> This should fix a fairly serious interactive performance issue with > DragonFly when it is also running cpu-bound programs. The bug has > been in DragonFly since 2006. Corecode reported it. > > The fix is in HEAD atm. Please test for a bit, and we can MFC it to > stable if people report good results.
Which reminds me, wasn't there a a new CPU scheduler being written as a part of SoC? How is that progressing? Petr
