2005-01-07 02:29:27 +00:00
|
|
|
/*-
|
1994-05-24 10:09:53 +00:00
|
|
|
* Copyright (c) 1991, 1993
|
|
|
|
* The Regents of the University of California. All rights reserved.
|
|
|
|
*
|
|
|
|
* This code is derived from software contributed to Berkeley by
|
|
|
|
* The Mach Operating System project at Carnegie-Mellon University.
|
|
|
|
*
|
|
|
|
* Redistribution and use in source and binary forms, with or without
|
|
|
|
* modification, are permitted provided that the following conditions
|
|
|
|
* are met:
|
|
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer.
|
|
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
|
|
* documentation and/or other materials provided with the distribution.
|
|
|
|
* 4. Neither the name of the University nor the names of its contributors
|
|
|
|
* may be used to endorse or promote products derived from this software
|
|
|
|
* without specific prior written permission.
|
|
|
|
*
|
|
|
|
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
|
|
|
|
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
|
|
|
|
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
|
|
|
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
|
|
|
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
|
|
|
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
|
|
|
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
|
|
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
|
|
|
* SUCH DAMAGE.
|
|
|
|
*
|
1997-02-10 02:22:35 +00:00
|
|
|
* @(#)vm_map.h 8.9 (Berkeley) 5/17/95
|
1994-05-24 10:09:53 +00:00
|
|
|
*
|
|
|
|
*
|
|
|
|
* Copyright (c) 1987, 1990 Carnegie-Mellon University.
|
|
|
|
* All rights reserved.
|
|
|
|
*
|
|
|
|
* Authors: Avadis Tevanian, Jr., Michael Wayne Young
|
These changes embody the support of the fully coherent merged VM buffer cache,
much higher filesystem I/O performance, and much better paging performance. It
represents the culmination of over 6 months of R&D.
The majority of the merged VM/cache work is by John Dyson.
The following highlights the most significant changes. Additionally, there are
(mostly minor) changes to the various filesystem modules (nfs, msdosfs, etc) to
support the new VM/buffer scheme.
vfs_bio.c:
Significant rewrite of most of vfs_bio to support the merged VM buffer cache
scheme. The scheme is almost fully compatible with the old filesystem
interface. Significant improvement in the number of opportunities for write
clustering.
vfs_cluster.c, vfs_subr.c
Upgrade and performance enhancements in vfs layer code to support merged
VM/buffer cache. Fixup of vfs_cluster to eliminate the bogus pagemove stuff.
vm_object.c:
Yet more improvements in the collapse code. Elimination of some windows that
can cause list corruption.
vm_pageout.c:
Fixed it, it really works better now. Somehow in 2.0, some "enhancements"
broke the code. This code has been reworked from the ground-up.
vm_fault.c, vm_page.c, pmap.c, vm_object.c
Support for small-block filesystems with merged VM/buffer cache scheme.
pmap.c vm_map.c
Dynamic kernel VM size, now we dont have to pre-allocate excessive numbers of
kernel PTs.
vm_glue.c
Much simpler and more effective swapping code. No more gratuitous swapping.
proc.h
Fixed the problem that the p_lock flag was not being cleared on a fork.
swap_pager.c, vnode_pager.c
Removal of old vfs_bio cruft to support the past pseudo-coherency. Now the
code doesn't need it anymore.
machdep.c
Changes to better support the parameter values for the merged VM/buffer cache
scheme.
machdep.c, kern_exec.c, vm_glue.c
Implemented a seperate submap for temporary exec string space and another one
to contain process upages. This eliminates all map fragmentation problems
that previously existed.
ffs_inode.c, ufs_inode.c, ufs_readwrite.c
Changes for merged VM/buffer cache. Add "bypass" support for sneaking in on
busy buffers.
Submitted by: John Dyson and David Greenman
1995-01-09 16:06:02 +00:00
|
|
|
*
|
1994-05-24 10:09:53 +00:00
|
|
|
* Permission to use, copy, modify and distribute this software and
|
|
|
|
* its documentation is hereby granted, provided that both the copyright
|
|
|
|
* notice and this permission notice appear in all copies of the
|
|
|
|
* software, derivative works or modified versions, and any portions
|
|
|
|
* thereof, and that both notices appear in supporting documentation.
|
These changes embody the support of the fully coherent merged VM buffer cache,
much higher filesystem I/O performance, and much better paging performance. It
represents the culmination of over 6 months of R&D.
The majority of the merged VM/cache work is by John Dyson.
The following highlights the most significant changes. Additionally, there are
(mostly minor) changes to the various filesystem modules (nfs, msdosfs, etc) to
support the new VM/buffer scheme.
vfs_bio.c:
Significant rewrite of most of vfs_bio to support the merged VM buffer cache
scheme. The scheme is almost fully compatible with the old filesystem
interface. Significant improvement in the number of opportunities for write
clustering.
vfs_cluster.c, vfs_subr.c
Upgrade and performance enhancements in vfs layer code to support merged
VM/buffer cache. Fixup of vfs_cluster to eliminate the bogus pagemove stuff.
vm_object.c:
Yet more improvements in the collapse code. Elimination of some windows that
can cause list corruption.
vm_pageout.c:
Fixed it, it really works better now. Somehow in 2.0, some "enhancements"
broke the code. This code has been reworked from the ground-up.
vm_fault.c, vm_page.c, pmap.c, vm_object.c
Support for small-block filesystems with merged VM/buffer cache scheme.
pmap.c vm_map.c
Dynamic kernel VM size, now we dont have to pre-allocate excessive numbers of
kernel PTs.
vm_glue.c
Much simpler and more effective swapping code. No more gratuitous swapping.
proc.h
Fixed the problem that the p_lock flag was not being cleared on a fork.
swap_pager.c, vnode_pager.c
Removal of old vfs_bio cruft to support the past pseudo-coherency. Now the
code doesn't need it anymore.
machdep.c
Changes to better support the parameter values for the merged VM/buffer cache
scheme.
machdep.c, kern_exec.c, vm_glue.c
Implemented a seperate submap for temporary exec string space and another one
to contain process upages. This eliminates all map fragmentation problems
that previously existed.
ffs_inode.c, ufs_inode.c, ufs_readwrite.c
Changes for merged VM/buffer cache. Add "bypass" support for sneaking in on
busy buffers.
Submitted by: John Dyson and David Greenman
1995-01-09 16:06:02 +00:00
|
|
|
*
|
|
|
|
* CARNEGIE MELLON ALLOWS FREE USE OF THIS SOFTWARE IN ITS "AS IS"
|
|
|
|
* CONDITION. CARNEGIE MELLON DISCLAIMS ANY LIABILITY OF ANY KIND
|
1994-05-24 10:09:53 +00:00
|
|
|
* FOR ANY DAMAGES WHATSOEVER RESULTING FROM THE USE OF THIS SOFTWARE.
|
These changes embody the support of the fully coherent merged VM buffer cache,
much higher filesystem I/O performance, and much better paging performance. It
represents the culmination of over 6 months of R&D.
