| === asac_ is now known as asac | ||
| AnAnt_ | Hello, I noticed that there are new linux-image-2.6.24-20* & linux-restricted-modules-common packages for Hardy, yet linux-image-generic is still at version 2.6.24.19.21, is that correct ? | 09:06 |
|---|---|---|
| ivoks | zul: hi there | 12:49 |
| zul | hey ivoks | 12:49 |
| AnAnt_ | Hello, I noticed that there are new linux-image-2.6.24-20* & linux-restricted-modules-common packages for Hardy, yet linux-image-generic is still at version 2.6.24.19.21, is that correct ? | 13:23 |
| ivoks | it is | 13:27 |
| AnAnt_ | so I can upgrade the linux-restricted-modules-common without expecting any problems ? | 13:28 |
| ivoks | wait for the meta package | 13:29 |
| AnAnt_ | the linux-restricted-modules-common new version is 2.6.24-20 | 13:29 |
| AnAnt_ | ok, thanks | 13:29 |
| Kano | hi BenC , do you want to keep 686 setting? k6-2/3 are only 586 and usually fast enough too | 15:14 |
| Kano | like the hardy kernel, intrepid is 686 | 15:14 |
| BenC | Kano: 586 will probably work still | 15:15 |
| Kano | BenC: current kernel definitely does not boot on k6-2 | 15:19 |
| Kano | due to: CONFIG_M686=y | 15:19 |
| Kano | before it was CONFIG_M586=y | 15:22 |
| mkrufky | is cogito available in hardy? i dont see it in apt :-/ | 15:35 |
| bboschman | Hi | 15:52 |
| bboschman | I'm compiling my own kernel using make-kpkg - just extracted linux-source.tar.bz2; copied /boot/config-*server .config and run make-kpkg kernel-image | 15:53 |
| bboschman | this results in a 180MB linux-image*.deb | 15:53 |
| bboschman | why is this so large? | 15:53 |
| abogani | bboschman: Please don't you use the Debian way to build Ubuntu kernel: it is completely unsupported. The Ubuntu way ("fakeroot debian/rules binary-server") should be works great. | 15:58 |
| bboschman | abogani, but I'm not sure how to make config changes the ubuntu way | 15:59 |
| _ruben | anyone know why im seeing only 4gb of ram on a system which has 6gb? .. this is what i get in dmesg: | 16:06 |
| _ruben | [ 40.008864] Memory: 3993236k/6291456k available (2295k kernel code, 135080k reserved, 1238k data, 308k init) | 16:06 |
| _ruben | its a dual 1st gen opteron running 64bits gutsy gibbon | 16:06 |
| Kano | maybe wrong bios settings? | 16:10 |
| _ruben | post shows 6gigs | 16:11 |
| _ruben | and it does detect it, but cant use it somehow | 16:11 |
| stgraber | _ruben: what kernel are you using ? | 16:12 |
| _ruben | Linux vn-vm03 2.6.22-15-server #1 SMP Fri Jul 11 19:20:50 UTC 2008 x86_64 GNU/Linux | 16:13 |
| bboschman | can I just overwrite debian/config/i386/config.server with a Full .config file? | 16:13 |
| maks_ | mkrufky: cogito is gone, use plain git | 16:14 |
| maks_ | pblumbing disappears and porcelaine of git itslef gets good! | 16:14 |
| smb_tp_ | bboschman, configs are combined config and config.server (in your case) you might want to empty one and place everything else into the other | 16:14 |
| Blinny | bboschman: FYI, "the Ubuntu way" is described here: https://help.ubuntu.com/community/Kernel/Compile | 16:15 |
| _ruben | most 4gb related stuff i see on google is 32bits related .. not much 64bits stuff there | 16:16 |
| bboschman | Blinny, I know that it is described there, but not how the config files are defined (or even a own branch is set) | 16:16 |
| bboschman | I'd like to build something like linux-image-*-preempt | 16:17 |
| mkrufky | thanks, maks_ | 16:18 |
| _ruben | i'll try with hardy tomorrow, perhaps the hw is just too old (HP DL145 G1) to handle 4+ GB properly .. gotta go now | 16:19 |
| bboschman | does config.server override options from config? | 16:20 |
| smb_tp_ | bboschman, AFAIK it is cat'ed together so normally it would by name sorting... | 16:23 |
| bboschman | let's see ... :) | 16:51 |
| rtg | zul: did you see the note on the kernel team mailing list regarding Xen: "Lates XEN Kernel do not come up" | 17:48 |
| zul | rtg: yeah its something to do with his raid | 17:48 |
| zul | we didnt make any changes to the xen bits | 17:49 |
| rtg | zul: perhaps you could respond to him. | 17:49 |
| zul | rtg: sure | 17:49 |
| _ruben | lets see if upgrading to hardy fixes the 4g/6g prob | 18:56 |
| _ruben | will see in the morning afterall .. box doesnt come up after reboot ;) | 19:24 |
| BenC | _ruben: what 4g/6g issue do you mean? | 19:43 |
| RAdams | I need to compile 2.6.24-19 AND 2.6.26-4.10 with acpi debugging AND tracing to track a known issue on Dell Latitude d800 and so laptops. This issue occurs in both kernels, I am certain. I am familiar with how to do this with a vanilla kernel, but I want to confirm what the most useful/correct way to do this for this Ubuntu-specific build. I would love to help finally nail down this problem and thus contribute to the improvement of Ubuntu, | 19:55 |
| smb_tp_ | RAdams, have you already looked at https://help.ubuntu.com/community/Kernel/Compile? | 19:57 |
| BenC | RAdams: if it occurs on vanilla kernel too, best bet is to start there | 20:10 |
