Incorporate the enhanced ffs_dirpref() by Grigoriy Orlov, as found in
FreeBSD (three commits; the initial work, man page updates, and a fix
to ffs_reload()), with the following differences:
- Be consistent between newfs(8) and tunefs(8) as to the options which
set and control the tuning parameters for this work (avgfilesize & avgfpdir)
- Use u_int16_t instead of u_int8_t to keep track of the number of
contiguous directories (suggested by Chuck Silvers)
- Work within our FFS_EI framework
- Ensure that fs->fs_maxclusters and fs->fs_contigdirs don't point to
the same area of memory
The new algorithm has a marked performance increase, especially when
performing tasks such as untarring pkgsrc.tar.gz, etc.
The original FreeBSD commit messages are attached:
=====
mckusick 2001/04/10 01:39:00 PDT
Directory layout preference improvements from Grigoriy Orlov <gluk@ptci.ru>.
His description of the problem and solution follow. My own tests show
speedups on typical filesystem intensive workloads of 5% to 12% which
is very impressive considering the small amount of code change involved.
------
One day I noticed that some file operations run much faster on
small file systems then on big ones. I've looked at the ffs
algorithms, thought about them, and redesigned the dirpref algorithm.
First I want to describe the results of my tests. These results are old
and I have improved the algorithm after these tests were done. Nevertheless
they show how big the perfomance speedup may be. I have done two file/directory
intensive tests on a two OpenBSD systems with old and new dirpref algorithm.
The first test is "tar -xzf ports.tar.gz", the second is "rm -rf ports".
The ports.tar.gz file is the ports collection from the OpenBSD 2.8 release.
It contains 6596 directories and 13868 files. The test systems are:
1. Celeron-450, 128Mb, two IDE drives, the system at wd0, file system for
test is at wd1. Size of test file system is 8 Gb, number of cg=991,
size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current
from Dec 2000 with BUFCACHEPERCENT=35
2. PIII-600, 128Mb, two IBM DTLA-307045 IDE drives at i815e, the system
at wd0, file system for test is at wd1. Size of test file system is 40 Gb,
number of cg=5324, size of cg is 8m, block size = 8k, fragment size = 1k
OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=50
You can get more info about the test systems and methods at:
http://www.ptci.ru/gluk/dirpref/old/dirpref.html
Test Results
tar -xzf ports.tar.gz rm -rf ports
mode old dirpref new dirpref speedup old dirprefnew dirpref speedup
First system
normal 667 472 1.41 477 331 1.44
async 285 144 1.98 130 14 9.29
sync 768 616 1.25 477 334 1.43
softdep 413 252 1.64 241 38 6.34
Second system
normal 329 81 4.06 263.5 93.5 2.81
async 302 25.7 11.75 112 2.26 49.56
sync 281 57.0 4.93 263 90.5 2.9
softdep 341 40.6 8.4 284 4.76 59.66
"old dirpref" and "new dirpref" columns give a test time in seconds.
speedup - speed increasement in times, ie. old dirpref / new dirpref.
------
Algorithm description
The old dirpref algorithm is described in comments:
/*
* Find a cylinder to place a directory.
*
* The policy implemented by this algorithm is to select from
* among those cylinder groups with above the average number of
* free inodes, the one with the smallest number of directories.
*/
A new directory is allocated in a different cylinder groups than its
parent directory resulting in a directory tree that is spreaded across
all the cylinder groups. This spreading out results in a non-optimal
access to the directories and files. When we have a small filesystem
it is not a problem but when the filesystem is big then perfomance
degradation becomes very apparent.
What I mean by a big file system ?
1. A big filesystem is a filesystem which occupy 20-30 or more percent
of total drive space, i.e. first and last cylinder are physically
located relatively far from each other.
2. It has a relatively large number of cylinder groups, for example
more cylinder groups than 50% of the buffers in the buffer cache.
The first results in long access times, while the second results in
many buffers being used by metadata operations. Such operations use
cylinder group blocks and on-disk inode blocks. The cylinder group
block (fs->fs_cblkno) contains struct cg, inode and block bit maps.
It is 2k in size for the default filesystem parameters. If new and
parent directories are located in different cylinder groups then the
system performs more input/output operations and uses more buffers.
On filesystems with many cylinder groups, lots of cache buffers are
used for metadata operations.
My solution for this problem is very simple. I allocate many directories
in one cylinder group. I also do some things, so that the new allocation
method does not cause excessive fragmentation and all directory inodes
will not be located at a location far from its file's inodes and data.
The algorithm is:
/*
* Find a cylinder group to place a directory.
*
* The policy implemented by this algorithm is to allocate a
* directory inode in the same cylinder group as its parent
* directory, but also to reserve space for its files inodes
* and data. Restrict the number of directories which may be
* allocated one after another in the same cylinder group
* without intervening allocation of files.
*
* If we allocate a first level directory then force allocation
* in another cylinder group.
*/
My early versions of dirpref give me a good results for a wide range of
file operations and different filesystem capacities except one case:
those applications that create their entire directory structure first
and only later fill this structure with files.
My solution for such and similar cases is to limit a number of
directories which may be created one after another in the same cylinder
group without intervening file creations. For this purpose, I allocate
an array of counters at mount time. This array is linked to the superblock
fs->fs_contigdirs[cg]. Each time a directory is created the counter
increases and each time a file is created the counter decreases. A 60Gb
filesystem with 8mb/cg requires 10kb of memory for the counters array.
The maxcontigdirs is a maximum number of directories which may be created
without an intervening file creation. I found in my tests that the best
performance occurs when I restrict the number of directories in one cylinder
group such that all its files may be located in the same cylinder group.
