lfs_subr.c revision 1.91 1 1.90 maya /* $NetBSD: lfs_subr.c,v 1.91 2017/04/06 03:12:48 maya Exp $ */
2 1.2 cgd
3 1.8 perseant /*-
4 1.31 perseant * Copyright (c) 1999, 2000, 2001, 2002, 2003 The NetBSD Foundation, Inc.
5 1.8 perseant * All rights reserved.
6 1.8 perseant *
7 1.8 perseant * This code is derived from software contributed to The NetBSD Foundation
8 1.8 perseant * by Konrad E. Schroder <perseant (at) hhhh.org>.
9 1.8 perseant *
10 1.8 perseant * Redistribution and use in source and binary forms, with or without
11 1.8 perseant * modification, are permitted provided that the following conditions
12 1.8 perseant * are met:
13 1.8 perseant * 1. Redistributions of source code must retain the above copyright
14 1.8 perseant * notice, this list of conditions and the following disclaimer.
15 1.8 perseant * 2. Redistributions in binary form must reproduce the above copyright
16 1.8 perseant * notice, this list of conditions and the following disclaimer in the
17 1.8 perseant * documentation and/or other materials provided with the distribution.
18 1.8 perseant *
19 1.8 perseant * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
20 1.8 perseant * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
21 1.8 perseant * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
22 1.8 perseant * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
23 1.8 perseant * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
24 1.8 perseant * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
25 1.8 perseant * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
26 1.8 perseant * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
27 1.8 perseant * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
28 1.8 perseant * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
29 1.8 perseant * POSSIBILITY OF SUCH DAMAGE.
30 1.8 perseant */
31 1.1 mycroft /*
32 1.1 mycroft * Copyright (c) 1991, 1993
33 1.1 mycroft * The Regents of the University of California. All rights reserved.
34 1.1 mycroft *
35 1.1 mycroft * Redistribution and use in source and binary forms, with or without
36 1.1 mycroft * modification, are permitted provided that the following conditions
37 1.1 mycroft * are met:
38 1.1 mycroft * 1. Redistributions of source code must retain the above copyright
39 1.1 mycroft * notice, this list of conditions and the following disclaimer.
40 1.1 mycroft * 2. Redistributions in binary form must reproduce the above copyright
41 1.1 mycroft * notice, this list of conditions and the following disclaimer in the
42 1.1 mycroft * documentation and/or other materials provided with the distribution.
43 1.43 agc * 3. Neither the name of the University nor the names of its contributors
44 1.1 mycroft * may be used to endorse or promote products derived from this software
45 1.1 mycroft * without specific prior written permission.
46 1.1 mycroft *
47 1.1 mycroft * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
48 1.1 mycroft * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
49 1.1 mycroft * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
50 1.1 mycroft * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
51 1.1 mycroft * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
52 1.1 mycroft * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
53 1.1 mycroft * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
54 1.1 mycroft * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
55 1.1 mycroft * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
56 1.1 mycroft * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
57 1.1 mycroft * SUCH DAMAGE.
