=== doko_ [n=doko@dslb-084-059-078-133.pools.arcor-ip.net] has joined #ubuntu-kernel [12:36] jbailey: there are already [12:36] linux-headers-2.6.15-6-386, for example [12:39] mjg59: ping [12:42] ispiked: Hi [12:43] mjg59: I hate to bug you, but I'm wondering why fn+f1 doesn't put my laptop to sleep. the keycode appears to be registered, and I could've sworn this used to work. [12:50] ispiked: What hardware? [12:50] mjg59: the same as you. [12:50] mjg59: dell d610. [12:50] Have you got the latest acpi-support ? [12:51] yeah. did that break it? === dilinger [i=dilinger@mouth.voxel.net] has joined #ubuntu-kernel [12:51] I'm pretty sure I can suspend using logout > suspend with the latest acpi-support package. [12:52] I'm headed out now. be back later. [12:53] Hm. Not sure, then. === trevilor [i=trevilor@p549CDD4C.dip.t-dialin.net] has joined #ubuntu-kernel [01:28] BenC: So, is the idea to switch to the libata PATA drivers where possible? === mgalvin [n=mgalvin@cpe-69-205-38-37.nycap.res.rr.com] has joined #ubuntu-kernel [04:23] mjg59: where is that button's action configured? [04:23] ispiked: /etc/acpi/events [04:25] mjg59: same deal with the logout action? [04:25] http://rafb.net/paste/results/lcTK1z80.html [04:25] ispiked: Uhm. Which logout action? [04:26] "suspend the computer" [04:26] Oh, the Gnome one? No, that triggers gdm to execute pmi [04:27] mjg59: what is that? [04:27] It's a script that ends up executing /etc/acpi/sleep.sh, so it's broadly the same [04:34] it'd be nice if scroll lock would work (with the light, too). [04:34] might file a bug for that. [05:18] BenC: ata_piix + added PCI ID drives my Thinkpad fine [06:06] morning === JaneW [n=JaneW@196.36.161.235] has joined #ubuntu-kernel [08:21] mjg59 : Who do I complain to that my 'screen close' button no longer triggers a screen lock? === jane_ [n=JaneW@196.36.161.235] has joined #ubuntu-kernel === chmj [n=chmj@dsl-146-143-211.telkomadsl.co.za] has joined #ubuntu-kernel === dhpeterson_ [n=dhpeters@ppp8-85.lns1.syd7.internode.on.net] has joined #ubuntu-kernel === dhpeterson_ is now known as dhpeterson [09:16] hi ... breezy install fails for me ... it's a uhci_hcd issue ... can anyone here help? [09:16] i have checked ubuntu bugzilla and debian BTS and it's not a dupe [09:16] i want to raise a bug report but i don't know which package and where to do so, etc [09:27] file it on bugzilla.ubuntu.com against 'linux' [09:27] okay [09:27] what severity ... both hoary and breezy installers don't complete. is that a critical? [09:27] btw same for the debian sarge installer [09:28] interestingly, knoppix 2.6 _does_ boot successfully [09:28] Does our livecd boot? [09:29] neither live nor install works [09:29] Saying "a livecd boots" and "the installer fails" don't relate much. [09:29] live cd and install cd fail in the same place [09:29] Okay, if the livecd doesn't work either, then please, file a bug, lots of fun debug output, blah blah. [09:29] sorry if i was ambiguous [09:29] Severity is normal. [09:30] okay i will prepare a bug report, thanks [10:25] BenC: [10:25] [ 1067.366233] bcm43xx: Invalid PHY Revision 7 [10:25] [ 1199.264579] bcm43xx: FIXME: Possibly broken code in bcm43xx_phy_setupg() at drivers/net/wireless/bcm43xx/bcm43xx_phy.c:346 [10:25] the driver loads and the board is initialized [10:25] but i can't get it to associate with my AP [10:26] i think i am getting that error when trying to force to use 11M [10:27] it's probably the model that's not fully supported for my understanding of the code === CataEnry [n=cataenry@host149-23.pool8250.interbusiness.it] has joined #ubuntu-kernel === jbailey [n=jbailey@modemcable139.249-203-24.mc.videotron.ca] has joined #ubuntu-kernel === CataEnry [n=Enrico@host149-23.pool8250.interbusiness.it] has joined #ubuntu-kernel === trevilor [i=gnu@p549CE2DA.dip.t-dialin.net] has joined #ubuntu-kernel [02:04] fabbione, BenC: Next time one