=== cheddarcheese is now known as cheddacheese
[00:22] <cheddacheese> when will ext4 be in the kernel?
[00:26] <johanbr> cheddacheese: I don't know, but ext4 is still pretty experimental. I wouldn't use it on anything important.
[00:26] <cheddacheese> well its in the vanilla kernel.
[00:26] <cheddacheese> im using it right now, but its not in the ubuntu kernel
[00:30] <johanbr> If it's in the vanilla kernel, I'm sure it'll be synced in for Hardy+1.
[00:50] <mjg59> cheddacheese: There's no guarantee that the on-disk format is stable yet
[00:50] <mjg59> Hence ext4dev rather than ext4
[00:51] <cheddacheese> cant it be included as an expermental option? the user has to add test_fs to there existing ext3 partitions in order to mount as ext4dev.
[00:52] <mjg59> No. There's no reason to be running ext4 now unless you're developing it, and if you are then you'll need to be building new kernels anyway.
[00:54] <cheddacheese> o but ubuntu includes xfs which imo is not even stsble as ext4dev
[00:55] <mjg59> xfs has a stable on-disk format
[00:56] <cheddacheese> d=the=then y does it corrupt easy then
[00:57] <crimsun> where are you getting metrics for "it corrupt[s] easy"?
[00:57] <cheddacheese> every hard poweroff corrupts it for me.
[00:58] <mjg59> cheddacheese: That's not what stable on-disk format means
[00:58] <cheddacheese> duh
[01:00] <mjg59> When the ext4 format has been stabalised, then it's likely it'll be enabled
[01:01] <mjg59> Until then, it's not useful for it to be
[01:01] <cheddacheese> is there a way to enable it without compiling a kernel?
[01:04] <mjg59> No. But, like I said, it's no use as a filesystem right now unless you're willing to back up and restore on every kernel upgrade.
[01:04] <mjg59> So you should only be using it if you're developing the code
[01:05] <cheddacheese> im using ext4dev right now as a testfs for some stuff and it has survived more crashes than xfs.
[01:07] <cheddacheese> so yeah maybe xfs should be removed until it gets more stablised if it ever does
[01:10] <mjg59> The xfs on-disk format is perfectly stable
[01:10] <cheddacheese> maybe for you but not for me
[01:11] <mjg59> No. Really. You don't know what you're talking about.
[01:11] <mjg59> The way in which data is arranged on the disk in xfs does not change with new kernel versions. As a result, new kernels can read partitions formatted with older kernels.
[01:12] <mjg59> That is not currently guaranteed with ext4
[01:15] <cheddacheese> i do know wat i am talking about
[01:15] <cheddacheese> when they devoled ext3 they did everything possible not to change the on disk layout to make it backwards compatiable
[01:17] <mjg59> See http://lkml.org/lkml/2006/6/28/454
[01:17] <mjg59> Note the last paragraph
[01:21] <cheddacheese> i see that
[01:23] <mjg59> So, until the on-disk format has been stabalised and will no longer change, it can be enabled in Ubuntu
[01:28] <cheddacheese> that wont happen for a long time as most of the stuff that will change the layout r in da mm kernel.
[01:34] <mjg59> That's not a problem
[01:34] <cheddacheese> it is for ext3 users
[01:35] <mjg59> Having to backup and restore their filesystems after a kernel upgrade would be more of a problem
[01:38] <cheddacheese> im sure you dont know the limitations of ext3
[01:39] <mjg59> Oddly enough, I do
[01:40] <cheddacheese> wat r they den
[01:42] <mjg59> ...
[01:42] <cheddacheese> nm
[01:44] <mjg59> (Even if I didn't know, it's hardly difficult to find the ext4 entry on Wikipedia)
[01:56] <cheddacheese> I know more about filesystems than any ubuntu kernel dev
[04:19] <ripps> I just installed the kernel 2.6.25 using kernelcheck in Ubuntu 8.04. How do I propertly get sound working in the custom kernel? Not having much help in #ubuntu and #alsa.