58 1.1 mycroft *
59 1.6 fvdl * @(#)lfs_subr.c 8.4 (Berkeley) 5/8/95
60 1.1 mycroft */
61 1.20 lukem
62 1.20 lukem #include <sys/cdefs.h>
63 1.90 maya __KERNEL_RCSID(0, "$NetBSD: lfs_subr.c,v 1.91 2017/04/06 03:12:48 maya Exp $");
64 1.1 mycroft
65 1.1 mycroft #include <sys/param.h>
66 1.3 christos #include <sys/systm.h>
67 1.1 mycroft #include <sys/namei.h>
68 1.1 mycroft #include <sys/vnode.h>
69 1.1 mycroft #include <sys/buf.h>
70 1.1 mycroft #include <sys/mount.h>
71 1.1 mycroft #include <sys/malloc.h>
72 1.1 mycroft #include <sys/proc.h>
73 1.61 perseant #include <sys/kauth.h>
74 1.1 mycroft
75 1.78 dholland #include <ufs/lfs/ulfs_inode.h>
76 1.1 mycroft #include <ufs/lfs/lfs.h>
77 1.84 dholland #include <ufs/lfs/lfs_accessors.h>
78 1.80 dholland #include <ufs/lfs/lfs_kernel.h>
79 1.1 mycroft #include <ufs/lfs/lfs_extern.h>
80 1.1 mycroft
81 1.31 perseant #include <uvm/uvm.h>
82 1.31 perseant
83 1.50 perseant #ifdef DEBUG
84 1.53 christos const char *lfs_res_names[LFS_NB_COUNT] = {
85 1.31 perseant "summary",
86 1.31 perseant "superblock",
87 1.52 perseant "file block",
88 1.31 perseant "cluster",
89 1.31 perseant "clean",
90 1.52 perseant "blkiov",
91 1.31 perseant };
92 1.31 perseant #endif
93 1.31 perseant
94 1.31 perseant int lfs_res_qty[LFS_NB_COUNT] = {
95 1.31 perseant LFS_N_SUMMARIES,
96 1.31 perseant LFS_N_SBLOCKS,
97 1.31 perseant LFS_N_IBLOCKS,
98 1.31 perseant LFS_N_CLUSTERS,
99 1.31 perseant LFS_N_CLEAN,
100 1.52 perseant LFS_N_BLKIOV,
101 1.31 perseant };
102 1.31 perseant
103 1.31 perseant void
104 1.31 perseant lfs_setup_resblks(struct lfs *fs)
105 1.31 perseant {
106 1.31 perseant int i, j;
107 1.31 perseant int maxbpp;
108 1.31 perseant
109 1.51 perseant ASSERT_NO_SEGLOCK(fs);
110 1.81 dholland fs->lfs_resblk = malloc(LFS_N_TOTAL * sizeof(res_t), M_SEGMENT,
111 1.81 dholland M_WAITOK);
112 1.31 perseant for (i = 0; i < LFS_N_TOTAL; i++) {
113 1.31 perseant fs->lfs_resblk[i].inuse = 0;
114 1.31 perseant fs->lfs_resblk[i].p = NULL;
115 1.31 perseant }
116 1.31 perseant for (i = 0; i < LFS_RESHASH_WIDTH; i++)
117 1.31 perseant LIST_INIT(fs->lfs_reshash + i);
118 1.31 perseant
119 1.31 perseant /*
120 1.31 perseant * These types of allocations can be larger than a page,
121 1.31 perseant * so we can't use the pool subsystem for them.
122 1.31 perseant */
123 1.31 perseant for (i = 0, j = 0; j < LFS_N_SUMMARIES; j++, i++)
124 1.83 dholland fs->lfs_resblk[i].size = lfs_sb_getsumsize(fs);
125 1.31 perseant for (j = 0; j < LFS_N_SBLOCKS; j++, i++)
126 1.34 perseant fs->lfs_resblk[i].size = LFS_SBPAD;
127 1.31 perseant for (j = 0; j < LFS_N_IBLOCKS; j++, i++)
128 1.82 dholland fs->lfs_resblk[i].size = lfs_sb_getbsize(fs);
129 1.31 perseant for (j = 0; j < LFS_N_CLUSTERS; j++, i++)
130 1.34 perseant fs->lfs_resblk[i].size = MAXPHYS;
131 1.31 perseant for (j = 0; j < LFS_N_CLEAN; j++, i++)
132 1.34 perseant fs->lfs_resblk[i].size = MAXPHYS;
133 1.52 perseant for (j = 0; j < LFS_N_BLKIOV; j++, i++)
134 1.52 perseant fs->lfs_resblk[i].size = LFS_MARKV_MAXBLKCNT * sizeof(BLOCK_INFO);
135 1.34 perseant
136 1.34 perseant for (i = 0; i < LFS_N_TOTAL; i++) {
137 1.34 perseant fs->lfs_resblk[i].p = malloc(fs->lfs_resblk[i].size,
138 1.34 perseant M_SEGMENT, M_WAITOK);
139 1.34 perseant }
140 1.31 perseant
141 1.31 perseant /*
142 1.31 perseant * Initialize pools for small types (XXX is BPP small?)
