[01:16] <jj-afk> back on later
[01:48] <jk-> ogasawara: ping?
[02:25] <BenC> I just saw that Canonical was supposed to have a session at the Freescale Technology Forum, but it seems to have been canceled...sad :(
[02:36] <lifeless> the forum or the session?
[04:16] <ogasawara> jk-: pong
[04:16] <jk-> ogasawara: hey
[04:17] <jk-> just been looking though lucid git
[04:18] <jk-> bcbfc24ead6935714a001ab6f1f725763075f98c in particular - is this yours?
[04:20] <jk->  ( http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-maverick.git;a=commitdiff;h=bcbfc24ead6935714a001ab6f1f725763075f98c )
[04:40] <ogasawara> jk-: indeed it appears that way as I'd signed off on it, however I'm slightly confused by the bug noted in the commit message
[04:40] <ogasawara> jk-: the commit for that bug should really be:
[04:40] <ogasawara> commit 5b4d91d21e6c3e57fdbd5b4c2354c6f98153a3f3
[04:40] <ogasawara> Author: Islam Amer <pharon@gmail.com>
[04:40] <ogasawara> Date:   Thu Jun 24 13:39:47 2010 -0400
[04:40] <ogasawara>     (pre-stable) dell-wmi: Add support for eject key on Dell Studio 1555
[04:41] <jk-> ok, i'm confused now
[04:41] <jk-> that patch is for drivers/media/video/
[04:41] <jk-> ah, yeah, buglink is wrong
[04:41] <jk-> anyhow
[04:42] <jk-> there are a couple of changes that aren't IR-related
[04:43] <jk-> - to drivers/media/video/cx23885/cx23885-core.c
[04:43] <jk-> is this intentional?
[04:46] <ogasawara> jk-: I want to say yes but I'd likely need to go back and review the email thread to completely refresh my memory
[04:46] <jk-> ogasawara: probably getting late there, I can put this in an email to kernel-team if you prefer
[04:46] <jk-> I'll check the thread out.
[04:51] <ogasawara> jk-: Subject: "LIRC 0.8.7 Fixes for Maverick"
[04:52] <jk-> ogasawara: cool, thanks
[07:32] <bullgard4> Ich bin mit einer Natty-Live-CD unterwegs. Warum öffnet 'Alt+F2 > Run a command > gparted <Enter>' nicht GParted? Ebenso nicht das Klicken auf  auf das Symbol »gparted«?  
[07:33] <bullgard4> Wie kann man in der Live-CD umschalten von Unity auf GNOME 2?
[07:34] <bullgard4> upps! Pardon me.
[08:19] <smb> morning
[08:19] <diwic> morning
[08:21]  * apw yawns ... morning
[08:22] <smb> apw, You sound like "need tea/coffee"
[08:37] <abogani> morning all
[08:38]  * apw needs t/c badly thank you ...
[08:50] <apw> smb you got a hardy box lying around still ?
[08:51] <smb> apw, You bet. Currently using it for xen test but I got a desktop install on it too.
[09:02] <apw> smb, http://people.canonical.com/~apw/CVE-2011-0726-hardy/
[09:02] <ubot2> apw: ** RESERVED ** This candidate has been reserved by an organization or individual that will use it when announcing a new security problem.  When the candidate has been publicized, the details for this candidate will be provided. (http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2011-0726)
[09:03] <YoBoY> hi
[09:03] <apw> smb, so we should expect the contents of /proc/<pid>/stat to have addresses 0'd out for stack, code start and code end
[09:03] <apw> with that kernel
[09:03] <apw> YoBoY, hi
[09:06] <YoBoY> last week I came here with a freeze problem on my computer on natty (since the release). I used ubuntu some days without effects like recommended and no freeze problem anymore. I also tried since yesterday the new 2.6.39 kernel and the freeze problem came back, any idea what I can try to correct that and keep visual effects on my computer ?
[09:08] <apw> YoBoY, there is nothing other than a hard hang?  you are unable to contact the machine over the network when it hangs ?
[09:10] <YoBoY> apw: I can go to a tty and restart gdm
[09:10] <apw> YoBoY, and things work find after restarting gdm ?