There may be some deterioration in performance if all the file inodes
are in the same cylinder group as its containing directory, but their
data partially resides in a different cylinder group. The maxcontigdirs
value is calculated to try to prevent this condition. Since there is
no way to know how many files and directories will be allocated later
I added two optimization parameters in superblock/tunefs. They are:
int32_t fs_avgfilesize; /* expected average file size */
int32_t fs_avgfpdir; /* expected # of files per directory */
These parameters have reasonable defaults but may be tweeked for special
uses of a filesystem. They are only necessary in rare cases like better
tuning a filesystem being used to store a squid cache.
I have been using this algorithm for about 3 months. I have done
a lot of testing on filesystems with different capacities, average
filesize, average number of files per directory, and so on. I think
this algorithm has no negative impact on filesystem perfomance. It
works better than the default one in all cases. The new dirpref
will greatly improve untarring/removing/coping of big directories,
decrease load on cvs servers and much more. The new dirpref doesn't
speedup a compilation process, but also doesn't slow it down.
Obtained from: Grigoriy Orlov <gluk@ptci.ru>
=====
=====
iedowse 2001/04/23 17:37:17 PDT
Pre-dirpref versions of fsck may zero out the new superblock fields
fs_contigdirs, fs_avgfilesize and fs_avgfpdir. This could cause
panics if these fields were zeroed while a filesystem was mounted
read-only, and then remounted read-write.
Add code to ffs_reload() which copies the fs_contigdirs pointer
from the previous superblock, and reinitialises fs_avgf* if necessary.
Reviewed by: mckusick
=====
=====
nik 2001/04/10 03:36:44 PDT
Add information about the new options to newfs and tunefs which set the
expected average file size and number of files per directory. Could do
with some fleshing out.
=====
This commit is contained in:
lukem
parent
8b59971a0d
commit
5c2ee5861d
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: setup.c,v 1.48 2001/09/03 14:53:31 lukem Exp $ */
|
||||
/* $NetBSD: setup.c,v 1.49 2001/09/06 02:16:00 lukem Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright (c) 1980, 1986, 1993
|
||||
@ -38,7 +38,7 @@
|
||||
#if 0
|
||||
static char sccsid[] = "@(#)setup.c 8.10 (Berkeley) 5/9/95";
|
||||
#else
|
||||
__RCSID("$NetBSD: setup.c,v 1.48 2001/09/03 14:53:31 lukem Exp $");
|
||||
__RCSID("$NetBSD: setup.c,v 1.49 2001/09/06 02:16:00 lukem Exp $");
|
||||
#endif
|
||||
#endif /* not lint */
|
||||
|
||||
@ -592,9 +592,14 @@ cmpsblks(const struct fs *sb, struct fs *asb)
|
||||
asb->fs_rotdelay = sb->fs_rotdelay;
|
||||
asb->fs_maxbpg = sb->fs_maxbpg;
|
||||
memmove(asb->fs_ocsp, sb->fs_ocsp, sizeof sb->fs_ocsp);
|
||||
asb->fs_contigdirs = sb->fs_contigdirs;
|
||||
asb->fs_csp = sb->fs_csp;
|
||||
asb->fs_maxcluster = sb->fs_maxcluster;
|
||||
memmove(asb->fs_fsmnt, sb->fs_fsmnt, sizeof sb->fs_fsmnt);
|
||||
memmove(asb->fs_snapinum,
|
||||
sb->fs_snapinum, sizeof sb->fs_snapinum);
|
||||
asb->fs_avgfilesize = sb->fs_avgfilesize;
|
||||
asb->fs_avgfpdir = sb->fs_avgfpdir;
|
||||
memmove(asb->fs_sparecon,
|
||||
sb->fs_sparecon, sizeof sb->fs_sparecon);
|
||||
/*
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: extern.h,v 1.5 2001/08/30 14:37:26 lukem Exp $ */
|
||||
/* $NetBSD: extern.h,v 1.6 2001/09/06 02:16:01 lukem Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright (c) 1997 Christos Zoulas. All rights reserved.
|
||||
@ -58,6 +58,8 @@ extern int maxbpg; /* maximum blocks per file in a cyl group */
|
||||
extern int nrpos; /* # of distinguished rotational positions */
|
||||
extern int bbsize; /* boot block size */
|
||||
extern int sbsize; /* superblock size */
|
||||
extern int avgfilesize; /* expected average file size */
|
||||
extern int avgfpdir; /* expected number of files per directory */
|
||||
extern u_long memleft; /* virtual memory available */
|
||||
extern caddr_t membase; /* start address of memory based filesystem */
|
||||
extern int needswap; /* Filesystem not in native byte order */
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: mkfs.c,v 1.54 2001/09/02 01:58:32 lukem Exp $ */
|
||||
/* $NetBSD: mkfs.c,v 1.55 2001/09/06 02:16:01 lukem Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright (c) 1980, 1989, 1993
|
||||
@ -38,7 +38,7 @@
|
||||
#if 0
|
||||
static char sccsid[] = "@(#)mkfs.c 8.11 (Berkeley) 5/3/95";
|
||||
#else
|
||||
__RCSID("$NetBSD: mkfs.c,v 1.54 2001/09/02 01:58:32 lukem Exp $");
|
||||
__RCSID("$NetBSD: mkfs.c,v 1.55 2001/09/06 02:16:01 lukem Exp $");
|
||||
#endif
|
||||
#endif /* not lint */
|
||||
|
||||
@ -164,6 +164,17 @@ mkfs(struct partition *pp, const char *fsys, int fi, int fo)
|
||||
printf("preposterous ntrak %d\n", sblock.fs_ntrak), exit(14);
|
||||
if (sblock.fs_nsect <= 0)
|
||||
printf("preposterous nsect %d\n", sblock.fs_nsect), exit(15);
|
||||
/*
|
||||
* collect and verify the filesystem density info
|
||||
*/
|
||||
sblock.fs_avgfilesize = avgfilesize;
|
||||
sblock.fs_avgfpdir = avgfpdir;
|
||||
if (sblock.fs_avgfilesize <= 0)
|
||||
printf("illegal expected average file size %d\n",
|
||||
sblock.fs_avgfilesize), exit(14);
|
||||
if (sblock.fs_avgfpdir <= 0)
|
||||
printf("illegal expected number of files per directory %d\n",
|
||||
sblock.fs_avgfpdir), exit(15);
|
||||
/*
|
||||
* collect and verify the block and fragment sizes
|
||||
*/
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
.\" $NetBSD: newfs.8,v 1.32 2001/08/31 08:03:27 lukem Exp $
|
||||
.\" $NetBSD: newfs.8,v 1.33 2001/09/06 02:16:01 lukem Exp $
|
||||
.\"
|
||||
.\" Copyright (c) 1983, 1987, 1991, 1993, 1994
|
||||
.\" The Regents of the University of California. All rights reserved.
