[00:32] hyperair: yes, the default ubuntu kernel supports it [00:33] i'm testing out one of the ubuntu kernel team's 2.6.30 kernels and it works great except I don't have any audio, can someone steer me in the right direction please? Perhaps I need to compile in another module or something? I have a "Intel Corporation 82801G (ICH7 Family) High Definition Audio Controller" [00:40] eshaase: maybe it just be muted, have you run alsamixer to check it? === mcasadevall is now known as NCommander [01:36] AceLan: yeah, i've checked and its not muted [01:42] eshaase: Have you tried with kernel 2.6.31, or a recent alsa-driver snapshot? [01:53] TheMuso: i'll try with 2.6.31 [01:53] TheMuso: i know what alsa is but what is alsa-driver? [02:22] eshaase: Its the component of alsa that deals with the kernel drivers. [02:22] eshaase: Anyway try 2.6.31 as that should pretty much have the latest alsa code. [02:23] bar a few bits that are not in the mainline kernel yet, but I don't think they will matter. [02:46] ok, i'll give that a shot, brb [08:48] hi [08:48] can anyone tell me when jaunty uses kexec to boot ? I am trying to debug a problem where grub is skipped after a poweroff and it looks suspiciously like kexec is being used by jaunty [08:49] but I could be wrong [08:49] Is the kexec-tools package installed? [08:49] andersk: no [08:50] Is there an /etc/init.d/kexec-load? [08:50] no [08:50] Wait--it's skipped after a _poweroff_, not after a reboot? That isn't kexec, then. [08:50] that's right [08:51] it's very odd [08:51] restart works as normal [08:51] but after shutdown -P or -h grub is skipped [08:51] and I/we have no idea how to start debugging it [08:52] Maybe this is some kind of grub savedefault thing? I don't remember how that works. [08:53] interesting.. it does take 15 seconds to boot as well which just generally seemed fast to me [08:53] according to dmesg that is [08:57] andersk: I don't think it's that... my menu.1st is quite boring and has timeout 10 , http://pastebin.com/f607db10f [09:00] Your `default 8` is out of range; you have 7 entries (0-6). [09:00] andersk: aargh.. I just gave you the wrong file sorry [09:00] let me fix that [09:08] andersk: ok http://launchpadlibrarian.net/28500841/menu.lst [09:08] andersk: all the info is in the bug report https://bugs.launchpad.net/ubuntu/+source/gnome-power-manager/+bug/389930 [09:08] Malone bug 389930 in grub "grub menu skipped after shutdown" [Low,New] [09:12] andersk: sorry for the delay [09:16] hi tgpraveen [09:17] hi lesshaste [09:17] tgpraveen: do you happen to know anything about the kernel? :) Everyone I ask turns out to be a newbie [09:18] lol i am sorry i too am a newbie [09:18] anyways ask what u want to and i see a lot of expereienced ppl in the chat list [09:18] so someone will answer [09:18] tgpraveen: they don't reply :( [09:19] lesshaste: hmm they might be busy. my advice try the mailing list [09:19] is there a ubuntu kernel mailing list? [09:21] ah yes.. I just found it [09:21] maybe a non newbie will arrive ... [09:21] tgpraveen: out of interest.. are you here to ask a question too? [09:22] not really just to browse around [09:22] u know see what others are talking abt [09:22] and out of intrest what is your question [09:23] well.. after a poweroff (shutdown -h or shutdown -P) my laptop skips grub when it restarts [09:23] but after a restart (shutdown -r) it doesn't [09:24] hmm i am not sure but that doesnt really sound like a kernel issue though i might be wrong [09:24] maybe u should try #ubuntu-devel [09:26] ok [09:26] tgpraveen: the problem is that there are just facts I am missing [09:27] tgpraveen: does ubuntu have some mechanism for kick starting the kernel extra quick that could be causing this [09:27] tgpraveen: I also notice the boot time is very fast [09:28] lesshaste: karmic or jaunty? [09:28] jaunty [09:28] hmm dont really know. dont ask in #ubuntu-devel for jaunty [09:29] i guess u should file a bug report on launchpad [09:29] https://bugs.launchpad.net/ubuntu/+source/gnome-power-manager/+bug/389930 [09:29] and wait for somebody to answer here [09:29] Malone bug 389930 in grub "grub menu skipped after shutdown" [Low,New] [09:29] :) [09:29] it got assigned to the wrong place I think [09:30] yes gnome power mgr isnot the right package [09:30] nor is grub [09:30] where it is assigned now === sonicmctails is now known as mcasadvall [09:50] smb, we expect