| bullgard4 | [Symmetrical Multiprocessing, SMP]: Is there a tool that shows if an application program uses both processors on a particular two-processor SMP machine? | 03:58 |
|---|---|---|
| lifeless | bullgard4: do you mean 'does it have threads' ? | 04:07 |
| bullgard4 | lifeless: No. I imagine that an application may either be processed by only one processor or by both. | 04:10 |
| stgraber | no, if your application has a single process it can't be using two CPUs | 04:11 |
| lifeless | stgraber: it can get scheduled to both, just not concurrently :P | 04:11 |
| stgraber | lifeless: right :) | 04:12 |
| lifeless | bullgard4: your question is essentially ill-defined in kernel terms, heck, even application isn't that well defined (I know single apps that have 10's of different images with 100's of processes active concurrently) | 04:12 |
| lifeless | bullgard4: I'm sure you are asking this as part of answering some larger issue/project - perhaps some context about that will help us answer you | 04:13 |
| bullgard4 | May be referring to a more specific case will help me better to understand. Does pbzip2 load the two processor kernels while gzip, bzip2 do load only one? | 04:17 |
| lifeless | bullgard4: pbzip2 is probably threaded, so the kernel can schedule it onto multiple CPU's at once | 04:19 |
| lifeless | gzip and bzip2 are single threaded and can only be scheduled on a single cpu at a time | 04:20 |
| bullgard4 | lifeless, stgraber: Thank you very much for your help. | 04:23 |
| === smb_tp_ is now known as smb_tp | ||
| Kano | hi, why does 2.6.28-8 depend on wireless-crda which is only in universe? | 10:00 |
| Kano | remove that depend please... | 10:05 |
| ivoks | is there a change hardy's latest kernel was compiled with an older version of gcc that we have in repository? | 10:05 |
| ivoks | which would make building kernel modules impossible? | 10:05 |
| persia | ivoks, Older version, no. Older revision, yes. It was built with the gcc that was available at the time it was built. | 10:18 |
| ivoks | persia: i'm aware of that... | 10:19 |
| ivoks | but, we can't build kernel modules if kernel doesn't get rebuild | 10:19 |
| ivoks | gcc in hardy is 4.2.4 | 10:20 |
| ivoks | it was updates couple of months ago | 10:20 |
| ivoks | yet, the kernel is compiled with 4.2.3 | 10:20 |
| persia | Hrm? | 10:20 |
| * persia looks | 10:20 | |
| Kano | depends, you can even build the kernel with gcc 4.1.1 and then compile modules with 4.3 | 10:20 |
| Kano | just not the other way around | 10:20 |
| persia | ivoks, Which arch? | 10:20 |
| ivoks | persia: i386 | 10:20 |
| Kano | but then the linux-headers are not installable | 10:21 |
| persia | ivoks, For some reason I can't find the build log for 23.48, apparently in part because it's a security update. Sorry not to be able to help. | 10:26 |
| ivoks | persia: right, that was security update | 10:26 |
| ivoks | persia: bug 292381 | 10:29 |
| ubot3 | Malone bug 292381 in linux "[hardy] kernel 2.6.24-21-generic and gcc version mismatch" [High,Triaged] https://launchpad.net/bugs/292381 | 10:29 |
| persia | ivoks, It may well have been built against 4.2.3 then, as I believe -security only has 4.2.3 and -updates has 4.2.4. | 10:29 |
| persia | ivoks, Trick being that for people only tracking -security, I think 4.2.3 is correct. | 10:29 |
| ivoks | i understand that | 10:29 |
| ivoks | that's a gcc bug | 10:29 |
| ivoks | gcc should never get version change during release | 10:30 |
| persia | Or at least never get any sort of ABI change. | 10:30 |
| persia | The version string isn't the important part. | 10:30 |
| ivoks | well | 10:36 |
| ivoks | i just did more digging | 10:36 |
| ivoks | kernel's 64bit buildd doesn't have gcc 4.2.4 | 10:37 |
| ivoks | while 32bit buildd does | 10:37 |
| persia | ivoks, Found it: http://launchpadlibrarian.net/21747228/buildlog_ubuntu-hardy-i386.linux_2.6.24-23.48_FULLYBUILT.txt.gz shows gcc-4.2_4.2.3-2ubuntu7 : it's entirely due to skew between -updates and -security. | 10:47 |
