summaryrefslogtreecommitdiffstats
path: root/sys/vm
Commit message (Expand)AuthorAgeFilesLines
* MFC r303773alc2016-08-101-1/+1
* MFC r303448:kib2016-08-101-40/+30
* MFC r303492alc2016-08-051-1/+0
* MFC r303446:kib2016-08-041-4/+4
* MFC r303356 and r303465alc2016-08-011-15/+12
* MFC r302567:kib2016-07-251-0/+4
* MFC r303101alc2016-07-231-0/+9
* MFC r302980alc2016-07-211-39/+76
* Change the type of the map entry's next_read field from a vm_pindex_t to aalc2016-07-073-10/+10
* Autotune the number of pages set aside for UMA startup based on the numbercperciva2016-07-072-0/+17
* Replace a number of conflations of mp_ncpus and mp_maxid with eithernwhitehorn2016-07-062-3/+4
* Clarify the vnode_destroy_vobject() logic handling for already terminatedkib2016-07-051-3/+14
* Change type of the 'dead' variable to boolean.kib2016-07-031-2/+2
* If the vm_fault() handler raced with the vm_object_collapse()kib2016-06-271-4/+11
* In vm_page_xunbusy_maybelocked(), add fast path for unbusy when nokib2016-06-231-4/+22
* Add a comment noting locking regime for vm_page_xunbusy().kib2016-06-231-0/+1
* Fix a LOR between vnode locks and allproc_lock.kib2016-06-222-15/+34
* The vmtotal sysctl handler marks active vm objects to calculatekib2016-06-212-46/+37
* Fix inconsistent locking of the swap pager named objects list.kib2016-06-131-63/+56
* Explicitely initialize sw_alloc_sx. Currently it is not initializedkib2016-06-131-0/+1
* Reset the page busy lock state after failing to insert into the object.markj2016-06-021-0/+2
* Don't preserve the page's object linkage in vm_page_insert_after().markj2016-06-021-6/+2
* Fix memguard(9) in kernels with INVARIANTS enabled.markj2016-06-012-9/+30
* If the fast path unbusy in vm_page_replace() fails, slow path needs tokib2016-06-011-15/+20
* Do not leak the vm object lock when swap reservation failed, inkib2016-05-291-0/+1
* The flag "vm_pages_needed" has long served two distinct purposes: (1) toalc2016-05-273-42/+63
* Use vm_page_replace_checked() instead of vm_page_rename() for implementingalc2016-05-271-17/+10
* Prevent parallel object collapses. Both vm_object_collapse_scan() andkib2016-05-261-0/+5
* Style changes to some most outrageous violations in vm_object_collapse().kib2016-05-261-9/+6
* In vm_page_cache(), only drop the vnode after radix insert failurekib2016-05-241-1/+2
* In vm_page_alloc_contig(), on vm_page_insert() failure, mark eachkib2016-05-241-1/+3
* Remove Giant around allocation of the swap pager with non-NULL handle.kib2016-05-241-3/+0
* Correct an error in a comment: One of the conditions for page allocationalc2016-05-231-9/+8
* Mark swap-related proc sysctls as not requiring Giant.kib2016-05-221-3/+5
* Replace hand-made exclusive lock, protecting against parallelkib2016-05-221-24/+10
* Remove false claim. Giant is dropped by mi_startup() before passingkib2016-05-221-2/+0
* When descending a shadow chain of objects, it makes no sense to updatealc2016-05-211-1/+2
* Add new bus methods for mapping resources.jhb2016-05-201-2/+3
* Clean up the handling of errors from vm_pager_get_pages(). Mostly, thisalc2016-05-191-26/+18
* vm/vm_page.h: Fix trivial '-Wpointer-sign' warningcem2016-05-192-3/+3
* Add implementation of robust mutexes, hopefully close enough to thekib2016-05-173-1/+3
* Move vm_domain_rr_selectdomain() under #ifdef VM_NUMA_ALLOC.jhb2016-05-101-4/+2
* sys/vm: minor spelling fixes in comments.pfg2016-05-029-16/+16
* Avoid duplicated calls to pmap_page_get_memattr().kib2016-05-011-5/+13
* Don't require write locks on the VM object for vm_page_prev/next.jhb2016-04-291-2/+2
* Trim redundant message.jhb2016-04-271-1/+1
* Cleanup redundant parenthesis from existing howmany()/roundup() macro uses.pfg2016-04-221-1/+1
* sys: use our roundup2/rounddown2() macros when param.h is available.pfg2016-04-213-10/+10
* Remove slightly used const values that can be replaced with nitems().pfg2016-04-211-3/+1
* Add more fine-grained kernel options for NUMA support.jhb2016-04-094-14/+39
OpenPOWER on IntegriCloud