[04:23] <johanbr> ripps: That's hard to summarize in a few lines. You just have to make sure you have all the config options you need and then it should just work.
[04:25] <ripps> Do I install Alsa in Kernel, or reinstall the Alsa package?
[04:26] <johanbr> In the kernel probably, if you're building a kernel.org kernel.
[04:29] <ripps> Another question: Why are custom built kernels so much larger than distrubution kernels? The intrd image was about 45mb, while all the other intrd's were around 8mb
[04:30] <crimsun> they aren't stripped.
[04:31] <ripps> crimsum: what's that mean?
[04:31] <crimsun> and I would build alsa-driver separately unless you enabled all the config options, to which I presume johanbr referred, in your .config.
[04:32] <crimsun> ripps: you probably want to see strip(1)
[04:32] <ripps> I think the problems I was experiencing was because I accidently built both alsa and oss into the kernel. I'm rebuilding it without oss.
[04:35] <crimsun> be aware that 2.6.25 ships with 1.0.16rc2, which is older than what hardy ships in lum.
[04:35] <crimsun> (not to mention the various additions and quirks added in lum)
[04:43] <johanbr> ripps: Why are you compiling your own kernel anyway>
[04:43] <johanbr> ?
[04:45] <ripps> johanbr: Just want to play around. Why not? Linux is mostly just my mess around OS anyway.
[04:46] <johanbr> okay :)
[04:47] <ripps> My system was working TOO well... I didn't have anything to play around with. So I decided to install a new kernel. That would give me something to do.
[04:47] <ripps> That's why I'm going to install Intrepid Ibex when it comes to beta
[04:48] <johanbr> I thought it was the opposite - that something wasn't working.
[07:10] <dacresni> any talk here of porting ksplice to ppc?
[11:22] <holinx> hi
[11:23] <holinx> Linux 2.6.24-16-generic/i686 is alright for now?
[11:23] <holinx> anything i should keep in mind?
[11:23] <holinx> or any suggestions
[15:04] <JanC> seems like some CPUs don't have CPU frequency scaling enabled: https://bugs.launchpad.net/ubuntu/+source/linux-meta/+bug/177646 ?
[15:04] <ubotu> Launchpad bug 177646 in linux-meta "Celeron M530, no frequence scaling" [Undecided,New] 
[15:05] <wesley> hi
[15:20] <wesley> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/177646 will you fix this ?
=== mkrufky is now known as mkrufky-away
=== mkrufky-away is now known as mkrufky
[20:06] <tormod> Hi, in dmesg I have Clocksource tsc unstable (delta = -58695238058 ns). Is this bug #221351? There's half(?) minute delay while discovering the boot disk, can this be related?
[20:06] <ubotu> Launchpad bug 221351 in linux "TSC Clocksource can cause hangs and time jumps" [Medium,Fix committed] https://launchpad.net/bugs/221351
[20:24] <mjg59> tormod: No
[20:26] <tormod> mjg59: thanks. no on both questions?
[20:28] <mjg59> tormod: No to the latter - it's possible that the message is related to 221351, but it's not causing your boot delay
[20:29] <tormod> is the boot delay a known issue? couldn't find a bug report so far.
[20:30] <tormod> seems like it's stuck between discovering the disk and discovering the parititions - unless the console messages are buffered - dmesg timestamps don't show much delay.
[20:30] <mjg59> I've seen it, but don't know if a bug is filed
[20:37] <mjg59> I suspect it's hanging during IDE probing
[20:37] <mjg59> Some machines respond very slowly when that's going on
[20:40] <tormod> I reported it anyway in bug #223235
[20:40] <ubotu> Launchpad bug 223235 in linux "discovery of partitions delays booting by 1/2 minute" [Undecided,New] https://launchpad.net/bugs/223235
[23:46] <osmosis> 3ware raid controller seems slow on hardy 2.6.24-16-server