| === hggdh_ is now known as hggdh | ||
| bryceh | ogasawara, almost hate to mention it but I got my hands on a system that dislikes rc6 - lp #954660 | 02:01 |
|---|---|---|
| ubot2` | Launchpad bug 954660 in linux "Intel DH67GD board w/ i5-2405S cpu fails to boot with rc6 enabled" [Undecided,Confirmed] https://launchpad.net/bugs/954660 | 02:01 |
| ah- | thanks a lot apw that's exactly what i was looking for | 02:04 |
| ogasawara | bryceh: thanks, I'll contact eugeni | 03:21 |
| smb | Another 70M since yesterday... yay... | 08:09 |
| apw | smb, 70 ... sigh | 08:16 |
| apw | smb, 77 for me, and another new unity | 08:17 |
| smb | Must have some different packages and/or other things installed. Btw, we should have another kernel. The last one is so long ago... :-P | 08:18 |
| ppisati | moin | 08:19 |
| smb | ppisati, ciao | 08:19 |
| apw | smb, so last week | 08:19 |
| ppisati | 90 new pkgs | 08:21 |
| ppisati | cool | 08:21 |
| smb | "thank you for joining the mozilla test pilot team"?!? wtf! | 08:22 |
| apw | smb, oh i saw that on one of my test boxes yesterday evening too. thought i'd clicked something by mistake ... perhaps not | 08:23 |
| smb | apw, No I guess because they gave us a beta version... bah | 08:24 |
| apw | why so i want a beta version in an LTS | 08:24 |
| smb | because you can? | 08:24 |
| apw | oh sorry silly me, rules don't apply to everyone, lets upload 3.3-rc7 today | 08:25 |
| smb | apw, There are rules? rules are so last week... :-P | 08:26 |
| apw | so last cycle | 08:27 |
| apw | smb, interesting, just started firefox here (not my normal browser) and not such popup | 08:27 |
| smb | apw, The test pilot thing was a tab, but weird anyway | 08:28 |
| smb | !#$&, do I only have to reboot to apply the changes or did they steal alt-tab | 08:36 |
| ubot2` | smb: Error: I am only a bot, please don't think I'm intelligent :) | 08:36 |
| * smb reboots to see whether that brings back alt-tab | 08:41 | |
| apw | well i guess it boots | 08:43 |
| smb | It seems to... | 08:45 |
| smb | alt-tab is lost though | 08:45 |
| jjohansen | apw: how would you like updated apparmor kernel patches? tree with reverts, and patches that match versions that I am push upstream, or just the diffs. There will be a rebase against what gets upstream coming at some point too | 10:11 |
| apw | jjohansen, for P i think reverts and new patches is simplest then we can rebase the old ones out of existance | 10:12 |
| apw | anything shiney in this one ? | 10:12 |
| jjohansen | apw: that is what I figured, nothing shiny just bug fixes and changes based on upstream review | 10:12 |
| jjohansen | the shiny stuff will all be in M | 10:16 |
| apw | jj M ? | 10:24 |
| apw | jjohansen, ^^ | 10:25 |
| jjohansen | apw: haha, oops, I sorry I was looking at a maverick security/sru stuff when I wrote that | 10:27 |
| jjohansen | apw: P+1, so Q | 10:27 |
| jjohansen | :) | 10:27 |
| * ppisati -> lunch | 12:22 | |
| === smb` is now known as smb | ||
| === bladernr_afk is now known as bladernr_ | ||
| jdstrand | jjohansen: wowo, 'M'. That is the *long* view :) | 13:13 |
| jjohansen | jdstrand: well I like to plan ahead | 13:14 |
| jdstrand | 'wowo'? | 13:14 |
| jdstrand | jjohansen: :) | 13:14 |
| ogra_ | apw, poke | 13:56 |
| ogra_ | apw, i ran into bug 954243 yesterday, do you think it would be possible to make devtmpfs mount /dev/pts ? | 13:57 |
| ubot2` | Launchpad bug 954243 in linux-ti-omap4 "booting pandaboard without initrd dies with "init: no available ptys"" [Undecided,New] https://launchpad.net/bugs/954243 | 13:57 |
| ogra_ | (or tgardner ^^^) | 13:57 |
| apw | ogra_, right ... it assumes you setup your environment before upstarts runs, it _assumes_ you have an initrd | 14:02 |
| ogra_ | exactly | 14:03 |
| ogra_ | that makes initrdless boot impossible | 14:03 |
| apw | all this to save .1s | 14:03 |
| ogra_ | heh | 14:03 |
| ogra_ | did you measure ? | 14:03 |
| apw | when boot currently is in the 50-70 range cause of a bunch of problems | 14:03 |
