xref: /netbsd-src/share/man/man9/uvm.9 (revision 89c5a767f8fc7a4633b2d409966e2becbb98ff92)
1.\"	$NetBSD: uvm.9,v 1.7 1999/05/06 12:04:50 hwr Exp $
2.\"
3.\" Copyright (c) 1998 Matthew R. Green
4.\" All rights reserved.
5.\"
6.\" Redistribution and use in source and binary forms, with or without
7.\" modification, are permitted provided that the following conditions
8.\" are met:
9.\" 1. Redistributions of source code must retain the above copyright
10.\"    notice, this list of conditions and the following disclaimer.
11.\" 2. Redistributions in binary form must reproduce the above copyright
12.\"    notice, this list of conditions and the following disclaimer in the
13.\"    documentation and/or other materials provided with the distribution.
14.\" 3. The name of the author may not be used to endorse or promote products
15.\"    derived from this software without specific prior written permission.
16.\"
17.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
18.\" IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
19.\" OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
20.\" IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
21.\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
22.\" BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
23.\" LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
24.\" AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
25.\" OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27.\" SUCH DAMAGE.
28.\"
29.\" XXX this manual sets nS to 1 or 0 in the description, to obtain
30.\" synopsis-like function prototypes.  any better way?
31.\"
32.Dd April 10, 1998
33.Dt UVM 9
34.Os
35.Sh NAME
36.Nm UVM virtual memory system
37.Nd external interface
38.Sh SYNOPSIS
39.Fd #include <sys/param.h>
40.Fd #include <vm/vm.h>
41.Fd #include <uvm/uvm.h>
42.Sh DESCRIPTION
43The UVM virtual memory system manages access to the computer's memory
44resources.  User processes and the kernel access these resources through
45UVM's exteral interface.  UVM's external interface includes functions that:
46.Pp
47.Bl -hyphen -compact
48.It
49initialise UVM sub-systems
50.It
51manage virtual address spaces
52.It
53resolve page faults
54.It
55memory map files and devices
56.It
57perform uio-based I/O to virtual memory
58.It
59allocate and free kernel virtual memory
60.It
61allocate and free physical memory
62.El
63.Pp
64In addition to exporting these services, UVM has two kernel-level processes:
65pagedaemon and swapper.   The pagedaemon process sleeps until physical memory
66becomes scarce.  When that happens, pagedaemon is awoken.   It scans physical
67memory, paging out and freeing memory that has not been recently used.  The
68swapper process swaps in runnable processes that are currently swapped out,
69if there is room.
70.Pp
71There are also several miscellaneous functions.
72
73.Pp
74.Sh INITIALISATION
75
76.nr nS 1
77.Pp
78.Ft void
79.Fn uvm_init
80.Ft void
81.Fn uvm_init_limits "struct proc *p"
82.Ft void
83.Fn uvm_setpagesize
84.Ft void
85.Fn uvm_swap_init
86.nr nS 0
87
88.Pp
89.Fn uvm_init
90sets up the UVM system at system boot time, after the
91copyright has been printed.  It initialises
92global state, the page, map, kernel virtual memory state,
93machine-dependent physical map, kernel memory allocator,
94pager and anonymous memory sub-systems, and then enables
95paging of kernel objects.
96
97.Pp
98.Fn uvm_init_limits
99initialises process limits for the named process.  This is for use by
100the system startup for process zero, before any other processes are
101created.
102
103.Pp
104.Fn uvm_setpagesize
105initialises the uvmexp members pagesize (if not already done by
106machine-dependent code), pageshift and pagemask.  It should be called by
107machine-dependent code early in the
108.Xr pmap_init 9
109call.
110
111.Pp
112.Fn uvm_swap_init
113initialises the swap sub-system.
