=== zul [n=chuck@ubuntu/member/zul] has joined #ubuntu-kernel [12:14] where can I get the .config that is is used to buiild the normal ubuntu kernel? [12:14] CarlFK: debian/config/ [12:14] all I want to do is update a wifi driver using http://acx100.sourceforge.net/ [12:14] thanks [12:14] CarlFK: barring that, there's a copy in /boot/config-$(uname -r) [12:15] how handy [12:16] should it be called .config or .configure or ? [12:18] depends, are you copying it to $source/. ? [12:18] if so, $source/.config [12:19] it actually depends how you're going to compile it [12:19] I was going to type make ;) [12:20] juser@dhcp36:~/src/ubuntu-2.6$ cp debian/config/i386/config.686 .config [12:20] make [12:21] few lines, then # using defaults found in .config [12:21] ah, no [12:26] are there any building instructions somewhere? [12:26] sorry, my wifi connection died [12:26] erase all the configs you're not going to use from debian/config/ [12:26] copy the one you want to use there [12:27] then debuild binary [12:27] into debian/config/ ? [12:27] yes [12:27] so this: juser@dhcp36:~/src/ubuntu-2.6$ cp debian/config/i386/config.686 debian/config/i386/.config [12:28] hmm, not exacly what you said... [12:29] CarlFK: nah, you can just erase everything save config.686 [12:29] err [12:29] sorry, I forgot you're not using your /own/ [12:29] ah, otherwise it would build one for each config? [12:29] you'll just want to erase config.{k7,server*} [12:30] yeah, I want to get a vanilla build to work before I start hacking on it [12:33] -bash: debuild: command not found [12:34] devscripts [12:36] should that be part of linux-kernel-devel ? [12:36] (no) [12:36] you could ``fakeroot debian/rules binary'', too [12:38] this is a form of "hack comfortably on the kernel." that I am not familiar with ;) [12:40] http://paste.foxshare.net:8888/732 - dpkg-checkbuilddeps: Unmet build dependencies: debhelper [12:40] you'll need that, too. === Keybuk [n=scott@quest.netsplit.com] has joined #ubuntu-kernel [12:46] @#$@$ [12:46] Can't exec "fakeroot": No such file or directory at /usr/bin/debuild line 845. [12:47] right, install that, too. [12:47] couldn't exec fakeroot debian/rules: [12:48] do I need to sudo debuild binary ? [12:49] http://paste.foxshare.net:8888/733 [12:49] something is still missing [12:52] I have 10 min to get to the3 bank... be back in 30 [12:55] CarlFK: it's recommended that you install fakeroot and use that instead of using sudo [01:13] back [01:14] I did install fakeroot [01:14] I didn't use sudo [01:14] but I didn't use fakeroot either [01:14] guessing I am building a kernel.deb? [01:15] linux-image-foo.deb [01:17] is this what I should do? juser@dhcp36:~/src/ubuntu-2.6$ fakeroot debuild binary [01:18] just ``debuild binary'' should suffice [01:19] http://paste.foxshare.net:8888/733 [01:20] ah [01:20] I rm'ed archmap [01:20] basically [01:21] this fixed it: juser@dhcp36:~/src/ubuntu-2.6$ cp debian/config.all/archmap debian/config/ [01:21] (should be obvious what I did...) [01:21] arg! [01:21] it is still prompting me for .config options [01:21] Prompt for development and/or incomplete code/drivers (EXPERIMENTAL) [N/y/?] (NEW) [01:22] do I need debian/config.all/i386/config.386 also? [01:23] yes. config and config.386 /unless/ you just copy your existing /boot/config-$(uname -r) over as config.686 [01:23] then you'll probably want to trim archmap [01:25] is this really the recommend way for someone to tweak there kernel? [01:26] there is no "recommended way" that I know of [01:26] I'm surprised. [01:27] not totally - just a little ;) === doko [n=doko@201.160.20.136.cableonline.com.mx] has joined #ubuntu-kernel [01:31] dpkg-checkbuilddeps: error: syntax error in control file debian/control at line 0: empty file [01:31] yup [01:31] -rw-r--r-- 1 juser juser 0 2006-05-15 18:30 debian/control [01:32] who's problem is that? [01:32] btw - if I ever get a clean build, Ill do it all again