[09:10] <apw> s/find/fine
[09:10] <YoBoY> yes until the next freeze
[09:11] <YoBoY> the hang occurs when the computer go to screensaver by the way (forget to tell)
[09:11] <YoBoY> and not each times
[09:11] <apw> ok thats very odd, it is cirtainly unusual for restarting gdm to recover from a kernel grpahics driver failure
[09:12] <apw> and this only occurs when 'effects' are enabled yes ?
[09:12] <YoBoY> yes
[09:12] <apw> i am suspicious this is not a kernel issue, but a compiz one
[09:12] <apw> my first suggestion would be to reproduce the issue again and then switch to VT1
[09:12] <YoBoY> vt1 ?
[09:13] <apw> from there, first take a copy of the kernel dmesg, running: dmsg >DMESG.FROZEN
[09:13] <YoBoY> tty ?
[09:13] <apw> (vt1 == tty1, ctrl-alt-F1)
[09:13] <YoBoY> ok ^^"
[09:13] <apw> and then try restarting compiz,  i think that the incantation is:  DISPLAY=:0.0 compiz --replace
[09:14] <YoBoY> I can do that, next time :)
[09:14] <apw> and then switch back to X and see if that unfreezes things
[09:14] <apw> if the compiz replace works i am suspicious that we have a compiz hang and need to report against that, otherwise a report it against the kernel
[09:19] <YoBoY> I tried the 2.6.39 kernel beacause I thought my problem was similar to the bug 740126
[09:19] <ubot2> Launchpad bug 740126 in linux "Disabling an output can cause vblank events to be missed" [High,Fix released] https://launchpad.net/bugs/740126
[09:20] <apw> http://awhitcroft.blogspot.com/2011/05/union-file-systems-again.html
[09:22] <YoBoY> thanks apw i'll come back with the results
[09:22] <apw> YoBoY, sounds good
[10:50] <apw> ppisati, i am assuming you're noticong that when i am doing CVEs now i am sending out fsm-imx51 etc patches too, so you don't have to worry about those CVEs separatly
[10:56] <ppisati> apw: yep, saw it
[10:57] <apw> ppisati, maybe one day you and i will meet in the middle of the table
[10:57] <apw> ppisati, and i think we deserve some beer when we get there
[10:57] <ppisati> that's a good excuse to booze :)
[11:30] <ppisati> bug 800121
[11:30] <ubot2> Launchpad bug 800121 in linux-ti-omap4 "CVE 2010-4649" [Undecided,New] https://launchpad.net/bugs/800121
[11:30] <ppisati> please accept the nominations
[12:20] <apw> ppisati, you should mention my nick ... and done
[12:46]  * apw pops ou
[12:46] <apw> out
[12:49] <jjohansen> apw: thanks for setting up the planet.ubuntu syndication
[13:00] <jvargas> hi
[13:01] <jvargas> I want to downgrade the 11.04 kernel to the one shipped with 10.10, so my network card could work at a very decent speed.
[13:02] <jvargas> how can I proceed? I do not find info aboiut that..
[13:26] <jj-afk> back on in a few
[13:35]  * apw whines about poor internet
[14:44] <bjf> ##
[14:44] <bjf> ## Ubuntu Kernel Team Meeting - Today @ 17:00 UTC - #ubuntu-meeting
[14:44] <bjf> ##      agenda: https://wiki.ubuntu.com/KernelTeam/Meeting
[14:44] <bjf> ##
[15:02]  * ogasawara back in 20
[15:09] <apw> tgardner, would you cast an eye over the oneiric ti-omap4 binaries in ckt PPA, they are oneiric2 as the ppa wouldn't allow me to reuse the name, but would be oneiric1 for upload to OO
[15:09] <apw> (then i can delete them before i get moaned at by those whining stable people)
[15:09] <tgardner> apw, cast an eye to what end ?
[15:10] <apw> tgardner, sanity check the version numbers for me, i don't want another libc abi hack or similar
[15:10] <tgardner> apw, ah
[15:10] <apw> they look ok to me, but i am too familiar with them now
[15:10] <apw> the way they are made it should be impossible to be anything other than the same as natty ... but i've been blind to it before
[15:14] <tgardner> apw, you don't have a ti-omap4 branch in oneiric yet ?
