[00:12] sorry about the bug spam, folks. I can't really prevent people from filing bugs clearly caused by the grub configuration not being properly updated. [02:40] !ops [02:40] Help! lamont, zul, T-Bone, mdz, or jdub [02:41] !ops [02:41] Help! lamont, zul, T-Bone, mdz, or jdub [02:42] um, what do you need? [02:42] ban me [02:42] why don't you just /part ? [02:42] why not ban me [02:42] sigh, /ignore === cwillu_ is now known as cwillu [03:16] hi guys, I read https://help.ubuntu.com/community/Kernel/Compile but am still a bit confused about how to update config files. After "make menuconfig", how to merge what I changed in .config to debian/config? I saw /debian/config/config.common and debian/config/ARCH/config.xxx. Should I manually modify all files in /debian/config or replace them with the new .config? [03:17] which way is recommended to change them? Thanks! [03:43] [answer my own question] found it. just copy .config to either debian/config/xxx and run debian/rules updateconfigs and then Mr. Proper. [03:44] maybe the wiki need update a little bit to clarify this "copy" step. I saw lots of guys asking the same question in LP and other maillist. :) [03:45] eggonlea_: its a wiki, you can do that === bjf is now known as bjf-afk [08:55] smb, did we get any testing feedback on the -virtual kernel issues? [08:55] apw, Yes, luckily even good ones [08:57] I guess the plan is to prepare and push everything for upload and then plan for it being up Thursday [08:57] apw, ^ [09:04] smb, yeah sounds about right. we should consider the m586_tsc thing too [09:05] apw, Oh, ok. So I probably should delay the finalizing commit to the kernel to get that in as well [09:05] i guess it could go in with the stables ... as noone has yet mentioned it [09:08] apw, Right. it would be slightly more preferable to stay with the current finalized version. At least for non-lava issues [09:08] smb, then that sounds like a plan [09:09] apw, ack [09:10] hei all [09:10] hehe, I like that term: non-lava issue. (assuming it wasn't a typo) [09:10] hi i need some help, can someone oblige [09:11] amitk, no I meant anything that is not melting and very hot [09:11] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/425756 [09:11] Malone bug 425756 in linux "[regression karmic] cd/dvd drive not detected" [Medium,Triaged] [09:12] smb: yeah, thats what I thought. I propose we change our bugs from critical to lava and non-lava [09:12] :) [09:13] amitk, It would be a nice change :) [09:15] indus, I am not sure there is any advice that can be quickly given. The report seems to cover potentially various issues. I believe to remember something about detection in CDs but it might be different too. apw, do you have more recollection there? [09:16] smb, i was just reading same, and i concur there is nothing easy to follow in there. i thought some cd stuff was improved generally but clearly not fixing this person [09:16] smb: hmm, sorry i dont quite understand [09:16] indus, what are your symptoms? [09:16] indus, And in your case since when (Jaunty, or before)? [09:17] yep, when did it last work [09:17] ok once i put in the live cd, the cd reaches menu for booting ubuntu etc , then when i select 'try ubuntu without change to your computer' it starts looking at HDD instead of cd drive and later fails with initramfs [09:17] it last worked in Hardy heron [09:18] once i used some boot options pci=nomsi, rootdelay bla bla etcand it worked once to twice but no idea what it does [09:18] works in hardy nicely [09:18] i am not sure if that is the same bug or not, the bug referenced implies a normal boot without cd [09:18] I'd suspect something with libata, as that was one of the major changes after hardy [09:19] so ... i would recommend getting a new personal bug filed with the machine info on it [09:19] and try and get a picture of the failure into initramfs [09:19] apw: yeah the bug is normal boot, but live cd also doesnt work, i instaleld it with usb [09:19] and a dmesg output from there if you can get it [09:19] ... [09:19] wait i give you another live cd bug [09:19] i have one for casper intrepid [09:19] ok if you have a normal boot without CD then what is in the dmesg when it fails [09:20] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/425756 [09:20] Malone bug 425756 in linux "[regression karmic] cd/dvd drive not detected" [Medium,Triaged] [09:20] no no sorry [09:20] wait' [09:20] https://bugs.launchpad.net/ubuntu/+source/casper/+bug/303654 [09:20] Malone bug 303654 in casper "[intrepid] Live CD cannot mount - panic Unable to find a medium containing a live file system" [High,Triaged] [09:21] apw: dmesg says, and also at boot i have a message , libata1 soft reset failed , expecting a reference package element, device not ready [09:22] one problem might be related to rootdelay, I remember there had been issues with waiting long enough for the root fs in some cases. But together with the other behaviour it might be ata related [09:22] so we need to get that log element into the bug so we can see its exact form [09:22] right [09:22] also we need to know the exact hardware you have both mobo / disk interface / drive [09:22] dmesg 3.972516] ata1: softreset failed (device not ready) [09:22] [ 3.972519] ata1: failed due to HW bug, retry pmp=0 [09:23] asus M2A VM motherboard amd 690 g chipset / sata HDD /IDE samsung combo drive [09:23] I would say booting with an usb-cdrom and replacing "quiet splash" with "debug" on the grub line, and then posting the whole dmesg (maybe try to access the internal drive too) [09:23] smb: i dont have usb cd rom :) [09:24] if i couldget some kernel parameters to boot , i could try [09:24] rootdelay , so many things i had tried [09:24] but its a hit and miss mostly [09:24] indus, was something with all_generic_ide amongst those [09:24] smb: of course , but it doesnt work always [09:25] i tried it too [09:25] its like, instead of reading the cd drive, it reads from HDD [09:25] hdd light stays on after cd boots from menu [09:25] indus, How did you say you secceeded booting? From a usb stick? [09:25] smb: yeah [09:26] flash [09:26] if the cd is missing then its not unlikely it falls back to the hdd [09:26] no , cd is present when live cd boot i use [09:26] it boots from cd, but once i press menu options, it starts reading HDD [09:26] not to the kernel its not is it ... else you'd not be reportng a missing cd dirve [09:26] Ok, if you do that again and get up (with using the debug line), then gather dmesg, cat /proc/interrupts [09:26] ya ok [09:27] kernel cant see or tries to look in wrong location i believe [09:27] smb: hmm what do i need to do? [09:27] smb: you saw that bug report, i have dmesg in there [09:27] its with the live cd , also have casper.log [09:28] indus, you boot from the usb stick, then on the boot menu you would go to advanced options and edit the line to remove splash and quiet and put debug there instead [09:28] then boot the live image [09:28] smb: ok and ? but with usb it boots fine. [09:28] indus, ok where is your dmesg, the one in the bug you first referenced does not have your message in it [09:29] apw: wait which dmesg you need, normal boot or live cd? [09:29] indus, still it would try to access your cd and hdd and the errors should be there [09:29] smb: here is dmesg from live cd intrepid http://launchpadlibrarian.net/34279545/dmesg.txt [09:29] as the problem is likely the same i am not sure it matters. debugging on your booted system is probabally easier i'd assume [09:30] http://launchpadlibrarian.net/34279424/casper.log this is casper [09:30] indus, that dmesg is from a .27 kernel [09:31] apw: yeah have had the problem since [09:31] apw: ill give you karmic dmesg then [09:32] apw: http://launchpadlibrarian.net/31444487/BootDmesg.txt [09:32] which is also a bit outdated [09:33] aah ok [09:33] so you need the current kernel [09:33] indus, Have you tried with the release version? [09:33] smb: of course, iam fully updated to release [09:34] basically, as i stated earrlier, hasnt worked in any kernel after hardy 8.04 [09:34] indus, Just the log you had was from 2.6.31-9. And at the current point, when looking at the problem it helps to start of as recent as possible [09:34] aah damn it, iam at work now [09:34] diff system [09:35] ill give you dmesg with latest one in evening then [09:35] anyother way i could get cd rom to be read? [09:35] temporary solutions [09:36] secondly, i have a general question about multi threaded applications [09:36] nvm [09:37] oh so this sm600 system isn't the right one [09:37] apw: ? [09:38] nothing springs to mind without knowing more about it. a current dmesg can't hurt and if you have a dmesg from hardy that might be useful to se ehow it was recognised in that kernel [09:38] indus, Sure. And with those kind of problems, it also helps when we have a good summary on how things are connected (should the drives be on the sata port, or ata, are the sata ports set to ahci or ide) all that stuff. Thought info can be found in the dmesg it is slower to gather it manually. [09:38] smb: in bios its set to ide controller [09:39] i interepreted what you suggested there as saying the second dmesg is not from the system [09:39] smb: HDD is on sata port [09:39] apw: no no all were filed from my system, iam just chatting with you from work system [09:39] indus, from the log it seems ahci is firing up [09:39] smb: hmm i must have fiddled iwth bios settings that time [09:40] smb: but itsdefault on ide controller,there are 2 other options raid and ihce [09:40] but i use ide [09:40] also, the cd drive is connected with this wide and thin cable , which ironically says 'asus HDD cable' [09:40] i believe its ide cable [09:41] indus, Ok, so this is all stuff that we should have in the bug report. The better we know how things should be, the better one can spot differences. [09:42] smb: hmm where do i add it [09:42] smb: apw ill try to give you a 8.04 dmesg, and ill attach it today [09:42] indus, You either edit the description section of you bug or add a comment with it [09:42] smb: ok [09:43] but is that really necessary? 8.04, man ill have to install it and all that [09:43] just reinstalled 9.04 a few days ago [09:43] indus, you could boot the livecd for 8.04 and take it from there [09:43] wait let me get an old bug with hardy then [09:43] as the cd works in 8.04 [09:43] apw: hmm , from live cd when it boots into desktop? [09:44] yep ... it must have detected and used the CD [09:44] yes ok ill surely do that today, [09:44] apw: so from a terminal i type dmesg and paste it there in report? [09:45] i mean, no need to install right? [09:45] indus, you actually should have network from the live cd and could add it to the bug report [09:45] indus, No need to install [09:45] smb: ya i know, i have network, ill add it [09:45] Its just to get a dmesg from a working case as compared to a non-working [09:45] let me see if i can do this in an hour or so [09:46] so i have a question,is the kernel in C? [09:48] the majority of the kernel is in C yes [09:49] apw: smb i have an ubuntu 8.04.1 cd , is that ok? [09:49] if the cd works yes [09:50] what apw says [09:50] also one more important thing here, some peopel had this problem befoer hardy and some patch fixed it, so it fixed for some and broke for me [09:50] not sure if its related though [09:51] nvm [09:51] ill brb and upload dmesg [09:51] 5 min [09:52] indus, i note that your karmic kernel is an old one 2.6.31-9.29 ... the current one is -14.28 or so [09:54] apw, Not sure about the .28 but certainly -14 [09:55] -14.48 I would guess [09:55] doh ... typo [10:21] ok got the dmesg [10:22] apw: i attached the kernel [10:27] its ATAPI cd drive , is that different from ide [10:27] guys are you there? smb apw [10:28] indus, yes. not so much difference that it should not work [10:29] so now what :) [10:29] indus, you happen to know whether the cd driver is specifically set to master or slave? [10:30] smb: ooh dont remember, wait i came back to office now, :) [10:30] smb: slave i think though [10:30] smb: isnt it in the bios post screen? [10:30] and exactly where has the dmesg gone? [10:30] smb: dmesg i attached to bug report [10:30] indus, you added to the first or second bug [10:31] smb: hmm karmic bug report [10:31] indus, ok [10:31] indus, And the drive definitely should be set to master [10:31] smb: hmm ok how do i do that [10:31] it is the only drive on the pata bus [10:32] I would not know except by opening the case and look at the drive [10:32] in theory there should be a selector [10:33] sometimes its cable select, so it depends which connector on the cable its connected to [10:33] smb, how did you tell it was a slave from the dmesg? [10:33] apw, No I am working on indus statement atm [10:34] I have not got to the dmesg itself yet [10:34] i had the feeling it was a slave too, and am wondering what i saw to trigger that [10:34] if that is visible there anyways [10:35] * smb wonders whether it would be visible in the bios. maybe, depending on how good/bad that is [10:36] * apw shakes his head at all the 'sb600' needs workaround messages in the karmic dmesg ... a lemon of a machine i recon [10:36] i think its visible in my bios , i remember the words [10:37] its a beautiful chipset actually :) onboard graphics are really good [10:37] in the first post messages, it mentions HDD first ,then cd drive is probably slave , ugh i dont really remember [10:38] apw: its a 1 mm thin 5 cm wide cable connected to mobo [10:38] Well, as your hdd is sata that would be different worlds [10:38] ya hdd is sata sure, so no master slave business [10:39] the cable might provide cable select ability but this needs to be supported by the drive [10:39] cable select, its a simple cable,a wire [10:39] [ 1.000067] ata5.01: NODEV after polling detection [10:40] smb, looking at other messages i think that that can be .00 and .01 and .01 is slave [10:40] (this is from karmic dmesg) [10:40] indus, its one of the cables deliberatly broken after one of the connectors [10:40] smb: sorry i dont know this one [10:41] there were some changes in the final karmic kernel to handle better selection of speeds for pata disks ... speed of the io over the cable. which might be related to the issue [10:41] So there might be disagreement between controller, cable and drive what should be master and slave [10:41] aah internal fighting :) [10:42] i wish they documented what these damn messages really are meant to mean for humans [10:43] apw, Also the dmesg of hardy shows the cdrom as hdb (which would be another indicator of it being slave) [10:43] yeah hdb [10:43] i noticed [10:43] yeah smb i think thats pretty definative... though that means it can work [10:43] and i think the .01 failure indicates it found something on the cable and rejected it as a broken thing [10:44] where is this .01? in 8.04 dmesg? or new one [10:44] indus, yes that was in the new dmesgs [10:45] hmm which line is this? lots of .01, just interested in knowing [10:45] [ 1.000067] ata5.01: NODEV after polling detection [10:45] pretty sure ata1-4 are your sata ports and ata5-6 are your pata ports [10:45] i think the .01 can be .00 and .01 and i think they are master and slave indicators in this context [10:46] not yet confirmed that but it seems more than conincidental that is emmitted [10:46] and that message means 'i talked to it and it was junk so i am ignoring it' [10:47] indus, What I would suggest, when you get home. Try taking out the drive and check, whether it can be manually set to master or slave. I would suspect it can, as you said you use the hdd cable and probably the last connector (leaving the middle one free) and I believe those are wired to set cable select capable drives to master [10:47] smb: i think it can be [10:47] smb: some jumper settings which iam a little scaredof [10:48] i note that the bios is setting the drive to pio mode in the hardy dump, and the NODEV message in karmic is sometihng which can only be emitted in ATA PIO mode: [10:48] hopefully documented on the drive and not in the documentation one usually discards quite soon [10:48] /* If diagnostic failed and this is [10:48] * IDENTIFY, it's likely a phantom [10:48] * device. Mark hint. [10:49] so its very likely that line indicates that the drive is being dropped as it failed to respond appropriatly [10:49] for some definition of appropriatly [10:50] smb: hmm there is a connector in the middle of cable you mean? hmm i will check this [10:51] smb: http://www.cksinfo.com/clipart/electronics/computers/cables/ide-cable.png [10:51] this one? [10:51] indus, those ata flat cable connectors I know usually have three connectors... [10:51] whats the middle one for [10:52] indus, yep, iirc blue goes into the board side and black and gray would be master and slave drives [10:52] middle is generally slave, IIRC [10:52] jk-, Yeah, so do I [10:53] http://www.novell.com/coolsolutions/img/16837_ide-cable.gif [10:53] make up your minds [10:53] :) [10:53] *joke* [10:54] ok let me get this right, i set cd drive to master, but connect the cables as it were? [10:54] indus, Well doesn't that prove us. :) [10:54] yeah heh [10:54] what if i let cd drive remain slave and connect that middle grey one? just wondering [10:54] indus, yes, I believe there had been issue running slave only setups like this [10:55] as a general rule, if there is only one drive on the cable it should run as master [10:55] smb: hmm but ther is only 1 drive [10:56] smb: i hada floppy, on another port which i removed after karmic got stuck with floppy mesages [10:56] indus, soooo? ;-) [10:56] smb: i mean, since ther is only 1 drive, shouldnt it run as master [10:56] bah nvm , iam speculating [10:56] I thought thats what I said :) [10:57] smb: anyways ill set it to master manually [10:57] smb: will you guys be around here in a few hours? [10:57] indus, Chances are good [10:57] excellen [10:57] t [10:57] now,i have a general question for you guys/ladies [10:58] i have this game called quake4 which has an executable called quake4-smp which takes use of dual cores [10:58] ever since karmic alpha 4 it crashes now, is this a kernel thing or an application thing [10:59] i guess this is too little information for you , but i wonder what the problem could be [11:00] indus, An application crashing is certainly an application thing. [11:00] i get a segmentation fault [11:00] was wine since dapper drake [11:01] fine [11:01] funny thing is , if i run the regular non smp file and enable smp in console it runs fine [11:02] ok but can this qualify for a regression? [11:02] you could try stracing the -smp version and see what it does just before [11:02] its likely its something that the app is running to find out how many cpus there are or something [11:02] something which is likely allowed to change, and something userspace is meant to cope with [11:03] coredumping would clearly be a fail on the applications behalf [11:04] s/clearly/very highly likely/ ... if i am totally fair about it [11:04] apw: ok , but if this application worked fine for 2 years, and started crashing in the newer karmic os, what is the solution [11:04] i wrote to id software, but unlikely they will do anything now [11:05] newer karmic kernel and user space ... well if the app is broken in some way which is now exposed by karmic thats not likely something we can fix for them [11:05] hmm [11:05] if its breakage in the kernel/userspace then we can address that [11:05] figureing out which with a closed source app would be very hard [11:06] though i would start by stracing it and seeing what it did last [11:06] apw: ok ,hmm iam going to try running this gaem with hardy today again to make sure its not a corrupt data file [11:06] apw: running strace ill try reading more on that [11:07] apw: could you explain userspace [11:07] apw: also, what is it i could do for you to address the kernel/userspace issue in relation to this application? [11:07] if any* [11:08] userspace == everything thats not the kernel and not your application in this case [11:09] so libraries and the like... they could have had a bug fixed which the app is relying on for instance [11:36] oops sorry [11:36] i got disconnected [11:41] apw smb hello again [11:41] hi [11:43] any further comments on my bug? [11:45] apw: setting jumper will involve physically dismounting the drive isnt it? [11:46] i would expect so, the jumper docs are bnormally on the bottom of the drive [11:47] i am building a test kernel with some additional diagnostics information to try and find out which of three cases are triggering the drive reject ... which may help understanding [11:48] the three cases would be? [11:48] this is all so interesting, i should learn some C i guess [11:50] also, would you like me to test this kernel you are building? [11:59] the three cases are a controller diagnostic failure, an HSM violation, and a stuck register check [12:00] when the kernel is built and pushed up i'll add some instructions and commenty to the bug with the details [12:26] bye now and thanks [12:26] ill drop in later in 3 hours or so [14:31] apw smb hooray it works when i changed it to master [14:31] so now what [14:31] :) [14:32] i want to thank you for this help [14:32] indus, heh well thats good new. first thing is to report that in the bug [14:32] did you boot my kernel at all? before that? might be interesting to have that data [14:32] ya cool isnt it, and i finally found out what a jumper is [14:32] but with the drive as slave [14:32] which kernel? havent checked the report [14:32] you added ? [14:32] i just came home [14:33] i pushed a kernel to try and find out why it errored there [14:33] * indus is happily browsing through the 8.04 cd [14:33] very nice [14:33] iam really greatful (i hope i spelled that right) [14:33] finally iam gonna rent some dvd's [14:34] Im not very positive having a single drive forced to slave is a use case which would be thought of supported [14:34] heh well please do write up that you change it to master and it started working [14:34] It might have worked with the old driver but ... [14:34] smb, hmm? [14:34] * apw isn't at all convinced there is any techinical reason that