The majority of the merged VM/cache work is by John Dyson.
The following highlights the most significant changes. Additionally, there are
(mostly minor) changes to the various filesystem modules (nfs, msdosfs, etc) to
support the new VM/buffer scheme.
vfs_bio.c:
Significant rewrite of most of vfs_bio to support the merged VM buffer cache
scheme. The scheme is almost fully compatible with the old filesystem
interface. Significant improvement in the number of opportunities for write
clustering.
vfs_cluster.c, vfs_subr.c
Upgrade and performance enhancements in vfs layer code to support merged
VM/buffer cache. Fixup of vfs_cluster to eliminate the bogus pagemove stuff.
vm_object.c:
Yet more improvements in the collapse code. Elimination of some windows that
can cause list corruption.
vm_pageout.c:
Fixed it, it really works better now. Somehow in 2.0, some "enhancements"
broke the code. This code has been reworked from the ground-up.
vm_fault.c, vm_page.c, pmap.c, vm_object.c
Support for small-block filesystems with merged VM/buffer cache scheme.
pmap.c vm_map.c
Dynamic kernel VM size, now we dont have to pre-allocate excessive numbers of
kernel PTs.
vm_glue.c
Much simpler and more effective swapping code. No more gratuitous swapping.
proc.h
Fixed the problem that the p_lock flag was not being cleared on a fork.
swap_pager.c, vnode_pager.c
Removal of old vfs_bio cruft to support the past pseudo-coherency. Now the
code doesn't need it anymore.
machdep.c
Changes to better support the parameter values for the merged VM/buffer cache
scheme.
machdep.c, kern_exec.c, vm_glue.c
Implemented a seperate submap for temporary exec string space and another one
to contain process upages. This eliminates all map fragmentation problems
that previously existed.
ffs_inode.c, ufs_inode.c, ufs_readwrite.c
Changes for merged VM/buffer cache. Add "bypass" support for sneaking in on
busy buffers.
Submitted by: John Dyson and David Greenman
1995-01-09 16:06:02 +00:00
|
|
|
*
|
1994-05-24 10:09:53 +00:00
|
|
|
* Carnegie Mellon requests users of this software to return to
|
|
|
|
*
|
|
|
|
* Software Distribution Coordinator or Software.Distribution@CS.CMU.EDU
|
|
|
|
* School of Computer Science
|
|
|
|
* Carnegie Mellon University
|
|
|
|
* Pittsburgh PA 15213-3890
|
|
|
|
*
|
|
|
|
* any improvements or extensions that they make and grant Carnegie the
|
|
|
|
* rights to redistribute these changes.
|
1994-08-02 07:55:43 +00:00
|
|
|
*
|
1999-08-28 01:08:13 +00:00
|
|
|
* $FreeBSD$
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Virtual memory map module definitions.
|
|
|
|
*/
|
|
|
|
#ifndef _VM_MAP_
|
|
|
|
#define _VM_MAP_
|
|
|
|
|
2002-04-28 23:12:52 +00:00
|
|
|
#include <sys/lock.h>
|
2004-07-30 09:10:28 +00:00
|
|
|
#include <sys/sx.h>
|
2003-01-01 00:13:01 +00:00
|
|
|
#include <sys/_mutex.h>
|
2001-05-03 11:33:51 +00:00
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
|
|
|
* Types defined:
|
|
|
|
*
|
|
|
|
* vm_map_t the high-level address map data structure.
|
|
|
|
* vm_map_entry_t an entry in an address map.
|
|
|
|
*/
|
|
|
|
|
2003-08-11 07:14:08 +00:00
|
|
|
typedef u_char vm_flags_t;
|
2000-02-28 04:10:35 +00:00
|
|
|
typedef u_int vm_eflags_t;
|
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
|
|
|
* Objects which live in maps may be either VM objects, or
|
|
|
|
* another map (called a "sharing map") which denotes read-write
|
|
|
|
* sharing with other maps.
|
|
|
|
*/
|
|
|
|
union vm_map_object {
|
These changes embody the support of the fully coherent merged VM buffer cache,
much higher filesystem I/O performance, and much better paging performance. It
represents the culmination of over 6 months of R&D.
The majority of the merged VM/cache work is by John Dyson.
The following highlights the most significant changes. Additionally, there are
(mostly minor) changes to the various filesystem modules (nfs, msdosfs, etc) to
support the new VM/buffer scheme.
vfs_bio.c:
Significant rewrite of most of vfs_bio to support the merged VM buffer cache
scheme. The scheme is almost fully compatible with the old filesystem
interface. Significant improvement in the number of opportunities for write
clustering.
vfs_cluster.c, vfs_subr.c
Upgrade and performance enhancements in vfs layer code to support merged
VM/buffer cache. Fixup of vfs_cluster to eliminate the bogus pagemove stuff.
vm_object.c:
Yet more improvements in the collapse code. Elimination of some windows that
can cause list corruption.
vm_pageout.c:
Fixed it, it really works better now. Somehow in 2.0, some "enhancements"
broke the code. This code has been reworked from the ground-up.
vm_fault.c, vm_page.c, pmap.c, vm_object.c
Support for small-block filesystems with merged VM/buffer cache scheme.
pmap.c vm_map.c
Dynamic kernel VM size, now we dont have to pre-allocate excessive numbers of
kernel PTs.
vm_glue.c
Much simpler and more effective swapping code. No more gratuitous swapping.
proc.h
Fixed the problem that the p_lock flag was not being cleared on a fork.
swap_pager.c, vnode_pager.c
Removal of old vfs_bio cruft to support the past pseudo-coherency. Now the
code doesn't need it anymore.
machdep.c
Changes to better support the parameter values for the merged VM/buffer cache
scheme.
machdep.c, kern_exec.c, vm_glue.c
Implemented a seperate submap for temporary exec string space and another one
to contain process upages. This eliminates all map fragmentation problems
that previously existed.
ffs_inode.c, ufs_inode.c, ufs_readwrite.c
Changes for merged VM/buffer cache. Add "bypass" support for sneaking in on
busy buffers.