114
115.Pp
116.Sh VIRTUAL ADDRESS SPACE MANAGEMENT
117
118.Pp
119.nr nS 1
120.Ft int
121.Fn uvm_map "vm_map_t map" "vaddr_t *startp" "vsize_t size" "struct uvm_object *uobj" "vaddr_t uoffset" "uvm_flag_t flags"
122.Ft int
123.Fn uvm_map_pageable "vm_map_t map" "vaddr_t start" "vaddr_t end" "boolean_t new_pageable"
124.Ft boolean_t
125.Fn uvm_map_checkprot "vm_map_t map" "vaddr_t start" "vaddr_t end" "vm_prot_t protection"
126.Ft int
127.Fn uvm_map_protect "vm_map_t map" "vaddr_t start" "vaddr_t end" "vm_prot_t new_prot" "boolean_t set_max"
128.Ft int
129.Fn uvm_deallocate "vm_map_t map" "vaddr_t start" "vsize_t size"
130
131.Ft struct vmspace *
132.Fn uvmspace_alloc "vaddr_t min" "vaddr_t max" "int pageable"
133.Ft void
134.Fn uvmspace_exec "struct proc *p"
135.Ft struct vmspace *
136.Fn uvmspace_fork "struct vmspace *vm"
137.Ft void
138.Fn uvmspace_free "struct vmspace *vm1"
139.Ft void
140.Fn uvmspace_share "struct proc *p1" "struct proc *p2"
141.Ft void
142.Fn uvmspace_unshare "struct proc *p"
143.nr nS 0
144
145.Pp
146.Fn uvm_map
147establishes a valid mapping in map
148.Fa map ,
149which must be unlocked.  The new mapping has size
150.Fa size ,
151which must be in
152.Dv PAGE_SIZE
153units.  The
154.Fa uobj
155and
156.Fa uoffset
157arguments can have four meanings.  When
158.Fa uobj
159is
160.Dv NULL
161and
162.Fa uoffset
163is
164.Dv UVM_UNKNOWN_OFFSET ,
165.Fn uvm_map
166does not use the machine-dependant
167.Dv PMAP_PREFER
168function.  If
169.Fa uoffset
170is any other value, it is used as the hint to
171.Dv PMAP_PREFER .
172When
173.Fa uobj
174is not
175.Dv NULL
176and
177.Fa uoffset
178is
179.Dv UVM_UNKNOWN_OFFSET ,
180.Fn uvm_map
181finds the offset based upon the virtual address, passed as
182.Fa startp .
183If
184.Fa uoffset
185is any other value, we are doing a normal mapping at this offset.  The
186start address of the map will be returned in
187.Fa startp .
188.Pp
189.Fa flags
190passed to
191.Fn uvm_map
192are typically created using the
193.Fn UVM_MAPFLAG "vm_prot_t prot" "vm_prot_t maxprot" "vm_inherit_t inh" "int advice" "int flags"
194macro, which uses the following values.
195The
196.Fa prot
197and
198.Fa maxprot
199can take are:
200.Bd -literal
201#define UVM_PROT_MASK   0x07    /* protection mask */
202#define UVM_PROT_NONE   0x00    /* protection none */
203#define UVM_PROT_ALL    0x07    /* everything */
204#define UVM_PROT_READ   0x01    /* read */
205#define UVM_PROT_WRITE  0x02    /* write */
206#define UVM_PROT_EXEC   0x04    /* exec */
207#define UVM_PROT_R      0x01    /* read */
208#define UVM_PROT_W      0x02    /* write */
209#define UVM_PROT_RW     0x03    /* read-write */
210#define UVM_PROT_X      0x04    /* exec */
211#define UVM_PROT_RX     0x05    /* read-exec */
212#define UVM_PROT_WX     0x06    /* write-exec */
213#define UVM_PROT_RWX    0x07    /* read-write-exec */
214.Ed
215.Pp
216The values that
217.Fa inh
218can take are:
219.Bd -literal
220#define UVM_INH_MASK    0x30    /* inherit mask */
221#define UVM_INH_SHARE   0x00    /* "share" */
222#define UVM_INH_COPY    0x10    /* "copy" */
223#define UVM_INH_NONE    0x20    /* "none" */
224#define UVM_INH_DONATE  0x30    /* "donate" << not used */
225.Ed
226.Pp
227The values that
228.Fa advice
229can take are:
230.Bd -literal
231#define UVM_ADV_NORMAL  0x0     /* 'normal' */
232#define UVM_ADV_RANDOM  0x1     /* 'random' */
233#define UVM_ADV_SEQUENTIAL 0x2  /* 'sequential' */
234#define UVM_ADV_MASK    0x7     /* mask */
235.Ed
236.Pp
237The values that
238.Fa flags
239can take are:
240.Bd -literal
241#define UVM_FLAG_FIXED   0x010000 /* find space */
242#define UVM_FLAG_OVERLAY 0x020000 /* establish overlay */
243#define UVM_FLAG_NOMERGE 0x040000 /* don't merge map entries */
244#define UVM_FLAG_COPYONW 0x080000 /* set copy_on_write flag */
245#define UVM_FLAG_AMAPPAD 0x100000 /* for bss: pad amap to reduce malloc() */
246#define UVM_FLAG_TRYLOCK 0x200000 /* fail if we can not lock map */
247.Ed
248.Pp
249The
250.Dv UVM_MAPFLAG
251macro arguments can be combined with an or operator.  There are
252several special purpose macros for checking protection combinations, e.g. the
253.Dv UVM_PROT_WX
254macro.  There are also some additional macros to extract bits from
255the flags.  The
256.Dv UVM_PROTECTION ,
257.Dv UVM_INHERIT ,
258.Dv UVM_MAXPROTECTION
259and
260.Dv UVM_ADVICE
261macros return the protection, inheritance, maximum protection and advice,
262respectively.