the slave should not work on its own [14:34] but ... its good to know how to make it work [14:34] its possible that the error from the master failing is getting in the way of the slave identify [14:35] * cking is bemused by thus too [14:35] its possible its an obvious thing to upstream [14:35] If I recall that correctly there is some relationship between detection sequence and used ports [14:35] apw yeah, not all people will be able to drop in here and do the master slave thing [14:36] lucky i had a screwdriver [14:36] indus, if you have the energy to put it back as slave then test my kernel it may help us in deciding what if anything to report upstream. if not not the end of the world [14:36] but do report your success either way [14:36] apw, what is the size of kernel? [14:36] 25MB or so [14:36] hmm [14:36] So they go somewhat master, then slave. But when no master is found they might stop. Maybe even the drive keeps silent unless it "heard" his master [14:37] ok so just download and install the kernel? [14:37] i think in this case we get a controller level error on the master and then that is still pending when the slave error occurs and that moves that drive to bust [14:38] possibly incorrectly. though i am cirtian that master is the preferred place for a dive [14:38] drive [14:38] indus, is only worth doing that kernel if the drive moves back to slave ... else it'll report nothing new [14:38] ya ill move it back to slave [14:39] but which kernel exaactly? i see a few [14:39] oops sorry headers and image, so i need 64 bit and all.deb? [14:39] if you don't have any binary drivers then just the kernel image is enough [14:40] i have nvidia [14:40] actually iam not sure what you mean [14:40] so you need the headers too and the all [14:41] so headers amd64,image amd64 and all.deb [14:41] downloading [14:41] indeed so [14:41] thanks for testing [14:41] no problem at all [14:42] so i install it now? or move slave and then? [14:42] yikes i already installed [14:42] aah boot time i guess [14:42] will this appear in grub? [14:46] brb [15:24] hi [15:24] smb: help === root is now known as indus [15:24] indus, what's up? [15:25] i installed those kernel and now my screen flickers , but i dont see that kernel in grub [15:26] wher is apw [15:26] man [15:26] ? [15:26] indus, it would be a bit strange when a kernel, you not even see in grub would affect your system. sure there was no looseing of cables while switching back to slave [15:26] what kernels do you see [15:27] the kernel should have been a -15 right? [15:27] which should be separate from your real kernels [15:28] hmm he is gone [15:29] Loose power cable? [15:34] smb: just do me a favour, please go to that bug report and tell me name of kernel which andy gave [15:34] i have to remove it [15:35] x wont start for me now [15:35] indus, So nvidia module does not get rebuild correctly... a sec [15:36] yeah but i cant see that kernel in grub either, i booted normal kernel, i guess x starts after that doh ! [15:37] should be linux-image-2.6.31-15-generic_2.6.31-15.49~lp425756apw1 [15:37] as the package name [15:38] how come i dont see it in grub? [15:39] indus, Its a bit strange as the nvidia modules should get rebuild when you install that. All the kernels before were -14. And it should be in grub as well [15:39] hmm i have a 15, but it doesnt say fullk name [15:40] but i booted 14 and that too wotn get to x [15:40] The latest release kernel was -14.48, so everything saying -15 is new [15:40] anyways ill brb [15:44] ok solved [15:44] smb, but when i did apt-get remove and pressed tab, it only showed upto 15 , not the full name [15:44] phew [15:45] Glad, to hear. Though I cannot see how installing a -15 kernel would kill your -14 version... [15:45] indus, so what happened there? [15:45] smb, i swear it killled 14 too [15:46] apw, display trouble [15:46] yeah installing -15 would not normally have any way to touch -14 [15:46] apw, well , it happened i swear [15:46] question: to enable bluetooth on my laptop, i need to load a module. the problem is i have to do it everytime i reboot. is there anyway to load it automaticly? [15:46] indus, well lets call it day on testing, and just get you to report 'moving to master fixed it for me' to the bug [15:47] apw, :) [15:47] we can let the others play with -15 if they can't move to master [15:47] Darkness, Tried to add it to /etc/modules? [15:47] apw, iam willing to try it [15:47]