Submitted by: John Dyson and David Greenman
1995-01-09 16:06:02 +00:00
|
|
|
struct vm_object *vm_object; /* object object */
|
|
|
|
struct vm_map *sub_map; /* belongs to another map */
|
1994-05-24 10:09:53 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Address map entries consist of start and end addresses,
|
|
|
|
* a VM object (or sharing map) and offset into that object,
|
|
|
|
* and user-exported inheritance and protection information.
|
|
|
|
* Also included is control information for virtual copy operations.
|
|
|
|
*/
|
|
|
|
struct vm_map_entry {
|
These changes embody the support of the fully coherent merged VM buffer cache,
much higher filesystem I/O performance, and much better paging performance. It
represents the culmination of over 6 months of R&D.
The majority of the merged VM/cache work is by John Dyson.
The following highlights the most significant changes. Additionally, there are
(mostly minor) changes to the various filesystem modules (nfs, msdosfs, etc) to
support the new VM/buffer scheme.
vfs_bio.c:
Significant rewrite of most of vfs_bio to support the merged VM buffer cache
scheme. The scheme is almost fully compatible with the old filesystem
interface. Significant improvement in the number of opportunities for write
clustering.
vfs_cluster.c, vfs_subr.c
Upgrade and performance enhancements in vfs layer code to support merged
VM/buffer cache. Fixup of vfs_cluster to eliminate the bogus pagemove stuff.
vm_object.c:
Yet more improvements in the collapse code. Elimination of some windows that
can cause list corruption.
vm_pageout.c:
Fixed it, it really works better now. Somehow in 2.0, some "enhancements"
broke the code. This code has been reworked from the ground-up.
vm_fault.c, vm_page.c, pmap.c, vm_object.c
Support for small-block filesystems with merged VM/buffer cache scheme.
pmap.c vm_map.c
Dynamic kernel VM size, now we dont have to pre-allocate excessive numbers of
kernel PTs.
vm_glue.c
Much simpler and more effective swapping code. No more gratuitous swapping.
proc.h
Fixed the problem that the p_lock flag was not being cleared on a fork.
swap_pager.c, vnode_pager.c
Removal of old vfs_bio cruft to support the past pseudo-coherency. Now the
code doesn't need it anymore.
machdep.c
Changes to better support the parameter values for the merged VM/buffer cache
scheme.
machdep.c, kern_exec.c, vm_glue.c
Implemented a seperate submap for temporary exec string space and another one
to contain process upages. This eliminates all map fragmentation problems
that previously existed.
ffs_inode.c, ufs_inode.c, ufs_readwrite.c
Changes for merged VM/buffer cache. Add "bypass" support for sneaking in on
busy buffers.
Submitted by: John Dyson and David Greenman
1995-01-09 16:06:02 +00:00
|
|
|
struct vm_map_entry *prev; /* previous entry */
|
|
|
|
struct vm_map_entry *next; /* next entry */
|
2002-09-22 04:33:43 +00:00
|
|
|
struct vm_map_entry *left; /* left child in binary search tree */
|
|
|
|
struct vm_map_entry *right; /* right child in binary search tree */
|
These changes embody the support of the fully coherent merged VM buffer cache,
much higher filesystem I/O performance, and much better paging performance. It
represents the culmination of over 6 months of R&D.
The majority of the merged VM/cache work is by John Dyson.
The following highlights the most significant changes. Additionally, there are
(mostly minor) changes to the various filesystem modules (nfs, msdosfs, etc) to
support the new VM/buffer scheme.
vfs_bio.c:
Significant rewrite of most of vfs_bio to support the merged VM buffer cache
scheme. The scheme is almost fully compatible with the old filesystem
interface. Significant improvement in the number of opportunities for write
clustering.
vfs_cluster.c, vfs_subr.c
Upgrade and performance enhancements in vfs layer code to support merged
VM/buffer cache. Fixup of vfs_cluster to eliminate the bogus pagemove stuff.
vm_object.c:
Yet more improvements in the collapse code. Elimination of some windows that
can cause list corruption.
vm_pageout.c:
Fixed it, it really works better now. Somehow in 2.0, some "enhancements"
broke the code. This code has been reworked from the ground-up.
vm_fault.c, vm_page.c, pmap.c, vm_object.c
Support for small-block filesystems with merged VM/buffer cache scheme.
pmap.c vm_map.c
Dynamic kernel VM size, now we dont have to pre-allocate excessive numbers of
kernel PTs.
vm_glue.c
Much simpler and more effective swapping code. No more gratuitous swapping.
proc.h
Fixed the problem that the p_lock flag was not being cleared on a fork.
swap_pager.c, vnode_pager.c
Removal of old vfs_bio cruft to support the past pseudo-coherency. Now the
code doesn't need it anymore.
machdep.c
Changes to better support the parameter values for the merged VM/buffer cache
scheme.
machdep.c, kern_exec.c, vm_glue.c
Implemented a seperate submap for temporary exec string space and another one
to contain process upages. This eliminates all map fragmentation problems
that previously existed.
ffs_inode.c, ufs_inode.c, ufs_readwrite.c
Changes for merged VM/buffer cache. Add "bypass" support for sneaking in on
busy buffers.
Submitted by: John Dyson and David Greenman
1995-01-09 16:06:02 +00:00
|
|
|
vm_offset_t start; /* start address */
|
|
|
|
vm_offset_t end; /* end address */
|
1999-01-06 23:05:42 +00:00
|
|
|
vm_offset_t avail_ssize; /* amt can grow if this is a stack */
|
2004-08-13 08:06:34 +00:00
|
|
|
vm_size_t adj_free; /* amount of adjacent free space */
|
|
|
|
vm_size_t max_free; /* max free space in subtree */
|
These changes embody the support of the fully coherent merged VM buffer cache,
much higher filesystem I/O performance, and much better paging performance. It
represents the culmination of over 6 months of R&D.
The majority of the merged VM/cache work is by John Dyson.
The following highlights the most significant changes. Additionally, there are
(mostly minor) changes to the various filesystem modules (nfs, msdosfs, etc) to
support the new VM/buffer scheme.
vfs_bio.c:
Significant rewrite of most of vfs_bio to support the merged VM buffer cache
scheme. The scheme is almost fully compatible with the old filesystem
interface. Significant improvement in the number of opportunities for write
clustering.
vfs_cluster.c, vfs_subr.c
Upgrade and performance enhancements in vfs layer code to support merged
VM/buffer cache. Fixup of vfs_cluster to eliminate the bogus pagemove stuff.
vm_object.c:
Yet more improvements in the collapse code. Elimination of some windows that
can cause list corruption.
vm_pageout.c:
Fixed it, it really works better now. Somehow in 2.0, some "enhancements"
broke the code. This code has been reworked from the ground-up.
vm_fault.c, vm_page.c, pmap.c, vm_object.c
Support for small-block filesystems with merged VM/buffer cache scheme.
pmap.c vm_map.c
Dynamic kernel VM size, now we dont have to pre-allocate excessive numbers of
kernel PTs.
vm_glue.c
Much simpler and more effective swapping code. No more gratuitous swapping.
proc.h
Fixed the problem that the p_lock flag was not being cleared on a fork.
swap_pager.c, vnode_pager.c
Removal of old vfs_bio cruft to support the past pseudo-coherency. Now the
code doesn't need it anymore.
machdep.c
Changes to better support the parameter values for the merged VM/buffer cache
scheme.
machdep.c, kern_exec.c, vm_glue.c
Implemented a seperate submap for temporary exec string space and another one
to contain process upages. This eliminates all map fragmentation problems
that previously existed.
ffs_inode.c, ufs_inode.c, ufs_readwrite.c
Changes for merged VM/buffer cache. Add "bypass" support for sneaking in on
busy buffers.