[15:14] <apw> tgardner, we have nothing newer than what is in natty no
[15:14] <apw> tgardner, i've pushed what i am proposing to upload into the repo for now
[15:14] <apw> tgardner, when we do finally get a real one then it will just get pushed over and go away
[15:15] <apw> last i asked it was still at least weeks away before we see anything
[15:16] <apw> and i am not confident, we'll get it then, plus the fact its behind natty is starting to impact testing
[15:16] <tgardner> apw, just did an update on oneiric. now I see the ti-omap4 branch
[15:17] <apw> tgardner, cool 
[15:19] <tgardner> apw, so, isn't the issue a namespace collision? e.g., the binary package names will collide in the archive. right now you're creating identicle package names for 2 pockets - http://ports.ubuntu.com/pool/main/l/linux-ti-omap4/linux-image-2.6.38-1209-omap4_2.6.38-1209.13_armel.deb
[15:20] <tgardner> 2 releases, rather (natty and oneiric)
[15:20] <apw> they are at different version number however
[15:20] <tgardner> thats the reason I've had to change the ABI series in the past
[15:21]  * apw looks
[15:21] <tgardner> apw, is simple version skew sufficient?
[15:22] <brendand> bjf, sconklin - any news on the Natty -proposed kernel?
[15:23] <apw> 	linux-image-2.6.38-1209-omap4_2.6.38-1209.13oneiric2_armel.deb
[15:23] <apw> tgardner, thats the one in the pool in the PPA, so the filename is different
[15:23] <apw> i thought we needed the abi to be different between all releases in the same pocket
[15:23] <apw> so that the common header don't collide
[15:24] <apw> and update each other
[15:24] <tgardner> apw, there is that as well
[15:24] <apw> ie all the armel common headers would be -10.X and swap about randomly
[15:24] <apw> but version skew is enough if you only have that abi range once as far as i can tell
[15:25] <apw> we can't do it this way going back to lucid for the lts backports as they are elective only
[15:26] <tgardner> the difference with the LTS backports is that the source package name changes
[15:26] <apw> right so we avoid the update
[15:26] <apw> but here we'd want it
[15:27] <tgardner> apw, but we do it with the meta package, _not_ through versioning
[15:27] <tgardner> versioning of the kernel binary, that is
[15:29] <apw> right, but the update would be ok
[15:29] <apw> safe i mean
[15:30] <apw> i could bump abi and release it on .39 comining in, but i dont think we need to
[15:31] <tgardner> apw, AKAIK .39 doesn't work yet, so there isn't much use in releasing on .39
[15:31] <tgardner> AFAIK*
[15:31] <apw> right, that we are waiting  on a working one hense wnting a better .38
[15:32] <tgardner> apw, but I think the ABI issue is separate. I'd be more comfortable with a unique ABI for each release.
[15:40] <apw> tgardner, ok i'll respin it with an updated abi
[15:41] <apw> should be trivial
[15:41] <tgardner> apw, thanks.
[15:41] <sconklin> brendand: No. We're totally dependent on several people who reported regression in the -proposed kernel to help test  , and are not getting responses from them. So it will sit until we get more information.
[15:42]  * apw notes just how close together 'Delete PPA' and 'Package details', somewhat different risks for clicking too
[15:44] <tgardner> sconklin, is that reasonable? how long do we wait on resources we don't own ?
[15:45] <sconklin> tgardner: we have no way of knowing which of several hundred patches caused these, it's not a simple case of a known revert.
[15:45] <sconklin> If we are to expend resources, the next step would be to assign someone to try to come up with reproducers
[16:35] <tgardner> herton, https://bugs.launchpad.net/ubuntu/+source/linux/+bug/794096/comments/8
[16:35] <ubot2> Ubuntu bug 794096 in linux "SMTP and posting to a web-form time out (probably due to netfilter changes)" [Undecided,In progress]
[16:37] <herton> tgardner: I think not, didn't tried reverting that
[16:38] <herton> tgardner: other think I thought about that bug, the guy has a p54 card, and from 9.43 to 10.44 kernel it should have started to work again
[16:38] <tgardner> herton, of the natty commits since the last release, its about the only patch that could have a bad effect on netfilter: 'git log -p --reverse Ubuntu-2.6.38-9.43..HEAD -- net'
[16:39] <herton> so we could also have a conflict in his setup, from one of the logs seems the wifi is associating also on that machine
[16:39] <herton> I'll revert also this commit and ask for him to test
[16:40] <tgardner> herton, I frequently have that situation in Maverick on my laptop, e.g., 2 routes to the same place when I plug in an ethernet
[16:40] <tgardner> herton, didn't lamont complain about a network regression also ?
