/srv/irclogs.ubuntu.com/2012/01/30/#ubuntu-kernel.txt

shawn1Hello01:55
shawn1My kernel doesn't seem to be allowing me to use a multi-gpu card (GTX295).  It will only detect one core.  I checked lspci and that should show two entries for the GTX 295, but it only shows one.  I did lspci -A intel-conf2, though, and found "00:08.0 Non-VGA unclassified device: Device aebc:75c5 (rev 40)".  That item didn't show up on any of the other lspci detection lists.01:57
ohsixdo you get any messages about pci routing in dmesg? there's a few methods you can try if the bios doesn't set things up right02:00
unixpro1970Shawn2, some gtx295 cards are 1 core while others are 2 core.06:54
unixpro1970The single core gtx295 came later06:54
=== smb` is now known as smb
pip!ciao10:38
ubot2`Factoid 'ciao' not found10:38
pip!list10:38
ubot2`This is not a file sharing channel (or network); be sure to read the channel topic. If you're looking for information about me, type « /msg ubottu !bot ». If you're looking for a channel, see « /msg ubottu !alis ».10:38
abdoHello guys10:58
abdoone of my servers had a kernel panic10:58
abdoI can collect "linux-image-2.6.35-26.0.crash.gz" file from /var/crash10:59
abdoI tried to use:10:59
abdogdb vmlinuz-2.6.35-26-server linux-image-2.6.35-26.0.crash11:01
abdoand got:11:01
abdo/tmp/linux-image-2.6.35-26.0.crash" is not a core dump: File format not recognized11:01
abdocan any one help me and put me on the right path? thanks in advance.11:01
* ppisati -> starvation avoidance12:32
rbasakI'm trying to cross compile an armhf kernel out of the git source for bisecting purposes. Is this easiest way? I found instructions for using xdeb but they break when when xdeb 404s on http://ports.ubuntu.com/ubuntu-ports/pool/main/z/zlib/lib32z1_1.2.3.4.dfsg-3ubuntu3_armel.deb. And the instructions at https://wiki.ubuntu.com/Kernel/BuildYourOwnKernel don't cover cross compiling. Some help, please?13:00
ppisatirbasak: export $(dpkg-architecture -aarmel); export CROSS_COMPILE=arm-linux-gnueabi-13:01
ppisatirbasak: fakeroot debian/rules clean binary-omap413:01
ppisatirbasak: to get an armhf kernel, substitute "-aarmel" with "-aarmhf"13:01
ppisatirbasak: yep, i should update those instructions13:02
rbasakawesome, thanks!13:02
rbasakI have an automated test rig now. I should be able to do an automatic bisection13:02
ppisatirbasak: btw, what are you looking for with your bisection?13:03
rbasakright now, the random panics. but I also wanted to have an arrangement ready to do it in the future13:03
ppisatirbasak: random panic???13:04
rbasakthe installer fails randomly right now13:04
rbasakI had some logs saved from friday, hang on13:04
rbasakppisati: eg. http://paste.ubuntu.com/819098/13:05
ppisatirbasak: well, that's with the previous kernel13:07
ppisatirbasak: but the panic looks strange'13:08
rbasakppisati: ah yes - I take it the installer hasn't been updated with the latest kernel yet13:08
ppisatirbasak: tell me if it's reproducible with the new one13:09
rbasakppisati: I'd love a script to give me a working uImage and uInitrd out of the latest kernel. How would I do that?13:12
rbasakI'm not yet very familiar with the tooling around the kernel builds13:12
ppisatirbasak: uImage is definitely inside the .deb, so you can get it easily13:14
ppisatirbasak: while initrd IIRC is built at install time13:14
rbasakwhat about the installer initrd? that's the one I'm after13:14
ppisatirbasak: so probably the easiest way is to insall the kernel somewhere and get the two files13:14
rbasakIf the uImage is in there I can extract that part, thanks13:14
rbasakIs the installer uInitrd the same as an installed machine's uInitrd?13:14
