summaryrefslogtreecommitdiff
path: root/sys
AgeCommit message (Collapse)Author
2005-03-17fix ral_enable_tsf_sync() for hostap mode.Damien Bergamini
2005-03-17do not drop ToDS frames when acting as an access point.Damien Bergamini
fix beacon expired interrupt handler in hostap mode.
2005-03-17give ipv6 a chance. pointed out by Paul de Weerd.Damien Bergamini
2005-03-17enable tx retries for unicast frames.Damien Bergamini
ok and tweaks from dlg@
2005-03-17fix for big endian archsDavid Gwynne
ok damien@
2005-03-17- fix ralink urlDamien Bergamini
- add usb events for attach/detach - pre-allocate rx buffers so they are not allocated in an interrupt context - let ural_free_rx_list() do the cleanup
2005-03-17+vge, ciphy; requests by pvalchevTheo de Raadt
2005-03-16oops, this needs the ural classifierTheo de Raadt
2005-03-16ural will be here tooTheo de Raadt
2005-03-16permit ramdisks to compileTheo de Raadt
2005-03-16syncDamien Bergamini
2005-03-16move ral vendor/product Ids into usbdevsDamien Bergamini
2005-03-16enable uralTheo de Raadt
2005-03-16Driver for Ralink Technology RT2500USB wireless adapters.Damien Bergamini
ok deraadt@
2005-03-16Use biowait() instead of rolling our own.Miod Vallat
tested krw@ moritz@, ok deraadt@
2005-03-16+rtw, ral, but a few rare ethernet devices go away..Theo de Raadt
2005-03-16ath, atw, ral, rtw on CD ramdiskTheo de Raadt
2005-03-16more wireless chipsetsTheo de Raadt
2005-03-16play the relied-upon-xlation proper; deraadt@ okMichael Shalayeff
2005-03-16don't account packet's twice; from joel@; ok deraadt@, henning@Markus Friedl
2005-03-16amd64 also fails copyoutstr regress, missed in the first round.Ted Unangst
need to checker the to address. patch from netbsd. ok deraadt mickey
2005-03-15s/IPL_BIO/IPL_TTY. ok drahn@, deraadt@Uwe Stuehler
2005-03-15tag for release, more things coming, but ports needs thisTheo de Raadt
2005-03-15enable hotplug on macppc, been using it for some time, forgot it wasn't enabled.Dale Rahn
2005-03-15Skip empty pci ranges rather than aborting probe, as done on sparc64.Miod Vallat
ok drahn@
2005-03-15Only warn if pci_{io,mem}_find fails if the returned value is not ENOENT - emptyMiod Vallat
ranges are skipped and considered harmless.
2005-03-15Only print warning messages in pci_{io,mem}_find() if option DEBUG; instead,Miod Vallat
do not return 1 on failure but a real errno the caller can check for proper action. tweaks and ok mickey@
2005-03-15A potpourri of cgfourteen fixes and improvements:Miod Vallat
- use the autoincremented luts for faster colormap programming. - colormap changes are now deferred to the vertical retrace interrupt, causes less display flickering under X11 (still some visible at boot time because interrupts are disabled at this point). - if the resolution choosen does not allow the VSIMM to run in 32bit mode, do not allow mmap()ing past the frame buffer area. - print the VSIMM revision at boot time, and be more friendly to revision 0, which needs to be handled in a slightly different way.
2005-03-15If no "intr" property exists, correctly handle the "interrupts" property forMiod Vallat
non-SBus devices, such as cgfourteen@obio.
2005-03-15Nuke matchbyname(), which isn't used anymore.Miod Vallat
ok deraadt@
2005-03-15Take care of updating the PROMs view of the cursor position in the common fbMiod Vallat
code, rather than doing this in only a subset of the fb drivers.
2005-03-15byte order of mss, only affects synproxy code path, from John L. ScarfoneDaniel Hartmeier
2005-03-15use IFQ_ rather than IF_ macro here too, maybe helps altq; ok henningPeter Valchev
2005-03-15use IFQ_* macros instead of hand-rolling or using IF_*, makes ALTQ work onHenning Brauer
re, from Armin Wolfermann <aw@osn.de>, ok dlg markus pval
2005-03-15Seperate delay for discharge and charge, otherwise discharge period may notDale Rahn
be long enough. should fix most of the extra keys on rollover.
2005-03-15- stop tearing down routes when the carp interface goes to backup in theRyan Thomas McBride
carpdev case; block outgoing packets in carp_output() instead. - unbreak the "bow out on 'ifconfig down'" behaviour, allows instant failover on graceful halt/reboot to work again. ok mpf@ henning@
2005-03-15Make attach errors appear on new lines and makeJonathan Gray
dmesg output less ugly while here.
2005-03-15no need to flush a few fpregs since soon the whole pcb get flushedMichael Shalayeff
2005-03-15hafta flush pcb once active (and a tush tush tush)Michael Shalayeff
2005-03-15revise lockf limit policy. after exceeding "soft limit", start enforcingTed Unangst
a hard limit of 2x. add a comment to this effect. reviewed by miod
2005-03-15Try harder to prevent spurious touch screen 'down' events.Dale Rahn
2005-03-15Restructure the keyboard code so that the keys in rawkey are always sentDale Rahn
as expected. zero both keystate structures, keystate may not be fully initialized. worked on with uwe.
2005-03-14Fix typo in commentTom Cosgrove
2005-03-14regenPedro Martelletto
2005-03-14Add GeForce4 MX 4000, okay deraadt@Pedro Martelletto
2005-03-14don't display an error message when re-associating.Damien Bergamini
this prevents annoying 'unknown association state 9' from happening.
2005-03-14syncTheo de Raadt
2005-03-14support jvc mp-prx1; iedowseTheo de Raadt
2005-03-14more Yukon models and Yukon EC revs.Brad Smith
2005-03-13Add missed WSDISPLAYIO_[GS]VIDEO ioctl handling.Miod Vallat