Hello,
some my statements in Part 1 were wrong.
1) I said OK to this rule:
ENV{MODALIAS}=="?*", RUN+="/sbin/modprobe $env{MODALIAS}"
Unfortunately, due to udev (and kernel?) bug, this doesn't work for
input devices even with 2.6.15 + gregkh patch. A working configuration
requires patching udev to accept a comma (",") as a trusted character in
sysfs attribute values. Patch is attached.
Then, the rule has to be rewritten as:
SYSFS{modalias}=="?*", RUN+="/sbin/modprobe $sysfs{modalias}"
Otherwise, udev looks at the wrong modalias for input devices.
2) I said OK to the following piece of the initscript:
list=$(echo /sys/bus/*/devices/*/uevent)
list="$list $(echo /sys/class/*/*/uevent)"
list="$list $(echo /sys/block/*/uevent /sys/block/*/*/uevent)"
The following line should be added for input subdevices to work in
certain semi-non-modular configurations:
list="$list $(echo /sys/class/*/*/*/uevent)"
With those changes, it is indeed possible to kill the input.sh hack and
its rule.
3) I forgot to say that, after commenting out udevstart, you may get
some errors due to /dev/null not being available. Avoid them by issuing
the following commands:
mknod -m 0666 /lib/udev/devices/null c 1 3
mknod -m 0600 /lib/udev/devices/console c 5 1
4) Also, I promised to talk about debugging.
I have a suspicion that the current waiting procedure is not good
enough, and an uevent can slip through. I ask all who will test Jim's
package to compile the test program:
gcc -o /lib/udev/bug bug.c
This program will append the whole uevent environment to the /dev/bug
file if it exists and the event is not related to known slow busses.
The compiled program is used instead of a shell script in order to
minimize timing differences between situations with and without the
program. Then, add the following rule to the end of 05-udev-early.rules:
RUN+="bug"
and replace the udev initscript with the attached one. Then reboot a few
times.
If the initscript tells you to report a bug, please do so. My statement
about unbootable system is VERY serious. I did have some uevents in the
/dev/bugreport file, all of them were USB-related. In order to avoid
SPAM, in the attached version of bug.c I disabled reporting of all USB
and IEEE1394 uevents that unexpectedly get past the loop.
To Matthew Burgess: this "bug.c" test also applies to you if you
implement hotplug removal independently.
--
Alexander E. Patrakov
Don't mail to [EMAIL PROTECTED]: the server is off until 2006-01-11
Use my GMail or linuxfromscratch address instead
Submitted By: Alexander E. Patrakov
Date: 2006-01-06
Initial Package Version: 079
Upstream Status: Not submitted yet, need confirmation from Jim Gifford
Origin: Alexander E. Patrakov
Description: Allows a comma in sysfs attribute values.
Needed for loading input modules properly.
--- udev-079/udev_utils_string.c 2005-12-22 23:51:30.000000000 +0000
+++ udev-079/udev_utils_string.c 2006-01-06 17:32:51.000000000 +0000
@@ -245,7 +245,7 @@
if ((str[i] >= '0' && str[i] <= '9') ||
(str[i] >= 'A' && str[i] <= 'Z') ||
(str[i] >= 'a' && str[i] <= 'z') ||
- strchr(" #$%+-./:[EMAIL PROTECTED]", str[i])) {
+ strchr(" #$%+-./:[EMAIL PROTECTED],", str[i])) {
i++;
continue;
}
#define _GNU_SOURCE
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <stdlib.h>
#include <argz.h>
int main(int argc, char * argv[])
{
char * envz;
char * devpath;
size_t len;
int bug;
devpath = getenv("DEVPATH");
if (devpath) {
/* Avoid SPAM on the list.
USB and IEEE1394 are known troublemakers.
