87241caa43
Firstly, now our read-ahead clustering is on a file descriptor basis and not on a per-vnode basis. This will allow multiple processes reading the same file to take advantage of read-ahead clustering. Secondly, there previously was a problem with large reads still using the ramp-up algorithm. Of course, that was bogus, and now we read the entire "chunk" off of the disk in one operation. The read-ahead clustering algorithm should use less CPU than the previous also (I hope :-)).
333 lines
8.6 KiB
C
333 lines
8.6 KiB
C
/*
|
|
* modified for Lites 1.1
|
|
*
|
|
* Aug 1995, Godmar Back (gback@cs.utah.edu)
|
|
* University of Utah, Department of Computer Science
|
|
*/
|
|
/*
|
|
* Copyright (c) 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.
|
|
*
|
|
* @(#)ufs_readwrite.c 8.7 (Berkeley) 1/21/94
|
|
*/
|
|
|
|
#if !defined(__FreeBSD__)
|
|
#include "diagnostic.h"
|
|
#endif
|
|
|
|
#define BLKSIZE(a, b, c) blksize(a, b, c)
|
|
#define FS struct ext2_sb_info
|
|
#define I_FS i_e2fs
|
|
#define READ ext2_read
|
|
#define READ_S "ext2_read"
|
|
#define WRITE ext2_write
|
|
#define WRITE_S "ext2_write"
|
|
|
|
/*
|
|
* Vnode op for reading.
|
|
*/
|
|
/* ARGSUSED */
|
|
static int
|
|
READ(ap)
|
|
struct vop_read_args /* {
|
|
struct vnode *a_vp;
|
|
struct uio *a_uio;
|
|
int a_ioflag;
|
|
struct ucred *a_cred;
|
|
} */ *ap;
|
|
{
|
|
register struct vnode *vp;
|
|
register struct inode *ip;
|
|
register struct uio *uio;
|
|
register FS *fs;
|
|
struct buf *bp;
|
|
daddr_t lbn, nextlbn;
|
|
off_t bytesinfile;
|
|
long size, xfersize, blkoffset;
|
|
int error;
|
|
u_short mode;
|
|
|
|
vp = ap->a_vp;
|
|
ip = VTOI(vp);
|
|
mode = ip->i_mode;
|
|
uio = ap->a_uio;
|
|
|
|
#ifdef DIAGNOSTIC
|
|
if (uio->uio_rw != UIO_READ)
|
|
panic("%s: mode", READ_S);
|
|
|
|
if (vp->v_type == VLNK) {
|
|
if ((int)ip->i_size < vp->v_mount->mnt_maxsymlinklen)
|
|
panic("%s: short symlink", READ_S);
|
|
} else if (vp->v_type != VREG && vp->v_type != VDIR)
|
|
panic("%s: type %d", READ_S, vp->v_type);
|
|
#endif
|
|
fs = ip->I_FS;
|
|
#if 0
|
|
if ((u_quad_t)uio->uio_offset > fs->fs_maxfilesize)
|
|
return (EFBIG);
|
|
#endif
|
|
|
|
for (error = 0, bp = NULL; uio->uio_resid > 0; bp = NULL) {
|
|
if ((bytesinfile = ip->i_size - uio->uio_offset) <= 0)
|
|
break;
|
|
lbn = lblkno(fs, uio->uio_offset);
|
|
nextlbn = lbn + 1;
|
|
size = BLKSIZE(fs, ip, lbn);
|
|
blkoffset = blkoff(fs, uio->uio_offset);
|
|
|
|
xfersize = fs->s_frag_size - blkoffset;
|
|
if (uio->uio_resid < xfersize)
|
|
xfersize = uio->uio_resid;
|
|
if (bytesinfile < xfersize)
|
|
xfersize = bytesinfile;
|
|
|
|
if (lblktosize(fs, nextlbn) >= ip->i_size)
|
|
error = bread(vp, lbn, size, NOCRED, &bp);
|
|
else if (doclusterread)
|
|
error = cluster_read(vp,
|
|
ip->i_size, lbn, size, NOCRED,
|
|
uio->uio_resid, (ap->a_ioflag >> 16), &bp);
|
|
else if (lbn - 1 == vp->v_lastr) {
|
|
int nextsize = BLKSIZE(fs, ip, nextlbn);
|
|
error = breadn(vp, lbn,
|
|
size, &nextlbn, &nextsize, 1, NOCRED, &bp);
|
|
} else
|
|
error = bread(vp, lbn, size, NOCRED, &bp);
|
|
if (error) {
|
|
brelse(bp);
|
|
bp = NULL;
|
|
break;
|
|
}
|
|
vp->v_lastr = lbn;
|
|
|
|
/*
|
|
* We should only get non-zero b_resid when an I/O error
|
|
* has occurred, which should cause us to break above.
