summaryrefslogtreecommitdiff
path: root/sys
AgeCommit message (Expand)Author
2009-11-03don't take playback blocksize constraints into account when roundingJacob Meuser
2009-11-03don't claim to support full-duplex if the device doesn't support bothJacob Meuser
2009-11-03provide a get_default_params method. tries aucat(1) defaults firstJacob Meuser
2009-11-03rewrite supported audio parameter matching codeJacob Meuser
2009-11-03Simplify and remove debug cruft from interrupt handler. Original diff fromMarco Peereboom
2009-11-02Don't write bogus values to reply_fifo_host_signalling_addr. This registerMarco Peereboom
2009-11-02s/hz/Hz/ on multiples of the SI unit hertz other than MHz.Igor Sobrado
2009-11-02Got the RX threshold interrupt working, does not help RX performance though;Miod Vallat
2009-11-02Remove file that should never have been there.Mark Kettenis
2009-11-02Remove unused member(s) of `struct acpi_softc'.Mark Kettenis
2009-11-02A few fixes and improvements:Miod Vallat
2009-11-02When the iec(4) Ethernet address is unkown, pass ff:ff:ff:ff:ff:ff instead ofMiod Vallat
2009-11-02Add support for ``M-Audio Revolution 5.1'' cards, based on envy24HT.Alexandre Ratchov
2009-11-02correct signature. no change as function is if 0'd.Federico G. Schwindt
2009-11-01Move IS[12]BYTEMSG and ISEXTMSG defines to a common place.Federico G. Schwindt
2009-11-01Remove bogus #define __HAVE_GENERIC_SOFT_INTERRUPTS. No code change.Mark Kettenis
2009-11-01Call selwakeup()/KNOTE() even if the queue has emptied completely.Nicholas Marriott
2009-11-01kqueue support for midi(4).Nicholas Marriott
2009-11-01Looks like I forgot to add files again. These are the most important filesMiod Vallat
2009-11-01Driver for the sgi IOC3 onboard Ethernet interface. Tested on Octane only soMiod Vallat
2009-11-01call powerhook_disestablish() and pci_intr_disestablish() earlier.Damien Bergamini
2009-11-01- fix rx chain mask in scan commandDamien Bergamini
2009-11-01add support for the ati sb900^Whudson2 chips.David Gwynne
2009-10-31A chunk from my WIP bootblocks tree went in by mistake in previous commit;Miod Vallat
2009-10-31Establish a shutdown hook to disable the watchdog timer to prevent watchdogMark Kettenis
2009-10-31Make sure we don't insert and entry into the list of pending commends twice.Mark Kettenis
2009-10-31in the absense of link state handling, drivers should still be silent; ok miodTheo de Raadt
2009-10-31KNOTE() after selwakeup(); ok miodTheo de Raadt
2009-10-31Do not override AS and LD, to be cross compilation friendly; noticed by syuu@Miod Vallat
2009-10-31s/Mhz/MHz/, MHz is a multiple of the SI unit hertz (whose symbol is Hz).Igor Sobrado
2009-10-31Use suser when possible. Suggested by miod@.Federico G. Schwindt
2009-10-31reorder operations in detach(), call powerhook_disestablish() earlier,Damien Bergamini
2009-10-31Add missing KNOTE() calls after selwakeup(), until we decide if the KNOTE()Theo de Raadt
2009-10-31Make sure the descriptor is writable for some operations.Federico G. Schwindt
2009-10-31Blinkenlichten on Octane.Miod Vallat
2009-10-30missing \n in log() message.Damien Bergamini
2009-10-30correct interrupt status bits; patfbsd@davenulle.orgTheo de Raadt
2009-10-30KNOTE() after selwakeup()Theo de Raadt
2009-10-30pipeclose() calls pileseltimeout() which does the KNOTE(); therefore callingTheo de Raadt
2009-10-30Add missing KNOTE() calls after selwakeup() in ptcread() (to wake up writersNicholas Marriott
2009-10-30Don't do a double free, sys_getsockopt() will take care of freeing theClaudio Jeker
2009-10-30Support IP30 secondary cpu bootup. ok miod@Takuya ASADA
2009-10-29When probing devices at boot, compare against the full device driver nameJoel Sing
2009-10-29add isp(4), but disable it since on most machines the pci bus is earlierTheo de Raadt
2009-10-29I missed updating a line for the AF_INET6 route case whenJonathan Gray
2009-10-29Make sure the descriptor is writable for some operations. Pointed by miod@.Federico G. Schwindt
2009-10-29iha(4), just like on GENERIC; from bradTheo de Raadt
2009-10-28a few leftovers from yesterday's SCCS/RCS-ID removal;Ingo Schwarze
2009-10-28LOMlite seems to get wedged from time to time; add some code to unwedge it.Mark Kettenis
2009-10-28delete balony comment; ok kettenisTheo de Raadt