JFo | baptistemm, apologies, I got pulled away doing a favor for a friend | 02:13 |
---|---|---|
JFo | will chat with you tomorrow | 02:16 |
AceLan | cooloney: should I file a new bug for the ftrace config changes? | 02:23 |
AceLan | cooloney: or just give you the patch? | 02:23 |
cooloney | AceLan: great, please file a new bug and send me the patch? | 02:31 |
AceLan | cooloney: got it | 02:48 |
cooloney | AceLan: thanks a lot. | 02:49 |
=== rsalveti_ is now known as rsalveti | ||
cooloney | AceLan: could you give me a hand | 03:26 |
cooloney | AceLan: please try to set the mac address on your fsl-imx51 hardware | 03:26 |
cooloney | AceLan: maybe that will cause kernel panic | 03:27 |
AceLan | cooloney: yo | 04:00 |
AceLan | cooloney: my fsl-imx51 h/w doesn't have fec, only wireless | 04:01 |
cooloney | AceLan: oh, too bad | 04:08 |
smb | moin | 08:27 |
smb | psurbhi, morning, have you seen the mail from kees ? | 08:33 |
psurbhi | smb, no | 08:33 |
* psurbhi checks | 08:33 | |
psurbhi | smb, morning :) | 08:33 |
smb | psurbhi, He has found some problems with your sync patch apparently | 08:34 |
psurbhi | smb, ok | 08:34 |
psurbhi | let me check.. | 08:34 |
* smb good-mornings cking | 08:42 | |
cking | hiya | 08:42 |
cooloney | morning smb psurbhi and cking | 08:43 |
* abogani waves | 08:43 | |
psurbhi | hi cooloney, cking | 08:43 |
smb | Hey abogani. Reminds me you asked for some sponsoring and a forgot. You got a page somewhere I could add things to? | 08:45 |
smb | I forgot, even | 08:45 |
smb | Heya apw | 08:45 |
apw | smb, morning you ... | 08:46 |
* apw waves to psurbhi | 08:46 | |
* psurbhi waves back to apw | 08:47 | |
cooloney | apw: morning, | 08:48 |
abogani | smb: The link is https://wiki.ubuntu.com/AlessioIgorBogani/linux-rtPPUApplication. It isn't really necessary anymore because DMB have choosen to refuse me upload rights on linux-rt package. Evidently more than 3 years of work and 7 Ubuntu supported releases don't matter nothing :-) | 08:48 |
apw | o/ cooloney | 08:48 |
apw | abogani, have people been uploading it for you as in sponsoring the uploads though? | 08:49 |
psurbhi | smb, thanks for pointing this out | 08:50 |
abogani | smb: In any case I really appreciate your comments about me and my work. In this I can bring up my moral (if comments are positive obviously) :-) | 08:50 |
smb | abogani, Hm, sad. What exactly was their reasoning? | 08:50 |
abogani | apw: Since Intrepid Luke Yelavich | 08:50 |
abogani | smb: No one sponsor me. | 08:50 |
apw | abogani, thought as much ... | 08:50 |
apw | so was it just lack of sponsors getting behind you, probabally need some social engineering to line them up before hand | 08:51 |
smb | abogani, Doh. Well on my side it was mainly because timing and forgetting | 08:51 |
smb | close to release and some sprints | 08:51 |
abogani | smb: No worry I know that you (like other UKT members are really busy). | 08:52 |
abogani | apw: I'm starting to think that linux-rt don't have a lot of users... :-| | 08:52 |
mase_wk | hmm maybe i should try the RT kernel | 08:53 |
mase_wk | never really looked at it | 08:53 |
* abogani wonders if someone of UKT would want "adopt" linux-rt package. | 08:53 | |
smb | abogani, You usually find out one hour after removing the package completely :-P | 08:54 |
abogani | smb: Indeed | 08:54 |
smb | abogani, No thank you. Its nothing about the package, but I already have enough O_o | 08:54 |
abogani | smb: :) | 08:55 |
apw | abogani, busy for release for sure, and a lot of traveling about just there as well, and ash clouds and ... mad times indeed | 08:57 |
cking | don't mention the a** word | 08:58 |
abogani | apw: Thanks anyway. As I already said above I know than you really are busy. :-) | 09:00 |
smb | cking, the v* word then? | 09:00 |
cking | volcano is fine by me ;-) | 09:00 |
* smb pictures an eruption under cking 's chair | 09:01 | |
cking | that's not a particularly pleasant image | 09:04 |
abogani | smb: Please don't forget to add sponsor to above mentioned wiki page... I really appreciated it :-) | 09:05 |
smb | cking, No. But you mentioned the v* word. Well, I hope for the best. Lucky you don't fly to UDS. I heard some closures in Ireland and Scotland have been done | 09:06 |
smb | abogani, I try to finally remember this time. | 09:06 |
psurbhi | apw, we could give btrfs as an experimental support in maverick | 09:11 |
apw | smb, maybe a*h down here this afternoon ... | 09:12 |
cking | urk | 09:12 |
apw | psurbhi, we generally don't offer installs on EXPERIMENTAL marked filesystems, the major barrier is whether the on disk format has stopped changing or not | 09:12 |
psurbhi | though the wiki says it has not, it seems it has | 09:13 |
psurbhi | but we would find out later if it changes.. as of now.. it does not seem to have changed for a long time | 09:13 |
apw | i think that may be a gating issue, if they have stopped changing it, we may be able to offer it as a 'at your own risk' job | 09:13 |
psurbhi | the wiki says "its still experimental and susceptible to change" | 09:13 |
psurbhi | apw, yes! 'at your own risk' exactly | 09:14 |
apw | a UDS topic for sure, to work out the constraints on whether we can and get people to go investigate them | 09:14 |
psurbhi | ya | 09:16 |
apw | psurbhi, i note that there is an upstream potential fix in that mainline bug linked from the unmount is a slow as snot bug we were just discussing | 09:17 |
apw | you might want to revert our patch and shove that on and test the result | 09:17 |
psurbhi | apw, ok.. for lucid? | 09:18 |
psurbhi | or for maverick? | 09:18 |
psurbhi | apw, i mean do we revert it for lucid too? | 09:19 |
smb | psurbhi, for lucid as your workaround probably has not been copied to maverik, but otherwise both | 09:19 |