|
|
* However, if the short read did not cause an error,
|
|
* then we want to ensure that we do not uiomove bad
|
|
* or uninitialized data.
|
|
*/
|
|
size -= bp->b_resid;
|
|
if (size < xfersize) {
|
|
if (size == 0)
|
|
break;
|
|
xfersize = size;
|
|
}
|
|
if (uio->uio_segflg != UIO_NOCOPY)
|
|
ip->i_flag |= IN_RECURSE;
|
|
error =
|
|
uiomove((char *)bp->b_data + blkoffset, (int)xfersize, uio);
|
|
if (uio->uio_segflg != UIO_NOCOPY)
|
|
ip->i_flag &= ~IN_RECURSE;
|
|
if (error)
|
|
break;
|
|
#if !defined(__FreeBSD__)
|
|
if (S_ISREG(mode) && (xfersize + blkoffset == fs->s_frag_size ||
|
|
uio->uio_offset == ip->i_size))
|
|
bp->b_flags |= B_AGE;
|
|
#endif
|
|
|
|
bqrelse(bp);
|
|
}
|
|
if (bp != NULL)
|
|
bqrelse(bp);
|
|
if (!(vp->v_mount->mnt_flag & MNT_NOATIME))
|
|
ip->i_flag |= IN_ACCESS;
|
|
return (error);
|
|
}
|
|
|
|
/*
|
|
* Vnode op for writing.
|
|
*/
|
|
static int
|
|
WRITE(ap)
|
|
struct vop_write_args /* {
|
|
struct vnode *a_vp;
|
|
struct uio *a_uio;
|
|
int a_ioflag;
|
|
struct ucred *a_cred;
|
|
} */ *ap;
|
|
{
|
|
register struct vnode *vp;
|
|
register struct uio *uio;
|
|
register struct inode *ip;
|
|
register FS *fs;
|
|
struct buf *bp;
|
|
struct proc *p;
|
|
daddr_t lbn;
|
|
off_t osize;
|
|
int blkoffset, error, flags, ioflag, resid, size, xfersize;
|
|
struct timeval tv;
|
|
|
|
ioflag = ap->a_ioflag;
|
|
uio = ap->a_uio;
|
|
vp = ap->a_vp;
|
|
ip = VTOI(vp);
|
|
|
|
#ifdef DIAGNOSTIC
|
|
if (uio->uio_rw != UIO_WRITE)
|
|
panic("%s: mode", WRITE_S);
|
|
#endif
|
|
|
|
switch (vp->v_type) {
|
|
case VREG:
|
|
if (ioflag & IO_APPEND)
|
|
uio->uio_offset = ip->i_size;
|
|
if ((ip->i_flags & APPEND) && uio->uio_offset != ip->i_size)
|
|
return (EPERM);
|
|
/* FALLTHROUGH */
|
|
case VLNK:
|
|
break;
|
|
case VDIR:
|
|
if ((ioflag & IO_SYNC) == 0)
|
|
panic("%s: nonsync dir write", WRITE_S);
|
|
break;
|
|
default:
|
|
panic("%s: type", WRITE_S);
|
|
}
|
|
|
|
fs = ip->I_FS;
|
|
#if 0
|
|
if (uio->uio_offset < 0 ||
|
|
(u_quad_t)uio->uio_offset + uio->uio_resid > fs->fs_maxfilesize)
|
|
return (EFBIG);
|
|
#endif
|
|
/*
|
|
* Maybe this should be above the vnode op call, but so long as
|
|
* file servers have no limits, I don't think it matters.