compat-wireless to be newer than the kernel in general right? [09:50] newer but still compatible [09:51] apw, a) yes (I would) and b) hopefully (as in compat). [09:51] apw, why? [09:53] am trying to update lbm, but it seems behind mainline for one accessor [09:53] so i suspect its not been updates since the merge window closed [09:53] so i guess fixing it (i think its gonna be easy) is worth it else we cannot upload it, nor linux-meta [09:55] apw, Have you checked with the compat-wireless sources or just with what we have in lbm? [09:56] this is after using the update script to get the latest version from the repo [09:56] Ok, which driver do we talk of? [09:58] updates/compat-wireless-2.6/drivers/net/wireless/mac80211_hwsim.c [09:58] Probably the other approach would be to go with the slightly outdated driver in lbm until the sources get updated. After all lbm is only best-effort and opt-in [09:58] is the file with the error, sounds like a simulator too [09:58] Oh, it actually fails to build [09:58] well i think the problem is that _all_ versions would not compile, cause its out of date wtf to the mainline kernel [09:58] yeah fails to build right now [09:59] i may have fixed it, but i am just trying to gauge if it is correct to fix it in this context, i think it likely is [10:00] I would say yes. Better to have a compiling package at that stage [10:00] We would likely pick up an updated upstream as soon as there is one [10:00] (upstream being the compat-wireless sourcE) [10:01] right ... it working is the bit that i see as best effort. we need something in the package which builds and installs at least [10:02] ok .. good then i think i am going the right way :) thanks [10:02] Yes, that sounds much sensible. I guess so. :) NP [10:08] hi all [10:09] I have sort of fixed my weird grub error but ... what part of linux is running when the grub menu is showing? [10:09] presumably not the kernel yet [10:10] <_ruben> grub is [10:10] <_ruben> bios -> grub -> linux [10:11] _ruben: ok... so my problem is that grub doesn't display on the screen after a poweroff [10:11] _ruben: so you can't see the menu [10:11] _ruben: it does display after a restart (shutdown -r now) [10:11] <_ruben> sounds like a bios issue [10:11] _ruben: it is running in both cases I now realise [10:13] is there anything grub could do in theory to initialise whatever the BIOS has failed to? [10:13] actually.. is it worth trying grub2? [10:16] <_ruben> well .. if grub doesnt show up at all, there's nothign grub can do .. unless grub breaks after getting started but prior to outputing anything [10:17] lesshaste, the kernel is definatly not involved in the loading the kernel, all that is grub code [10:17] i assume this behaviour is not affected by kexec being installed/not install. ie. if you remove kexec nothing changes [10:18] (it cannot be related of course as kexec only executed once the kernel is up) [10:33] * kdub is having a hard time finding a bug to be useful with :( [10:35] could i persuade someone here to compile acpi-cpufreq as a module? [10:36] as opposed to compiled into the kernel [10:36] bug #355232 [10:36] Malone bug 355232 in linux "acpi-cpufreq/powernow-k8 should not be built-in into the kernel image" [Low,Triaged] https://launchpad.net/bugs/355232 [10:36] kdub, what feels like the main problem? Too unspecific data or unknown areas? [10:44] smb: unknown areas in working with all the tools, i suppose [10:44] apw: thanks...it's all a little mysterious to me [10:45] and managing the configurations, and ways to test [10:45] smb: it feels like the headache that always happens from going from theory to practice, if you know what i mean [10:46] kdub ... heh don't be dissheartened, its a steep curve getting into the kernel for sure [10:46] kdub, I guess so. Probably the way to go is pick one and ask for things on how to proceed here [10:46] therre are few packages the same size ... ooffice, X server, little else [10:47] smb, perhaps have a sneak look and see if there is something simple on there, a nice cherry-pick or something [10:47] lesshaste, Mysteriously I think grub just uses the simple text mode that should be initialized by the BIOS. Do you get any visual feedback (like bios screen)? [10:48] smb: I can enter the bios setup by pressing esc or f1... but after a poweroff I just see the Toshiba logo and actually there is a slight change to the screeen when I assume