarts, switch to a console (<ctrl><alt><F1>), login, and type "sudo oem-config-remove && sudo reboot".
Dec 19 17:54:16 <GrueMaster> There is a dbus race condition that we are working to resolve.
Dec 19 17:54:55 <GrueMaster> Won't fix in Oneiric (would require image respin), but we want it fixed in Precise.
Dec 19 17:55:03 <doko> GrueMaster, could you ping me (and lamont) on test results? trying to convince to install the kernel from -proposed once it's tested
Dec 19 17:55:36 <RoyK> GrueMaster: thanks
Dec 19 17:57:22 <GrueMaster> Doko It is in testing now (since 1.5 hours ago).  Should have results in ~1 hour (The current test takes 2.5 hours according to my success history from when I tested the mmap patch).
Dec 19 18:36:31 <infinity> ogra_: The oem-config-remove thing (A) only happens with the GTK frontend, and (B) only happens sporadically (hasn't happened to me yet this cycle).
Dec 19 18:55:09 <GrueMaster> doko:  Kernel SRU tests passed.
Dec 19 19:17:38 <parin> when try to install omap4 extras is syas available from "ti-ompa4-ppa" source, i hit use this source it says not found
Dec 19 19:17:50 <parin> it says there is not a software package called "ubuntu-omap4-extras" in your current software source
Dec 19 19:18:08 <parin> i aslo tried installing from command line same result
Dec 19 19:21:04 <GrueMaster> parin: Which release are you trying to use?
Dec 19 19:25:11 <parin> I am using Ubuntu 11.10
Dec 19 19:25:59 <GrueMaster> Hmmm.  Did you to an apt-get update?
Dec 19 19:27:25 <parin> i dont exactly remember i did that
Dec 19 19:27:33 <parin> but i can try it now
Dec 19 19:31:35 <parin> ok now i di update and getting package dependencies error
Dec 19 19:33:44 <RoyK> where can I find the ubuntu-omap4-extras* packages?
Dec 19 19:38:39 <GrueMaster> Can you type "apt-cache madison ubuntu-omap4-extras"?
Dec 19 19:39:18 <GrueMaster> Have you tried the ti link on the unity launcher?  It should setup the ppa and install the packages for you.
Dec 19 19:44:45 <RoyK> seems there was an apt-get update missing after the repo was added
Dec 19 19:49:35 <GrueMaster> ah.
Dec 19 20:00:12 <parin> i added ppa:tiomap-dev/release, applied apt-get update still not able to install
Dec 19 20:03:00 <parin> if i do sudo apt-get install libegl1-sgx-omap4 libgles1-sgx-omap4 libgles2-sgx-omap4
Dec 19 20:03:23 <parin> it asks me to delete some libegli-mesa packages
Dec 19 20:03:27 <parin> is that ok?
Dec 19 20:16:58 <infinity> Yes.
Dec 19 20:33:49 <parin> now after installing omap4-extras i restared the pandabaord and i am gettinng errors
Dec 19 20:34:25 <parin> ti_st_open st_register  failed -22
Dec 19 20:34:36 <parin> and only command line access
Dec 20 01:51:20 <cdahmedeh> hello friends! any recommendations for an arm development board? it needs to have hdmi output
Dec 20 01:51:47 <cdahmedeh> ideally it should be easy to find (ie, don't have to wait 5 months for order), and not more than 150$
Dec 20 01:52:14 <twb> development as in for a hobby, or with an eye to building something production later on?
Dec 20 01:52:21 <cdahmedeh> hobby
Dec 20 01:53:06 <twb> Dunno about the HDMI, the most important thing is to make sure it's arm v7
Dec 20 01:53:18 <cdahmedeh> why is that so?
Dec 20 01:53:31 <twb> ubuntu doesn't support earlier versions of the architecture
Dec 20 01:53:41 <cdahmedeh> alright
Dec 20 01:54:14 <twb> Which basically means not sheeva/guruplug, everything else is new enough design
Dec 20 01:54:26 <cdahmedeh> ok
Dec 20 01:54:27 <twb> A lot of people here seem to be using beagle/panda boards
Dec 20 01:54:51 <twb> (I'm only here because my netbook is arm, so don't trust anything I say particularly :p)
Dec 20 01:55:12 <cdahmedeh> hmm.. arm netbook would be an interesting alternative
Dec 20 01:56:50 <cdahmedeh> the raspberry pi seems interesting, but i have a feeling they're going to be really hard to get at first
Dec 20 01:56:56 <cdahmedeh> that thing is going to sell like hotcakes
Dec 20 02:06:39 <lilstevie> raspberry pi is not armv7
Dec 20 02:07:35 <cdahmedeh> i don't plan to run ubuntu specifically, debian would do
Dec 20 02:07:54 <lilstevie> ok, well debian will
Dec 20 02:08:01 <cdahmedeh> it's armv6 isn't it?
