freebsd-dev/sys/ufs/ffs/fs.h

766 lines
30 KiB
C
Raw Normal View History

/*-
1994-05-24 10:09:53 +00:00
* Copyright (c) 1982, 1986, 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.
* 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.
*
* @(#)fs.h 8.13 (Berkeley) 3/21/95
1999-08-28 01:08:13 +00:00
* $FreeBSD$
1994-05-24 10:09:53 +00:00
*/
#ifndef _UFS_FFS_FS_H_
#define _UFS_FFS_FS_H_
1994-05-24 10:09:53 +00:00
/*
2002-05-16 21:28:32 +00:00
* Each disk drive contains some number of filesystems.
* A filesystem consists of a number of cylinder groups.
1994-05-24 10:09:53 +00:00
* Each cylinder group has inodes and data.
*
2002-05-16 21:28:32 +00:00
* A filesystem is described by its super-block, which in turn
1994-05-24 10:09:53 +00:00
* describes the cylinder groups. The super-block is critical
* data and is replicated in each cylinder group to protect against
* catastrophic loss. This is done at `newfs' time and the critical
* super-block data does not change, so the copies need not be
* referenced further unless disaster strikes.
*
2002-05-16 21:28:32 +00:00
* For filesystem fs, the offsets of the various blocks of interest
1994-05-24 10:09:53 +00:00
* are given in the super block as:
* [fs->fs_sblkno] Super-block
* [fs->fs_cblkno] Cylinder group block
* [fs->fs_iblkno] Inode blocks
* [fs->fs_dblkno] Data blocks
* The beginning of cylinder group cg in fs, is given by
* the ``cgbase(fs, cg)'' macro.
*
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
* Depending on the architecture and the media, the superblock may
* reside in any one of four places. For tiny media where every block
* counts, it is placed at the very front of the partition. Historically,
* UFS1 placed it 8K from the front to leave room for the disk label and
* a small bootstrap. For UFS2 it got moved to 64K from the front to leave
* room for the disk label and a bigger bootstrap, and for really piggy
* systems we check at 256K from the front if the first three fail. In
* all cases the size of the superblock will be SBLOCKSIZE. All values are
* given in byte-offset form, so they do not imply a sector size. The
* SBLOCKSEARCH specifies the order in which the locations should be searched.
*/
#define SBLOCK_FLOPPY 0
#define SBLOCK_UFS1 8192
#define SBLOCK_UFS2 65536
#define SBLOCK_PIGGY 262144
#define SBLOCKSIZE 8192
#define SBLOCKSEARCH \
{ SBLOCK_UFS2, SBLOCK_UFS1, SBLOCK_FLOPPY, SBLOCK_PIGGY, -1 }
/*
* Max number of fragments per block. This value is NOT tweakable.
1994-05-24 10:09:53 +00:00
*/
#define MAXFRAG 8
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
1994-05-24 10:09:53 +00:00
/*
* Addresses stored in inodes are capable of addressing fragments
1995-05-30 08:16:23 +00:00
* of `blocks'. File system blocks of at most size MAXBSIZE can
1994-05-24 10:09:53 +00:00
* be optionally broken into 2, 4, or 8 pieces, each of which is
* addressable; these pieces may be DEV_BSIZE, or some multiple of
1994-05-24 10:09:53 +00:00
* a DEV_BSIZE unit.
*
* Large files consist of exclusively large data blocks. To avoid
* undue wasted disk space, the last data block of a small file may be
* allocated as only as many fragments of a large block as are
2002-05-16 21:28:32 +00:00
* necessary. The filesystem format retains only a single pointer
1994-05-24 10:09:53 +00:00
* to such a fragment, which is a piece of a single large block that
* has been divided. The size of such a fragment is determinable from
* information in the inode, using the ``blksize(fs, ip, lbn)'' macro.
*
2002-05-16 21:28:32 +00:00
* The filesystem records space availability at the fragment level;
1994-05-24 10:09:53 +00:00
* to determine block availability, aligned fragments are examined.
*/
/*
* MINBSIZE is the smallest allowable block size.
* In order to insure that it is possible to create files of size
* 2^32 with only two levels of indirection, MINBSIZE is set to 4096.
* MINBSIZE must be big enough to hold a cylinder group block,
* thus changes to (struct cg) must keep its size within MINBSIZE.
* Note that super blocks are always of size SBLOCKSIZE,
* and that both SBLOCKSIZE and MAXBSIZE must be >= MINBSIZE.
1994-05-24 10:09:53 +00:00
*/
#define MINBSIZE 4096
/*
2002-05-16 21:28:32 +00:00
* The path name on which the filesystem is mounted is maintained
1995-05-30 08:16:23 +00:00
* in fs_fsmnt. MAXMNTLEN defines the amount of space allocated in
1994-05-24 10:09:53 +00:00
* the super block for this name.
*/
#define MAXMNTLEN 468
/*
* The volume name for this filesystem is maintained in fs_volname.
* MAXVOLLEN defines the length of the buffer allocated.
*/
#define MAXVOLLEN 32
/*
* 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 a few
* 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
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>
2001-04-10 08:38:59 +00:00
* 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. A fourth pointer, fs_active, is used when creating
* snapshots; it points to a bitmap of cylinder groups for which the
* free-block bitmap has changed since the snapshot operation began.