Submitted by: John Dyson and David Greenman
1995-01-09 16:06:02 +00:00
|
|
|
union vm_map_object object; /* object I point to */
|
1995-12-11 04:58:34 +00:00
|
|
|
vm_ooffset_t offset; /* offset into object */
|
2000-02-28 04:10:35 +00:00
|
|
|
vm_eflags_t eflags; /* map entry flags */
|
These changes embody the support of the fully coherent merged VM buffer cache,
much higher filesystem I/O performance, and much better paging performance. It
represents the culmination of over 6 months of R&D.
The majority of the merged VM/cache work is by John Dyson.
The following highlights the most significant changes. Additionally, there are
(mostly minor) changes to the various filesystem modules (nfs, msdosfs, etc) to
support the new VM/buffer scheme.
vfs_bio.c:
Significant rewrite of most of vfs_bio to support the merged VM buffer cache
scheme. The scheme is almost fully compatible with the old filesystem
interface. Significant improvement in the number of opportunities for write
clustering.
vfs_cluster.c, vfs_subr.c
Upgrade and performance enhancements in vfs layer code to support merged
VM/buffer cache. Fixup of vfs_cluster to eliminate the bogus pagemove stuff.
vm_object.c:
Yet more improvements in the collapse code. Elimination of some windows that
can cause list corruption.
vm_pageout.c:
Fixed it, it really works better now. Somehow in 2.0, some "enhancements"
broke the code. This code has been reworked from the ground-up.
vm_fault.c, vm_page.c, pmap.c, vm_object.c
Support for small-block filesystems with merged VM/buffer cache scheme.
pmap.c vm_map.c
Dynamic kernel VM size, now we dont have to pre-allocate excessive numbers of
kernel PTs.
vm_glue.c
Much simpler and more effective swapping code. No more gratuitous swapping.
proc.h
Fixed the problem that the p_lock flag was not being cleared on a fork.
swap_pager.c, vnode_pager.c
Removal of old vfs_bio cruft to support the past pseudo-coherency. Now the
code doesn't need it anymore.
machdep.c
Changes to better support the parameter values for the merged VM/buffer cache
scheme.
machdep.c, kern_exec.c, vm_glue.c
Implemented a seperate submap for temporary exec string space and another one
to contain process upages. This eliminates all map fragmentation problems
that previously existed.
ffs_inode.c, ufs_inode.c, ufs_readwrite.c
Changes for merged VM/buffer cache. Add "bypass" support for sneaking in on
busy buffers.
Submitted by: John Dyson and David Greenman
1995-01-09 16:06:02 +00:00
|
|
|
vm_prot_t protection; /* protection code */
|
|
|
|
vm_prot_t max_protection; /* maximum protection */
|
|
|
|
vm_inherit_t inheritance; /* inheritance */
|
|
|
|
int wired_count; /* can be paged if = 0 */
|
1999-09-17 05:40:17 +00:00
|
|
|
vm_pindex_t lastr; /* last read */
|
2010-12-02 17:37:16 +00:00
|
|
|
struct ucred *cred; /* tmp storage for creator ref */
|
1994-05-24 10:09:53 +00:00
|
|
|
};
|
|
|
|
|
2000-02-28 04:10:35 +00:00
|
|
|
#define MAP_ENTRY_NOSYNC 0x0001
|
|
|
|
#define MAP_ENTRY_IS_SUB_MAP 0x0002
|
|
|
|
#define MAP_ENTRY_COW 0x0004
|
|
|
|
#define MAP_ENTRY_NEEDS_COPY 0x0008
|
|
|
|
#define MAP_ENTRY_NOFAULT 0x0010
|
|
|
|
#define MAP_ENTRY_USER_WIRED 0x0020
|
|
|
|
|
|
|
|
#define MAP_ENTRY_BEHAV_NORMAL 0x0000 /* default behavior */
|
|
|
|
#define MAP_ENTRY_BEHAV_SEQUENTIAL 0x0040 /* expect sequential access */
|
|
|
|
#define MAP_ENTRY_BEHAV_RANDOM 0x0080 /* expect random access */
|
|
|
|
#define MAP_ENTRY_BEHAV_RESERVED 0x00C0 /* future use */
|
1997-01-16 04:16:22 +00:00
|
|
|
|
2000-02-28 04:10:35 +00:00
|
|
|
#define MAP_ENTRY_BEHAV_MASK 0x00C0
|
1999-08-01 06:05:09 +00:00
|
|
|
|
2002-06-07 18:34:23 +00:00
|
|
|
#define MAP_ENTRY_IN_TRANSITION 0x0100 /* entry being changed */
|
|
|
|
#define MAP_ENTRY_NEEDS_WAKEUP 0x0200 /* waiters in transition */
|
2000-02-28 04:10:35 +00:00
|
|
|
#define MAP_ENTRY_NOCOREDUMP 0x0400 /* don't include in a core */
|
1999-08-01 06:05:09 +00:00
|
|
|
|
2003-08-30 21:25:23 +00:00
|
|
|
#define MAP_ENTRY_GROWS_DOWN 0x1000 /* Top-down stacks */
|
|
|
|
#define MAP_ENTRY_GROWS_UP 0x2000 /* Bottom-up stacks */
|
|
|
|
|
2009-04-10 10:16:03 +00:00
|
|
|
#define MAP_ENTRY_WIRE_SKIPPED 0x4000
|
|
|
|
|
2002-06-01 16:59:30 +00:00
|
|
|
#ifdef _KERNEL
|
2003-11-03 16:14:45 +00:00
|
|
|
static __inline u_char
|
2002-06-01 16:59:30 +00:00
|
|
|
vm_map_entry_behavior(vm_map_entry_t entry)
|
2003-11-03 16:14:45 +00:00
|
|
|
{
|
2002-06-01 16:59:30 +00:00
|
|
|
return (entry->eflags & MAP_ENTRY_BEHAV_MASK);
|
|
|
|
}
|
2004-08-09 19:52:29 +00:00
|
|
|
|
|
|
|
static __inline int
|
|
|
|
vm_map_entry_user_wired_count(vm_map_entry_t entry)
|
|
|
|
{
|
|
|
|
if (entry->eflags & MAP_ENTRY_USER_WIRED)
|
|
|
|
return (1);
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static __inline int
|
|
|
|
vm_map_entry_system_wired_count(vm_map_entry_t entry)
|
|
|
|
{
|
|
|
|
return (entry->wired_count - vm_map_entry_user_wired_count(entry));
|
|
|
|
}
|
2002-06-01 16:59:30 +00:00
|
|
|
#endif /* _KERNEL */
|
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
2002-09-22 04:33:43 +00:00
|
|
|
* A map is a set of map entries. These map entries are
|
|
|
|
* organized both as a binary search tree and as a doubly-linked
|
|
|
|
* list. Both structures are ordered based upon the start and
|
|
|
|
* end addresses contained within each map entry. Sleator and
|
|
|
|
* Tarjan's top-down splay algorithm is employed to control
|
2003-11-03 16:14:45 +00:00
|
|
|
* height imbalance in the binary search tree.