| persia | Either needs two different kernels (one for -updates and one for -security) or to not have impactful variation. | 10:48 |
| ivoks | correct | 10:48 |
| lukehasnoname | I would be interested in someone could solve this problem: | 12:36 |
| lukehasnoname | http://www.nabble.com/Yaird-throws-error-about-bad-sr0-device-link-td21449825.html | 12:36 |
| lukehasnoname | I know it's not Ubuntu-kernel related, but perhaps it's a slow morning for someone who wants to help me out | 12:37 |
| soren | Why use yaird? What does it do that mkinitramfs doesn't? | 12:41 |
| lukehasnoname | >_> I'm using a video tutorial which stated using that tool. I'm not familiar with kernel building. | 13:00 |
| lukehasnoname | I googled the error I was getting and found that ml archive | 13:00 |
| alex_joni | lukehasnoname: googling I found some people having to update to a newer yaird version (with erros similar to yours), but I would do as soren suggest.. use mkinitramfs | 13:05 |
| lukehasnoname | I'll try it later. I've been awake 24 hours in studying for tests and finishing up a project proposal | 13:19 |
| lukehasnoname | thanks | 13:19 |
| ivoks | hi... are there any plans to support more than 8 cpus in hardy? | 13:54 |
| rtg | ivoks: debian/config/amd64/config.server:CONFIG_NR_CPUS=64 | 13:55 |
| ivoks | ah, right | 13:55 |
| ivoks | so, only i386 supports 8 | 13:56 |
| rtg | correct | 13:56 |
| ivoks | thanks | 13:56 |
| persia | rtg, WIth the lpia kernel merge, does linux-lpia now have the same modules and udebs as i386/amd64 or are there significant differences? | 13:56 |
| rtg | persia: udebs should be pretty close, but I'm not sure about all of the modules. The configs came over intact, so that likely hasn't changed. | 13:57 |
| persia | Thanks. I'm thinking some bugs I was thinking of filing may have become obsolete, but I'll retest, and maybe file. | 13:58 |
| rtg | persia: who is the archive admin today? I need to get wireless-crda promoted to main | 13:59 |
| persia | rtg, You already have an approved MIR? | 14:00 |
| rtg | yep, hang on. | 14:00 |
| persia | https://wiki.ubuntu.com/ArchiveAdministration#Archive%20days says nobody is the offical archive-admin on Fridays. | 14:00 |
| rtg | persia: which is why I asked | 14:01 |
| rtg | https://bugs.launchpad.net/ubuntu/+bug/325801 | 14:01 |
| ubot3 | Malone bug 325801 in ubuntu "Main inclusion request: wireless-crda" [Undecided,In progress] | 14:01 |
| persia | So just beg in #u-d and any of the team might be able to help (including the few who don't have days). | 14:01 |
| persia | Theoretically anyone in https://launchpad.net/~ubuntu-archive/+members ought to be able to do it. | 14:03 |
| rtg | persia: thanks, I'll see whom I can annoy. | 14:03 |
| rtg | actually, I thought you were on that list. | 14:04 |
| persia | No. I signed up for enough of those "it sounded like a good idea at the time, but it's a bunch of work" things already :) | 14:05 |
| rtg | I hear yuo :) | 14:05 |
| === thunderstruck is now known as gnomefreak | ||
| Kano | rtg: you forgot my patch, you only need to apply it then you get a quickcam.ko as you still have it enabled in the config | 15:25 |
| rtg | I did not forget, I just didn't do it yet. | 15:26 |
| Kano | also i had to revert one config | 15:26 |
| Kano | as you depend on something now which is only in jaunty universe | 15:27 |
| Kano | wireless-crda is not even in main... | 15:28 |
| Kano | of course i used =y | 15:28 |
| Kano | and removed that depend | 15:29 |
| === Kano is now known as Kano32 | ||
| === Kano32 is now known as Kano | ||
| === Nicke_ is now known as Nicke | ||
| === wgrant_ is now known as wgrant | ||
| BUGabundo | hi | 23:24 |
| BUGabundo | is any kernel dev awake around here? | 23:24 |
| BUGabundo | if it wasn't to much trouble | 23:25 |
| BUGabundo | could some one take a look at https://bugs.launchpad.net/ubuntu/+source/linux/+bug/329254 | 23:25 |
| ubot3 | Malone bug 329254 in linux "kernel: [ 341.030356] Pid: 0, comm: swapper Tainted: P W 2.6.28-7-generic #20-Ubuntu" [Undecided,New] | 23:25 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!