Dec 20 02:08:17 <lilstevie> raspberry pi?
Dec 20 02:08:35 <lilstevie> debian is armv4t IIRC
Dec 20 02:08:44 <lilstevie> and yes, the pi is armv6
Dec 20 02:08:55 <cdahmedeh> thanks, you answered my next question (-:
Dec 20 02:09:11 <cdahmedeh> anyways, do you have any recommendations for an arm development board?
Dec 20 02:09:16 <lilstevie> but really these armv7 devices are faster
Dec 20 02:09:28 <lilstevie> beagle/panda
Dec 20 02:09:33 <lilstevie> for dev boards
Dec 20 02:09:37 <cdahmedeh> ok
Dec 20 02:09:50 <lilstevie> beagle is omap3 panda is omap4
Dec 20 02:10:11 <lilstevie> so probably over all I would say go the panda
Dec 20 02:10:27 <cdahmedeh> sounds neat
Dec 20 02:11:36 * cdahmedeh looks at ship date
Dec 20 02:11:37 <cdahmedeh> oh wow
Dec 20 02:11:42 <cdahmedeh> jan 18 2012
Dec 20 02:11:45 <cdahmedeh> (-:
Dec 20 02:11:54 <lilstevie> heh
Dec 20 02:12:02 <lilstevie> you can get the panda from digikey
Dec 20 02:12:09 <cdahmedeh> that's where i checked
Dec 20 02:12:18 <lilstevie> heh
Dec 20 02:12:31 <cdahmedeh> sounds like these arm boards are tricky to find aren't they?
Dec 20 02:12:38 <lilstevie> they can be
Dec 20 02:12:44 <lilstevie> they aren't the most common things
Dec 20 02:12:51 <lilstevie> I have the trimslice myself
Dec 20 02:13:02 <lilstevie> but that is a fair bit more than your budget
Dec 20 02:13:10 <lilstevie> panda is only just over it
Dec 20 02:13:23 <cdahmedeh> yes exactly
Dec 20 02:13:27 <cdahmedeh> the platforms are very interesting
Dec 20 02:13:41 <cdahmedeh> the mali 400 sounds really nice, but so far the boards i am finding are expensive
Dec 20 02:14:32 <lilstevie> yeah
Dec 20 02:14:44 <lilstevie> well the mali-400 really are only in the exynos dev boards
Dec 20 02:14:54 <lilstevie> and samsung devel boards are really expensive
Dec 20 02:15:47 <cdahmedeh> wow, very expensive
Dec 20 02:15:55 <cdahmedeh> wow these cost more than i thought they would
Dec 20 02:16:06 <lilstevie> cdahmedeh: which do?
Dec 20 02:16:13 <lilstevie> exynos, or trimslice
Dec 20 02:16:24 <lilstevie> or...
Dec 20 02:16:54 <cdahmedeh> the boards with the mali 400
Dec 20 02:17:02 <lilstevie> ah yeah
Dec 20 02:17:09 <lilstevie> they are like 300-400 IIRC
Dec 20 02:17:43 <cdahmedeh> yeah something like that
Dec 20 02:18:32 <lilstevie> tegra devel boards are terrible
Dec 20 02:18:45 <lilstevie> that said, the ones from nvidia even have an lcd
Dec 20 02:18:57 <lilstevie> but they are $1000
Dec 20 02:19:04 <cdahmedeh> crazy
Dec 20 02:19:13 <cdahmedeh> though tegra sounds like a great platform
Dec 20 02:19:34 <lilstevie> trimslice is a tegra2 board and that will set you between 200-400
Dec 20 02:19:41 <lilstevie> tegra isn't that great
Dec 20 02:22:02 <cdahmedeh> it's a bit slower than the sgx 540 isn't it?