and make a wiki page [01:33] -rw-r--r-- 1 crimsun crimsun 39016 2006-05-12 12:44 control [01:33] -rw-r--r-- 1 crimsun crimsun 39016 2006-05-12 12:44 control.stub [01:39] -rw-r--r-- 2 juser juser 39016 2006-05-15 17:47 ./debian/build/build-686/debian/control [01:40] sh: line 2: cd: debian/build/build-386: No such file or directory [01:40] @#$@#$@ [01:42] found some docs: https://wiki.ubuntu.com/KernelCompileHowto?highlight=%28CategoryKernel%29 [01:44] "At this point, you need to change your kernel's configuration to statically include your bus, disk, and filesystem drivers. This can be rather difficult if you don't know what you're doing. Use "make menuconfig" (or "make xconfig", gconfig, etc.) to change the config." [01:44] why is that? [01:47] that's if you don't want to muck with initramfs === Keybuk wonders whether dapper boots without an initramfs === lifeless doubts it [01:50] Keybuk: Not for me, given my IDE driver's on it. :) [01:54] hmm, my goal may not be needed at all [01:54] https://launchpad.net/distros/ubuntu/+source/linux-source-2.6.15/+bug/22238 [01:58] hmm, last Changelog change is 2005 12 02, but there are a few 2006 versions in http://www.cmartin.tk/acx/ [01:59] I am going to rebuild the box and follow https://wiki.ubuntu.com/KernelBuildpackageHowto [02:00] see ya tomorrow [02:00] thanks for the help [02:04] man, I need a "bashing initng" channel [02:04] there's always ubuntu-users. [02:04] ooh, oops. Probably not. :-) [02:05] this thing is seriously scary [02:05] the fact people rely on it to boot their computer [02:05] for a start, it's totally backwards [02:06] rather than "begin with startup, and work your way through the list of things one by one until you run out of things to run" [02:06] it's "here's the last thing to run, now work out everything you need to get there" [02:06] eww [02:07] I like the fact that you just get a flashing cursor on tty1 too [02:07] it doesn't bother spawning, say, getty [02:07] the author clearly decided that was "too hard" [02:07] (it spawns it on tty2 through tty6 just fine) [02:08] lol [02:11] Keybuk: WEll, it's a dependacy based init system. [02:11] it's a crack based [02:11] no, it's worse [02:11] it's cornflour based :p [02:11] Dude, it's no worse than coding in Prolog ;) [02:11] And a frightening amount of stuff is written in prolog. =) [02:11] jbailey: try cobol [02:12] it just seems very strange to me to have a "goal" for booting [02:13] goal: gdm. [02:13] Do what it takes to get user to gdm. [02:13] everything I've learned suggests that you want the exact opposite; booting is just a state of mind [02:13] right [02:13] so what do you do about the stuff that's not in gdm's critical path? :p [02:13] do you run it, or do you not run it? [02:13] (initNG chooses not to run it ) [02:13] Don't run it. [02:13] It's not something you've said you're interested in. [02:13] yeah [02:14] How much cruft would we have saved from our bootup over the years if we did that? [02:14] it means you have to set a *huge* list of goals [02:14] indeed, initNG comes with a shell script that just sets the list of goals to the list of init scripts [02:14] simply because it doesn't work otherwise [02:15] I think the opposite is better [02:15] "right, we've started -- does anything care?" [02:15] "a network device just got plugged in -- did anything need one of those?" [02:16] Autconfiguring every network device that shows up isn't necessarily a secure system. [02:16] On a user based system that might be more correct. [02:16] On a server based system it's arguably not. [02:16] a network device showing up kinda implies physical access :p [02:18] True, and trusted armed guards are the final line of defense. True, [02:18] wtf? [02:19] *giggles* [02:19] oh, wow [02:19] initNG includes a kernel patch [02:19] Eh? [02:19] to allow you to attach gdb to process 1 [02:19] *gawp* [02:19] ummm...okie dokie [02:19] there's a REASON the kernel