1994-05-24 10:09:53 +00:00
*/
#define NOCSPTRS ((128 / sizeof(void *)) - 4)
1994-05-24 10:09:53 +00:00
/*
* A summary of contiguous blocks of various sizes is maintained
* in each cylinder group. Normally this is set by the initial
* value of fs_maxcontig. To conserve space, a maximum summary size
* is set by FS_MAXCONTIG.
*/
#define FS_MAXCONTIG 16
/*
2002-05-16 21:28:32 +00:00
* MINFREE gives the minimum acceptable percentage of filesystem
1994-05-24 10:09:53 +00:00
* blocks which may be free. If the freelist drops below this level
* only the superuser may continue to allocate blocks. This may
* be set to 0 if no reserve of free blocks is deemed necessary,
2002-05-16 21:28:32 +00:00
* however throughput drops by fifty percent if the filesystem
1994-05-24 10:09:53 +00:00
* is run at between 95% and 100% full; thus the minimum default
* value of fs_minfree is 5%. However, to get good clustering
* performance, 10% is a better choice. hence we use 10% as our
* default value. With 10% free space, fragmentation is not a
* problem, so we choose to optimize for time.
*/
1995-03-10 22:18:16 +00:00
#define MINFREE 8
1994-05-24 10:09:53 +00:00
#define DEFAULTOPT FS_OPTTIME
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>
2001-04-10 08:38:59 +00:00
/*
* 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
2004-05-31 16:55:12 +00:00
* being used for squid caches or news spools.
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>
2001-04-10 08:38:59 +00:00
*/
#define AVFILESIZ 16384 /* expected average file size */
#define AFPDIR 64 /* expected number of files per directory */
/*
* The maximum number of snapshot nodes that can be associated
* with each filesystem. This limit affects only the number of
* snapshot files that can be recorded within the superblock so
* that they can be found when the filesystem is mounted. However,
* maintaining too many will slow the filesystem performance, so
* having this limit is a good idea.
*/
#define FSMAXSNAP 20
/*
* Used to identify special blocks in snapshots:
*
* BLK_NOCOPY - A block that was unallocated at the time the snapshot
* was taken, hence does not need to be copied when written.
* BLK_SNAP - A block held by another snapshot that is not needed by this
* snapshot. When the other snapshot is freed, the BLK_SNAP entries
* are converted to BLK_NOCOPY. These are needed to allow fsck to
* identify blocks that are in use by other snapshots (which are
* expunged from this snapshot).
*/
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
#define BLK_NOCOPY ((ufs2_daddr_t)(1))
#define BLK_SNAP ((ufs2_daddr_t)(2))
/*
* Sysctl values for the fast filesystem.
*/
#define FFS_ADJ_REFCNT 1 /* adjust inode reference count */
#define FFS_ADJ_BLKCNT 2 /* adjust inode used block count */
#define FFS_BLK_FREE 3 /* free range of blocks in map */
#define FFS_DIR_FREE 4 /* free specified dir inodes in map */
#define FFS_FILE_FREE 5 /* free specified file inodes in map */
#define FFS_SET_FLAGS 6 /* set filesystem flags */
#define FFS_ADJ_NDIR 7 /* adjust number of directories */
#define FFS_ADJ_NBFREE 8 /* adjust number of free blocks */
#define FFS_ADJ_NIFREE 9 /* adjust number of free inodes */
#define FFS_ADJ_NFFREE 10 /* adjust number of free frags */
#define FFS_ADJ_NUMCLUSTERS 11 /* adjust number of free clusters */
Background: When renaming a directory it passes through several intermediate states. First its new name will be created causing it to have two names (from possibly different parents). Next, if it has different parents, its value of ".." will be changed from pointing to the old parent to pointing to the new parent. Concurrently, its old name will be removed bringing it back into a consistent state. When fsck encounters an extra name for a directory, it offers to remove the "extraneous hard link"; when it finds that the names have been changed but the update to ".." has not happened, it offers to rewrite ".." to point at the correct parent. Both of these changes were considered unexpected so would cause fsck in preen mode or fsck in background mode to fail with the need to run fsck manually to fix these problems. Fsck running in preen mode or background mode now corrects these expected inconsistencies that arise during directory rename. The functionality added with this update is used by fsck running in background mode to make these fixes. Solution: This update adds three new fsck sysctl commands to support background fsck in correcting expected inconsistencies that arise from incomplete directory rename operations. They are: setcwd(dirinode) - set the current directory to dirinode in the filesystem associated with the snapshot. setdotdot(oldvalue, newvalue) - Verify that the inode number for ".." in the current directory is oldvalue then change it to newvalue. unlink(nameptr, oldvalue) - Verify that the inode number associated with nameptr in the current directory is oldvalue then unlink it. As with all other fsck sysctls, these new ones may only be used by processes with appropriate priviledge. Reported by: jeff Security issues: rwatson
2010-01-11 20:44:05 +00:00
#define FFS_SET_CWD 12 /* set current directory */
#define FFS_SET_DOTDOT 13 /* set inode number for ".." */
#define FFS_UNLINK 14 /* remove a name in the filesystem */
#define FFS_SET_INODE 15 /* update an on-disk inode */
#define FFS_SET_BUFOUTPUT 16 /* set buffered writing on descriptor */
#define FFS_MAXID 16 /* number of valid ffs ids */
/*
* Command structure passed in to the filesystem to adjust filesystem values.