[16:41] <lamont> I did.
[16:41] <lamont> testing it is somewhat problematic
[16:41] <lamont> both reproducing it reliably to the point of having a good test, and then making time to run the tests on the one place I've seen it thus far
[16:42] <tgardner> lamont, would it be in an environment that experienced fragmentation?
[16:43] <lamont> PPPoE connection, so definitely subject to PMTU discovery and/or fragmentation
[16:43] <tgardner> lamont, could you run one of herton's kernels with a specific patch reverted for awhile ?
[16:44] <lamont> sure
[16:44] <herton> lamont: you're running on lucid correct?
[16:45]  * lamont checks on exactly when he next needs his network to really work
[16:45] <tgardner> herton, it should be a natty kernel.
[16:45] <lamont> herton lucid/i386
[16:45] <lamont> though 'tever on the actual kernel you give me, as long as it's happy with a lucid/i386 userspace
[16:45] <lamont> (the box is i386 only)
[16:46] <tgardner> lamont, I thought we were talking natty. do you have a lucid version that you _know_ works, and the one that doesn't ?
[16:46] <lamont> I have about 60 minutes before I'll need to yank the kernel back to something known-to-work
[16:48] <lamont> 2.6.32-31-generic-pae works, the upgrade from about 4 days before I filed my bug to -32 broke things
[16:48] <lamont> pae or non-pae, doesn't seem to matter
[16:48] <lamont> symptom was "web sites don't work", with what looked to be perfectly normal tcpdump of the traffic
[16:51] <herton> tgardner: lamont filled bug 791512 about it on lucid
[16:51] <ubot2> Launchpad bug 791512 in linux "tcp connections hang in forwarding machine" [Undecided,New] https://launchpad.net/bugs/791512
[16:51] <lamont> yeah - that bug
[16:51] <herton> so it may well be that commit, which should also be in lucid
[16:51] <lamont> herton: any chance you have a lucid kernel with that commit reverted?
[16:52] <tgardner> herton, I'm not seeing that commit in Luciod
[16:54] <herton> tgardner: hmm indeed, so the issue gets more strange, may be it isn't the same bug
[16:55] <tgardner> herton, the only common patches between Lucid and Natty updates are 'dccp: handle invalid feature options length' and 'af_unix: Only allow recv on connected seqpacket sockets.'. The second one deals with fragments.
[16:56] <tgardner> herton, nm, the second patch deals with sequences, not fragments.
[16:57] <tgardner> still, those are the 2 patches that are common between the 2 releases, plus we have similar symptoms on natty.
[16:58]  * lamont tries one of the two kernels he got before
[16:59] <tgardner> herton, how about producing a kernel for lamont to try with those 2 patches reverted?
[16:59] <herton> the dccp change I also can't see how it would cause the issue
[16:59] <herton> tgardner: yep, will do this then
[16:59] <herton> also revert on natty and ask MacSlow to test it
[17:00] <tgardner> yeah, the dccp change looks benign, but you never know... the rest appear protocol specific.
[17:00] <lamont> brb
[17:04] <bjf> ##
[17:04] <bjf> ## Kernel team meeting in one hour
[17:04] <bjf> ##
[17:07] <lamont> linux-image-2.6.32-32-generic-pae_2.6.32-32.63~01spc823e22f_i386.deb <-- this one is fai
[17:07] <lamont> l
[17:07]  * lamont tries the other
[17:08] <apw> bjf, i hope it'll be over in an hour from then :)
[17:10] <bjf> apw, is editmoin working for you since the wiki upgrade?
[17:12] <apw> bjf they say they revoked all the key things, so you'd need to get the new one
[17:12] <apw> if you can remember how
[17:13] <bjf> apw, looking for that info now 
[17:14] <apw> error: body information not found
[17:14] <bjf> apw, indexing must be working, the search returned results, fast
[17:14] <apw> bjf do yoou get htat ?
[17:14] <apw> if so i think thats a different problem
[17:15] <lamont> hrmpf.  did my last bit make it through?
[17:15] <lamont> 585c_i386.deb <-- appears to be win
 sconklin: ^^
[17:15] <lamont> just in case
 linux-image-2.6.32-32-generic-pae_2.6.32-32.63~01spce92585c_i386.deb <-- appears to be win
[17:17] <tgardner> sconklin, what patches got reverted for that kernel ?