psurbhi | smb, ok | 09:19 |
apw | psurbhi, the important thing at this stage is to get the suggested patch tested | 09:20 |
smb | psurbhi, That is the thing you wanted to check when i asked about kees mail. Remember? ;-) | 09:20 |
psurbhi | smb, thats what started the discussion :) | 09:20 |
apw | and report back in the upstream bug if it works for you, perhaps get kees to test as well as he has some more test cases | 09:21 |
psurbhi | thanks for that | 09:21 |
psurbhi | apw, the suggested patch ? | 09:21 |
psurbhi | apw, ya.. get it | 09:21 |
psurbhi | i will do it and report back | 09:22 |
apw | psurbhi, most excellent | 09:30 |
apw | pgraner, about? | 10:10 |
pgraner | apw: yep | 10:11 |
* cking notes that w/o wiggle my life would be painful | 10:33 | |
apw | cking, it is a blessing indeed | 10:35 |
apw | RAOF, hey ... | 10:35 |
apw | cking, can you remind me the incantation on the new CD's to get to nomodeset? | 10:56 |
apw | you hit a key in that initial purple thing don't you | 10:59 |
cking | apw, nay, cannot remember the key | 11:18 |
cking | apw, to get to grub? or what? | 11:19 |
apw | no when we had the new CDs they go purple with a strange thingy at the bottom then boot i think, if you do something while the odd thingy is visible you get the older menu with function keys | 11:19 |
cking | press enter? | 11:20 |
apw | possible will have to make a cd to be sure | 11:25 |
cking | sigh, another day, another ACPI bug... | 11:36 |
cking | 3 cheers for acpiexec - the tool that allows me to single step AML | 11:39 |
apw | cking, sounds amazing ... whos is that? | 11:56 |
cking | apw, http://smackerelofopinion.blogspot.com/2010/03/debugging-acpi-using-acpiexec.html | 11:58 |
apw | cking, geek | 11:59 |
cking | yep | 11:59 |
* apw notes the greeks are attacking their own parliament | 12:00 | |
cking | politics the greek way | 12:01 |
apw | they seem more militant than usual ... lots of throwing things at the moment | 12:01 |
=== cking is now known as cking-afk | ||
=== cking-afk is now known as cking | ||
h00k | I am not familiar with the innerworkings of the kernel, but I'm seeing this in dmesg, I'm just wondering what it means: CE: hpet increasing min_delta_ns to 33750 nsec | 13:35 |
h00k | Turns out there is bug 270798 related to it. | 13:37 |
ubot3 | Malone bug 270798 in linux "lockups with default (hpet) clocksource on 2.6.27-2-generic 64-bit" [Medium,Confirmed] https://launchpad.net/bugs/270798 | 13:37 |
smb | h00k, Just that the hpet could not be programmed with a smaller interval and is now using a bigger one | 13:37 |
h00k | smb: what is hpet? | 13:37 |
smb | high precision event timer | 13:38 |
h00k | alright | 13:38 |
h00k | it appears to keep increasing and over time, my laptop gets slower. I guess I'll mark this bug as affecting me | 13:38 |
smb | h00k, The successor of the old timer with a higher precision (and more bugs apparently ) | 13:38 |
h00k | smb: heh. | 13:39 |
smb | Does it only show this message once or over and over with higher numbers? | 13:39 |
h00k | No, over and over with higher numbers | 13:39 |
h00k | 15000, 22500, 33750 | 13:40 |
h00k | so far, after 529 seconds of uptime. | 13:40 |
h00k | http://pastebin.ubuntu.com/428263/ basically | 13:40 |
h00k | 2.6.32-21-generic | 13:41 |
smb | In theory it should stop at some point. But there might also be a chance that failing to program it is not because it needs more delay but for some other reason | 13:41 |
h00k | I'll keep an eye out, I suppose | 13:41 |
smb | Yeah, if it stays at a certain number it still might be ok | 13:42 |
smb | Not having looked into the bug above I am not sure what lockup means in that context | 13:43 |
h00k | I'll see what it ends up doing after today. | 13:44 |
smb | ... and seeing 2.6.27 as the start of it, I wonder whether it would not make sense to open a new bug for 2.6.32 problems | 13:44 |
h00k | If it keeps increasing? Should I open something new? | 13:44 |
h00k | Ah, okay. | 13:44 |
smb | h00k, If it never stops I definitely would open a new thing | 13:44 |
h00k | smb: oh, upstream might be aware: https://bugzilla.kernel.org/show_bug.cgi?id=14426 | 13:45 |
ubot3 | bugzilla.kernel.org bug 14426 in Realtime Clock "CE: hpet increasing min_delta_ns flood" [Normal,New] | 13:45 |
smb | h00k, Ok, so if you want to open a lp bug make sure to link the upstream bug to it | 13:46 |
smb | If there is not already one linked | 13:46 |
smb | somewhere else | 13:46 |
h00k | alright, i'll check | 13:46 |
h00k | I don't see anything on LP regarding this particular kernel, I'm opening a new bug and I'll link it upstream | 13:54 |
smb | h00k, cool, thanks | 13:55 |
h00k | smb: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/575774 | 14:06 |
ubot3 | Malone bug 575774 in linux "2.6.32-21-generic CE: hpet increasing min_delta_ns flood" [Undecided,New] | 14:06 |
bigcx2 | hey all, is there a way to disable APIC in 2.6.32.11? | 14:11 |
bigcx2 | there's no option for it in menuconfig | 14:11 |
bigcx2 | and when i try to set CONFIG_X86_LOCAL_APIC as no in my config, as soon as i do make oldconfig it overwrites it with yes | 14:12 |
bigcx2 | and then, trying to set it as no in arch/x86/Kconfig makes my kernel build fail with | 14:13 |
bigcx2 | arch/x86/kernel/tsc.c: In function ‘unsynchronized_tsc’: | 14:13 |
bigcx2 | arch/x86/kernel/tsc.c:827: error: implicit declaration of function ‘apic_is_clustered_box’ | 14:13 |
bigcx2 | the same thing happens with 2.6.32.7 | 14:14 |
mjg59 | nolapic on the kernel command line | 14:14 |
mjg59 | Or noapic if you want to disable the i/o apic | 14:14 |
bigcx2 | mjg59: i tried that...the problem is i'm trying to build a RTAI kernel | 14:15 |
