[01:18] * ogra has massive probs trying to build amitk's linux-lpia package, wasnt the bi check dropped in intrepid ? [01:20] http://launchpadlibrarian.net/18547672/buildlog_ubuntu-intrepid-lpia.linux-lpia_2.6.27-4.6_FAILEDTOBUILD.txt.gz shows that it fails in abi-check-lpia even though i added appropriate ignore files [01:27] * ogra gives up and hopes someone else will then build a -4 kernel for lpia [02:14] Hola [02:15] alguien sabe donde descargar el 2.6.27 x64 deb [02:28] hi [02:28] are there some test versions for 2.6.27 yet? im on 8.04 adm64 === rafal is now known as limcore [03:08] limcore, Test versions for 2.6.27 exist only in 8.10. 8.04 is exceedingly unlikely to receive them. [03:10] can I somehow test 8,10's kernel and perhaps drivers, while remaining (other software) at 8.04? [03:19] hello [03:19] my question is about kernel devel but about the kernel i think [03:19] i just updated my kernel and now my printer doesn't work === bdmurray_ is now known as bdmurray [07:39] I have some questions on how the ubuntu git tree rebases new releases into the tree [07:39] * NCommander is attemtping to figure out how to rebase the linux-ports tree [07:40] The log on some of the tags suggests its merged and not rebased on a new release, but I'm not sure if thats right or not === mdz_ is now known as mdz [10:32] can anyone tell me what this means? [10:32] ACPI: EC: missing confirmations, switch off interrupt mode. [10:32] ACPI Exception (evregion-0419): AE_TIME, Returned by Handler for [EmbeddedControl] [20080609] [10:32] kernel problem or BIOS problem? === asac_ is now known as asac [11:02] mdz, they don't exactly have a lot of helpful comments in there do they [11:03] Anyone here a kernel git wizard? [11:03] * NCommander is trying to figure out how to rebase the git repo without it exploding in my face [11:04] which repo you looking at [11:06] linux-ports [11:06] * NCommander kicks xchat [11:06] I'm trying to rebase from 2.5.24 which is the basis of that repo against linus 2.6.27 tree [11:07] But I don't think I"m doing it right since I either seem to pull old patches into the tree, or it stops every patch with a conflicts [11:07] thats a big jump [11:07] can you paste the url you cloned the original from [11:07] then i can try it and see what it does to me [11:08] Yeah [11:08] But the ports tree has relatively few Ubuntu specific patches [11:09] apw, http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-intrepid-ports.git;a=summary [11:09] (thats the ubuntu/ubuntu-intrepid-ports.git branch) [11:09] Personally, my gut instinct is to merge, but thats a bad idea since the Ubuntu patches will get pushed into the history instead of sitting ontop [11:10] my gut concurs [11:10] * NCommander notes he has a dislike for git [11:11] Damn tool nearly as painful to use as CVS when it comes to preventing yourself from shooting yourself in the foot [11:11] and it used to be close to arch in terms of user friendlyness :-P [11:13] they do appear to be merging the stable branch in there [11:14] The ports repo has seen no love this release [11:15] Myself and TheMuso are working to make at least powerpc ports not suck [11:20] my rebase stops on the first patch with what appears to be a trivial merge which i would have expected git to detect and merge [11:21] yeah [11:21] Same problem [11:21] It looks like they did SOMETHING to break the repos history [11:21] But of which I'm am unsure [11:22] I'm tempted just to take the debian folder, and start with a fresh 2.6.27 tree [11:24] hey BenC [11:25] BenC, I was wondering if you can help solve a rebasing mystery with linux-ports (and help create a kernel.u.c account) [11:47] NCommander: went through the whole rebase, can't say i was very careful but the main pain is near t [11:47] the beginning, most of the merges seemed pretty and most of the last bunch of patches rebased ok === BenC1 is now known as BenC === pgraner_ is now known as pgraner [14:59] rtg: the patch is attached to the email I sent to the kernel team