grub is loaded [10:48] apw, Maybe. I have not looked into the community section but probably a lot there are old ones without much feedback... [10:48] smb: little bits of the top of the lettering become blacked out :) [10:48] smb: did I misunderstand the question? [10:49] lesshaste, normally there is a shift of font, to my eye the font gets 'shorter' vertically when the kernel loads the console fonts up [10:50] lesshaste, No, that was what I meant. Is there any control in the BIOS about which gfx should get initialized. Not sure this is a dead lead, but on some laptops I could select at least whether lcp or crt should be primary. On some destops there was a selection about agp or pci sometimes... [10:51] smb, sometimes they allow disabling the 'onboard' graphics so that if you stick in a second card it uses that ... perhaps thats got turned on when there is not one installed [10:52] the only "display" options I see are "power on display = auto-selected" which can be changed to LCD+Analog RGB [10:53] there is a mysterious "execute-disable bit capability" ?! :) [10:53] everything else is fairly clear [10:54] lesshaste, that enabled NX support in the processor [10:54] or more specifically prevents disabling of the NX support in the processor [10:54] to allow buggy OSs to boot in the face of the new support [10:54] NX allows us to mark the stack non-executable to prevent overflows being so useful [10:55] thanks [10:55] there is also an option for virtualisation technology I see :) [10:56] yep same thing, turns off support for vmx extensions to prevent OSs which don't expect it from blue-screening, erm, i mean crashing :) [10:56] :) [10:56] bioses should really come with man pages [10:56] :) [10:56] what make of machine is this? [10:56] toshiba tecra a9 [10:57] nnng. what is toshiba up to [10:57] I was going to post to the grub mailing list [10:57] but I see they don't support anything before grub2 anymore [10:59] lesshaste, One desperate thing to try might be to uncomment the default line in the grub menu.lst that does some colouring. Not that Toshiba did something nice as setting default colours to black on black... [11:00] But I cannot remember anything in grub to enable a specific video mode... apw or cking do you? [11:01] there is a terminal command i think, never seen it required tho [11:01] do you need to change colours in grub 0.97? [11:02] cking, It is unclear whether colours are enough. Just brought up that and there would be a default line in the standard grub menu [11:03] The general symptom is that grub menu does not display after cold start but after reboot [11:03] smb: interesting! [11:04] I'm just checking if there is a bios upgrade too [11:09] kdub, We will try to find one bug to start of tomorrow, if that i ok with you... [12:05] smb: well... it turns out it has been reported before https://bugs.launchpad.net/ubuntu/+source/grub/+bug/374823 [12:05] Malone bug 374823 in grub "GRUB menu not showing up at cold boot" [Undecided,New] [12:30] lesshaste: from what I can recall, grub does not twiddle with BIOS video modes at all, so this is part of the problem. [12:30] cking: yes I think the solution may be to try grub2 [12:30] cking: as no one at grub will fix a legacy bug in any case [12:31] unless there is a really compelling reason to put the effort into addressing this for legacy grub [12:31] cking: just another annoying bug :) You can always just memorise the order of entries in menu.1st :) [12:31] or boot and then restart [12:32] yeah, but that's unpleasant. [12:32] right [12:32] but I somehow suspect it won't come under "compelling" for the developers [13:19] hi all [13:20] Is it possible that the 2.6.30-10 initramfs script has changed? I always got the wrong keyboard layout with it while 2.6.30-9 works fine. The keyboard layout entry in console-setup seems to be the same in both initrds. Any idea? [13:23] I need the correct keyboard layout for dm-crypt [13:32] Unggnu: known bug, if you want to figure out what's going wrong feel free as I haven't had a chance yet. bug 392218 [13:32] Malone bug 392218 in console-setup "Console font not accepted after boot up" [High,Triaged] https://launchpad.net/bugs/392218 [13:32] (the description is a bit too specific, don't worry about that ...) [13:32] lesshaste: I rather suspect grub2 *will* work better, as it initialises a graphical mode by default and therefore is bound to twiddle