lfs_bio.c revision 1.52 1 1.52 yamt /* $NetBSD: lfs_bio.c,v 1.52 2002/12/28 14:39:08 yamt Exp $ */
2 1.2 cgd
3 1.7 perseant /*-
4 1.27 perseant * Copyright (c) 1999, 2000 The NetBSD Foundation, Inc.
5 1.7 perseant * All rights reserved.
6 1.7 perseant *
7 1.7 perseant * This code is derived from software contributed to The NetBSD Foundation
8 1.7 perseant * by Konrad E. Schroder <perseant (at) hhhh.org>.
9 1.7 perseant *
10 1.7 perseant * Redistribution and use in source and binary forms, with or without
11 1.7 perseant * modification, are permitted provided that the following conditions
12 1.7 perseant * are met:
13 1.7 perseant * 1. Redistributions of source code must retain the above copyright
14 1.7 perseant * notice, this list of conditions and the following disclaimer.
15 1.7 perseant * 2. Redistributions in binary form must reproduce the above copyright
16 1.7 perseant * notice, this list of conditions and the following disclaimer in the
17 1.7 perseant * documentation and/or other materials provided with the distribution.
18 1.7 perseant * 3. All advertising materials mentioning features or use of this software
19 1.7 perseant * must display the following acknowledgement:
20 1.7 perseant * This product includes software developed by the NetBSD
21 1.7 perseant * Foundation, Inc. and its contributors.
22 1.7 perseant * 4. Neither the name of The NetBSD Foundation nor the names of its
23 1.7 perseant * contributors may be used to endorse or promote products derived
24 1.7 perseant * from this software without specific prior written permission.
25 1.7 perseant *
26 1.7 perseant * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
27 1.7 perseant * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
28 1.7 perseant * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
29 1.7 perseant * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
30 1.7 perseant * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
31 1.7 perseant * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
32 1.7 perseant * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
33 1.7 perseant * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
34 1.7 perseant * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
35 1.7 perseant * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
36 1.7 perseant * POSSIBILITY OF SUCH DAMAGE.
37 1.7 perseant */
38 1.1 mycroft /*
39 1.1 mycroft * Copyright (c) 1991, 1993
40 1.1 mycroft * The Regents of the University of California. All rights reserved.
41 1.1 mycroft *
42 1.1 mycroft * Redistribution and use in source and binary forms, with or without
43 1.1 mycroft * modification, are permitted provided that the following conditions
44 1.1 mycroft * are met:
45 1.1 mycroft * 1. Redistributions of source code must retain the above copyright
46 1.1 mycroft * notice, this list of conditions and the following disclaimer.
47 1.1 mycroft * 2. Redistributions in binary form must reproduce the above copyright
48 1.1 mycroft * notice, this list of conditions and the following disclaimer in the
49 1.1 mycroft * documentation and/or other materials provided with the distribution.
50 1.1 mycroft * 3. All advertising materials mentioning features or use of this software
51 1.1 mycroft * must display the following acknowledgement:
52 1.1 mycroft * This product includes software developed by the University of
53 1.1 mycroft * California, Berkeley and its contributors.
54 1.1 mycroft * 4. Neither the name of the University nor the names of its contributors
55 1.1 mycroft * may be used to endorse or promote products derived from this software
56 1.1 mycroft * without specific prior written permission.
57 1.1 mycroft *
58 1.1 mycroft * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
59 1.1 mycroft * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
60 1.1 mycroft * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
61 1.1 mycroft * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
62 1.1 mycroft * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
63 1.1 mycroft * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
64 1.1 mycroft * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
65 1.1 mycroft * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
66 1.1 mycroft * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
67 1.1 mycroft * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
68 1.1 mycroft * SUCH DAMAGE.
69 1.1 mycroft *
70 1.6 fvdl * @(#)lfs_bio.c 8.10 (Berkeley) 6/10/95
71 1.1 mycroft */
72 1.39 lukem
73 1.39 lukem #include <sys/cdefs.h>
74 1.52 yamt __KERNEL_RCSID(0, "$NetBSD: lfs_bio.c,v 1.52 2002/12/28 14:39:08 yamt Exp $");
75 1.1 mycroft
76 1.1 mycroft #include <sys/param.h>
77 1.5 christos #include <sys/systm.h>
78 1.1 mycroft #include <sys/proc.h>
79 1.1 mycroft #include <sys/buf.h>
80 1.1 mycroft #include <sys/vnode.h>
81 1.1 mycroft #include <sys/resourcevar.h>
82 1.1 mycroft #include <sys/mount.h>
83 1.1 mycroft #include <sys/kernel.h>
84 1.1 mycroft
85 1.1 mycroft #include <ufs/ufs/inode.h>
86 1.1 mycroft #include <ufs/ufs/ufsmount.h>
87 1.5 christos #include <ufs/ufs/ufs_extern.h>
88 1.1 mycroft
89 1.7 perseant #include <sys/malloc.h>
90 1.1 mycroft #include <ufs/lfs/lfs.h>
91 1.1 mycroft #include <ufs/lfs/lfs_extern.h>
92 1.1 mycroft
93 1.7 perseant /* Macros to clear/set/test flags. */
94 1.7 perseant # define SET(t, f) (t) |= (f)
95 1.7 perseant # define CLR(t, f) (t) &= ~(f)
96 1.7 perseant # define ISSET(t, f) ((t) & (f))
97 1.7 perseant
98 1.1 mycroft /*
99 1.1 mycroft * LFS block write function.
