/srv/irclogs.ubuntu.com/2008/01/02/#ubuntu-kernel.txt

=== asac__ is now known as asac
=== asac_ is now known as asac
=== asac_ is now known as asac
goldphishIs there a way to use menuconfig to configure a kernel and still use debian/rules to build it?02:37
johanbrgoldphish: I usually run menuconfig (or whatever) first and then do "fakeroot make-kpkg --initrd kernel_headers kernel_image".02:45
goldphishI tried that but then linux-restricted-modules and linux-ubuntu-modules fails in random ways02:47
goldphishThis whole process is way more complex than it needs to be. All I was is a single module that debian was too stubborn to include in their kernel. (it's in the vanilla)02:48
johanbrgoldphish: So just build that single module, then.02:55
goldphishhow?02:59
johanbrWhich module is it?03:00
goldphishkeyspan03:00
goldphishit's in the kernel source03:01
goldphishso there is no seperate package for it03:01
goldphishthere was some political drama over licensing and Debian decided not to include it. Everyone else is fine with it03:01
mjg59goldphish: That implies that this isn't really the right place to be asking...03:02
goldphishmjg59: why not?03:02
mjg59Or is it also disabled in Ubuntu?03:02
mjg59We don't use Debian's kernels03:02
goldphishyeah03:02
goldphishwhoops, I meant to say it's not included in Ubuntu kernels either03:03
goldphishbut the problem originated with Debian03:03
johanbrThere is a module named keyspan in the Feisty kernel I'm running right now.03:03
goldphishinteresting; it's missing in the gutsy kernel03:04
goldphishaka 2.6.22-14-generic03:04
johanbrBug report, including a link to a deb someone built: https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.22/+bug/13210603:06
ubotuLaunchpad bug 132106 in linux-source-2.6.22 "[gutsy] keyspan serial adapter not detected" [Low,Triaged] 03:06
goldphishnice03:07
shenkihow would one go about finding what git revision a released kernel comes from?04:44
goldphishI created a kernel using make-kpkg. How do I go about rebuilding linux-restricted-modules? It wants to build for "flavours" though there really isn't one05:38
=== XSource_ is now known as XSource
=== asac_ is now known as asac
=== fabbione is now known as thegodfather
krautmoin08:05
=== XSource_ is now known as XSource
=== luka74 is now known as Lure
=== amit1 is now known as amitk
=== chuck_ is now known as zul
zulum...http://launchpadlibrarian.net/11049559/buildlog_ubuntu-hardy-powerpc.linux-restricted-modules-2.6.24_2.6.24.2-2.8_FAILEDTOBUILD.txt.gz13:15
=== \sh_away is now known as \sh
sorenIt seems https://lists.ubuntu.com/archives/kernel-team/2007-November/001891.html was never applied. Could someone do that please?15:40
sorenBenC said it's ok (in the e-mail I'm replying to at https://lists.ubuntu.com/archives/kernel-team/2007-November/001895.html)15:41
rtgsoren: got it15:44
sorenrtg: Thanks very much!15:46
=== chuck is now known as zul
BenCsbader: Welcome...how was your holiday?16:41
sbaderThanks. Holiday was quite well (but catched a cold at the end)16:42
rtgsbader: glad to have you on board.16:49
=== \sh is now known as \sh_away
crispinBenC__: I'm trying to get the fusion driver working in dapper on a Sun x4100, but have found what I think is a bug in the linux-backports-modules package - the mutex_init() wrapper calls sema_init(..., 0), that locks up the booting for me16:52
crispinshouldn't it be sema_init(..., 1) ?16:52
BenC__crispin: that's possible...any chance you could test that?16:53
=== BenC__ is now known as BenC
crispinyeah, it boots16:53
crispin(with 1)16:53
BenCdoes the driver work though?16:53
crispinnot quite - it claims I have a 'generic sg1 type 0' scsi device16:53
crispinit doesn't find the partitions16:54
crispinwith the 0 in the sema_init it locks up after a line like: "scsi0: ioc0: LSISAS1064: FwRev=...., Ports=1, MaxQ=511, IRQ=16"16:55
Kanohi, did somebody test a kernel with # CONFIG_IDE is not set17:03
Kanothats the one and only way to boot for some systems17:03
Kanokanotix thorhammer rc7 uses this kind of kernel, modified 2.6.24-3 kernel, patched to 2.6.24-rc6-git7, additional dmraid patch and secperm17:05