Dec 20 02:25:25 <cdahmedeh> though i don't think i will need that much gpu to play with
Dec 20 02:26:54 <cdahmedeh> there are no arm devices with regular open gl support right?
Dec 20 02:28:14 <lilstevie> correct
Dec 20 02:31:34 <cdahmedeh> i think i'm going to go with a pandaboard
Dec 20 02:32:00 <cdahmedeh> the connectivity options are great, has two hdmi/dvi outputs
Dec 20 02:32:03 <lilstevie> cool
Dec 20 02:32:08 <lilstevie> yeah
Dec 20 02:32:25 <lilstevie> and is supported by canonical
Dec 20 02:32:36 <cdahmedeh> perfect
Dec 20 02:33:03 <cdahmedeh> so that thing obviously needs a power supply which it does not come with
Dec 20 02:33:12 <cdahmedeh> anything other accessories that i miss?
Dec 20 02:33:41 <lilstevie> no idea, I don't have one
Dec 20 02:36:35 <cdahmedeh> well thanks alot for your help lilstevie
Dec 20 02:39:42 <twb> 13:01 <cdahmedeh> i don't plan to run ubuntu specifically, debian would do
Dec 20 02:39:53 <twb> cdahmedeh: debian's new armhf arch is armv6
Dec 20 02:39:57 <twb> *armv7
Dec 20 02:40:11 <twb> So if you want to run Debian I would suggest still aiming for armv7
Dec 20 02:41:05 <cdahmedeh> i'll take that into account then
Dec 20 02:41:19 <twb> IME the most important thing is to buy a device that lots of other hackers use
Dec 20 02:41:50 <cdahmedeh> and the pandaboard/beagleboard are the most popular ones right now. correct?
Dec 20 02:41:57 <twb> Like back in the day, you would try to get a thinkpad or a powerbook -- not so much because they were good (though they were), but because they had a huge linux userbase and people actually made sure they worked
Dec 20 02:42:26 <cdahmedeh> i noticed that
Dec 20 02:43:31 <lilstevie> yes panda would be the target
Dec 20 02:43:42 <lilstevie> most of the guys have pandas or beagles AFAIK
Dec 20 02:45:15 <cdahmedeh> ok
Dec 20 02:46:15 <twb> What's a panda cost, ballpark?
Dec 20 02:46:30 <cdahmedeh> listed price is 174$ i think
Dec 20 02:46:37 <cdahmedeh> sounds about right to me
Dec 20 02:46:47 <cdahmedeh> plus you need to buy power supply
Dec 20 02:46:49 <twb> US$?
Dec 20 02:46:52 <cdahmedeh> shipping and taxes
Dec 20 02:46:52 <cdahmedeh> yes
Dec 20 02:46:56 <twb> Righto
Dec 20 02:47:51 <cdahmedeh> like i was toying around with qemu, it's awesome, but there's no way to emulate opengles
Dec 20 02:48:37 <twb> lilstevie: currently when I just naively open up a GTK2 browser and try to do multi-finger scroll on the touchpad, it doesn't work.  What bit(s) aren't configured for that yet: the touchscreen driver, GTK, the browser, ... ?
Dec 20 02:49:02 <cdahmedeh> don't you need to configure x somehow?
Dec 20 02:49:33 <cdahmedeh> maybe create/edit /usr/share/X11/xorg.conf.d/50-synaptics.conf
Dec 20 02:49:56 <twb> I didn't think tf101 was using synaptics hw
Dec 20 02:50:25 <cdahmedeh> oh
Dec 20 02:50:50 <twb> Hm, pandaboard is listed as using powervr GPU -- I thought they only worked with shitty binary drivers
Dec 20 02:51:49 <cdahmedeh> isn't that the case for any 3d accelerated arm gpu?
Dec 20 02:52:36 <GrueMaster> True for most any decent 3D gpu, regardless of arch.
Dec 20 02:53:11 <cdahmedeh> the only completely usable 3d open source drivers are on intel as far as i know
Dec 20 02:53:20 <cdahmedeh> and maybe some older ati chips
Dec 20 02:53:25 <twb> cdahmedeh: and matrox!