bigcx2 | and RTAI complains about it being enabled in the config and then not being available/enabled | 14:15 |
bigcx2 | so that doesn't work for me | 14:15 |
bigcx2 | so there's no way to build any of the latest kernels without apic? | 14:15 |
bigcx2 | that's what it looks like | 14:16 |
mjg59 | Only if you turn off SMP. But I believe that that's always been the case. | 14:16 |
bigcx2 | hm. | 14:17 |
bigcx2 | i don't need smp, lemme give that a shot | 14:17 |
mjg59 | Also, it has to be a 32-bit build | 14:17 |
bigcx2 | it is | 14:17 |
bigcx2 | mjg59: i think that worked, thanks | 14:23 |
apw | psurbhi, hey ... that unmount patch upstream suggested, any idea when you might get that tested? | 15:04 |
psurbhi | apw, i will do it today evening/today morning.. shall send it out soon | 15:04 |
apw | psurbhi, cool, only as ask kees is asking about reverting the current work around as it has some other side effects he doesn't like, and if we make a change i'd like to do both together to make the SRU people not explode | 15:05 |
psurbhi | apw, ack | 15:05 |
pmatulis | is this the proper procedure for updating an initramfs using a live cd?: | 15:07 |
pmatulis | boot, mount the partition containing /boot, mount it (say on /mnt), then 'chroot /mnt update-initramfs -u -k <kernel version>' | 15:07 |
apw | smb, what was the thing about running update-initramfs which was 'negative' ? | 15:07 |
smb | apw, Someone *claimed* it would then not be done on normal updates, but I really did not think this happened to me ever | 15:08 |
apw | ah ok, then i suspect that is the right way to do it | 15:08 |
pmatulis | what if /usr is on a separate partition? | 15:11 |
smb | I thinks so. I don't see any reason why this should be bad, given that it initrd is _always_ automatically build this way. Maybe I am wrong but I have never received a good explanation on why | 15:11 |
apw | gah, pmatulis you would need to mount all the normal things in there | 15:11 |
apw | if /usr is separate mount that, likely you need to mount /proc too | 15:11 |
smb | I guess at least proc and /dev too | 15:12 |
apw | do people still split /usr these days? | 15:12 |
pmatulis | apw: so enter chroot and then mount those partitions? | 15:12 |
apw | yeah after the chroot | 15:12 |
smb | apw, Sounds even more anal than I am | 15:12 |
apw | indeed, something not to be tollerated :) | 15:12 |
smb | Heh, well, maybe there could be reasons, like to make it ro | 15:13 |
pmatulis | but the shell is not available in the /boot chroot | 15:14 |
smb | But at least for a while I did it but ended always up with too little space in either /usr or /opt and then just gave up. Maybe by now online expansion of file system works... | 15:14 |
smb | pmatulis, Is even /bin in another partition or only /usr? | 15:15 |
pmatulis | only /usr | 15:16 |
smb | pmatulis, I guess you need to mount everything into /mnt the intentional / (maybe not /usr) and /proc and /dev | 15:17 |
smb | and /boot | 15:17 |
smb | Thats what I usually do: | 15:18 |
smb | mount /dev/x /mnt (root partition) | 15:18 |
smb | mount /dev/y /mnt/boot | 15:18 |
smb | mount --bind /dev /mnt/dev | 15:19 |
smb | mount --bind /proc /mnt/proc | 15:19 |
smb | chroot /mnt | 15:19 |
pmatulis | ok, that looks reasonable. all this b/c apparently there is a problem with 2.6.32-22 whereby initramfs is not updated after kernel upgrade. still not confirmed | 15:20 |
pmatulis | smb: 6 | 15:20 |
pmatulis | smb: heh, ^^^ | 15:21 |
* smb fall out of his parser at b/c and 6 | 15:22 | |
* psurbhi got a dr appointment - quits for today | 15:23 | |
pmatulis | smb: ? | 15:23 |
apw | b/c == because | 15:23 |
smb | pmatulis, I fail to translate b/c and have no idea what 6 meant | 15:23 |
smb | apw, ahh | 15:23 |
apw | and 6 is what ^ is without shift accidentally | 15:24 |
apw | on a UK/US keyboard | 15:24 |
pmatulis | apw: you rock | 15:24 |
apw | i have seen smb's keyboard :) | 15:24 |
smb | apw, You get the personal translator award. :) | 15:24 |
apw | heh ... :) | 15:24 |
smb | apw, Its all sensible (at least to me) | 15:25 |
apw | i work with a lot of foreign people with non-native english speakers predominating | 15:25 |
apw | pmatulis, not seen that here for sure on any of my machines | 15:26 |
apw | as they really wounldn't boot well without initramds | 15:26 |
smb | pmatulis, Ok, so to get back to the issue... :-P I have the feeling this happens from time to time (not kernel related) but cannot say what is the trigger | 15:26 |
smb | Actually I rather saw this happen with update-grub | 15:26 |
apw | smb, actually i do have a theory ... | 15:27 |
apw | when i did this recent upgrade it did install more than one kernel | 15:27 |
* smb is all ears | 15:27 | |
apw | arn't initramfss generated as a 'delayed' dpkg triggery thing? | 15:27 |
pmatulis | maybe it has to do with update-grub then | 15:27 |
apw | and that actually only does the 'latest' one | 15:27 |
apw | so we might in theory miss one therre, though not the one you should be using of course | 15:27 |
apw | hrm | 15:27 |
smb | I think that would be the intention. | 15:28 |
smb | Its probably hard to get all the special cases right | 15:29 |
smb | as the initrd update is also triggered by essential userspace changing | 15:29 |
smb | like installing lvm or updating udev | 15:29 |
smb | That you would only want to do on the 'current' kernel | 15:30 |
smb | but installing a new kernel obviously needs it all the time | 15:30 |
smb | pmatulis, My update-grub issue would 'only' be not seeing the new kernel. The initrd was there as far as I remember | 15:32 |
smb | So it might be two things | 15:32 |
pmatulis | smb: forgot to mention something | 15:32 |
pmatulis | smb: the problem purportedly manifests itself when using LVM | 15:32 |