rbasakThe one you gave me for the test the other day didn't work13:15
ppisatirbasak: don't know, you should ask the #arm guys (ogra's angels)13:15
rbasakOK I'll take it to #ubuntu-arm13:15
abdoCan any one help me on my question?13:29
rbasakppisati: make: *** No rule to make target `binary-omap4'.  Stop.13:37
rbasakppisati: this is debian/rules out of ubuntu-precise.git, right?13:37
ppisatirbasak: did you checkout the omap4 branch first? and did you clean?13:42
rbasakah, there's an omap4 branch? :-)13:43
ppisatirbasak: git checkout ti-omap413:43
rbasakppisati: now working, thanks :)13:43
ppisatirbasak: tell me how it goes13:44
rbasakwill do!13:44
ppisatia bug in Incoplete status will vanish after 60days of no activity, right?14:12
ppisati*Incomplete14:12
=== bladernr_afk is now known as bladernr_
hertonppisati, for Ubuntu I think that's correct, there are more details: https://help.launchpad.net/Bugs/Expiry (assuming this is uptodate)14:30
=== tgardner is now known as tgardner-afk
ppisatiwhy bug 732628 is still showing up in linux-ti-omap4?15:27
ubot2`Launchpad bug 732628 in linux-ti-omap4 "TOCTOU in mount.ecryptfs_private" [Low,Fix committed] https://launchpad.net/bugs/73262815:27
* ogasawara back in 2015:31
jsalisburyogasawara, A possible regression with ath9k devices in 3.2.0-12: bug 923512 and 92360516:05
ubot2`Launchpad bug 923512 in linux "ath9k wireless stopped working after kernel upgrade to 3.2.0-12" [High,Incomplete] https://launchpad.net/bugs/92351216:05
ubot2`Launchpad bug 923605 in linux "Wireless network card unable to connect to router with 3.2.0-12 kernel (dup-of: 923512)" [Undecided,Incomplete] https://launchpad.net/bugs/92360516:05
=== bladernr_ is now known as bladernr_afk
=== yofel_ is now known as yofel
jsalisburyogasawara, I can bisect between the two kernels to identify the bad commit.16:09
ogasawarajsalisbury: ok thanks.  the only thing that would come to mind is something in the 3.2.2 stable update...16:12
ogasawarajsalisbury: everything else would seem unrelated16:12
=== bladernr_afk is now known as bladernr_
ogasawarajsalisbury: you'll want to find out specifically what version them mean when they say "working fine with the previous kernel version". do they mean 3.2.0-11.19 or an Oneiric kernel or ...16:15
jsalisburyogasawara, ahh, right.  thanks.16:23
smbtgardner-afk, Not sure whether someone fixed something under our back or it makes a difference between bare metal and kvm-vm or is just random, but right now an cobbler install of precise-amd64 did work (with the usual whining about unloadable modules)17:00
=== CharlieMike is now known as ayan
ogasawaradiwic: is there a specific tag you want us to use to flag audio bugs which are likely due to the the new jack detection patches?  or do you want us to just ping you with the bug #'s?17:15
ogasawaraeg bug 92331617:15
ubot2`Launchpad bug 923316 in linux "[regression] Speakers are not muted while plugin the headphone " [Undecided,Confirmed] https://launchpad.net/bugs/92331617:15
ogasawarabug 92340917:16
ubot2`Launchpad bug 923409 in linux "No sound" [Undecided,Confirmed] https://launchpad.net/bugs/92340917:16
ogasawarajsalisbury: ^^ just fyi, you'll want to keep an eye out for any audio regressions as of 3.2.0-11.19 due to the new jack detection patches17:19
=== tgardner-afk is now known as tgardner
* ppisati disappears for a bit...17:35
diwicogasawara, ouch :-( 17:51
diwicogasawara, I guess pinging me or assigning bugs to me will be best, thanks17:52
ogasawaradiwic: ack, will do17:53
=== chrisccoulson is now known as chrisccoulson2
jsalisburyogasawara, I'll keep an eye out, thanks18:01
=== chrisccoulson2 is now known as chrisccoulson
