My example: If you use gpm and it or your computer crashes, gpm will just bitch about gpm running already as a /var/run/gpm.pid exists, /var/run/*pid should be cleaned up before services are started during boot! Reproducible: Always Steps to Reproduce:
It will be the case if you try to restart it without rebooting. Do this test: # rc-update del gpm # /etc/init.d/gpm stop zap # touch /var/run/gpm.pid Now reboot, and check if /var/run/gpm.pid still exists. You should see that it is gone. The only reason it will be there, is if you have gpm to start at boot, but it crashes, and the gpm init script did not get a fail status from gpm, so marked it as started.
as I cannot reproduce it anymore it does appear to be a brainbug (or config-problem) ;]
Gpm was prob started on that reboot, but crashed. I have some pid checking that I need to finish and implement, but that is another bug.