pmatulis | smb: it's the lvm2 module that is missing from the initrd | 15:33 |
pmatulis | smb: using ext4, not sure if that is important | 15:33 |
smb | Is that installing lvm2 without a kernel update? | 15:34 |
smb | If the kernel is also updated at the same time, this might be a third problem we sometimes see: kernel is updated/installed->initrd is (re-)generated. lvm2 is installed at the same time but initrd is not regenerated because it thinks this is already done. | 15:36 |
pmatulis | smb: no, LVM is in use on disk | 15:36 |
smb | pmatulis, Then we should wait for the results of running update-initramfs manually. If that still fails to include lvm2 it is a problem there | 15:37 |
pmatulis | smb: alright, will report back | 15:38 |
JFo | ogasawara, interested in this bug? It's an old one. bug 45021 | 15:46 |
ubot3 | Malone bug 45021 in linux-source-2.6.22 "including "omnibook" module would help with ACPI support on HP laptops" [Medium,Won't fix] https://launchpad.net/bugs/45021 | 15:46 |
JFo | anyone have an idea who needs to see bug 574184 | 15:46 |
ubot3 | Malone bug 574184 in linux "Bad signatures for 10.04 installer validation: MD5SUM SHA1SUM SHA256SUM" [Undecided,Confirmed] https://launchpad.net/bugs/574184 | 15:46 |
JFo | apparently the DVD ISOs have bad signatures | 15:46 |
smb | JFo, maybe cjwatson or slangasek directly? | 15:48 |
JFo | smb, ta | 15:49 |
apw | JFo, yeah bring that up on ubuntu-devel, | 15:51 |
JFo | apw, done | 15:52 |
JFo | apw, or were you thinking that was a bad idea? | 15:55 |
apw | no it is the right thing. thats a case of someone using linux to mean 'the whole thing' not the 'kernel' | 15:55 |
JFo | right | 15:55 |
JFo | thought so\ | 15:55 |
JFo | here is an interesting bug 575518 | 15:56 |
ubot3 | Malone bug 575518 in linux "linux kernel contains GPL violations" [Undecided,Incomplete] https://launchpad.net/bugs/575518 | 15:56 |
* smb closes eyes and ears | 15:56 | |
mjg59 | JFo: It's entirely unclear whether including those objects is something other than mere aggregation | 15:58 |
apw | smb, its the remaining inbuilt firmwares | 15:58 |
apw | as they come directly from linus' tree i am less worried about it | 15:58 |
JFo | same here | 15:58 |
JFo | but I wanted to bring it up | 15:58 |
persia | Isn't fixing that 90% just editing debian/copyright? | 15:58 |
apw | JFo, pgraner has the lead on all firmware related issues so i would get him on the case | 15:59 |
persia | so that the entirety of the source *isn't* claimed GPL? | 15:59 |
JFo | apw, that is the plan :) | 15:59 |
* persia thought someone went through that pain about 18 months ago already | 15:59 | |
mjg59 | persia: No, large quantities of the source tree aren't GPL | 15:59 |
mjg59 | persia: THere's a lot of it that's BSD, for instance | 15:59 |
JFo | pgraner, you there my captain? | 15:59 |
mjg59 | And some dual GPL/MPL | 15:59 |
apw | persia, for the majority of consumers i suspect so yes, though i suspect from the reports tenor that they will not be happy to install a kernel which pulls them in | 15:59 |
JFo | yeah, sounds like they want the option to leave them out | 15:59 |
apw | persia, someone did for the main firmware packages, not for the small inbuilt ones | 16:00 |
persia | mjg59: I still think it's largely a documentation issue, but I may be mistaken. That said, most BSD stuff becomes GPL when integrated with GPL stuff, doesn't it? | 16:00 |
persia | apw: Ah, that makes sense. | 16:00 |
mjg59 | persia: No, BSD stuff remains BSD | 16:00 |
persia | OK. | 16:00 |
mjg59 | persia: But that's fine, because the BSD license doesn't confer any restrictions above and beyond the GPL | 16:00 |
apw | only doing GPL things with BSD ensures you don't violate BSD cause its essentially a superset of the GPL rights | 16:01 |
persia | Right. I'm confusing license-of-the-compiled-binary-package and license-of-the-compiled-binary-objects. | 16:01 |
pmatulis | smb: bad news: | 16:02 |
pmatulis | mount: unknown filesystem type 'LVM2_member' | 16:02 |
apw | all thats said i thought there was a driver to pull them all out to loadable status, not sure if that has occured yet | 16:02 |
mjg59 | apw: It's happening gradually | 16:02 |
apw | thought as much, not sure when it slated to be 'done' | 16:02 |
smb | pmatulis, You need to install lvm2 to the live image | 16:02 |
apw | as us trying to preceed is plain bonkers :) | 16:02 |
mjg59 | apw: That said, I think that analysis is wrong in places | 16:02 |
ogra | apw, will anyone from the kernel team merge initramfs-tools ? | 16:03 |
apw | ogra, do what ? | 16:03 |
apw | ogra, as in sync it from debian ? | 16:03 |
ogra | apw, merge the initramfs-tools package | 16:03 |
ogra | yes | 16:03 |
mjg59 | apw: In fact, it *is* wrong in places | 16:03 |
ogra | not *sync* | 16:03 |
ogra | that needs heavy-lifting merge work :) | 16:03 |
apw | normally foundations has done that IIRC | 16:03 |
mjg59 | drivers/scsi/aic7xxx/aic79xx.seq is *source* and comes with a GPL compatible license | 16:03 |
ogra | ah, kernel-team is the owner since lucid | 16:04 |
mjg59 | The toolchain for it is *included in the kernel source* | 16:04 |
apw | yeah it does appear to be a list of all files not called .c | 16:04 |
apw | and following his links its to the radical end of the alarmist world | 16:05 |
mjg59 | drivers/scsi/sym53c8xx_2/sym_fw1.h looks pretty like source to me | 16:06 |
apw | 'its not free software because i cannot understand it' | 16:06 |
apw | not 'its not free because i am not allowed to distribute it freely' | 16:06 |
ogasawara | JFo: I took care of that bug - https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/45021/comments/34 | 16:08 |