Dec 20 02:53:34 <cdahmedeh> i did not know that
Dec 20 02:53:35 <twb> That was probably technically 3d
Dec 20 02:53:49 <twb> cdahmedeh: they stopped making cards in like 1992
Dec 20 02:53:58 <cdahmedeh> 1492*
Dec 20 02:53:59 <cdahmedeh> (-:
Dec 20 02:54:22 <lilstevie> twb: all arm devices use "shitty binary drivers"
Dec 20 02:54:30 <twb> :-(
Dec 20 02:54:32 <lilstevie> and same as x86
Dec 20 02:54:46 <GrueMaster> Intel may make open source 3D drivers, but they still suck.
Dec 20 02:54:55 <cdahmedeh> they're acceptable
Dec 20 02:54:58 <twb> Well, AMD x86 yes, intel boards are pretty good
Dec 20 02:55:06 <twb> as far as free-ness goes that is
Dec 20 02:55:11 <GrueMaster> Depends on what you are doing.
Dec 20 02:55:23 <cdahmedeh> it tends to die off when you use the more obscure opengl stuff
Dec 20 02:55:23 <lilstevie> nvidia don't /really/ FOSS their drivers either
Dec 20 02:55:24 <lilstevie> :p
Dec 20 02:55:26 <twb> Except for poulsbo and iwl
Dec 20 02:55:39 <cdahmedeh> yeah thoses are still closed
Dec 20 02:55:41 <cdahmedeh> iwl isn't that great
Dec 20 02:55:47 <twb> iwl blows
Dec 20 02:55:52 <twb> it has closed firmware blobs
Dec 20 02:55:56 <cdahmedeh> i have the strangest with iwl
Dec 20 02:55:58 <twb> ath9k wtf
Dec 20 02:55:59 <lilstevie> but twb as for the touchpad, it is not working properly
Dec 20 02:56:01 <twb> *ftw
Dec 20 02:56:08 <cdahmedeh> i have to get myself an ath9k
Dec 20 02:56:09 <GrueMaster> Poulsbo is an odd case.  It is produced by Intel, but it is PowerVR IP (so Intel doesn't have access to TRM).
Dec 20 02:56:09 <twb> lilstevie: driver issue?
Dec 20 02:56:18 <lilstevie> multitouch is pretty much disabled in the touchpad driver :)
Dec 20 02:56:24 <twb> GrueMaster: that's what caused my original question
Dec 20 02:56:40 <lilstevie> the only thing you can do with it is two finger tap
Dec 20 02:56:44 <twb> lilstevie: er, I'm talking about the touchscreen, not the thing next to the keyboard
Dec 20 02:57:00 <lilstevie> oh
Dec 20 02:57:12 <lilstevie> well you said on the touchpad
Dec 20 02:57:17 <twb> Sorry braino
Dec 20 02:57:19 <GrueMaster> It is worse than not having PowerVR drivers.  Intel can't develop Poulsbo in-house due to competitive reasons.
Dec 20 02:57:59 <cdahmedeh> does poulsbo work at all under linux?
Dec 20 02:58:00 <GrueMaster> At least Ti has the TRM to develop decent drivers, even if PowerVR won't let them open source the drivers.
Dec 20 02:58:05 <twb> GrueMaster: pity, because atom Z has hardware VT; whereas the other atoms dont
Dec 20 02:58:18 <GrueMaster> Poulsbo used to.
Dec 20 02:58:25 <twb> cdahmedeh: yes using a crappy old kernel that supports the powervr gpu, or using vesa
Dec 20 02:58:40 <GrueMaster> twb: I know.  I used to contract for Intel during the Moblin 1.0 days.
Dec 20 02:58:56 <cdahmedeh> twb: sounds wonderful!
Dec 20 02:59:08 <GrueMaster> Flash worked better running in XP in a VT on Linux than native at the time.
Dec 20 02:59:15 <twb> Stupid intel "vt is a price diffentiator"
Dec 20 02:59:27 <lilstevie> twb: I would guess the browser is not configured with utouch/gies
Dec 20 02:59:30 <lilstevie> geis*
Dec 20 02:59:34 <twb> GrueMaster: eh, who the fuck cares about flash
Dec 20 02:59:45 <twb> lilstevie: yeah, definitely not :-)
Dec 20 02:59:53 <GrueMaster> Heh.  Well, Moblin 1.0 desktop was flash based, so....
**** ENDING LOGGING AT Tue Dec 20 02:59:57 2011