mailing list [15:00] zul: i saw that, but I'm lazy and git pulls take 2 commands. [15:00] rtg: im lazy too and I dont have the git tree setup for intrepid [15:01] zul: so what is this: http://kernel.ubuntu.com/git?p=chucks/ubuntu-intrepid.git;a=commit;h=73d6bbb5e0dcdb902c37ef2bc2925de600dd0172 [15:01] old junk git tree [15:03] Hi guys, i have a problem... i have installed Kernel-2.6.24.21-rt on Ubuntu Hardy by linux-rt package. After a while time, the system crashes and generate I/O Error, and i can not run any process. Sorry for my bad English. [15:07] D2_WON: Can you report exactly the I/O Error? === limcore is now known as limcore_ [16:02] persia, rtg : I just uploaded a synced version of the rt patchset on my PPA (against 2.6.27 7.11) together to lrm-rt package. Moreover I have tested rt kernel with DKMS packages (in my case with nvidia video driver) and it works. [16:42] Is there a new feature in intrepid to change the clock speed of my nvidia GPU? or memory? [17:39] BenC: so what is the new firmware package? my i3945 isn't working too well after an update a few minutes ago. [18:12] rtg: linux-firmware, but it isn't up yet [18:13] BenC: its breaking my rice bowl. no firmware, no wireless. [18:41] BenC: I sure like this USB stick as an install medium. cuts it down to about 8 minutes. [18:43] I need to set mine back up [18:44] I wonder when the next publisher run is...linux-firmware is still pending [18:44] rtg: do you want me to just send you the firmware file? [18:44] BenC: I've been using Evan's usb-creator package. [18:45] BenC: I'm just reinstalling. I had a version that worked, so I'll just avoid updating. the i3945 issue is related to kernel and not much else. [18:45] rtg: zinc:~bcollins/ if you need it [18:45] BenC: 'tanks [18:57] What source package provides linux-firmware? My test server (running the server kernel) can't seem to find it and so is a bit stuck at the moment. [18:58] ScottK: wait until the archive catches up. its in a bit of flux right now. [18:58] OK. === ogra_ is now known as ogra [19:47] has beenc been around? [19:47] benc [19:53] hey ben :) [19:53] speaking of the devil :) [19:53] i know him [19:53] its been quite awhile since i spoke to him tho [19:53] so my sentence is even more true :) [19:53] yeah! [19:53] (he hides his little horns under a basecap since i know him ;) ) [19:54] heheheh [19:54] right now i am thinking of fooling around with the kernel as usual :) [19:54] fix this nvidia bug once and for all [19:54] its driving me crazy [19:55] just get better HW :) [19:55] * ogra is all intel on his laptop [19:55] hehehehe [19:55] i wish :P [19:56] if i had the cash, i would buy all amd [19:56] and use a company that supports linux [19:56] heh [19:57] timecop on that other network i chat on, will be mad that i ditched windows [19:57] but vista sucks [19:57] this is much nicer [19:58] yeah [19:59] timecop? i think that rings a bell [19:59] yeah ben and i know him [19:59] he is on efnet [19:59] the GNAA guy? [19:59] i hardly ever go on efnet anymore [19:59] yeah [19:59] oh god ;) [20:00] i dont op on his channel lwz anymore [20:00] i dont know if its apropriate to use teh full channel name here on freenode [20:00] dont know the aup [20:01] the network i chat on now, does not like people to mention anything about p2p and other grey area things.. [20:01] laga_: did you talk on lwz on efnet before? [20:02] no. [20:02] i used to op there many moons agao [20:02] before timecop was in GNAA [20:04] laga: i quit going there because my interests are not in computers as much anymore [20:18] rtg: I tried unloading and reloading iwl3945 and its dependencies a few hundred times on davidm's machine, and can't get it to crash [20:19] mdz: BenC and I are speculating its an interrupt problem. He pointed out a bug report that says something else: http://www.intellinuxwireless.org/bugzilla/show_bug.cgi?id=1778#c15 [20:20] mdz: It seems to be a race with other modules loading in parallel [20:20] guessing on the same IRQ (not