On 04/28/2014 01:50 PM, Harshavardhana wrote: >> >> That's fine, but until we get more slaves running regression I can't see >> adding 15+ minutes to each regression run. > > Why isn't this added to local patch submission script itself in > ./rfc.sh ?- and reduces the load on the server. Every patch submission > runs 'cppcheck' on the changeset which was committed locally and if > its fails hard error and the developer has to fix it. That's certainly an option. Although our developers might not like the 15+ minutes it adds to their submission. That said, I'd venture to say that 'good' developers were already running things like valgrind, coverity, cppcheck, other compilers, etc., against the code. But doing those things take time. Except for cppcheck, which I wasn't aware of, I've done the others occasionally. And I plan to start running them routinely in cron on the lab machines we have coming on line now. ;-) -- Kaleb