143 1.31 perseant */
144 1.46 simonb pool_init(&fs->lfs_clpool, sizeof(struct lfs_cluster), 0, 0, 0,
145 1.68 ad "lfsclpl", &pool_allocator_nointr, IPL_NONE);
146 1.46 simonb pool_init(&fs->lfs_segpool, sizeof(struct segment), 0, 0, 0,
147 1.68 ad "lfssegpool", &pool_allocator_nointr, IPL_NONE);
148 1.86 dholland /* XXX: should this int32 be 32/64? */
149 1.83 dholland maxbpp = ((lfs_sb_getsumsize(fs) - SEGSUM_SIZE(fs)) / sizeof(int32_t) + 2);
150 1.82 dholland maxbpp = MIN(maxbpp, lfs_segsize(fs) / lfs_sb_getfsize(fs) + 2);
151 1.46 simonb pool_init(&fs->lfs_bpppool, maxbpp * sizeof(struct buf *), 0, 0, 0,
152 1.68 ad "lfsbpppl", &pool_allocator_nointr, IPL_NONE);
153 1.31 perseant }
154 1.31 perseant
155 1.31 perseant void
156 1.31 perseant lfs_free_resblks(struct lfs *fs)
157 1.31 perseant {
158 1.31 perseant int i;
159 1.31 perseant
160 1.31 perseant pool_destroy(&fs->lfs_bpppool);
161 1.31 perseant pool_destroy(&fs->lfs_segpool);
162 1.31 perseant pool_destroy(&fs->lfs_clpool);
163 1.31 perseant
164 1.72 ad mutex_enter(&lfs_lock);
165 1.31 perseant for (i = 0; i < LFS_N_TOTAL; i++) {
166 1.39 perseant while (fs->lfs_resblk[i].inuse)
167 1.72 ad mtsleep(&fs->lfs_resblk, PRIBIO + 1, "lfs_free", 0,
168 1.72 ad &lfs_lock);
169 1.31 perseant if (fs->lfs_resblk[i].p != NULL)
170 1.31 perseant free(fs->lfs_resblk[i].p, M_SEGMENT);
171 1.31 perseant }
172 1.31 perseant free(fs->lfs_resblk, M_SEGMENT);
173 1.72 ad mutex_exit(&lfs_lock);
174 1.31 perseant }
175 1.31 perseant
176 1.31 perseant static unsigned int
177 1.31 perseant lfs_mhash(void *vp)
178 1.31 perseant {
179 1.31 perseant return (unsigned int)(((unsigned long)vp) >> 2) % LFS_RESHASH_WIDTH;
180 1.31 perseant }
181 1.31 perseant
182 1.31 perseant /*
183 1.31 perseant * Return memory of the given size for the given purpose, or use one of a
184 1.31 perseant * number of spare last-resort buffers, if malloc returns NULL.
185 1.49 perry */
186 1.31 perseant void *
187 1.31 perseant lfs_malloc(struct lfs *fs, size_t size, int type)
188 1.31 perseant {
189 1.31 perseant struct lfs_res_blk *re;
190 1.31 perseant void *r;
191 1.91 maya int i, start;
192 1.31 perseant unsigned int h;
193 1.31 perseant
194 1.51 perseant ASSERT_MAYBE_SEGLOCK(fs);
195 1.34 perseant r = NULL;
196 1.34 perseant
197 1.31 perseant /* If no mem allocated for this type, it just waits */
198 1.34 perseant if (lfs_res_qty[type] == 0) {
199 1.34 perseant r = malloc(size, M_SEGMENT, M_WAITOK);
200 1.34 perseant return r;
201 1.34 perseant }
202 1.31 perseant
203 1.31 perseant /* Otherwise try a quick malloc, and if it works, great */
204 1.34 perseant if ((r = malloc(size, M_SEGMENT, M_NOWAIT)) != NULL) {
205 1.31 perseant return r;
206 1.34 perseant }
207 1.31 perseant
208 1.31 perseant /*
209 1.31 perseant * If malloc returned NULL, we are forced to use one of our
210 1.31 perseant * reserve blocks. We have on hand at least one summary block,
211 1.31 perseant * at least one cluster block, at least one superblock,
212 1.31 perseant * and several indirect blocks.
