Description of problem: I've just installed fedora 5 from the installation CDs on my IBM Lenovo Thinkpad T60p, I am experiencing a hard hang after seeing the "Starting udev" text. I followed all the hints from Bugzilla Bug 186164 รข udev hang on boot but none of them did remedy the problem. Version-Release number of selected component (if applicable): Stock udev from FC5 DVD Steps to Reproduce: 1. Install FC5, accepting all defaults 2. Reboot 3. Observe hang How reproducible: Always Actual results: Computer hangs Expected results: Computer runs Additional info:
In the meantime I've updated the kernel to 2.6.16-1.2096_FC5smp and xorg to xorg-x11-server-Xorg-1.0.1-9. I've also installed the pcmciautils-012-1.i386.rpm package. I get the message: Starting udev: Wait timeout. Will continue in the background The computer is responsive, but it waits until eternity... :-( By the way: adding the "acpi=off" ended in Kernel panic, because "ata1 failed to respond (30 secs)"
Created attachment 128083 [details] /etc/modprobe.d/blacklist-compat Does it help, if you copy this file to /etc/modprobe.d/blacklist-compat?
No, unfortumately not. The behaviour is the same as before: Starting udev: Wait timeout. Will continue in the background The computer is responsive, but it waits until eternity... :-(
My computer is still waiting ... Can I do something to isolate the error further? I need my Linux box very urgent!!!
hmm.. /etc/udev/udev.conf ... log to debug
udev_log="debug"
There are too many lines too fast going through the screen, but I recognize a strange messaage: pass_env_to_socket: passing -1 bytes to ... /org/...
comment out the line in 50-udev.rules: ACTION=="add", SUBSYSTEM=="?*", ENV{MODALIAS}=="?*", RUN+="/sbin/modprobe $env{MODALIAS}" This will prevent module loading and retry... I suspect a module will get loaded, which lets the kernel to stall. Then you could load all the modules by hand and see which stalls..
Created attachment 128317 [details] the log from system-config-soundcard
Sorry, I've made a mistake, so my previous message should be appended after this one: Your trick worked fine, my Linux box is up and running (first without a net connection, which worked after a modprobe e1000 ). Now I have still two/three problems: 1) the soundcard isn't working at all 2) my usb stick isn't recognised 3) at the moment I don't know if the finger print reader is working (I will look at it when the previous problems are solved) For 1) I've sent the log from system-config-soundcard already, additionally I've tried to modprobe snd_hda_intel the coresponding terminal stalled, the output from ps aux | grep modprobe is root 3093 0.0 0.0 1752 608 pts/4 S+ 19:30 0:00 modprobe snd_hda_intel root 3106 0.0 0.0 4400 988 pts/4 S+ 19:30 0:00 sh -c /sbin/modprobe --ignore-install snd-pcm && /sbin/modprobe snd-pcm-oss && /sbin/modprobe snd-seq-device && /sbin/modprobe snd-seq-oss root 3126 0.0 0.0 1808 576 pts/4 D+ 19:30 0:00 /sbin/modprobe snd_seq_oss root 3149 0.0 0.0 1660 552 ? D< 19:30 0:00 /sbin/modprobe -q -- snd_card_0 root 3263 0.0 0.0 3940 752 pts/5 S+ 19:33 0:00 grep modprobe and on the text terminal <Ctrl><Alt><F1> I see many messages like printk: 11 messages suppressed azx_get_response timeout For 2) I've looked at my older Fedora 4 installation, there are a lot of modules that include a "usb" in their names. Is there a module which can be loaded first and the rest is done automatically or is there a list of modules that should be loaded piece by piece?
to work without sound remove from modules.conf: alias snd-card-0 snd-hda-intel options snd-card-0 index=0 options snd-hda-intel index=0 remove snd-hda-intel { /usr/sbin/alsactl store 0 >/dev/null 2>&1 || : ; }; /sbin/modprobe -r --ignore-remove snd-hda-intel and add snd-hda-intel to /etc/modprobe.d/blacklist. restore the udev modprobe rule and try to restart. If you can boot, we will reassign this bug to kernel.
to work without sound remove from modules.conf: ------- I think should read as modprobe I can boot, the usb stick is up and running. Thank you for all your efforts! If I can help I'll will do that according to your advices.
if you still have the problem with snd_hda_intel, you can file a new bug against the component kernel. Closing this bug with "NOTABUG" regarding udev.