d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
ba4ad1fcea
freebsd-dev/sys/ufs/ffs/ffs_vnops.c

275 lines
8.0 KiB
C
Raw Normal View History

BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Copyright (c) 1982, 1986, 1989, 1993
* The Regents of the University of California. All rights reserved.
*
* 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.
*
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
* @(#)ffs_vnops.c 8.15 (Berkeley) 5/14/95
$Id$ -> $FreeBSD$
1999-08-28 01:08:13 +00:00
* $FreeBSD$
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/resourcevar.h>
Cosmetics. make gcc less noisy. Still some way to go here.
1994-10-10 01:04:55 +00:00
#include <sys/signalvar.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <sys/kernel.h>
#include <sys/stat.h>
#include <sys/buf.h>
#include <sys/proc.h>
#include <sys/mount.h>
#include <sys/vnode.h>
Decommision miscfs/specfs/specdev.h. Most of it goes into <sys/conf.h>, a few lines into <sys/vnode.h>. Add a few fields to struct specinfo, paving the way for the fun part.
1999-08-08 18:43:05 +00:00
#include <sys/conf.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Moved some #includes from <sys/param.h> nearer to where they are actually used.
1998-03-28 10:33:27 +00:00
#include <machine/limits.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <vm/vm.h>
Changes from John Dyson and myself: Fixed remaining known bugs in the buffer IO and VM system. vfs_bio.c: Fixed some race conditions and locking bugs. Improved performance by removing some (now) unnecessary code and fixing some broken logic. Fixed process accounting of # of FS outputs. Properly handle NFS interrupts (B_EINTR). (various) Replaced calls to clrbuf() with calls to an optimized routine called vfs_bio_clrbuf(). (various FS sync) Sync out modified vnode_pager backed pages. ffs_vnops.c: Do two passes: Sync out file data first, then indirect blocks. vm_fault.c: Fixed deadly embrace caused by acquiring locks in the wrong order. vnode_pager.c: Changed to use buffer I/O system for writing out modified pages. This should fix the problem with the modification date previous not getting updated. Also dramatically simplifies the code. Note that this is going to change in the future and be implemented via VOP_PUTPAGES(). vm_object.c: Fixed a pile of bugs related to cleaning (vnode) objects. The performance of vm_object_page_clean() is terrible when dealing with huge objects, but this will change when we implement a binary tree to keep the object pages sorted. vm_pageout.c: Fixed broken clustering of pageouts. Fixed race conditions and other lockup style bugs in the scanning of pages. Improved performance.
1995-04-09 06:03:56 +00:00
#include <vm/vm_page.h>
#include <vm/vm_object.h>
Untangled the vm.h include file spaghetti.
1995-12-07 12:48:31 +00:00
#include <vm/vm_extern.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <ufs/ufs/quota.h>
#include <ufs/ufs/inode.h>
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
#include <ufs/ufs/ufsmount.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <ufs/ufs/ufs_extern.h>
#include <ufs/ffs/fs.h>
#include <ufs/ffs/ffs_extern.h>
Staticize.
1995-12-17 21:14:36 +00:00
static int ffs_fsync __P((struct vop_fsync_args *));
static int ffs_getpages __P((struct vop_getpages_args *));
In the author's words: These diffs implement the first stage of a VOP_{GET|PUT}PAGES pushdown for local media FS's. See ffs_putpages in /sys/ufs/ufs/ufs_readwrite.c for implementation details for generic *_{get|put}pages for local media FS's. Support is trivial to add for any FS that formerly relied on the default behaviour of the vnode_pager in in EOPNOTSUPP cases (just copy the ffs_getpages() code for the FS in question's *_{get|put}pages). Obviously, it would be better if each local media FS implemented a more optimal method, instead of calling an exported interface from the /sys/vm/vnode_pager.c, but this is a necessary first step in getting the FS's to a point where they can be supplied with better implementations on a case-by-case basis. Obviously, the cd9660_putpages() can be rather trivial (since it is a read-only FS type 8-)). A slight (temporary) modification is made to print a diagnostic message in the case where the underlying filesystem attempts to engage in the previous behaviour. Failure is likely to be ungraceful. Submitted by: terry@freebsd.org (Terry Lambert)
1998-02-26 06:39:59 +00:00
static int ffs_putpages __P((struct vop_putpages_args *));
Staticize.
1995-12-17 21:14:36 +00:00
static int ffs_read __P((struct vop_read_args *));
static int ffs_write __P((struct vop_write_args *));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/* Global vfs data structures for ufs. */
Introduced a type `vop_t' for vnode operation functions and used it 1138 times (:-() in casts and a few more times in declarations. This change is null for the i386. The type has to be `typedef int vop_t(void *)' and not `typedef int vop_t()' because `gcc -Wstrict-prototypes' warns about the latter. Since vnode op functions are called with args of different (struct pointer) types, neither of these function types is any use for type checking of the arg, so it would be preferable not to use the complete function type, especially since using the complete type requires adding 1138 casts to avoid compiler warnings and another 40+ casts to reverse the function pointer conversions before calling the functions.
1995-11-09 08:17:23 +00:00
vop_t **ffs_vnodeop_p;
Staticize.
1995-12-17 21:14:36 +00:00
static struct vnodeopv_entry_desc ffs_vnodeop_entries[] = {
vnops megacommit 1. Use the default function to access all the specfs operations. 2. Use the default function to access all the fifofs operations. 3. Use the default function to access all the ufs operations. 4. Fix VCALL usage in vfs_cache.c 5. Use VOCALL to access specfs functions in devfs_vnops.c 6. Staticize most of the spec and fifofs vnops functions. 7. Make UFS panic if it lacks bits of the underlying storage handling.
1997-10-15 13:24:07 +00:00
{ &vop_default_desc, (vop_t *) ufs_vnoperate },
Stylistic overhaul of vnops tables. 1. Remove comment stating the blatantly obvious. 2. Align in two columns. 3. Sort all but the default element alphabetically. 4. Remove XXX comments pointing out entries not needed.
1997-10-15 09:22:02 +00:00
{ &vop_fsync_desc, (vop_t *) ffs_fsync },
Hmm, realign the vnops into two columns.