213 1.31 perseant */
214 1.51 perseant
215 1.72 ad mutex_enter(&lfs_lock);
216 1.31 perseant /* skip over blocks of other types */
217 1.31 perseant for (i = 0, start = 0; i < type; i++)
218 1.31 perseant start += lfs_res_qty[i];
219 1.31 perseant while (r == NULL) {
220 1.31 perseant for (i = 0; i < lfs_res_qty[type]; i++) {
221 1.31 perseant if (fs->lfs_resblk[start + i].inuse == 0) {
222 1.31 perseant re = fs->lfs_resblk + start + i;
223 1.31 perseant re->inuse = 1;
224 1.31 perseant r = re->p;
225 1.34 perseant KASSERT(re->size >= size);
226 1.31 perseant h = lfs_mhash(r);
227 1.31 perseant LIST_INSERT_HEAD(&fs->lfs_reshash[h], re, res);
228 1.72 ad mutex_exit(&lfs_lock);
229 1.31 perseant return r;
230 1.31 perseant }
231 1.31 perseant }
232 1.51 perseant DLOG((DLOG_MALLOC, "sleeping on %s (%d)\n",
233 1.51 perseant lfs_res_names[type], lfs_res_qty[type]));
234 1.72 ad mtsleep(&fs->lfs_resblk, PVM, "lfs_malloc", 0,
235 1.72 ad &lfs_lock);
236 1.51 perseant DLOG((DLOG_MALLOC, "done sleeping on %s\n",
237 1.51 perseant lfs_res_names[type]));
238 1.31 perseant }
239 1.31 perseant /* NOTREACHED */
240 1.72 ad mutex_exit(&lfs_lock);
241 1.31 perseant return r;
242 1.31 perseant }
243 1.31 perseant
244 1.31 perseant void
245 1.65 christos lfs_free(struct lfs *fs, void *p, int type)
246 1.31 perseant {
247 1.31 perseant unsigned int h;
248 1.31 perseant res_t *re;
249 1.31 perseant
250 1.51 perseant ASSERT_MAYBE_SEGLOCK(fs);
251 1.31 perseant h = lfs_mhash(p);
252 1.72 ad mutex_enter(&lfs_lock);
253 1.31 perseant LIST_FOREACH(re, &fs->lfs_reshash[h], res) {
254 1.31 perseant if (re->p == p) {
255 1.32 yamt KASSERT(re->inuse == 1);
256 1.31 perseant LIST_REMOVE(re, res);
257 1.31 perseant re->inuse = 0;
258 1.31 perseant wakeup(&fs->lfs_resblk);
259 1.72 ad mutex_exit(&lfs_lock);
260 1.31 perseant return;
261 1.31 perseant }
262 1.31 perseant }
263 1.90 maya
264 1.90 maya for (int i = 0; i < LFS_N_TOTAL; i++) {
265 1.91 maya KDASSERTMSG(fs->lfs_resblk[i].p == p,
266 1.90 maya "lfs_free: inconsistent reserved block");
267 1.32 yamt }
268 1.90 maya
269 1.72 ad mutex_exit(&lfs_lock);
270 1.90 maya
271 1.31 perseant /*
272 1.31 perseant * If we didn't find it, free it.
273 1.31 perseant */
274 1.31 perseant free(p, M_SEGMENT);
275 1.31 perseant }
276 1.1 mycroft
277 1.1 mycroft /*
278 1.1 mycroft * lfs_seglock --
279 1.1 mycroft * Single thread the segment writer.