[17:17] <sconklin> lamont: ok, those two commits (Sha1 in the version) bound the netfilter changes in the kernel
[17:17] <sconklin> I was hoping one would pass and one would fail. That should take the bisection down to about 3 tries
[17:17] <sconklin> tgardner: ^^
[17:18] <tgardner> sconklin, ack
[17:18] <tgardner> sconklin, are you bisecting the whole mess, or just the net directory ?
[17:19] <apw> bjf, shame that login, and writing to pages is as slow as hell
[17:19] <sconklin> tgardner: well, those were not bisections, just builds from those commits. 
[17:19] <sconklin> hold on let me look at the repo
[17:20] <tgardner> sconklin, that SHA1 includes the nf patches
[17:21] <sconklin> yeah, it appears to be backwards from what I would expect
[17:21] <sconklin> so I have no clue
[17:21] <tgardner> sconklin, I do. your 'win' kernel excludes 'af_unix: Only allow recv on connected seqpacket sockets' which I think is the likely culprit.
[17:22] <sconklin> ok, but not sure that explains the "lose' kernel
[17:23] <tgardner> sconklin, herton is busily building a version with that and one other patch reverted. which one was the 'lose' kernel ?
[17:23] <lamont> linux-headers-2.6.32-32-generic-pae_2.6.32-32.63~01spc823e22f_i386.deb
[17:23] <sconklin> 823e22f was a faila ccording to lamont
[17:24] <lamont> sconklin: note that I wish I had a testcase that was 100% reliable
[17:24] <bjf> apw, did you get your's setup again, I can't find the "user info page"
[17:24] <sconklin> Thanks Lamont - noted that it's not 100%.
[17:24] <apw> bjf,  no it seemed broken and i was relying on you to get there first :)
[17:25] <bjf> apw, i'll get it sorted and get back to you
[17:25] <lamont> fwiw, in about 10 minutes, I go dark-for-testing for a couple hours minimum
[17:25] <apw> heh, how lucky is that
[17:27] <tgardner> lamont, I'll have herton post his kernel in bug #791512 please remember to check for it when you light up again
[17:27] <ubot2> Launchpad bug 791512 in linux "tcp connections hang in forwarding machine" [Undecided,New] https://launchpad.net/bugs/791512
[17:28] <lamont> tgardner: will do.  I'll be able to fetch it and all, just can't get away with rebooting my "test" box for a while
[17:29] <apw> lamont, you'll be trying to convince us you have a day job next :)
[17:29] <lamont> heh
[17:29] <tgardner> lamont, no problem. now I'm off to track down Mirco who had similar issues, but with natty.
[17:29] <herton> lamont: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/791512/comments/4
[17:29] <ubot2> Ubuntu bug 791512 in linux "tcp connections hang in forwarding machine" [Undecided,New]
[17:29] <lamont> test box actually gets used by other people, and they have needs
[17:30] <herton> going to lunch here, will come back in ~30min
[17:32] <lamont> Generating grub.cfg ...
[17:32] <lamont> /usr/sbin/grub-probe: error: raid5rec is not loaded.
[17:32] <lamont> User postinst hook script [/usr/sbin/update-grub] exited with value 1
[17:33] <lamont> this is my other problem child: lucid kernel on a ESXi 4 vm
[17:33] <lamont> with software rade
[17:33] <lamont> raid. wow. where'd thatcome from
[17:34] <apw> dunno but its like totally rad man
[17:35] <apw> ogasawara, are you rebasing to -rc4 ?  i fancy a new shiney kernel, this one is getting old and tired
[17:36] <ogasawara> apw: I am, ran into a minor bump, but test build should be finishing momentarily
[17:36] <apw> ogasawara, yay, then i simply need to wait :)  
[17:36] <ogasawara> apw: I'll ping you when it's ready
[17:37] <apw> i am seeing horrible disk slowness, but i can't work out if that is cause its worse, or just because my bug mailbox has hit 60k messages .... /me applies a liberal dollop of D
[17:37] <tgardner> herton, will you have a kernel for bug #794096 soon? I'll track down Mirco and get him to test it when its ready.