263.Fn uvm_map
264returns a standard UVM return value.
265
266.Pp
267.Fn uvm_map_pageable
268changes the pageability of the pages in the range from
269.Fa start
270to
271.Fa end
272in map
273.Fa map
274to
275.Fa new_pageable .
276.Fn uvm_map_pageable
277returns a standard UVM return value.
278
279.Pp
280.Fn uvm_map_checkprot
281checks the protection of the range from
282.Fa start
283to
284.Fa end
285in map
286.Fa map
287against
288.Fa protection .
289This returns either
290.Dv TRUE
291or
292.Dv FALSE .
293
294.Pp
295.Fn uvm_map_protect
296changes the protection
297.Fa start
298to
299.Fa end
300in map
301.Fa map
302to
303.Fa new_prot ,
304also setting the maximum protection to the region to
305.Fa new_prot
306if
307.Fa set_max
308is non-zero.  This function returns a standard UVM return value.
309
310.Pp
311.Fn uvm_deallocate
312deallocates kernel memory in map
313.Fa map
314from address
315.Fa start
316to
317.Fa start + size .
318
319.Pp
320.Fn uvmspace_alloc
321allocates and returns a new address space, with ranges from
322.Fa min
323to
324.Fa max ,
325setting the pageability of the address space to
326.Fa pageable.
327
328.Pp
329.Fn uvmspace_exec
330either reuses the address space of process
331.Fa p
332if there are no other references to it, or creates
333a new one with
334.Fn uvmspace_alloc .
335
336.Pp
337.Fn uvmspace_fork
338creates and returns a new address space based upon the
339.Fa vm1
340address space, typically used when allocating an address space for a
341child process.
342
343.Pp
344.Fn uvmspace_free
345lowers the reference count on the address space
346.Fa vm ,
347freeing the data structures if there are no other references.
348
349.Pp
350.Fn uvmspace_share
351causes process
352.Pa p2
353to share the address space of
354.Fa p1 .
355
356.Pp
357.Fn uvmspace_unshare
358ensures that process
359.Fa p
360has its own, unshared address space, by creating a new one if
361necessary by calling
362.Fn uvmspace_fork .
363
364.Pp
365.Sh PAGE FAULT HANDLING
366
367.Pp
368.nr nS 1
369.Ft int
370.Fn uvm_fault "vm_map_t orig_map" "vaddr_t vaddr" "vm_fault_t fault_type" "vm_prot_t access_type"
371.nr nS 0
372.Pp
373.Fn uvm_fault
374is the main entry point for faults.  It takes
375.Fa orig_map
376as the map the fault originated in, a
377.Fa vaddr
378offset into the map the fault occured,
379.Fa fault_type
380describing the type of fault, and
381.Fa access_type
382describing the type of access requested.
383.Fn uvm_fault
384returns a standard UVM return value.