280 1.1 mycroft */
281 1.31 perseant int
282 1.18 perseant lfs_seglock(struct lfs *fs, unsigned long flags)
283 1.1 mycroft {
284 1.1 mycroft struct segment *sp;
285 1.49 perry
286 1.72 ad mutex_enter(&lfs_lock);
287 1.7 thorpej if (fs->lfs_seglock) {
288 1.58 perseant if (fs->lfs_lockpid == curproc->p_pid &&
289 1.58 perseant fs->lfs_locklwp == curlwp->l_lid) {
290 1.1 mycroft ++fs->lfs_seglock;
291 1.1 mycroft fs->lfs_sp->seg_flags |= flags;
292 1.74 mlelstv mutex_exit(&lfs_lock);
293 1.31 perseant return 0;
294 1.38 perseant } else if (flags & SEGM_PAGEDAEMON) {
295 1.72 ad mutex_exit(&lfs_lock);
296 1.31 perseant return EWOULDBLOCK;
297 1.51 perseant } else {
298 1.51 perseant while (fs->lfs_seglock) {
299 1.72 ad (void)mtsleep(&fs->lfs_seglock, PRIBIO + 1,
300 1.72 ad "lfs_seglock", 0, &lfs_lock);
301 1.51 perseant }
302 1.51 perseant }
303 1.7 thorpej }
304 1.49 perry
305 1.1 mycroft fs->lfs_seglock = 1;
306 1.1 mycroft fs->lfs_lockpid = curproc->p_pid;
307 1.58 perseant fs->lfs_locklwp = curlwp->l_lid;
308 1.72 ad mutex_exit(&lfs_lock);
309 1.36 perseant fs->lfs_cleanind = 0;
310 1.36 perseant
311 1.51 perseant LFS_ENTER_LOG("seglock", __FILE__, __LINE__, 0, flags, curproc->p_pid);
312 1.89 maya
313 1.27 perseant /* Drain fragment size changes out */
314 1.66 ad rw_enter(&fs->lfs_fraglock, RW_WRITER);
315 1.27 perseant
316 1.31 perseant sp = fs->lfs_sp = pool_get(&fs->lfs_segpool, PR_WAITOK);
317 1.31 perseant sp->bpp = pool_get(&fs->lfs_bpppool, PR_WAITOK);
318 1.1 mycroft sp->seg_flags = flags;
319 1.1 mycroft sp->vp = NULL;
320 1.26 perseant sp->seg_iocount = 0;
321 1.1 mycroft (void) lfs_initseg(fs);
322 1.49 perry
323 1.1 mycroft /*
324 1.1 mycroft * Keep a cumulative count of the outstanding I/O operations. If the
325 1.1 mycroft * disk drive catches up with us it could go to zero before we finish,
326 1.1 mycroft * so we artificially increment it by one until we've scheduled all of
327 1.1 mycroft * the writes we intend to do.
328 1.1 mycroft */
329 1.72 ad mutex_enter(&lfs_lock);
330 1.1 mycroft ++fs->lfs_iocount;
331 1.82 dholland fs->lfs_startseg = lfs_sb_getcurseg(fs);
332 1.72 ad mutex_exit(&lfs_lock);
333 1.31 perseant return 0;
334 1.1 mycroft }
335 1.8 perseant
336 1.31 perseant static void lfs_unmark_dirop(struct lfs *);
337 1.31 perseant
338 1.31 perseant static void
339 1.31 perseant lfs_unmark_dirop(struct lfs *fs)
340 1.31 perseant {
341 1.31 perseant struct inode *ip, *nip;
342 1.31 perseant struct vnode *vp;
343 1.40 perseant int doit;
344 1.31 perseant
345 1.51 perseant ASSERT_NO_SEGLOCK(fs);
346 1.72 ad mutex_enter(&lfs_lock);
347 1.40 perseant doit = !(fs->lfs_flags & LFS_UNDIROP);
348 1.40 perseant if (doit)
349 1.40 perseant fs->lfs_flags |= LFS_UNDIROP;
350 1.51 perseant if (!doit) {
351 1.72 ad mutex_exit(&lfs_lock);
352 1.40 perseant return;
353 1.51 perseant }
354 1.40 perseant
355 1.31 perseant for (ip = TAILQ_FIRST(&fs->lfs_dchainhd); ip != NULL; ip = nip) {
356 1.31 perseant nip = TAILQ_NEXT(ip, i_lfs_dchain);
357 1.31 perseant vp = ITOV(ip);
358 1.77 perseant if ((ip->i_flag & (IN_ADIROP | IN_CDIROP)) == IN_CDIROP) {
359 1.31 perseant --lfs_dirvcount;
360 1.59 perseant --fs->lfs_dirvcount;
361 1.71 ad vp->v_uflag &= ~VU_DIROP;
362 1.31 perseant TAILQ_REMOVE(&fs->lfs_dchainhd, ip, i_lfs_dchain);
363 1.31 perseant wakeup(&lfs_dirvcount);
364 1.31 perseant fs->lfs_unlockvp = vp;
365 1.72 ad mutex_exit(&lfs_lock);
366 1.76 hannken vrele(vp);
367 1.72 ad mutex_enter(&lfs_lock);
368 1.31 perseant fs->lfs_unlockvp = NULL;
369 1.77 perseant ip->i_flag &= ~IN_CDIROP;
370 1.72 ad }
371 1.31 perseant }
372 1.40 perseant
373 1.40 perseant fs->lfs_flags &= ~LFS_UNDIROP;
374 1.51 perseant wakeup(&fs->lfs_flags);
375 1.72 ad mutex_exit(&lfs_lock);
376 1.31 perseant }
377 1.31 perseant
378 1.31 perseant static void
379 1.31 perseant lfs_auto_segclean(struct lfs *fs)
380 1.31 perseant {
381 1.48 perseant int i, error, s, waited;
382 1.31 perseant
383 1.51 perseant ASSERT_SEGLOCK(fs);
384 1.31 perseant /*
385 1.31 perseant * Now that we've swapped lfs_activesb, but while we still
386 1.31 perseant * hold the segment lock, run through the segment list marking
387 1.31 perseant * the empty ones clean.