100 1.1 mycroft *
101 1.1 mycroft * XXX
102 1.1 mycroft * No write cost accounting is done.
103 1.1 mycroft * This is almost certainly wrong for synchronous operations and NFS.
104 1.1 mycroft */
105 1.7 perseant int locked_queue_count = 0; /* XXX Count of locked-down buffers. */
106 1.7 perseant long locked_queue_bytes = 0L; /* XXX Total size of locked buffers. */
107 1.7 perseant int lfs_writing = 0; /* Set if already kicked off a writer
108 1.1 mycroft because of buffer space */
109 1.7 perseant extern int lfs_dostats;
110 1.7 perseant
111 1.27 perseant /*
112 1.51 yamt * reserved number/bytes of locked buffers
113 1.51 yamt */
114 1.51 yamt int locked_queue_rcount = 0;
115 1.51 yamt long locked_queue_rbytes = 0L;
116 1.51 yamt
117 1.51 yamt int lfs_fits_buf(struct lfs *, int, int);
118 1.52 yamt int lfs_reservebuf(struct lfs *, struct vnode *vp, struct vnode *vp2,
119 1.52 yamt int, int);
120 1.52 yamt int lfs_reserveavail(struct lfs *, struct vnode *vp, struct vnode *vp2, int);
121 1.51 yamt
122 1.51 yamt int
123 1.51 yamt lfs_fits_buf(struct lfs *fs, int n, int bytes)
124 1.51 yamt {
125 1.51 yamt int count_fit =
126 1.51 yamt (locked_queue_count + locked_queue_rcount + n < LFS_WAIT_BUFS);
127 1.51 yamt int bytes_fit =
128 1.51 yamt (locked_queue_bytes + locked_queue_rbytes + bytes < LFS_WAIT_BYTES);
129 1.51 yamt
130 1.51 yamt #ifdef DEBUG_LFS
131 1.51 yamt if (!count_fit) {
132 1.51 yamt printf("lfs_fits_buf: no fit count: %d + %d + %d >= %d\n",
133 1.51 yamt locked_queue_count, locked_queue_rcount,
134 1.51 yamt n, LFS_WAIT_BUFS);
135 1.51 yamt }
136 1.51 yamt if (!bytes_fit) {
137 1.51 yamt printf("lfs_fits_buf: no fit bytes: %ld + %ld + %d >= %d\n",
138 1.51 yamt locked_queue_bytes, locked_queue_rbytes,
139 1.51 yamt bytes, LFS_WAIT_BYTES);
140 1.51 yamt }
141 1.51 yamt #endif /* DEBUG_LFS */
142 1.51 yamt
143 1.51 yamt return (count_fit && bytes_fit);
144 1.51 yamt }
145 1.51 yamt
146 1.52 yamt /* ARGSUSED */
147 1.51 yamt int
148 1.52 yamt lfs_reservebuf(struct lfs *fs, struct vnode *vp, struct vnode *vp2,
149 1.52 yamt int n, int bytes)
150 1.51 yamt {
151 1.51 yamt KASSERT(locked_queue_rcount >= 0);
152 1.51 yamt KASSERT(locked_queue_rbytes >= 0);
153 1.51 yamt
154 1.51 yamt while (n > 0 && !lfs_fits_buf(fs, n, bytes)) {
155 1.51 yamt int error;
156 1.51 yamt
157 1.51 yamt ++fs->lfs_writer;
158 1.51 yamt lfs_flush(fs, 0);
159 1.51 yamt if (--fs->lfs_writer == 0)
160 1.51 yamt wakeup(&fs->lfs_dirops);
161 1.51 yamt
162 1.51 yamt error = tsleep(&locked_queue_count, PCATCH | PUSER,
163 1.51 yamt "lfsresbuf", hz * LFS_BUFWAIT);
164 1.51 yamt if (error && error != EWOULDBLOCK)
165 1.51 yamt return error;
166 1.51 yamt }
167 1.51 yamt
168 1.51 yamt locked_queue_rcount += n;
169 1.51 yamt locked_queue_rbytes += bytes;
170 1.51 yamt
171 1.51 yamt KASSERT(locked_queue_rcount >= 0);
172 1.51 yamt KASSERT(locked_queue_rbytes >= 0);
173 1.51 yamt
174 1.51 yamt return 0;
175 1.51 yamt }
176 1.51 yamt
177 1.51 yamt /*
178 1.27 perseant * Try to reserve some blocks, prior to performing a sensitive operation that
179 1.27 perseant * requires the vnode lock to be honored. If there is not enough space, give
180 1.27 perseant * up the vnode lock temporarily and wait for the space to become available.