vanhoofogasawara: been otp for a bit, but tgardner let me know you found the prob already w/ cw-3.2, happy to test things out here for ya18:38
ogasawaravanhoof: perfect, I'm just getting the bug filed etc.  Will get a patch thrown together and have you test.18:39
vanhoofogasawara: for reference: http://paste.ubuntu.com/822889/18:39
* tgardner -> lunch18:56
=== arges is now known as arges_
jwijsalisbury: i think bug 923094 is just about shipping that specific module in /debian.master/d-i20:00
ubot2`Launchpad bug 923094 in linux "Debian Installer kernel (PowerPC) is missing therm_adt746x module" [Medium,Incomplete] https://launchpad.net/bugs/92309420:00
ogasawaravanhoof: bug 923900, http://people.canonical.com/~ogasawara/lp923900/20:15
ubot2`Launchpad bug 923900 in linux-backports-modules-3.0.0 "Enable iwlwifi drivers for compat-wireless v3.2 backports" [Medium,Triaged] https://launchpad.net/bugs/92390020:15
vanhoofogasawara: thanks20:16
vanhoofwill give it a go20:16
jsalisburyjwi, right20:21
vanhoofogasawara: looks good, gonna do some testing now w/ new(er) cards20:27
vanhoofogasawara: http://paste.ubuntu.com/823004/20:27
vanhoofogasawara: thank you :D20:27
=== arges_ is now known as arges
smagounA question on kernel ABI numbers. The ABI numbers need to be unique within a particular topic branch, but not globally - correct? So for example if you have a topic branch foo with ABI base 3.2-100, and another topic branch bar with ABI base 3.2-100, would there be package collisions if packages from each branch were in the same archive? (Not that this is good form)21:39
tgardnersmagoun, they need to be unique within an archive since all of the binary packages share the same directory.21:41
smagountgardner: So we'd potentially wind up with linux-image-3.2.0-100-foo and linux-image-3.2.0-100-bar? Which isn't useful if you're trying to keep the two separate. Is that correct?21:45
=== bladernr_ is now known as bladernr_afk
tgardnersmagoun, it might be confusing if the binaries are unrelated ('cause its the form we have right now, e.g., linux-image-3.2.12-virtual and linux-image-3.2.12-generic)21:48
tgardnernote that I didn't get those example version numbers quite correct. it should have been linux-image-3.2.0-12-generic, etc21:50
smagountgardner: makes sense. Is there a 'safe' ABI base that thirdparties can adopt as an ABI base if they want to avoid collisions with Ubuntu kernels for the forseeable future (say through 14.04)? -2000, for example?21:50
tgardnersmagoun, yeah, anything 3 digits with a unique flavour name is likely gonna be OK21:51
smagountgardner: thanks. Is 4 digits off-limits, or just more typing than 3 digits and therefore less desirable?21:52
tgardnersmagoun, checkout how we've named the ti-omap4 stuff: https://launchpad.net/ubuntu/+source/linux-ti-omap4. 4 digits is fine (I mis-spoke)21:52
* smagoun looks21:53
smagountgardner: Is there a specific reason for the jump in ABI between 3.0 and 3.2? I see 3.0.0-1205.10 in 11.10, and 3.2.0-1405.7 in 12.0421:53
tgardnersmagoun, since the kernel version is embedded in the package name there really wasn't a need to also rev the ABI series, but it might reduce confusion. I dunno... 21:55
smagoununderstood, thanks21:56
tgardnerso, no, I don't think there was a specific reason21:56
* tgardner is EOD21:56
hallynhm, is bug 923900 why my laptop lost 23:03
ubot2`Launchpad bug 923900 in linux-backports-modules-3.0.0 "Enable iwlwifi drivers for compat-wireless v3.2 backports" [Medium,Triaged] https://launchpad.net/bugs/92390023:03
hallynwireless with latest update?23:03
hallynprobably not...~.23:04

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!