|
|
*/
|
|
p = uio->uio_procp;
|
|
if (vp->v_type == VREG && p &&
|
|
uio->uio_offset + uio->uio_resid >
|
|
p->p_rlimit[RLIMIT_FSIZE].rlim_cur) {
|
|
psignal(p, SIGXFSZ);
|
|
return (EFBIG);
|
|
}
|
|
|
|
resid = uio->uio_resid;
|
|
osize = ip->i_size;
|
|
flags = ioflag & IO_SYNC ? B_SYNC : 0;
|
|
|
|
for (error = 0; uio->uio_resid > 0;) {
|
|
lbn = lblkno(fs, uio->uio_offset);
|
|
blkoffset = blkoff(fs, uio->uio_offset);
|
|
xfersize = fs->s_frag_size - blkoffset;
|
|
if (uio->uio_resid < xfersize)
|
|
xfersize = uio->uio_resid;
|
|
|
|
#if defined(__FreeBSD__)
|
|
if (uio->uio_offset + xfersize > ip->i_size)
|
|
vnode_pager_setsize(vp, uio->uio_offset + xfersize);
|
|
#endif
|
|
|
|
if (fs->s_frag_size > xfersize)
|
|
flags |= B_CLRBUF;
|
|
else
|
|
flags &= ~B_CLRBUF;
|
|
|
|
error = ext2_balloc(ip,
|
|
lbn, blkoffset + xfersize, ap->a_cred, &bp, flags);
|
|
if (error)
|
|
break;
|
|
|
|
if (uio->uio_offset + xfersize > ip->i_size) {
|
|
ip->i_size = uio->uio_offset + xfersize;
|
|
#if !defined(__FreeBSD__)
|
|
vnode_pager_setsize(vp, (u_long)ip->i_size);
|
|
#endif
|
|
}
|
|
#if !defined(__FreeBSD__)
|
|
(void)vnode_pager_uncache(vp);
|
|
#endif
|
|
|
|
size = BLKSIZE(fs, ip, lbn) - bp->b_resid;
|
|
if (size < xfersize)
|
|
xfersize = size;
|
|
|
|
if (uio->uio_segflg != UIO_NOCOPY)
|
|
ip->i_flag |= IN_RECURSE;
|
|
error =
|
|
uiomove((char *)bp->b_data + blkoffset, (int)xfersize, uio);
|
|
if (uio->uio_segflg != UIO_NOCOPY)
|
|
ip->i_flag &= ~IN_RECURSE;
|
|
if (ioflag & IO_VMIO)
|
|
bp->b_flags |= B_RELBUF;
|
|
|
|
if (ioflag & IO_SYNC) {
|
|
(void)bwrite(bp);
|
|
} else if (xfersize + blkoffset == fs->s_frag_size) {
|
|
if (doclusterwrite) {
|
|
#if defined(__FreeBSD__)
|
|
bp->b_flags |= B_CLUSTEROK;
|
|
#endif
|
|
cluster_write(bp, ip->i_size);
|
|
} else {
|
|
#if !defined(__FreeBSD__)
|
|
bp->b_flags |= B_AGE;
|
|
#endif
|
|
bawrite(bp);
|
|
}
|
|
} else {
|
|
#if defined(__FreeBSD__)
|
|
bp->b_flags |= B_CLUSTEROK;
|
|
#endif
|
|
bdwrite(bp);
|
|
}
|
|
if (error || xfersize == 0)
|
|
break;
|
|
ip->i_flag |= IN_CHANGE | IN_UPDATE;
|
|
}
|
|
/*
|
|
* If we successfully wrote any data, and we are not the superuser
|
|
* we clear the setuid and setgid bits as a precaution against
|
|
* tampering.
|
|
*/
|
|
if (resid > uio->uio_resid && ap->a_cred && ap->a_cred->cr_uid != 0)
|
|
ip->i_mode &= ~(ISUID | ISGID);
|
|
if (error) {
|
|
if (ioflag & IO_UNIT) {
|
|
(void)VOP_TRUNCATE(vp, osize,
|
|
ioflag & IO_SYNC, ap->a_cred, uio->uio_procp);
|
|
uio->uio_offset -= resid - uio->uio_resid;
|
|
uio->uio_resid = resid;
|
|
}
|
|
} else if (resid > uio->uio_resid && (ioflag & IO_SYNC)) {
|
|
#if !defined(__FreeBSD__)
|
|
get_time(&tv);
|
|
#else
|
|
tv = time;
|
|
#endif
|
|
error = VOP_UPDATE(vp, &tv, &tv, 1);
|
|
}
|
|
return (error);
|
|
}
|