|
1999-08-23 18:08:34 +00:00
|
|
|
*
|
2002-04-27 22:01:37 +00:00
|
|
|
* List of locks
|
|
|
|
* (c) const until freed
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
|
|
|
struct vm_map {
|
These changes embody the support of the fully coherent merged VM buffer cache,
much higher filesystem I/O performance, and much better paging performance. It
represents the culmination of over 6 months of R&D.
The majority of the merged VM/cache work is by John Dyson.
The following highlights the most significant changes. Additionally, there are
(mostly minor) changes to the various filesystem modules (nfs, msdosfs, etc) to
support the new VM/buffer scheme.
vfs_bio.c:
Significant rewrite of most of vfs_bio to support the merged VM buffer cache
scheme. The scheme is almost fully compatible with the old filesystem
interface. Significant improvement in the number of opportunities for write
clustering.
vfs_cluster.c, vfs_subr.c
Upgrade and performance enhancements in vfs layer code to support merged
VM/buffer cache. Fixup of vfs_cluster to eliminate the bogus pagemove stuff.
vm_object.c:
Yet more improvements in the collapse code. Elimination of some windows that
can cause list corruption.
vm_pageout.c:
Fixed it, it really works better now. Somehow in 2.0, some "enhancements"
broke the code. This code has been reworked from the ground-up.
vm_fault.c, vm_page.c, pmap.c, vm_object.c
Support for small-block filesystems with merged VM/buffer cache scheme.
pmap.c vm_map.c
Dynamic kernel VM size, now we dont have to pre-allocate excessive numbers of
kernel PTs.
vm_glue.c
Much simpler and more effective swapping code. No more gratuitous swapping.
proc.h
Fixed the problem that the p_lock flag was not being cleared on a fork.
swap_pager.c, vnode_pager.c
Removal of old vfs_bio cruft to support the past pseudo-coherency. Now the
code doesn't need it anymore.
machdep.c
Changes to better support the parameter values for the merged VM/buffer cache
scheme.
machdep.c, kern_exec.c, vm_glue.c
Implemented a seperate submap for temporary exec string space and another one
to contain process upages. This eliminates all map fragmentation problems
that previously existed.
ffs_inode.c, ufs_inode.c, ufs_readwrite.c
Changes for merged VM/buffer cache. Add "bypass" support for sneaking in on
busy buffers.
Submitted by: John Dyson and David Greenman
1995-01-09 16:06:02 +00:00
|
|
|
struct vm_map_entry header; /* List of entries */
|
2004-07-30 09:10:28 +00:00
|
|
|
struct sx lock; /* Lock for map data */
|
2002-12-31 19:38:04 +00:00
|
|
|
struct mtx system_mtx;
|
These changes embody the support of the fully coherent merged VM buffer cache,
much higher filesystem I/O performance, and much better paging performance. It
represents the culmination of over 6 months of R&D.
The majority of the merged VM/cache work is by John Dyson.
The following highlights the most significant changes. Additionally, there are
(mostly minor) changes to the various filesystem modules (nfs, msdosfs, etc) to
support the new VM/buffer scheme.
vfs_bio.c:
Significant rewrite of most of vfs_bio to support the merged VM buffer cache
scheme. The scheme is almost fully compatible with the old filesystem
interface. Significant improvement in the number of opportunities for write
clustering.
vfs_cluster.c, vfs_subr.c
Upgrade and performance enhancements in vfs layer code to support merged
VM/buffer cache. Fixup of vfs_cluster to eliminate the bogus pagemove stuff.
vm_object.c:
Yet more improvements in the collapse code. Elimination of some windows that
can cause list corruption.
vm_pageout.c:
Fixed it, it really works better now. Somehow in 2.0, some "enhancements"
broke the code. This code has been reworked from the ground-up.
vm_fault.c, vm_page.c, pmap.c, vm_object.c
Support for small-block filesystems with merged VM/buffer cache scheme.
pmap.c vm_map.c
Dynamic kernel VM size, now we dont have to pre-allocate excessive numbers of
kernel PTs.
vm_glue.c
Much simpler and more effective swapping code. No more gratuitous swapping.
proc.h
Fixed the problem that the p_lock flag was not being cleared on a fork.
swap_pager.c, vnode_pager.c
Removal of old vfs_bio cruft to support the past pseudo-coherency. Now the
code doesn't need it anymore.
machdep.c
Changes to better support the parameter values for the merged VM/buffer cache
scheme.
machdep.c, kern_exec.c, vm_glue.c
Implemented a seperate submap for temporary exec string space and another one
to contain process upages. This eliminates all map fragmentation problems
that previously existed.
ffs_inode.c, ufs_inode.c, ufs_readwrite.c
Changes for merged VM/buffer cache. Add "bypass" support for sneaking in on
busy buffers.
