LinuxQuestions.org - Lenny boots slowly on iBook G4

- Debian (https://www.linuxquestions.org/questions/debian-26/)

- - Lenny boots slowly on iBook G4 (https://www.linuxquestions.org/questions/debian-26/lenny-boots-slowly-on-ibook-g4-705832/)

Lenny boots slowly on iBook G4

Hello,

I have updated to Lenny yesterday. Everything is working fine but there's a little thing that caught my attention and I'd like to get rid of. Here it :

Code:

debian:/home/stefan# dmesg  

... some output cut ...

[  26.032778] usbcore: registered new interface driver appletouch

[  26.570544] Broadcom 43xx driver loaded [ Features: PMLR, Firmware-ID: FW13 ]

[  26.590843] udev: renamed network interface wmaster0 to eth2

[  27.644382] snd-aoa-fabric-layout: found bus with layout 80

[  27.652076] snd-aoa-fabric-layout: Using PMF GPIOs

[  27.702081] snd-aoa-codec-tas: found tas3004

[  27.702127] snd-aoa-fabric-layout: can use this codec

[  27.775889] snd-aoa-codec-tas: tas found, addr 0x35 on /pci@f2000000/mac-io@17/i2c@18000/i2c-bus@0/codec@6a

[  58.595745] Adding 387452k swap on /dev/hda5.  Priority:-1 extents:1 across:387452k

[  59.235305] EXT3 FS on hda4, internal journal

[  59.915206] loop: module loaded

... some output cut...

The strange thing is that after the last snd-aoa-code-tas line (in bold), there's a 30 seconds pause.

It is repeatable. Why is there such a long pause ? When the boot is running, this pause is quite visible.

I've checked the logs (as far as I know that stuff) and nothing seems wrong, no time out, no error and the system is doing fine.

Nevertheless, I'd like to get a top speed boot :-)

Although I don't expect any hard answer, I'd be glad to learn how to get more boot-time information so that I can track the issue better.

Thanks,

Stefan

I've investigated some more. While the boot is running, I read :

waiting for /dev to be fully populated

this doesn't show up in any logs afterwards, it's just shown on the console while booting.
After that line, there's some of the stuff I've shown in my previous message and then, there's :

Done.

So I guess the problem lies there and has nothing to do with snd-aoa stuff.

So the next question is : what is this /dev population thing and do I debug it (or at least, make it more verbose).

stF

Oh btw, on the very same machine, I've used Etch for a year and never had such an issue...

Activating udev debugging to a file is simple (when you know :-))

in /etc/init.d/udev

I set :

Code:

    log_daemon_msg "Starting the hotplug events dispatcher" "udevd"

    if udevd --daemon --debug >> /dev/.udev/udevd.log; then  # debug added by stefan

        log_end_msg $?

    else

        log_end_msg $?

    fi

et voilà... No I have to make sens of all that information.

stF

Made some progress, by commenting out this line :

# SUBSYSTEM=="net", DRIVERS=="?*", ATTRS{address}=="00:14:51:85:13:6e", NAME="eth2"

in

/etc/udev/rules.d/70-persistent-net.rules

and now it works like a charm... But I'm sure that disabling this line is breaking some functionality....

stF

There was also this method:
Edit /etc/udev/udev.conf and change priority
dpkg-reconfigure udev
Reboot, debug should be in syslog

Thx for the advice, it is much cleaner than my direct editing...

stF