of you are twiddling the kernel configs, CONFIG_CRAMFS=y can probably go away now. [02:05] Need to double check with Colin, but I don't think d-i uses it at all. [02:10] ok [02:11] hey BenC [02:11] i was just talking with Kamion [02:11] i did fix a couple of kernel d-i things for him [02:11] mind to pull from my repo? [02:12] doing that now [02:12] thanks [02:14] hmmm [02:14] if you have time it would be a good idea to fire up a full build on sparc [02:14] mine is chaoking on OOo2, gcc-3.4 and gcj-4.1 [02:14] these d-i changes are always hairy without a build [02:14] the other arches are ok [02:15] I do full builds on all 6 arches before release [02:16] ah cool [02:16] even better [02:16] sparc64 is done in 12 minutes, so it's easiest :) [02:16] BenC: i was wondering.. can i get access to your sparc somehow that doesn't distrurb you? [02:16] disturb [02:16] yeah, I can try setting that up today [02:16] since you use the box for testing [02:17] i could just use it while you sleep to push some big pkgs to a fine tuned buildd there [02:17] and skip stuff like gcc-* here [02:17] i will send you my ssh key [02:17] no need to run TODAY [02:17] sometimes, when you have time [02:18] sure thing, I'll setup dchroot for you and sudo access inside that chroot [02:18] nah [02:18] i will send you a chroot :) [02:18] ok, just send a url, and I'll unpack the chroot :) [02:20] BenC: sure.. i just need to pack it first ;) [02:21] don't forget your rootkit :) [02:21] of course :) [02:22] it's funny when people tell us: "no you can't get access to my machine to debug this kernel problem.." [02:22] they really make me laugh [02:22] they don't realize we do actually build that kernel for them :) [02:22] we have kernel privileges! [02:22] and they don't see it :P [02:23] yeah, lol [02:24] "if I wanted your machine, it'd already be DoS'ing the whitehouse by now" [02:25] https://wiki.ubuntu.com/FabioMassimoDiNitto <- [02:25] that was like: zul meets fabbione, fabbione meets zul === BenC lol === zul [n=chuck@ubuntu/member/zul] has joined #ubuntu-kernel [02:43] hey sorry i havent been around. [02:43] BenC: chinstrap:/home/fabbione/sparc-chroot-dapper.tar.bz2 [02:43] BenC: and i am off for a nap :) [02:43] hey zul [02:44] later fellas [02:44] cya fabbione [02:46] bye fabbione [02:46] hey fabbione, bye fabbione [02:56] infinity: Oh, yeah, that one's my fault === mgalvin [n=mgalvin@ubuntu/member/mgalvin] has joined #ubuntu-kernel [03:13] BenC: About? [03:13] yeah [03:14] mjg59: pong [03:15] BenC: What's the long-term plan with all these libata PATA drivers? [03:15] see how they go, and if they start breaking shit, back them out [03:15] unless I can fix them (or jgarzik can help me) [03:16] The obvious breakage is that drives will move from hda to sda [03:16] I don't think he's keen on me using them in the dist [03:16] hmm [03:16] On the other hand, it means we can do hotswap properly [03:16] upgrades will get messy though [03:17] maybe I can default the combined_mode to IDE === CataEnry [n=Enrico@host149-23.pool8250.interbusiness.it] has joined #ubuntu-kernel [03:17] until lately I hadn't realized how much of a mess things were between ide and ata/sata/pata [03:21] what do you suggest? [03:22] mjg59: re: s/hda/sda/ - I wonder if there's any reasonable way to move systems to by-uuid for booting and mounting? [03:23] I've always liked that method [03:23] Part of udevroadmap is to do that for new installs in d-i anyway, so it's just transitioning people to the New World Order. [03:23] same thing with iftab for network interfaces [03:23] I don't know iftab [03:23] BenC: The libata drivers are pretty untested, but /ought/ to work [03:23] iftab gives network interfaces their names based on mac address [03:23] But when we upload a kernel with them in, there's going to be breakage [03:24] I think I'll disable them for now, until we get this sorted out === CataEnry [n=Enrico@host149-23.pool8250.interbusiness.it] has joined #ubuntu-kernel === lamont [n=lamont@mib.fc.hp.com] has joined #ubuntu-kernel === CataEnry [n=cataenry@host149-23.pool8250.interbusiness.it] has joined #ubuntu-kernel [03:51] mjg59 : Is this new behaviour intentional, or a bug? :) [03:51] mjg59 : (I've always used my "screen close" button as a quick shortcut to "I'm walking away from the laptop for 5 minutes, lock the screen") [03:53] infinity: bug [03:53] I fucked up the environment variable check === BenC [n=bcollins@debian/developer/bcollins] has joined #ubuntu-kernel === chmj [n=chmj@dsl-146-143-211.telkomadsl.co.za] has joined #ubuntu-kernel === CataEnry [n=cataenry@host149-23.pool8250.interbusiness.it] has joined #ubuntu-kernel === ispiked [n=ispiked@Lee-12-238.rh.ncsu.edu] has joined #ubuntu-kernel === jbailey_ [n=jbailey@modemcable139.249-203-24.mc.videotron.ca] has joined #ubuntu-kernel [04:50] BenC: Just had my X hang on ppc64 again. Interesting that this time a kill -9 of Xorg doesn't make it go away. [04:50] BenC: Anything I can usefully collect for you for a bug report before I hit the shiny switch? [04:50] check dmesg [04:51] and "cat /proc//wchan" if the X pid [04:51] no oops in dmesg [04:51] I have a feeling you are hitting a drm issue [04:51] I did the cat a dozen times, each in .__start [04:52] hmm, that's an odd place [04:52] is the process in D state this time? [04:52] 5603 root 25 0 0 0 0 R 100 0.0 11:06.14 Xorg [04:52] (from top) [04:52] It was taking up a bunch of ram before I kill -9'd it. [04:52] is that 100% CPU usage? [04:53] Yes. [04:53] all the memory usage is gone, and it's hanging [04:53] that's an odd one :) [04:53] Shall I strace it? [04:53] yeah [04:53] # strace -p 5603 [04:53] attach: ptrace(PTRACE_ATTACH, ...): Operation not permitted [04:54] Should I try the ppc32 kernel to see if it recurs there? === jbailey [n=jbailey@modemcable139.249-203-24.mc.videotron.ca] has joined #ubuntu-kernel [05:19] infinity: ping [05:39] pong... ish. [05:39] 3:30am and heading to bed. [05:39] So, uhh.... make it quick. :) === mdz_ [n=mdz@studiocity-motorola-bsr1-70-36-194-85.vnnyca.adelphia.net] has joined #ubuntu-kernel [05:47] BenC : Alright, whatever you were pinging for, you missed your chance. ail me, or catch me tomorrow. === infinity -> bed, finally. [05:47] hey [05:48] sorry [05:48] will dapper kernels work on breezy? [05:48] Oh. You are here. [05:48] i'm gonna try getting this smp opteron box stable w/ some 2.6.15 love [05:48] On breezy servers or desktops? [05:48] infinity: next time prepend "BenC" to the pong so I actually get an audio feedback :) [05:48] udev will probably blow up on desktops. [05:48] infinity: both [05:48] infinity: so i'd just need to backport udev? [05:48] dilinger: you'll have to sync a few packages to dapper aswell [05:49] mainly udev, and ditch hotplug [05:49] BenC : Well?.. :) [05:49] ok [05:49] thanks [05:49] i'll give it a shot on my laptop, first [05:49] and once i'm locked out of a usable system, i'll try the opteron server ;) [05:49] infinity: should/could I add server-{low,high}end to lrm? [05:50] Hrm. [05:50] Are we building wireless drivers for -server targets? [05:50] dilinger: you should be able to add dapper to sources.list, and install the kernel, and then remove it [05:51] infinity: right now, -server is just a tuned version of the normal targets (HZ, disable-preempt) [05:51] same hw support [05:51] I assume that might change? [05:51] For now, I wouldn't bother. We'll discuss it later. [05:51] highend supports numa/bigsmp/summit/es7000 targets [05:51] ok [05:51] If we want atheros and acx_pci support, we'll have to do it. [05:52] I may want to tweak things to not bother including the video crap on -server targets. [05:52] well, all the kernel video crap is in there too [05:52] I hope that will go away. [05:52] me too [05:53] But, yeah. Let's discuss it tomorrow. [05:53] If you're doing an LRM/linux-meta ABI bump tonight, just leave restricted out of the server stuff in linux-meta for now. [05:53] We'll sort it later when we're sure what we want. === ..