much reason for a race otherwise) [20:22] mdz: can you 'cat /proc/interrupts' on Davids machine? [20:22] chrisf_: if you want to go all opensource, buy intel :) [20:22] yeah :) [20:22] ben: i guess iam behind the times LOL [20:22] BenC: Long time no see! :) [20:24] rtg: http://paste.ubuntu.com/58006/ [20:24] 217: 99246 0 PCI-MSI-edge iwl3945 [20:25] mdz: ah, try booting with nomsi (after removing the /etc/modprobe.d load delay) [20:26] hmm...MSI [20:26] mdz: rather, thats 'pci=nomsi' [20:27] we had to disable MSI by default on a previous release, but I think it was affecting e1000 [20:27] chrisf_: yeah, been awhile [20:27] yeah, but the 3945 is about the same vintage [20:29] ben: congratulations in being a part of the ubuntu team :) [20:29] rtg: oooh...someone on the original bug report did mention that disabling iwl3945(rfkill) and using e1000 resulted in a hang as well [20:30] chrisf_: Thanks, been here more than 3 years now :) [20:30] ben: now i have two things i want to play with ubuntu muhahha installing WoW and getting nvidia's piece of trash top run on this [20:31] i am already having WoW withdrawals [20:31] davidm: ^^ [20:31] chrisf_: install Ubuntu, and you should get nvidia enabled without much problem [20:31] yeah i have ibex on here now [20:32] and nvidia is being a little *** to me [20:32] chrisf_: nvidia-glx-177 ? [20:32] chrisf_: System => Administration => Hardware Drivers [20:32] naww card is too old [20:32] nvidia-glx-96 [20:32] chrisf_: that panel will tell you exactly which one you need, and install it for you [20:33] rtg: I've spoken to davidm, and have him trying two things [20:33] yeah it wont install or compile the module [20:34] chrisf_: what's the compile log say? [20:34] mdz: make sure he's remove the modprobe load delay. [20:34] rtg: 1. reproducing the failure with an instrumented modprobe, so that we can try to see which modules are being loaded in which order at the point when it hangs [20:34] rtg: 2. pci=nomsi [20:34] rtg: I already did [20:34] mdz: cool [20:34] mdz: that's likely to be enough to offset the race to not happen [20:34] BenC: I told him how to back it out in case it doesn't trigger anymore [20:37] Error! Bad return status for module build on kernel: 2.6.27-7-generic (i686) [20:37] Consult the make.log in the build directory [20:37] /var/lib/dkms/nvidia/96.43.05/build/ for more information. [20:37] Installing initial module [20:37] Error! Could not locate nvidia.ko for module nvidia in the DKMS tree. [20:37] You must run a dkms build for kernel 2.6.27-7-generic (i686) first. [20:38] Done. [20:38] chrisf_: see this bug report: https://bugs.launchpad.net/bugs/251107 [20:38] it doesn't depend on us [20:39] yeah nvidia [20:39] hasnt fixed there stuff [20:47] BenC,rtg: he's just hung it with pci=nomsi [20:47] drat [20:47] ben: hehe i am no longer in florida [20:47] this place is for the birds [20:48] i hate the cold weather! [20:51] rtg,BenC: is there any way to get the kernel to log a message whenever a module is loaded? that would be handy here [20:53] mdz: I'm not seeing anything. [20:58] rtg,BenC: I've just noticed that david has a modified boot process which is loading modules in parallel with the rest of the boot process [20:58] mdz: how is his different then Jane's ? [20:59] rtg: I don't have access to Jane's system at the moment, but I highly doubt she would have that change [20:59] it might be a useful way to try to trigger the problem though, dunno [20:59] mdz: I'm sure her's is stock. [21:01] rtg: you can try it by commenting out the logic which calls udevsettle in /etc/init.d/udev [21:03] rtg,BenC: do we have any strong indications that either silbs or davidm's problem is actually related to 3945? [21:03] mdz: yes [21:04] mdz: sleeping 5 seconds before loading 3945 solves it [21:04] BenC: given that it's a race, that workaround you provided could easily throw things out of whack and cause it to not trigger [21:04] even if the problem isn't in 3945 [21:04] mdz: the bug report has 10 dupes and