[00:47] soren: Thanks. === Qtpaxa2 is now known as Qtpaxa [05:55] Anyone know the minimum version of gcc required to compile kernel 2.6.28? [06:00] Documentation/Changes lists 3.2 [06:00] thanks === TheMuso_ is now known as TheMuso [13:10] hi rtg , there are several issues with your 2.6.30 kernel [13:11] first [13:12] drivers/staging/Kconfig b/drivers/staging/Kconfig [13:12] why on earth did you disable 3 drivers? [13:12] funnyly you even patched one of those disabled drivers [13:13] rt2870 got an added id [13:13] but it never was compiled [13:13] that added id conflicts merging with [13:14] http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=a9ed12e5c149b05adf13bea5f1e96d68c2127faf [13:14] thats after rc5,but add it, then you could add the added it you have in your tree [13:15] i hate to resolve it everytime [13:15] next [13:15] i found the issue why the kernel did not boot on my systems with saa7136 tv card [13:16] the error was s626 module [13:17] can be solved when you do disable CONFIG_COMEDI, which is useless for me anyway [13:18] last thing is [13:18] if you want to be able to use fglrx with all tricks then you need [13:18] http://kanotix.com/files/excalibur/linux-2.6.30-generic/source/2.6.30-export-flush_tlb_page.patch [13:18] and if you do not want to use a very huge patch then use [13:18] http://kanotix.com/files/excalibur/linux-2.6.30-generic/source/2.6.29-ubuntu-copy-headers-for-fglrx.patch [13:19] Host/Kernel/OS "Kanotix" running Linux 2.6.30-4-generic i686 [ Kanotix Excalibur 20090506-17:27 ] [13:19] CPU Info (1) Intel Core2 Duo E8400 @ clocked at [ 2999.000 MHz ] [13:19] (2) Intel Core2 Duo E8400 @ clocked at [ 1999.000 MHz ] [13:19] Videocard ATI Mobility Radeon HD 3450 X.Org 1.4.2 [ 1680x1050 ] [13:19] Processes 140 | Uptime 1:03 | Memory 438.3/1977.9MB | HDD Size 1000GB (6%used) | GLX Renderer ATI Radeon HD 3450 | GLX Version 2.1.8591 | Client Konversation 1.1 | Infobash v2.67.1 [13:19] with those patches it is possible to run fglrx [13:19] the export is absolutely required [13:20] Kano: Why exactly is it that you don't report this on Launchpad? IRC is a horrible, horrible bug tracker. [13:21] Horrible. [13:21] soren: there is no 2.6.30 in launchpad [13:21] as it is only git [13:21] or where are the packages [13:21] a) yes, there is. [13:21] b) that does not make IRC a better bug tracker. [13:21] Kano, you are reporting issues with karmics kernel... as you say 'your 2.6.30' [13:22] soren: not my problem, i fixed the issues, fix em or leave it [13:22] therefore there is a package to report it against [13:22] apw: i use git [13:22] as i require changes [13:22] which are not that important for u [13:22] i use keyboards that doesn't mean that there isn't a bug tracker [13:22] fix it or not [13:23] fun [13:23] Aw. [13:25] there really are packages of the karmic kernel in the karmic release pocket [13:27] * TheMuso reads the above exchange, and finds himself thinking that if were on the kernel team, he'd ignore kano's "bug reports" out of spite. :)( [13:38] heh, no we just tend to lose things that are only in irc [13:40] apw: I know that. kano refuses to use a bug tracker period. I've seen some things fixed that he's brought up in here, but I am sure there are many that aren' [13:40] apw: I know that. kano refuses to use a bug tracker period. I've seen some things fixed that he's brought up in here, but I am sure there are many that aren't. [14:54] smb, can you remind me what was triggered when libusual was enabled [14:55] apw, With libusual enabled this would try to handle registration of usb storage devices, which would try to load usb-storage rather early as it was built in [14:56] ahh yes, usb storange, thanks [14:56] no worries :) [15:02] cking, your machine which is showing the bad sound ... what kernel is it running? [15:05] 2.6.28-12.43 [15:10] apw: I don't think it's a kernel issue. The error I get is a pulseaudio big. "The CPU load limiter only becomes active as result of PA entering an [15:10] endless loop of some kind. It's only symptom of a bug somewhere, but [15:10] gives no hint where it actually is." [15:10] i think that means you have the appropriate 'fixes' for the pointers stuff as i understand it, which might be triggering confused pulse [15:11] I suppose I can