> Hum, I just tested. Yes the atime is updated on a write, and testing > PGO_OVERWRITE doesn't fix the problem :( i think testing access_type works for -current. it doesn't work for netbsd-2, though. YAMAMOTO Takashi