=== jjohansen is now known as jj-afk === smb` is now known as smb [07:06] Morning .* === jj-afk is now known as jjohansen [07:41] * apw groans [07:42] * smb prepares coffee infusions [07:55] hi [07:55] is ubuntu kernel tree git bisectable? [07:56] cause with 3.0.0-10 my router fans are on 4500rpm instead of <2000rpm like it was on 3.0-1 [08:06] hrw, https://wiki.ubuntu.com/Kernel/KernelBisection - or flirt with apw to make test kernels for you ;-) [08:08] hrw: yes and no [08:08] will read [08:08] hrw: it is bisectable from the import point forward, otherwise you need to bisect against upstream with the ubuntu patches thrown on top [08:09] jjohansen: in other words: ubuntu kernel tree is rebased often? [08:10] During development: yes [08:10] hrw: yes its rebased up until the kernel freeze (5 days ago) [08:12] ok, will stay then with 3.1.0-rc6 self built instead probably [08:13] or will bisect upstream clean kernel [08:14] Not, sure, are there probably 3.0.x mainline builds as well to compare against? [08:16] I am fine with bisecting on my own [08:19] Sure, I find them still useful for narrowing the range for bisection quickly [08:23] right - http://kernel.ubuntu.com/~kernel-ppa/mainline/ exists ;D [08:24] will have to connect monitor/keyboard to machine for grub selection === BenC__ is now known as BenC [09:21] hi again [09:21] 3.0.4 mainline == 1917rpm fan. 3.0.4 ubuntu = 4500 rpm [09:22] hrw, an interesting result indeed [09:22] d510mo board (atom d510) [09:23] hrw and is there any cpu load when running ubuntu kernel? [09:23] apw: load <1.0 [09:23] comparible to mainline ? [09:24] apw: same load - routing packets + irssi session [09:24] and the reported load is comparible [09:24] yes [09:24] clearly the machine is doing more, else it wouldn't be so very hot [09:24] it had 4500rpm 23°C [09:24] as we can assume the environmental load is the same [09:24] now it has 2000rpm 46°C which is also fine [09:26] i'd be interested in comparing powertop output on both [09:26] looking at wakeups and also looking at c-state residency and frequency state residency [09:27] 1.13/oneiric will be fine? [09:27] if there is nothing obvious then a bisect through the the ubuntu delta will be the next logical step [09:27] 1.13 has most of what you want yes [09:28] D510 has no power management afaik but the wakeups per second might be interesting [09:29] Wakeups-from-idle per second : 191,1 interval: 15,0s [09:29] Top causes for wakeups: 31,0% (204,1) [ath9k] [09:29] this is router/accesspoint [09:30] its is the comparison from one to the other that is interesting so collect them all on both [09:30] hows that comparing between mainline and the ubuntu kernel [09:30] I will reboot to ubuntu and collect 'powertop -d' output [09:30] brb === chrisccoulson_ is now known as chrisccoulson [09:47] hi [09:47] 3.0.0-11-server and also 1757rpm 49°C [09:48] http://krzys.juszkiewicz.com.pl/~hrw/powertop/ has powertop dumps [09:59] hrw, so which kernel was bad ? [10:03] 3.0.0-8 probably [10:03] if you have it somewhere (amd64 -server) then I can check [10:05] hrw if its working fine and in line with mainline in the -11 kernel then i think we are all happy [10:06] hrw, though 3.0.0-8.11 was based on 3.0.1 mainline not 3.0.4 ? [10:10] https://launchpad.net/ubuntu/oneiric/+source/linux/3.0.0-8.11 [10:10] hrw, you can get old binaries from the launchpad librarian for example as above [10:17] I am fine with just leaving it as it is ;) === yofel_ is now known as yofel === Quintasan_ is now known as Quintasan [13:41] * ogasawara back in 20 [14:22] smb, do any of your machines have real serial ports on them ? [14:22] apw, sure [14:23] smb, that you use as consoles ? [14:23] apw, not directly. I may do. but normally use sol [14:24] smb, sol appears as a serial console to the kernel anyhow doesn't it ? [14:24] apw, correct. on the server its just a serial port [14:25] to linux [14:27] apw, So as long as it is not the real hw part of the cable connection my server is already using the serial console for boot [14:27] ## [14:27] ## Ubuntu Kernel Team Meeting - Today @ 17:00 UTC - #ubuntu-meeting [14:27] ## agenda: https://wiki.ubuntu.com/KernelTeam/Meeting [14:27] ## [15:02] apw, kernel.org is resolving again. maybe some of the git repos will begin to reappear [15:03] tgardner, looks the same to me, all 'down for maint' and git.k.o is non-existant [15:03] apw, uh, yep. was just coming to the same conclusion [15:21] tgardner: Does oneiric's udev require a minimum kernel level? [15:21] BenC, not to my knowledge, though I've been more concerned with backporting recent kernels to Lucid. [15:22] udevd[1162]: unable to receive ctrl connection: Function not implemented [15:22] I get flooded with that if I run oneiric's udev on a 2.6.34 kernel, but backing udev down to natty's version makes it all happy [15:23] BenC, hmm, I haven't the faintest idea. [15:23] Ok, thanks [15:29] tgardner: - Some architectures might need a later kernel, that supports accept4(), or need to backport the accept4() syscall wiring in the kernel. [15:29] tgardner: usage of accept4() is the cause [15:30] just in case you ever hear about it again [15:30] BenC, ok. the only arches that I care about for LTS backports (so far) are x86'en [15:31] I wasn't suggesting you backport anything, just trying to track it down :) [15:31] BenC, understood. [16:01] ## [16:01] ## Kernel team meeting in one hour [16:01] ## [16:50] bjf, You may already know (I wasn't around last week) but the meeting-bot now seems to want the name of the meeting as an argument to startmeeting [16:58] smb, the bot can screw itself [16:59] bjf, Won't mind. Just thought I relate to you how the other meeting succeded in convincing it to do something [16:59] ## [16:59] ## Meeting starting now [16:59] ## [17:02] Hm, seems to work as good without. Only difference is that the topic is screwed more or less. *shrug* [17:07] ppisati: can you rebase git://kernel.ubuntu.com/ppisati/ubuntu-maverick.git ti-omap4? it got two more patches today, doesn't pull clean [17:10] herton: ok, i'll do [17:14] apw, we're using overlayfs for oneiric live CDs now, right ? [17:16] herton: don [17:16] e [17:16] tgardner, yes [17:32] * tgardner --> lunch === med_out is now known as medberry [19:14] since *xorg.conf no longer exists how to i drop back to default ati drivers? removeing fglrx doesnt help [19:15] i was told kernel takes over and sets drivers since that file no longer exists [19:18] gnomefreak, that is normally the plan, check for radeon being blacklisted in /etc/modprobe.d [19:18] though are you sure there isn't an xorg.conf selecting fgrlx when installed, i thought there was [19:19] apw: i cant find it in /etc/X11 [19:20] gnomefreak, worth asking on #ubuntu-x, they should know [19:21] there is a fglrx.conf in /etc/modeprobe.d [19:22] is it normal for modprobe to take this long on startup - https://launchpadlibrarian.net/80443468/farnsworth-oneiric-20110920-9.png? [19:22] seb128 doesn't see the same issue on his laptop, and we have quite similar hardware - http://people.canonical.com/~seb128/seb-e6410-oneiric-20110920-9.png [19:23] apw: thanks ill ask them when i get back. thanks you for the help [19:41] * jjohansen -> lunch [20:38] hi apw === medberry is now known as med_out === kentb is now known as kentb-out === rsalveti` is now known as rsalveti