*/
if (strstr(devpath, "usb"))
return 0;
if (strstr(devpath, "1394"))
return 0;
}
devpath = getenv("PHYSDEVPATH");
if (devpath) {
if (strstr(devpath, "usb"))
return 0;
if (strstr(devpath, "1394"))
return 0;
}
bug = open("/dev/bug", O_WRONLY | O_APPEND);
if (bug == -1)
return 0;
setenv("_SEPARATOR", "--------------------------------------", 1);
argz_create(environ, &envz, &len);
argz_stringify(envz, len, '\n');
envz[len-1]='\n';
write(bug, envz, len);
close(bug);
free(envz);
return 0;
}
#!/bin/sh
########################################################################
# Begin $rc_base/init.d/udev
#
# Description : Udev Boot Script
#
# Authors : Based on Open Suse Udev Rules
# [EMAIL PROTECTED]
#
# Adapted to : Jim Gifford
# LFS
#
# Version : 00.00
#
# Notes :
#
########################################################################
. /etc/sysconfig/rc
. ${rc_functions}
trigger_device_events() {
# generate events with the sysfs trigger
list=$(echo /sys/bus/*/devices/*/uevent)
list="$list $(echo /sys/class/*/*/uevent)"
list="$list $(echo /sys/class/*/*/*/uevent)"
list="$list $(echo /sys/block/*/uevent /sys/block/*/*/uevent)"
for i in $list; do
case "$i" in
*/device/uevent|*\**)
# skip followed device symlinks
continue
;;
*/class/mem/*|*/class/tty/*)
first="$first $i"
;;
*/block/md*)
last="$last $i"
;;
*/*)
default="$default $i"
;;
esac
done
# trigger the sorted events
for i in $first $default $last; do
echo "add" > "$i"
done
}
case "$1" in
start)
boot_mesg "Creating /dev in tmpfs..."
mount -n -t tmpfs tmpfs /dev -o mode=755,size=4M
evaluate_retval
boot_mesg "Copying static entries..."
cp -ar /lib/udev/devices/* /dev
evaluate_retval
# boot_mesg "Executing udevstart..."
# udevstart
# evaluate_retval
# disable hotplug helper, udevd listens to netlink
echo "" > /proc/sys/kernel/hotplug
# start udevd
boot_mesg "Starting udevd..."
/sbin/udevd --daemon
evaluate_retval
# cleanup some stuff
# rm -f /var/run/sysconfig/network
# rm -rf /events/*
# start coldplugging
boot_mesg "Performing Coldplugging..."
# unlikely, but we may be faster than the first event
mkdir -p /dev/.udev/queue
# configure all devices
trigger_device_events
# until we know how to do better, just wait for _all_ events to finish
loop=300
while test -d /dev/.udev/queue; do
sleep 0.1;
test "$loop" -gt 0 || break
loop=$(($loop - 1))
done
>/dev/bug
test "$loop" -gt 0
evaluate_retval
sleep 2
if test -s /dev/bug; then
mv /dev/bug /dev/bugreport
boot_mesg "Please mail the contents of /dev/bugreport to
lfs-dev@linuxfromscratch.org" ${WARNING}
boot_mesg "Otherwise, the next version of LFS will be unbootable on
your system!"
echo_failure
sleep 10
else
rm -f /dev/bug
fi
;;
stop)
boot_mesg "Stopping udevd..."
echo "/sbin/hotplug" > /proc/sys/kernel/hotplug
killproc /sbin/udevd
;;
restart)
boot_mesg "Restarting udevd..."
killproc /sbin/udevd
loadproc /sbin/udevd --daemon
evaluate_retval
;;
status)
statusproc /sbin/udevd
;;
reload)
boot_mesg "Reloading udev rules..."
udevcontrol reload_rules
cp --preserve=all --recursive --update /lib/udev/devices/* /dev
evaluate_retval
;;
force-reload)
boot_mesg "Updating all available device nodes in /dev..."
udevcontrol reload_rules
rm -rf /dev/.udev /dev/disk
cp --preserve=all --recursive --update /lib/udev/devices/* /dev
trigger_device_events
evaluate_retval
;;
*)
echo "Usage: $0 {start|stop|restart|status|reload|force-reload}"
exit 1
;;
esac
--
http://linuxfromscratch.org/mailman/listinfo/lfs-dev
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page
