summaryrefslogtreecommitdiff
path: root/sys
AgeCommit message (Expand)Author
2015-01-19Remove dead store causing clang to warn; dhillMiod Vallat
2015-01-19use SYMLOOP_MAX rather than MAXSYMLINKSTheo de Raadt
2015-01-19mikeb points out that 'struct ipsec_policy' can also be hidden by _KERNELTheo de Raadt
2015-01-19SYMLOOP_MAX == MAXSYMLINKS, so use SYMLOOP_MAXTheo de Raadt
2015-01-19First step of hiding many kernel-only parts of <netinet/ip_ipsp.h>Theo de Raadt
2015-01-19Make use of an msr available on recent Intel processors to obtain theJonathan Gray
2015-01-19Make some $OpenBSD$ lines prettier/standardier by eliminatingKenneth R Westerback
2015-01-19<netinet6/in6_ifattach.h> is not needed here.Martin Pieuchot
2015-01-19pass the size of the softc to free.David Gwynne
2015-01-19white space fixes. no binary change.David Gwynne
2015-01-19unneccessary cmask variable; ok guentherTheo de Raadt
2015-01-18Back out previous. HOST_NAME_MAX is only defined for newer POSIXTodd C. Miller
2015-01-18Forward declare 'struct file' so that libkvm and others that indirectly pullPhilip Guenther
2015-01-18Do not trust the content of event TRBs coming from the hardware andMartin Pieuchot
2015-01-18When restoring spsr values when handling traps use spsr_fsxc insteadJonathan Gray
2015-01-18Switch some uses of msr that only deal with interrupts/mode to useJonathan Gray
2015-01-18Do not even try to dereference a NULL pointer.Martin Pieuchot
2015-01-18Complete synchronous abort method modeled after the existing ones.Martin Pieuchot
2015-01-18Do not try to free xfers before aborting the pipes, otherwise a clownMartin Pieuchot
2015-01-18Add commented-out necsb(4) and audio(4) configuration to GENERIC.Kenji Aoyama
2015-01-18Revert 1.166 (but keep the bufq_wait() interface change), for this is wrongMiod Vallat
2015-01-18The 'mrs' instruction only deals with the whole register withoutJonathan Gray
2015-01-18Since we are no longer resetting rings when a Babble or Stall conditionMartin Pieuchot
2015-01-18The 'mrs' instruction only deals with the whole register withoutJonathan Gray
2015-01-18unifdef IPKDB. These codepaths are holdouts from NetBSD code and areJonathan Gray
2015-01-18Per POSIX, <sys/socket.h> needs to expose struct iovec, and may do soPhilip Guenther
2015-01-17Split the consumer & producer logic into two different functions inMartin Pieuchot
2015-01-17use NGROUPS_MAXTheo de Raadt
2015-01-17guard usingacpi variable inside NACPI. from oyvind jaegtnesTed Unangst
2015-01-17Define MAXHOSTNAMELEN as HOST_NAME_MAX+1. OK deraadt@Todd C. Miller
2015-01-17Delete an extra tab.Kenji Aoyama
2015-01-17Add an ascii bit/field diagram for armv7-a psr to match theJonathan Gray
2015-01-17resume_randomness() before spinning up other cpus...Theo de Raadt
2015-01-17More complete gpio reset sequences for SABRE Lite/Nitrogen6X.Jonathan Gray
2015-01-16increase namecache to maxvnodes again now that the n^2 loop is no more.Ted Unangst
2015-01-16Revert forcing a rachitic `c' slice size when the drive geometry is not known,Miod Vallat
2015-01-16Don't bother to initialize %g1; we don't do this on sparc either.Mark Kettenis
2015-01-16disklabel_sun_to_bsd() will nicely set the disk size if it is zero, but it isMiod Vallat
2015-01-16increasing the size of the namecache suddenly made the commentTed Unangst
2015-01-16Less code, more better. No longer need to worry about what mysteriousTed Unangst
2015-01-16Binary code patching on amd64Stefan Fritsch
2015-01-16Move to using PATH_MAXTheo de Raadt
2015-01-15include <sys/syslimits.h> and switch to use of LOGIN_NAME_MAX, makingTheo de Raadt
2015-01-15include <sys/syslimits.h> to get access to (and then use) NGROUPS_MAXTheo de Raadt
2015-01-15Stop including <sys/param.h> and other stuff. Assume this will be includedTheo de Raadt
2015-01-15Do not include 5 header files, including <sys/param.h>Theo de Raadt
2015-01-15remove static version stringTed Unangst
2015-01-15simplify des headers by stripping out all the unnecessary userland bitsTed Unangst
2015-01-15Remove a variable no longer used in re_tick().Brad Smith
2015-01-15Rework the disk label handling once again. Disk label is read at attach timeMiod Vallat