summaryrefslogtreecommitdiff
path: root/sys/kern
AgeCommit message (Expand)Author
2009-08-08two things:Bob Beck
2009-08-06reintroduce the uvm_tree commit.Owain Ainsworth
2009-08-03timeout_add -> timeout_add_msecBret Lambert
2009-08-02Dynamic buffer cache support - a re-commit of what was backed outBob Beck
2009-07-30We enable POOL_DEBUG (except in a release)Theo de Raadt
2009-07-21Handle the case where so_pcb is NULL.Todd C. Miller
2009-07-19Hook up "text" (executable) output and implement for fuser too.Todd C. Miller
2009-07-19clalloc() can't fail, so there's no need to handle failure cases.Bret Lambert
2009-07-09Remove the VREF() macro and replaces all instances with a call to verf(),Thordur I. Bjornsson
2009-06-25backout the buf_acquire() does the bremfree() since all callersThordur I. Bjornsson
2009-06-24Remove a semicolon at the end of an if statement that made the threadJonathan Gray
2009-06-24Remove extra psignal/wakeup in exit1() which can cause the parent toKurt Miller
2009-06-24turn off POOL_DEBUG as we go into release; pointed out by mpfTheo de Raadt
2009-06-22sync m_inithdr() with m_gethdr() after theThordur I. Bjornsson
2009-06-17Revert bufq's. this is inline with the major midlayer reverts thatThordur I. Bjornsson
2009-06-16Backout all changes to uvm after pmemrange (which will be backed outOwain Ainsworth
2009-06-15No you cannot remove a sysctl mib once you add it. It becomes ABI. ThatTheo de Raadt
2009-06-15syncTheo de Raadt
2009-06-15Back out all the buffer cache changes I committed during c2k9. This reverts t...Bob Beck
2009-06-14Don't mess with the bounds in the lp if *partoffp is non-NULL, since thisTheo de Raadt
2009-06-12rework pool_get() a bit so that if you call if with a constructor setOwain Ainsworth
2009-06-09Fix userland regression tests by providing a #define for pool_setipl()Owain Ainsworth
2009-06-08must calculate iovlen or ktrace panics the machine; diff from blambertTheo de Raadt
2009-06-08some vnodes have a NULL mountpoint, so strlcpy of the f_mntonname if not advisedTheo de Raadt
2009-06-07Add KERN_FILE2 sysctl analogous to KERN_PROC2 but for file structures,Todd C. Miller
2009-06-06We need miod's no-coredumps-on-ramdisks diff, because we have grown theTheo de Raadt
2009-06-06All caller of buf_acquire were doing bremfree before the call.Artur Grabowski
2009-06-05abort any time adjustment in progress if the clock is set with settimeofdayChris Kuethe
2009-06-05Dynamic buffer cache sizing.Bob Beck
2009-06-05%zu is not meant for size_t in the kernel. Change to %lu.Ray Lai
2009-06-05Clamp the ending bound to the size of the disk. This makes disklabel -ATheo de Raadt
2009-06-05Initial support for routing domains. This allows to bind interfaces toClaudio Jeker
2009-06-04obvious shrinkage.Owain Ainsworth
2009-06-04Recycle four ancient fields in the disklabel structure, replacing them withTheo de Raadt
2009-06-04add $OpenBSD$ tagThordur I. Bjornsson
2009-06-04POOL_DEBUG and DIAGNOSTIC should be better friendsTheo de Raadt
2009-06-04the POOL_DEBUG checks needed to be more friendly with DIAGNOSTICTheo de Raadt
2009-06-04enable POOL_DEBUG again just for the hackathon.Owain Ainsworth
2009-06-04unfuck msleep - fixed by art and ariane after much horror and teeth gnashingBob Beck
2009-06-04don't use splvm ourselves, use pool_setipl on the pool we're allocatingOwain Ainsworth
2009-06-04Put readv/writev changes back in, as they no longer hang ckuethe's ntpd.Bret Lambert
2009-06-03add a flexible buffer queue (bufq) api, based on the never usedThordur I. Bjornsson
2009-06-03add kern.bufcachepercent sysctl to allow adjusting the buffer cacheBob Beck
2009-06-03xfs -> nnpfsThordur I. Bjornsson
2009-06-03syncJanne Johansson
2009-06-03remove bad #ifdef from syscall.masterJanne Johansson
2009-06-03Arla client rename from xfs to nnpfs for later upgrades. Tested on various ar...Janne Johansson
2009-06-03Add a callback specifically for softraid. If you touch this kittens willMarco Peereboom
2009-06-03Change bufhash from the old grotty hash table to red-black trees hangingBob Beck
2009-06-03Revert readv/writev changes, as they trigger an apparent file descriptorBret Lambert