385
386.Pp
387.Sh MEMORY MAPPING FILES AND DEVICES
388
389.Pp
390.nr nS 1
391.Ft struct uvm_object *
392.Fn uvn_attach "void *arg" "vm_prot_t accessprot"
393.Ft void
394.Fn uvm_vnp_setsize "struct vnode *vp" "u_quad_t newsize"
395.Ft void
396.Fn uvm_vnp_sync "struct mount *mp"
397.Ft void
398.Fn uvm_vnp_terminate "struct vnode *vp"
399.Ft boolean_t
400.Fn uvm_vnp_uncache "struct vnode *vp"
401.nr nS 0
402
403.Pp
404.Fn uvn_attach
405attaches a UVM object to vnode
406.Fa arg ,
407creating the object if necessary.  The object is returned.
408
409.Pp
410.Fn uvm_vnp_setsize
411sets the size of vnode
412.Fa vp
413to
414.Fa newsize .
415Caller must hold a reference to the vnode.  If the vnode shrinks, pages
416no longer used are discarded.  This function will be removed when the
417filesystem and VM buffer caches are merged.
418
419.Pp
420.Fn uvm_vnp_sync
421flushes dirty vnodes from either the mount point passed in
422.Fa mp ,
423or all dirty vnodes if
424.Fa mp
425is
426.Dv NULL .
427This function will be removed when the filesystem and VM buffer caches
428are merged.
429
430.Pp
431.Fn uvm_vnp_terminate
432frees all VM resources allocated to vnode
433.Fa vp .
434If the vnode still has references, it will not be destroyed; however
435all future operations using this vnode will fail.  This function will be
436removed when the filesystem and VM buffer caches are merged.
437
438.Pp
439.Fn uvm_vnp_uncache
440disables vnode
441.Fa vp
442from persisting when all references are freed.  This function will be
443removed when the file-system and UVM caches are unified.  Returns
444true if there is no active vnode.
445
446.Pp
447.Sh VIRTUAL MEMORY I/O
448
449.Pp
450.nr nS 1
451.Ft int
452.Fn uvm_io "vm_map_t map" "struct uio *uio"
453.nr nS 0
454
455.Pp
456.Fn uvm_io
457performs the I/O described in
458.Fa uio
459on the memory described in
460.Fa map .
461
462.Pp
463.Sh ALLOCATION OF KERNEL MEMORY
464
465.Pp
466.nr nS 1
467.Ft vaddr_t
468.Fn uvm_km_alloc "vm_map_t map" "vsize_t size"
469.Ft vaddr_t
470.Fn uvm_km_zalloc "vm_map_t map" "vsize_t size"
471.Ft vaddr_t
472.Fn uvm_km_alloc1 "vm_map_t map" "vsize_t size" "boolean_t zeroit"
473.Ft vaddr_t
474.Fn uvm_km_kmemalloc "vm_map_t map" "struct uvm_object *obj" "vsize_t size" "int flags"
475.Ft vaddr_t
476.Fn uvm_km_valloc "vm_map_t map" "vsize_t size"
477.Ft vaddr_t
478.Fn uvm_km_valloc_wait "vm_map_t map" "vsize_t size"
479.Ft struct vm_map *
480.Fn uvm_km_suballoc "vm_map_t map" "vaddr_t *min" "vaddr_t *max " "vsize_t size" "boolean_t pageable" "boolean_t fixed" "vm_map_t submap"
481.Ft void
482.Fn uvm_km_free "vm_map_t map" "vaddr_t addr" "vsize_t size"
483.Ft void
484.Fn uvm_km_free_wakeup "vm_map_t map" "vaddr_t addr" "vsize_t size"
485.nr nS 0
486
487.Pp
488.Fn uvm_km_alloc
489and
490.Fn uvm_km_zalloc
491allocate
492.Fa size
493bytes of wired kernel memory in map
494.Fa map .
495In addition to allocation,
496.Fn uvm_km_zalloc
497zeros the memory.  Both of these functions are defined as macros in
498terms of
499.Fn uvm_km_alloc1 ,
500and should almost always be used in preference to
501.Fn uvm_km_alloc1 .
502
503.Pp
504.Fn uvm_km_alloc1
505allocates and returns
506.Fa size
507bytes of wired memory in the kernel map, zeroing the memory if the
508.Fa zeroit
509argument is non-zero.