ubot3 | Malone bug 45021 in linux-source-2.6.22 "including "omnibook" module would help with ACPI support on HP laptops" [Medium,Won't fix] | 16:08 |
apw | JFo, we can talk to pete next week about it i recon | 16:08 |
apw | i don't see it being a major issue | 16:08 |
JFo | cool | 16:08 |
JFo | yeah, no big, just wanted to provide the head's up | 16:08 |
apw | yeah i think most of it is non-pragmatic and we are not in any real danger there, but its for the lawyers not 'us' to take a position and thats petes end of the world | 16:09 |
manjo | ogasawara, omnibook is under ubuntu/ | 16:11 |
ogasawara | manjo: right, which is why I marked it Fix Released | 16:12 |
ogasawara | manjo: and posted the reasoning in the comment | 16:12 |
* apw waves to manjo | 16:12 | |
* manjo waves back to apw | 16:12 | |
manjo | ogasawara, right. | 16:13 |
pgraner | JFo: you can close that bug | 16:14 |
pgraner | JFo: we created linux-firmware-nonfree to address that | 16:15 |
pgraner | JFo: I forgot to close it | 16:15 |
JFo | pgraner, I think apw was indicating it was things that were not in nonfree | 16:15 |
JFo | ogasawara, thanks, just now saw your response above | 16:16 |
manjo | apw, any plans for suspend resume this cycle ? except for freq based reporting which I moved to this cycle ? | 16:16 |
apw | i suspect that given his desire to point out nonfree in his bug, that that would placate him | 16:16 |
JFo | apw, cool | 16:16 |
apw | curtainly worth saying we now have that same as debian, and firmware has been cleaned up and recategorised | 16:16 |
JFo | I'll write up something | 16:16 |
pgraner | apw: ack | 16:16 |
* pgraner is off to find food | 16:17 | |
JFo | enjoy | 16:17 |
=== pgraner is now known as pgraner-afk | ||
JFo | apw, just reading your e-mail re:KMS Bugs... dude, you read my mind and actually created the tag I was thinking of using... scary. :) | 16:38 |
apw | jfo you need a thicker tin helmet | 16:39 |
JFo | indeed | 16:39 |
JFo | and some aluminium lining :) | 16:39 |
cking | great minds and all that | 16:44 |
kees | apw: which patch did surbhi see as helping fix the umount stalls? | 16:46 |
smb | kees, The one you found causing the slow downs afaik | 16:54 |
kees | smb: no, I mean, what's the proposed new solution? | 16:54 |
apw | kees, thats not been found, mearly me requesting she test the patch as proposed at the end of the upstream bug | 16:54 |
smb | kees, She is currently looking into the patch that was proposed in the upstream bugzilla | 16:55 |
apw | to see if it is the fix or not | 16:55 |
kees | okay, right, I wanted to direct attention there if it wasn't already known, but it is, so good. | 16:55 |
kees | I'm glad to at least understand the trigger (boundaries) | 16:55 |
ogasawara | smb: what was the keybinding incantation you had in your .vimrc file? | 16:56 |
achiang | there's this too: http://www.spinics.net/lists/linux-ext4/msg18780.html | 16:56 |
achiang | kees: ^^ | 16:56 |
smb | ogasawara, A second (if apw is not again faster) | 16:56 |
smb | ogasawara, map =F :call Finalise()<ctrl-m> | 16:58 |
ogasawara | smb: awesome, thanks | 16:58 |
kees | achiang: ooh, yeah | 16:58 |
smb | ogasawara, The ctrl-m is ctrl-v+ctrl-m | 16:58 |
smb | ogasawara, map =N :call NewVersion()^M^[:call Finalise()^M''A | 16:59 |
smb | ogasawara, Thats probably better to describe what you see. ^M and ^[ are control sequences | 17:01 |
pmatulis | re fix-committed for karmic in bug 513848, wondering when it will become fix-released | 17:07 |
ubot3 | Malone bug 513848 in linux "[karmic] CPU load not being reported accurately" [Low,Fix committed] https://launchpad.net/bugs/513848 | 17:07 |
apw | smb, that one i think is acked and applied, when is the next -proposed going in? | 17:07 |
smb | apw, Somewhen after the next round of security | 17:08 |
smb | IOW don't know, yet | 17:08 |
apw | ahh a while then | 17:08 |
smb | I hope not too long | 17:08 |
pmatulis | a few weeks? | 17:08 |
smb | pmatulis, yeah, given UDS interruption | 17:09 |
pmatulis | smb: alrighty then | 17:09 |
ogasawara | smb, apw: I think my ping yesterday missed you guys, but did we accidentally lose this commit in Lucid? - "UBUNTU: [Config] Add atl1c to nic-modules udeb" | 17:10 |
ogasawara | smb, apw: kernel-team ml thread was "Lucid pull request, SRU lp557130" from rtg | 17:11 |
apw | ogasawara, i thought that was Fix Committed | 17:11 |
smb | ogasawara, Thought I did that. lemme check i messed up to push | 17:11 |
smb | ogasawara, Hm, I might have forgotten to push. i see it locally but not remotely | 17:13 |
smb | ogasawara, I am in the middle of a am sequence the. Give me a few min and I push it | 17:14 |
ogasawara | smb: ok cool. no hurry. | 17:14 |
smb | ogasawara, ok, pushed. Sorry for that | 17:19 |
ogasawara | smb: thanks | 17:19 |
JFo | you go ogasawara crack that whip ;-0 | 17:20 |
* JFo grumbles about his shift key | 17:20 | |
ogasawara | JFo: heh, I'd have not even noticed it had tim not nudged me when he sent the patch mentioning I'd want it for Maverick | 17:21 |
JFo | ah | 17:21 |
cking | shift key? | 17:23 |
vanhoof | pgraner-afk: ping | 17:25 |
apw | cking, 0 -> ) | 17:29 |
JFo | heh, sorry, should have been more clear | 17:30 |
smb | apw is universal keyboard translator today | 17:30 |
JFo | indeed | 17:30 |
* JFo applauds apw | 17:30 | |
smb | In my world this would have been 0 -> = | 17:31 |
apw | smb, :) | 17:31 |
* smb is away to have some fun | 17:32 | |
=== smb is now known as smb-afk | ||
cking | heh, 3 hours to print a map on my Brother laser printer. That's what I call painful | 17:35 |
manjo | cking, how big is the map ? | 17:37 |
cking | manjo, not that big - my laser printer has a small buffer, and rendering images takes forever | 17:39 |
cking | my fail for buying a lame printer | 17:39 |