the video mode [13:33] cjwatson: thx === BenSee is now known as BenC [13:37] Btw. I got around 10 seconds boot time with the 2.6.30-8 in Karmic I guess [13:37] with fastboot [13:37] can someone tell me what the stable-review patches on kernel.org are [13:37] but after -9 I got around 14 [13:39] and what is fastboot? [13:39] a boot option [13:39] Around 2 seconds faster in my case [13:39] can it be used in jaunty? [13:39] no [13:39] Since 2.6.29 [13:39] even with a new kernel? [13:39] If it is >=2.6.29 yes [13:40] cool. i'll have to try that [13:40] Isn't much anyway. I just wondered why there is a difference of four seconds or around 40% between this two kernel releases [13:41] masterkernel: http://lwn.net/Articles/314808/ [13:44] thanks Unggnu [13:51] <_ruben> hrm .. just did a dist-upgrade (including kernel) on my jaunty system .. now it fails to boot because it cant find my lvm [14:00] smb, do you know if it is safe to upload -lbm when the kernel's packages are not yet accepted? [14:00] the only mention of fastboot is the BSDism, meaning do not check the disks on reboot [14:01] apw, As far as I know it is save as it will depwait for the kernel [14:01] ok cool then ... smb i have a karmic -lbm for upload ;) [14:02] apw, I suspected that much. Will have a look [14:02] in the normal place :) [14:03] * smb thinks that sound like conspiracy :) [14:04] smb, ok the git tree is pushed too [14:04] iit is also built in my ppa: https://edge.launchpad.net/~apw/+archive/daily/ [14:05] heh yeah, i am out to confuse you with all the changing about [14:05] Well I can see in debdiff that is a little scary [14:08] * apw has a look [14:09] smb, you got a link to the debdiff or doing that locally? [14:10] * apw feels he is missing a trick somewhere [14:10] apw, I did it locally but slightly wrong. So it is probably not as scary as the diff between that and Jaunty ;-) Doing it right in a sec [14:10] heheh [14:13] * apw does the same, the diff looks ok i think [14:15] apw, Mostly yes. I am a bit confused by that strange change of compat-wireless master fom 2009-06-12 to 2009-06-10... [14:15] that was an error in my previous version, a missunderstanding of how thye were named [14:15] it should always have been -10 [14:15] i did it on the 12th but the tree i took was the 10th tree, and i miss recorded that [14:17] apw, And your changeslog in the package might be different from the tree ? [14:17] how so? [14:18] I just pulled and in git 1-7 has two lines, while in the package diff it looks like 5 [14:20] [ Andy Whitcroft ] [14:20] * Update compat-wireless to master-2009-06-19 [14:20] * Bump ABI [14:20] * update compat-wireless to track changes to skb dst accessors [14:20] * update compat-wireless to track changes to ALIGN(foo, NETDEV_ALIGN) [14:20] * update compat-wireless to track removal of FIRMWARE_NAME_MAX [14:23] <_ruben> hmm .. booting the -11 kernel does work .. must be something flakey with -13 or its initrd [14:49] smb, hi again [14:49] smb, I fixed the grub problem in the end [14:50] lesshaste, So what did you change? [14:50] smb, I upgraded to grub2, in short [14:50] smb, which has completely different display routines it seems [14:51] smb, upgrading to grub 2 is not exactly user friendly at present, at least on jaunty [14:51] but with a little editing it worked [14:52] lesshaste, Hm, guess that counts as a work-around. I wonder what the heck is then worng with the grub1-yourbios interaction... [14:53] smb, it's a good question.. did I show you the duplicate bug with the people talking about workarounds? [14:57] smb, I feel I fixed this bug more or less on my own.. I hope someone at ubuntu will at least read https://bugs.launchpad.net/ubuntu/+source/gnome-power-manager/+bug/389930 now! [14:57] smb: grub1 doesn't reinitialise the video mode itself *shrug* [14:57] Malone bug 389930 in grub "grub menu does not display after cold boot" [Low,New] [14:57] cjwatson, the sudo apt-get install grub-pc command made a real mess of menu.lst [14:57] lesshaste: huh? several people have read it. It now sounds like updating to grub2 is sufficient, and since grub2 will be the default in karmic, it sounds as though there's effectively nothing to do [14:57] lesshaste: made a real mess how? [14:58] cjwatson, there is a point where it says it has used "kopt" to find a command line and asks you to check it [14:58] cjwatson, it presented a blank line [14:58] cjwatson, then