1997-10-15 10:05:29 +00:00
{ &vop_getpages_desc, (vop_t *) ffs_getpages },
In the author's words: These diffs implement the first stage of a VOP_{GET|PUT}PAGES pushdown for local media FS's. See ffs_putpages in /sys/ufs/ufs/ufs_readwrite.c for implementation details for generic *_{get|put}pages for local media FS's. Support is trivial to add for any FS that formerly relied on the default behaviour of the vnode_pager in in EOPNOTSUPP cases (just copy the ffs_getpages() code for the FS in question's *_{get|put}pages). Obviously, it would be better if each local media FS implemented a more optimal method, instead of calling an exported interface from the /sys/vm/vnode_pager.c, but this is a necessary first step in getting the FS's to a point where they can be supplied with better implementations on a case-by-case basis. Obviously, the cd9660_putpages() can be rather trivial (since it is a read-only FS type 8-)). A slight (temporary) modification is made to print a diagnostic message in the case where the underlying filesystem attempts to engage in the previous behaviour. Failure is likely to be ungraceful. Submitted by: terry@freebsd.org (Terry Lambert)
1998-02-26 06:39:59 +00:00
{ &vop_putpages_desc, (vop_t *) ffs_putpages },
Stylistic overhaul of vnops tables. 1. Remove comment stating the blatantly obvious. 2. Align in two columns. 3. Sort all but the default element alphabetically. 4. Remove XXX comments pointing out entries not needed.
1997-10-15 09:22:02 +00:00
{ &vop_read_desc, (vop_t *) ffs_read },
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
{ &vop_balloc_desc, (vop_t *) ffs_balloc },
Stylistic overhaul of vnops tables. 1. Remove comment stating the blatantly obvious. 2. Align in two columns. 3. Sort all but the default element alphabetically. 4. Remove XXX comments pointing out entries not needed.
1997-10-15 09:22:02 +00:00
{ &vop_reallocblks_desc, (vop_t *) ffs_reallocblks },
{ &vop_write_desc, (vop_t *) ffs_write },
Introduced a type `vop_t' for vnode operation functions and used it 1138 times (:-() in casts and a few more times in declarations. This change is null for the i386. The type has to be `typedef int vop_t(void *)' and not `typedef int vop_t()' because `gcc -Wstrict-prototypes' warns about the latter. Since vnode op functions are called with args of different (struct pointer) types, neither of these function types is any use for type checking of the arg, so it would be preferable not to use the complete function type, especially since using the complete type requires adding 1138 casts to avoid compiler warnings and another 40+ casts to reverse the function pointer conversions before calling the functions.
1995-11-09 08:17:23 +00:00
{ NULL, NULL }
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
};
Staticize.
1995-12-17 21:14:36 +00:00
static struct vnodeopv_desc ffs_vnodeop_opv_desc =
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{ &ffs_vnodeop_p, ffs_vnodeop_entries };
Introduced a type `vop_t' for vnode operation functions and used it 1138 times (:-() in casts and a few more times in declarations. This change is null for the i386. The type has to be `typedef int vop_t(void *)' and not `typedef int vop_t()' because `gcc -Wstrict-prototypes' warns about the latter. Since vnode op functions are called with args of different (struct pointer) types, neither of these function types is any use for type checking of the arg, so it would be preferable not to use the complete function type, especially since using the complete type requires adding 1138 casts to avoid compiler warnings and another 40+ casts to reverse the function pointer conversions before calling the functions.
1995-11-09 08:17:23 +00:00
vop_t **ffs_specop_p;
Staticize.
1995-12-17 21:14:36 +00:00
static struct vnodeopv_entry_desc ffs_specop_entries[] = {
vnops megacommit 1. Use the default function to access all the specfs operations. 2. Use the default function to access all the fifofs operations. 3. Use the default function to access all the ufs operations. 4. Fix VCALL usage in vfs_cache.c 5. Use VOCALL to access specfs functions in devfs_vnops.c 6. Staticize most of the spec and fifofs vnops functions. 7. Make UFS panic if it lacks bits of the underlying storage handling.
1997-10-15 13:24:07 +00:00
{ &vop_default_desc, (vop_t *) ufs_vnoperatespec },
Stylistic overhaul of vnops tables. 1. Remove comment stating the blatantly obvious. 2. Align in two columns. 3. Sort all but the default element alphabetically. 4. Remove XXX comments pointing out entries not needed.
1997-10-15 09:22:02 +00:00
{ &vop_fsync_desc, (vop_t *) ffs_fsync },
Introduced a type `vop_t' for vnode operation functions and used it 1138 times (:-() in casts and a few more times in declarations. This change is null for the i386. The type has to be `typedef int vop_t(void *)' and not `typedef int vop_t()' because `gcc -Wstrict-prototypes' warns about the latter. Since vnode op functions are called with args of different (struct pointer) types, neither of these function types is any use for type checking of the arg, so it would be preferable not to use the complete function type, especially since using the complete type requires adding 1138 casts to avoid compiler warnings and another 40+ casts to reverse the function pointer conversions before calling the functions.
1995-11-09 08:17:23 +00:00
{ NULL, NULL }
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
};
Staticize.
1995-12-17 21:14:36 +00:00
static struct vnodeopv_desc ffs_specop_opv_desc =
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{ &ffs_specop_p, ffs_specop_entries };
Introduced a type `vop_t' for vnode operation functions and used it 1138 times (:-() in casts and a few more times in declarations. This change is null for the i386. The type has to be `typedef int vop_t(void *)' and not `typedef int vop_t()' because `gcc -Wstrict-prototypes' warns about the latter. Since vnode op functions are called with args of different (struct pointer) types, neither of these function types is any use for type checking of the arg, so it would be preferable not to use the complete function type, especially since using the complete type requires adding 1138 casts to avoid compiler warnings and another 40+ casts to reverse the function pointer conversions before calling the functions.
1995-11-09 08:17:23 +00:00
vop_t **ffs_fifoop_p;
Staticize.
1995-12-17 21:14:36 +00:00
static struct vnodeopv_entry_desc ffs_fifoop_entries[] = {
vnops megacommit 1. Use the default function to access all the specfs operations. 2. Use the default function to access all the fifofs operations. 3. Use the default function to access all the ufs operations. 4. Fix VCALL usage in vfs_cache.c 5. Use VOCALL to access specfs functions in devfs_vnops.c 6. Staticize most of the spec and fifofs vnops functions. 7. Make UFS panic if it lacks bits of the underlying storage handling.