[17:37] <ubot2> Launchpad bug 794096 in linux "SMTP and posting to a web-form time out (probably due to netfilter changes)" [Undecided,In progress] https://launchpad.net/bugs/794096
[17:37] <apw> tgardner, last herton messages says back in 30, 7 mins ago
[17:37] <tgardner> apw, right, I just didn't want him to forget what he was doing
[17:38] <apw> :) ahh
[17:41] <apw> bjf what was that weechat ish thingy called
[17:42] <bjf> apw, weechat
[17:43] <lamont> herton: you do promise this'll boot, right?
[17:44] <bjf> booting is overrated
[17:45] <lamont> bah.  your -headers package fails to install
[17:45]  * lamont reboots again
[17:46] <smb> lamont, Something completely unrelated and mostly out of curiosity: do you know whether the ppa builders run 32bit domU on 64bit servers or only matching architectures?
[17:49] <bjf> apw, got new moin ids from IS but still busted, and they don't support editmoin, so i'm still working it
[17:49] <apw> from is ?
[17:49] <apw> i thoight they were personal, and you got them from your broswer
[17:49] <lamont> herton: your kernel appears to be non-fail
[17:49] <lamont> thank you
[17:50] <bjf> apw, thought so also, still investigating
[17:50] <apw> i think it was pitti who first introduced me to it
[17:50] <herton> tgardner: yes it's built now, will copy it now
[17:50] <lamont> smb: ppa builders are *wavehands* 64-bit everything where they can be (hw supports 64-bit), and 32-bit only where they need to be for hardware.
[17:50] <herton> lamont: hmm ok, so one of the reverts probably worked
[17:51] <lamont> with maybe 5 machines that make that statement a total lie. <-- smb
[17:51] <herton> I'll build just with the af_unix change reverted, and ask you to test again, it's the most likely one
[17:51] <lamont> herton: and dark-to-testing for a couple hours, but can do another one after that if you want to try unreverting one of those
[17:51] <lamont> heh. nice collision there
[17:52] <tgardner> herton, please detail in the LP report _which_ patches were reverted.
[17:52] <apw> bjf, pitti says its broken and he's not managed to fix it
[17:52] <herton> tgardner: ok
[17:52] <bjf> apw, sweet!
[17:52] <smb> lamont, Hm ok. Just asking cause I found "xm save" crashing when trying to save a 32bit domU on a 64bit kernel. And that seemed to be the same at least for the previous abi release.
[17:52] <apw> bjf, see #ubuntu-devel
[17:53] <ogasawara> apw: linux-image-3.0-2-generic_3.0-2.3_amd64.deb on tangerine in my oneiric-amd64 dir
[17:53] <ogasawara> apw: i386 build in progress...
[17:53] <apw> ogasawara, looks exciting 
[17:53] <tgardner> ogasawara, close enough to push ?
[17:53]  * apw updates his amd64 crasher pre-install
[17:54] <ogasawara> tgardner: just booted it on an amd64 box so it must work everywhere :) which means I'm gonna push it
[17:54] <tgardner> good enough for me
[17:54] <apw> ogasawara, its those boots, hardcore
[17:54] <lamont> smb: dom0 and domU are the same on all of them
[17:55] <smb> lamont, Ah, that explains a lot. :)
[17:56] <ogasawara> tgardner, apw: pushed to master-next
[17:56] <bjf> apw, is got the ids from the server logs for my session, i don't know how to find them myself anymore, couldn't find them in firefox
[17:56] <apw> bjf, oh i see, hrm
[17:56] <apw> but they don't work anymore anyhow ?
[17:56] <bjf> apw, right
[17:57]  * apw focuses winning vibes on bjf
[17:59] <bjf> ##
[17:59] <bjf> ## Meeting starting now
[17:59] <bjf> ##
[18:01] <apw> ppisati, should we consider using a specific older version of the compiler for the kernel on arm ?
[18:02] <ppisati> apw: can we switch compiler just for the kernel package?
[18:03] <apw> ppisati, i think it is possible, not desirable normally but possible
[18:03] <jjohansen> ppisati: what was the bug # on that one again
[18:03] <ppisati> lp#791552
[18:04] <ppisati> bug 791552
[18:04] <ubot2> Launchpad bug 791552 in linux "No USB support on beagle/beagleXM" [High,New] https://launchpad.net/bugs/791552
[18:04] <ppisati> in one of the message i attached the discussion where they acked the bug and proposed some workarounds (that didn't work here)
[18:05] <ppisati> apw: goign back to 4.5 as long as this bug is not fixed is ok for me, but i fear that we could miss many early problems if we switch back even for the entire userland
[18:06] <apw> ppisati, this would be only for the kernel
[18:08] <tgardner> we used to use a specific kernel compiler for PARISC
[18:10] <herton> tgardner: http://people.canonical.com/~herton/lp794096/reverts/
[18:10] <herton> the kernel for Mirco to check
[18:10] <apw> tgardner, thats what i thought, i think its a matter of setting GCC and adding the right build-dep
[18:11] <maco> i was just reading victor's blog post. system tap works on ubuntu?
