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
|
|
|
* 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.
|
|
|
|
* 3. All advertising materials mentioning features or use of this software
|
|
|
|
* must display the following acknowledgement:
|
|
|
|
* This product includes software developed by the University of
|
|
|
|
* California, Berkeley and its contributors.
|
|
|
|
* 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>
|
2002-03-18 15:08:09 +00:00
|
|
|
#include <sys/lockmgr.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.
|
|
|
|
*/
|
|
|
|
|
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 */
|
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
|
|
|
/* Only in task maps: */
|
|
|
|
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 */
|
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
|
|
|
|
2002-06-01 16:59:30 +00:00
|
|
|
#ifdef _KERNEL
|
|
|
|
static __inline u_char
|
|
|
|
vm_map_entry_behavior(vm_map_entry_t entry)
|
|
|
|
{
|
|
|
|
return (entry->eflags & MAP_ENTRY_BEHAV_MASK);
|
|
|
|
}
|
|
|
|
#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
|
|
|
|
* height imbalance in the binary search tree.
|
1999-08-23 18:08:34 +00:00
|
|
|
*
|
|
|
|
* Note: the lock structure cannot be the first element of vm_map
|
|
|
|
* because this can result in a running lockup between two or more
|
|
|
|
* system processes trying to kmem_alloc_wait() due to kmem_alloc_wait()
|
|
|
|
* and free tsleep/waking up 'map' and the underlying lockmgr also
|
|
|
|
* sleeping and waking up on 'map'. The lockup occurs when the map fills
|
|
|
|
* up. The 'exec' map, for example.
|
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 */
|
2002-03-18 15:08:09 +00:00
|
|
|
struct lock 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 */
|
2002-07-11 02:39:24 +00:00
|
|
|
u_char needs_wakeup;
|
1999-08-23 18:16:05 +00:00
|
|
|
u_char system_map; /* Am I a system map? */
|
2001-03-14 06:48:53 +00:00
|
|
|
u_char infork; /* Am I in fork processing? */
|
2002-09-22 04:33:43 +00:00
|
|
|
vm_map_entry_t root; /* Root of a binary search tree */
|
1998-01-17 09:17:02 +00:00
|
|
|
unsigned int timestamp; /* Version number */
|
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_map_entry_t first_free; /* First free space hint */
|
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) */
|
1994-05-24 10:09:53 +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);
|
|
|
|
}
|
|
|
|
#endif /* _KERNEL */
|
|
|
|
|
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 */
|
|
|
|
struct pmap vm_pmap; /* private physical map */
|
|
|
|
int vm_refcnt; /* number of references */
|
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
|
|
|
/* we copy from vm_startcopy to the end of the structure on fork */
|
|
|
|
#define vm_startcopy vm_rssize
|
|
|
|
segsz_t vm_rssize; /* current resident set size in pages */
|
|
|
|
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 */
|
2002-12-15 18:50:04 +00:00
|
|
|
#define vm_endcopy vm_exitingcnt
|
|
|
|
int vm_exitingcnt; /* several processes zombied in exit1 */
|
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);
|
|
|
|
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);
|
|
|
|
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);
|
2002-07-11 02:39:24 +00:00
|
|
|
int vm_map_unlock_and_wait(vm_map_t map, boolean_t user_wait);
|
|
|
|
void vm_map_wakeup(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)
|
|
|
|
#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)
|
|
|
|
#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);
|
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
|
1997-08-05 00:02:08 +00:00
|
|
|
#define MAX_MAPENT 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 */
|
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 */
|
|
|
|
#define VM_FAULT_USER_WIRE 2 /* Likewise, but for user purposes */
|
|
|
|
#define VM_FAULT_WIRE_MASK (VM_FAULT_CHANGE_WIRING|VM_FAULT_USER_WIRE)
|
|
|
|
#define VM_FAULT_DIRTY 8 /* Dirty the page */
|
1996-12-14 17:54:17 +00:00
|
|
|
|
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);
|
1994-05-24 10:09:53 +00:00
|
|
|
struct pmap;
|
2001-07-04 20:15:18 +00:00
|
|
|
vm_map_t vm_map_create (struct pmap *, vm_offset_t, vm_offset_t);
|
|
|
|
int vm_map_delete (vm_map_t, vm_offset_t, vm_offset_t);
|
|
|
|
int vm_map_find (vm_map_t, vm_object_t, vm_ooffset_t, vm_offset_t *, vm_size_t, boolean_t, vm_prot_t, vm_prot_t, int);
|
|
|
|
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);
|
|
|
|
void vm_map_init (struct vm_map *, vm_offset_t, vm_offset_t);
|
|
|
|
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 *);
|
|
|
|
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 *);
|
|
|
|
int vm_map_clean (vm_map_t, vm_offset_t, vm_offset_t, boolean_t, boolean_t);
|
|
|
|
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);
|
|
|
|
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,
|
|
|
|
boolean_t user_unwire);
|
2002-06-08 07:32:38 +00:00
|
|
|
int vm_map_wire(vm_map_t map, vm_offset_t start, vm_offset_t end,
|
|
|
|
boolean_t user_wire);
|
2001-07-04 20:15:18 +00:00
|
|
|
int vmspace_swap_count (struct vmspace *vmspace);
|
2002-05-05 22:42:40 +00:00
|
|
|
int vm_uiomove(vm_map_t, vm_object_t, off_t, int, vm_offset_t, int *);
|
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_ */
|