doesn't let you do that! [02:20] Keybuk: cap_ptrace will stop you doing that. [02:20] So they'd have to be taking out chunks of security/. [02:21] /* [02:21] * You may not mess with init [02:21] */ [02:21] if (pid == 1) [02:21] return -EPERM; [02:21] it's not so much just security ... it's the fact the kernel relies on being able to attach processes to init and have them reaped [02:21] if init is stopped in ptrace, you'd dead lock the kernel [02:22] also seclvl.c: * Explicitely disallow ptrace'ing the init process. [02:37] sweet [02:37] I like it when people not only imitate djb's coding style, but also even his lack of web page design [02:51] oh, and this replacement init is written in scheme! [02:52] proof that initNG isn't the most crackful thing in its field [02:52] jbailey probably likes this one :p [02:52] Hmm? [02:52] hehe [02:52] Oh dear. [02:52] Tell me they compiled it? [02:53] it says it's intended for the Hurd [02:53] which explains a lot :p [02:53] Oh, haha. [02:53] Is that the one written by Jeroen? [02:54] Wolfgang Jhrling [02:54] Ah, I met him at LinuxTag a week ago finally. [02:55] Anyhow, I need to clean up. houseguests in two days. [04:16] jbailey: ping [04:50] shouldn't linux-kernel-devel be somewhat related to one of the https://wiki.ubuntu.com/KernelCompileHowto pages? [04:54] apt-get source linux-source-2.6.8.1 [04:55] is there a meta package for 'current version' ? [04:55] linux-source-2.6. [04:56] thanks === bluefoxicy [n=bluefox@c-68-33-112-13.hsd1.md.comcast.net] has joined #ubuntu-kernel === doko [n=doko@201.160.17.131.cableonline.com.mx] has joined #ubuntu-kernel [05:06] I forgot, how do I build a kernel package from source again? apt-get source ...? [05:09] pick one of these https://wiki.ubuntu.com/KernelCompileHowto [05:09] I am using https://wiki.ubuntu.com/KernelHowto [05:09] sudo apt-get install linux-source [05:10] that got me the source in /usr/src [05:12] BenC: Pong, but about to head to bed. [05:13] BenC: 'sup? [05:21] hey [05:21] jbailey: any crashes? [05:22] Not so far. [05:22] I've done the usual things today, I think. [05:22] cool, just wanted to check in [05:22] I'm about to upload [05:22] Congrats if you've licked this one. =) [05:23] I've reenabled the screensaver for overnight, too. [05:23] It didn't stop the crashes before, but who knows. [05:23] That way it'll get a bit of a GL workout, too. [05:27] alright [05:27] going to try patching this http://people.redhat.com/mingo/exec-shield/exec-shield-nx-2.6.15.patch [05:28] Into ubuntu's kernel via https://wiki.ubuntu.com/KernelBuildpackageHowto [05:28] If it works I'll make note on https://wiki.ubuntu.com/UbuntuDownUnder/BOFs/ProactiveSecurityRoadmap and I guess we'll see what happens in Edgy or such. [05:29] the patch itself needs testing; other softwares' reactions to this is thoroughly tested already, as i386 ubuntu on any architecture with a real NX bit already turns the NX bit on and signifies the PROT_EXEC protection [05:30] Not enough time to get a good thorough test cycle in before Dapper though. === doko_ [n=doko@201.160.17.131.cableonline.com.mx] has joined #ubuntu-kernel [05:39] ah screw it, I'll get it tomorrow or something [05:39] need to sleep now. === _human_blip_ [n=mike@220.157.65.236] has joined #ubuntu-kernel === ivoks [n=ivoks@ubuntu/member/ivoks] has joined #ubuntu-kernel === kimo [n=ahmed@82.129.234.122] has joined #ubuntu-kernel === _human_blip_ [n=mike@220.157.65.236] has joined #ubuntu-kernel [09:42] I'm still having problem powering of my toshiba A105-S361 laptop (Bug 31993). It's very irritating every time I turn off my laptop. I am willing to help anyone by testing any fixes on my machine.. === lloydinho [n=andreas@rosinante.egmont-kol.dk] has joined #ubuntu-kernel === _human_blip_ [n=mike@220.157.65.236] has joined #ubuntu-kernel === ivoks [n=ivoks@ubuntu/member/ivoks] has joined #ubuntu-kernel === ivoks_ [n=ivoks@lns01-0505.dsl.iskon.hr] has