|
||||
@ -33,7 +33,7 @@
|
||||
.\"
|
||||
.\" @(#)newfs.8 8.6 (Berkeley) 5/3/95
|
||||
.\"
|
||||
.Dd August 31, 2001
|
||||
.Dd September 6, 2001
|
||||
.Dt NEWFS 8
|
||||
.Os
|
||||
.Sh NAME
|
||||
@ -52,6 +52,8 @@
|
||||
.Op Fl d Ar rotdelay
|
||||
.Op Fl e Ar maxbpg
|
||||
.Op Fl f Ar frag-size
|
||||
.Op Fl g Ar avgfilesize
|
||||
.Op Fl h Ar avgfpdir
|
||||
.Op Fl i Ar bytes-per-inode
|
||||
.Op Fl k Ar skew
|
||||
.Op Fl l Ar interleave
|
||||
@ -203,6 +205,10 @@ See
|
||||
for more details on how to set this option.
|
||||
.It Fl f Ar frag-size
|
||||
The fragment size of the file system in bytes.
|
||||
.It Fl g Ar avgfilesize
|
||||
The expected average file size for the file system.
|
||||
.It Fl h Ar avgfpdir
|
||||
The expected average number of files per directory on the file system.
|
||||
.It Fl i Ar bytes-per-inode
|
||||
This specifies the density of inodes in the file system.
|
||||
The default is to create an inode for each 4096 bytes of data space.
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: newfs.c,v 1.46 2001/08/30 14:37:26 lukem Exp $ */
|
||||
/* $NetBSD: newfs.c,v 1.47 2001/09/06 02:16:01 lukem Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright (c) 1983, 1989, 1993, 1994
|
||||
@ -43,7 +43,7 @@ __COPYRIGHT("@(#) Copyright (c) 1983, 1989, 1993, 1994\n\
|
||||
#if 0
|
||||
static char sccsid[] = "@(#)newfs.c 8.13 (Berkeley) 5/1/95";
|
||||
#else
|
||||
__RCSID("$NetBSD: newfs.c,v 1.46 2001/08/30 14:37:26 lukem Exp $");
|
||||
__RCSID("$NetBSD: newfs.c,v 1.47 2001/09/06 02:16:01 lukem Exp $");
|
||||
#endif
|
||||
#endif /* not lint */
|
||||
|
||||
@ -179,6 +179,8 @@ int maxcontig = 0; /* max contiguous blocks to allocate */
|
||||
int rotdelay = ROTDELAY; /* rotational delay between blocks */
|
||||
int maxbpg; /* maximum blocks per file in a cyl group */
|
||||
int nrpos = NRPOS; /* # of distinguished rotational positions */
|
||||
int avgfilesize = AVFILESIZ;/* expected average file size */
|
||||
int avgfpdir = AFPDIR; /* expected number of files per directory */
|
||||
int bbsize = BBSIZE; /* boot block size */
|
||||
int sbsize = SBSIZE; /* superblock size */
|
||||
int mntflags = MNT_ASYNC; /* flags to be passed to mount */
|
||||
@ -225,8 +227,8 @@ main(int argc, char *argv[])
|
||||
errx(1, "insane maxpartitions value %d", maxpartitions);
|
||||
|
||||
opstring = mfs ?
|
||||
"NT:a:b:c:d:e:f:i:m:o:s:" :
|
||||
"B:FNOS:T:Za:b:c:d:e:f:i:k:l:m:n:o:p:r:s:t:u:x:";
|
||||
"NT:a:b:c:d:e:f:g:h:i:m:o:s:" :
|
||||
"B:FNOS:T:Za:b:c:d:e:f:g:h:i:k:l:m:n:o:p:r:s:t:u:x:";
|
||||
while ((ch = getopt(argc, argv, opstring)) != -1)
|
||||
switch (ch) {
|
||||
case 'B':
|
||||
@ -288,6 +290,14 @@ main(int argc, char *argv[])
|
||||
fsize = strsuftoi("fragment size",
|
||||
optarg, 1, INT_MAX);
|
||||
break;
|
||||
case 'g':
|
||||
avgfilesize = strsuftoi("average file size",
|
||||
optarg, 1, INT_MAX);
|
||||
break;
|
||||
case 'h':
|
||||
avgfpdir = strsuftoi("expected files per directory",
|
||||
optarg, 1, INT_MAX);
|
||||
break;
|
||||
case 'i':
|
||||
density = strsuftoi("bytes per inode",
|
||||
optarg, 1, INT_MAX);
|
||||
@ -865,6 +875,8 @@ usage(void)
|
||||
fprintf(stderr, "\t-e maxbpg\tmaximum blocks per file "
|
||||
"in a cylinder group\n");
|
||||
fprintf(stderr, "\t-f fsize\tfragment size\n");
|
||||
fprintf(stderr, "\t-g average file size\n");
|
||||
fprintf(stderr, "\t-h average files per directory\n");
|
||||
fprintf(stderr, "\t-i density\tnumber of bytes per inode\n");
|
||||
if (!mfs) {
|
||||
fprintf(stderr,
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
.\" $NetBSD: tunefs.8,v 1.25 2001/09/03 16:30:02 lukem Exp $
|
||||
.\" $NetBSD: tunefs.8,v 1.26 2001/09/06 02:16:01 lukem Exp $
|
||||
.\"
|
||||
.\" Copyright (c) 1983, 1991, 1993
|
||||
.\" The Regents of the University of California. All rights reserved.
