| * m4t still investigating this kernel issue on maverick | 00:19 | |
| m4t | even without kernel-package, using a standard 'make' and booting the bzImage with grub results in an identical hang | 00:19 |
|---|---|---|
| m4t | FAIL: gcc.c-torture/compile/limits-blockid.c -O0 (internal compiler error) | 01:57 |
| m4t | that's not it :( | 02:02 |
| m4t | it hangs in qemu too :-] | 03:38 |
| m4t | 2.6.36 compiled with lucid toolchain bzImage boots to kernel panic w/no rootfs | 03:39 |
| m4t | the one compiled with maverick toolchain freezes at rtc init | 03:40 |
| m4t | well, a bit more to it than that | 04:39 |
| m4t | i'd like to send over a .config using CONFIG_MPENTIUMM=y and see if someone else can't reproduce this hang in qemu on their system | 04:40 |
| === yofel_ is now known as yofel | ||
| jovaro | Hi, I use a Via Epia board with TV-out and after updating from 9.10 to 10.04 the image flickers like crazy, before starting X even | 15:38 |
| jovaro | I asked in #openchrome and they said it was a kernel issue | 15:38 |
| jovaro | so is there someone here that can help perhaps? | 15:38 |
| jovaro | the flickering starts right after the fsck's when booting | 15:39 |
| jovaro | ok nevermind, it appears that the viafb just doesn't work anymore. I now use the uvesafb and that works much better | 16:02 |
| === emma_ is now known as emma | ||
| CarlFK | wget http://archive.ubuntu.com/ubuntu/dists/maverick/main/installer-amd64/current/images/netboot/ubuntu-installer/amd64/linux | 18:12 |
| CarlFK | sudo kexec --load linux | 18:12 |
| CarlFK | ELF rel parse failed | 18:12 |
| CarlFK | um, wut? | 18:12 |
| === BenC__ is now known as BenC | ||
| alex88 | hi guys, recompiling a server using only necessary modules etc how can help the speed of the pc? i mean, is it useful? like boot time and other things? | 22:40 |
| CarlFK | alex88: probably not. lots of effort has gone into speeding things up so small chance that you will find anything significant. | 23:08 |
| alex88 | CarlFK, thank you, so it's much more a waste of time :) | 23:28 |
| CarlFK | alex88: if you want to invest time in something like that, look into the upstart stuff. it is new enough that there is probably lots of productive work needing to be done | 23:31 |
| CarlFK | Here is what just bit me: https://bugs.launchpad.net/ubuntu/+source/squid-deb-proxy/+bug/666014 | 23:34 |
| ubot2 | Launchpad bug 666014 in squid-deb-proxy (Ubuntu) "Avahi service for squid-deb-proxy does not start (affects: 3) (heat: 262)" [Undecided,New] | 23:34 |
| alex88 | you mean helping in bug solving? | 23:37 |
| CarlFK | alex88: basically. I have a feeling digging into that bug will reveal some other bugs, perhaps in the upstart framework or whatever it's called | 23:39 |
| alex88 | well about kernel i was thinking about configuring removing modules and then recompiling, not modifying source | 23:40 |
| alex88 | bytheway i'll spend time on bugs.. | 23:41 |
| CarlFK | alex88: assuming you did shave some boot time, you would then have to watch for kernel updates and port your stuff. yuck. | 23:42 |
| alex88 | well, just needed to recompile new kernel with same config and check if something is changed | 23:43 |
| CarlFK | pretty sure everything is a module that can be, and having them on disk doesn't slow things down | 23:43 |
| CarlFK | building them into the kernel would speed things up by .0001 second. and change the order things are loaded in, so maybe have side effects. yuck :) | 23:44 |
| alex88 | oh got it, i think that time is better on bugs... | 23:46 |
| alex88 | bytheway, just a questions, modules are all loaded by default or just when needed to a device? | 23:46 |
| CarlFK | juser@dhcp232:/etc/gdm$ cat /etc/modules | 23:47 |
| CarlFK | # /etc/modules: kernel modules to load at boot time. | 23:47 |
| CarlFK | # This file contains the names of kernel modules that should be loaded | 23:47 |
| CarlFK | # at boot time, one per line. Lines beginning with "#" are ignored. | 23:47 |
| CarlFK | loop | 23:47 |
| CarlFK | lp | 23:47 |
| CarlFK | rtc | 23:47 |
| CarlFK | and then something (name escapes me) will load stuff as needed | 23:47 |
| CarlFK | so no, not all are loaded. | 23:49 |
| CarlFK | lsmod will show what is currently loaded. my box has 72. | 23:50 |
| CarlFK | desktop box | 23:50 |
| CarlFK | server box has 17 | 23:50 |
| alex88 | ok that's all..thank you for your time :) | 23:52 |
| CarlFK | https://launchpad.net/hotplug | 23:53 |
| CarlFK | that's the something that loads modules | 23:53 |
| CarlFK | i think... | 23:53 |
| alex88 | think that too | 23:55 |
| JanC | I think most modules are loaded by the kernel itself or by udev rules, certainly hotplug isn't used anymore (unless you have a *very* old Ubuntu version) | 23:56 |
| alex88 | but some time ago i've seen that in config you can choose the generic cpu or the p4, core2duo, what that affect? | 23:59 |
| CarlFK | udev - that's what I was thinking | 23:59 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!