Submitted by: John Dyson and David Greenman
1995-01-09 16:06:02 +00:00
|
|
|
int nentries; /* Number of entries */
|
|
|
|
vm_size_t size; /* virtual size */
|
2003-08-13 03:13:22 +00:00
|
|
|
u_int timestamp; /* Version number */
|
2002-07-11 02:39:24 +00:00
|
|
|
u_char needs_wakeup;
|
2008-12-31 05:44:05 +00:00
|
|
|
u_char system_map; /* (c) Am I a system map? */
|
2003-08-11 07:14:08 +00:00
|
|
|
vm_flags_t flags; /* flags for this vm_map */
|
2002-09-22 04:33:43 +00:00
|
|
|
vm_map_entry_t root; /* Root of a binary search tree */
|
2002-04-27 22:01:37 +00:00
|
|
|
pmap_t pmap; /* (c) Physical map */
|
|
|
|
#define min_offset header.start /* (c) */
|
|
|
|
#define max_offset header.end /* (c) */
|
2010-12-09 21:02:22 +00:00
|
|
|
int busy;
|
1994-05-24 10:09:53 +00:00
|
|
|
};
|
|
|
|
|
2003-08-11 07:14:08 +00:00
|
|
|
/*
|
|
|
|
* vm_flags_t values
|
|
|
|
*/
|
|
|
|
#define MAP_WIREFUTURE 0x01 /* wire all future pages */
|
2010-12-09 21:02:22 +00:00
|
|
|
#define MAP_BUSY_WAKEUP 0x02
|
2003-08-11 07:14:08 +00:00
|
|
|
|
2002-04-27 22:01:37 +00:00
|
|
|
#ifdef _KERNEL
|
|
|
|
static __inline vm_offset_t
|
|
|
|
vm_map_max(vm_map_t map)
|
|
|
|
{
|
|
|
|
return (map->max_offset);
|
|
|
|
}
|
|
|
|
|
|
|
|
static __inline vm_offset_t
|
|
|
|
vm_map_min(vm_map_t map)
|
|
|
|
{
|
|
|
|
return (map->min_offset);
|
|
|
|
}
|
|
|
|
|
|
|
|
static __inline pmap_t
|
|
|
|
vm_map_pmap(vm_map_t map)
|
|
|
|
{
|
|
|
|
return (map->pmap);
|
|
|
|
}
|
2003-08-11 07:14:08 +00:00
|
|
|
|
|
|
|
static __inline void
|
|
|
|
vm_map_modflags(vm_map_t map, vm_flags_t set, vm_flags_t clear)
|
|
|
|
{
|
|
|
|
map->flags = (map->flags | set) & ~clear;
|
|
|
|
}
|
2002-04-27 22:01:37 +00:00
|
|
|
#endif /* _KERNEL */
|
|
|
|
|
2003-11-03 16:14:45 +00:00
|
|
|
/*
|
1995-12-07 12:48:31 +00:00
|
|
|
* Shareable process virtual address space.
|
2002-06-25 18:14:38 +00:00
|
|
|
*
|
|
|
|
* List of locks
|
|
|
|
* (c) const until freed
|
1995-12-07 12:48:31 +00:00
|
|
|
*/
|
|
|
|
struct vmspace {
|
|
|
|
struct vm_map vm_map; /* VM address map */
|
2002-07-22 16:22:27 +00:00
|
|
|
struct shmmap_state *vm_shm; /* SYS5 shared memory private data XXX */
|
1995-12-07 12:48:31 +00:00
|
|
|
segsz_t vm_swrss; /* resident set size before last swap */
|
|
|
|
segsz_t vm_tsize; /* text size (pages) XXX */
|
|
|
|
segsz_t vm_dsize; /* data size (pages) XXX */
|
|
|
|
segsz_t vm_ssize; /* stack size (pages) */
|
2002-06-25 18:14:38 +00:00
|
|
|
caddr_t vm_taddr; /* (c) user virtual address of text */
|
|
|
|
caddr_t vm_daddr; /* (c) user virtual address of data */
|
1995-12-07 12:48:31 +00:00
|
|
|
caddr_t vm_maxsaddr; /* user VA at max stack growth */
|
2010-10-21 17:29:32 +00:00
|
|
|
volatile int vm_refcnt; /* number of references */
|
2008-03-01 22:54:42 +00:00
|
|
|
/*
|
|
|
|
* Keep the PMAP last, so that CPU-specific variations of that
|
|
|
|
* structure on a single architecture don't result in offset
|
|
|
|
* variations of the machine-independent fields in the vmspace.
|
|
|
|
*/
|
|
|
|
struct pmap vm_pmap; /* private physical map */
|
1995-12-07 12:48:31 +00:00
|
|
|
};
|
|
|
|
|
2002-06-01 22:41:43 +00:00
|
|
|
#ifdef _KERNEL
|
|
|
|
static __inline pmap_t
|
|
|
|
vmspace_pmap(struct vmspace *vmspace)
|
|
|
|
{
|
|
|
|
return &vmspace->vm_pmap;
|
|
|
|
}
|
|
|
|
#endif /* _KERNEL */
|
|
|
|
|
2001-05-19 01:28:09 +00:00
|
|
|
#ifdef _KERNEL
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
|
|
|
* Macros: vm_map_lock, etc.
|
|
|
|
* Function:
|
1999-08-16 18:21:09 +00:00
|
|
|
* Perform locking on the data portion of a map. Note that
|
|
|
|
* these macros mimic procedure calls returning void. The
|
|
|
|
* semicolon is supplied by the user of these macros, not
|
|
|
|
* by the macros themselves. The macros can safely be used
|
|
|
|
* as unbraced elements in a higher level statement.