510
511.Pp
512.Fn uvm_km_kmemalloc
513allocates and returns
514.Fa size
515bytes of wired kernel memory into
516.Fa obj .
517The flags can be any of:
518.Bd -literal
519#define UVM_KMF_NOWAIT  0x1                     /* matches M_NOWAIT */
520#define UVM_KMF_VALLOC  0x2                     /* allocate VA only */
521#define UVM_KMF_TRYLOCK UVM_FLAG_TRYLOCK        /* try locking only */
522.Ed
523.Pp
524.Dv UVM_KMF_NOWAIT
525causes
526.Fn uvm_km_kmemalloc
527to return immediately if no memory is available.
528.Dv UVM_KMF_VALLOC
529causes no pages to be allocated, only a virtual address.
530.Dv UVM_KMF_TRYLOCK
531causes
532.Fn uvm_km_kmemalloc
533to use
534.Fn simple_lock_try
535when locking maps.
536
537.Pp
538.Fn uvm_km_valloc
539and
540.Fn uvm_km_valloc_wait
541return a newly allocated zero-filled address in the kernel map of size
542.Fa size .
543.Fn uvm_km_valloc_wait
544will also wait for kernel memory to become available, if there is a
545memory shortage.
546
547.Pp
548.Sh ALLOCATION OF PHYSICAL MEMORY
549
550.Pp
551.nr nS 1
552.Ft struct vm_page *
553.Fn uvm_pagealloc "struct uvm_object *uobj" "vaddr_t off" "struct vm_anon *anon"
554.Ft void
555.Fn uvm_pagerealloc "struct vm_page *pg" "struct uvm_object *newobj" "vaddr_t newoff"
556.Ft void
557.Fn uvm_pagefree "struct vm_page *pg"
558.Ft int
559.Fn uvm_pglistalloc "psize_t size" "paddr_t low" "paddr_t high" "paddr_t alignment" "paddr_t boundary" "struct pglist *rlist" "int nsegs" "int waitok"
560.Ft void
561.Fn uvm_pglistfree "struct pglist *list"
562.Ft void
563.Fn uvm_page_physload "vaddr_t start" "vaddr_t end" "vaddr_t avail_start" "vaddr_t avail_end"
564.nr nS 0
565
566.Pp
567.Fn uvm_pagealloc
568allocates a page of memory at virtual address
569.Fa off
570in either the object
571.Fa uobj
572or the anonymous memory
573.Fa anon ,
574which must be locked by the caller.  Only one of
575.Fa off
576and
577.Fa uobj
578can be non
579.Dv NULL .
580Returns
581.Dv NULL
582when no page can be found.
583
584.Pp
585.Fn uvm_pagerealloc
586reallocates page
587.Fa pg
588to a new object
589.Fa newobj ,
590at a new offset
591.Fa newoff .
592
593.Pp
594.Fn uvm_pagefree
595free's the physical page
596.Fa pg .
597
598.Pp
599.Fn uvm_pglistalloc
600allocates a list of pages for size
601.Fa size
602byte under various constraints.
603.Fa low
604and
605.Fa high
606describe the lowest and highest addresses acceptable for the list.  If
607.Fa alignment
608is non-zero, it describes the required alignment of the list, in
609power-of-two notation.  If
610.Fa boundary
611is non-zero, no segment of the list may cross this power-of-two
612boundary, relative to zero.
613The
614.Fa nsegs
615and
616.Fa waitok
617arguments are currently ignored.
618
619.Pp
620.Fn uvm_pglistfree
621frees the list of pages pointed to by
622.Fa list .
623
624.Pp
625.Fn uvm_page_physload
626loads physical memory segments into VM space.  It must be called at system
627boot time to setup physical memory management pages.  The arguments describe
628the
629.Fa start
630and
631.Fa end
632of the physical addresses of the segment, and the available start and end
633addresses of pages not already in use.
634.\" XXX expand on "system boot time"!
635
636.Pp
637.Fn uvm_km_suballoc
638allocates submap from
639.Fa map ,
640creating a new map if
641.Fa submap
642is
643.Dv NULL .