1997-10-15 13:24:07 +00:00
{ &vop_default_desc, (vop_t *) ufs_vnoperatefifo },
Stylistic overhaul of vnops tables. 1. Remove comment stating the blatantly obvious. 2. Align in two columns. 3. Sort all but the default element alphabetically. 4. Remove XXX comments pointing out entries not needed.
1997-10-15 09:22:02 +00:00
{ &vop_fsync_desc, (vop_t *) ffs_fsync },
Introduced a type `vop_t' for vnode operation functions and used it 1138 times (:-() in casts and a few more times in declarations. This change is null for the i386. The type has to be `typedef int vop_t(void *)' and not `typedef int vop_t()' because `gcc -Wstrict-prototypes' warns about the latter. Since vnode op functions are called with args of different (struct pointer) types, neither of these function types is any use for type checking of the arg, so it would be preferable not to use the complete function type, especially since using the complete type requires adding 1138 casts to avoid compiler warnings and another 40+ casts to reverse the function pointer conversions before calling the functions.
1995-11-09 08:17:23 +00:00
{ NULL, NULL }
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
};
Staticize.
1995-12-17 21:14:36 +00:00
static struct vnodeopv_desc ffs_fifoop_opv_desc =
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{ &ffs_fifoop_p, ffs_fifoop_entries };
Implemented loadable VFS modules, and made most existing filesystems loadable. (NFS is a notable exception.)
1994-09-21 03:47:43 +00:00
VNODEOP_SET(ffs_vnodeop_opv_desc);
VNODEOP_SET(ffs_specop_opv_desc);
VNODEOP_SET(ffs_fifoop_opv_desc);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <ufs/ufs/ufs_readwrite.c>
/*
* Synch an open file.
*/
/* ARGSUSED */
Staticize.
1995-12-17 21:14:36 +00:00
static int
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
ffs_fsync(ap)
struct vop_fsync_args /* {
struct vnode *a_vp;
struct ucred *a_cred;
int a_waitfor;
struct proc *a_p;
} */ *ap;
{
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
struct vnode *vp = ap->a_vp;
struct buf *bp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct buf *nbp;
Several performance improvements for soft updates have been added: 1) Fastpath deletions. When a file is being deleted, check to see if it was so recently created that its inode has not yet been written to disk. If so, the delete can proceed to immediately free the inode. 2) Background writes: No file or block allocations can be done while the bitmap is being written to disk. To avoid these stalls, the bitmap is copied to another buffer which is written thus leaving the original available for futher allocations. 3) Link count tracking. Constantly track the difference in i_effnlink and i_nlink so that inodes that have had no change other than i_effnlink need not be written. 4) Identify buffers with rollback dependencies so that the buffer flushing daemon can choose to skip over them.
2000-01-10 00:24:24 +00:00
int s, error, wait, passes, skipmeta;
Make our v_usecount vnode reference count work identically to the original BSD code. The association between the vnode and the vm_object no longer includes reference counts. The major difference is that vm_object's are no longer freed gratuitiously from the vnode, and so once an object is created for the vnode, it will last as long as the vnode does. When a vnode object reference count is incremented, then the underlying vnode reference count is incremented also. The two "objects" are now more intimately related, and so the interactions are now much less complex. When vnodes are now normally placed onto the free queue with an object still attached. The rundown of the object happens at vnode rundown time, and happens with exactly the same filesystem semantics of the original VFS code. There is absolutely no need for vnode_pager_uncache and other travesties like that anymore. A side-effect of these changes is that SMP locking should be much simpler, the I/O copyin/copyout optimizations work, NFS should be more ponderable, and further work on layered filesystems should be less frustrating, because of the totally coherent management of the vnode objects and vnodes. Please be careful with your system while running this code, but I would greatly appreciate feedback as soon a reasonably possible.
1998-01-06 05:26:17 +00:00
daddr_t lbn;
Several performance improvements for soft updates have been added: 1) Fastpath deletions. When a file is being deleted, check to see if it was so recently created that its inode has not yet been written to disk. If so, the delete can proceed to immediately free the inode. 2) Background writes: No file or block allocations can be done while the bitmap is being written to disk. To avoid these stalls, the bitmap is copied to another buffer which is written thus leaving the original available for futher allocations. 3) Link count tracking. Constantly track the difference in i_effnlink and i_nlink so that inodes that have had no change other than i_effnlink need not be written. 4) Identify buffers with rollback dependencies so that the buffer flushing daemon can choose to skip over them.
2000-01-10 00:24:24 +00:00
wait = (ap->a_waitfor == MNT_WAIT);
Give vn_isdisk() a second argument where it can return a suitable errno. Suggested by: bde
2000-01-10 12:04:27 +00:00
if (vn_isdisk(vp, NULL)) {
Make our v_usecount vnode reference count work identically to the original BSD code. The association between the vnode and the vm_object no longer includes reference counts. The major difference is that vm_object's are no longer freed gratuitiously from the vnode, and so once an object is created for the vnode, it will last as long as the vnode does. When a vnode object reference count is incremented, then the underlying vnode reference count is incremented also. The two "objects" are now more intimately related, and so the interactions are now much less complex. When vnodes are now normally placed onto the free queue with an object still attached. The rundown of the object happens at vnode rundown time, and happens with exactly the same filesystem semantics of the original VFS code. There is absolutely no need for vnode_pager_uncache and other travesties like that anymore. A side-effect of these changes is that SMP locking should be much simpler, the I/O copyin/copyout optimizations work, NFS should be more ponderable, and further work on layered filesystems should be less frustrating, because of the totally coherent management of the vnode objects and vnodes. Please be careful with your system while running this code, but I would greatly appreciate feedback as soon a reasonably possible.
1998-01-06 05:26:17 +00:00
lbn = INT_MAX;
Add a hook to ffs_fsync to allow soft updates to get first chance at doing a sync on the block device for the filesystem. That allows it to push the bitmap blocks before the inode blocks which greatly reduces the number of inode rollbacks that need to be done.
