Arch Linux

Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines

Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.

REPEAT: Do NOT report bugs for outdated packages!
Tasklist

FS#3995 - Virtually useless MODULES array

Attached to Project: Arch Linux
Opened by Jaroslaw Swierczynski (swiergot) - Saturday, 18 February 2006, 11:02 GMT
Task Type Bug Report
Category Packages: Testing
Status Closed
Assigned To Tobias Powalowski (tpowa)
Architecture not specified
Severity Medium
Priority Normal
Reported Version 0.7.1 Noodle
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Since loading modules from rc.conf's MODULES array has been moved before udev, the array is mostly useless. I have lirc-serial and em8300 in the array and here is my modprobe.conf:

install lirc-serial setserial /dev/ttyS0 uart none ; modprobe -i lirc-serial ; true
install em8300 modprobe adv717x ; modprobe -i em8300 ; true

It all worked before but initscripts 0.7.1-21 broke it. When setserial is executed, /dev/ttyS0 does not exist although serial port support is built into kernel. As for em8300, it can't create device files because udev is not running (FYI, em8300 is not detected by udev). As a result, I might as well remove the modules from the MODULES array because I have to load them anew in rc.local anyway.

I fail to understand why this was done but obviously it's wrong.
This task depends upon

Closed by  Tobias Powalowski (tpowa)
Tuesday, 21 February 2006, 07:41 GMT
Reason for closing:  Fixed
Comment by Tobias Powalowski (tpowa) - Sunday, 19 February 2006, 09:33 GMT
ok we are working on a fix for that
Comment by Tobias Powalowski (tpowa) - Sunday, 19 February 2006, 13:34 GMT
could you please try if these 2 packages fix your problem?
Comment by Jaroslaw Swierczynski (swiergot) - Sunday, 19 February 2006, 14:31 GMT
First of all, there is a bug in rc.sysinit and booting stops. It's line 52:

if [ `pidof -o %PPID /sbin/udevd` ]; then

pidof gives more than one number which causes syntax error. The fix is simple:

if [ "`pidof -o %PPID /sbin/udevd`" ]; then

Now, the answer to you question is yes and no. Yes - the em8300 module creates all devices. No - /dev/ttyS0 still doesn't exist and setserial fails. May I ask why the MODULES array is processed before starting udev?
Comment by Tobias Powalowski (tpowa) - Sunday, 19 February 2006, 14:35 GMT
hmm, loading modules should happen between start udev and laoding uevents.
is module laoding still triggered before starting udev?
Comment by Jaroslaw Swierczynski (swiergot) - Sunday, 19 February 2006, 14:48 GMT
No, starting udev just doesn't create many devices, just fd, stdin, stdout, stderr, core, pts, shm, nothing more.
Comment by Tobias Powalowski (tpowa) - Sunday, 19 February 2006, 14:55 GMT
well yes then modules are loaded and should create your ttyS0 device, this is it not doing right?
Comment by Jaroslaw Swierczynski (swiergot) - Sunday, 19 February 2006, 15:05 GMT
Well yes, but serial port support isn't built as module, is it?
Comment by Tobias Powalowski (tpowa) - Sunday, 19 February 2006, 16:18 GMT
could you please try these packages?
Comment by Jaroslaw Swierczynski (swiergot) - Sunday, 19 February 2006, 17:26 GMT
Unfortunately, still no ttyS0 :(
Comment by Tobias Powalowski (tpowa) - Sunday, 19 February 2006, 18:02 GMT
coud you try this start_udev please?
Comment by Jaroslaw Swierczynski (swiergot) - Sunday, 19 February 2006, 18:28 GMT
Tobias, you're the man! :) Thanks!

Loading...