644The addresses of the submap can be specified exactly by setting the
645.Fa fixed
646argument to non-zero, which causes the
647.Fa min
648argument specify the beginning of the address in thes submap.  If
649.Fa fixed
650is zero, any address of size
651.Fa size
652will be allocated from
653.Fa map
654and the start and end addresses returned in
655.Fa min
656and
657.Fa max .
658If
659.Fa pageable
660is non-zero, entries in the map may be paged out.
661
662.Pp
663.Fn uvm_km_free
664and
665.Fn uvm_km_free_wakeup
666free
667.Fa size
668bytes of memory in the kernal map, starting at address
669.Fa addr .
670.Fn uvm_km_free_wakeup
671calls
672.Fn thread_wakeup
673on the map before unlocking the map.
674
675.Pp
676.Sh PROCESSES
677
678.Pp
679.nr nS 1
680.Ft void
681.Fn uvm_pageout
682.Ft void
683.Fn uvm_scheduler
684.Ft void
685.Fn uvm_swapin "struct proc *p"
686.Ft void
687
688.Pp
689.Fn uvm_pageout
690is the main loop for the page daemon.
691
692.Pp
693.Fn uvm_scheduler
694is the process zero main loop, which is to be called after the
695system has finished starting other processes.  It handles the
696swapping in of runnable, swapped out processes in priority
697order.
698
699.Pp
700.Fn uvm_swapin
701swaps in the named process.
702
703.Pp
704.Sh MISCELLANEOUS FUNCTIONS
705
706.nr nS 1
707.Pp
708.Ft struct uvm_object *
709.Fn uao_create "vsize_t size" "int flags"
710.Ft void
711.Fn uao_detach "struct uvm_object *uobj"
712.Ft void
713.Fn uao_reference "struct uvm_object *uobj"
714
715.Ft boolean_t
716.Fn uvm_chgkprot "caddr_t addr" "size_t len" "int rw"
717.Ft void
718.Fn uvm_kernacc "caddr_t addr" "size_t len" "int rw"
719.Ft boolean_t
720.Fn uvm_useracc "caddr_t addr" "size_t len" "int rw"
721
722.Ft void
723.Fn uvm_vslock "struct proc *p" "caddr_t addr" "size_t len"
724.Ft void
725.Fn uvm_vsunlock "struct proc *p" "caddr_t addr" "size_t len"
726
727.Ft void
728.Fn uvm_meter
729.Ft int
730.Fn uvm_sysctl "int *name" "u_int namelen" "void *oldp" "size_t *oldlenp" "void *newp " "size_t newlen" "struct proc *p"
731
732.Ft void
733.Fn uvm_fork "struct proc *p1" "struct proc *p2" "boolean_t shared"
734.Ft int
735.Fn uvm_grow "struct proc *p" "vaddr_t sp"
736.Ft int
737.Fn uvm_coredump "struct proc *p" "struct vnode *vp" "struct ucred *cred" "struct core *chdr"
738.nr nS 0
739
740.Pp
741The
742.Fn uao_create ,
743.Fn uao_detach
744and
745.Fn uao_reference
746functions operate on anonymous memory objects, such as those used to support
747System V shared memory.
748.Fn uao_create
749returns an object of size
750.Fa size
751with flags:
752.Bd -literal
753#define UAO_FLAG_KERNOBJ        0x1     /* create kernel object */
754#define UAO_FLAG_KERNSWAP       0x2     /* enable kernel swap */
755.Pp
756.Ed
757which can only be used once each at system boot time.
758.Fn uao_reference
759creates an additional reference to the named anonymous memory object.
760.Fn uao_detach
761removes a reference from the named anonymous memory object, destroying
762it if removing the last reference.
763
764.Pp
765.Fn uvm_chgkprot
766changes the protection of kernel memory from
767.Fa addr
768to
769.Fa addr + len
770to the value of
771.Fa rw .
772This is primarily useful for debuggers, for setting breakpoints.
773This function is only available with options
774.Dv KGDB .
775
776.Pp
777.Fn uvm_kernacc
778and
779.Fn uvm_useracc
780check the access at address
781.Fa addr
782to
783.Fa addr + len
784for
785.Fa rw
786access, in the kernel address space, and the current process'
787address space respectively.
788
789.Pp
790.Fn uvm_vslock
791and
792.Fn uvm_vsunlock
793control the wiring and unwiring of pages for process
794.Fa p
795from
796.Fa addr
797to
798.Fa addr + len .