181 1.27 perseant *
182 1.44 perseant * Called with vp locked. (Note nowever that if fsb < 0, vp is ignored.)
183 1.46 yamt *
184 1.47 yamt * XXX YAMT - it isn't safe to unlock vp here
185 1.47 yamt * because the node might be modified while we sleep.
186 1.47 yamt * (eg. cached states like i_offset might be stale,
187 1.47 yamt * the vnode might be truncated, etc..)
188 1.47 yamt * maybe we should have a way to restart the vnode op. (EVOPRESTART?)
189 1.27 perseant */
190 1.27 perseant int
191 1.52 yamt lfs_reserveavail(struct lfs *fs, struct vnode *vp, struct vnode *vp2, int fsb)
192 1.27 perseant {
193 1.27 perseant CLEANERINFO *cip;
194 1.27 perseant struct buf *bp;
195 1.27 perseant int error, slept;
196 1.27 perseant
197 1.27 perseant slept = 0;
198 1.45 yamt while (fsb > 0 && !lfs_fits(fs, fsb + fs->lfs_ravail) &&
199 1.45 yamt vp != fs->lfs_unlockvp) {
200 1.49 yamt #if 0
201 1.49 yamt /*
202 1.49 yamt * XXX ideally, we should unlock vnodes here
203 1.49 yamt * because we might sleep very long time.
204 1.49 yamt */
205 1.27 perseant VOP_UNLOCK(vp, 0);
206 1.52 yamt if (vp2 != NULL) {
207 1.52 yamt VOP_UNLOCK(vp2, 0);
208 1.52 yamt }
209 1.50 yamt #else
210 1.50 yamt /*
211 1.50 yamt * XXX since we'll sleep for cleaner with vnode lock holding,
212 1.50 yamt * deadlock will occur if cleaner wants to acquire the vnode
213 1.50 yamt * lock.
214 1.50 yamt * (eg. lfs_markv -> lfs_fastvget -> getnewvnode -> vclean)
215 1.50 yamt */
216 1.49 yamt #endif
217 1.27 perseant
218 1.30 perseant if (!slept) {
219 1.30 perseant #ifdef DEBUG
220 1.27 perseant printf("lfs_reserve: waiting for %ld (bfree = %d,"
221 1.30 perseant " est_bfree = %d)\n",
222 1.44 perseant fsb + fs->lfs_ravail, fs->lfs_bfree,
223 1.30 perseant LFS_EST_BFREE(fs));
224 1.30 perseant #endif
225 1.30 perseant }
226 1.27 perseant ++slept;
227 1.27 perseant
228 1.27 perseant /* Wake up the cleaner */
229 1.27 perseant LFS_CLEANERINFO(cip, fs, bp);
230 1.31 perseant LFS_SYNC_CLEANERINFO(cip, fs, bp, 0);
231 1.27 perseant wakeup(&lfs_allclean_wakeup);
232 1.27 perseant wakeup(&fs->lfs_nextseg);
233 1.27 perseant
234 1.27 perseant error = tsleep(&fs->lfs_avail, PCATCH | PUSER, "lfs_reserve",
235 1.27 perseant 0);
236 1.49 yamt #if 0
237 1.27 perseant vn_lock(vp, LK_EXCLUSIVE | LK_RETRY); /* XXX use lockstatus */
238 1.52 yamt vn_lock(vp2, LK_EXCLUSIVE | LK_RETRY); /* XXX use lockstatus */
239 1.49 yamt #endif
240 1.27 perseant if (error)
241 1.27 perseant return error;
242 1.27 perseant }
243 1.44 perseant #ifdef DEBUG
244 1.27 perseant if (slept)
245 1.27 perseant printf("lfs_reserve: woke up\n");
246 1.44 perseant #endif
247 1.44 perseant fs->lfs_ravail += fsb;
248 1.51 yamt
249 1.27 perseant return 0;
250 1.27 perseant }
251 1.27 perseant
252 1.51 yamt #ifdef DIAGNOSTIC
253 1.51 yamt int lfs_rescount;
254 1.51 yamt int lfs_rescountdirop;
255 1.51 yamt #endif
256 1.51 yamt
257 1.51 yamt int
258 1.52 yamt lfs_reserve(struct lfs *fs, struct vnode *vp, struct vnode *vp2, int fsb)
259 1.51 yamt {
260 1.51 yamt int error;
261 1.51 yamt int cantwait;
262 1.51 yamt
263 1.52 yamt KASSERT(fsb < 0 || VOP_ISLOCKED(vp));
264 1.52 yamt KASSERT(vp2 == NULL || fsb < 0 || VOP_ISLOCKED(vp2));
265 1.52 yamt
266 1.52 yamt /*
267 1.52 yamt * XXX
268 1.52 yamt * vref vnodes here so that cleaner doesn't try to reuse them.
