summaryrefslogtreecommitdiffstats
path: root/sys/vm
Commit message (Expand)AuthorAgeFilesLines
* Don't bother calling pmap_create() when creating the temporary map.dg1995-02-151-2/+2
* YF fix.phk1995-02-142-7/+8
* YF Fix.phk1995-02-141-3/+1
* Fixed problem with msync causing a panic.dg1995-02-141-17/+19
* Carefully choose the value for vm_object_cache_max. The previous calculationdg1995-02-121-2/+5
* Killed MACHVMCOMPAT function prototypes as the functions don't exist.dg1995-02-091-5/+1
* Killed MACHVMCOMPAT code. It doesn't compile, and in its present statedg1995-02-091-100/+1
* Minor algorithmic adjustments that reduce the CPU consumption of thedg1995-02-091-72/+12
* Fixed bmap run-length brokeness.dg1995-02-031-62/+53
* swap_pager.c:dg1995-02-0210-287/+373
* Completed the fix for attempting to page out pages via the device_pager.dg1995-01-281-1/+4
* Use the VM_PAGE_BITS_ALL in a place it can be used.dg1995-01-261-4/+4
* Don't attempt to clean device_pager backed objects at terminate time.dg1995-01-251-2/+3
* Added ability to detect sequential faults and DTRT. (swap_pager.c)dg1995-01-2411-77/+104
* Moved some splx's down a few lines in vm_page_insert and vm_page_removedg1995-01-151-3/+3
* Protect a qcollapse call with an object lock before calling. The locksdg1995-01-131-8/+13
* Improve my previous change to use the same tests as are used in qcollapse.dg1995-01-111-2/+3
* Fixed a panic that Garrett reported to me...the OBJ_INTERNAL flag wasn'tdg1995-01-112-5/+17
* Kill VM_PAGE_INIT macro as it is only used once and makes the code moredg1995-01-102-22/+15
* Fixed some formatting weirdness that I overlooked in the previous commit.dg1995-01-1011-60/+60
* These changes embody the support of the fully coherent merged VM buffer cache,dg1995-01-0934-4710/+4582
* Make sure that the object being collapsed doesn't go away on us...bydg1995-01-051-3/+13
* Submitted by: Ben Jacksonats1995-01-021-2/+2
* Clean up previous commits (format for 80 columns...).bde1994-12-301-7/+8
* Do vm_page_rename more conservatively in rcollapse and qcollapse, anddg1994-12-231-4/+9
* Initialize b_vnbuf.le_next before returning a new buffer in getpbuf anddg1994-12-232-4/+8
* Fixed a benign off by one error.dg1994-12-221-2/+2
* Don't ever clear B_BUSY on a pbuf (or any other flag for that matter).dg1994-12-192-7/+5
* Fixed multiple bogons with the map entry handling.dg1994-12-181-8/+8
* Fixed bug where statically allocated map entries might be freed to thedg1994-12-181-4/+7
* Change swapping policy to be a bit more aggressive about finding adg1994-12-181-5/+6
* Protect kmem_map modifications with splhigh() to work around a problem withdg1994-12-151-3/+19
* Don't put objects that have no parent on the reverse_shadow_list. Problemdg1994-12-111-2/+3
* These changes fix a couple of lingering VM problems:dg1994-11-252-107/+4
* Don't try to page to a vnode that had it's filesystem unmounted.dg1994-11-241-7/+23
* Preallocate the first swap block to work around a failure with swap startingdg1994-11-221-2/+2
* Don't ever try to kill off process 1 - even if we are out of swap spacedg1994-11-171-2/+3
* Remove a peice of commented out code that was left over from the earlygibbs1994-11-171-3/+2
* pmap.h:bde1994-11-142-8/+9
* Set laundry flag when transitioning an inactive page from clean to dirty.dg1994-11-141-2/+4
* Fixed bug where a read-behind to a negative offset would occur if thedg1994-11-132-19/+25
* Fixed bugs in accounting of swap space that resulted in the pager thinkingdg1994-11-131-62/+19
* Implemented swap locking via P_SWAPPING flag. It was possible for a processdg1994-11-131-5/+11
* Fixed return status from pagers. Ahem...the previous method would manufacturedg1994-11-063-17/+25
* Added support for starting the experimental "vmdaemon" system process.dg1994-11-063-80/+358
* Improved I/O error reporting.dg1994-10-251-5/+19
* #if 0'd out the object cache trimming code - there are multiple waysdg1994-10-251-1/+3
* Fixed object cache trimming policy so it actually works.dg1994-10-231-7/+5
* Adjusted reserved levels to fix a deadlock condition.dg1994-10-231-3/+3
* Changed a thread_sleep into an spl protected tsleep. A deadlock can occurdg1994-10-232-9/+12
OpenPOWER on IntegriCloud