IrDA baud_rate not configurable at boot

[Daniele Napolitano]

Binary package hint: irda-utils

I made the baud rate of SIR trasmission configurable by /etc/default/irda-utils

I Ubuntized the irda-utils init script with use of "log_daemon_msg" and "log_end_msg".

[Steve Langasek (vorlon)]

Hi Daniele,

A few points:

- please submit patches using unidiff format (diff -u) instead of classic diff; classic diffs don't include any context to ensure that the patch will apply correctly if the target file has changed.
- why have you provided copies of log_action_begin_msg() and log_action_end_msg() that are never used? Even if these are meant to be log_daemon_msg() and log_end_msg(), the correct approach is to make irda-utils depend on the package that provides /lib/lsb/init-functions, not to work around its absence.
- why should irda-utils call sysctl at all from an init script, instead of installing a file under /etc/sysctl.d? Since the file in /etc/sysctl.d would be a conffile, users could edit that instead of /etc/default/irda-utils to get the same result.

Unsubscribing ubuntu-main-sponsors for the moment.

[Daniele Napolitano (dnax88)]

There is a problems to add the dependency to lsb-base?

About use of /etc/init.d/irda-utils instead of /etc/sysctl.conf is for more simplicity and coherence. If I want to change the infrared settings, edit files provided by irda-utils.

[pstadelmann (patrick-stadelmann)]

I still have problems with IrDA in Ubuntu 8.10, it seems the baudrate is not taken into account.I've applied the patch above, and in irdadump, I can see that the specified maximum baud rate is indeed used in the SNRM frame sent by the Ubuntu PC to the secondary device. If the max of both devices is 9600, the communication works fine. However, if it is e.g. 115200, it doesn't work. I found that if I modify the secondary device to use 9600 even though it said 115200 in the UA frame when responding to the SNRM frame, it works ! So it seems that somehow the negotiated baudrate is not used by the Ubuntu PC.

[pstadelmann (patrick-stadelmann)]

Well, it's a known bug in kernel 2.6.27

http://bugzilla.kernel.org/show_bug.cgi?id=11795

[Daniele Napolitano (dnax88)]

Updated Debian package imported.