269 1.52 yamt */
270 1.52 yamt lfs_vref(vp);
271 1.52 yamt if (vp2 != NULL) {
272 1.52 yamt lfs_vref(vp2);
273 1.52 yamt }
274 1.52 yamt
275 1.51 yamt cantwait = (VTOI(vp)->i_flag & IN_ADIROP);
276 1.51 yamt #ifdef DIAGNOSTIC
277 1.51 yamt if (cantwait) {
278 1.51 yamt if (fsb > 0)
279 1.51 yamt lfs_rescountdirop++;
280 1.51 yamt else if (fsb < 0)
281 1.51 yamt lfs_rescountdirop--;
282 1.51 yamt if (lfs_rescountdirop < 0)
283 1.51 yamt panic("lfs_rescountdirop");
284 1.51 yamt }
285 1.51 yamt else {
286 1.51 yamt if (fsb > 0)
287 1.51 yamt lfs_rescount++;
288 1.51 yamt else if (fsb < 0)
289 1.51 yamt lfs_rescount--;
290 1.51 yamt if (lfs_rescount < 0)
291 1.51 yamt panic("lfs_rescount");
292 1.51 yamt }
293 1.51 yamt #endif
294 1.51 yamt if (cantwait)
295 1.51 yamt return 0;
296 1.51 yamt
297 1.52 yamt error = lfs_reserveavail(fs, vp, vp2, fsb);
298 1.51 yamt if (error)
299 1.52 yamt goto done;
300 1.51 yamt
301 1.51 yamt /*
302 1.51 yamt * XXX just a guess. should be more precise.
303 1.51 yamt */
304 1.52 yamt error = lfs_reservebuf(fs, vp, vp2,
305 1.52 yamt fragstoblks(fs, fsb), fsbtob(fs, fsb));
306 1.51 yamt if (error)
307 1.52 yamt lfs_reserveavail(fs, vp, vp2, -fsb);
308 1.52 yamt
309 1.52 yamt done:
310 1.52 yamt lfs_vunref(vp);
311 1.52 yamt if (vp2 != NULL) {
312 1.52 yamt lfs_vunref(vp2);
313 1.52 yamt }
314 1.51 yamt
315 1.51 yamt return error;
316 1.51 yamt }
317 1.1 mycroft
318 1.1 mycroft int
319 1.36 perseant lfs_bwrite(void *v)
320 1.5 christos {
321 1.1 mycroft struct vop_bwrite_args /* {
322 1.1 mycroft struct buf *a_bp;
323 1.5 christos } */ *ap = v;
324 1.17 augustss struct buf *bp = ap->a_bp;
325 1.7 perseant
326 1.7 perseant #ifdef DIAGNOSTIC
327 1.23 perseant if (VTOI(bp->b_vp)->i_lfs->lfs_ronly == 0 && (bp->b_flags & B_ASYNC)) {
328 1.7 perseant panic("bawrite LFS buffer");
329 1.16 perseant }
330 1.7 perseant #endif /* DIAGNOSTIC */
331 1.7 perseant return lfs_bwrite_ext(bp,0);
332 1.7 perseant }
333 1.7 perseant
334 1.7 perseant /*
335 1.44 perseant * Determine if there is enough room currently available to write fsb
336 1.44 perseant * blocks. We need enough blocks for the new blocks, the current
337 1.44 perseant * inode blocks (including potentially the ifile inode), a summary block,
338 1.44 perseant * and the segment usage table, plus an ifile block.
339 1.7 perseant */
340 1.32 perseant int
341 1.36 perseant lfs_fits(struct lfs *fs, int fsb)
342 1.7 perseant {
343 1.26 perseant int needed;
344 1.26 perseant
345 1.36 perseant needed = fsb + btofsb(fs, fs->lfs_sumsize) +
346 1.44 perseant ((howmany(fs->lfs_uinodes + 1, INOPB(fs)) + fs->lfs_segtabsz +
347 1.44 perseant 1) << (fs->lfs_blktodb - fs->lfs_fsbtodb));
348 1.27 perseant
349 1.26 perseant if (needed >= fs->lfs_avail) {
350 1.27 perseant #ifdef DEBUG
351 1.36 perseant printf("lfs_fits: no fit: fsb = %d, uinodes = %d, "
352 1.26 perseant "needed = %d, avail = %d\n",
353 1.36 perseant fsb, fs->lfs_uinodes, needed, fs->lfs_avail);
354 1.26 perseant #endif
355 1.7 perseant return 0;
356 1.7 perseant }
357 1.7 perseant return 1;
358 1.7 perseant }
359 1.7 perseant
360 1.7 perseant int
361 1.44 perseant lfs_availwait(struct lfs *fs, int fsb)
362 1.32 perseant {
363 1.32 perseant int error;
364 1.32 perseant CLEANERINFO *cip;
365 1.32 perseant struct buf *cbp;
366 1.32 perseant
367 1.44 perseant while (!lfs_fits(fs, fsb)) {
368 1.32 perseant /*
369 1.32 perseant * Out of space, need cleaner to run.
370 1.32 perseant * Update the cleaner info, then wake it up.
371 1.32 perseant * Note the cleanerinfo block is on the ifile
372 1.32 perseant * so it CANT_WAIT.
