summaryrefslogtreecommitdiff
path: root/sys/arch
AgeCommit message (Collapse)Author
1999-04-27make this compile for !MNN case. (which is the default). DUH.Artur Grabowski
1999-04-24Use the generic m68k param.h.Jason Downs
1999-04-23correct keycap.pcvt pathTodd C. Miller
1999-04-23last part of UVM stuff for pmap (aborted yesterday by a routing loop).Artur Grabowski
Just changed names of variables and functions
1999-04-23mmmmap() bounds check.Jason Downs
1999-04-23Kcore dump, from NetBSD.Jason Downs
1999-04-23Add kcore.hJason Downs
1999-04-23Fix kcore initialization.Jason Downs
1999-04-22MACHINE_NEW_NONCONTIG code. From NetBSD. Needed by UVMArtur Grabowski
1999-04-22UVM stuff. Mostly name changes.Artur Grabowski
1999-04-22UVM stuff. vm_fault -> uvm_faultArtur Grabowski
1999-04-22UVM kernacc -> uvm_kernaccArtur Grabowski
1999-04-22UVM fixes. Mostly changed function names.Artur Grabowski
New buffer allocation from NetBSD
1999-04-22kcopy needed by UVM. From NetBSDArtur Grabowski
1999-04-22UVM fixes, cnt -> uvmexpArtur Grabowski
1999-04-22constants for UVMArtur Grabowski
1999-04-22[vp]{addr,size}_t needed by UVMArtur Grabowski
1999-04-22implement pmap_{,de}activateArtur Grabowski
1999-04-22protect from multiple includeArtur Grabowski
1999-04-22UVM fixes. vm_map_create -> uvm_map_create and kmem_alloc -> uvm_km_allocArtur Grabowski
1999-04-22UVM fixes, just a few useracc -> uvm_useraccArtur Grabowski
1999-04-22fix device selector, spif isn't an DV_NET, missing sbus_establishJason Wright
1999-04-22Bug fixes from NetBSD, port-powerpc/7240 and port-powerpc/7243Dale S. Rahn
The first fixes a major problem seen _too_ many times to record. If a task that is busy doing disk io is interrupted with a ^C, its would frequently put the system into a mode where it would very shortly crash, It was known that it was a signal delivery that was occurring incorrectly, but had not been diagnosed where. It is no longer possible to crash the machine on whim. (This probably should be recorded as a stability patch for 25). The second was not a bug that had ever been seen to be identified, but looking at the code it was obvious that the onfault variable was not being reset before the function exit. The code differs between NetBSD and OpenBSD somewhat, however only two instances of the code was reported, and three instances exist in the OpenBSD code. (all were fixed).
1999-04-22Follow proper OpenBSD shutdown/power off semantics.Jason Downs
1999-04-20disable some stuffMichael Shalayeff
1999-04-20one moreMichael Shalayeff
1999-04-20establish irq handlerMichael Shalayeff
1999-04-20note a couple of typical io layouts for referenceMichael Shalayeff
1999-04-20name changesMichael Shalayeff
1999-04-20syscall(), fix interrupt handlingMichael Shalayeff
1999-04-20do lazy blok mapping of io space in bus_*Michael Shalayeff
add copy{in,out}*, kcopy, etc boot routines cpu-id-hpux-compat stuff rest of scheduler and such uvm
1999-04-20no longer blockmap the whole io spaceMichael Shalayeff
fix bootstrapping uvm
1999-04-20flesh out, who cares about single-step anywayMichael Shalayeff
1999-04-20some cpu_* routines, uvmMichael Shalayeff
1999-04-20map mainbus spaceMichael Shalayeff
1999-04-20export some moreMichael Shalayeff
1999-04-20more devicesMichael Shalayeff
1999-04-20dumpconf()Michael Shalayeff
1999-04-20scheduler, syscalls, {fs}u{,s}wordsMichael Shalayeff
1999-04-20workaroun ld overflow bug; general cleanup, alignments, pads, and needed ↵Michael Shalayeff
sections/symbols
1999-04-20factor out common devices/optionsMichael Shalayeff
1999-04-20map the io spaceMichael Shalayeff
1999-04-20print cpu ID (type) if availableMichael Shalayeff
1999-04-20s/MINIOSIZ/IODC_MINIOSIZ/Michael Shalayeff
1999-04-20move rd_root_{size,image} definitions to the end of the file to workaround ↵Michael Shalayeff
linkr overflow bug
1999-04-20call any vegetable, call it by nameMichael Shalayeff
1999-04-20syncMichael Shalayeff
1999-04-20most of cpu boards defined nowMichael Shalayeff
all the bus types and all the rest of small thingies (; some more boards, add more by driver progress new framework
1999-04-20move load address to 0x700000, so we fit w/ ramdiskMichael Shalayeff
heap limit on 0xc0000 remove networking stuff
1999-04-20fix prev commit blowup:Michael Shalayeff
no ldsym for now s/sd/dk/ since it's not always a disk, not even scsi (: bump a version, things changed