| apw | no, but frankly it cannot be more than 5s | 14:04 |
| ogra_ | on arm its more like 15s | 14:04 |
| apw | out of what 4m ? | 14:04 |
| ogra_ | given the bootloader needs >10s to read and shuffle the initrd in ram | 14:04 |
| ogra_ | my ac100 currently boots in 25s | 14:04 |
| ogra_ | omap4 is rather like 90-120s though | 14:05 |
| ogra_ | thanks to the slow bootloader and bad IO on SD cards | 14:06 |
| apw | ogra_, ok ... acdording o jodh then if you add --no-log to the command line upstart should stop using pts and in theory you should be able to boot, can you test that for me? | 14:11 |
| ogra_ | apw, sure | 14:12 |
| ogra_ | will take a moment as my current install is totally trashed | 14:13 |
| * ogasawara back in 20 | 14:19 | |
| === josephliu is now known as josephl | ||
| === josephl is now known as josephliu | ||
| ogra_ | apw, seems to help | 14:53 |
| tgardner | ogasawara, rebooting gomeisa for kernel update | 14:57 |
| ogasawara | tgardner: ack | 14:57 |
| tgardner | lets hope it comes back this time. I ran update-initramfs by hand | 14:57 |
| ogasawara | hrm, am I really the only one in the mumble channel | 15:08 |
| apw | ogasawara, nope you arn't in mine | 15:08 |
| apw | cking, whats the story with your thinkpad fan investigation | 15:09 |
| tgardner | ogasawara, are really there even now ? | 15:10 |
| cking | apw, ongoing | 15:10 |
| apw | cking, james hunt is also seeing issues, he might be able to test | 15:10 |
| tgardner | ogasawara, we can hear you | 15:10 |
| ogasawara | tgardner, apw: we still looking to drop non-smp powerpc flavor? If so, I best craft an email asap and try to get it dropped in the next upload. | 15:20 |
| tgardner | ogasawara, yeah, please do | 15:20 |
| tgardner | bjf, did cobbler_web get hosed on your servers after a recent update ? | 15:30 |
| bjf | tgardner: no, i can update today and see if that changes | 15:30 |
| bjf | tgardner: fully updated, still cobblering | 15:39 |
| tgardner | bjf, hmm. | 15:39 |
| GrueMaster | ppisati: Did something change with the 3.2 omap kernel after Beta1 that would screw with the serial console? I am not getting much output after kernel boot, and it doesn't appear to start oem-config. | 15:42 |
| tgardner | bjf, do you remember where cobbler/orchestra keeps its local mirror setting ? | 15:45 |
| smb | tgardner, The ocherstra_proxy snippet? | 15:45 |
| tgardner | taht might be it | 15:45 |
| tgardner | smb, ok, where the hell is the proxy snippet? orchestra and cobbler have their bits splattered evrywhere. | 15:48 |
| smb | tgardner, I normally go there via web but I guess you said its broken right now | 15:49 |
| smb | give me a sec | 15:49 |
| tgardner | yeah, for some reason | 15:49 |
| smb | tgardner, Sorry took a bit. /var/lib/cobbler/snippets | 15:54 |
| GrueMaster | ppisati: The problem seems to appear with 3.2.0-18.28. The ubuntu-server 20120304/precise-preinstalled-server-armhf+omap.img.gz works, 20120305/precise-preinstalled-server-armhf+omap.img.gz fails. Only difference is the kernel according to the manifests. | 15:58 |
| GrueMaster | See http://paste.ubuntu.com/883454/ | 16:08 |
| tgardner | jsalisbury, henrix,jj: rebooting tangerine for kernel update | 16:08 |
| GrueMaster | http://paste.ubuntu.com/883465/ is a good boot log. | 16:08 |
| henrix | tgardner: ack | 16:08 |
| ppisati | GrueMaster: ok, let me check | 16:16 |
| ppisati | GrueMaster: so, what was the first kernel with this regression? | 16:16 |
| GrueMaster | Note that this appears to be serial console specific and omap specific. omap4 images work fine, as does the omap desktop images. | 16:17 |
| GrueMaster | 3.2.0-18.28 | 16:17 |
| GrueMaster | linux-image-3.2.0-18-omap | 16:17 |
| GrueMaster | here's the manifest diff. http://paste.ubuntu.com/883486/ | 16:18 |
| ppisati | GrueMaster: ah, so it's an omap3-only think | 16:19 |
| ppisati | *thing | 16:19 |
| ppisati | ok | 16:19 |
| GrueMaster | yes. | 16:19 |