388 1.31 perseant * XXX - do we really need to do them all at once?
389 1.31 perseant */
390 1.48 perseant waited = 0;
391 1.83 dholland for (i = 0; i < lfs_sb_getnseg(fs); i++) {
392 1.31 perseant if ((fs->lfs_suflags[0][i] &
393 1.31 perseant (SEGUSE_ACTIVE | SEGUSE_DIRTY | SEGUSE_EMPTY)) ==
394 1.31 perseant (SEGUSE_DIRTY | SEGUSE_EMPTY) &&
395 1.31 perseant (fs->lfs_suflags[1][i] &
396 1.31 perseant (SEGUSE_ACTIVE | SEGUSE_DIRTY | SEGUSE_EMPTY)) ==
397 1.31 perseant (SEGUSE_DIRTY | SEGUSE_EMPTY)) {
398 1.31 perseant
399 1.48 perseant /* Make sure the sb is written before we clean */
400 1.72 ad mutex_enter(&lfs_lock);
401 1.48 perseant s = splbio();
402 1.48 perseant while (waited == 0 && fs->lfs_sbactive)
403 1.72 ad mtsleep(&fs->lfs_sbactive, PRIBIO+1, "lfs asb",
404 1.72 ad 0, &lfs_lock);
405 1.48 perseant splx(s);
406 1.72 ad mutex_exit(&lfs_lock);
407 1.48 perseant waited = 1;
408 1.48 perseant
409 1.31 perseant if ((error = lfs_do_segclean(fs, i)) != 0) {
410 1.50 perseant DLOG((DLOG_CLEAN, "lfs_auto_segclean: lfs_do_segclean returned %d for seg %d\n", error, i));
411 1.31 perseant }
412 1.31 perseant }
413 1.31 perseant fs->lfs_suflags[1 - fs->lfs_activesb][i] =
414 1.31 perseant fs->lfs_suflags[fs->lfs_activesb][i];
415 1.31 perseant }
416 1.31 perseant }
417 1.31 perseant
418 1.1 mycroft /*
419 1.1 mycroft * lfs_segunlock --
420 1.1 mycroft * Single thread the segment writer.
421 1.1 mycroft */
422 1.1 mycroft void
423 1.18 perseant lfs_segunlock(struct lfs *fs)
424 1.1 mycroft {
425 1.1 mycroft struct segment *sp;
426 1.1 mycroft unsigned long sync, ckp;
427 1.22 perseant struct buf *bp;
428 1.40 perseant int do_unmark_dirop = 0;
429 1.49 perry
430 1.15 perseant sp = fs->lfs_sp;
431 1.15 perseant
432 1.72 ad mutex_enter(&lfs_lock);
433 1.72 ad KASSERT(LFS_SEGLOCK_HELD(fs));
434 1.15 perseant if (fs->lfs_seglock == 1) {
435 1.77 perseant if ((sp->seg_flags & (SEGM_PROT | SEGM_CLEAN)) == 0)
436 1.40 perseant do_unmark_dirop = 1;
437 1.72 ad mutex_exit(&lfs_lock);
438 1.1 mycroft sync = sp->seg_flags & SEGM_SYNC;
439 1.1 mycroft ckp = sp->seg_flags & SEGM_CKP;
440 1.57 perseant
441 1.57 perseant /* We should have a segment summary, and nothing else */
442 1.57 perseant KASSERT(sp->cbpp == sp->bpp + 1);
443 1.57 perseant
444 1.57 perseant /* Free allocated segment summary */
445 1.82 dholland lfs_sb_suboffset(fs, lfs_btofsb(fs, lfs_sb_getsumsize(fs)));
446 1.57 perseant bp = *sp->bpp;
447 1.57 perseant lfs_freebuf(fs, bp);
448 1.8 perseant
449 1.31 perseant pool_put(&fs->lfs_bpppool, sp->bpp);
450 1.18 perseant sp->bpp = NULL;
451 1.36 perseant
452 1.36 perseant /*
453 1.36 perseant * If we're not sync, we're done with sp, get rid of it.