| RAdams | smb_tp_: yes. Would using the steps under "Alternate Build Method: The Old-Fashioned Debian Way" yield a good enough kernel/debugging to help trace this issue? The only way I know how to enable acpi debugging and tracing is with make menuconfig, which I apparently don't get the opportunity to do using the first method | 20:11 |
| RAdams | benc: that would be more helpful? I can do that. But won't I have to obtain/rebuild drivers for hardware? what other changes can I expect I'll need to make to get a workable system for testing? It doesn't have to have every feature, just enough to replicate the problem | 20:12 |
| RAdams | benc: I am not certain that the problem exists in vanilla, so the first thing I would have to do is verify that it does :o | 20:12 |
| BenC | RAdams: cp /boot/config-`uname -r` .config | 20:13 |
| RAdams | Also, passing acpi=off stops the problem from happening, so it's something with /proc/acpi as it's implemented into the ubuntu kernels | 20:13 |
| BenC | RAdams: in a vanilla kernel, and start from there | 20:13 |
| RAdams | benc: will do | 20:13 |
| smb_tp_ | RAdams, Otherwise, if you want to use the git source, you can do. For test purposes you could put you could put your config file to debian /config/<yourarch>/config and empty the config.generic there. | 20:16 |
| RAdams | which vanilla kernel should I use? http://kernel.org/pub/linux/kernel/v2.6/patch-2.6.26.bz2? any need to build http://kernel.org/pub/linux/kernel/v2.6/patch-2.6.24.bz2 as well, or is the current sufficient, so the problem gets fixed upstream (if that's where the problem is) | 20:17 |
| RAdams | I have 0 git experience, btw, so if it's best to go that route, I'll have to read up on git I suppose o.O | 20:18 |
| smb_tp_ | RAdams, You would need t know only one command "git-clone http://kernel.ubuntu.com/ubuntu/ubuntu-hardy.git <yourdirname>" | 20:21 |
| RAdams | smb_tp_: that will get me the current devel version of the kernel? | 20:21 |
| smb_tp_ | RAdams, Just don't ask for "mrproper" He is too clean | 20:21 |
| smb_tp_ | RAdams, Replace hardy with intrepid and you get the current intrepid tree. If you mean Linus tree with devel version, that would be a different url | 20:23 |
| RAdams | oic | 20:23 |
| RAdams | so the hardy one will get me the current hardy kernel in use? | 20:23 |
| RAdams | .24-19? | 20:23 |
| smb_tp_ | RAdams, Current Hardy: yes. 24-19 no. 24-20 is somewhere rolling | 20:24 |
| RAdams | oh | 20:24 |
| RAdams | ok ty | 20:24 |
| === smb_tp_ is now known as smb_tp | ||
| === chuck_ is now known as zul | ||
| === macd_ is now known as macd | ||
| netzwurm | hi | 22:24 |
| netzwurm | I am trying to patch the server kernel. I am confused, however. If I get linux-source-2.6.24 and copy the config from /boot/config-2.6.24-19-server to the source, the source doesn't build because arch/xen/Makefile doesn't exist. that same file is also neither in the linux-headers packages for 2.6.24-19 or 2.6.24-19-server. Is that a bug? | 22:26 |
| netzwurm | Obviously the server-kernel is patched with xen, but how do I build that kernel? where does the patch come from? | 22:27 |
| smb | netzwurm, You should not need that. I am not sure what you did. I just tried a "fakeroot debian/rules binary-server" and a similar out-of-tree build. both seem to run happily | 23:02 |
| netzwurm | smb: oh. that's in the debian source package? What I am asking about is whether there is a way to build the server-kernel from within the linux-source-2.6.24/ dir. | 23:04 |
| netzwurm | smb: you took the apt-get source linux-image-2.6.24-19-server i take it? | 23:05 |
| smb | netzwurm, yes | 23:05 |
| netzwurm | smb: hm. how do i add my custom patches to that then? | 23:05 |
| netzwurm | oh. i just patch the source. | 23:05 |
| netzwurm | whatever. | 23:05 |
| netzwurm | thanks. | 23:05 |
| smb | netzwurm, I would guess so. :) np | 23:06 |
| netzwurm | smb: although that's still complicated with modifying the config. | 23:07 |
| netzwurm | i mean, i am trying to find the patchset for the server-kernel now. | 23:07 |
| smb | netzwurm, Practically you could replace the debian/config/<arch>/config{,.server} file(s) | 23:08 |
| netzwurm | smb: right. | 23:09 |
| smb | smb, Will you need to distribute a patchset later or would just getting a kernel package compiled be enough | 23:10 |
| smb | netzwurm, Anyways, I am not sure this gets repeated too often but https://help.ubuntu.com/community/Kernel/Compile should be of help... | 23:14 |
| netzwurm | smb: what i'd really like would be a configured kernel-source with all patches. I don't really want to build the kernel, but i need the build-source. | 23:25 |
| netzwurm | smb: because the headers aren't enough. | 23:25 |
| smb | netzwurm, the linux-headers-server package is not enough? maybe the missing headers from lum... but sorry I have to go. bbiab | 23:30 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!