| ppisati | ok, i | 16:19 |
| ppisati | ll take a look | 16:19 |
| jodh | hi - I mentioned to apw earlier that bug 751689 is still impacting me on precise with 3.2.0-18-generic-pae kernel. | 16:23 |
| ubot2` | Launchpad bug 751689 in linux "ThinkPads overheat due to slow fans when on 'auto'" [Critical,In progress] https://launchpad.net/bugs/751689 | 16:23 |
| apw | cking, ^^ | 16:23 |
| jodh | If I fire up say 4 kvm instances or try to build eglibc, the machine just performs a controlled shutdown. It's been freaking me out a little actually as I didn't realize it was this kernel issue re-occuring. | 16:23 |
| cking | apw, /me on the phone | 16:24 |
| jsalisbury | cking, I was able to overheat and crash my system. I'll send you the data. | 16:24 |
| jodh | This is on a Lenovo T410 thinkpad with Nvidia graphics. FTR, I have not touched the BIOS since I bought it (around October 2010). | 16:25 |
| jodh | dmidecode says BIOS version is "Version: 6IET72WW (1.32 )" dated 08/27/2010. | 16:25 |
| cking | jodh, let me send you an email with a description of a tool to try out to capture the failure mode | 16:35 |
| jodh | cking: that'd be great, thanks. | 16:43 |
| cking | jodh, once we've got some data I can see what's going on and feed this back to Lenovo | 16:49 |
| jodh | OK. It looks like the kernel emits an acpi netlink event when the less critical "hot" state is reached - should some other part of the system be handling that and popping up a window explaining to the user, or maybe trying to forcibly ramp fans, etc? | 16:50 |
| jodh | Having the system just die on you is kinda alarming ;-D | 16:50 |
| apw | jodh, wuss | 16:54 |
| jodh | maybe the kernel needs a OOT killer - shutdown the processes that are chewing too much power to try and avoid the forced system shutdown? | 16:58 |
| tseliot | apw, cking: any ideas as to why trying to build a drm module out of the tree would result in "make" ignoring the driver's makefile? I'm getting this "make -f scripts/Makefile.build obj=/var/lib/dkms/drm-dkms/3.2/build/staging/cdv" and then "(cat /dev/null; ) > /var/lib/dkms/drm-dkms/3.2/build/staging/cdv/modules.order" | 16:59 |
| tseliot | apw, cking: full output http://paste.ubuntu.com/883504/ | 17:00 |
| apw | tseliot, what is missing ? | 17:00 |
| tseliot | apw: cdv should build as the gma500 module does (they're both in staging and enabled in the config file) | 17:01 |
| apw | can we see the Makefile | 17:02 |
| tseliot | apw: my staging/Makefile: http://pastebin.ubuntu.com/883554/ | 17:03 |
| apw | and in the cdv directory ? | 17:04 |
| tseliot | apw: the main Makefile: http://pastebin.ubuntu.com/883557/ | 17:04 |
| tseliot | apw: staging/cdv/Makefile: http://paste.ubuntu.com/883561/ | 17:05 |
| tseliot | apw: ignore the default target, I get the same error without it | 17:05 |
| apw | obj-$(CONFIG_DRM_INTEL_CDV) += cedarview_gfx.o | 17:06 |
| apw | try making that line obj-y += | 17:06 |
| tseliot | ok | 17:06 |
| apw | you may not have that config set in the third level | 17:06 |
| apw | as they normally come in via the config, and you are cheating i think | 17:07 |
| tseliot | apw: same result. And it's a bit weird to see that I can enable gma500 in my Makefile but it won't work with cdv | 17:09 |
| cking | jodh, I suspect we should handle that event and crank the fan up to full tilt | 17:09 |
| apw | tseliot, can i have the fma500 Makefile too pls | 17:09 |
| apw | cking, there is a framework for acpi events to shell commands which could do it, but its not very reliable compared to doing it in the kernel | 17:10 |
| tseliot | apw: staging/gma500/Makefile: http://paste.ubuntu.com/883572/ | 17:10 |
| tseliot | apw: and I'm building with "sudo make KERNELRELEASE=3.2.2-030202-generic all KVER=3.2.2-030202-generic V=1" | 17:12 |
| apw | and what does this do on its own: | 17:14 |
| apw | make -f scripts/Makefile.build obj=/var/lib/dkms/drm-dkms/3.2/build/staging/cdv | 17:14 |
| apw | and indeed this: | 17:15 |
