freebsd-dev/sys/ufs/lfs/lfs_balloc.c

242 lines
7.3 KiB
C
Raw Normal View History

1994-05-24 10:09:53 +00:00
/*
* Copyright (c) 1989, 1991, 1993
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the University of
* California, Berkeley and its contributors.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* @(#)lfs_balloc.c 8.4 (Berkeley) 5/8/95
* $Id: lfs_balloc.c,v 1.10 1997/02/22 09:47:17 peter Exp $
1994-05-24 10:09:53 +00:00
*/
#include <sys/param.h>
#include <sys/systm.h>
1994-05-24 10:09:53 +00:00
#include <sys/buf.h>
#include <sys/proc.h>
#include <sys/vnode.h>
#include <sys/mount.h>
#include <sys/resourcevar.h>
#include <miscfs/specfs/specdev.h>
#include <ufs/ufs/quota.h>
#include <ufs/ufs/inode.h>
#include <ufs/ufs/ufsmount.h>
#include <ufs/ufs/ufs_extern.h>
1994-05-24 10:09:53 +00:00
#include <ufs/lfs/lfs.h>
#include <ufs/lfs/lfs_extern.h>
int
lfs_balloc(vp, offset, iosize, lbn, bpp)
1994-05-24 10:09:53 +00:00
struct vnode *vp;
int offset;
1994-05-24 10:09:53 +00:00
u_long iosize;
ufs_daddr_t lbn;
1994-05-24 10:09:53 +00:00
struct buf **bpp;
{
struct buf *ibp, *bp;
struct inode *ip;
struct lfs *fs;
struct indir indirs[NIADDR+2];
ufs_daddr_t daddr, lastblock;
int bb; /* number of disk blocks in a block disk blocks */
int error, frags, i, nsize, osize, num;
1994-05-24 10:09:53 +00:00
ip = VTOI(vp);
fs = ip->i_lfs;
1995-05-30 08:16:23 +00:00
/*
1994-05-24 10:09:53 +00:00
* Three cases: it's a block beyond the end of file, it's a block in
* the file that may or may not have been assigned a disk address or
* we're writing an entire block. Note, if the daddr is unassigned,
* the block might still have existed in the cache (if it was read
* or written earlier). If it did, make sure we don't count it as a
* new block or zero out its contents. If it did not, make sure
* we allocate any necessary indirect blocks.
* If we are writing a block beyond the end of the file, we need to
* check if the old last block was a fragment. If it was, we need
* to rewrite it.
1994-05-24 10:09:53 +00:00
*/
*bpp = NULL;
if (error = ufs_bmaparray(vp, lbn, &daddr, &indirs[0], &num, NULL, NULL ))
1994-05-24 10:09:53 +00:00
return (error);
/* Check for block beyond end of file and fragment extension needed. */
lastblock = lblkno(fs, ip->i_size);
if (lastblock < NDADDR && lastblock < lbn) {
osize = blksize(fs, ip, lastblock);
if (osize < fs->lfs_bsize && osize > 0) {
if (error = lfs_fragextend(vp, osize, fs->lfs_bsize,
lastblock, &bp))
return(error);
ip->i_size = (lastblock + 1) * fs->lfs_bsize;
vnode_pager_setsize(vp, (u_long)ip->i_size);
ip->i_flag |= IN_CHANGE | IN_UPDATE;
VOP_BWRITE(bp);
}
}
1994-05-24 10:09:53 +00:00
bb = VFSTOUFS(vp->v_mount)->um_seqinc;
if (daddr == UNASSIGNED)
/* May need to allocate indirect blocks */
for (i = 1; i < num; ++i)
if (!indirs[i].in_exists) {
ibp = getblk(vp, indirs[i].in_lbn, fs->lfs_bsize,
0, 0);
1994-05-24 10:09:53 +00:00
if (!(ibp->b_flags & (B_DONE | B_DELWRI))) {
if (!ISSPACE(fs, bb, curproc->p_ucred)){
ibp->b_flags |= B_INVAL;
brelse(ibp);
error = ENOSPC;
} else {
ip->i_blocks += bb;
ip->i_lfs->lfs_bfree -= bb;
vfs_bio_clrbuf(ibp);
1994-05-24 10:09:53 +00:00
error = VOP_BWRITE(ibp);
}
} else
panic ("Indirect block should not exist");
if (!ISSPACE(fs, bb, curproc->p_ucred)){
ibp->b_flags |= B_INVAL;
brelse(ibp);
return(ENOSPC);
} else {
ip->i_blocks += bb;
ip->i_lfs->lfs_bfree -= bb;
clrbuf(ibp);
if(error = VOP_BWRITE(ibp))
return(error);
}
1994-05-24 10:09:53 +00:00
}
/*
* If the block we are writing is a direct block, it's the last
* block in the file, and offset + iosize is less than a full
* block, we can write one or more fragments. There are two cases:
* the block is brand new and we should allocate it the correct
* size or it already exists and contains some fragments and
* may need to extend it.
