cvs commit: src/sys/dev/disk/nata ata-all.c

["Thomas E. Spanjaard" <tgen@xxxxxxxxxxxxxxxxxxxxxxx>]

tgen        2007/01/09 13:17:00 PST

DragonFly src repository

  Modified files:
    sys/dev/disk/nata    ata-all.c 
  Log:
  Add a temporary fix to ata_interrupt() to ignore weird interrupts seen on
  SMP+ATAPI+DMA. For some reason, the ATAPI device asserts INTRQ while it's
  otherwise completely ready to fulfill the request. DRQ is set in the status
  register, and no matter how I read the specs, this just should never
  happen. Is this actually a race condition? I have no idea :(. Ignoring the
  interrupt, however, does not seem to produce adverse results (tested with
  a DragonFlyBSD livecd with SMP kernel).
  
  Revision  Changes    Path
  1.7       +14 -0     src/sys/dev/disk/nata/ata-all.c


http://www.dragonflybsd.org/cvsweb/src/sys/dev/disk/nata/ata-all.c.diff?r1=1.6&r2=1.7&f=u