*/
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
#define FFS_CMD_VERSION 0x19790518 /* version ID */
struct fsck_cmd {
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t version; /* version of command structure */
int32_t handle; /* reference to filesystem to be changed */
int64_t value; /* inode or block number to be affected */
int64_t size; /* amount or range to be adjusted */
int64_t spare; /* reserved for future use */
};
1994-05-24 10:09:53 +00:00
/*
* Per cylinder group information; summarized in blocks allocated
* from first cylinder group data blocks. These blocks have to be
* read in from fs_csaddr (size fs_cssize) in addition to the
* super block.
*/
struct csum {
int32_t cs_ndir; /* number of directories */
int32_t cs_nbfree; /* number of free blocks */
int32_t cs_nifree; /* number of free inodes */
int32_t cs_nffree; /* number of free frags */
1994-05-24 10:09:53 +00:00
};
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
struct csum_total {
int64_t cs_ndir; /* number of directories */
int64_t cs_nbfree; /* number of free blocks */
int64_t cs_nifree; /* number of free inodes */
int64_t cs_nffree; /* number of free frags */
int64_t cs_numclusters; /* number of free clusters */
int64_t cs_spare[3]; /* future expansion */
};
1994-05-24 10:09:53 +00:00
/*
2002-05-16 21:28:32 +00:00
* Super block for an FFS filesystem.
1994-05-24 10:09:53 +00:00
*/
struct fs {
2002-05-16 21:28:32 +00:00
int32_t fs_firstfield; /* historic filesystem linked list, */
int32_t fs_unused_1; /* used for incore super blocks */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t fs_sblkno; /* offset of super-block in filesys */
int32_t fs_cblkno; /* offset of cyl-block in filesys */
int32_t fs_iblkno; /* offset of inode-blocks in filesys */
int32_t fs_dblkno; /* offset of first data after cg */
int32_t fs_old_cgoffset; /* cylinder group offset in cylinder */
int32_t fs_old_cgmask; /* used to calc mod fs_ntrak */
int32_t fs_old_time; /* last time written */
int32_t fs_old_size; /* number of blocks in fs */
int32_t fs_old_dsize; /* number of data blocks in fs */
u_int32_t fs_ncg; /* number of cylinder groups */
int32_t fs_bsize; /* size of basic blocks in fs */
int32_t fs_fsize; /* size of frag blocks in fs */
int32_t fs_frag; /* number of frags in a block in fs */
1994-05-24 10:09:53 +00:00
/* these are configuration parameters */
int32_t fs_minfree; /* minimum percentage of free blocks */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t fs_old_rotdelay; /* num of ms for optimal next block */
int32_t fs_old_rps; /* disk revolutions per second */
1994-05-24 10:09:53 +00:00
/* these fields can be computed from the others */
int32_t fs_bmask; /* ``blkoff'' calc of blk offsets */
int32_t fs_fmask; /* ``fragoff'' calc of frag offsets */
int32_t fs_bshift; /* ``lblkno'' calc of logical blkno */
int32_t fs_fshift; /* ``numfrags'' calc number of frags */
1994-05-24 10:09:53 +00:00
/* these are configuration parameters */
int32_t fs_maxcontig; /* max number of contiguous blks */
int32_t fs_maxbpg; /* max number of blks per cyl group */
1994-05-24 10:09:53 +00:00
/* these fields can be computed from the others */
int32_t fs_fragshift; /* block to frag shift */
int32_t fs_fsbtodb; /* fsbtodb and dbtofsb shift constant */
int32_t fs_sbsize; /* actual size of super block */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t fs_spare1[2]; /* old fs_csmask */
/* old fs_csshift */
int32_t fs_nindir; /* value of NINDIR */
u_int32_t fs_inopb; /* value of INOPB */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t fs_old_nspf; /* value of NSPF */
1994-05-24 10:09:53 +00:00
/* yet another configuration parameter */
int32_t fs_optim; /* optimization preference, see below */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t fs_old_npsect; /* # sectors/track including spares */
int32_t fs_old_interleave; /* hardware sector interleave */
int32_t fs_old_trackskew; /* sector 0 skew, per track */
int32_t fs_id[2]; /* unique filesystem id */
1994-05-24 10:09:53 +00:00
/* sizes determined by number of cylinder groups and their sizes */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t fs_old_csaddr; /* blk addr of cyl grp summary area */
int32_t fs_cssize; /* size of cyl grp summary area */
int32_t fs_cgsize; /* cylinder group size */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t fs_spare2; /* old fs_ntrak */
int32_t fs_old_nsect; /* sectors per track */
int32_t fs_old_spc; /* sectors per cylinder */
int32_t fs_old_ncyl; /* cylinders in filesystem */
int32_t fs_old_cpg; /* cylinders per group */
u_int32_t fs_ipg; /* inodes per group */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t fs_fpg; /* blocks per group * fs_frag */
1994-05-24 10:09:53 +00:00
/* this data must be re-computed after crashes */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
struct csum fs_old_cstotal; /* cylinder summary information */
1994-05-24 10:09:53 +00:00
/* these fields are cleared at mount time */
int8_t fs_fmod; /* super block modified flag */
2002-05-16 21:28:32 +00:00
int8_t fs_clean; /* filesystem is clean flag */