[topic/#ubuntu-kernel:BenC] : Ubuntu kernel development discussion ONLY | New git tree for dapper: https://wiki.ubuntu.com/KernelGitGuide | 2.6.15-7.9 uploaded (The "My momma made me do it" release) [05:54] ok [05:54] ehhehe [05:54] cool [05:54] new crack :) [05:54] BenC: btw.. it's GO on ppc.. or at least it was last pull i did [05:54] dunno if you read the scrollback === ispiked [n=ispiked@Lee-12-238.rh.ncsu.edu] has joined #ubuntu-kernel [05:54] i did also test the Airport Extreme2 driver [05:54] but it fails on me [05:55] mjg59 got it working [05:55] BenC: are you getting any feedback on CD-ROM/DMA issues since we switched to 2.6.15? [05:55] you have to do some manual shit [05:55] fabbione: I had to do: [05:55] mdz: zero [05:55] interesting [05:55] only thing I've heard about cdrom was the atapi stuff with sata controllers [05:55] and that was a seperate issue [05:56] I thought we'd fixed that long ago [05:56] iwconfig foo essid bar; iwconfig foo channel baz; ifconfig foo up; iwlist scan; iwconfig foo essid bang [05:56] BenC : Keybuk fed you a patch for ide stuff that fixed mdz's DMA issues, no? [05:56] infinity: yes [05:56] (And the CDROM on sata thing should be fixed in this upload too, right?) [05:56] BenC: is that in -7.9? [05:56] mjg59: trying now.. but i have no idea what chan i am using [05:56] it wasn't dma directly, it was ide-generic/modular-ide crack [05:56] mjg59: how can i figure that out? [05:57] BenC: is modular ide a bad idea? why isn't it upstream? [05:57] mdz_: IDE maintainer is a pain [05:57] BenC : For LRM's ABI bump, you should be able to just bump "abi_version" in debian/rules, add a new changelog entry, and build the source package. (make sure you have kernel-wedge installed) [05:57] BenC : Should be smooth sailing. [05:57] breezy's ide was that it loaded all the ide controller drivers, and then ide-generic, and let them fight it out, then removed any that didn't win [05:57] that doesn't work for dapper [05:58] infinity: yeah, I have it ready to upload once I see a success report for atleast one arch :) [05:58] BenC : Rock. [05:58] Kay, I'm off to bed, then. We'll revisit the LRM-on-server-kernels thing later. [05:58] mdz: the way breezy handled drivers grabbing devices was bad [05:58] mjg59: my T42 with 2.6.15-6 failed to come back from STR today [05:59] dapper will low the correct drivers instead of doing a free-for-all [05:59] s/low/load/ [05:59] mdz_: When suspended, did it have the hard drive LED on? [05:59] (previously /sys/power/state was empty, but that's fixed now) [05:59] mjg59: it wasn't on when I tried to resume at least [05:59] if anything doesn't get loaded, or is loaded incorrectly, Keybuk assures me he will correct it in initramfs :) [06:00] the moon light was on, though the fan was still running [06:00] mdz_: If the fan was still running, it hadn't suspended [06:00] or maybe it was just radiating heat [06:00] anyway it was hot [06:00] Right [06:00] Sounds like it froze on the way down [06:00] I've had my Thinkpad do something like that. I'll look into it. [06:00] any testing I can do? [06:01] mjg59: do you get this in your dmesg: [26120.102096] bcm43xx: FIXME: Possibly broken code in bcm43xx_phy_setupg() at drivers/net/wireless/bcm43xx/bcm43xx_phy.c:346 [06:01] mdz: btw, I don't think we are going to do kernel-wedge and kernel-package merging for dapper [06:01] See if it's reproducible? [06:01] fabbione: Yeah [06:01] mjg59: hmmmm [06:01] BenC: something scary there? [06:01] what channel should i use? [06:01] things work for us now, and kernel-package 10 in debian is way way way different [06:01] fabbione: Whatever channel your AP is on [06:01] mjg59: i have no clue... [06:01] we have patches in wedge and