*/
if (lbn < NDADDR && lblkno(fs, ip->i_size) == lbn) {
nsize = fragroundup(fs, offset + iosize);
frags = numfrags(fs, nsize);
bb = fragstodb(fs, frags);
if (lblktosize(fs, lbn) == ip->i_size)
/* Brand new block or fragment */
*bpp = bp = getblk(vp, lbn, nsize, 0, 0);
else {
/* Extend existing block */
if (error = lfs_fragextend(vp, (int)blksize(fs, ip, lbn),
nsize, lbn, &bp))
return(error);
*bpp = bp;
}
} else {
/*
* Get the existing block from the cache either because the
* block is 1) not a direct block or because it's not the last
* block in the file.
*/
frags = dbtofrags(fs, bb);
*bpp = bp = getblk(vp, lbn, blksize(fs, ip, lbn), 0, 0);
}
1994-05-24 10:09:53 +00:00
/*
* The block we are writing may be a brand new block
* in which case we need to do accounting (i.e. check
* for free space and update the inode number of blocks.
*/
1994-05-24 10:09:53 +00:00
if (!(bp->b_flags & (B_CACHE | B_DONE | B_DELWRI))) {
1995-05-30 08:16:23 +00:00
if (daddr == UNASSIGNED)
1994-05-24 10:09:53 +00:00
if (!ISSPACE(fs, bb, curproc->p_ucred)) {
bp->b_flags |= B_INVAL;
brelse(bp);
return(ENOSPC);
} else {
ip->i_blocks += bb;
ip->i_lfs->lfs_bfree -= bb;
if (iosize != fs->lfs_bsize)
vfs_bio_clrbuf(bp);
1994-05-24 10:09:53 +00:00
}
else if (iosize == fs->lfs_bsize)
/* Optimization: I/O is unnecessary. */
bp->b_blkno = daddr;
1994-05-24 10:09:53 +00:00
else {
/*
* We need to read the block to preserve the
* existing bytes.
*/
1994-05-24 10:09:53 +00:00
bp->b_blkno = daddr;
bp->b_flags |= B_READ;
These changes embody the support of the fully coherent merged VM buffer cache, much higher filesystem I/O performance, and much better paging performance. It represents the culmination of over 6 months of R&D. The majority of the merged VM/cache work is by John Dyson. The following highlights the most significant changes. Additionally, there are (mostly minor) changes to the various filesystem modules (nfs, msdosfs, etc) to support the new VM/buffer scheme. vfs_bio.c: Significant rewrite of most of vfs_bio to support the merged VM buffer cache scheme. The scheme is almost fully compatible with the old filesystem interface. Significant improvement in the number of opportunities for write clustering. vfs_cluster.c, vfs_subr.c Upgrade and performance enhancements in vfs layer code to support merged VM/buffer cache. Fixup of vfs_cluster to eliminate the bogus pagemove stuff. vm_object.c: Yet more improvements in the collapse code. Elimination of some windows that can cause list corruption. vm_pageout.c: Fixed it, it really works better now. Somehow in 2.0, some "enhancements" broke the code. This code has been reworked from the ground-up. vm_fault.c, vm_page.c, pmap.c, vm_object.c Support for small-block filesystems with merged VM/buffer cache scheme. pmap.c vm_map.c Dynamic kernel VM size, now we dont have to pre-allocate excessive numbers of kernel PTs. vm_glue.c Much simpler and more effective swapping code. No more gratuitous swapping. proc.h Fixed the problem that the p_lock flag was not being cleared on a fork. swap_pager.c, vnode_pager.c Removal of old vfs_bio cruft to support the past pseudo-coherency. Now the code doesn't need it anymore. machdep.c Changes to better support the parameter values for the merged VM/buffer cache scheme. machdep.c, kern_exec.c, vm_glue.c Implemented a seperate submap for temporary exec string space and another one to contain process upages. This eliminates all map fragmentation problems that previously existed. ffs_inode.c, ufs_inode.c, ufs_readwrite.c Changes for merged VM/buffer cache. Add "bypass" support for sneaking in on busy buffers. Submitted by: John Dyson and David Greenman
1995-01-09 16:06:02 +00:00
vfs_busy_pages(bp, 0);
1994-05-24 10:09:53 +00:00
VOP_STRATEGY(bp);
return(biowait(bp));
}
}
return (0);
}
lfs_fragextend(vp, osize, nsize, lbn, bpp)
struct vnode *vp;
int osize;
int nsize;
daddr_t lbn;
struct buf **bpp;
{
struct inode *ip;
struct lfs *fs;
long bb;
int error;
ip = VTOI(vp);
fs = ip->i_lfs;
bb = (long)fragstodb(fs, numfrags(fs, nsize - osize));
if (!ISSPACE(fs, bb, curproc->p_ucred)) {
return(ENOSPC);
}
if (error = bread(vp, lbn, osize, NOCRED, bpp)) {
brelse(*bpp);
return(error);
}
#ifdef QUOTA
if (error = chkdq(ip, bb, curproc->p_ucred, 0)) {
brelse(*bpp);
return (error);
}
#endif
ip->i_blocks += bb;
ip->i_flag |= IN_CHANGE | IN_UPDATE;
fs->lfs_bfree -= fragstodb(fs, numfrags(fs, (nsize - osize)));
allocbuf(*bpp, nsize);
bzero((char *)((*bpp)->b_data) + osize, (u_int)(nsize - osize));
return(0);
1994-05-24 10:09:53 +00:00
}