int8_t fs_ronly; /* mounted read-only flag */
int8_t fs_old_flags; /* old FS_ flags */
u_char fs_fsmnt[MAXMNTLEN]; /* name mounted on */
u_char fs_volname[MAXVOLLEN]; /* volume name */
u_int64_t fs_swuid; /* system-wide uid */
int32_t fs_pad; /* due to alignment of fs_swuid */
1994-05-24 10:09:53 +00:00
/* 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_int8_t *fs_contigdirs; /* (u) # of contig. allocated dirs */
struct csum *fs_csp; /* (u) cg summary info buffer */
int32_t *fs_maxcluster; /* (u) max cluster in each cyl group */
u_int *fs_active; /* (u) used by snapshots to track fs */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t fs_old_cpc; /* cyl per cycle in postbl */
int32_t fs_maxbsize; /* maximum blocking factor permitted */
int64_t fs_unrefs; /* number of unreferenced inodes */
int64_t fs_sparecon64[16]; /* old rotation block list head */
int64_t fs_sblockloc; /* byte offset of standard superblock */
struct csum_total fs_cstotal; /* (u) cylinder summary information */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
ufs_time_t fs_time; /* last time written */
int64_t fs_size; /* number of blocks in fs */
int64_t fs_dsize; /* number of data blocks in fs */
ufs2_daddr_t fs_csaddr; /* blk addr of cyl grp summary area */
int64_t fs_pendingblocks; /* (u) blocks being freed */
u_int32_t fs_pendinginodes; /* (u) inodes being freed */
uint32_t fs_snapinum[FSMAXSNAP];/* list of snapshot inode numbers */
u_int32_t fs_avgfilesize; /* expected average file size */
u_int32_t fs_avgfpdir; /* expected # of files per directory */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t fs_save_cgsize; /* save real cg size to use fs_bsize */
ufs_time_t fs_mtime; /* Last mount or fsck time. */
int32_t fs_sujfree; /* SUJ free list */
int32_t fs_sparecon32[23]; /* reserved for future constants */
int32_t fs_flags; /* see FS_ flags below */
int32_t fs_contigsumsize; /* size of cluster summary array */
int32_t fs_maxsymlinklen; /* max length of an internal symlink */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t fs_old_inodefmt; /* format of on-disk inodes */
u_int64_t fs_maxfilesize; /* maximum representable file size */
int64_t fs_qbmask; /* ~fs_bmask for use with 64-bit size */
int64_t fs_qfmask; /* ~fs_fmask for use with 64-bit size */
int32_t fs_state; /* validate fs_clean field */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t fs_old_postblformat; /* format of positional layout tables */
int32_t fs_old_nrpos; /* number of rotational positions */
int32_t fs_spare5[2]; /* old fs_postbloff */
/* old fs_rotbloff */
int32_t fs_magic; /* magic number */
1994-05-24 10:09:53 +00:00
};
/* Sanity checking. */
#ifdef CTASSERT
CTASSERT(sizeof(struct fs) == 1376);
#endif
1994-05-24 10:09:53 +00:00
/*
* Filesystem identification
1994-05-24 10:09:53 +00:00
*/
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
#define FS_UFS1_MAGIC 0x011954 /* UFS1 fast filesystem magic number */
#define FS_UFS2_MAGIC 0x19540119 /* UFS2 fast filesystem magic number */
#define FS_BAD_MAGIC 0x19960408 /* UFS incomplete newfs magic number */
1994-05-24 10:09:53 +00:00
#define FS_OKAY 0x7c269d38 /* superblock checksum */
#define FS_42INODEFMT -1 /* 4.2BSD inode format */
#define FS_44INODEFMT 2 /* 4.4BSD inode format */
1994-05-24 10:09:53 +00:00
/*
* Preference for optimization.
*/
#define FS_OPTTIME 0 /* minimize allocation time */
#define FS_OPTSPACE 1 /* minimize disk fragmentation */
/*
* Filesystem flags.
*
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
* The FS_UNCLEAN flag is set by the kernel when the filesystem was
* mounted with fs_clean set to zero. The FS_DOSOFTDEP flag indicates
* that the filesystem should be managed by the soft updates code.
* Note that the FS_NEEDSFSCK flag is set and cleared only by the
* fsck utility. It is set when background fsck finds an unexpected
* inconsistency which requires a traditional foreground fsck to be
* run. Such inconsistencies should only be found after an uncorrectable
* disk error. A foreground fsck will clear the FS_NEEDSFSCK flag when
* it has successfully cleaned up the filesystem. The kernel uses this
* flag to enforce that inconsistent filesystems be mounted read-only.
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
* The FS_INDEXDIRS flag when set indicates that the kernel maintains
* on-disk auxiliary indexes (such as B-trees) for speeding directory
* accesses. Kernels that do not support auxiliary indicies clear the
* flag to indicate that the indicies need to be rebuilt (by fsck) before
* they can be used.
*
* FS_ACLS indicates that POSIX.1e ACLs are administratively enabled
* for the file system, so they should be loaded from extended attributes,
* observed for access control purposes, and be administered by object
* owners. FS_NFS4ACLS indicates that NFSv4 ACLs are administratively
* enabled. This flag is mutually exclusive with FS_ACLS. FS_MULTILABEL
* indicates that the TrustedBSD MAC Framework should attempt to back MAC
* labels into extended attributes on the file system rather than maintain
* a single mount label for all objects.
