[12:12] <Mithrandir> BenC: ps3-kboot> uh, another copy of the linux kernel in the archive?
[12:13] <varkatope> can i ask for a specific kernel driver module availability in feisty here?
[12:16] <cjwatson> Mithrandir: it's necessary unfortunately - kboot doesn't work with 2.6.20 yet
[12:16] <cjwatson> Ben tried fairly hard to make that work
[12:16] <Mithrandir> cjwatson: hm, but it doesn't seem to actually build the kernel at any point?
[12:17] <Mithrandir> sorry, it does.
[12:17] <Mithrandir> I guess he'll want this in main too?
[12:19] <cjwatson> yeah, it's needed for PS3s
[12:19] <cjwatson> I think Ben's still working on making it work with 2.6.20
[12:19] <Mithrandir> ok.
[12:20] <Mithrandir> I'm not happy with it, but I see the point of it not working with 2.6.20 yet.  I wonder if having a policy that such packages should build-depend on the current linux-source and then ship the patches to go down to the version they want, then building with that would make sense or just be utterly insane.
[12:21] <cjwatson> I think the patch might well be about as big as the kernel, so I don't see the point in this kind of case
[12:22] <cjwatson> the only result of that would be that it would be a lot less comprehensible
[12:22] <Mithrandir> extra incentive to get stuff working for the latest kernel because of the pain involved. :-)
[12:22] <cjwatson> I understand the incentive, but I think the method is misplaced
[12:25] <varkatope> where to ask/look if a specific chipset is supported by the kernel of feisty fawn?
[12:27] <cjwatson> varkatope: look> apt-get source linux-source-2.6.20, debian/config/wherever; ask> bug report on linux-source-2.6.20
[12:28] <varkatope> cjwatson: thx
[01:34] <gilligan_> hi
[01:48] <doko> bug 89851
 doko: the heartbeat problem might be a missing kernel header..
 ia64: checking for linux/icmpv6.h... no
 i386: checking for linux/icmpv6.h... yes
 doko: ^^ this one is the problem for sure.. you can see in the i386 build log that it then builds IPv6addr
[01:49] <fabbione> doko: i said probably.. you need to check the header packages
[01:49] <doko> on the list, but low prio
[01:54] <gilligan_> http://gilligan.neuecouch.de/nl_socket.txt <-- could someone do me a favor and check if he can see what the problem with this simple netlink socket test is ?
[01:55] <gilligan_> I don't get it
[02:22] <BenC> Mithrandir: has to be a certain version. The 2.6.20 can't kexec kernels
[02:49] <zul> besides sometimes its not trivial to go one from one version to the next.
[03:05] <zul> there is a meeting today isnt there?
[03:07] <gilligan_> hm.. still looking for some help on a netlink problem.. anyone?
[03:22] <zul> hah...http://lists.xensource.com/archives/html/xen-devel/2007-03/msg00166.html
[03:22] <zul> 13 year olds..
[03:33] <Keybuk> zul: why does Xen need such an old kernel version?
[03:33] <mjg59> Because Xen isn't in the main kernel tree?
[03:34] <zul> Keybuk: Xensource doesnt even track the latest kernels and the one we have is from redhat
[03:34] <Keybuk> right, I just mean has there been some known difficulty updating it to 2.6.20 or is it just that nobody's tried yet
[03:35] <zul> Keybuk: there is some dificulty because of the workqueue changes and the paravirt-ops changes but opensuse only ported to 2.6.20 very recently and its not really tested yet
[03:35] <zul> quintela from redhat is still in the process of porting 3.0.3 to 2.6.20
[03:35] <Keybuk> *nods*
[03:36] <zul> and Im in the process of porting xen-unstable from 2.6.18 to 2.6.21
[03:36] <Keybuk> the reason that check is there is because we had some problems getting the feisty udev to work with older kernels than 2.6.19
[03:36] <Keybuk> that could be fixed now though
[03:36] <zul> cool
[03:36] <Keybuk> I'll investigate and see whether we can't drop that back to a minimum 2.6.17 dep
[03:37] <zul> that would be nice for the xen users who still want to use their compiled kernels as well
[03:37] <Keybuk> I'm not sure how broken feisty would be with an older kernel though
[03:38] <Keybuk> things like apport might not work, and may fail badly
[03:38] <zul> true..Im hoping for feisty to base it off the proper kernel
[03:42] <zul> mjg59: I doubt Xensource is looking for inclusion now
[03:42] <kylem> xensource are <censored by the CoC>
[03:43] <zul> heh...a bunch of <censored by the CoC>? :)
[03:44] <kylem> jeremy has been doing a decent job getting xen-paravirt up to snuff.
[03:44] <kylem> i'd be happy to do a backport in my spare time if you need help.
[03:44] <zul> kylem: sure I was toying around with it this weekend
[05:58] <zul> *sigh* https://launchpad.net/bugs/89665
[06:01] <gilligan_> wow..what an idiot
[11:12] <Kazz> Hey, does anyone know if there are problems with LDM (for accessing windows dynamic disks)?  It's part of the kernel, so I hope you don't mind me asking here...it seems to be compiled in, but it doesn't seem to be doing its job.
[11:14] <Kazz> It wasn't working with 2.6.17-10-generic and it's still just showing everything as 1 SFS partition with -11 too when it should be showing the individual partitions inside of the "dynamic disk".