| apw | make -f scripts/Makefile.build obj=/var/lib/dkms/drm-dkms/3.2/build/staging/gma500 | 17:15 |
| apw | tseliot, ^^ | 17:15 |
| tseliot | apw: make: scripts/Makefile.build: No such file or directory | 17:15 |
| apw | oh you'd need to run those in /lib/modules/`uname -r`/source | 17:15 |
| apw | or is it /build | 17:15 |
| tseliot | apw: right, /lib/modules/3.2.2-030202-generic/build/scripts/ | 17:16 |
| apw | tseliot, ok my next debug attempt would be to mv cvd cvd-keep; cp -rp gma500 cdv | 17:20 |
| apw | and see if it builds | 17:20 |
| * henrix will be out for ~20min | 17:20 | |
| tseliot | apw: sudo make -f scripts/Makefile.build obj=/var/lib/dkms/drm-dkms/3.2/build/staging/gma500 doesn't seem to do anything (I created a link to the path to scripts) | 17:20 |
| tseliot | ok | 17:21 |
| apw | tseliot, perhaps :%s/psb_gfx/psb_gfx2 | 17:21 |
| apw | so you get two different modules | 17:21 |
| tseliot | apw: ah, ok | 17:21 |
| apw | tseliot, that should tell us if its the subdir that the issue, or the staging levle | 17:22 |
| tseliot | right | 17:24 |
| tseliot | apw: yes, the module was built in ./staging/cdv/psb_gfx2.ko | 17:26 |
| tseliot | so I guess it's just the subdir | 17:26 |
| apw | ok so its something to do with the makefile in cvd then | 17:26 |
| apw | so i'd start deleting bits of it now | 17:26 |
| apw | you got rid of default: already yes | 17:27 |
| tseliot | apw: yes | 17:27 |
| apw | tseliot, check the filenames are sensible ... i am a little worrried that the ENVDIR etc might need $(src)/ on the front | 17:28 |
| apw | i am unsure what would happen if none of the filenames point anywhere in ceadaview_gfx-y ... that might do nothing | 17:29 |
| tseliot | apw: good point, let me try | 17:30 |
| === yofel_ is now known as yofel | ||
| tseliot | apw: nothing seems to have changed and I added a staging/cdv prefix to those variables: http://paste.ubuntu.com/883613/ | 17:34 |
| tseliot | I could've used $(INCDIR) but it shouldn't change much | 17:35 |
| jodh | cking: agreed. I've raised bug 955287. Am running tp-thermstat now... | 17:36 |
| ubot2` | Launchpad bug 955287 in acpi-support "Ubuntu should handle "hot" CPUs by taking preemptive action and warning users" [Undecided,New] https://launchpad.net/bugs/955287 | 17:36 |
| cking | jodh, ack, let me see what else we can do in the driver too | 17:36 |
| === tgardner is now known as tgardner-lunch | ||
| * ppisati -> goes out to get some stuff | 17:41 | |
| tseliot | apw: thanks for your help | 17:51 |
| apw | tseliot, did you win ? | 17:51 |
| apw | ccccccbgbcivibcrdlbefnuhnilivbfndklfhuigdbvd | 17:51 |
| tseliot | apw: no, but you tried to help and at least now I can exclude one case | 17:52 |
| ogra_ | apw, manually generating md5sums ? | 17:52 |
| tseliot | :D | 17:52 |
| apw | ogra_, heh no playing with a password generator | 17:52 |
| ogra_ | use a cat on the kbd ... gets you the safest PWs in the world ;) | 17:53 |
| apw | and unity is helping out by letting my output go to the most annoying window | 17:53 |
| tseliot | hehe | 17:53 |
| tseliot | apw: progress!!! I had to change the last line to obj-m | 18:05 |
| apw | tseliot, erp | 18:07 |
| apw | oh you wanted a _module_ of course | 18:07 |
| apw | so you want the FOOO=y -> FOOO=m | 18:07 |
| apw | the CONFIG_xxx_CDV=m | 18:07 |
| tseliot | apw: yes, I don't know why I haven't thought about it before | 18:07 |
| apw | tseliot, cause it is tooo obvious | 18:08 |
| tseliot | apw: good point :D | 18:08 |
| === tgardner-lunch is now known as tgardner | ||
| ogasawara | tgardner: I'm out tomorrow and friday, but I'll still plan to upload sometime friday. Andy's going to cover the release meeting for me. Could you however keep an eye on that powerpc thread for me. | 18:28 |
| tgardner | ogasawara, can do. | 18:33 |
| * ogasawara lunch | 19:22 | |
| * cking --> EOD | 19:40 | |
| * tgardner -> EOD | 20:29 | |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!