joined #ubuntu-kernel === jbailey [n=jbailey@modemcable139.249-203-24.mc.videotron.ca] has joined #ubuntu-kernel [01:00] BenC: ping when you're awake. Interesting twist to the bug this time. [01:04] jbailey: radeon bug? [01:04] airlied: I *think* so. [01:04] But different than it's ever been. [01:04] different bug then :-) [01:05] what does it do now? [01:05] Usually either my machine is locked solid, or X is pushed to 100% utilisation and ssh;ing in to kill it can take 15 minutes. [01:05] yeah, that more like a real hang.. anything in particular cause it? [01:05] No. That's the problem. [01:06] I can't force it to reproduce, although it does so consistantly within a day or so, and often within an hour. [01:06] When the bug first started happening it could be when I was typing. Now adays it's more likely when I'm idle. [01:06] jbailey: what chip is it? the M10? [01:06] I had disabled the screensavers and stuff to see if it was GL triggered, but that hadn't stopped it. [01:07] 0000:f0:10.0 VGA compatible controller: ATI Technologies Inc RV350 AR [Radeon 9600] (prog-if 00 [VGA] ) [01:07] This time what happened what that I cam back to my machine and tapped my mouse. Black screen. [01:07] Went to the other machine, and ssh'd in. on of the glx screensavers was running, 60% utilisation. Nothing big. [01:07] kill worked fine. [01:07] (didn't even need kill -9) [01:08] like I usually do with X [01:08] Came back to my machine and the keyed up C-M-F1 to switch consoles had taken effect. [01:08] When I switched back to my graphical session, I noticed that my IRC sessions had all hung at the time when I'd moved the mouse. [01:08] So a bunch of them, the last traffic I saw was 06h58. [01:09] and then disconnected. [01:09] reconnected at 06h59 when I killed the screensaver. [01:10] I have no conclusions as to what the problem is, only that it's now different. [01:10] so the X server didn't crash out .. wierd.. [01:10] Right, like I'm still on that session now. [01:13] do you have DRI working? i.e.does glxinfo give Direct Rendering: Yes [01:14] direct rendering: Yes [01:15] I do get a strance warning as well,. though: [01:15] *********************************WARN_ONCE********************************* [01:15] File r300_state.c function r300Enable line 456 [01:15] TODO - double side stencil ! [01:15] *************************************************************************** [01:15] And glxgears doesn't print frame rates. It just prints that warning as well. [01:16] ah gears has fps disabled.. do you remember what screensaver it was? === smurf [n=smurf@debian/developer/smurf] has joined #ubuntu-kernel === zul [n=chuck@ubuntu/member/zul] has joined #ubuntu-kernel [02:13] heylo === jbailey [n=jbailey@modemcable139.249-203-24.mc.videotron.ca] has joined #ubuntu-kernel [02:18] BenC, airlied: Had my first full machine hang. === ivoks [n=ivoks@ubuntu/member/ivoks] has joined #ubuntu-kernel === zul decides to take the rest of the week off === tuxmaniac [n=aanjhan@60.254.67.17] has joined #ubuntu-kernel === human_blip [n=mike@220.157.65.236] has joined #ubuntu-kernel === mgalvin [n=mgalvin@ubuntu/member/mgalvin] has joined #ubuntu-kernel === ivoks [n=ivoks@ubuntu/member/ivoks] has joined #ubuntu-kernel === CarlFK [n=carl@c-67-163-39-124.hsd1.il.comcast.net] has joined #ubuntu-kernel [03:30] CC [M] drivers/usb/net/zd1211/zdmic.o [03:30] DEVLIST drivers/usb/net/zd1211/zddevlist.h [03:30] make[4] : *** [drivers/usb/net/zd1211/zddevlist.h] Error 1 === ivoks [n=ivoks@ubuntu/member/ivoks] has joined #ubuntu-kernel [03:31] that was from linux-source-2.6.15.tar.bz2 [03:34] #error "Error in source file, line 35" [03:35] huh? === tuxmaniac [n=aanjhan@60.254.67.17] has joined #ubuntu-kernel [04:01] jbailey: suckage [04:04] crimsun: ping [04:04] Mithrandir: ping [04:04] CarlFK: Install the linux-source-2.6.15 build-deps [04:04] BenC: pong [04:05] BenC: flight 8 is cancelled due to the sudden