|
||||
@ -33,7 +33,7 @@
|
||||
.\"
|
||||
.\" @(#)tunefs.8 8.3 (Berkeley) 5/3/95
|
||||
.\"
|
||||
.Dd September 3, 2001
|
||||
.Dd September 6, 2001
|
||||
.Dt TUNEFS 8
|
||||
.Os
|
||||
.Sh NAME
|
||||
@ -45,6 +45,8 @@
|
||||
.Op Fl a Ar maxcontig
|
||||
.Op Fl d Ar rotdelay
|
||||
.Op Fl e Ar maxbpg
|
||||
.Op Fl g Ar avgfilesize
|
||||
.Op Fl h Ar avgfpdir
|
||||
.Op Fl k Ar trackskew
|
||||
.Op Fl m Ar minfree
|
||||
.Bk -words
|
||||
@ -111,6 +113,10 @@ more frequently than if they were allowed to allocate all the blocks
|
||||
in a cylinder group before seeking elsewhere.
|
||||
For file systems with exclusively large files,
|
||||
this parameter should be set higher.
|
||||
.It Fl g Ar avgfilesize
|
||||
This specifies the expected average file size.
|
||||
.It Fl h Ar avgfpdir
|
||||
This specifies the expected number of files per directory.
|
||||
.It Fl k Ar trackskew
|
||||
This specifies the skew in sectors from one track to the next in a cylinder.
|
||||
The default value is zero, indicating that each track in a cylinder begins at
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: tunefs.c,v 1.23 2001/09/03 15:04:39 lukem Exp $ */
|
||||
/* $NetBSD: tunefs.c,v 1.24 2001/09/06 02:16:01 lukem Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright (c) 1983, 1993
|
||||
@ -43,7 +43,7 @@ __COPYRIGHT("@(#) Copyright (c) 1983, 1993\n\
|
||||
#if 0
|
||||
static char sccsid[] = "@(#)tunefs.c 8.3 (Berkeley) 5/3/95";
|
||||
#else
|
||||
__RCSID("$NetBSD: tunefs.c,v 1.23 2001/09/03 15:04:39 lukem Exp $");
|
||||
__RCSID("$NetBSD: tunefs.c,v 1.24 2001/09/06 02:16:01 lukem Exp $");
|
||||
#endif
|
||||
#endif /* not lint */
|
||||
|
||||
@ -95,9 +95,9 @@ main(int argc, char *argv[])
|
||||
{
|
||||
#ifdef TUNEFS_SOFTDEP
|
||||
int softdep;
|
||||
#define OPTSTRING "AFNa:d:e:k:m:n:o:t:"
|
||||
#define OPTSTRING "AFNa:d:e:g:h:k:m:n:o:t:"
|
||||
#else
|
||||
#define OPTSTRING "AFNa:d:e:k:m:o:t:"
|
||||
#define OPTSTRING "AFNa:d:e:g:h:k:m:o:t:"
|
||||
#endif
|
||||
struct stat st;
|
||||
int i, ch, Aflag, Fflag, Nflag;
|
||||
@ -105,9 +105,11 @@ main(int argc, char *argv[])
|
||||
const char *special, *chg[2];
|
||||
char device[MAXPATHLEN];
|
||||
int maxbpg, maxcontig, minfree, rotdelay, optim, trackskew;
|
||||
int avgfilesize, avgfpdir;
|
||||
|
||||
Aflag = Fflag = Nflag = 0;
|
||||
maxbpg = maxcontig = minfree = rotdelay = optim = trackskew = -1;
|
||||
avgfilesize = avgfpdir = -1;
|
||||
#ifdef TUNEFS_SOFTDEP
|
||||
softdep = -1;
|
||||
#endif
|
||||
@ -146,6 +148,16 @@ main(int argc, char *argv[])
|
||||
1, INT_MAX);
|
||||
break;
|
||||
|
||||
case 'g':
|
||||
avgfilesize = getnum(optarg,
|
||||
"average file size", 1, INT_MAX);
|
||||
break;
|
||||
|
||||
case 'h':
|
||||
avgfpdir = getnum(optarg,
|
||||
"expected number of files per directory",
|
||||
1, INT_MAX);
|
||||
break;
|
||||
|
||||
case 'm':
|
||||
minfree = getnum(optarg,
|
||||
@ -271,6 +283,10 @@ main(int argc, char *argv[])
|
||||
}
|
||||
CHANGEVAL(sblock.fs_trackskew, trackskew,
|
||||
"track skew in sectors", "");
|
||||
CHANGEVAL(sblock.fs_avgfilesize, avgfilesize,
|
||||
"average file size", "");
|
||||
CHANGEVAL(sblock.fs_avgfpdir, avgfpdir,
|
||||
"expected number of files per directory", "");
|
||||
|
||||
if (Nflag) {
|
||||
fprintf(stdout, "tunefs: current settings of %s\n", special);
|
||||
@ -290,6 +306,11 @@ main(int argc, char *argv[])
|
||||
#endif
|
||||
fprintf(stdout, "\toptimization preference: %s\n",
|
||||
chg[sblock.fs_optim]);
|
||||
fprintf(stdout, "\taverage file size: %d\n",
|
||||
sblock.fs_avgfilesize);
|
||||
fprintf(stdout,
|
||||
"\texpected number of files per directory: %d\n",
|
||||
sblock.fs_avgfpdir);
|
||||
fprintf(stdout, "\ttrack skew %d sectors\n",
|
||||
sblock.fs_trackskew);
|
||||
fprintf(stdout, "tunefs: no changes made\n");
|
||||
@ -336,6 +357,8 @@ usage(void)
|
||||
fprintf(stderr, "\t-a maximum contiguous blocks\n");
|
||||
fprintf(stderr, "\t-d rotational delay between contiguous blocks\n");
|
||||
fprintf(stderr, "\t-e maximum blocks per file in a cylinder group\n");