373 1.32 perseant */
374 1.32 perseant LFS_CLEANERINFO(cip, fs, cbp);
375 1.32 perseant LFS_SYNC_CLEANERINFO(cip, fs, cbp, 0);
376 1.32 perseant
377 1.32 perseant printf("lfs_availwait: out of available space, "
378 1.32 perseant "waiting on cleaner\n");
379 1.32 perseant
380 1.32 perseant wakeup(&lfs_allclean_wakeup);
381 1.32 perseant wakeup(&fs->lfs_nextseg);
382 1.32 perseant #ifdef DIAGNOSTIC
383 1.32 perseant if (fs->lfs_seglock && fs->lfs_lockpid == curproc->p_pid)
384 1.32 perseant panic("lfs_availwait: deadlock");
385 1.32 perseant #endif
386 1.32 perseant error = tsleep(&fs->lfs_avail, PCATCH | PUSER, "cleaner", 0);
387 1.32 perseant if (error)
388 1.32 perseant return (error);
389 1.32 perseant }
390 1.32 perseant return 0;
391 1.32 perseant }
392 1.32 perseant
393 1.32 perseant int
394 1.36 perseant lfs_bwrite_ext(struct buf *bp, int flags)
395 1.7 perseant {
396 1.1 mycroft struct lfs *fs;
397 1.1 mycroft struct inode *ip;
398 1.51 yamt int fsb, s;
399 1.48 yamt
400 1.48 yamt KASSERT(bp->b_flags & B_BUSY);
401 1.48 yamt KASSERT(flags & BW_CLEAN || !(bp->b_flags & B_CALL));
402 1.48 yamt
403 1.1 mycroft /*
404 1.16 perseant * Don't write *any* blocks if we're mounted read-only.
405 1.16 perseant * In particular the cleaner can't write blocks either.
406 1.16 perseant */
407 1.40 chs if (VTOI(bp->b_vp)->i_lfs->lfs_ronly) {
408 1.32 perseant bp->b_flags &= ~(B_DELWRI | B_READ | B_ERROR);
409 1.32 perseant LFS_UNLOCK_BUF(bp);
410 1.40 chs if (bp->b_flags & B_CALL)
411 1.16 perseant bp->b_flags &= ~B_BUSY;
412 1.16 perseant else
413 1.16 perseant brelse(bp);
414 1.16 perseant return EROFS;
415 1.16 perseant }
416 1.16 perseant
417 1.16 perseant /*
418 1.1 mycroft * Set the delayed write flag and use reassignbuf to move the buffer
419 1.1 mycroft * from the clean list to the dirty one.
420 1.1 mycroft *
421 1.1 mycroft * Set the B_LOCKED flag and unlock the buffer, causing brelse to move
422 1.1 mycroft * the buffer onto the LOCKED free list. This is necessary, otherwise
423 1.1 mycroft * getnewbuf() would try to reclaim the buffers using bawrite, which
424 1.1 mycroft * isn't going to work.
425 1.1 mycroft *
426 1.1 mycroft * XXX we don't let meta-data writes run out of space because they can
427 1.1 mycroft * come from the segment writer. We need to make sure that there is
428 1.1 mycroft * enough space reserved so that there's room to write meta-data
429 1.1 mycroft * blocks.
430 1.1 mycroft */
431 1.1 mycroft if (!(bp->b_flags & B_LOCKED)) {
432 1.1 mycroft fs = VFSTOUFS(bp->b_vp->v_mount)->um_lfs;
433 1.36 perseant fsb = fragstofsb(fs, numfrags(fs, bp->b_bcount));
434 1.7 perseant
435 1.7 perseant ip = VTOI(bp->b_vp);
436 1.48 yamt if (flags & BW_CLEAN) {
437 1.26 perseant LFS_SET_UINO(ip, IN_CLEANING);
438 1.7 perseant } else {
439 1.36 perseant LFS_SET_UINO(ip, IN_MODIFIED);
440 1.36 perseant if (bp->b_lblkno >= 0)
441 1.36 perseant LFS_SET_UINO(ip, IN_UPDATE);
442 1.7 perseant }
443 1.36 perseant fs->lfs_avail -= fsb;
444 1.32 perseant bp->b_flags |= B_DELWRI;
445 1.32 perseant
446 1.32 perseant LFS_LOCK_BUF(bp);
447 1.43 perseant bp->b_flags &= ~(B_READ | B_DONE | B_ERROR);
448 1.14 fvdl s = splbio();
449 1.1 mycroft reassignbuf(bp, bp->b_vp);
450 1.1 mycroft splx(s);
451 1.1 mycroft }
452 1.7 perseant
453 1.40 chs if (bp->b_flags & B_CALL)
454 1.7 perseant bp->b_flags &= ~B_BUSY;
455 1.7 perseant else
456 1.7 perseant brelse(bp);
457 1.7 perseant
458 1.1 mycroft return (0);
459 1.1 mycroft }
460 1.1 mycroft
461 1.24 perseant void
462 1.36 perseant lfs_flush_fs(struct lfs *fs, int flags)
463 1.11 perseant {
464 1.40 chs if (fs->lfs_ronly == 0 && fs->lfs_dirops == 0)
465 1.11 perseant {
466 1.11 perseant /* disallow dirops during flush */
467 1.24 perseant fs->lfs_writer++;
468 1.11 perseant
469 1.11 perseant /*
470 1.11 perseant * We set the queue to 0 here because we
471 1.11 perseant * are about to write all the dirty
472 1.11 perseant * buffers we have. If more come in
473 1.11 perseant * while we're writing the segment, they
474 1.11 perseant * may not get written, so we want the
475 1.11 perseant * count to reflect these new writes
476 1.11 perseant * after the segwrite completes.