*/
#define FS_UNCLEAN 0x0001 /* filesystem not clean at mount */
#define FS_DOSOFTDEP 0x0002 /* filesystem using soft dependencies */
#define FS_NEEDSFSCK 0x0004 /* filesystem needs sync fsck before mount */
#define FS_SUJ 0x0008 /* Filesystem using softupdate journal */
#define FS_ACLS 0x0010 /* file system has POSIX.1e ACLs enabled */
#define FS_MULTILABEL 0x0020 /* file system is MAC multi-label */
#define FS_GJOURNAL 0x0040 /* gjournaled file system */
#define FS_FLAGS_UPDATED 0x0080 /* flags have been moved to new location */
#define FS_NFS4ACLS 0x0100 /* file system has NFSv4 ACLs enabled */
#define FS_INDEXDIRS 0x0200 /* kernel supports indexed directories */
#define FS_TRIM 0x0400 /* issue BIO_DELETE for deleted blocks */
1994-05-24 10:09:53 +00:00
/*
* Macros to access bits in the fs_active array.
*/
#define ACTIVECGNUM(fs, cg) ((fs)->fs_active[(cg) / (NBBY * sizeof(int))])
#define ACTIVECGOFF(cg) (1 << ((cg) % (NBBY * sizeof(int))))
#define ACTIVESET(fs, cg) do { \
if ((fs)->fs_active) \
ACTIVECGNUM((fs), (cg)) |= ACTIVECGOFF((cg)); \
} while (0)
#define ACTIVECLEAR(fs, cg) do { \
if ((fs)->fs_active) \
ACTIVECGNUM((fs), (cg)) &= ~ACTIVECGOFF((cg)); \
} while (0)
1994-05-24 10:09:53 +00:00
/*
* The size of a cylinder group is calculated by CGSIZE. The maximum size
* is limited by the fact that cylinder groups are at most one block.
1995-05-30 08:16:23 +00:00
* Its size is derived from the size of the maps maintained in the
1994-05-24 10:09:53 +00:00
* cylinder group and the (struct cg) size.
*/
#define CGSIZE(fs) \
/* base cg */ (sizeof(struct cg) + sizeof(int32_t) + \
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
/* old btotoff */ (fs)->fs_old_cpg * sizeof(int32_t) + \
/* old boff */ (fs)->fs_old_cpg * sizeof(u_int16_t) + \
1994-05-24 10:09:53 +00:00
/* inode map */ howmany((fs)->fs_ipg, NBBY) + \
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
/* block map */ howmany((fs)->fs_fpg, NBBY) +\
1994-05-24 10:09:53 +00:00
/* if present */ ((fs)->fs_contigsumsize <= 0 ? 0 : \
/* cluster sum */ (fs)->fs_contigsumsize * sizeof(int32_t) + \
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
/* cluster map */ howmany(fragstoblks(fs, (fs)->fs_fpg), NBBY)))
/*
* The minimal number of cylinder groups that should be created.
*/
#define MINCYLGRPS 4
1994-05-24 10:09:53 +00:00
/*
* Convert cylinder group to base address of its global summary info.
*/
#define fs_cs(fs, indx) fs_csp[indx]
1994-05-24 10:09:53 +00:00
/*
2002-05-16 21:28:32 +00:00
* Cylinder group block for a filesystem.
1994-05-24 10:09:53 +00:00
*/
#define CG_MAGIC 0x090255
struct cg {
int32_t cg_firstfield; /* historic cyl groups linked list */
int32_t cg_magic; /* magic number */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t cg_old_time; /* time last written */
u_int32_t cg_cgx; /* we are the cgx'th cylinder group */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int16_t cg_old_ncyl; /* number of cyl's this cg */
int16_t cg_old_niblk; /* number of inode blocks this cg */
u_int32_t cg_ndblk; /* number of data blocks this cg */
struct csum cg_cs; /* cylinder summary information */
u_int32_t cg_rotor; /* position of last used block */
u_int32_t cg_frotor; /* position of last used frag */
u_int32_t cg_irotor; /* position of last used inode */
u_int32_t cg_frsum[MAXFRAG]; /* counts of available frags */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
int32_t cg_old_btotoff; /* (int32) block totals per cylinder */
int32_t cg_old_boff; /* (u_int16) free block positions */
u_int32_t cg_iusedoff; /* (u_int8) used inode map */
u_int32_t cg_freeoff; /* (u_int8) free block map */
u_int32_t cg_nextfreeoff; /* (u_int8) next available space */
u_int32_t cg_clustersumoff; /* (u_int32) counts of avail clusters */
u_int32_t cg_clusteroff; /* (u_int8) free cluster map */
u_int32_t cg_nclusterblks; /* number of clusters this cg */
u_int32_t cg_niblk; /* number of inode blocks this cg */
u_int32_t cg_initediblk; /* last initialized inode */
u_int32_t cg_unrefs; /* number of unreferenced inodes */
int32_t cg_sparecon32[2]; /* reserved for future use */
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
ufs_time_t cg_time; /* time last written */
int64_t cg_sparecon64[3]; /* reserved for future use */
u_int8_t cg_space[1]; /* space for cylinder group maps */
1994-05-24 10:09:53 +00:00
/* actually longer */
};
1994-05-24 10:09:53 +00:00
/*
* Macros for access to cylinder group array structures
*/
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
#define cg_chkmagic(cgp) ((cgp)->cg_magic == CG_MAGIC)
1994-05-24 10:09:53 +00:00
#define cg_inosused(cgp) \
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
((u_int8_t *)((u_int8_t *)(cgp) + (cgp)->cg_iusedoff))
1994-05-24 10:09:53 +00:00
#define cg_blksfree(cgp) \
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
((u_int8_t *)((u_int8_t *)(cgp) + (cgp)->cg_freeoff))
1994-05-24 10:09:53 +00:00
#define cg_clustersfree(cgp) \
((u_int8_t *)((u_int8_t *)(cgp) + (cgp)->cg_clusteroff))
1994-05-24 10:09:53 +00:00
#define cg_clustersum(cgp) \
((int32_t *)((uintptr_t)(cgp) + (cgp)->cg_clustersumoff))
1994-05-24 10:09:53 +00:00
/*
2002-05-16 21:28:32 +00:00
* Turn filesystem block numbers into disk block addresses.