1999-05-14 01:26:46 +00:00
if (vp->v_specmountpoint != NULL &&
(vp->v_specmountpoint->mnt_flag & MNT_SOFTDEP))
softdep_fsync_mountdev(vp);
Make our v_usecount vnode reference count work identically to the original BSD code. The association between the vnode and the vm_object no longer includes reference counts. The major difference is that vm_object's are no longer freed gratuitiously from the vnode, and so once an object is created for the vnode, it will last as long as the vnode does. When a vnode object reference count is incremented, then the underlying vnode reference count is incremented also. The two "objects" are now more intimately related, and so the interactions are now much less complex. When vnodes are now normally placed onto the free queue with an object still attached. The rundown of the object happens at vnode rundown time, and happens with exactly the same filesystem semantics of the original VFS code. There is absolutely no need for vnode_pager_uncache and other travesties like that anymore. A side-effect of these changes is that SMP locking should be much simpler, the I/O copyin/copyout optimizations work, NFS should be more ponderable, and further work on layered filesystems should be less frustrating, because of the totally coherent management of the vnode objects and vnodes. Please be careful with your system while running this code, but I would greatly appreciate feedback as soon a reasonably possible.
1998-01-06 05:26:17 +00:00
} else {
struct inode *ip;
ip = VTOI(vp);
lbn = lblkno(ip->i_fs, (ip->i_size + ip->i_fs->fs_bsize - 1));
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Flush all dirty buffers associated with a vnode.
*/
On our final pass through ffs_fsync, do all I/O synchronously so that we can find out if our flush is failing because of write errors. This change avoids a "flush failed" panic during unrecoverable disk errors.
1999-06-18 05:49:46 +00:00
passes = NIADDR + 1;
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
skipmeta = 0;
Several performance improvements for soft updates have been added: 1) Fastpath deletions. When a file is being deleted, check to see if it was so recently created that its inode has not yet been written to disk. If so, the delete can proceed to immediately free the inode. 2) Background writes: No file or block allocations can be done while the bitmap is being written to disk. To avoid these stalls, the bitmap is copied to another buffer which is written thus leaving the original available for futher allocations. 3) Link count tracking. Constantly track the difference in i_effnlink and i_nlink so that inodes that have had no change other than i_effnlink need not be written. 4) Identify buffers with rollback dependencies so that the buffer flushing daemon can choose to skip over them.
2000-01-10 00:24:24 +00:00
if (wait)
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
skipmeta = 1;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
s = splbio();
When fsync'ing a file on a filesystem using soft updates, we first try to write all the dirty blocks. If some of those blocks have dependencies, they will be remarked dirty when the I/O completes. On systems with really fast I/O systems, it is possible to get in an infinite loop trying to flush the buffers, because the I/O finishes before we can get all the dirty buffers off the v_dirtyblkhd list and into the I/O queue. (The previous algorithm looped over the v_dirtyblkhd list writing out buffers until the list emptied.) So, now we mark each buffer that we try to write so that we can distinguish the ones that are being remarked dirty from those that we have not yet tried to flush. Once we have tried to push every buffer once, we then push any associated metadata that is causing the remaining buffers to be redirtied. Submitted by: Matthew Dillon <dillon@apollo.backplane.com>
1999-03-02 04:04:31 +00:00
loop:
for (bp = TAILQ_FIRST(&vp->v_dirtyblkhd); bp;
bp = TAILQ_NEXT(bp, b_vnbufs))
bp->b_flags &= ~B_SCANNED;
Use TAILQ macros for clean/dirty block list processing. Set b_xflags rather than abusing the list next pointer with a magic number.
1998-10-31 15:31:29 +00:00
for (bp = TAILQ_FIRST(&vp->v_dirtyblkhd); bp; bp = nbp) {
nbp = TAILQ_NEXT(bp, b_vnbufs);
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
/*
Several performance improvements for soft updates have been added: 1) Fastpath deletions. When a file is being deleted, check to see if it was so recently created that its inode has not yet been written to disk. If so, the delete can proceed to immediately free the inode. 2) Background writes: No file or block allocations can be done while the bitmap is being written to disk. To avoid these stalls, the bitmap is copied to another buffer which is written thus leaving the original available for futher allocations. 3) Link count tracking. Constantly track the difference in i_effnlink and i_nlink so that inodes that have had no change other than i_effnlink need not be written. 4) Identify buffers with rollback dependencies so that the buffer flushing daemon can choose to skip over them.
2000-01-10 00:24:24 +00:00
* Reasons to skip this buffer: it has already been considered
* on this pass, this pass is the first time through on a
* synchronous flush request and the buffer being considered
* is metadata, the buffer has dependencies that will cause
* it to be redirtied and it has not already been deferred,
* or it is already being written.
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
*/
Several performance improvements for soft updates have been added: 1) Fastpath deletions. When a file is being deleted, check to see if it was so recently created that its inode has not yet been written to disk. If so, the delete can proceed to immediately free the inode. 2) Background writes: No file or block allocations can be done while the bitmap is being written to disk. To avoid these stalls, the bitmap is copied to another buffer which is written thus leaving the original available for futher allocations. 3) Link count tracking. Constantly track the difference in i_effnlink and i_nlink so that inodes that have had no change other than i_effnlink need not be written. 4) Identify buffers with rollback dependencies so that the buffer flushing daemon can choose to skip over them.
2000-01-10 00:24:24 +00:00
if ((bp->b_flags & B_SCANNED) != 0)
continue;
bp->b_flags |= B_SCANNED;
if ((skipmeta == 1 && bp->b_lblkno < 0))
continue;
if (!wait && LIST_FIRST(&bp->b_dep) != NULL &&
(bp->b_flags & B_DEFERRED) == 0 &&
bioops.io_countdeps && (*bioops.io_countdeps)(bp, 0)) {
bp->b_flags |= B_DEFERRED;
continue;
}
if (BUF_LOCK(bp, LK_EXCLUSIVE | LK_NOWAIT))
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
continue;
if ((bp->b_flags & B_DELWRI) == 0)
panic("ffs_fsync: not dirty");
Several performance improvements for soft updates have been added: 1) Fastpath deletions. When a file is being deleted, check to see if it was so recently created that its inode has not yet been written to disk. If so, the delete can proceed to immediately free the inode. 2) Background writes: No file or block allocations can be done while the bitmap is being written to disk. To avoid these stalls, the bitmap is copied to another buffer which is written thus leaving the original available for futher allocations. 3) Link count tracking. Constantly track the difference in i_effnlink and i_nlink so that inodes that have had no change other than i_effnlink need not be written. 4) Identify buffers with rollback dependencies so that the buffer flushing daemon can choose to skip over them.
