[00:01] Package: linux-image-2.6.15-51-686 [00:01] Version: 2.6.15-51.64 [00:01] that's not the one with the fix [00:01] it was fixed in 51.65 [00:01] make sure you have all the latest updates from dapper-security [00:01] reboot, and retest [00:02] i do have all the updates from dapper-security... [00:05] either you're mistaken, or that upload has not been released yet...either way, you aren't testing the version that has the fix [00:05] I'm checking the upload status, give me a minute [00:05] thanks === reynaldo_ is now known as reynaldo [00:09] ln-: looks like that upload is pending [00:09] ln-: kees would know more [00:13] do you think there's something wrong in ubuntu's "process," if a thing broken in Oct 2006 doesn't get fixed by Feb 2008, even if the fix is known and is trivial? [00:15] ln-: considering there's a work around and there is no data-loss involved, it's a rather trivial issue to begin with [00:15] ln-: kees handles security, so you really should talk to him === doko_ is now known as doko [00:16] i'll try to ping him again tomorrow. [00:25] ... or why bother. [00:27] ln-: well, if there's is an issue with the process, kees would be willing to explain or possibly improve [00:27] ln-: I'm sure he would appreciate input, or being put on the spot :) [00:28] ok, i'll ping him then.. [00:49] laga: somehow this kernel has problems to be used with live-helper [01:04] so something else is wrong.. === asac_ is now known as asac === antdedyet is now known as E_NOMAD === E_NOMAD is now known as antdedyet === antdedyet is now known as technomad === technomad is now known as technoant === technoant is now known as technogeek === technogeek is now known as technoant [08:34] moin [09:13] hi, current aufs in lum works for me [09:13] but [09:15] with standard ubuntu config (too many old legacy drivers) i get this [09:15] http://paste.debian.net/48033 [09:15] which results in no dma for hdc [09:16] a pure libata kernel has no problem at all with it... [09:18] also, i see no reason to try blacklists on those 2 modules, debian has even bcm43xx completely disabled, you could do that too [09:18] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/185719 [09:18] Launchpad bug 185719 in linux "remove duplicate pci ids bcm43xx/ssb prism54/p54pci" [Medium,Triaged] [09:18] as soon as tools like discover are installed blacklisting does not work [09:19] hostap_pci is completely unstable btw. dont know why [09:19] p54pci works fine [09:26] did anybody try ath5k? [10:16] hey juliank [10:17] juliank: aufs doesn't work with nfs branches. it was complaining that i need to apply the lhash patch (and add the config option) [10:18] laga: Are you running the latest git kernel? [10:18] hi laga , well it worked for live mode, but you can add those ;) [10:19] juliank: "latest" as in "pulled right after rtg committed your patches", yes. [10:19] Kano: great :) [10:19] did you get the link in pm? [10:19] yes [10:20] it has some problems shutting down, but before i got kde boot errors [10:20] haven't tried it yet, i don't need live disks at the moment :) [10:20] oh [10:20] you can use vbox or vmware to boot it too [10:20] laga: the latest ubuntu-hardy.git + ubuntu-hardy-lum.git should work [10:20] you will see a huge difference when you boot with union=unionfs [10:20] Kano: speed-wise? [10:20] then network-manager does not work [10:21] i dont get why the same kernel works on hardy with networkmanager [10:21] with unionfs [10:23] maybe somebody changed networkmanager [10:23] my nick comes up #2 in google when you search for "unionfs sucks", so don't expect an answer from me [10:24] good point ;) [10:25] note: i've changed my opinion since then. current unionfs is probably just fine with a vanilla kernel, but i guess i'll never find out :) [10:25] well not for me, with the test i did... [10:27] juliank: the changes which have been committed since then are not relevant - except for the missing entries in debian/config/i386 which i added myself [10:28] i even had the strange issue with unionfs that in a vm a simple du -h /dir returned an error - a real pc did not have that error however [10:28] odd. [10:28] laga: do cat /sys/fs/aufs/config [10:28] unionfs doesn't work well with bonnie either. not sure what's causing that. [10:28] could never trace it why this happend, aufs worked [10:31] laga: what does not work? build or mount? [10:31] juliank: mounting. [10:31] juliank: i'm booting the VM now.. [10:31] it builds, thats not the problem [10:31] do you want a sample iso? [10:32] Kano: You have an ISO? [10:32] of course [10:32] Kano: size? [10:32] with that kernel, but based on etch [10:32] live-helper created [10:32] until some other things are patched in u i will not use it [10:33] juliank: CONFIG_AUFS=m, CONFIG_AUFS_FAKE_DM=y, CONFIG_AUFS_BRANCH_MAX_127=y, CONFIG_AUFS_SYSAUFS=y, CONFIG_AUFS_RR_SQUASHFS=y [10:33] juliank: that's /sys/fs/aufs/config [10:35] if fails with the error message: aufs test_add:357:exe[2033]: NFS branch is not supported, try