when I just pushed on it added several lines with 80 or so copies of some ansi character that looks like an A [14:58] to menu.lst [14:59] nothing wrong with that [14:59] this apart from it still using root instead of uuid [14:59] it's a divider [14:59] the uuid thing was broken in jaunty but is fixed in karmic [14:59] cjwatson, well someone could close the bug I suppose :) [14:59] and the duplicate [14:59] <_ruben> ouch .. i found out why my -13 kernel wouldnt boot .. i used tasksel to remove "virtual machine host" .. which removed lvm2 !! [15:00] _ruben, eek! [15:00] <_ruben> reinstalled lvm2 .. it updated initrd .. et voila .. it boots [15:01] cjwatson, the blank line is correct? It's quite confusing for a poor unsuspecting user [15:01] lesshaste: looks like it means you have no special parameters beyond the defaults [15:01] cjwatson, sort of... it also added back in default parameters [15:02] cjwatson, I had removed quiet and splash [15:02] cjwatson, and it added them back it seems [15:02] feel free to file a bug. hardly qualifies as making a real mess though. [15:02] the only thing that seems like a real problem is the root/uuid stuff, which like I say is fixed in karmic. [15:02] ok [15:03] (real problem: by which I mean that otherwise it should still have booted fine) [15:03] cjwatson, no I thought the A characters were the mess. I didn't realise they would be harmless and was too scared to test it [15:03] they're a horizontal line when interpreted in the relevant character encoding [15:03] also, I see that it was expecting kopt= lines now [15:03] title `echo ───────────────────────────────────────────────────────────────────── | iconv -f utf-8 -t cp437` [15:04] ... is where that divider comes from [15:04] as opposed to parameters after the kernel name [15:04] CP437 being one of the old DOS code pages [15:04] lesshaste: yes, that's how you were supposed to edit the grub1 menu.lst file [15:04] yes [15:04] well sort of.. :) [15:04] the default isn't like that [15:04] so you can't blame the user for copying the default [15:05] if you just edit the bit after the kernel name and not kopt=, then the next update-grub will overwrite your changes [15:05] right [15:05] I don't blame the user, I blame the incredibly dreadful old menu.lst format. [15:05] :) [15:05] but there's only a limited amount we can do about that [15:05] I'd rather focus effort on making the new world better [15:06] cjwatson, oh about your comment in the other channel, having just edited menu.lst by hand I was a little nervous it wouldn't boot at all :) [15:07] in any case, grub 2 works so all is great... I assume the docs will be tidied up by the time karmic comes out... [15:07] now onto the next bug :) [15:08] we probably ought to do something with defoptions= in menu.lst [15:08] cjwatson, more important would be to fix the upgrade script for jaunty [15:08] cjwatson, is it much more work than just copying the karmic fix? [15:08] I'm not interested in fixing the upgrade script for jaunty, really [15:08] it was a proof of concept so that we could get some degree of testing [15:09] would you mind if someone else fixed it? [15:09] we aren't really intending the majority of people to upgrade to grub2 in jaunty, just a few intrepid (ahem) souls [15:09] not especially [15:09] :) [15:09] but I'm not sure it would meet the stable update guidelines [15:09] karmic and beyond is what we're actually interested in for grub2 [15:10] it's unlikely that you'll be able to do a really good job of jaunty without a much more sweeping update, which would definitely not meet the stable update guidelines [15:10] ok..maybe it should just be removed for jaunty [15:10] as it is broken [15:10] it's not broken [15:10] and added in some super flaky repository [15:10] it's just not as good as it could be [15:10] but that applies to lots of software [15:10] we can't remove packages once they land in a release anyway [15:10] well broken in the sense that a normal user might not be able to boot their computer after using it [15:10] we used the jaunty package for a systematic test on lots of hardware [15:11] the results were actually very good [15:11] so I'm not all that worried [15:11] ok.. I know when I am defeated :) [15:11] I'd be fine with us fixing the root/uuid thing in a jaunty update [15:11] that's fairly obvious and self-contained [15:11] but I don't think we should commit to