You are not logged in.
I've got a more or less new Arch system installed with more or less only the "basic" packages installed (base+internet stuff+gnome) at the moment. After I rebooted when the latest kernel update "kernel26-2.6.23.9-1" was installed I'm stuck at the bootup screen for looking at "Loading UDev uevent" staying in [BUSY] mode for over a minute before the system goes ahead and loads further. What can cause this "time-out"/waiting? And why now with just a ".8-1 to .9-1" update?
Linux archbox 2.6.23-ARCH #1 SMP PREEMPT Mon Nov 26 21:15:02 UTC 2007 i686 Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz GenuineIntel GNU/Linux
/etc/rc.conf:
LOCALE="en_US.utf8"
HARDWARECLOCK="localtime"
TIMEZONE="Europe/Oslo"
KEYMAP="no-latin1"
CONSOLEFONT="default8x16"
CONSOLEMAP=
USECOLOR="yes"
MOD_AUTOLOAD="yes"
MOD_BLACKLIST=()
MODULES=(atl1 mii slhc snd-mixer-oss snd-pcm-oss snd-hwdep snd-page-alloc snd-pcm snd-timer snd snd-hda-intel soundcore fuse uinput)
USELVM="no"
HOSTNAME="armor.joffer.net"
eth0="dhcp"
INTERFACES=(eth0)
gateway="default gw 192.168.0.1"
ROUTES=(!gateway)
DAEMONS=(syslog-ng network netfs crond dbus hal fam alsa sshd gpm avahi-daemon)
/etc/mkinitcpio.conf:
MODULES="pata_jmicron ata_generic ahci ata_piix"
BINARIES=""
FILES=""
HOOKS="base udev autodetect pata scsi sata usbinput keymap resume filesystems"
Offline
Can confirm this, my udev-uevents takes about 50 seconds, what is really annoying! My mkinitcpio.conf looks like this:
MODULES="pata_via ata_generic"
BINARIES=""
FILES=""
HOOKS="base udev autodetect pata keymap uresume filesystems"
now with 80% more sax-appeal!
"I hacked the Phrak, and all I got was this lousy signature"
Offline
This happens to me as well
Offline
Please file a bug, http://bugs.archlinux.org . Most developers do not actively follow these forums, filing a bug will ensure that the problem is forwarded to the correct people.
Offline
Will do. I'll edit this post when it's done.
Update:
Task # 8878
Last edited by Joffer (2007-12-09 13:05:08)
Offline
Just confirming that I also have a slow UDev loadtime.
Offline
I don't think that the arch devs can helps you because the change in the kernel package looks minimal so the reason for this can only found in the kernel changes. Sorry, my systems works as before. But there is one thing if i look at your configs; It seems that you have a similar pc as mine (MSI 975X Powerup Edition, SATA Harddisk, IDE CDROM) but my working mkinitcpio.conf is shorter than yours:
MODULES="ahci ext2 jfs fan"
BINARIES=""
FILES=""
HOOKS="base udev sata"
Perhaps it will helps if you copy your mkinitcpio.conf (the wiki about it is excellent) to mkinitcpio-test.conf and play a little bit with the parameters. After this you can create your own /boot/kernel26-test.img and boot it if you press "e" at the grub prompt (i hope you use grup). I'm afraid to say that in the worst case you have to start with a minimal rc.conf/mkinitcpio.conf to find out what is the reason for it, so sorry again that i have no exactly hint and good luck.
Offline
I don't think that the arch devs can helps you because the change in the kernel package looks minimal so the reason for this can only found in the kernel changes. Sorry, my systems works as before. But there is one thing if i look at your configs; It seems that you have a similar pc as mine (MSI 975X Powerup Edition, SATA Harddisk, IDE CDROM) but my working mkinitcpio.conf is shorter than yours:
MODULES="ahci ext2 jfs fan" BINARIES="" FILES="" HOOKS="base udev sata"
Perhaps it will helps if you copy your mkinitcpio.conf (the wiki about it is excellent) to mkinitcpio-test.conf and play a little bit with the parameters. After this you can create your own /boot/kernel26-test.img and boot it if you press "e" at the grub prompt (i hope you use grup). I'm afraid to say that in the worst case you have to start with a minimal rc.conf/mkinitcpio.conf to find out what is the reason for it, so sorry again that i have no exactly hint and good luck.
There are some discussion on in the bug post, so something may happen. Anyway, i find it strange that is just suddenly behaves this way. This was a fresh install and worked fine for several boots, and one morning the kernel was update so after a pacman -Syu and reboot i got this condition. I don't boot that often, so the problem is only the rare occation I'm in a hurry and my computer is off. But still, there has to be an explanation.. we'll see.