2000-01-10 00:24:24 +00:00
if (vp != bp->b_vp)
panic("ffs_fsync: vp != vp->b_vp");
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
/*
Several performance improvements for soft updates have been added: 1) Fastpath deletions. When a file is being deleted, check to see if it was so recently created that its inode has not yet been written to disk. If so, the delete can proceed to immediately free the inode. 2) Background writes: No file or block allocations can be done while the bitmap is being written to disk. To avoid these stalls, the bitmap is copied to another buffer which is written thus leaving the original available for futher allocations. 3) Link count tracking. Constantly track the difference in i_effnlink and i_nlink so that inodes that have had no change other than i_effnlink need not be written. 4) Identify buffers with rollback dependencies so that the buffer flushing daemon can choose to skip over them.
2000-01-10 00:24:24 +00:00
* If this is a synchronous flush request, or it is not a
* file or device, start the write on this buffer immediatly.
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
*/
Several performance improvements for soft updates have been added: 1) Fastpath deletions. When a file is being deleted, check to see if it was so recently created that its inode has not yet been written to disk. If so, the delete can proceed to immediately free the inode. 2) Background writes: No file or block allocations can be done while the bitmap is being written to disk. To avoid these stalls, the bitmap is copied to another buffer which is written thus leaving the original available for futher allocations. 3) Link count tracking. Constantly track the difference in i_effnlink and i_nlink so that inodes that have had no change other than i_effnlink need not be written. 4) Identify buffers with rollback dependencies so that the buffer flushing daemon can choose to skip over them.
2000-01-10 00:24:24 +00:00
if (wait || (vp->v_type != VREG && vp->v_type != VBLK)) {
Make our v_usecount vnode reference count work identically to the original BSD code. The association between the vnode and the vm_object no longer includes reference counts. The major difference is that vm_object's are no longer freed gratuitiously from the vnode, and so once an object is created for the vnode, it will last as long as the vnode does. When a vnode object reference count is incremented, then the underlying vnode reference count is incremented also. The two "objects" are now more intimately related, and so the interactions are now much less complex. When vnodes are now normally placed onto the free queue with an object still attached. The rundown of the object happens at vnode rundown time, and happens with exactly the same filesystem semantics of the original VFS code. There is absolutely no need for vnode_pager_uncache and other travesties like that anymore. A side-effect of these changes is that SMP locking should be much simpler, the I/O copyin/copyout optimizations work, NFS should be more ponderable, and further work on layered filesystems should be less frustrating, because of the totally coherent management of the vnode objects and vnodes. Please be careful with your system while running this code, but I would greatly appreciate feedback as soon a reasonably possible.
1998-01-06 05:26:17 +00:00
Changes from John Dyson and myself: Fixed remaining known bugs in the buffer IO and VM system. vfs_bio.c: Fixed some race conditions and locking bugs. Improved performance by removing some (now) unnecessary code and fixing some broken logic. Fixed process accounting of # of FS outputs. Properly handle NFS interrupts (B_EINTR). (various) Replaced calls to clrbuf() with calls to an optimized routine called vfs_bio_clrbuf(). (various FS sync) Sync out modified vnode_pager backed pages. ffs_vnops.c: Do two passes: Sync out file data first, then indirect blocks. vm_fault.c: Fixed deadly embrace caused by acquiring locks in the wrong order. vnode_pager.c: Changed to use buffer I/O system for writing out modified pages. This should fix the problem with the modification date previous not getting updated. Also dramatically simplifies the code. Note that this is going to change in the future and be implemented via VOP_PUTPAGES(). vm_object.c: Fixed a pile of bugs related to cleaning (vnode) objects. The performance of vm_object_page_clean() is terrible when dealing with huge objects, but this will change when we implement a binary tree to keep the object pages sorted. vm_pageout.c: Fixed broken clustering of pageouts. Fixed race conditions and other lockup style bugs in the scanning of pages. Improved performance.
1995-04-09 06:03:56 +00:00
/*
On our final pass through ffs_fsync, do all I/O synchronously so that we can find out if our flush is failing because of write errors. This change avoids a "flush failed" panic during unrecoverable disk errors.
1999-06-18 05:49:46 +00:00
* On our final pass through, do all I/O synchronously
* so that we can find out if our flush is failing
* because of write errors.
Changes from John Dyson and myself: Fixed remaining known bugs in the buffer IO and VM system. vfs_bio.c: Fixed some race conditions and locking bugs. Improved performance by removing some (now) unnecessary code and fixing some broken logic. Fixed process accounting of # of FS outputs. Properly handle NFS interrupts (B_EINTR). (various) Replaced calls to clrbuf() with calls to an optimized routine called vfs_bio_clrbuf(). (various FS sync) Sync out modified vnode_pager backed pages. ffs_vnops.c: Do two passes: Sync out file data first, then indirect blocks. vm_fault.c: Fixed deadly embrace caused by acquiring locks in the wrong order. vnode_pager.c: Changed to use buffer I/O system for writing out modified pages. This should fix the problem with the modification date previous not getting updated. Also dramatically simplifies the code. Note that this is going to change in the future and be implemented via VOP_PUTPAGES(). vm_object.c: Fixed a pile of bugs related to cleaning (vnode) objects. The performance of vm_object_page_clean() is terrible when dealing with huge objects, but this will change when we implement a binary tree to keep the object pages sorted. vm_pageout.c: Fixed broken clustering of pageouts. Fixed race conditions and other lockup style bugs in the scanning of pages. Improved performance.
1995-04-09 06:03:56 +00:00
*/
Several performance improvements for soft updates have been added: 1) Fastpath deletions. When a file is being deleted, check to see if it was so recently created that its inode has not yet been written to disk. If so, the delete can proceed to immediately free the inode. 2) Background writes: No file or block allocations can be done while the bitmap is being written to disk. To avoid these stalls, the bitmap is copied to another buffer which is written thus leaving the original available for futher allocations. 3) Link count tracking. Constantly track the difference in i_effnlink and i_nlink so that inodes that have had no change other than i_effnlink need not be written. 4) Identify buffers with rollback dependencies so that the buffer flushing daemon can choose to skip over them.
2000-01-10 00:24:24 +00:00
if (passes > 0 || !wait) {
if ((bp->b_flags & B_CLUSTEROK) && !wait) {
Convert buffer locking from using the B_BUSY and B_WANTED flags to using lockmgr locks. This commit should be functionally equivalent to the old semantics. That is, all buffer locking is done with LK_EXCLUSIVE requests. Changes to take advantage of LK_SHARED and LK_RECURSIVE will be done in future commits.