|
||||
fprintf(stderr, "\t-g average file size\n");
|
||||
fprintf(stderr, "\t-h expected number of files per directory\n");
|
||||
fprintf(stderr, "\t-k track skew in sectors\n");
|
||||
fprintf(stderr, "\t-m minimum percentage of free space\n");
|
||||
#ifdef TUNEFS_SOFTDEP
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: ffs_alloc.c,v 1.49 2001/08/31 03:38:45 lukem Exp $ */
|
||||
/* $NetBSD: ffs_alloc.c,v 1.50 2001/09/06 02:16:01 lukem Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright (c) 1982, 1986, 1989, 1993
|
||||
@ -62,7 +62,7 @@ static ufs_daddr_t ffs_alloccg __P((struct inode *, int, ufs_daddr_t, int));
|
||||
static ufs_daddr_t ffs_alloccgblk __P((struct inode *, struct buf *,
|
||||
ufs_daddr_t));
|
||||
static ufs_daddr_t ffs_clusteralloc __P((struct inode *, int, ufs_daddr_t, int));
|
||||
static ino_t ffs_dirpref __P((struct fs *, ino_t));
|
||||
static ino_t ffs_dirpref __P((struct inode *));
|
||||
static ufs_daddr_t ffs_fragextend __P((struct inode *, int, long, int, int));
|
||||
static void ffs_fserr __P((struct fs *, u_int, char *));
|
||||
static u_long ffs_hashalloc
|
||||
@ -642,12 +642,24 @@ ffs_valloc(v)
|
||||
if (fs->fs_cstotal.cs_nifree == 0)
|
||||
goto noinodes;
|
||||
|
||||
ipref = pip->i_number;
|
||||
if ((mode & IFMT) == IFDIR)
|
||||
ipref = ffs_dirpref(fs, ipref);
|
||||
ipref = ffs_dirpref(pip);
|
||||
else
|
||||
ipref = pip->i_number;
|
||||
if (ipref >= fs->fs_ncg * fs->fs_ipg)
|
||||
ipref = 0;
|
||||
cg = ino_to_cg(fs, ipref);
|
||||
/*
|
||||
* Track number of dirs created one after another
|
||||
* in a same cg without intervening by files.
|
||||
*/
|
||||
if ((mode & IFMT) == IFDIR) {
|
||||
if (fs->fs_contigdirs[cg] < 65535)
|
||||
fs->fs_contigdirs[cg]++;
|
||||
} else {
|
||||
if (fs->fs_contigdirs[cg] > 0)
|
||||
fs->fs_contigdirs[cg]--;
|
||||
}
|
||||
ino = (ino_t)ffs_hashalloc(pip, cg, (long)ipref, mode, ffs_nodealloccg);
|
||||
if (ino == 0)
|
||||
goto noinodes;
|
||||
@ -680,68 +692,112 @@ noinodes:
|
||||
}
|
||||
|
||||
/*
|
||||
* Find a cylinder in which to place a directory.
|
||||
* Find a cylinder group in which to place a directory.
|
||||
*
|
||||
* The policy implemented by this algorithm is to select from among
|
||||
* those cylinder groups with above the average number of free inodes
|
||||
* and a "reasonable" number of free blocks, the one with the smallest
|
||||
* number of directories. If there are no cylinder groups with a
|
||||
* reasonable number of free blocks, we select a CG with *any* free
|
||||
* blocks or free frags.
|
||||
* The policy implemented by this algorithm is to allocate a
|
||||
* directory inode in the same cylinder group as its parent
|
||||
* directory, but also to reserve space for its files inodes
|
||||
* and data. Restrict the number of directories which may be
|
||||
* allocated one after another in the same cylinder group
|
||||
* without intervening allocation of files.
|
||||
*
|
||||
* "Reasonable" here is arbitrarily defined as "at least 25% of the
|
||||
* average amount of free space."
|
||||
*
|
||||
* This complex policy is intended to avoid pathological (linear
|
||||
* search) allocation performance when a filesystem contains many
|
||||
* small cylinder groups with few directory inodes and no free blocks;
|
||||
* this was observed in practice with the old allocation policy (which
|
||||
* ignored the distribution of free blocks). Under the old policy,
|
||||
* when a new filesystem is populated with a number of files somewhat
|
||||
* larger than the CG size, and then a second tree containing a large
|
||||
* number of files and directories is created, mkdir() performance
|
||||
* would degrade catastrophically, taking many seconds and involving
|
||||
* thousands of disk reads to complete.
|
||||
*
|
||||
* XXX TODO: we currently ignore our "ipref" argument; we may want to
|
||||
* add a heuristic to determine whether to place a directory in the
|
||||
* same CG as its parent to reduce the amount of seeking required in
|
||||
* the course of tree-walks.