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
|
|
|
|
2002-04-28 23:12:52 +00:00
|
|
|
void _vm_map_lock(vm_map_t map, const char *file, int line);
|
|
|
|
void _vm_map_unlock(vm_map_t map, const char *file, int line);
|
2010-09-19 17:43:22 +00:00
|
|
|
int _vm_map_unlock_and_wait(vm_map_t map, int timo, const char *file, int line);
|
2002-04-28 23:12:52 +00:00
|
|
|
void _vm_map_lock_read(vm_map_t map, const char *file, int line);
|
|
|
|
void _vm_map_unlock_read(vm_map_t map, const char *file, int line);
|
|
|
|
int _vm_map_trylock(vm_map_t map, const char *file, int line);
|
2003-03-12 23:13:16 +00:00
|
|
|
int _vm_map_trylock_read(vm_map_t map, const char *file, int line);
|
2002-04-28 23:12:52 +00:00
|
|
|
int _vm_map_lock_upgrade(vm_map_t map, const char *file, int line);
|
|
|
|
void _vm_map_lock_downgrade(vm_map_t map, const char *file, int line);
|
2009-01-01 00:31:46 +00:00
|
|
|
int vm_map_locked(vm_map_t map);
|
2002-07-11 02:39:24 +00:00
|
|
|
void vm_map_wakeup(vm_map_t map);
|
2010-12-09 21:02:22 +00:00
|
|
|
void vm_map_busy(vm_map_t map);
|
|
|
|
void vm_map_unbusy(vm_map_t map);
|
|
|
|
void vm_map_wait_busy(vm_map_t map);
|
2002-04-28 23:12:52 +00:00
|
|
|
|
|
|
|
#define vm_map_lock(map) _vm_map_lock(map, LOCK_FILE, LOCK_LINE)
|
|
|
|
#define vm_map_unlock(map) _vm_map_unlock(map, LOCK_FILE, LOCK_LINE)
|
2010-09-19 17:43:22 +00:00
|
|
|
#define vm_map_unlock_and_wait(map, timo) \
|
|
|
|
_vm_map_unlock_and_wait(map, timo, LOCK_FILE, LOCK_LINE)
|
2002-04-28 23:12:52 +00:00
|
|
|
#define vm_map_lock_read(map) _vm_map_lock_read(map, LOCK_FILE, LOCK_LINE)
|
|
|
|
#define vm_map_unlock_read(map) _vm_map_unlock_read(map, LOCK_FILE, LOCK_LINE)
|
|
|
|
#define vm_map_trylock(map) _vm_map_trylock(map, LOCK_FILE, LOCK_LINE)
|
2003-03-12 23:13:16 +00:00
|
|
|
#define vm_map_trylock_read(map) \
|
|
|
|
_vm_map_trylock_read(map, LOCK_FILE, LOCK_LINE)
|
2002-04-28 23:12:52 +00:00
|
|
|
#define vm_map_lock_upgrade(map) \
|
|
|
|
_vm_map_lock_upgrade(map, LOCK_FILE, LOCK_LINE)
|
|
|
|
#define vm_map_lock_downgrade(map) \
|
|
|
|
_vm_map_lock_downgrade(map, LOCK_FILE, LOCK_LINE)
|
2001-07-04 20:15:18 +00:00
|
|
|
|
|
|
|
long vmspace_resident_count(struct vmspace *vmspace);
|
2003-10-06 01:47:12 +00:00
|
|
|
long vmspace_wired_count(struct vmspace *vmspace);
|
2001-05-19 01:28:09 +00:00
|
|
|
#endif /* _KERNEL */
|
1994-05-24 10:09:53 +00:00
|
|
|
|
1999-02-19 14:25:37 +00:00
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/* XXX: number of kernel maps and entries to statically allocate */
|
|
|
|
#define MAX_KMAP 10
|
1994-05-25 09:21:21 +00:00
|
|
|
#define MAX_KMAPENT 128
|
1994-05-24 10:09:53 +00:00
|
|
|
|
1996-01-19 04:00:31 +00:00
|
|
|
/*
|
|
|
|
* Copy-on-write flags for vm_map operations
|
|
|
|
*/
|
2000-02-28 04:10:35 +00:00
|
|
|
#define MAP_UNUSED_01 0x0001
|
|
|
|
#define MAP_COPY_ON_WRITE 0x0002
|
|
|
|
#define MAP_NOFAULT 0x0004
|
|
|
|
#define MAP_PREFAULT 0x0008
|
|
|
|
#define MAP_PREFAULT_PARTIAL 0x0010
|
|
|
|
#define MAP_DISABLE_SYNCER 0x0020
|
|
|
|
#define MAP_DISABLE_COREDUMP 0x0100
|
2001-10-31 03:06:33 +00:00
|
|
|
#define MAP_PREFAULT_MADVISE 0x0200 /* from (user) madvise request */
|
2003-09-27 22:28:14 +00:00
|
|
|
#define MAP_STACK_GROWS_DOWN 0x1000
|
|
|
|
#define MAP_STACK_GROWS_UP 0x2000
|
Implement global and per-uid accounting of the anonymous memory. Add
rlimit RLIMIT_SWAP that limits the amount of swap that may be reserved
for the uid.
The accounting information (charge) is associated with either map entry,
or vm object backing the entry, assuming the object is the first one
in the shadow chain and entry does not require COW. Charge is moved
from entry to object on allocation of the object, e.g. during the mmap,
assuming the object is allocated, or on the first page fault on the
entry. It moves back to the entry on forks due to COW setup.
The per-entry granularity of accounting makes the charge process fair
for processes that change uid during lifetime, and decrements charge
for proper uid when region is unmapped.
The interface of vm_pager_allocate(9) is extended by adding struct ucred *,
that is used to charge appropriate uid when allocation if performed by
kernel, e.g. md(4).
Several syscalls, among them is fork(2), may now return ENOMEM when
global or per-uid limits are enforced.
In collaboration with: pho
Reviewed by: alc
Approved by: re (kensmith)
2009-06-23 20:45:22 +00:00
|
|
|
#define MAP_ACC_CHARGED 0x4000
|
|
|
|
#define MAP_ACC_NO_CHARGE 0x8000
|
1996-01-19 04:00:31 +00:00
|
|
|
|
1996-12-14 17:54:17 +00:00
|
|
|
/*
|
|
|
|
* vm_fault option flags
|
|
|
|
*/
|
1997-04-06 02:29:45 +00:00
|
|
|
#define VM_FAULT_NORMAL 0 /* Nothing special */
|
|
|
|
#define VM_FAULT_CHANGE_WIRING 1 /* Change the wiring as appropriate */
|
1996-12-14 17:54:17 +00:00
|
|
|
|
2008-05-10 18:55:35 +00:00
|
|
|
/*
|
|
|
|
* The following "find_space" options are supported by vm_map_find()
|
|
|
|
*/
|
|
|
|
#define VMFS_NO_SPACE 0 /* don't find; use the given range */
|
|
|
|
#define VMFS_ANY_SPACE 1 /* find a range with any alignment */
|
|
|
|
#define VMFS_ALIGNED_SPACE 2 /* find a superpage-aligned range */
|
2010-04-18 22:32:07 +00:00
|
|
|
#if defined(__mips__)
|
|
|
|
#define VMFS_TLB_ALIGNED_SPACE 3 /* find a TLB entry aligned range */
|
|
|
|
#endif
|
2008-05-10 18:55:35 +00:00
|
|
|
|
2003-08-11 07:14:08 +00:00
|
|
|
/*
|
|
|
|
* vm_map_wire and vm_map_unwire option flags
|
|
|
|
*/
|
|
|
|