[18:11] <tgardner> apw, yep. I think Hardy has an example
[18:11] <apw> maco, beleive so yes, that is all ckings work, but he is off today
[18:12] <ppisati> apw: if a fix doesn't appera before alpha2, then yes, because wihout usb our SOCs are not that userful: we loose eth, keyboard, mouse, etcetc
[18:12] <maco> apw: ok. the system tap wiki claims it doesnt without lots of hacking around. one of the dinosaur book authors was looking at using ubuntu in the classroom and jumped to fedora after coming across that
[18:13] <apw> maco, i'd not heard that, i would recommend you touch cking tommorrow and confirm he didn't have to jump hoop to make it work, i didn't think so
[18:13] <tgardner> herton, Mirco won't know to try your new kernel unless you update the LP report. since he's subscribed he'll get an email
[18:13] <apw> and if he says its working then we need to find and kill the systemtap errorage
[18:13] <herton> yep, adding a comment there
[18:13] <maco> will do
[18:13] <apw> maco, thanks
[18:16] <ppisati> apw: while here, can we have a meeting about arm kernel and the platform rally?
[18:16] <ppisati> s/arm kernel/arm kernels/
[18:16] <apw> ppisati, sure whatsup
[18:17] <ppisati> apw: well, basically to pull all people involved together and try to define some kind of a schedule
[18:17] <ppisati> apw: e.g. TI first said they were going .39
[18:18] <ppisati> apw: last week said they moved to 3.0 (and they only told us when we asked about an ETA for the BSP)
[18:18] <apw> ppisati, the rally is normally pretty free form, but if you need to get people together thats great
[18:18] <ppisati> apw: but thert's no ETA, and there's no consesus on what to do
[18:18] <apw> ppisati, are you suggesting at the rally or before the rally to prepare for the rally ?
[18:18] <apw> ppisati, if at the rally just add it to the agenda
[18:18] <ppisati> apw: if there's a slot at the rally
[18:19] <ppisati> apw: ok, i'll talk with the arm and linaro people
[18:19] <apw> ppisati, rally has about 10 meetings total there is plenty of time ... normally we just walk to people and sit on their laps till they talk to us
[18:20] <ppisati> apw: cool, than i'll inform all the parties involved
[18:20] <apw> ppisati, i think we already had said we would get with davidm etc and work out what we are doing with the ti kernels
[18:20] <apw> ppisati, but reminding them can't hurt
[18:20] <ppisati> apw: cool, then we are already set
[18:21] <ppisati> but i'll ping people
[18:42] <apw> apw_, oi
[18:43] <apw> apw_, oi
[18:47] <herton> lamont: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/791512/comments/6
[18:47] <ubot2> Ubuntu bug 791512 in linux "tcp connections hang in forwarding machine" [Undecided,New]
[18:48] <herton> when you finish testing the current kernel, just try that one with only af_unix change reverted
[18:52] <lamont> herton: fetching both
[19:35] <apw> apw
[19:38] <lamont> herton: so are we expecting love or hate from r3?
[19:39] <herton> lamont: love, but who knows :)
[19:40]  * lamont stabs the machine to find out
[19:45] <lamont> herton: the kernel is love
[19:45]  * lamont boots in to -33.68
[19:45] <herton> yeah, would be good to know now if -33.68 stock in proposed really has the issue
[19:46] <herton> if 33.68 stock fails, that af_unix patch really is the culprit
[19:50] <Quintasan> http://paste.kde.org/85561
[19:50] <Quintasan> How not normal is the above output is?
[19:51] <apw> Quintasan: depends how long it goes on for and what exactly the machine was doing when it occured
[19:52] <Quintasan> apw: I'm doing linaro-media-create magic on my i.MX53 Quick Start board
[19:52] <Quintasan> last output is Populating rootfs partition. Be patient, this may take a few minutes
[19:52] <apw> and does the machine recover from that ?