477 1.11 perseant */
478 1.40 chs if (lfs_dostats)
479 1.11 perseant ++lfs_stats.flush_invoked;
480 1.24 perseant lfs_segwrite(fs->lfs_ivnode->v_mount, flags);
481 1.11 perseant
482 1.11 perseant /* XXX KS - allow dirops again */
483 1.40 chs if (--fs->lfs_writer == 0)
484 1.24 perseant wakeup(&fs->lfs_dirops);
485 1.11 perseant }
486 1.11 perseant }
487 1.11 perseant
488 1.1 mycroft /*
489 1.1 mycroft * XXX
490 1.1 mycroft * This routine flushes buffers out of the B_LOCKED queue when LFS has too
491 1.1 mycroft * many locked down. Eventually the pageout daemon will simply call LFS
492 1.1 mycroft * when pages need to be reclaimed. Note, we have one static count of locked
493 1.1 mycroft * buffers, so we can't have more than a single file system. To make this
494 1.1 mycroft * work for multiple file systems, put the count into the mount structure.
495 1.1 mycroft */
496 1.1 mycroft void
497 1.36 perseant lfs_flush(struct lfs *fs, int flags)
498 1.1 mycroft {
499 1.17 augustss struct mount *mp, *nmp;
500 1.7 perseant
501 1.40 chs if (lfs_dostats)
502 1.7 perseant ++lfs_stats.write_exceeded;
503 1.40 chs if (lfs_writing && flags == 0) {/* XXX flags */
504 1.16 perseant #ifdef DEBUG_LFS
505 1.16 perseant printf("lfs_flush: not flushing because another flush is active\n");
506 1.16 perseant #endif
507 1.1 mycroft return;
508 1.16 perseant }
509 1.1 mycroft lfs_writing = 1;
510 1.7 perseant
511 1.6 fvdl simple_lock(&mountlist_slock);
512 1.6 fvdl for (mp = mountlist.cqh_first; mp != (void *)&mountlist; mp = nmp) {
513 1.6 fvdl if (vfs_busy(mp, LK_NOWAIT, &mountlist_slock)) {
514 1.6 fvdl nmp = mp->mnt_list.cqe_next;
515 1.6 fvdl continue;
516 1.6 fvdl }
517 1.40 chs if (strncmp(&mp->mnt_stat.f_fstypename[0], MOUNT_LFS, MFSNAMELEN) == 0)
518 1.24 perseant lfs_flush_fs(((struct ufsmount *)mp->mnt_data)->ufsmount_u.lfs, flags);
519 1.6 fvdl simple_lock(&mountlist_slock);
520 1.6 fvdl nmp = mp->mnt_list.cqe_next;
521 1.6 fvdl vfs_unbusy(mp);
522 1.1 mycroft }
523 1.6 fvdl simple_unlock(&mountlist_slock);
524 1.7 perseant
525 1.43 perseant LFS_DEBUG_COUNTLOCKED("flush");
526 1.7 perseant
527 1.1 mycroft lfs_writing = 0;
528 1.1 mycroft }
529 1.1 mycroft
530 1.41 perseant #define INOCOUNT(fs) howmany((fs)->lfs_uinodes, INOPB(fs))
531 1.41 perseant #define INOBYTES(fs) ((fs)->lfs_uinodes * DINODE_SIZE)
532 1.41 perseant
533 1.1 mycroft int
534 1.36 perseant lfs_check(struct vnode *vp, ufs_daddr_t blkno, int flags)
535 1.1 mycroft {
536 1.1 mycroft int error;
537 1.10 perseant struct lfs *fs;
538 1.20 perseant struct inode *ip;
539 1.13 perseant extern int lfs_dirvcount;
540 1.10 perseant
541 1.1 mycroft error = 0;
542 1.20 perseant ip = VTOI(vp);
543 1.7 perseant
544 1.7 perseant /* If out of buffers, wait on writer */
545 1.7 perseant /* XXX KS - if it's the Ifile, we're probably the cleaner! */
546 1.20 perseant if (ip->i_number == LFS_IFILE_INUM)
547 1.20 perseant return 0;
548 1.20 perseant /* If we're being called from inside a dirop, don't sleep */
549 1.20 perseant if (ip->i_flag & IN_ADIROP)
550 1.7 perseant return 0;
551 1.7 perseant
552 1.20 perseant fs = ip->i_lfs;
553 1.20 perseant
554 1.20 perseant /*
555 1.20 perseant * If we would flush below, but dirops are active, sleep.
