mpab timeout on port builds

[dweber@…]

I was running mpab until it hit a snag on a port that was failing in an infinite loop (guile or something). It looks like mpab needs a process to monitor the time of each port build with a timeout limit to kill a build and move on to the next build.

[blb@…]

Having some sort of watchdog would be nice, but difficult to do properly. Since it just kicks off a port process for each build, it isn't watching processes come and go directly, so it can only watch that one port process. A simple timeout would be bad for ports that are known to take time (eg, ghc, the gcc* ports, etc) as we may otherwise be stopping a build that was going just fine.

Perhaps some sort of monitoring of the log file being created, as output should be going there on a regular basis, and if not then it may very well be stalled. Are there ports which, even in debug mode, output very little during a successful build?

[(none)]

Milestone Port Bugs deleted

[tobypeterson]

This should be a base bug, I think...