package for powerpc stuff, and custom update-initramfs usage [06:02] fabbione: iwlist scan should tell you === fabbione tries again from scratch [06:02] mdz: it's not impossible to merge those into the new packages, just that there's no pressing need to do so, that I can see [06:02] BenC: bugfixes? [06:02] that's the usual reason [06:03] most of the bugs fixes between our version and 10 in debian, is fixing things that broke when they split the debian/rules file up [06:05] mdz: -7.9 has -server kernels for i386, btw [06:05] BenC: so it'll need queue/new processing on i386? [06:05] yes [06:06] it's an abi bump anyway [06:06] ah, right [06:06] BenC: what's new and cool in the server flavour? [06:06] Keybuk: could you take a look at debian #342057, BusLogic not loaded because !sysfs (seems needed for vmware testing of initramfs-tools). [06:06] abi tracking will start with this upload too [06:06] mdz: HZ is down to 100 (suggested for server), and preempt is disables (enabled for desktop) [06:06] that's for low/high end server [06:07] highend also has generic bigsmp/numa/summit/es7000 support [06:07] 64CPU max, and 64Gig max [06:08] BenC: sounds like material for https://wiki.ubuntu.com/DapperReleaseNotes [06:08] BenC: please add an entry there under the server section [06:08] ok [06:09] mdz_: the -server kernels will land only on the -server CD [06:09] we will install by default the low end, with high end available [06:10] would be nice if we could detect highend [06:12] mjg59: do you have a 54M net at home? [06:12] fabbione: No [06:12] did you force the speed? [06:12] Yup [06:12] Sorry, forgot to mention that === fabbione hits mjg59 [06:12] ;) [06:12] It doesn't do speed autonegotiation properly yet [06:12] where in that process? [06:12] or just after [06:13] my ap doesn't feel the mac at all [06:14] actually [06:14] did you use dhcp on top? [06:14] or just static === fabbione mumbles a bit [06:16] Either [06:16] dmesg should show that it's authenticated if it's been successful [06:16] mdz: added [06:17] [ 450.499743] SoftMAC: Sending Authentication Request to 00:04:e2:d9:0a:dc. [06:17] it doesn't when i am in 11MB [06:17] fabbione: I think it ought to get a response, too [06:17] that's my neibourgh AP [06:17] 802.11g speeds currently don't work [06:17] that's at 54MB [06:17] Set it to 11 and then set the essid again? [06:17] o [06:17] maybe the AP is configured for g-only? [06:19] nope [06:19] i don't see any outgoing request to/from my AP [06:19] not even in dmesg [06:19] BenC: my ap is 11 only [06:19] it definetely can't do more than that [06:19] ah [06:19] fabbione: Does iwconfig show the correct essid and frequency? [06:19] i was trying my neighbour only becuase it's 54MB :) [06:20] i can see the frequency keep changing [06:20] otherwise the essid is fine [06:20] Uhm. [06:20] i wonder if the fact that i am not broadcasting the essid might confuse the driver [06:20] Have you explicitly set the channel? [06:20] yes [06:20] Yes, that might currently break it, I guess === fabbione changes [06:22] *cough*fuck cisco*cough* [06:25] i can't even find how to do it again [06:40] woo [06:41] mjg59: i got the AP to see the interface and viceversa [06:41] Any joy? [06:41] but no IP joy [06:41] Cool [06:41] [ 557.190722] SoftMAC: Sending Authentication Request to 00:02:8a:21:18:59. [06:41] Try configuring it manually rather than using dhcp? [06:41] yes [06:41] that's what i am doing right now [06:43] no joy either [06:43] it complains about authentication [06:43] but i have no authentication config on my AP [06:43] only filters by mac-address [06:43] so it's open, no wep/wap? [06:45] nope [06:47] if i run iwscan list a few times in a raw i see the AP at intervals.. [06:47] but that might be normal [06:48] How does it complain? [06:51] BenC: getting bugreports on debian initramfs-tools of !sysfs drivers that are fixed in your tree? [06:51] SoftMAC: Rates not sorted! [06:52] SoftMAC: Authentication timed out with 00:02:8a:21:18:59 [06:52] BenC: could you push those to linus, are they already in mm? [06:52] makx: uh, say that again? :) [06:52] I'm not sure what you mean [06:53] buslogic has sysfs support in your tree for example. :) [06:53] mjg59: i will fight with it tomorrow again.. gotta cook dinner [06:53] s/sysfs/module-device-table/ you mean? [06:53] indeed magic keyword. [06:54] it's not proper support, it's just a hack to get udev/hotplug to load the driver [06:54] for linus to take it, the driver(s) would also need to be converted over to the PCI callbacks for hotplug support [06:55] right now they only probe when loaded, if you insert a new device, the module has to be unloaded/loaded again to probe it [06:55] these are the same things we had in breezy [06:55] BenC: Is he being all prissy now about correctness? Sheesh. ;) [06:56] it's a convenience, but IMO, it's lying to have a module-device-table and not be able to support the hotplug callbacks for which is was designed for :) [06:57] makx: what are the chances that debian would want to clone ubuntu's git tree, and sync non debian/* changes? [06:58] the major thing would be some of the non-free stuff that debian would want to ignore (and purge from it's branch) [07:04] BenC: i assume the git tree includes all the external drivers and patches? [07:04] debian doesn't really want to maintain that stuff, they'd only be interested in bugfixes === CataEnry [n=Enrico@host149-23.pool8250.interbusiness.it] has joined #ubuntu-kernel [07:14] Hi === jbailey_ [n=jbailey@modemcable139.249-203-24.mc.videotron.ca] has joined #ubuntu-kernel [07:18] BenC: you didn't pull the last changes to d-i? [07:23] they are not in the changelog === fabbione goes for dinner [07:29] BenC: if you can do that Kamion will love you [07:29] perhpas fire up a 7.10 right away? [07:29] bbl [07:29] I did a pull and saw two changes from you in git-log [07:29] yes [07:30] but the changelog in the archive doesn't have them [07:30] * Stop providing ext2-modules and ship it for all arches. [07:30] not sure why, did you commit it? [07:30] it's missing [07:30] oh crap [07:30] ext2-modules is not listed in debian/control, so I assume all is well [07:31] i didn't commit the changelog [07:31] heh, shame on you :) [07:31] if you check debian/d-i/i386/package-list [07:32] # This file is used to build up the control file. The kernel version and [07:32] # "-di" are appended to the package names. Section can be left out. So can [07:32] # architecture, which is derived from the files in the modules directory. [07:32] # It overwrites specifications from /usr/share/kernel-wedge/package-list. [07:32] # [07:32] Package: kernel-image [07:32] Package: fat-modules [07:32] Priority: standard [07:32] it should look like that [07:33] yeah, it does [07:33] ok [07:33] do you have d-i/sparc/modules/sparc64/ext2-modules ? [07:33] yep [07:34] great [07:34] than it's missing only the changelog entry === fabbione goes and sits in a corner [07:34] thanks dude [07:34] sorry for the panix [07:34] panic [07:35] no problem :) [07:36] i did add the changelog entry and committed it for the sake of completness [07:36] it should be in my archive if you want to pull it [07:36] and i must run to cook now :) [07:36] later [07:37] later === mdz_ [n=mdz@69-175-232-197.vnnyca.adelphia.net] has joined #ubuntu-kernel === mdz_ [n=mdz@69-175-232-197.vnnyca.adelphia.net] has joined #ubuntu-kernel === trevilor [i=gnu@p549CE2DA.dip.t-dialin.net] has joined #ubuntu-kernel === mdz_ [n=mdz@69-175-232-197.vnnyca.adelphia.net] has joined #ubuntu-kernel === jbailey [n=jbailey@modemcable139.249-203-24.mc.videotron.ca] has joined #ubuntu-kernel === mdz_ [n=mdz@69-175-232-197.vnnyca.adelphia.net] has joined #ubuntu-kernel [11:14] BenC: zero. [11:14] heh [11:14] you tree went in completly different direction. [11:14]