556 1.20 perseant * Note that a dirop cannot ever reach this code!
557 1.20 perseant */
558 1.20 perseant while (fs->lfs_dirops > 0 &&
559 1.41 perseant (locked_queue_count + INOCOUNT(fs) > LFS_MAX_BUFS ||
560 1.41 perseant locked_queue_bytes + INOBYTES(fs) > LFS_MAX_BYTES ||
561 1.20 perseant lfs_dirvcount > LFS_MAXDIROP || fs->lfs_diropwait > 0))
562 1.20 perseant {
563 1.20 perseant ++fs->lfs_diropwait;
564 1.20 perseant tsleep(&fs->lfs_writer, PRIBIO+1, "bufdirop", 0);
565 1.20 perseant --fs->lfs_diropwait;
566 1.20 perseant }
567 1.10 perseant
568 1.41 perseant if (locked_queue_count + INOCOUNT(fs) > LFS_MAX_BUFS ||
569 1.41 perseant locked_queue_bytes + INOBYTES(fs) > LFS_MAX_BYTES ||
570 1.20 perseant lfs_dirvcount > LFS_MAXDIROP || fs->lfs_diropwait > 0)
571 1.7 perseant {
572 1.10 perseant ++fs->lfs_writer;
573 1.10 perseant lfs_flush(fs, flags);
574 1.40 chs if (--fs->lfs_writer == 0)
575 1.10 perseant wakeup(&fs->lfs_dirops);
576 1.7 perseant }
577 1.11 perseant
578 1.41 perseant while (locked_queue_count + INOCOUNT(fs) > LFS_WAIT_BUFS
579 1.41 perseant || locked_queue_bytes + INOBYTES(fs) > LFS_WAIT_BYTES)
580 1.7 perseant {
581 1.40 chs if (lfs_dostats)
582 1.7 perseant ++lfs_stats.wait_exceeded;
583 1.36 perseant #ifdef DEBUG_LFS
584 1.16 perseant printf("lfs_check: waiting: count=%d, bytes=%ld\n",
585 1.16 perseant locked_queue_count, locked_queue_bytes);
586 1.16 perseant #endif
587 1.7 perseant error = tsleep(&locked_queue_count, PCATCH | PUSER,
588 1.7 perseant "buffers", hz * LFS_BUFWAIT);
589 1.33 perseant if (error != EWOULDBLOCK)
590 1.33 perseant break;
591 1.18 perseant /*
592 1.18 perseant * lfs_flush might not flush all the buffers, if some of the
593 1.33 perseant * inodes were locked or if most of them were Ifile blocks
594 1.33 perseant * and we weren't asked to checkpoint. Try flushing again
595 1.33 perseant * to keep us from blocking indefinitely.
596 1.18 perseant */
597 1.41 perseant if (locked_queue_count + INOCOUNT(fs) > LFS_MAX_BUFS ||
598 1.41 perseant locked_queue_bytes + INOBYTES(fs) > LFS_MAX_BYTES)
599 1.18 perseant {
600 1.18 perseant ++fs->lfs_writer;
601 1.33 perseant lfs_flush(fs, flags | SEGM_CKP);
602 1.40 chs if (--fs->lfs_writer == 0)
603 1.18 perseant wakeup(&fs->lfs_dirops);
604 1.18 perseant }
605 1.7 perseant }
606 1.7 perseant return (error);
607 1.7 perseant }
608 1.1 mycroft
609 1.7 perseant /*
610 1.7 perseant * Allocate a new buffer header.