really extensive work on jaunty's grub2 [15:11] I think that's what I was saying! [15:11] what did you think I meant? [15:12] you could have been talking about any of the things you were commenting on in the grub2 upgrade, such as the kopt/defoptions thing [15:12] ah no [15:12] just the root/uuid thing [15:12] or the fact that jaunty's installer doesn't know that grub2 can boot of LVM [15:12] off [15:12] or any of the other things we know are broken in jaunty [15:13] well in any case, thanks for the help [15:13] and good luck with the grub work in general [15:13] there is actually a bug open on jaunty for the root/uuid thing [15:13] so if somebody wants to have at that, they should go right ahead :-) [15:13] :) [15:13] I have a lot more serious bugs to report now :) [15:13] bug 376879 [15:13] Malone bug 376879 in grub2 "grub2 installer modifies grub 0.97 menu.lst incorrectly and fails to chainload grub2" [Medium,In progress] https://launchpad.net/bugs/376879 [15:14] like the one that stops my dvd player working at all [15:15] bye for now [15:15] and thanks again [15:16] apw: actually, ^- that's still assigned to you. fancy producing a diff against jaunty? [15:17] cjwatson, will check it out thanks for the prod [15:47] smb, I want to update stuff under ubuntu tree in karmic.. what is the usual procedure ? can you point me to a wiki ? [15:48] manjo, The current procedure would be ask apw [15:49] heh ... manjo yep i saw that mentioned in our 'whats next' email [15:49] let me get this damn package uploaded then we can talk about it [15:49] ok [15:49] as i have started looking at ndiswrapper [15:49] yeah I am doing the same [15:49] probably we are duplicating effort [15:49] manjo, i think i have it done already ndiswrapper [15:49] am uploading a test kernel now [15:50] ah cool... so I can take that process and rinse repeat [15:50] we obviously need to coordinate :) [15:50] ok will wait for you ping [15:50] cool as [15:52] * manjo breaks for coffee [16:25] I am sorry this is probably a faq but can I install jaunty kernel on hardy? [16:25] my dvd drive doesn't work [16:37] apw: have you got a repository for newer kernels for hardy by any chance? [16:38] lesshaste, jaunrty kernel will need other deps you don't have so ... unlikely to work [16:38] intrepid might work [16:38] apw: I was looking at https://launchpad.net/~apw/+archive/ppa and thinking that might be you [16:39] yeah that is me [16:40] nothing new in there for hardy tho [16:40] ah [16:40] I am trying to work out why my dvd drive doesn't work [16:40] I thought testing a new kernel might help [16:41] I get lots of lovely errors :) [16:42] I assume trying to install a jaunty kernel and just letting it resolve dependencies is a bad plan [16:43] the problems look similar to https://bugs.launchpad.net/ubuntu/+source/linux/+bug/223780 but I can show you dmesg if you like as it is full of juicy information and errors :) [16:43] Malone bug 223780 in linux "hdc: status error: status=0x59 { DriveReady SeekComplete DataRequest Error }" [High,Fix released] [16:46] rebooting... [16:59] this bug https://bugs.launchpad.net/ubuntu/+source/linux/+bug/223780 is pretty fatal for me.. my system completely locks if i try to use brasero for example [16:59] Malone bug 223780 in linux "hdc: status error: status=0x59 { DriveReady SeekComplete DataRequest Error }" [High,Fix released] [16:59] hard locks I should say [17:00] is hardy too old for anyone to worry about? [17:01] upgrading to jaunty looks somewhat painful :( [18:14] hardy is supported still. [18:23] What worries me a little bit is "As a result cable detection [18:23] becomes less reliable when compared with 2.6.24 but the affected drives are [18:23] useable again." [19:49] hmm [19:49] Linux had another one of those "oh noes! no memory!" fits [19:49] when there was plenty [20:09] Keybuk, did we use to record the dmidecode output somewhere, or did i imagine that [20:10] Keybuk, running i386? ZONE_NORMAL exhausted, or KVA ? [20:13] apw: I've seen it on i386 and amd64 [20:13] not sure what the latter bit means [20:13] apw: the kernel exports dmi information under /sys [20:13] kernel virtual address space [20:13] Keybuk, in hardy? [20:13] jaunty and karmic [20:14] yeah thats my memory too [20:14] just thought i remembered something early boot slurping it up for something, hal perhaps to use [20:14] /sys/devices/virtual/dmi/id/ [20:14] on older releases