|
||||
* If we allocate a first level directory then force allocation
|
||||
* in another cylinder group.
|
||||
*/
|
||||
static ino_t
|
||||
ffs_dirpref(fs, ipref)
|
||||
struct fs *fs;
|
||||
ino_t ipref;
|
||||
ffs_dirpref(pip)
|
||||
struct inode *pip;
|
||||
{
|
||||
int cg, minndir, mincg, avgifree, bfreethresh;
|
||||
int minndirf, mincgf;
|
||||
struct csum *cs;
|
||||
register struct fs *fs;
|
||||
int cg, prefcg, dirsize, cgsize;
|
||||
int avgifree, avgbfree, avgndir, curdirsize;
|
||||
int minifree, minbfree, maxndir;
|
||||
int mincg, minndir;
|
||||
int maxcontigdirs;
|
||||
|
||||
fs = pip->i_fs;
|
||||
|
||||
avgifree = fs->fs_cstotal.cs_nifree / fs->fs_ncg;
|
||||
bfreethresh = fs->fs_cstotal.cs_nbfree / fs->fs_ncg;
|
||||
bfreethresh >>= 2;
|
||||
minndir = fs->fs_ipg;
|
||||
minndirf = fs->fs_ipg;
|
||||
mincg = 0;
|
||||
mincgf = 0;
|
||||
for (cg = 0; cg < fs->fs_ncg; cg++) {
|
||||
cs = &fs->fs_cs(fs, cg);
|
||||
if (cs->cs_nifree >= avgifree) {
|
||||
if ((cs->cs_ndir < minndir) &&
|
||||
(cs->cs_nbfree > bfreethresh)) {
|
||||
avgbfree = fs->fs_cstotal.cs_nbfree / fs->fs_ncg;
|
||||
avgndir = fs->fs_cstotal.cs_ndir / fs->fs_ncg;
|
||||
|
||||
/*
|
||||
* Force allocation in another cg if creating a first level dir.
|
||||
*/
|
||||
if (ITOV(pip)->v_flag & VROOT) {
|
||||
prefcg = random() % fs->fs_ncg;
|
||||
mincg = prefcg;
|
||||
minndir = fs->fs_ipg;
|
||||
for (cg = prefcg; cg < fs->fs_ncg; cg++)
|
||||
if (fs->fs_cs(fs, cg).cs_ndir < minndir &&
|
||||
fs->fs_cs(fs, cg).cs_nifree >= avgifree &&
|
||||
fs->fs_cs(fs, cg).cs_nbfree >= avgbfree) {
|
||||
mincg = cg;
|
||||
minndir = cs->cs_ndir;
|
||||
minndir = fs->fs_cs(fs, cg).cs_ndir;
|
||||
}
|
||||
if ((cs->cs_ndir < minndirf) &&
|
||||
((cs->cs_nffree + cs->cs_nbfree) > 0)) {
|
||||
mincgf = cg;
|
||||
minndirf = cs->cs_ndir;
|
||||
for (cg = 0; cg < prefcg; cg++)
|
||||
if (fs->fs_cs(fs, cg).cs_ndir < minndir &&
|
||||
fs->fs_cs(fs, cg).cs_nifree >= avgifree &&
|
||||
fs->fs_cs(fs, cg).cs_nbfree >= avgbfree) {
|
||||
mincg = cg;
|
||||
minndir = fs->fs_cs(fs, cg).cs_ndir;
|
||||
}
|
||||
}
|
||||
return ((ino_t)(fs->fs_ipg * mincg));
|
||||
}
|
||||
if (minndir == fs->fs_ipg)
|
||||
mincg = mincgf;
|
||||
return ((ino_t)(fs->fs_ipg * mincg));
|
||||
|
||||
/*
|
||||
* Count various limits which used for
|
||||
* optimal allocation of a directory inode.
|
||||
*/
|
||||
maxndir = min(avgndir + fs->fs_ipg / 16, fs->fs_ipg);
|
||||
minifree = avgifree - fs->fs_ipg / 4;
|
||||
if (minifree < 0)
|
||||
minifree = 0;
|
||||
minbfree = avgbfree - fs->fs_fpg / fs->fs_frag / 4;
|
||||
if (minbfree < 0)
|
||||
minbfree = 0;
|
||||
cgsize = fs->fs_fsize * fs->fs_fpg;
|
||||
dirsize = fs->fs_avgfilesize * fs->fs_avgfpdir;
|
||||
curdirsize = avgndir ? (cgsize - avgbfree * fs->fs_bsize) / avgndir : 0;
|
||||
if (dirsize < curdirsize)
|
||||
dirsize = curdirsize;
|
||||
maxcontigdirs = min(cgsize / dirsize, 255);
|
||||
if (fs->fs_avgfpdir > 0)
|
||||
maxcontigdirs = min(maxcontigdirs,
|
||||
fs->fs_ipg / fs->fs_avgfpdir);
|
||||
if (maxcontigdirs == 0)
|
||||
maxcontigdirs = 1;
|
||||
|
||||
/*
|
||||
* Limit number of dirs in one cg and reserve space for
|
||||
* regular files, but only if we have no deficit in
|
||||
* inodes or space.