some configurations and patches included in aufs source CVS [10:35] laga: I'm fixing it. [10:36] juliank: great. [10:36] well you would need to patch+compile the kernel first to try [10:37] patch what? [10:37] Kano: The Ubuntu kernel has the patches. [10:38] juliank: fine, then it should be easy to update just lum [10:38] a few seconds ;) [10:39] Kano: One variable name is wrong, that's all. [10:40] good, can update lum with ease [10:40] scripted ;) [10:40] laga: Do you have a build log of l-u-m? [10:41] juliank: no, but i can make one real quick :) [10:41] laga: I'm building [10:43] laga: try to add KSRC=$(PWD) to the top of ubuntu/fs/aufs/Makefile and build it. [10:43] also could you enable aufs for powerpc too? [10:43] laga: I will send a patch soon [10:44] Kano: Soon. [10:46] did somebody have got no framebuffer support on amd64 live cd or is it just me? [10:47] i386 live cd has a splash [10:47] amd64 not [10:48] juliank: what do you think about a blacklist option for casper, i have that for kanotix [10:48] it writes to /etc/modprobe.d/custom-blacklist [10:48] before udev starts [10:52] juliank: i don't think it's working [10:52] same message [10:52] laga: I need a build log. [10:54] juliank: http://www.pastebin.ca/884456 [10:55] juliank: here's the KSRC patch i added: http://www.pastebin.ca/884457 [10:57] laga: rebuild with KSRC=$(srctree) [10:58] am i supposed to use that string literally or should i expand srctree? [10:59] laga: as written. it's a variable [11:05] oops. that's gonna take a few minutes, just deleted the squashfs image for my diskless client. [11:12] juliank: ok. CONFIG_AUFS_LHASH_PATCH and friends were picked up properly, but i'm still getting that error message. [11:12] juliank: ok. CONFIG_AUFS_LHASH_PATCH and friends were picked up properly, but i'm still getting that error message. [11:13] juliank: i don't see a config option for the put_filp patch in sysfs [11:13] laga: /sys/fs/aufs/config [11:14] laga: output of grep '^.*[[:space:]]put_filp[[:space:]]vmlinux[[:space:]]EXPORT_SYMBOL' Module.symvers [11:14] ? [11:14] laga: sorry, typo [11:14] AUFS=m, AUFS_BRANCH_SYSAUFS=y, AUFS_RR_SQUASHFS=Y, AUFS_SEC_PERM_PATCH=y, AUFS_SPLICE_PATCH=y, AUFS_KSIZE_PATCH=y [11:15] i left out the CONFIG_ part. [11:17] juliank: what do you need grepped? [11:17] laga: not needed. I know the problem now. [11:17] juliank: great [11:19] laga: the patch is at http://www.pastebin.ca/884477 - [11:24] thanks, building now [11:29] juliank: awesome, it's working. [11:29] laga: I'll send the patches to kernel-team [11:29] juliank: great [11:33] laga: patches sent [11:33] :) [11:34] it takes 27s from power-on to having X loaded with aufs in my diskless environment [11:34] that's very nice IMHO :) [11:34] laga: Could you try to use the rr flag for the nfs root? I need to know if this works. [11:35] laga: in the aufs mount options [11:35] juliank: i'm not using nfs root, i'm using a squashfs root with a nfs share on top. still want it? [11:36] laga: your nfs root is specified in dirs= or br=, then yes [11:37] laga: s/root/share/ [11:37] ok [11:39] I always kill my ssh proxy because I close the window ssh is running in. [11:40] juliank: well.. aufs does not fail, but of coruse the system won't boot because / is read-only :) [11:41] laga: that's good. [11:41] thanks [11:41] exit [12:20] laga: how about exporting it again via nfs and boot via network? === \sh_away is now known as \sh === gaspa_cribbio is now known as gaspa === technoant is now known as antdedyet === \sh is now known as \sh_away === \sh_away is now known as \sh [19:14] rtg: hi, didnt you say you remove the pci ids when i report the bug? [19:32] Kano: i think if you increase the frequency of your questions it'll get fixed faster [19:36] rtg: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/185719 [19:36] Launchpad bug 185719 in linux "remove duplicate pci ids bcm43xx/ssb prism54/p54pci" [Medium,Triaged] [19:37] i dislike that idea with blacklisting [19:55] laga: best tell him that this change is needed, also you could forget the bcm43xx patch when you disable that module (same for prism54), debian disabled bcm43xx [19:56] bye [19:59] laga: so what does Kano think is wrong with blacklisting? [20:03] i have no clue [20:03] in that ticket, he says that it'd break if "discover" was installed === \sh is now known as \sh_away [22:25] rtg: what version of ALSA is in the stock kernel? [22:26] amitk_: dunno, some mix of 1.0.15 plus. I've removed it from the kernel for -6.10 and put it in lum (1.0.16rc2) [22:28] rtg: aaah... [22:28] rtg: are we updating to a newer version of alsa in lum? [22:28] amitk_: is that the sound of something stuck in your throat ? :) [22:29] hehe [22:29] amitk_: I pulled ALSA this AM, so the version in lum is very current. [22:29] Its actually -rc2 plus 8 or 9 more patches. [22:30] amitk_: by the way, I'm still build testing... You mileage may vary. [22:33] Jaroslav (employed by Red Hat now) anticipates releasing 1.0.16 final next Tuesday.