roll back to an earlier kernel and test [15:12] I've just spotted: hda-intel: IRQ timing workaround is activated for card #0. Suggest a bigger bdl_pos_adj. === awe is now known as awe-afk === awe-afk is now known as awe === awe is now known as awe-afk === awe-afk is now known as awe === smb__ is now known as smb [16:53] rtg this FIRMWARE_IN_KERNEL config change, do we know how much other firmware this may bring into the kernel other than the keyspan stuff? [16:54] apw: I think you have to selectively pick the drivers that get firmware. I think the option just enables those selections [16:57] there appears to be a few other things which use fw-shipped-* in the Makefiles, which would be changed by it, as as soon as this option is turned on we longer put any firmware for any device which is =y into the firmware module from the kernel [16:57] s/module/result blob/ [16:57] so i assume all of that firmware must go into the kernel image as well [16:57] it doesn't look like _that_ much, just making sure its what you expect [16:58] mod-fw := $(fw-shipped-m) [16:58] # If CONFIG_FIRMWARE_IN_KERNEL isn't set, then install the [16:58] # firmware for in-kernel drivers too. [16:58] ifndef CONFIG_FIRMWARE_IN_KERNEL [16:58] mod-fw += $(fw-shipped-y) [16:58] endif [16:58] installed-mod-fw := $(addprefix $(INSTALL_FW_PATH)/,$(mod-fw)) [16:59] apw: shit. I would not have expected that enabling the option would adversely affect other drivers. [16:59] apw: so, should we put it in the linux-firmware package instead? [17:01] rtg, risk wise perhaps so for sru ... a difficult call. what happens to the firmware blob that the kernel builds during build of the kernel [17:01] i wonder if that should be making its way wholesale into the firmware package [17:01] rtg: when do you merge aufs2? [17:01] rtg: http://aufs.sourceforge.net/README.aufs2 [17:02] apw: dunno, lemme look at it again. [17:02] in the test live image is even squashfs 3.1 used, that will never work [17:02] you need a cvs snapshot [17:02] user space of course [17:03] as squashfs 4 is in the kernel since .29 [17:04] hmm you are lucky,there is even a final 4.0 [17:04] in sid [17:05] when i build my .29 test image i had to use cvs [17:06] well merge aufs2 (it is already .30 compatible) + update squashfs-tools in karmic, then you can hope ;) [17:08] rtg: do you want to update jaunty to 2.6.28.10 today? [17:08] jaunty is released now so anything like that would be an sru [17:08] that said i thought .28 was already in git [17:08] apw: i mean the jaunty branch [17:08] i do not use precompiled kernels [17:09] the jaunty git tree has .10 in it [17:09] commit 8bdd945b77faa2d7c63ad669d2d8b1a623c57e67 [17:09] Author: Greg Kroah-Hartman [17:09] Date: Sat May 2 11:54:43 2009 -0700 [17:09] Linux 2.6.28.10 [17:09] ok [17:09] will compile it now [17:11] or do you want to add some more fixes soon? [17:12] there are always commits in the pipeline [17:13] as for aufs, that pending a decision on how live cd's will be done for karmic [17:14] you can have it the kernel even if you dont use it ;) [17:14] as you can use it without a live system too [17:14] no point in carrying delta for something we don't use [17:15] so how do you want to create live images? anything new in casper yet [17:15] no decision that i am aware of no [17:15] well i create live images, you talk about em ;) [17:19] apw: did you rebase jaunty today? or since when is there .10? [17:19] jaunty doesn't rebase its released. the commits for .10 went in about the 2nd [17:20] ok, then i have it already [17:22] so all i have to try now is aufs2... bbl [20:06] Hi [20:07] I am trying to provide linux-vserver kernel packages based on the ubuntu kernel in my ppa at: [20:07] https://launchpad.net/~christoph-lukas/+archive/ppa [20:08] the vserver kernel image for jaunty is working fine, the vserver functionality is also working fine [20:08] but I get an kernel oops when I try to load the nvidia module 180.44 build via dkms for this kernel. [20:08] could anybody help me debugging this issue? [20:09] the oops is here: http://paste.ubuntu.com/169897/ === awe is now known as awe-afk === erle64- is now known as erle- [22:17] my linux-headers package seems to be broken which seems to break the nvidia module. [22:17] sorry for the noise.