799These functions are normally used to wire memory for I/O.
800
801.Pp
802.Fn uvm_meter
803calculates the load average and wakes up the swapper if necessary.
804
805.Pp
806.Fn uvm_sysctl
807provides support for the
808.Dv CTL_VM
809domain of the
810.Xr sysctl 3
811hierarchy.
812.Fn uvm_sysctl
813handles the
814.Dv VM_LOADAVG ,
815.Dv VM_METER
816and
817.Dv VM_UVMEXP
818calls, which return the current load averages, calculates current VM
819totals, and returns the uvmexp structure respectively.  The load averages
820are access from userland using the
821.Xr getloadavg 3
822function.  The uvmexp structure has all global state of the UVM system,
823and has the following members:
824.Bd -literal
825/* vm_page constants */
826int pagesize;   /* size of a page (PAGE_SIZE): must be power of 2 */
827int pagemask;   /* page mask */
828int pageshift;  /* page shift */
829
830/* vm_page counters */
831int npages;     /* number of pages we manage */
832int free;       /* number of free pages */
833int active;     /* number of active pages */
834int inactive;   /* number of pages that we free'd but may want back */
835int paging;     /* number of pages in the process of being paged out */
836int wired;      /* number of wired pages */
837int reserve_pagedaemon; /* number of pages reserved for pagedaemon */
838int reserve_kernel; /* number of pages reserved for kernel */
839
840/* pageout params */
841int freemin;    /* min number of free pages */
842int freetarg;   /* target number of free pages */
843int inactarg;   /* target number of inactive pages */
844int wiredmax;   /* max number of wired pages */
845
846/* swap */
847int nswapdev;   /* number of configured swap devices in system */
848int swpages;    /* number of PAGE_SIZE'ed swap pages */
849int swpginuse;  /* number of swap pages in use */
850int nswget;     /* number of times fault calls uvm_swap_get() */
851int nanon;      /* number total of anon's in system */
852int nfreeanon;  /* number of free anon's */
853
854/* stat counters */
855int faults;             /* page fault count */
856int traps;              /* trap count */
857int intrs;              /* interrupt count */
858int swtch;              /* context switch count */
859int softs;              /* software interrupt count */
860int syscalls;           /* system calls */
861int pageins;            /* pagein operation count */
862                        /* pageouts are in pdpageouts below */
863int swapins;            /* swapins */
864int swapouts;           /* swapouts */
865int pgswapin;           /* pages swapped in */
866int pgswapout;          /* pages swapped out */
867int forks;              /* forks */
868int forks_ppwait;       /* forks where parent waits */
869int forks_sharevm;      /* forks where vmspace is shared */
870
871/* fault subcounters */
872int fltnoram;   /* number of times fault was out of ram */
873int fltnoanon;  /* number of times fault was out of anons */
874int fltpgwait;  /* number of times fault had to wait on a page */
875int fltpgrele;  /* number of times fault found a released page */
876int fltrelck;   /* number of times fault relock called */
877int fltrelckok; /* number of times fault relock is a success */
878int fltanget;   /* number of times fault gets anon page */
879int fltanretry; /* number of times fault retrys an anon get */
880int fltamcopy;  /* number of times fault clears "needs copy" */
881int fltnamap;   /* number of times fault maps a neighbor anon page */
882int fltnomap;   /* number of times fault maps a neighbor obj page */
883int fltlget;    /* number of times fault does a locked pgo_get */
884int fltget;     /* number of times fault does an unlocked get */
885int flt_anon;   /* number of times fault anon (case 1a) */
886int flt_acow;   /* number of times fault anon cow (case 1b) */
887int flt_obj;    /* number of times fault is on object page (2a) */
888int flt_prcopy; /* number of times fault promotes with copy (2b) */
889int flt_przero; /* number of times fault promotes with zerofill (2b) */
890
891/* daemon counters */
892int pdwoke;     /* number of times daemon woke up */
893int pdrevs;     /* number of times daemon rev'd clock hand */
894int pdswout;    /* number of times daemon called for swapout */
895int pdfreed;    /* number of pages daemon freed since boot */
896int pdscans;    /* number of pages daemon scaned since boot */
897int pdanscan;   /* number of anonymous pages scanned by daemon */
898int pdobscan;   /* number of object pages scanned by daemon */
899int pdreact;    /* number of pages daemon reactivated since boot */
900int pdbusy;     /* number of times daemon found a busy page */
901int pdpageouts; /* number of times daemon started a pageout */
902int pdpending;  /* number of times daemon got a pending pagout */
903int pddeact;    /* number of pages daemon deactivates */
904.Ed
905
906.Pp
907.Fn uvm_fork
908forks a virtual address space for process' (old)
909.Fa p1
910and (new)
911.Fa p2 .