* This maps filesystem blocks to device size blocks.
1994-05-24 10:09:53 +00:00
*/
#define fsbtodb(fs, b) ((daddr_t)(b) << (fs)->fs_fsbtodb)
1994-05-24 10:09:53 +00:00
#define dbtofsb(fs, b) ((b) >> (fs)->fs_fsbtodb)
/*
* Cylinder group macros to locate things in cylinder groups.
2002-05-16 21:28:32 +00:00
* They calc filesystem addresses of cylinder group data structures.
1994-05-24 10:09:53 +00:00
*/
#define cgbase(fs, c) (((ufs2_daddr_t)(fs)->fs_fpg) * (c))
1994-05-24 10:09:53 +00:00
#define cgdmin(fs, c) (cgstart(fs, c) + (fs)->fs_dblkno) /* 1st data */
#define cgimin(fs, c) (cgstart(fs, c) + (fs)->fs_iblkno) /* inode blk */
#define cgsblock(fs, c) (cgstart(fs, c) + (fs)->fs_sblkno) /* super blk */
#define cgtod(fs, c) (cgstart(fs, c) + (fs)->fs_cblkno) /* cg block */
#define cgstart(fs, c) \
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
((fs)->fs_magic == FS_UFS2_MAGIC ? cgbase(fs, c) : \
(cgbase(fs, c) + (fs)->fs_old_cgoffset * ((c) & ~((fs)->fs_old_cgmask))))
1994-05-24 10:09:53 +00:00
/*
* Macros for handling inode numbers:
2002-05-16 21:28:32 +00:00
* inode number to filesystem block offset.
1994-05-24 10:09:53 +00:00
* inode number to cylinder group number.
2002-05-16 21:28:32 +00:00
* inode number to filesystem block address.
1994-05-24 10:09:53 +00:00
*/
#define ino_to_cg(fs, x) (((ino_t)(x)) / (fs)->fs_ipg)
1994-05-24 10:09:53 +00:00
#define ino_to_fsba(fs, x) \
((ufs2_daddr_t)(cgimin(fs, ino_to_cg(fs, (ino_t)(x))) + \
(blkstofrags((fs), ((((ino_t)(x)) % (fs)->fs_ipg) / INOPB(fs))))))
#define ino_to_fsbo(fs, x) (((ino_t)(x)) % INOPB(fs))
1994-05-24 10:09:53 +00:00
/*
2002-05-16 21:28:32 +00:00
* Give cylinder group number for a filesystem block.
* Give cylinder group block number for a filesystem block.
1994-05-24 10:09:53 +00:00
*/
#define dtog(fs, d) ((d) / (fs)->fs_fpg)
#define dtogd(fs, d) ((d) % (fs)->fs_fpg)
/*
* Extract the bits for a block from a map.
* Compute the cylinder and rotational position of a cyl block addr.
*/
#define blkmap(fs, map, loc) \
(((map)[(loc) / NBBY] >> ((loc) % NBBY)) & (0xff >> (NBBY - (fs)->fs_frag)))
/*
* The following macros optimize certain frequently calculated
* quantities by using shifts and masks in place of divisions
* modulos and multiplications.