|
||||
*/
|
||||
prefcg = ino_to_cg(fs, pip->i_number);
|
||||
for (cg = prefcg; cg < fs->fs_ncg; cg++)
|
||||
if (fs->fs_cs(fs, cg).cs_ndir < maxndir &&
|
||||
fs->fs_cs(fs, cg).cs_nifree >= minifree &&
|
||||
fs->fs_cs(fs, cg).cs_nbfree >= minbfree) {
|
||||
if (fs->fs_contigdirs[cg] < maxcontigdirs)
|
||||
return ((ino_t)(fs->fs_ipg * cg));
|
||||
}
|
||||
for (cg = 0; cg < prefcg; cg++)
|
||||
if (fs->fs_cs(fs, cg).cs_ndir < maxndir &&
|
||||
fs->fs_cs(fs, cg).cs_nifree >= minifree &&
|
||||
fs->fs_cs(fs, cg).cs_nbfree >= minbfree) {
|
||||
if (fs->fs_contigdirs[cg] < maxcontigdirs)
|
||||
return ((ino_t)(fs->fs_ipg * cg));
|
||||
}
|
||||
/*
|
||||
* This is a backstop when we are deficient in space.
|
||||
*/
|
||||
for (cg = prefcg; cg < fs->fs_ncg; cg++)
|
||||
if (fs->fs_cs(fs, cg).cs_nifree >= avgifree)
|
||||
return ((ino_t)(fs->fs_ipg * cg));
|
||||
for (cg = 0; cg < prefcg; cg++)
|
||||
if (fs->fs_cs(fs, cg).cs_nifree >= avgifree)
|
||||
break;
|
||||
return ((ino_t)(fs->fs_ipg * cg));
|
||||
}
|
||||
|
||||
/*
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: ffs_bswap.c,v 1.12 2001/09/03 14:52:17 lukem Exp $ */
|
||||
/* $NetBSD: ffs_bswap.c,v 1.13 2001/09/06 02:16:02 lukem Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright (c) 1998 Manuel Bouyer.
|
||||
@ -77,6 +77,8 @@ ffs_sb_swap(struct fs *o, struct fs *n)
|
||||
n32[i] = bswap32(o32[i]);
|
||||
|
||||
n->fs_cpc = bswap32(o->fs_cpc);
|
||||
n->fs_avgfilesize = bswap32(o->fs_avgfilesize);
|
||||
n->fs_avgfpdir = bswap32(o->fs_avgfpdir);
|
||||
n->fs_contigsumsize = bswap32(o->fs_contigsumsize);
|
||||
n->fs_maxsymlinklen = bswap32(o->fs_maxsymlinklen);
|
||||
n->fs_inodefmt = bswap32(o->fs_inodefmt);
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: ffs_vfsops.c,v 1.84 2001/09/02 01:58:30 lukem Exp $ */
|
||||
/* $NetBSD: ffs_vfsops.c,v 1.85 2001/09/06 02:16:02 lukem Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright (c) 1989, 1991, 1993, 1994
|
||||
@ -447,6 +447,7 @@ ffs_reload(mountp, cred, p)
|
||||
*/
|
||||
newfs->fs_csp = fs->fs_csp;
|
||||
newfs->fs_maxcluster = fs->fs_maxcluster;
|
||||
newfs->fs_contigdirs = fs->fs_contigdirs;
|
||||
newfs->fs_ronly = fs->fs_ronly;
|
||||
memcpy(fs, newfs, (u_int)fs->fs_sbsize);
|
||||
if (fs->fs_sbsize < SBSIZE)
|
||||
@ -455,6 +456,12 @@ ffs_reload(mountp, cred, p)
|
||||
free(newfs, M_UFSMNT);
|
||||
mountp->mnt_maxsymlinklen = fs->fs_maxsymlinklen;
|
||||
ffs_oldfscompat(fs);
|
||||
/* An old fsck may have zeroed these fields, so recheck them. */
|
||||
if (fs->fs_avgfilesize <= 0)
|
||||
fs->fs_avgfilesize = AVFILESIZ;
|
||||
if (fs->fs_avgfpdir <= 0)
|
||||
fs->fs_avgfpdir = AFPDIR;
|
||||
|
||||
ffs_statfs(mountp, &mountp->mnt_stat, p);
|
||||
/*
|
||||
* Step 3: re-read summary information from disk.
|
||||
@ -663,6 +670,7 @@ ffs_mountfs(devvp, mp, p)
|
||||
blks = howmany(size, fs->fs_fsize);
|
||||
if (fs->fs_contigsumsize > 0)
|
||||
size += fs->fs_ncg * sizeof(int32_t);
|
||||
size += fs->fs_ncg * sizeof(*fs->fs_contigdirs);
|
||||
space = malloc((u_long)size, M_UFSMNT, M_WAITOK);
|
||||
fs->fs_csp = space;
|
||||
for (i = 0; i < blks; i += fs->fs_frag) {
|
||||
@ -688,10 +696,20 @@ ffs_mountfs(devvp, mp, p)
|
||||
bp = NULL;
|
||||
}
|
||||
if (fs->fs_contigsumsize > 0) {
|
||||
fs->fs_maxcluster = lp = (int32_t *)space;
|
||||
fs->fs_maxcluster = lp = space;
|
||||
for (i = 0; i < fs->fs_ncg; i++)
|
||||
*lp++ = fs->fs_contigsumsize;
|
||||
space = lp;
|
||||
}
|
||||
size = fs->fs_ncg * sizeof(*fs->fs_contigdirs);
|
||||
fs->fs_contigdirs = space;
|
||||
space = (char *)space + size;
|
||||
memset(fs->fs_contigdirs, 0, size);
|
||||
/* Compatibility for old filesystems - XXX */
|
||||
if (fs->fs_avgfilesize <= 0)
|
||||
fs->fs_avgfilesize = AVFILESIZ;
|
||||
if (fs->fs_avgfpdir <= 0)
|
||||
fs->fs_avgfpdir = AFPDIR;
|
||||
mp->mnt_data = (qaddr_t)ump;
|
||||
mp->mnt_stat.f_fsid.val[0] = (long)dev;
|
||||
mp->mnt_stat.f_fsid.val[1] = makefstype(MOUNT_FFS);
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: fs.h,v 1.19 2001/09/03 14:52:17 lukem Exp $ */
|
||||
/* $NetBSD: fs.h,v 1.20 2001/09/06 02:16:02 lukem Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright (c) 1982, 1986, 1993
|
||||
@ -111,15 +111,16 @@
|
||||
/*
|
||||
* There is a 128-byte region in the superblock reserved for in-core
|
||||
* pointers to summary information. Originally this included an array
|
||||
* of pointers to blocks of struct csum; now there are just two
|
||||
* of pointers to blocks of struct csum; now there are just three
|
||||
* pointers and the remaining space is padded with fs_ocsp[].