912If the
913.Fa shared
914argument is non zero, p1 shares its address space with p2,
915otherwise a new address space is created.  This function
916currently has no return value, and thus cannot fail.  In
917the future, this function will changed to allowed it to
918fail in low memory conditions.
919
920.Pp
921.Fn uvm_grow
922increases the stack segment of process
923.Fa p
924to include
925.Fa sp .
926
927.Pp
928.Fn uvm_coredump
929generates a coredump on vnode
930.Fa vp
931for process
932.Fa p
933with credentials
934.Fa cred
935and core header description in
936.Fa chdr .
937
938.Sh STANDARD UVM RETURN VALUES
939This section documents the standard return values that callers of UVM
940functions can expect.  They are derived from the Mach VM values
941of the same function.  The full list of values can be seen below.
942.Bd -literal
943#define KERN_SUCCESS            0
944#define KERN_INVALID_ADDRESS    1
945#define KERN_PROTECTION_FAILURE 2
946#define KERN_NO_SPACE           3
947#define KERN_INVALID_ARGUMENT   4
948#define KERN_FAILURE            5
949#define KERN_RESOURCE_SHORTAGE  6
950#define KERN_NOT_RECEIVER       7
951#define KERN_NO_ACCESS          8
952#define KERN_PAGES_LOCKED       9
953.Ed
954.Pp
955Note that
956.Dv KERN_NOT_RECEIVER
957and
958.Dv KERN_PAGES_LOCKED
959values are not actually returned by the UVM code.
960
961.Sh NOTES
962These functions are only available with options
963.Dv UVM .
964.Pp
965.Fn uvm_chgkprot
966is only available if the kernel has been compiled with options
967.Dv KGDB .
968.Pp
969The include file
970.Pa <vm/vm.h>
971will be deprecated when then Mach VM system is obsoleted.  All structure
972and types whose names begin with ``vm_'' will be renamed to ``uvm_''.
973.Pp
974The
975.Xr pmap 9
976manual page is not yet written.
977
978.Sh HISTORY
979UVM is a new VM system developed at Washington University in St. Louis
980(Missouri).  UVM's roots lie partly in the Mach-based
981.Bx 4.4
982VM system, the FreeBSD VM system, and the SunOS4 VM system.  UVM's basic
983structure is based on the
984.Bx 4.4
985VM system.  UVM's new i386 machine-depenent layer includes several ideas
986from FreeBSD.  UVM's new anonymous memory system is based on the
987anonymous memory system found in SunOS4 VM (as described in papers by
988published Sun Microsystems, Inc.).  UVM also includes a number of feature
989new to BSD including page loanout, map entry passing, simplified
990copy-on-write, and clustered anonymous memory pageout.  UVM will be
991further documented in August 1998 in a dissertation by Charles D. Cranor.
992.Pp
993UVM appeared in
994.Nx 1.4 .
995
996.Sh AUTHOR
997Charles D. Cranor <chuck@ccrc.wustl.edu> designed and implemented UVM.
998.Pp
999Matthew Green <mrg@eterna.com.au> wrote the swap-space managemnt code
1000and handled the logistical issues involed with merging UVM into the
1001NetBSD source tree.
1002.Pp
1003Chuck Silvers <chuq@chuq.com> implemented the aobj pager, thus allowing
1004UVM to support System V shared memory and process swapping.
1005
1006.Sh SEE ALSO
1007.Xr getloadavg 3 ,
1008.Xr kvm 3 ,
1009.Xr sysctl 3 ,
1010.Xr ddb 4 ,
1011.Xr options 4
1012