[19:53] <Quintasan> apw: It does, but I was unable to interact for a few seconds
[19:53] <apw> if so then its likely you are just hammering crapola out of your usb stick and clogging up everything else
[19:53] <Quintasan> Oh I see.
[19:53] <apw> yep, thats a known "not really meant to react that way", but non-fatal
[19:53] <lamont> herton: rebooting back into r3 to verify that the site that doesn't work now does work there
[19:53] <apw> one shouldn't be able to starve you when doing such things, but you can with a lot of writes
[19:54] <Quintasan> apw: http://paste.kde.org/85567
[19:55] <herton> lamont: ok, so -33.68 stock failed then?
[19:55] <Quintasan> apw: I was also wondering if you could tell me what's this
[19:56] <apw> Quintasan: that implies the disk host controller lost the link to the drive and reset the link a couple of times to recover it
[19:57] <Quintasan> apw: What can be the cause of that?
[19:57] <apw> does it happen often, if it does something is up, could be a bug, could be bad h/w
[19:57] <apw> very hard to diagnose, unless its common on one kernel and not on another, then we might be able to acertain why
[19:58] <apw> i think its saying there was a parity type error on the transaction on the bus
[19:59] <lamont> herton: 33.68 stock is fail
[20:00] <lamont> though the test website did magically work on that one.  the other one was solidly fail, and works on r3
[20:00] <lamont> so I have r3 live
[20:01] <herton> hmm. well, seems that change then is causing things to break
[20:01] <herton> lamont: can you post your testing results on bug 791512?
[20:01] <ubot2> Launchpad bug 791512 in linux "tcp connections hang in forwarding machine" [Undecided,New] https://launchpad.net/bugs/791512
[20:03] <Quintasan> apw: Okay, is it possible that cable is bad?
[20:03] <lamont> herton: will do
[20:03] <apw> Quintasan: it is _possible_ if you have a spare it may be worth changing.  it depends how often you see that, if its once ever i'd not bother
[20:03] <Quintasan> more like three times a day
[20:03] <Quintasan> :S
[20:04]  * Quintasan goes to buy a new SATA cable
[20:05]  * jjohansen -> lunch
[20:10] <apw> tgardner-afk: you applied this 0726 commit to hardy/master and lucid/ti-omap4, but not to maverick/ti-omap4 ... is there a reason or just an oversight
[20:11] <tgardner> apw, checking
[20:12] <tgardner> apw, doh! I just missed Maverick ti-omap4, gimme a sec
[20:13] <apw> tgardner: heh np, the matrix _is_ some use after all
[20:15] <tgardner> apw, done
[20:18] <apw> tgardner: ok i've pushed up a replacement oneiric ti-omap4 and tag (also removed the old tag) the tree diff is just the update script change and abi+100
[20:18] <apw> tgardner: so i am inclined to just upload it
[20:20] <tgardner> apw, looks good to me
[20:21] <apw> tgardner: ok thanks :)
[20:22] <apw> ENO-bjf
[20:39] <herton> tgardner: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/791512/comments/7
[20:39] <ubot2> Ubuntu bug 791512 in linux "tcp connections hang in forwarding machine" [Undecided,New]
[20:39] <herton> seems that af_unix change is really at fault
[20:40] <tgardner> herton, good. how about annoying the upstream authors? perhaps mention to the various stable maintainers that we're gonna revert it.
[20:42] <herton> ok, I'll try to enter in contact with patch author and CC netdev, lets see what they have to say
[21:46] <tgardner> ogasawara, did you have to use some magic to get oneiric to build ? I think ubuntu/rtl8192se is borken.
[21:46] <ogasawara> tgardner: no magic needed for me
[21:46] <ogasawara> tgardner: used the usual build scripts
[21:47] <tgardner> ogasawara, hmm, a rebuild with that commented out in ubuntu/Makefile seems to be working better.
[21:48] <ogasawara> tgardner: what was the error you were seeing?  my build log is on tangerine:oneiric-amd64/build.log
[21:49] <tgardner> ogasawara, I've over written it already, but it was complaining about EXTRA_CFLAGS
[21:50] <tgardner> ogasawara, I've updated the chroot on tangerine since your build. perhaps you should try again.
[21:51] <ogasawara> tgardner: ah, will kick off another build then
[21:56]  * tgardner --> EOD