1999-06-26 02:47:16 +00:00
BUF_UNLOCK(bp);
Make our v_usecount vnode reference count work identically to the original BSD code. The association between the vnode and the vm_object no longer includes reference counts. The major difference is that vm_object's are no longer freed gratuitiously from the vnode, and so once an object is created for the vnode, it will last as long as the vnode does. When a vnode object reference count is incremented, then the underlying vnode reference count is incremented also. The two "objects" are now more intimately related, and so the interactions are now much less complex. When vnodes are now normally placed onto the free queue with an object still attached. The rundown of the object happens at vnode rundown time, and happens with exactly the same filesystem semantics of the original VFS code. There is absolutely no need for vnode_pager_uncache and other travesties like that anymore. A side-effect of these changes is that SMP locking should be much simpler, the I/O copyin/copyout optimizations work, NFS should be more ponderable, and further work on layered filesystems should be less frustrating, because of the totally coherent management of the vnode objects and vnodes. Please be careful with your system while running this code, but I would greatly appreciate feedback as soon a reasonably possible.
1998-01-06 05:26:17 +00:00
(void) vfs_bio_awrite(bp);
} else {
Fix vfs_bio_awrite usage, and correct vtruncbuf usage.
1998-03-19 22:49:44 +00:00
bremfree(bp);
splx(s);
Make our v_usecount vnode reference count work identically to the original BSD code. The association between the vnode and the vm_object no longer includes reference counts. The major difference is that vm_object's are no longer freed gratuitiously from the vnode, and so once an object is created for the vnode, it will last as long as the vnode does. When a vnode object reference count is incremented, then the underlying vnode reference count is incremented also. The two "objects" are now more intimately related, and so the interactions are now much less complex. When vnodes are now normally placed onto the free queue with an object still attached. The rundown of the object happens at vnode rundown time, and happens with exactly the same filesystem semantics of the original VFS code. There is absolutely no need for vnode_pager_uncache and other travesties like that anymore. A side-effect of these changes is that SMP locking should be much simpler, the I/O copyin/copyout optimizations work, NFS should be more ponderable, and further work on layered filesystems should be less frustrating, because of the totally coherent management of the vnode objects and vnodes. Please be careful with your system while running this code, but I would greatly appreciate feedback as soon a reasonably possible.
1998-01-06 05:26:17 +00:00
(void) bawrite(bp);
When fsync'ing a file on a filesystem using soft updates, we first try to write all the dirty blocks. If some of those blocks have dependencies, they will be remarked dirty when the I/O completes. On systems with really fast I/O systems, it is possible to get in an infinite loop trying to flush the buffers, because the I/O finishes before we can get all the dirty buffers off the v_dirtyblkhd list and into the I/O queue. (The previous algorithm looped over the v_dirtyblkhd list writing out buffers until the list emptied.) So, now we mark each buffer that we try to write so that we can distinguish the ones that are being remarked dirty from those that we have not yet tried to flush. Once we have tried to push every buffer once, we then push any associated metadata that is causing the remaining buffers to be redirtied. Submitted by: Matthew Dillon <dillon@apollo.backplane.com>
1999-03-02 04:04:31 +00:00
s = splbio();
Make our v_usecount vnode reference count work identically to the original BSD code. The association between the vnode and the vm_object no longer includes reference counts. The major difference is that vm_object's are no longer freed gratuitiously from the vnode, and so once an object is created for the vnode, it will last as long as the vnode does. When a vnode object reference count is incremented, then the underlying vnode reference count is incremented also. The two "objects" are now more intimately related, and so the interactions are now much less complex. When vnodes are now normally placed onto the free queue with an object still attached. The rundown of the object happens at vnode rundown time, and happens with exactly the same filesystem semantics of the original VFS code. There is absolutely no need for vnode_pager_uncache and other travesties like that anymore. A side-effect of these changes is that SMP locking should be much simpler, the I/O copyin/copyout optimizations work, NFS should be more ponderable, and further work on layered filesystems should be less frustrating, because of the totally coherent management of the vnode objects and vnodes. Please be careful with your system while running this code, but I would greatly appreciate feedback as soon a reasonably possible.
1998-01-06 05:26:17 +00:00
}
} else {
Fix vfs_bio_awrite usage, and correct vtruncbuf usage.
1998-03-19 22:49:44 +00:00
bremfree(bp);
splx(s);
On our final pass through ffs_fsync, do all I/O synchronously so that we can find out if our flush is failing because of write errors. This change avoids a "flush failed" panic during unrecoverable disk errors.
1999-06-18 05:49:46 +00:00
if ((error = bwrite(bp)) != 0)
return (error);
When fsync'ing a file on a filesystem using soft updates, we first try to write all the dirty blocks. If some of those blocks have dependencies, they will be remarked dirty when the I/O completes. On systems with really fast I/O systems, it is possible to get in an infinite loop trying to flush the buffers, because the I/O finishes before we can get all the dirty buffers off the v_dirtyblkhd list and into the I/O queue. (The previous algorithm looped over the v_dirtyblkhd list writing out buffers until the list emptied.) So, now we mark each buffer that we try to write so that we can distinguish the ones that are being remarked dirty from those that we have not yet tried to flush. Once we have tried to push every buffer once, we then push any associated metadata that is causing the remaining buffers to be redirtied. Submitted by: Matthew Dillon <dillon@apollo.backplane.com>
1999-03-02 04:04:31 +00:00
s = splbio();
Make our v_usecount vnode reference count work identically to the original BSD code. The association between the vnode and the vm_object no longer includes reference counts. The major difference is that vm_object's are no longer freed gratuitiously from the vnode, and so once an object is created for the vnode, it will last as long as the vnode does. When a vnode object reference count is incremented, then the underlying vnode reference count is incremented also. The two "objects" are now more intimately related, and so the interactions are now much less complex. When vnodes are now normally placed onto the free queue with an object still attached. The rundown of the object happens at vnode rundown time, and happens with exactly the same filesystem semantics of the original VFS code. There is absolutely no need for vnode_pager_uncache and other travesties like that anymore. A side-effect of these changes is that SMP locking should be much simpler, the I/O copyin/copyout optimizations work, NFS should be more ponderable, and further work on layered filesystems should be less frustrating, because of the totally coherent management of the vnode objects and vnodes. Please be careful with your system while running this code, but I would greatly appreciate feedback as soon a reasonably possible.
1998-01-06 05:26:17 +00:00
}
} else if ((vp->v_type == VREG) && (bp->b_lblkno >= lbn)) {
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
/*
* If the buffer is for data that has been truncated
* off the file, then throw it away.