*/
#define blkoff(fs, loc) /* calculates (loc % fs->fs_bsize) */ \
((loc) & (fs)->fs_qbmask)
#define fragoff(fs, loc) /* calculates (loc % fs->fs_fsize) */ \
((loc) & (fs)->fs_qfmask)
This commit adds basic support for the UFS2 filesystem. The UFS2 filesystem expands the inode to 256 bytes to make space for 64-bit block pointers. It also adds a file-creation time field, an ability to use jumbo blocks per inode to allow extent like pointer density, and space for extended attributes (up to twice the filesystem block size worth of attributes, e.g., on a 16K filesystem, there is space for 32K of attributes). UFS2 fully supports and runs existing UFS1 filesystems. New filesystems built using newfs can be built in either UFS1 or UFS2 format using the -O option. In this commit UFS1 is the default format, so if you want to build UFS2 format filesystems, you must specify -O 2. This default will be changed to UFS2 when UFS2 proves itself to be stable. In this commit the boot code for reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c) as there is insufficient space in the boot block. Once the size of the boot block is increased, this code can be defined. Things to note: the definition of SBSIZE has changed to SBLOCKSIZE. The header file <ufs/ufs/dinode.h> must be included before <ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and ufs_lbn_t. Still TODO: Verify that the first level bootstraps work for all the architectures. Convert the utility ffsinfo to understand UFS2 and test growfs. Add support for the extended attribute storage. Update soft updates to ensure integrity of extended attribute storage. Switch the current extended attribute interfaces to use the extended attribute storage. Add the extent like functionality (framework is there, but is currently never used). Sponsored by: DARPA & NAI Labs. Reviewed by: Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00
#define lfragtosize(fs, frag) /* calculates ((off_t)frag * fs->fs_fsize) */ \
(((off_t)(frag)) << (fs)->fs_fshift)
#define lblktosize(fs, blk) /* calculates ((off_t)blk * fs->fs_bsize) */ \
(((off_t)(blk)) << (fs)->fs_bshift)
/* Use this only when `blk' is known to be small, e.g., < NDADDR. */
#define smalllblktosize(fs, blk) /* calculates (blk * fs->fs_bsize) */ \
1994-05-24 10:09:53 +00:00
((blk) << (fs)->fs_bshift)
#define lblkno(fs, loc) /* calculates (loc / fs->fs_bsize) */ \
((loc) >> (fs)->fs_bshift)
#define numfrags(fs, loc) /* calculates (loc / fs->fs_fsize) */ \
((loc) >> (fs)->fs_fshift)
#define blkroundup(fs, size) /* calculates roundup(size, fs->fs_bsize) */ \
(((size) + (fs)->fs_qbmask) & (fs)->fs_bmask)
#define fragroundup(fs, size) /* calculates roundup(size, fs->fs_fsize) */ \
(((size) + (fs)->fs_qfmask) & (fs)->fs_fmask)
#define fragstoblks(fs, frags) /* calculates (frags / fs->fs_frag) */ \
((frags) >> (fs)->fs_fragshift)
#define blkstofrags(fs, blks) /* calculates (blks * fs->fs_frag) */ \
((blks) << (fs)->fs_fragshift)
#define fragnum(fs, fsb) /* calculates (fsb % fs->fs_frag) */ \
((fsb) & ((fs)->fs_frag - 1))
#define blknum(fs, fsb) /* calculates rounddown(fsb, fs->fs_frag) */ \
((fsb) &~ ((fs)->fs_frag - 1))
/*
* Determine the number of available frags given a
* percentage to hold in reserve.
1994-05-24 10:09:53 +00:00
*/
#define freespace(fs, percentreserved) \
(blkstofrags((fs), (fs)->fs_cstotal.cs_nbfree) + \
(fs)->fs_cstotal.cs_nffree - \
(((off_t)((fs)->fs_dsize)) * (percentreserved) / 100))
1994-05-24 10:09:53 +00:00
/*
2002-05-16 21:28:32 +00:00
* Determining the size of a file block in the filesystem.
1994-05-24 10:09:53 +00:00
*/
#define blksize(fs, ip, lbn) \
(((lbn) >= NDADDR || (ip)->i_size >= smalllblktosize(fs, (lbn) + 1)) \
1994-05-24 10:09:53 +00:00
? (fs)->fs_bsize \
: (fragroundup(fs, blkoff(fs, (ip)->i_size))))
#define sblksize(fs, size, lbn) \
(((lbn) >= NDADDR || (size) >= ((lbn) + 1) << (fs)->fs_bshift) \
? (fs)->fs_bsize \
: (fragroundup(fs, blkoff(fs, (size)))))
/*
* Number of indirects in a filesystem block.
*/
#define NINDIR(fs) ((fs)->fs_nindir)
/*
* Indirect lbns are aligned on NDADDR addresses where single indirects
* are the negated address of the lowest lbn reachable, double indirects
* are this lbn - 1 and triple indirects are this lbn - 2. This yields
* an unusual bit order to determine level.
*/
static inline int
lbn_level(ufs_lbn_t lbn)
{
if (lbn >= 0)
return 0;
switch (lbn & 0x3) {
case 0:
return (0);
case 1:
break;
case 2:
return (2);
case 3:
return (1);
default:
break;
}
return (-1);
}
static inline ufs_lbn_t
lbn_offset(struct fs *fs, int level)
{
ufs_lbn_t res;
for (res = 1; level > 0; level--)
res *= NINDIR(fs);
return (res);
}
1994-05-24 10:09:53 +00:00
/*
* Number of inodes in a secondary storage block/fragment.
1994-05-24 10:09:53 +00:00
*/
#define INOPB(fs) ((fs)->fs_inopb)
#define INOPF(fs) ((fs)->fs_inopb >> (fs)->fs_fragshift)
/*
* Softdep journal record format.
