1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
|
/* $OpenBSD: vfs_cluster.c,v 1.42 2014/07/12 18:43:32 tedu Exp $ */
/* $NetBSD: vfs_cluster.c,v 1.12 1996/04/22 01:39:05 christos Exp $ */
/*
* 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. 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.
*
* @(#)vfs_cluster.c 8.8 (Berkeley) 7/28/94
*/
#include <sys/param.h>
#include <sys/proc.h>
#include <sys/buf.h>
#include <sys/vnode.h>
#include <sys/mount.h>
#include <sys/malloc.h>
#include <sys/systm.h>
#include <sys/resourcevar.h>
void cluster_wbuild(struct vnode *, struct buf *, long, daddr_t, int,
daddr_t);
struct cluster_save *cluster_collectbufs(struct vnode *, struct cluster_info *,
struct buf *);
/*
* Do clustered write for FFS.
*
* Three cases:
* 1. Write is not sequential (write asynchronously)
* Write is sequential:
* 2. beginning of cluster - begin cluster
* 3. middle of a cluster - add to cluster
* 4. end of a cluster - asynchronously write cluster
*/
void
cluster_write(struct buf *bp, struct cluster_info *ci, u_quad_t filesize)
{
struct vnode *vp;
daddr_t lbn;
int maxclen, cursize;
vp = bp->b_vp;
lbn = bp->b_lblkno;
/* Initialize vnode to beginning of file. */
if (lbn == 0)
ci->ci_lasta = ci->ci_clen = ci->ci_cstart = ci->ci_lastw = 0;
if (ci->ci_clen == 0 || lbn != ci->ci_lastw + 1 ||
(bp->b_blkno != ci->ci_lasta + btodb(bp->b_bcount))) {
maxclen = MAXBSIZE / vp->v_mount->mnt_stat.f_iosize - 1;
if (ci->ci_clen != 0) {
/*
* Next block is not sequential.
*
* If we are not writing at end of file, the process
* seeked to another point in the file since its
* last write, or we have reached our maximum
* cluster size, then push the previous cluster.
* Otherwise try reallocating to make it sequential.
*/
cursize = ci->ci_lastw - ci->ci_cstart + 1;
if (((u_quad_t)(lbn + 1)) * bp->b_bcount != filesize ||
lbn != ci->ci_lastw + 1 || ci->ci_clen <= cursize) {
cluster_wbuild(vp, NULL, bp->b_bcount,
ci->ci_cstart, cursize, lbn);
} else {
struct buf **bpp, **endbp;
struct cluster_save *buflist;
buflist = cluster_collectbufs(vp, ci, bp);
endbp = &buflist->bs_children
[buflist->bs_nchildren - 1];
if (VOP_REALLOCBLKS(vp, buflist)) {
/*
* Failed, push the previous cluster.
*/
for (bpp = buflist->bs_children;
bpp < endbp; bpp++)
brelse(*bpp);
free(buflist, M_VCLUSTER, 0);
cluster_wbuild(vp, NULL, bp->b_bcount,
ci->ci_cstart, cursize, lbn);
} else {
/*
* Succeeded, keep building cluster.
*/
for (bpp = buflist->bs_children;
bpp <= endbp; bpp++)
bdwrite(*bpp);
free(buflist, M_VCLUSTER, 0);
ci->ci_lastw = lbn;
ci->ci_lasta = bp->b_blkno;
return;
}
}
}
/*
* Consider beginning a cluster.
* If at end of file, make cluster as large as possible,
* otherwise find size of existing cluster.
*/
if ((u_quad_t)(lbn + 1) * (u_quad_t)bp->b_bcount != filesize &&
(VOP_BMAP(vp, lbn, NULL, &bp->b_blkno, &maxclen) ||
bp->b_blkno == -1)) {
bawrite(bp);
ci->ci_clen = 0;
ci->ci_lasta = bp->b_blkno;
ci->ci_cstart = lbn + 1;
ci->ci_lastw = lbn;
return;
}
ci->ci_clen = maxclen;
if (maxclen == 0) { /* I/O not contiguous */
ci->ci_cstart = lbn + 1;
bawrite(bp);
} else { /* Wait for rest of cluster */
ci->ci_cstart = lbn;
bdwrite(bp);
}
} else if (lbn == ci->ci_cstart + ci->ci_clen) {
/*
* At end of cluster, write it out.
*/
cluster_wbuild(vp, bp, bp->b_bcount, ci->ci_cstart,
ci->ci_clen + 1, lbn);
ci->ci_clen = 0;
ci->ci_cstart = lbn + 1;
} else
/*
* In the middle of a cluster, so just delay the
* I/O for now.
*/
bdwrite(bp);
ci->ci_lastw = lbn;
ci->ci_lasta = bp->b_blkno;
}
/*
* The last lbn argument is the current block on which I/O is being
* performed. Check to see that it doesn't fall in the middle of
* the current block (if last_bp == NULL).
*/
void
cluster_wbuild(struct vnode *vp, struct buf *last_bp, long size,
daddr_t start_lbn, int len, daddr_t lbn)
{
struct buf *bp;
#ifdef DIAGNOSTIC
if (size != vp->v_mount->mnt_stat.f_iosize)
panic("cluster_wbuild: size %ld != filesize %u",
size, vp->v_mount->mnt_stat.f_iosize);
#endif
redo:
while ((!incore(vp, start_lbn) || start_lbn == lbn) && len) {
++start_lbn;
--len;
}
/* Get more memory for current buffer */
if (len <= 1) {
if (last_bp) {
bawrite(last_bp);
} else if (len) {
bp = getblk(vp, start_lbn, size, 0, 0);
/*
* The buffer could have already been flushed out of
* the cache. If that has happened, we'll get a new
* buffer here with random data, just drop it.
*/
if ((bp->b_flags & B_DELWRI) == 0)
brelse(bp);
else
bawrite(bp);
}
return;
}
bp = getblk(vp, start_lbn, size, 0, 0);
if (!(bp->b_flags & B_DELWRI)) {
++start_lbn;
--len;
brelse(bp);
goto redo;
}
++start_lbn;
--len;
bawrite(bp);
goto redo;
}
/*
* Collect together all the buffers in a cluster.
* Plus add one additional buffer.
*/
struct cluster_save *
cluster_collectbufs(struct vnode *vp, struct cluster_info *ci,
struct buf *last_bp)
{
struct cluster_save *buflist;
daddr_t lbn;
int i, len;
len = ci->ci_lastw - ci->ci_cstart + 1;
buflist = malloc(sizeof(struct buf *) * (len + 1) + sizeof(*buflist),
M_VCLUSTER, M_WAITOK);
buflist->bs_nchildren = 0;
buflist->bs_children = (struct buf **)(buflist + 1);
for (lbn = ci->ci_cstart, i = 0; i < len; lbn++, i++)
(void)bread(vp, lbn, last_bp->b_bcount,
&buflist->bs_children[i]);
buflist->bs_children[i] = last_bp;
buflist->bs_nchildren = i + 1;
return (buflist);
}
|
