GRASS monitor problems on Solaris Workstations

Cliff Wood cbwood at mail.vt.edu
Tue Aug 5 18:10:29 EDT 1997


I am currently attempting to integrate GRASS routines (shell scripts)
into an existing GIS project that I am developing. I am currently having
problems, however, with the GRASS graphics monitors.

It seems, when the host program (written in C, it uses system() to make
calls to the GRASS scripts...I do it this way to prevent having to
recompile the GRASS code) locks up, sometimes it locks whatever
resources the monitors use (my guess are the fifos) and after this
happens, no matter *what* I do (aside from a shutdown), I cannot get
those GRASSs monitors unlocked or killed. Sometimes using "killall -9
ftcXDRIVER" will work, but then sometimes it won't. This same issue
holds true for d.mon/mon.status programs. This problem persists across
sessions, and removing stale lock files doesn't seem to help.

A reboot fixes the problem, however what *causes* it in the first place?
Is it fixable? Is there a way to fix it WITHOUT having to down the
machine that GRASS resides on?


- Thanks for the time. Sorry I had so many questions.



More information about the grass-user mailing list