JFo | <-lunch | 18:00 |
* manjo looks towards the kitchen | 18:04 | |
cking | always thinking of food? | 18:05 |
manjo | :) | 18:09 |
* apw is at the moment (food!) | 18:10 | |
cking | bah, now I'm hungry | 18:10 |
apw | cking, it _is_ dinner time, yours must be burned by now | 18:10 |
cking | ..just one more bug... | 18:16 |
manjo | apw, I guess you missed my Q earlier... I am wondering if you have any thoughts for a suspend/resume blueprint? I can think of frequency based reporting that I moved to this cycke | 18:30 |
* apw notes that PPA's have a delete button all of a sudden | 18:30 | |
apw | i don't think we ahve any plans for detailed work do we? | 18:31 |
apw | seems excessive to have a whole blueprint for one item like that | 18:31 |
apw | we have talked about having a catchall blueprint for those | 18:31 |
apw | for those small items which would be too small to warrent their own | 18:32 |
cking | hehe, my neighbour just drove their car up the driver and nearly eliminated a tory councillor delivering leaflets | 18:32 |
cking | s/driver/drive/ | 18:33 |
apw | sounds worthy to me | 18:33 |
manjo | apw, yep that works for me | 18:34 |
apw | manjo, so just keep it in mind and i'll get one made | 18:39 |
kro | which special options do I need in my kernel to be able to boot a ubuntu 10.04 userland? | 18:40 |
kro | so far I found out I need devtmpfs | 18:40 |
kro | and I have to remove ureadahead | 18:40 |
apw | yep ureadahead needs some patches from the ubuntu kernel | 18:40 |
kro | but I still can't boot a vanilla kernel on a fresh 10.04 :-/ | 18:40 |
apw | i generally build mine with the lucid config | 18:41 |
kro | systems get stuck during mountall/upstart-udev-bridge/udev | 18:41 |
apw | perhaps you don't have unix domain sockets built in? | 18:41 |
kro | that's not an option here :-/ | 18:42 |
apw | why so? | 18:42 |
apw | pretty sure something like that is used for mountall | 18:43 |
kro | CONFIG_ what would that be? | 18:43 |
kro | because it's some sort of company kernel... I can askto have some options added or removed, but I can't change the whole conf | 18:44 |
* apw checks what it is that it uses | 18:44 | |
=== cking is now known as cking-afk | ||
kro | (running on several thousand servers) | 18:44 |
apw | so you can change the whole of userspace ? | 18:44 |
apw | kro CONFIG_UNIX would be unix domain sockets | 18:46 |
kro | ah, nope, that one's =y | 18:46 |
apw | hrm ok | 18:46 |
apw | probabally need to ask the upstart people what its needing | 18:47 |
apw | keybuk is the ubuntu maintainer there | 18:47 |
kro | but right, upstart-udev-bridge seems to hang on some socket related stuff... | 18:48 |
abogani | kro: CONFIG_CONNECTOR? | 18:48 |
kro | bingo. | 18:48 |
kro | do I need that? | 18:48 |
apw | thats =y for distro and =m for ports and i assume both boot | 18:49 |
apw | what you got there? | 18:49 |
kro | I got "is not set", i see lucid's kernel is =y | 18:49 |
* apw has no idea if that is used, abogani is that something known needed for upstart? | 18:50 | |
apw | oh yeah that looks like something upstart would need to wait for its children | 18:51 |
kro | Connector - unified userspace <-> kernelspace linker | 18:51 |
kro | sounds good | 18:51 |
kro | thx, I'll give it a try | 18:51 |
abogani | apw: I'm not sure but I noticed some days ago a strange use of netlink for upstart/udev comunication. | 18:53 |
apw | looking at the description i think that upstart uses it for all child instantiation and tracking | 18:53 |
abogani | Sorry for (big) off-topic. I hope that no one flame me for that. | 18:55 |
* abogani wonders if there are in this channel someone could offers him some sort of job. | 18:55 | |
abogani | Thanks in advance! | 18:55 |
apw | all our jobs are listed on the web site | 18:56 |
manjo | apw, if lsusb lists a bluetooth device, that means the kernel is able to recognize the device correct ? | 19:06 |
=== cnd_mini is now known as cnd | ||
cnd | apw, smb, ogasawara: does a config enforce addition require a bug and an SRU (for lucid)? | 19:23 |
cnd | JFo: I keep getting emails about bug 521967, so I decided to take a look | 19:23 |
ubot3 | Malone bug 521967 in linux-backports-modules-2.6.32 "support for new atheros wifi chipset - AR2427/ath9k" [Medium,Triaged] https://launchpad.net/bugs/521967 | 19:23 |
cnd | JFo: it seems there's an upstream patch that was CC'd to stable@kernel.org, but it hasn't been accepted for over two months | 19:23 |
apw | build system changes arn't SRU material, but a bug makes sense | 19:23 |
cnd | JFo: I'm kinda busy with hwe, but maybe you could find someone to take a quick look at it? | 19:23 |
JFo | nice | 19:23 |
JFo | I'll see what i can do cnd | 19:23 |
cnd | JFo: thanks | 19:24 |
JFo | my pleasure | 19:24 |
JFo | hmmm | 19:26 |
ogasawara | cnd: hrm, even though it does seem a bit overkill for enforcing a config option, I'd say it needs to follow the same rules for all changes being applied to Lucid. So yes, a bug and an SRU. | 19:26 |
ogasawara | cnd: but I'd check with smb first as he's the official gatekeeper for Lucid | 19:26 |
JFo | cnd, that ath9k bug would explain why I have seen a veritable flood of ath9k issues | 19:27 |
cnd | JFo: potentially | 19:27 |
cking-afk | by | 19:37 |
cking-afk | bye | 19:37 |
manjo | ogasawara, when apport collects lspci and lsusb outputs, May be we should collect it will -vv flags | 19:44 |
cnd | ogasawara: do I need to do anything to have the patch I just sent for lucid also applied to maverick? | 19:44 |
ogasawara | cnd: nah, just mention you want it applied for maverick and I'll pull it | 19:44 |
cnd | ogasawara: k, thanks | 19:45 |
ogasawara | manjo: I thought it did use the -vv flags for lspci (not sure about lsusb) | 19:45 |
