On 26 Apr, Gopal Narayanan wrote: > Miquel van Smoorenburg writes: >> >> In article <[EMAIL PROTECTED]>, >> Gopal Narayanan <[EMAIL PROTECTED]> wrote: >> >I upgraded one of our machines to slink from hamm. The upgrade went >> >well. But after the upgrade I keep getting segmentation violations when >> >I run minicom. We use minicom to talk to another local machine through >> >the serial port. It was working fine before the upgrade. I am running >> >kernal 2.0.34. >> >> Well from the message you are including below (a kernel oops), it >> is clear that it isn't minicom that crashes, it's the kernel that >> crashes! Fortunately only minicom is killed and not the whole machine.. >> >> >When I start up minicom, it is fine until I give it one keystroke, upon >> >which it segfaults. Any help will be much appreciated. Will a kernel >> >upgrade be helpful. If so what kernel should I upgrade to? >> >> Oh yes, that will probbaly help. 2.0.36 is the latest stable kernel >> in the 2.0.x series, try that one. >> > > Thanks for your message. I did upgrade to 2.0.36. Still no luck. Same > problem. >
I found the problem. Kernel upgrades did not help as mentioned above. It appears that during the upgrade of the setserial package, I said yes to the installation of a new /etc/rc.boot/0setserial. That seems to break serial port access. When I put back the old 0setserial, and rebooted, minicom worked fine. Inspecting the two scripts, the old and new, I see enough differences. I will try to understand this and turn in a bug report if warranted. Has anyone else had a problem with the 0setserial bootup script in setserial-2.14-3 ? Gopal. -- ---------------------------------------------------------------------------- Gopal Narayanan Ph #: (413) 545 0925 Five College Radio Astronomy Observatory Fax#: (413) 545 4223 University of Massachusetts e-mail: [EMAIL PROTECTED] Amherst MA 01003 ---------------------------------------------------------------------------