*/
Make our v_usecount vnode reference count work identically to the original BSD code. The association between the vnode and the vm_object no longer includes reference counts. The major difference is that vm_object's are no longer freed gratuitiously from the vnode, and so once an object is created for the vnode, it will last as long as the vnode does. When a vnode object reference count is incremented, then the underlying vnode reference count is incremented also. The two "objects" are now more intimately related, and so the interactions are now much less complex. When vnodes are now normally placed onto the free queue with an object still attached. The rundown of the object happens at vnode rundown time, and happens with exactly the same filesystem semantics of the original VFS code. There is absolutely no need for vnode_pager_uncache and other travesties like that anymore. A side-effect of these changes is that SMP locking should be much simpler, the I/O copyin/copyout optimizations work, NFS should be more ponderable, and further work on layered filesystems should be less frustrating, because of the totally coherent management of the vnode objects and vnodes. Please be careful with your system while running this code, but I would greatly appreciate feedback as soon a reasonably possible.
1998-01-06 05:26:17 +00:00
bremfree(bp);
Convert buffer locking from using the B_BUSY and B_WANTED flags to using lockmgr locks. This commit should be functionally equivalent to the old semantics. That is, all buffer locking is done with LK_EXCLUSIVE requests. Changes to take advantage of LK_SHARED and LK_RECURSIVE will be done in future commits.
1999-06-26 02:47:16 +00:00
bp->b_flags |= B_INVAL | B_NOCACHE;
Make our v_usecount vnode reference count work identically to the original BSD code. The association between the vnode and the vm_object no longer includes reference counts. The major difference is that vm_object's are no longer freed gratuitiously from the vnode, and so once an object is created for the vnode, it will last as long as the vnode does. When a vnode object reference count is incremented, then the underlying vnode reference count is incremented also. The two "objects" are now more intimately related, and so the interactions are now much less complex. When vnodes are now normally placed onto the free queue with an object still attached. The rundown of the object happens at vnode rundown time, and happens with exactly the same filesystem semantics of the original VFS code. There is absolutely no need for vnode_pager_uncache and other travesties like that anymore. A side-effect of these changes is that SMP locking should be much simpler, the I/O copyin/copyout optimizations work, NFS should be more ponderable, and further work on layered filesystems should be less frustrating, because of the totally coherent management of the vnode objects and vnodes. Please be careful with your system while running this code, but I would greatly appreciate feedback as soon a reasonably possible.
1998-01-06 05:26:17 +00:00
splx(s);
When fsync'ing a file on a filesystem using soft updates, we first try to write all the dirty blocks. If some of those blocks have dependencies, they will be remarked dirty when the I/O completes. On systems with really fast I/O systems, it is possible to get in an infinite loop trying to flush the buffers, because the I/O finishes before we can get all the dirty buffers off the v_dirtyblkhd list and into the I/O queue. (The previous algorithm looped over the v_dirtyblkhd list writing out buffers until the list emptied.) So, now we mark each buffer that we try to write so that we can distinguish the ones that are being remarked dirty from those that we have not yet tried to flush. Once we have tried to push every buffer once, we then push any associated metadata that is causing the remaining buffers to be redirtied. Submitted by: Matthew Dillon <dillon@apollo.backplane.com>
1999-03-02 04:04:31 +00:00
brelse(bp);
s = splbio();
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
} else {
Convert buffer locking from using the B_BUSY and B_WANTED flags to using lockmgr locks. This commit should be functionally equivalent to the old semantics. That is, all buffer locking is done with LK_EXCLUSIVE requests. Changes to take advantage of LK_SHARED and LK_RECURSIVE will be done in future commits.
1999-06-26 02:47:16 +00:00
BUF_UNLOCK(bp);
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
vfs_bio_awrite(bp);
}
When fsync'ing a file on a filesystem using soft updates, we first try to write all the dirty blocks. If some of those blocks have dependencies, they will be remarked dirty when the I/O completes. On systems with really fast I/O systems, it is possible to get in an infinite loop trying to flush the buffers, because the I/O finishes before we can get all the dirty buffers off the v_dirtyblkhd list and into the I/O queue. (The previous algorithm looped over the v_dirtyblkhd list writing out buffers until the list emptied.) So, now we mark each buffer that we try to write so that we can distinguish the ones that are being remarked dirty from those that we have not yet tried to flush. Once we have tried to push every buffer once, we then push any associated metadata that is causing the remaining buffers to be redirtied. Submitted by: Matthew Dillon <dillon@apollo.backplane.com>
1999-03-02 04:04:31 +00:00
/*
* Since we may have slept during the I/O, we need
* to start from a known point.
*/
nbp = TAILQ_FIRST(&vp->v_dirtyblkhd);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
/*
* If we were asked to do this synchronously, then go back for
* another pass, this time doing the metadata.
*/
if (skipmeta) {
skipmeta = 0;
When fsync'ing a file on a filesystem using soft updates, we first try to write all the dirty blocks. If some of those blocks have dependencies, they will be remarked dirty when the I/O completes. On systems with really fast I/O systems, it is possible to get in an infinite loop trying to flush the buffers, because the I/O finishes before we can get all the dirty buffers off the v_dirtyblkhd list and into the I/O queue. (The previous algorithm looped over the v_dirtyblkhd list writing out buffers until the list emptied.) So, now we mark each buffer that we try to write so that we can distinguish the ones that are being remarked dirty from those that we have not yet tried to flush. Once we have tried to push every buffer once, we then push any associated metadata that is causing the remaining buffers to be redirtied. Submitted by: Matthew Dillon <dillon@apollo.backplane.com>
1999-03-02 04:04:31 +00:00
goto loop;
Changes from John Dyson and myself: Fixed remaining known bugs in the buffer IO and VM system. vfs_bio.c: Fixed some race conditions and locking bugs. Improved performance by removing some (now) unnecessary code and fixing some broken logic. Fixed process accounting of # of FS outputs. Properly handle NFS interrupts (B_EINTR). (various) Replaced calls to clrbuf() with calls to an optimized routine called vfs_bio_clrbuf(). (various FS sync) Sync out modified vnode_pager backed pages. ffs_vnops.c: Do two passes: Sync out file data first, then indirect blocks. vm_fault.c: Fixed deadly embrace caused by acquiring locks in the wrong order. vnode_pager.c: Changed to use buffer I/O system for writing out modified pages. This should fix the problem with the modification date previous not getting updated. Also dramatically simplifies the code. Note that this is going to change in the future and be implemented via VOP_PUTPAGES(). vm_object.c: Fixed a pile of bugs related to cleaning (vnode) objects. The performance of vm_object_page_clean() is terrible when dealing with huge objects, but this will change when we implement a binary tree to keep the object pages sorted. vm_pageout.c: Fixed broken clustering of pageouts. Fixed race conditions and other lockup style bugs in the scanning of pages. Improved performance.