ogasawara | manjo: I'm sure pitti would take patches :) | 19:45 |
manjo | ogasawara, I think it does not collect lsusb right now | 19:46 |
manjo | ogasawara, also dmidecode information will be useful especially dealing with bios bugs | 19:46 |
apw | dmidecode is tricky cause we cannot run root things | 19:47 |
ogasawara | manjo: yep, what apw said | 19:47 |
apw | there is a file containing the output at boot think, and that may be taken i forget | 19:47 |
ogasawara | manjo: I thought it might inject some bits into the bug description | 19:47 |
manjo | ogasawara, ah yes it does | 19:48 |
manjo | lsusb -v needs to be run as root to collect complete info, other wise it will say permission denied for some info | 19:49 |
manjo | nvidia driver seems to be trash.. I see a lot of reports where it takes more than 8sec to resume | 20:01 |
apw | manjo, yep, radeon seems to take 10s here for me | 20:08 |
manjo | I am trying to understand the bits and pieces of nvidia in the src... there is drivers/video/nvidia and drivers/gpu/drm/nouveau | 20:10 |
abogani | I suspect than a lot of magic happens in blob... | 20:11 |
abogani | By the way, When we'll know what is the final version of kernel used in Maverick? | 20:11 |
ogasawara | abogani: next thurs, tbd at UDS | 20:12 |
abogani | ogasawara: Thanks. | 20:12 |
akgraner | apw et al you awesome kernel people's :-) 1500 UTC Kernel Q &A Session for open week :-) - https://wiki.ubuntu.com/UbuntuOpenWeek | 20:15 |
apw | ogasawara, we need to get a possy together for that, you making it ? | 20:16 |
ogasawara | apw: yep, I'll be there | 20:16 |
JFo | akgraner, that is tomorrow, yes? | 20:16 |
akgraner | yeppers | 20:16 |
JFo | I'll be there if that matters | 20:16 |
JFo | :) | 20:16 |
akgraner | JFo, that will be 1100 your time :-) | 20:16 |
JFo | k | 20:17 |
* JFo plans coffee early | 20:17 | |
apw | who is going to take point, akgraner i assume you are running the show | 20:17 |
akgraner | apw, nah I just help :-) | 20:17 |
JFo | snort* | 20:17 |
JFo | errr | 20:17 |
akgraner | but yeah I'll be there and can drive the bot if you all want me too | 20:17 |
apw | ogasawara, you've done one of these before haven't you ... i'm clueless | 20:18 |
ogasawara | hrm, we should have some talking points | 20:18 |
JFo | well, it is Q & A | 20:18 |
JFo | so they ask, we answer? | 20:18 |
akgraner | ogasawara, if you want to add some slides put them in pdf format and shoot me the link to them | 20:18 |
apw | we might want some basic questions lined up to kick things off | 20:18 |
ogasawara | JFo: yah, but they sometimes need a little push | 20:18 |
JFo | good point | 20:18 |
akgraner | we can get them added to your session as well | 20:18 |
apw | ogasawara, remember those slides we did for pete | 20:19 |
apw | that was all questions and answers | 20:19 |
cnd_mini | ogasawara: what do the different colors on the kernel track uds schedule mean? | 20:19 |
ogasawara | apw: hrm, I don't recall those | 20:19 |
apw | when you wake up normally UTC ? | 20:19 |
JFo | cnd_mini, there is a schedule? | 20:19 |
ogasawara | cnd_mini: the colors usually group by track - ie the kernel is that yellow/orange color | 20:19 |
* JFo would like to see | 20:20 | |
ogasawara | JFo: summit.ubuntu.com | 20:20 |
apw | i am not sure i can be bothered tonight, but i can look them out tommororow | 20:20 |
JFo | ogasawara, thanks | 20:20 |
ogasawara | apw, akgraner: I never usually prepared slides | 20:20 |
cnd_mini | ogasawara: ahhh, thanks | 20:20 |
apw | ogasawara, i didn't mean as slides, more as potential basic questions we might ask each other if noone else has any to start with | 20:20 |
JFo | odd that is still shows me as not registered to attend | 20:20 |
ogasawara | apw: I'm usually up around 14:30 UTC, but I'll get up a little earlier tomorrow so we're prepared | 20:21 |
akgraner | with classbot and lernid sessions can now have slides - was just letting ya know | 20:21 |
akgraner | :-) | 20:21 |
akgraner | just in case | 20:21 |
ogasawara | apw: I'll throw together some general Q & A in an email today | 20:21 |
apw | akgraner, do i need to do anything other trhan use xchat in those two channels? | 20:21 |
ogasawara | apw: then maybe we can mumble in the morning | 20:21 |
ogasawara | apw: that's all I ever needed | 20:22 |
apw | ogasawara, you are something of a hearo | 20:22 |
akgraner | apw, you may want to use lernid | 20:22 |
akgraner | that way you can see both channels at the same time | 20:22 |
akgraner | side by side | 20:22 |
apw | i can see both with xchat too :) | 20:22 |
akgraner | ok then you are good to go :-) | 20:22 |
ogasawara | akgraner: tomorrow could you op both apw and I so we both can post to #ubuntu-classroom | 20:25 |
JFo | sweet! | 20:25 |
akgraner | yep | 20:25 |
* JFo gets out of it | 20:25 | |
akgraner | kirkland, are you all good to go for the Server Q & A for 1600 UTC tomorrow as well? or do I need to pop into another channel and ask? | 20:26 |
* ogasawara lunch | 20:39 | |
=== oubiwann is now known as oubiwann_ | ||
=== oubiwann_ is now known as oubiwann | ||
cnd_mini | sconklin: I just sent you my key for the uds key signing session, but I had to do it using msmtp directly, so can you verify everything looks ok? | 21:18 |
sconklin | looking | 21:18 |
sconklin | looks ok. The mailer even recognized it and asked if I wanted to import it | 21:20 |
cnd_mini | ahh cool | 21:20 |
sconklin | If there's a problem I can always get it from you directly at UDS before Wed | 21:20 |
sconklin | I'm still frantically working on a demo of the patch tracking stuff and slides in case I get picked for a plenary presentation | 21:20 |
cnd_mini | well, now I know how to send attachments using the cmdline :) | 21:21 |