Kanosamsung x65 can not boot with a "generic" kernel17:06
Kanono way17:06
zulheylo17:11
=== allee_ is now known as allee
lamontBenC: did the care package arrive ok?18:01
=== \sh_away is now known as \sh
BenClamont: yes, thanks...I'm hoping to get time today to mess around with it18:42
lamontBenC: ok.  the machines both think that they're $mumble.lamont, and fetch an addr from dhcp18:59
lamontand if you want, I have mac addrs19:00
BenCnp, I have open dhcp on my lan19:00
rtgBenC: I see that ia64 has section conflict errors similar to powerpc64-smp19:09
BenCrtg: yeah, starting to sound like a 64-bit issue, but sparc64/hppa64 seem unaffected19:10
BenCrtg: possibly forcing gcc-4.1 would be a good idea on ia64 (which I can test once I get this new ia64 up and running)19:10
rtgBenC: sounds good. 19:11
BenCrtg: somehow the linux-headers-2.6.24-x-powerpc64 package needs to depend on gcc-4.119:11
BenCrtg: so we don't have to force that build-dep in lum/lrm/lbm19:12
BenCwhich is what I did temporarily19:12
rtgBenC: right.19:14
rtgBenC: should hppa64 have the same Build-Depends as powerpc64 in linux-headers? For example, gcc-4.1-hppa64 [hppa] and binutils-hppa64 [hppa]19:27
BenCrtg: yeah, it should19:27
BenCthen we can remove those hardcoded build-deps from lum/lrm/lbm too19:27
rtgBenC: well, we'll have to futz with the kernel arch Makefile for hppa in order to ensure it uses the correct compiler.19:28
BenCrtg: no, it is already setup that way19:29
bluefoxicyAre Ubuntu kernel modules digitally signed such that they will load if Ubuntu boots with "enforcemodulesig"20:59
tjaaltonBenC: I've got a new lrm readyish with updated fglrx & nvidia-glx-new (stable releases). I've been running the new nvidia for over a week now without too much trouble22:32
tjaaltonbut there's a bug report about nvidia not being uninstallable that I need to check before I'd upload it22:34
tjaaltonif you don't mind me doing that, I mean22:35
tjaaltonfglrx 7.12 seems to have some issues with widescreen resolutions though, but on the other hand the current version leaks memory22:46
=== Lure_ is now known as Lure
Kanotjaalton: well best wait for 8-123:16
tjaaltonKano: no idea what the marketing name is, but apparently 8.45 is the next "traditional" version23:20
Kanoi know23:20
Kanoi still would prefer 8.39.4, because gl2benchmark works with it23:21
Kanoand googleeath too23:21
tjaaltonwe have 8.43 already23:21
Kanobut thats a crappy version23:21
tjaaltonthey all are :)23:22
Kano8.40 had that googleearth bug23:22
Kanoall newer can not render correctly23:22
tjaaltonbesides, whatever the version, it needs to work with 2.6.24rc kernel and xserver-1.423:22
Kanoit does23:22
Kanomy script can show it to you ;)23:23
tjaaltonno point to go back23:23
Kanohttp://kanotix.com/files/install-fglrx-debian.sh23:23
Kanoinstall-fglrx-debian.sh -v 8.39.423:23
Kanoit can install ANY version23:23
Kanofor 2.6.24 down to 8.39.423:23
BenCtjaalton: excellent, thanks23:24
tjaaltonBenC: great, I'll upload it tomorrow, but without the new ati as it seems to have issues with latest 2.6.24rc's and some widescreen resolutions..23:24
Kanoand with 7-11 and below with firegl cards ;)23:25
Kanountil you patch it and replace the control file23:25
Kanobtw. with my new thorhammer iso you can do funny things like: fglrx=v:8.39.4 to dl and install that specific driver on the fly23:32
Kanoalso it has CONFIG_IDE unset..23:33
Kanoit seems you have fear to use this23:33
Kanobut it is the best fix to get around some issues like samsung x65 and a nforce 430 board that has only problems with 2.6.24 generic (but not 2.6.22)23:34
Kanohttp://kanotix.com/files/thorhammer/kernel-2.6.24/linux/linux_2.6.24-3.5+c0.kanotix.1.tar.gz23:36
Kanopatched up to 2.6.24-rc6-git7 - extra patches the custom kernel way23:36
Kanoall ubuntu standard kernels disabled23:37
Kanoalso i set the default to concurrency level 423:39
Kanofor faster compile with standard pbuilder23:39
Kano1 is a bad fallback23:40

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!