← Back to team overview

rohc team mailing list archive

Reboots with trunk rohc-lib in kernel module for embedded platform (arm)

 

Hi,

I have integrated the trunk version of rohc-lib into a kernel module for an
embedded arm platform running kernel version 2.6.20. The kernel module is a
network driver for radio hardware to form a point to point link. The goal
is to test rohc-lib 1.7 with the new TCP profile along with the IP, RTP and
UDP profiles on this embedded platform.

I enabled only the IP profile and performed a ping which causes a reboot
after a random number of successful pings. This seems similar to bug
#1213411 (https://bugs.launchpad.net/rohc/+bug/1213411) - a NULL pointer
deference. However they appear to happen at random points in either the
rohc-lib code or the kernel module and cause a reboot. Sometimes a kernel
oops is produced, but most times the system reboots without displaying
anything.

Running the kernel module without rohc-lib enabled does not cause reboots.
Using only the rohc-lib uncompressed profile also does not cause reboots.
So far I experience reboots with the UDP and IP profile.

Does this relate to bug #1213411?

Are there any known issues with kernel 2.6.20 that may cause this?

Or are there any suggestions as to what may be causing this?

Any help would be much appreciated.

Regards,

Sean

Follow ups