Flying Saxman has a shorter mkinitcpio.conf than me, almost like yours, and also experience slowdowns while booting.
Offline
One thing I just remembered, since it was updated once more just now (or at least for my system) is that I ran 'hwd -u' after the hwd package was updated. Not sure if it has something to do with my/our problem though...
[root@matrix hwdata]# hwd -u
Testing: kernel (2.6.23-ARCH) supports uevents
Update pci-, usb.ids, and xorgtable.
Are you connected and ready to update(y/n)? [n] y
...
(files are downloaded and replaced)
Hmm.. Looks like it's some testing?? "Testing:" is in bold and color...
Update: I copied back the original files from before my first ran the command and booted. Did not help Still about a minute of waiting
Last edited by Joffer (2007-12-10 20:40:59)
Offline
If you be in hurry forgot my suggestions because for this you need time and patience.:) On the other side more optimized config files could be even a good idea. And yes, it is very strange that it happens to you and Saxman from one moment to the next but there must be a way to avoid it because here all is fine. Good luck again.
Offline
Hmm.. Looks like it's some testing?? "Testing:" is in bold and color...
Only checking if kernel is newer than 2.6.15. Earlier versions don't support uevents. Hwd uses the uevents to read hardwares' module names.
Markku
Offline
Joffer wrote:Hmm.. Looks like it's some testing?? "Testing:" is in bold and color...
Only checking if kernel is newer than 2.6.15. Earlier versions don't support uevents. Hwd uses the uevents to read hardwares' module names.
Yeah, I was told hwd didn't have anything to do with uevents/udev. I just had to post just in case, since I did not know.
Offline
I know that this thread is a little bit old, but it describes my problem perfectly. With an kernel/udev (I don't know..) update from testing I got this problem too. Before this, my arch had a boottime from around 30 seconds, with this bug it takes 30 seconds only to load the udev uevents. I made a bootchart to prove my problem and yes: it is udev. Here the link to the bootchart:
Offline
I know that this thread is a little bit old
but (for me) the problem still exists...
now with 80% more sax-appeal!
"I hacked the Phrak, and all I got was this lousy signature"
Offline
It may be related to the loop in the uevent function and/or module-loading scripts. I only face this with compiled kernels.
I need real, proper pen and paper for this.
Offline
Yes, the current udev in [testing] is slow. That's why it's in testing. I will be fixing it today.
Offline
but on my computer this effect also appears, even if I don't use testing.
And - as far as I have read - the other ones complaining in this thread aren't using testing neither.
Additionally I face this problem with the stock kernel and also with the fbcondecor-patched-kernel...
now with 80% more sax-appeal!
"I hacked the Phrak, and all I got was this lousy signature"
Offline
Yes, the current udev in [testing] is slow. That's why it's in testing. I will be fixing it today.
Yes, and I use [testing] to test software and to look if there are bugs. I think this is the reason why testing exists.
Nice to hear that it will be fixed today, thank you very much!
Greets,
maschino
Offline
I noticed the new version of udev in Core today. Since I was struggeling with _slow_ boot times due the "loading udev events..."-issue, I was happy thinking that the new udev would fix it all. Unfortunately it doesn't. My arch still is booting slow as hell.
Why is that udev-thing taking so long? I'm using arch for years now, being proud of the boot-speed and the speed when booted. But since a while, the boot-speed is horrible...
Offline
Well, the new UDev fixed the issue for me From 40secs to just 5
Offline
What scripts do you have in /etc/udev/rules.d/ ?
Offline
The /lib/udev/mod-blacklist.sh and /etc/udev/rules.d/00-load-blacklist.rules files shouldn't exist anymore :
http://cvs.archlinux.org/cgi-bin/viewcv … sroot=Core
That's what caused the slowdown.
pacman roulette : pacman -S $(pacman -Slq | LANG=C sort -R | head -n $((RANDOM % 10)))
Offline
The problem is also on my PC still existent. It got even worse, UDev uevent now even takes 1min.20sec.! blacklist.sh and 00-load-blacklist.rules do not exist on my system.
now with 80% more sax-appeal!
"I hacked the Phrak, and all I got was this lousy signature"
Offline
Are you sure your system is up to date?
$pacman -Qs ^udev$
local/udev 118-5 (base)
The userspace dev tools (udev)
$pacman -Qs ^filesystem$
local/filesystem 2008.03-2 (base)
Base filesystem
Udev uevent takes about 1,5 seconds on my system after the updates. If your system is up to date then there must be some module that causes problems on your hardware, I guess...
Offline
Same output as you. Damnit, how can I find such a module that is causing the delay other than disable them all and try enabling them one by one?
Can somebody (who's system don't has the problem) list the rules in the rules.d directory?
Last edited by Ibex (2008-03-13 15:49:49)
Offline