lots of subs that all have found it to be iwl3945 (blacklisting other modules doesn't help) [21:05] 3945 is the common factor [21:05] BenC: the description says that -1 worked while -2 didn't. has anyone bisected? [21:05] or reviewed the diff? [21:06] hmm aptitude is pretty nice [21:06] I haven't [21:06] but then, I can't reproduce it either [21:07] ben: from what i am reading, nvidia really doesnt want to support legacy users [21:08] mdz, rtg: Got to get a kid off the bus and fix dinner... [21:13] iwl3945 didn't change at all between -1.2 and -2.3 [21:13] mdz: as far as I can tell, it's had no love in awhile. [21:14] well i gotta reboot, as i was speaking they updated the nvidia packages [21:14] reboot [21:15] mdz: 32 bit kernel? [21:21] rtg, in conversation with mdz I just realized that while I have the Intel Corporation PRO/Wireless 3945ABG hardware I am not really using it as my machine is hooked via Ethernet [21:21] davidm: its still loading, and that is when you are getting hung up. [21:22] davidm: that's a 32 bit kernel, right? [21:22] yes [21:22] 32 bit machine [21:22] davidm: maybe thats why I can't repro. I'm resinstalling with i386 [21:23] my Ethernet controller is Intel Corporation 82573L Gigabit Ethernet Controller [21:24] davidm: e1000e ? [21:25] Yes [21:26] This 'pci=nomsi' did not effect the bug, still happened. [21:26] davidm: yep, heard that. [21:26] rtg, OK, just making sure :-) [21:41] davidm: drat, I'm 10 for 10 on a 32 bit kernel. no hangs. I'm gonna blast off for awhile.back in a couple hours. [21:50] davidm: please put that info into the bug as well, bug 263059 seems closest [22:01] hold BenC [22:01] *hola [22:36] BenC,rtg: davidm just managed to reproduce without usplash, on a text console [22:39] BenC,rtg: seems like a hard hang, no sysrq possible [22:40] BenC, I sent you an email requesting a kernel.u.c account so I can work on the linux-ports tree for PowerPC. [22:40] I'm having him attach a photo to the bug [22:50] rtg: he has it reproducing on the console more often than not. maybe we should try instrumenting the driver to see where it's breaking? [22:51] hellp [22:51] hello [22:51] that is [22:51] rtg, my laptop is now doing 3+ out of 5 locking at boot with usplash not in boot so you can see much better what is going on. [22:52] rtg, do you or pete want my laptop for testing? I can fedex out tonight [22:59] rtg, pgraner either of you about? [23:00] Lovely. [23:00] davidm: tried phoning pgraner? [23:00] I did no answer home or cell [23:00] I'm about to call Tim [23:11] davidm: random theory: could you try temporarily disabling the hwclock init script, which runs directly after S10udev, and see if that affects it? [23:14] Will do [23:16] about to reboot. [23:30] OK that was fun, mdz no real change, first two boots good, next 4 NG then a good one and I'm here reporting. [23:30] davidm: thanks for checking [23:30] davidm: since you've had a few more boots, can you comment on https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/263059/comments/95 ? [23:30] NP, for sure it's more broken then not at this point. [23:31] Will do [23:31] davidm: if we can confirm that, it should narrow the field a lot, and perhaps rtg can chase it down [23:32] It stops in differnt places, sometimes at the channels line, sometimes past that into another driver [23:33] davidm: I know, but I'm only interested in the lines coming from 3945 and 80211 [23:33] davidm: the normal sequence is: [23:33] [ 17.379861] iwl3945: Detected Intel Wireless WiFi Link 3945ABG [23:33] [ 17.523620] iwl3945: Tunable channels: 11 802.11bg, 13 802.11a channels [23:33] [ 17.524272] phy0: Selected rate control algorithm 'iwl-3945-rs' [23:33] but I have never seen it get as far as that third line when it hangs [23:34] OK, I'll go punch the reboot button a bunch more and pay close attention to the lockup points. [23:34] back in a bit. [23:34] I have my shipping box by the by so I can send this unit out. [23:56] OK, documented a series of hangs one after the other.