|
=== asac_ is now known as asac |
|
[09:06] <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 ? |
|
[12:49] <ivoks> zul: hi there |
|
[12:49] <zul> hey ivoks |
|
[13:23] <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 ? |
|
[13:27] <ivoks> it is |
|
[13:28] <AnAnt_> so I can upgrade the linux-restricted-modules-common without expecting any problems ? |
|
[13:29] <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 |
|
[15:14] <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 |
|
[15:15] <BenC> Kano: 586 will probably work still |
|
[15:19] <Kano> BenC: current kernel definitely does not boot on k6-2 |
|
[15:19] <Kano> due to: CONFIG_M686=y |
|
[15:22] <Kano> before it was CONFIG_M586=y |
|
[15:35] <mkrufky> is cogito available in hardy? i dont see it in apt :-/ |
|
[15:52] <bboschman> Hi |
|
[15:53] <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? |
|
[15:58] <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. |
|
[15:59] <bboschman> abogani, but I'm not sure how to make config changes the ubuntu way |
|
[16:06] <_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 |
|
[16:10] <Kano> maybe wrong bios settings? |
|
[16:11] <_ruben> post shows 6gigs |
|
[16:11] <_ruben> and it does detect it, but cant use it somehow |
|
[16:12] <stgraber> _ruben: what kernel are you using ? |
|
[16:13] <_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? |
|
[16:14] <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 |
|
[16:15] <Blinny> bboschman: FYI, "the Ubuntu way" is described here: https://help.ubuntu.com/community/Kernel/Compile |
|
[16:16] <_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) |
|
[16:17] <bboschman> I'd like to build something like linux-image-*-preempt |
|
[16:18] <mkrufky> thanks, maks_ |
|
[16:19] <_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 |
|
[16:20] <bboschman> does config.server override options from config? |
|
[16:23] <smb_tp_> bboschman, AFAIK it is cat'ed together so normally it would by name sorting... |
|
[16:51] <bboschman> let's see ... :) |
|
[17:48] <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 |
|
[17:49] <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 |
|
[18:56] <_ruben> lets see if upgrading to hardy fixes the 4g/6g prob |
|
[19:24] <_ruben> will see in the morning afterall .. box doesnt come up after reboot ;) |
|
[19:43] <BenC> _ruben: what 4g/6g issue do you mean? |
|
[19:55] <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, |
|
[19:57] <smb_tp_> RAdams, have you already looked at https://help.ubuntu.com/community/Kernel/Compile? |
|
[20:10] <BenC> RAdams: if it occurs on vanilla kernel too, best bet is to start there |
|
[20:11] <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 |
|
[20:12] <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 |
|
[20:13] <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 |
|
[20:16] <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. |
|
[20:17] <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) |
|
[20:18] <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 |
|
[20:21] <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 |
|
[20:23] <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? |
|
[20:24] <smb_tp_> RAdams, Current Hardy: yes. 24-19 no. 24-20 is somewhere rolling |
|
[20:24] <RAdams> oh |
|
[20:24] <RAdams> ok ty |
|
=== smb_tp_ is now known as smb_tp |
|
=== chuck_ is now known as zul |
|
=== macd_ is now known as macd |
|
[22:24] <netzwurm> hi |
|
[22:26] <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? |
|
[22:27] <netzwurm> Obviously the server-kernel is patched with xen, but how do I build that kernel? where does the patch come from? |
|
[23:02] <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 |
|
[23:04] <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. |
|
[23:05] <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. |
|
[23:06] <smb> netzwurm, I would guess so. :) np |
|
[23:07] <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. |
|
[23:08] <smb> netzwurm, Practically you could replace the debian/config/<arch>/config{,.server} file(s) |
|
[23:09] <netzwurm> smb: right. |
|
[23:10] <smb> smb, Will you need to distribute a patchset later or would just getting a kernel package compiled be enough |
|
[23:14] <smb> netzwurm, Anyways, I am not sure this gets repeated too often but https://help.ubuntu.com/community/Kernel/Compile should be of help... |
|
[23:25] <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. |
|
[23:30] <smb> netzwurm, the linux-headers-server package is not enough? maybe the missing headers from lum... but sorry I have to go. bbiab |
|
|