1995-04-09 06:03:56 +00:00
}
Several performance improvements for soft updates have been added: 1) Fastpath deletions. When a file is being deleted, check to see if it was so recently created that its inode has not yet been written to disk. If so, the delete can proceed to immediately free the inode. 2) Background writes: No file or block allocations can be done while the bitmap is being written to disk. To avoid these stalls, the bitmap is copied to another buffer which is written thus leaving the original available for futher allocations. 3) Link count tracking. Constantly track the difference in i_effnlink and i_nlink so that inodes that have had no change other than i_effnlink need not be written. 4) Identify buffers with rollback dependencies so that the buffer flushing daemon can choose to skip over them.
2000-01-10 00:24:24 +00:00
if (wait) {
Back out John's changes 1.45 -> 1.46 Kirk confirms that the original semantic was what he wanted... (well, a very slight difference) May fix "dangling deps" panic with soft updates.
1998-06-10 19:27:56 +00:00
while (vp->v_numoutput) {
vp->v_flag |= VBWAIT;
(void) tsleep((caddr_t)&vp->v_numoutput,
PRIBIO + 4, "ffsfsn", 0);
}
Softdep_sync_metadata appears to expect that it is called at splbio, so make it so...
1998-03-21 05:16:09 +00:00
Back out John's changes 1.45 -> 1.46 Kirk confirms that the original semantic was what he wanted... (well, a very slight difference) May fix "dangling deps" panic with soft updates.
1998-06-10 19:27:56 +00:00
/*
* Ensure that any filesystem metatdata associated
* with the vnode has been written.
*/
splx(s);
if ((error = softdep_sync_metadata(ap)) != 0)
return (error);
Changes from John Dyson and myself: Fixed remaining known bugs in the buffer IO and VM system. vfs_bio.c: Fixed some race conditions and locking bugs. Improved performance by removing some (now) unnecessary code and fixing some broken logic. Fixed process accounting of # of FS outputs. Properly handle NFS interrupts (B_EINTR). (various) Replaced calls to clrbuf() with calls to an optimized routine called vfs_bio_clrbuf(). (various FS sync) Sync out modified vnode_pager backed pages. ffs_vnops.c: Do two passes: Sync out file data first, then indirect blocks. vm_fault.c: Fixed deadly embrace caused by acquiring locks in the wrong order. vnode_pager.c: Changed to use buffer I/O system for writing out modified pages. This should fix the problem with the modification date previous not getting updated. Also dramatically simplifies the code. Note that this is going to change in the future and be implemented via VOP_PUTPAGES(). vm_object.c: Fixed a pile of bugs related to cleaning (vnode) objects. The performance of vm_object_page_clean() is terrible when dealing with huge objects, but this will change when we implement a binary tree to keep the object pages sorted. vm_pageout.c: Fixed broken clustering of pageouts. Fixed race conditions and other lockup style bugs in the scanning of pages. Improved performance.
1995-04-09 06:03:56 +00:00
s = splbio();
Softdep_sync_metadata appears to expect that it is called at splbio, so make it so...
1998-03-21 05:16:09 +00:00
Use TAILQ macros for clean/dirty block list processing. Set b_xflags rather than abusing the list next pointer with a magic number.
1998-10-31 15:31:29 +00:00
if (!TAILQ_EMPTY(&vp->v_dirtyblkhd)) {
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
/*
* Block devices associated with filesystems may
* have new I/O requests posted for them even if
* the vnode is locked, so no amount of trying will
* get them clean. Thus we give block devices a
* good effort, then just give up. For all other file
* types, go around and try again until it is clean.
*/
if (passes > 0) {
passes -= 1;
When fsync'ing a file on a filesystem using soft updates, we first try to write all the dirty blocks. If some of those blocks have dependencies, they will be remarked dirty when the I/O completes. On systems with really fast I/O systems, it is possible to get in an infinite loop trying to flush the buffers, because the I/O finishes before we can get all the dirty buffers off the v_dirtyblkhd list and into the I/O queue. (The previous algorithm looped over the v_dirtyblkhd list writing out buffers until the list emptied.) So, now we mark each buffer that we try to write so that we can distinguish the ones that are being remarked dirty from those that we have not yet tried to flush. Once we have tried to push every buffer once, we then push any associated metadata that is causing the remaining buffers to be redirtied. Submitted by: Matthew Dillon <dillon@apollo.backplane.com>
1999-03-02 04:04:31 +00:00
goto loop;
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
}
#ifdef DIAGNOSTIC
Give vn_isdisk() a second argument where it can return a suitable errno. Suggested by: bde
2000-01-10 12:04:27 +00:00
if (!vn_isdisk(vp, NULL))
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
vprint("ffs_fsync: dirty", vp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#endif
Reviewed by: dyson@freebsd.org (john Dyson), dg@root.com (david greenman) Submitted by: Kirk McKusick (mcKusick@mckusick.com) Obtained from: WHistle development tree
1998-03-08 09:59:44 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Back out John's changes 1.45 -> 1.46 Kirk confirms that the original semantic was what he wanted... (well, a very slight difference) May fix "dangling deps" panic with soft updates.
1998-06-10 19:27:56 +00:00
splx(s);
Several performance improvements for soft updates have been added: 1) Fastpath deletions. When a file is being deleted, check to see if it was so recently created that its inode has not yet been written to disk. If so, the delete can proceed to immediately free the inode. 2) Background writes: No file or block allocations can be done while the bitmap is being written to disk. To avoid these stalls, the bitmap is copied to another buffer which is written thus leaving the original available for futher allocations. 3) Link count tracking. Constantly track the difference in i_effnlink and i_nlink so that inodes that have had no change other than i_effnlink need not be written. 4) Identify buffers with rollback dependencies so that the buffer flushing daemon can choose to skip over them.
2000-01-10 00:24:24 +00:00
return (UFS_UPDATE(vp, wait));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Reference in New Issue Copy Permalink