|
||||
*
|
||||
* NOCSPTRS determines the size of this padding. One pointer (fs_csp)
|
||||
* is taken away to point to a contiguous array of struct csum for
|
||||
* all cylinder groups; a second (fs_maxcluster) points to an array
|
||||
* of cluster sizes that is computed as cylinder groups are inspected.
|
||||
* of cluster sizes that is computed as cylinder groups are inspected;
|
||||
* and the third points to an array that tracks the creation of new
|
||||
* directories.
|
||||
*/
|
||||
#define NOCSPTRS ((128 / sizeof(void *)) - 2)
|
||||
#define NOCSPTRS ((128 / sizeof(void *)) - 3)
|
||||
|
||||
/*
|
||||
* A summary of contiguous blocks of various sizes is maintained
|
||||
@ -144,6 +145,18 @@
|
||||
#define MINFREE 5
|
||||
#define DEFAULTOPT FS_OPTTIME
|
||||
|
||||
/*
|
||||
* Grigoriy Orlov <gluk@ptci.ru> has done some extensive work to fine
|
||||
* tune the layout preferences for directories within a filesystem.
|
||||
* His algorithm can be tuned by adjusting the following parameters
|
||||
* which tell the system the average file size and the average number
|
||||
* of files per directory. These defaults are well selected for typical
|
||||
* filesystems, but may need to be tuned for odd cases like filesystems
|
||||
* being used for sqiud caches or news spools.
|
||||
*/
|
||||
#define AVFILESIZ 16384 /* expected average file size */
|
||||
#define AFPDIR 64 /* expected number of files per directory */
|
||||
|
||||
/*
|
||||
* Per cylinder group information; summarized in blocks allocated
|
||||
* from first cylinder group data blocks. These blocks have to be
|
||||
@ -230,11 +243,15 @@ struct fs {
|
||||
/* these fields retain the current block allocation info */
|
||||
int32_t fs_cgrotor; /* last cg searched */
|
||||
void *fs_ocsp[NOCSPTRS]; /* padding; was list of fs_cs buffers */
|
||||
u_int16_t *fs_contigdirs; /* # of contiguously allocated dirs */
|
||||
struct csum *fs_csp; /* cg summary info buffer for fs_cs */
|
||||
int32_t *fs_maxcluster; /* max cluster in each cyl group */
|
||||
int32_t *fs_maxcluster; /* max cluster in each cyl group */
|
||||
int32_t fs_cpc; /* cyl per cycle in postbl */
|
||||
int16_t fs_opostbl[16][8]; /* old rotation block list head */
|
||||
int32_t fs_sparecon[50]; /* RESERVED for future constants */
|
||||
int32_t fs_snapinum[20]; /* RESERVED for snapshot inode nums */
|
||||
int32_t fs_avgfilesize; /* expected average file size */
|
||||
int32_t fs_avgfpdir; /* expected # of files per directory */
|
||||
int32_t fs_sparecon[28]; /* RESERVED for future constants */
|
||||
int32_t fs_contigsumsize; /* size of cluster summary array */
|
||||
int32_t fs_maxsymlinklen; /* max length of an internal symlink */
|
||||
int32_t fs_inodefmt; /* format of on-disk inodes */
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: dumpfs.c,v 1.29 2001/09/02 01:58:32 lukem Exp $ */
|
||||
/* $NetBSD: dumpfs.c,v 1.30 2001/09/06 02:16:02 lukem Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright (c) 1983, 1992, 1993
|
||||
@ -43,7 +43,7 @@ __COPYRIGHT("@(#) Copyright (c) 1983, 1992, 1993\n\
|
||||
#if 0
|
||||
static char sccsid[] = "@(#)dumpfs.c 8.5 (Berkeley) 4/29/95";
|
||||
#else
|
||||
__RCSID("$NetBSD: dumpfs.c,v 1.29 2001/09/02 01:58:32 lukem Exp $");
|
||||
__RCSID("$NetBSD: dumpfs.c,v 1.30 2001/09/06 02:16:02 lukem Exp $");
|
||||
#endif
|
||||
#endif /* not lint */
|
||||
|
||||
@ -198,6 +198,8 @@ dumpfs(const char *name)
|
||||
(unsigned long long)afs.fs_maxfilesize);
|
||||
printf("nindir\t%d\tinopb\t%d\tnspf\t%d\n",
|
||||
afs.fs_nindir, afs.fs_inopb, afs.fs_nspf);
|
||||
printf("avgfilesize %d\tavgfpdir %d\n",
|
||||
afs.fs_avgfilesize, afs.fs_avgfpdir);
|
||||
printf("sblkno\t%d\tcblkno\t%d\tiblkno\t%d\tdblkno\t%d\n",
|
||||
afs.fs_sblkno, afs.fs_cblkno, afs.fs_iblkno, afs.fs_dblkno);
|
||||
printf("sbsize\t%d\tcgsize\t%d\toffset\t%d\tmask\t0x%08x\n",
|
||||
|
||||
Loading…
Reference in New Issue
Block a user