summaryrefslogtreecommitdiff
path: root/sys
AgeCommit message (Collapse)Author
1997-12-10emulate linux SIOCGIFHWADDR; stolen from freebsd; not tested yetTheo de Raadt
1997-12-10insert forgotten copyright + handle eject ioctl.Niels Provos
1997-12-09syncTheo de Raadt
1997-12-09eolTheo de Raadt
1997-12-09fix O_TRUNC and O_{SH,EX}LOCK properly; problem found by provosTheo de Raadt
1997-12-09when downgrading a filesystem to read-only, use softdep_flushfiles; mckusickTheo de Raadt
1997-12-09Intel P5 f00f workaround; weingart & who knows who elseTheo de Raadt
1997-12-08add kern.nosuidcoredump to lock things out even moreTheo de Raadt
1997-12-07handle linux cdrom ioctls. not that i play quake that often but it wasNiels Provos
the only binary around.
1997-12-06goodbye TIMEZONE and DSTkstailey
1997-12-05argh. this darn little bugger made CDIOCREADSUBCHANNEL unuseable, thisNiels Provos
should also fix some cdio problems with SCSI cdroms.
1997-12-04vfs lite2 changes things...Theo de Raadt
1997-12-03Move "struct head header" in struct ncb down to where it used toTodd C. Miller
be. Found by Ken, fixed by Niklas.
1997-12-03DST changekstailey
1997-12-03Prevent IPF and NAT configuration changes when securelevel > 1.kstailey
1997-12-03fix printfs in apm_power_printJim Rees
1997-12-02forgot to commit this too.Todd C. Miller
1997-12-02Fix deadlock in cd9660 code. NetBSD PR#2262Constantine Sapuntzakis
1997-12-02Unlock child temporarily while truncating directory. This preventsConstantine Sapuntzakis
a double locking bug in the soft updates code.
1997-12-02More splbio()'s added so that reassignbuf can do its thing.Constantine Sapuntzakis
1997-12-02Protect some more vnode operation with splbio. Needed for correct interactionConstantine Sapuntzakis
with soft updates code.
1997-12-02once moreMichael Shalayeff
1997-12-02Bye bye Innaccurate misleading fileTheo de Raadt
1997-12-02syncTheo de Raadt
1997-12-02If i386, and COMPAT_LINUX is enabled, if setegid() is called with an outTheo de Raadt
of range value recognize it as the Linux personality() system call, and switch into Linux emulation mode.... At least they could have picked an unused system call number!
1997-12-01Does not compile. Obviously untested code. RevertedTheo de Raadt
1997-12-01Newer ncr driver from NetBSD minus the scsipi stuff and withTodd C. Miller
changes to make it diff usefully against the FreeBSD driver. Tested on alpha and i386, not tested on powerpc.
1997-12-01rearrange memory computations, remove bogus MEM_COMPUTE hackMichael Shalayeff
1997-11-30use libkern's bcd convsMichael Shalayeff
1997-11-30missed, i'm 'what?', dunno, missed anywaysMichael Shalayeff
1997-11-30big diskinfo changes from toby@Michael Shalayeff
1997-11-30missed from last bcd2bin commitMichael Shalayeff
1997-11-30fix TIOCGPGRP in sunos emulationTheo de Raadt
1997-11-30use stackgap macrosTheo de Raadt
1997-11-30add bcd<-->bin conversions; from freebsdMichael Shalayeff
1997-11-30config apm even if power mgmt is disengagedJim Rees
1997-11-30Move in_addr_t and in_port_t to <sys/types.h> and add sa_family_tTodd C. Miller
and suseconds_t types for XPG4.2.
1997-11-30Pull in vm/vm.h in the appropriate order for simple_dat_lock_t in vm/pmap.hgene
1997-11-30verbose eisa/pciTheo de Raadt
1997-11-30vm/lock.h -> sys/lock.hgene
1997-11-30Canonify the _*_H_ namespace appropriately.gene
Also, clean up comments and wrap with _KERNEL correctly.
1997-11-30hmm. If out of ports, return EADDRNOTAVAILTheo de Raadt
1997-11-30Add multiple inclusion protection, clean up comments, and wrap with _KERNELgene
where appropriate. Note that this buys us notable speed boosts on a machine as slow as the mac68k.
1997-11-30add bios0 etcTheo de Raadt
1997-11-28one les tabMichael Shalayeff
1997-11-28Pull in header files needed to get definition of u_lbn_t ingene
ufs/ffs/ffs_extern.h. This must be fallout from the lite2 merge. Compiles again.
1997-11-27oh, connect() semantics are EVEN WORSETheo de Raadt
1997-11-26Optimized to perform well on small interleaves which makes the ccd driverNiklas Hallqvist
really useful as a performance booster. The interleave needs to be a multiple of the software page size (CLBYTES) and not have mirror components if the new optimized policy should be used, otherwise the old policy will be used still.
1997-11-26enforce VM_PROT_READ in mmapTheo de Raadt
1997-11-26fix shmctl IPC_STATTheo de Raadt