#define VM_MAP_WIRE_SYSTEM 0 /* wiring in a kernel map */
|
|
|
|
#define VM_MAP_WIRE_USER 1 /* wiring in a user map */
|
|
|
|
|
|
|
|
#define VM_MAP_WIRE_NOHOLES 0 /* region must not have holes */
|
|
|
|
#define VM_MAP_WIRE_HOLESOK 2 /* region may have holes */
|
|
|
|
|
1999-12-29 05:07:58 +00:00
|
|
|
#ifdef _KERNEL
|
2001-07-04 20:15:18 +00:00
|
|
|
boolean_t vm_map_check_protection (vm_map_t, vm_offset_t, vm_offset_t, vm_prot_t);
|
2003-06-15 07:28:33 +00:00
|
|
|
vm_map_t vm_map_create(pmap_t, vm_offset_t, vm_offset_t);
|
2009-02-24 20:57:43 +00:00
|
|
|
int vm_map_delete(vm_map_t, vm_offset_t, vm_offset_t);
|
2008-05-10 18:55:35 +00:00
|
|
|
int vm_map_find(vm_map_t, vm_object_t, vm_ooffset_t, vm_offset_t *, vm_size_t,
|
|
|
|
int, vm_prot_t, vm_prot_t, int);
|
2008-04-28 05:30:23 +00:00
|
|
|
int vm_map_fixed(vm_map_t, vm_object_t, vm_ooffset_t, vm_offset_t, vm_size_t,
|
|
|
|
vm_prot_t, vm_prot_t, int);
|
2001-07-04 20:15:18 +00:00
|
|
|
int vm_map_findspace (vm_map_t, vm_offset_t, vm_size_t, vm_offset_t *);
|
|
|
|
int vm_map_inherit (vm_map_t, vm_offset_t, vm_offset_t, vm_inherit_t);
|
2010-04-03 19:07:05 +00:00
|
|
|
void vm_map_init(vm_map_t, pmap_t, vm_offset_t, vm_offset_t);
|
2001-07-04 20:15:18 +00:00
|
|
|
int vm_map_insert (vm_map_t, vm_object_t, vm_ooffset_t, vm_offset_t, vm_offset_t, vm_prot_t, vm_prot_t, int);
|
|
|
|
int vm_map_lookup (vm_map_t *, vm_offset_t, vm_prot_t, vm_map_entry_t *, vm_object_t *,
|
|
|
|
vm_pindex_t *, vm_prot_t *, boolean_t *);
|
2004-08-12 20:14:49 +00:00
|
|
|
int vm_map_lookup_locked(vm_map_t *, vm_offset_t, vm_prot_t, vm_map_entry_t *, vm_object_t *,
|
|
|
|
vm_pindex_t *, vm_prot_t *, boolean_t *);
|
2001-07-04 20:15:18 +00:00
|
|
|
void vm_map_lookup_done (vm_map_t, vm_map_entry_t);
|
|
|
|
boolean_t vm_map_lookup_entry (vm_map_t, vm_offset_t, vm_map_entry_t *);
|
2004-04-24 03:46:44 +00:00
|
|
|
void vm_map_pmap_enter(vm_map_t map, vm_offset_t addr, vm_prot_t prot,
|
2003-06-29 23:32:55 +00:00
|
|
|
vm_object_t object, vm_pindex_t pindex, vm_size_t size, int flags);
|
2001-07-04 20:15:18 +00:00
|
|
|
int vm_map_protect (vm_map_t, vm_offset_t, vm_offset_t, vm_prot_t, boolean_t);
|
|
|
|
int vm_map_remove (vm_map_t, vm_offset_t, vm_offset_t);
|
|
|
|
void vm_map_startup (void);
|
|
|
|
int vm_map_submap (vm_map_t, vm_offset_t, vm_offset_t, vm_map_t);
|
2003-11-09 05:25:35 +00:00
|
|
|
int vm_map_sync(vm_map_t, vm_offset_t, vm_offset_t, boolean_t, boolean_t);
|
2001-07-04 20:15:18 +00:00
|
|
|
int vm_map_madvise (vm_map_t, vm_offset_t, vm_offset_t, int);
|
|
|
|
void vm_map_simplify_entry (vm_map_t, vm_map_entry_t);
|
|
|
|
void vm_init2 (void);
|
|
|
|
int vm_map_stack (vm_map_t, vm_offset_t, vm_size_t, vm_prot_t, vm_prot_t, int);
|
|
|
|
int vm_map_growstack (struct proc *p, vm_offset_t addr);
|
2002-06-07 18:34:23 +00:00
|
|
|
int vm_map_unwire(vm_map_t map, vm_offset_t start, vm_offset_t end,
|
2003-08-11 07:14:08 +00:00
|
|
|
int flags);
|
2002-06-08 07:32:38 +00:00
|
|
|
int vm_map_wire(vm_map_t map, vm_offset_t start, vm_offset_t end,
|
2003-08-11 07:14:08 +00:00
|
|
|
int flags);
|
2001-07-04 20:15:18 +00:00
|
|
|
int vmspace_swap_count (struct vmspace *vmspace);
|
2002-03-10 21:52:48 +00:00
|
|
|
#endif /* _KERNEL */
|
These changes embody the support of the fully coherent merged VM buffer cache,
much higher filesystem I/O performance, and much better paging performance. It
represents the culmination of over 6 months of R&D.
The majority of the merged VM/cache work is by John Dyson.
The following highlights the most significant changes. Additionally, there are
(mostly minor) changes to the various filesystem modules (nfs, msdosfs, etc) to
support the new VM/buffer scheme.
vfs_bio.c:
Significant rewrite of most of vfs_bio to support the merged VM buffer cache
scheme. The scheme is almost fully compatible with the old filesystem
interface. Significant improvement in the number of opportunities for write
clustering.
vfs_cluster.c, vfs_subr.c
Upgrade and performance enhancements in vfs layer code to support merged
VM/buffer cache. Fixup of vfs_cluster to eliminate the bogus pagemove stuff.
vm_object.c:
Yet more improvements in the collapse code. Elimination of some windows that
can cause list corruption.
vm_pageout.c:
Fixed it, it really works better now. Somehow in 2.0, some "enhancements"
broke the code. This code has been reworked from the ground-up.
vm_fault.c, vm_page.c, pmap.c, vm_object.c
Support for small-block filesystems with merged VM/buffer cache scheme.
pmap.c vm_map.c
Dynamic kernel VM size, now we dont have to pre-allocate excessive numbers of
kernel PTs.
vm_glue.c
Much simpler and more effective swapping code. No more gratuitous swapping.
proc.h
Fixed the problem that the p_lock flag was not being cleared on a fork.
swap_pager.c, vnode_pager.c
Removal of old vfs_bio cruft to support the past pseudo-coherency. Now the
code doesn't need it anymore.
machdep.c
Changes to better support the parameter values for the merged VM/buffer cache
scheme.
machdep.c, kern_exec.c, vm_glue.c
Implemented a seperate submap for temporary exec string space and another one
to contain process upages. This eliminates all map fragmentation problems
that previously existed.
ffs_inode.c, ufs_inode.c, ufs_readwrite.c
Changes for merged VM/buffer cache. Add "bypass" support for sneaking in on
busy buffers.
Submitted by: John Dyson and David Greenman
1995-01-09 16:06:02 +00:00
|
|
|
#endif /* _VM_MAP_ */
|