611 1.7 perseant */
612 1.34 perseant #ifdef MALLOCLOG
613 1.34 perseant # define DOMALLOC(S, T, F) _malloc((S), (T), (F), file, line)
614 1.34 perseant struct buf *
615 1.36 perseant lfs_newbuf_malloclog(struct lfs *fs, struct vnode *vp, ufs_daddr_t daddr, size_t size, char *file, int line)
616 1.34 perseant #else
617 1.35 perseant # define DOMALLOC(S, T, F) malloc((S), (T), (F))
618 1.7 perseant struct buf *
619 1.36 perseant lfs_newbuf(struct lfs *fs, struct vnode *vp, ufs_daddr_t daddr, size_t size)
620 1.34 perseant #endif
621 1.7 perseant {
622 1.7 perseant struct buf *bp;
623 1.7 perseant size_t nbytes;
624 1.7 perseant int s;
625 1.7 perseant
626 1.36 perseant nbytes = roundup(size, fsbtob(fs, 1));
627 1.7 perseant
628 1.34 perseant bp = DOMALLOC(sizeof(struct buf), M_SEGMENT, M_WAITOK);
629 1.7 perseant bzero(bp, sizeof(struct buf));
630 1.43 perseant if (nbytes) {
631 1.34 perseant bp->b_data = DOMALLOC(nbytes, M_SEGMENT, M_WAITOK);
632 1.7 perseant bzero(bp->b_data, nbytes);
633 1.7 perseant }
634 1.7 perseant #ifdef DIAGNOSTIC
635 1.40 chs if (vp == NULL)
636 1.7 perseant panic("vp is NULL in lfs_newbuf");
637 1.40 chs if (bp == NULL)
638 1.7 perseant panic("bp is NULL after malloc in lfs_newbuf");
639 1.1 mycroft #endif
640 1.7 perseant s = splbio();
641 1.7 perseant bgetvp(vp, bp);
642 1.7 perseant splx(s);
643 1.7 perseant
644 1.43 perseant bp->b_saveaddr = (caddr_t)fs;
645 1.7 perseant bp->b_bufsize = size;
646 1.7 perseant bp->b_bcount = size;
647 1.7 perseant bp->b_lblkno = daddr;
648 1.7 perseant bp->b_blkno = daddr;
649 1.7 perseant bp->b_error = 0;
650 1.7 perseant bp->b_resid = 0;
651 1.7 perseant bp->b_iodone = lfs_callback;
652 1.7 perseant bp->b_flags |= B_BUSY | B_CALL | B_NOCACHE;
653 1.7 perseant
654 1.7 perseant return (bp);
655 1.7 perseant }
656 1.7 perseant
657 1.34 perseant #ifdef MALLOCLOG
658 1.34 perseant # define DOFREE(A, T) _free((A), (T), file, line)
659 1.34 perseant void
660 1.36 perseant lfs_freebuf_malloclog(struct buf *bp, char *file, int line)
661 1.34 perseant #else
662 1.34 perseant # define DOFREE(A, T) free((A), (T))
663 1.7 perseant void
664 1.36 perseant lfs_freebuf(struct buf *bp)
665 1.34 perseant #endif
666 1.7 perseant {
667 1.7 perseant int s;
668 1.7 perseant
669 1.7 perseant s = splbio();
670 1.40 chs if (bp->b_vp)
671 1.7 perseant brelvp(bp);
672 1.7 perseant splx(s);
673 1.7 perseant if (!(bp->b_flags & B_INVAL)) { /* B_INVAL indicates a "fake" buffer */
674 1.34 perseant DOFREE(bp->b_data, M_SEGMENT);
675 1.7 perseant bp->b_data = NULL;
676 1.1 mycroft }
677 1.34 perseant DOFREE(bp, M_SEGMENT);
678 1.7 perseant }
679 1.1 mycroft
680 1.7 perseant /*
681 1.7 perseant * Definitions for the buffer free lists.
682 1.7 perseant */
683 1.7 perseant #define BQUEUES 4 /* number of free buffer queues */
684 1.7 perseant
685 1.7 perseant #define BQ_LOCKED 0 /* super-blocks &c */
686 1.7 perseant #define BQ_LRU 1 /* lru, useful buffers */
687 1.7 perseant #define BQ_AGE 2 /* rubbish */
688 1.7 perseant #define BQ_EMPTY 3 /* buffer headers with no memory */
689 1.7 perseant
690 1.7 perseant extern TAILQ_HEAD(bqueues, buf) bufqueues[BQUEUES];
691 1.7 perseant
692 1.7 perseant /*
693 1.7 perseant * Return a count of buffers on the "locked" queue.
694 1.27 perseant * Don't count malloced buffers, since they don't detract from the total.
695 1.7 perseant */
696 1.7 perseant void
697 1.43 perseant lfs_countlocked(int *count, long *bytes, char *msg)
698 1.7 perseant {
699 1.17 augustss struct buf *bp;
700 1.17 augustss int n = 0;
701 1.17 augustss long int size = 0L;
702 1.7 perseant
703 1.7 perseant for (bp = bufqueues[BQ_LOCKED].tqh_first; bp;
704 1.7 perseant bp = bp->b_freelist.tqe_next) {
705 1.27 perseant if (bp->b_flags & B_CALL) /* Malloced buffer */
706 1.27 perseant continue;
707 1.7 perseant n++;
708 1.7 perseant size += bp->b_bufsize;
709 1.27 perseant #ifdef DEBUG_LOCKED_LIST
710 1.27 perseant if (n > nbuf)
711 1.27 perseant panic("lfs_countlocked: this can't happen: more"
712 1.27 perseant " buffers locked than exist");
713 1.27 perseant #endif
714 1.7 perseant }
715 1.43 perseant #ifdef DEBUG_LOCKED_LIST
716 1.27 perseant /* Theoretically this function never really does anything */
717 1.27 perseant if (n != *count)
718 1.43 perseant printf("lfs_countlocked: %s: adjusted buf count from %d to %d\n",
719 1.43 perseant msg, *count, n);
720 1.27 perseant if (size != *bytes)
721 1.43 perseant printf("lfs_countlocked: %s: adjusted byte count from %ld to %ld\n",
722 1.43 perseant msg, *bytes, size);
723 1.27 perseant #endif
724 1.7 perseant *count = n;
725 1.7 perseant *bytes = size;
726 1.7 perseant return;
727 1.1 mycroft }
728