leave on Friday. [04:05] Mithrandir: ah, answers my question...I'll be doing a kernel upload today then [04:05] BenC: please do go ahead [04:06] sudden leave? [04:06] zul: Canonical employees and have been forced to take Friday off :) [04:06] ah...ok.. [04:06] zul: the distro team was told to take Friday off. Anybody who is active on ubuntu work that day and detected by mdz gets to buy mdz a drink in Paris [04:07] i thought it was something catastropic === ivoks [n=ivoks@ubuntu/member/ivoks] has joined #ubuntu-kernel [04:07] hehe...i have friday off as well [04:07] I'm going to make a small git push on Friday just to see if he's watching [04:07] heh [04:07] the catch is, if he notices, then he didn't take off either, so he'll owe me a drink too :) [04:11] BenC: think I could talk you into accepting a couple of patches? in launchpad on 4478[23] , a couple of bugfixes (one applied in 2.6.16 even) that are quite significant to us. [04:12] maswan: got both of them already [04:12] BenC: excellent! [04:12] BenC: what is "the linux-source-2.6.15 build-deps" ? [04:12] CarlFK: apt-get build-deps linux-source-2.6.15 [04:13] or maybe it's just build-dep [04:13] BenC: Should I update launchpad status, or do you want to fiddle with that yourself? [04:13] BenC: build-dep [04:13] BenC: When are you planning the kernel upload? [04:14] BenC: I wanted to make one last small tweak to vga16fb timings, but if I'm too late and miss the release, it's not the end of the world. [04:14] infinity: I'm getting ready to start my last build cycle before upload [04:14] so you've got about an hour or two [04:14] (since it more-or-less works for more people than it more-or-less worked for in breezy, I'm not fussed about fixing it for the last 2 or 3 people it's broken for) [04:15] maswan: I'll get it in a few minutes, thanks [04:15] BenC: Awesome. Thanks for the great work! [04:15] BenC: Og, an hour or two isn't long enough to test the changes I wanted to make and get 'em in, so don't worry about it. [04:15] s/Og/Oh/ [04:15] infinity: ok, I'm sure there will be one more upload before release, so there's be another chance [04:16] BenC: Alright, if I find the free time. It's currently prioritised below more important stuff, but it's been something I was "meaning to do before release" and never got to. [04:16] BenC: I suspect I'll have a lot of those left over on June 1st, though. :) [04:16] probably :) [04:39] wow [04:39] I might have almost passed linear algebra, I think. [04:39] I counted 126 points on my final, I need 126 to pass. o_o === mxpxpod [n=BryanFor@unaffiliated/mxpxpod] has joined #ubuntu-kernel === mxpxpod_ [n=BryanFor@wuw-ojr3gmca.dybb.com] has joined #ubuntu-kernel === mxpxpod__ [n=BryanFor@wuw-ojr3gmca.dybb.com] has joined #ubuntu-kernel [05:03] bluefoxicy: Relying on your counting abilities to guess if you passed a math related exam seems counter intuitive :) [05:31] if all I want to do is build a module, I just need the kernel headers, right? [05:32] yes [05:32] thanks [05:39] Is there a ubuntu package that is the kernel source without any patches? [05:39] (I don't want it, just trying to see how it is described) === CataEnry [n=cataenry@host46-6.pool80183.interbusiness.it] has joined #ubuntu-kernel [05:41] BenC: heh. [05:46] CarlFK: no such package, since kernel.org has all that stuff [05:47] would that be called "the Vanilla Kernel?" [05:48] yes... === tuxmaniac [n=aanjhan@60.254.67.17] has joined #ubuntu-kernel === mxpxpod_ [n=BryanFor@wuw-ojr3gmca.dybb.com] has joined #ubuntu-kernel [05:59] Mmm. vanilla. === jbailey wants icecream now. === zul is having lunch === ivoks [n=ivoks@ubuntu/member/ivoks] has joined #ubuntu-kernel [06:06] what is the difference between kernel-source-2.6.11 and kernel-tree-2.6.11 ? [06:06] nm [06:07] tree = source + patches [06:09] wait.. source = "Linux kernel source for version 2.6.11 with Debian patches" http://packages.ubuntu.com/breezy/devel/kernel-tree-2.6.11 [06:09] so what is the point of tree? [06:17] There is no more tree, so you can stop worrying about it. :) [06:17] good answer === CarlFK thinks "there is no spoon' [06:18] SPOOOON [06:18] - the tick [06:24] hehe... [06:24] spooooooooooon === mxpxpod [n=BryanFor@unaffiliated/mxpxpod] has joined #ubuntu-kernel [06:31] BenC: pong [06:33] crimsun: The one-liner patch failed to build [06:33] guh [06:34] BenC: what is the error message? [06:34] BenC: Line 608 of drivers/acpi/asus_acpi.c - can you remove the "Asus ACPI: Error reading LCD status" printk? [06:35] It doesn't work on some machines and they spew errors [06:35] crimsun: The member of that struct doesn't exist [06:35] mjg59_: ok [06:35] Ta [06:35] saves me sending you a patch :) [06:37] mjg59_: done [06:40] BenC: (just taking a sec to update here) [06:49] BenC: I see the proper fix, I'll push out a patch ASAP (ETA: 10 mins) [06:50] BenC: sorry about that [06:50] ok [06:50] np [07:06] if I build a module using linux-headers-2.6.15-22-386, what happens when a new kernel package is released? [07:06] You get to rebuild your module to match the new kernel. [07:07] thats what I thought. thanks [07:07] (Assuming the ABI changed, which we try very hard to avoid during stable releases) === ivoks [n=ivoks@ubuntu/member/ivoks] has joined #ubuntu-kernel [07:35] i have an rem song in my head [07:35] make it stop!! [07:37] what is the difference between: === Keybuk [n=scott@quest.netsplit.com] has joined #ubuntu-kernel [07:37] apt-get install linux-source-2.6.15 [07:37] and [07:37] apt-get source linux-image-2.6.15-22-386 [07:37] the actual source is the same [07:37] just one way installs the package, and the other downloads the files that actually build the images [07:37] so just the location ? [07:38] ah [07:39] later...im on my way home === Keybuk [n=scott@quest.netsplit.com] has joined #ubuntu-kernel [07:55] BenC: sent. === JaneW [n=JaneW@dsl-146-177-94.telkomadsl.co.za] has joined #ubuntu-kernel [08:08] BenC: Just had the screensaver hang agian. I think it might've been gleidoscope both times. [08:08] crimsun: wow, one-liner went to a pretty hefty patch :) [08:08] jbailey: Damn, there goes the hope for updated DRM fixing things :/ [08:09] BenC: yeah, I dunno what crack I was smoking the first time w/ grep struct [08:09] BenC: Yeah. I'll probably take it out on next reboot. === allee [n=ach@allee.exgal.mpe.mpg.de] has joined #ubuntu-kernel === BenC [n=bcollins@debian/developer/bcollins] has joined #ubuntu-kernel [08:51] BenC: you still only have 2.6.16 DRM, the radeon changes are only in my tree and 2.6.17.. [08:52] airlied: You haven't provided me with any info on how to obtain that, or a patch to update my tree [08:53] BenC: okay I'll do a patch later, my tree is is the drm-2.6 one on kernel.org.. [08:55] is it possible to pull from it without getting anything past 2.6.15? [08:56] BenC: probably not... [08:57] BenC: I'll cook a patch later when I'm actually awake as opposed to no when I should be asleep.. [08:59] can someone review https://wiki.ubuntu.com/KernelHowTos - it is a summary of what I have learned here over the last day [08:59] might be too late to get it in, but I'll try === JanC [n=janc@lugwv/member/JanC] has joined #ubuntu-kernel [09:37] http://acx100.sourceforge.net is a wifi nic module - it's README says I need to build it with the rest of the kernel: http://dev.personnelware.com/carl/temp/May16/a/README [09:38] shouldn't I be able to do it with just the headers? [10:12] yes [10:12] you do know we have acx100 in our kernels, right? [11:29] yes - but it is about 8 updates old === human_blip [n=mike@220.157.65.236] has joined #ubuntu-kernel [11:34] current is v0.3.35 [11:35] acx_config.h:#define ACX_RELEASE "v0.3.21" [11:35] .21 came from rsync.kernel.org/pub/scm/linux/kernel/git/bcollins/ubuntu-2.6.git === zul [n=chuck@ubuntu/member/zul] has joined #ubuntu-kernel [11:56] heylo