*/
#define JOP_ADDREF 1 /* Add a reference to an inode. */
#define JOP_REMREF 2 /* Remove a reference from an inode. */
#define JOP_NEWBLK 3 /* Allocate a block. */
#define JOP_FREEBLK 4 /* Free a block or a tree of blocks. */
#define JOP_MVREF 5 /* Move a reference from one off to another. */
#define JOP_TRUNC 6 /* Partial truncation record. */
Implement fully asynchronous partial truncation with softupdates journaling to resolve errors which can cause corruption on recovery with the old synchronous mechanism. - Append partial truncation freework structures to indirdeps while truncation is proceeding. These prevent new block pointers from becoming valid until truncation completes and serialize truncations. - On completion of a partial truncate journal work waits for zeroed pointers to hit indirects. - softdep_journal_freeblocks() handles last frag allocation and last block zeroing. - vtruncbuf/ffs_page_remove moved into softdep_*_freeblocks() so it is only implemented in one place. - Block allocation failure handling moved up one level so it does not proceed with buf locks held. This permits us to do more extensive reclaims when filesystem space is exhausted. - softdep_sync_metadata() is broken into two parts, the first executes once at the start of ffs_syncvnode() and flushes truncations and inode dependencies. The second is called on each locked buf. This eliminates excessive looping and rollbacks. - Improve the mechanism in process_worklist_item() that handles acquiring vnode locks for handle_workitem_remove() so that it works more generally and does not loop excessively over the same worklist items on each call. - Don't corrupt directories by zeroing the tail in fsck. This is only done for regular files. - Push a fsync complete record for files that need it so the checker knows a truncation in the journal is no longer valid. Discussed with: mckusick, kib (ffs_pages_remove and ffs_truncate parts) Tested by: pho
2011-06-10 22:48:35 +00:00
#define JOP_SYNC 7 /* fsync() complete record. */
#define JREC_SIZE 32 /* Record and segment header size. */
#define SUJ_MIN (4 * 1024 * 1024) /* Minimum journal size */
#define SUJ_MAX (32 * 1024 * 1024) /* Maximum journal size */
#define SUJ_FILE ".sujournal" /* Journal file name */
/*
* Size of the segment record header. There is at most one for each disk
* block in the journal. The segment header is followed by an array of
* records. fsck depends on the first element in each record being 'op'
* and the second being 'ino'. Segments may span multiple disk blocks but
* the header is present on each.
*/
struct jsegrec {
uint64_t jsr_seq; /* Our sequence number */
uint64_t jsr_oldest; /* Oldest valid sequence number */
uint16_t jsr_cnt; /* Count of valid records */
uint16_t jsr_blocks; /* Count of device bsize blocks. */
uint32_t jsr_crc; /* 32bit crc of the valid space */
ufs_time_t jsr_time; /* timestamp for mount instance */
};
/*
* Reference record. Records a single link count modification.
*/
struct jrefrec {
uint32_t jr_op;
uint32_t jr_ino;
uint32_t jr_parent;
uint16_t jr_nlink;
uint16_t jr_mode;
int64_t jr_diroff;
uint64_t jr_unused;
};
/*
* Move record. Records a reference moving within a directory block. The
* nlink is unchanged but we must search both locations.
*/
struct jmvrec {
uint32_t jm_op;
uint32_t jm_ino;
uint32_t jm_parent;
uint16_t jm_unused;
int64_t jm_oldoff;
int64_t jm_newoff;
};
/*
* Block record. A set of frags or tree of blocks starting at an indirect are
* freed or a set of frags are allocated.
*/
struct jblkrec {
uint32_t jb_op;
uint32_t jb_ino;
ufs2_daddr_t jb_blkno;
ufs_lbn_t jb_lbn;
uint16_t jb_frags;
uint16_t jb_oldfrags;
uint32_t jb_unused;
};
/*
* Truncation record. Records a partial truncation so that it may be
Implement fully asynchronous partial truncation with softupdates journaling to resolve errors which can cause corruption on recovery with the old synchronous mechanism. - Append partial truncation freework structures to indirdeps while truncation is proceeding. These prevent new block pointers from becoming valid until truncation completes and serialize truncations. - On completion of a partial truncate journal work waits for zeroed pointers to hit indirects. - softdep_journal_freeblocks() handles last frag allocation and last block zeroing. - vtruncbuf/ffs_page_remove moved into softdep_*_freeblocks() so it is only implemented in one place. - Block allocation failure handling moved up one level so it does not proceed with buf locks held. This permits us to do more extensive reclaims when filesystem space is exhausted. - softdep_sync_metadata() is broken into two parts, the first executes once at the start of ffs_syncvnode() and flushes truncations and inode dependencies. The second is called on each locked buf. This eliminates excessive looping and rollbacks. - Improve the mechanism in process_worklist_item() that handles acquiring vnode locks for handle_workitem_remove() so that it works more generally and does not loop excessively over the same worklist items on each call. - Don't corrupt directories by zeroing the tail in fsck. This is only done for regular files. - Push a fsync complete record for files that need it so the checker knows a truncation in the journal is no longer valid. Discussed with: mckusick, kib (ffs_pages_remove and ffs_truncate parts) Tested by: pho
2011-06-10 22:48:35 +00:00
* completed at check time. Also used for sync records.
*/
struct jtrncrec {
uint32_t jt_op;
uint32_t jt_ino;
int64_t jt_size;
uint32_t jt_extsize;
uint32_t jt_pad[3];
};
union jrec {
struct jsegrec rec_jsegrec;
struct jrefrec rec_jrefrec;
struct jmvrec rec_jmvrec;
struct jblkrec rec_jblkrec;
struct jtrncrec rec_jtrncrec;
};
#ifdef CTASSERT
CTASSERT(sizeof(struct jsegrec) == JREC_SIZE);
CTASSERT(sizeof(struct jrefrec) == JREC_SIZE);
CTASSERT(sizeof(struct jmvrec) == JREC_SIZE);
CTASSERT(sizeof(struct jblkrec) == JREC_SIZE);
CTASSERT(sizeof(struct jtrncrec) == JREC_SIZE);
CTASSERT(sizeof(union jrec) == JREC_SIZE);
#endif
1994-05-24 10:09:53 +00:00
extern int inside[], around[];
extern u_char *fragtbl[];
#endif