=== cnd_mini is now known as cnd | ||
cnd | I've been working nonstop on oem stuff... | 21:22 |
sconklin | cnd: I totally understand | 21:23 |
cnd | I think things should be quieted down now though, at least for a bit | 21:23 |
JFo | cnd, I have not yet gotten anyone to look over that bug | 21:28 |
JFo | <-slack | 21:28 |
cnd | JFo: np, consider it just a wish | 21:29 |
JFo | heh | 21:29 |
cnd | I don't really care other than to quiet the people down | 21:29 |
JFo | yeah | 21:29 |
JFo | I understand | 21:29 |
mrec | hey, does anyone know is there something like a broadcast message available before a system goes into hibernation? | 21:31 |
JFo | mrec, you mean to users? | 21:31 |
mrec | yes, I need to close and restart an application (as root) when this happens | 21:32 |
JFo | I'm not aware of anything | 21:32 |
JFo | but then, others are more knowledgeable | 21:32 |
mrec | it's a problem for live data.. | 21:32 |
mrec | as long as the notebook is down it will miss the data and after it wakes up the data is corrupted | 21:32 |
JFo | I see | 21:33 |
mjg59 | mrec: There isn't. The easiest thing is for you to drop a script fragment into pm-utils. | 21:35 |
mrec | hmm ok this works for ubuntu | 21:37 |
mrec | I guess this is absolutely messed up and every distribution has its own path for this | 21:38 |
mrec | but it helps alot for ubuntu thanks | 21:38 |
kamal | mjg59: Hi - yesterday you kindly pointed me to i915_opregion.c as a possible solution to the Dell Studio 1558's non-functional DSDT brightness control -- may I update you on my progress and where I'm stuck? | 22:15 |
mjg59 | Sure! | 22:15 |
* kamal pastes away... | 22:15 | |
kamal | Today I hacked up some test code where I arrange for i915_driver_irq_handler() to call asle_set_backlight() based on a global that I stuff in the acpi/video brightness code ... | 22:15 |
kamal | Additional printk's show that my hack does end up calling asle_set_backlight() in response to the brightness keys and that routine returns 0 (success) -- but alas, there is no effect on the actual backlight brightness. | 22:15 |
kamal | Should I expect asle_set_backlight() to "just work" on anything that supports opregion? Or perhaps there's some additional opregion setup steps that I'm supposed to be doing first? | 22:15 |
kamal | (Side note: I also tested moving the misplaced _BCM routine where I thought it was supposed to go -- that did make the acpi/video driver happier, but that _BCM routine doesn't actually affect the brightness either. Further experiments showed that the registers that _BCM sets just don't any visible effect). | 22:15 |
kamal | mjg59: ^^ Any advice will be most appreciated. | 22:15 |
mjg59 | kamal: Ok, unless Dell are doing something *very* odd then the writes should probably work | 22:16 |
mjg59 | kamal: Can you check which registers asle_set_backlight() is hitting? | 22:17 |
kirkland | akgraner: i cannot do that | 22:17 |
kirkland | akgraner: will need to get someone else | 22:17 |
kirkland | akgraner: try mathiaz, zul, or smoser | 22:17 |
mjg59 | Oh, hang on | 22:17 |
mjg59 | kamal: More to the point, what's getting passed to asle_set_backlight in the bclp argument? | 22:18 |
kamal | mjg59: :-) I was careful to pass in : ( (1<<31) | mybrightnessvalue ) where mybrightnessvalue ranges from e.g. 0 to 100 (I see that it accepts 0 to 255 range) | 22:19 |
kamal | mjg59: I do check that asle_set_backlight returns 0, which it only does if it gets all the way through. | 22:19 |
mjg59 | kamal: Yeah, ok. So I guess you get to make sure that it's writing the registers | 22:20 |
kamal | mjg59: ... this was all of course after missing the ASLE_BCLK_VALID check the first time ;-) but yes, its writing the registers | 22:20 |
mjg59 | kamal: Hrm. | 22:21 |
mjg59 | kamal: Oh... | 22:21 |
mjg59 | kamal: asle_set_backlight_ironlake() | 22:21 |
mjg59 | I bet that'll help things | 22:21 |
kamal | mjg59: Doh! I noticed that yesterday, and wondered, but then forgot about it! | 22:21 |
mjg59 | This is an i5? | 22:22 |
kamal | mjg59: yes, an i5. | 22:22 |
mjg59 | Yeah | 22:22 |
kamal | yeah | 22:22 |
mjg59 | Needs to be the ironlake version | 22:22 |
kamal | :-) Ok, I'll take another spin at it. I'll let you know! | 22:22 |
kamal | mjg59: Bingo! it works perfectly! | 22:34 |
kamal | (well, as perfectly as a horrid hack should work at least) | 22:35 |
mjg59 | kamal: Heh | 22:35 |
mjg59 | kamal: Ok, so now you just need to add a callback registration function to video.c and have i915 set that up before it calls acpi_video_register() | 22:36 |
mjg59 | Then if that's set, use it rather than calling _BCM or _BCL | 22:36 |
kamal | mjg59: ok got it -- I'll let you know when I have that ready for review -- thanks very much for your help with this. | 22:38 |
mjg59 | No problem | 22:38 |
mjg59 | Saves me having to write it | 22:38 |
kamal | mjg59: I'll need some guidance about how we can detect whether opregion-brightness-control is supported ... and will we want to let i915 take over brightness control by default? Anyway, I'll get working on the callback-based functionality and then we/you can figure out when to enable it. | 22:43 |
mjg59 | kamal: If you're in i915_opregion_init() then it's probably supported - there's some error checking through there | 22:44 |
kamal | mjg59: is it the case that all devices which support opregion will support the backlight control mechanism? | 22:44 |
mjg59 | You might want to check the spec | 22:45 |
mjg59 | I expect so, since it's about the only useful thing opregion gives you | 22:45 |
kamal | mjg59: :-) ok got it -- anyway I'll focus on getting the callback working. thanks again! | 22:46 |
=== kamal is now known as kamal-afk |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!