454 1.36 perseant * Otherwise, we keep a local copy around but free
455 1.36 perseant * fs->lfs_sp so another process can use it (we have to
456 1.36 perseant * wait but they don't have to wait for us).
457 1.36 perseant */
458 1.26 perseant if (!sync)
459 1.31 perseant pool_put(&fs->lfs_segpool, sp);
460 1.18 perseant fs->lfs_sp = NULL;
461 1.1 mycroft
462 1.1 mycroft /*
463 1.1 mycroft * If the I/O count is non-zero, sleep until it reaches zero.
464 1.1 mycroft * At the moment, the user's process hangs around so we can
465 1.1 mycroft * sleep.
466 1.1 mycroft */
467 1.72 ad mutex_enter(&lfs_lock);
468 1.63 christos if (--fs->lfs_iocount == 0) {
469 1.44 yamt LFS_DEBUG_COUNTLOCKED("lfs_segunlock");
470 1.63 christos }
471 1.36 perseant if (fs->lfs_iocount <= 1)
472 1.22 perseant wakeup(&fs->lfs_iocount);
473 1.72 ad mutex_exit(&lfs_lock);
474 1.1 mycroft /*
475 1.26 perseant * If we're not checkpointing, we don't have to block
476 1.26 perseant * other processes to wait for a synchronous write
477 1.26 perseant * to complete.
478 1.26 perseant */
479 1.26 perseant if (!ckp) {
480 1.51 perseant LFS_ENTER_LOG("segunlock_std", __FILE__, __LINE__, 0, 0, curproc->p_pid);
481 1.89 maya
482 1.72 ad mutex_enter(&lfs_lock);
483 1.26 perseant --fs->lfs_seglock;
484 1.26 perseant fs->lfs_lockpid = 0;
485 1.58 perseant fs->lfs_locklwp = 0;
486 1.72 ad mutex_exit(&lfs_lock);
487 1.26 perseant wakeup(&fs->lfs_seglock);
488 1.26 perseant }
489 1.26 perseant /*
490 1.1 mycroft * We let checkpoints happen asynchronously. That means
491 1.1 mycroft * that during recovery, we have to roll forward between
492 1.1 mycroft * the two segments described by the first and second
493 1.1 mycroft * superblocks to make sure that the checkpoint described
494 1.1 mycroft * by a superblock completed.
495 1.1 mycroft */
496 1.72 ad mutex_enter(&lfs_lock);
497 1.74 mlelstv while (ckp && sync && fs->lfs_iocount) {
498 1.72 ad (void)mtsleep(&fs->lfs_iocount, PRIBIO + 1,
499 1.72 ad "lfs_iocount", 0, &lfs_lock);
500 1.74 mlelstv DLOG((DLOG_SEG, "sleeping on iocount %x == %d\n", fs, fs->lfs_iocount));
501 1.74 mlelstv }
502 1.26 perseant while (sync && sp->seg_iocount) {
503 1.72 ad (void)mtsleep(&sp->seg_iocount, PRIBIO + 1,
504 1.72 ad "seg_iocount", 0, &lfs_lock);
505 1.50 perseant DLOG((DLOG_SEG, "sleeping on iocount %x == %d\n", sp, sp->seg_iocount));
506 1.26 perseant }
507 1.72 ad mutex_exit(&lfs_lock);
508 1.26 perseant if (sync)
509 1.31 perseant pool_put(&fs->lfs_segpool, sp);
510 1.36 perseant
511 1.1 mycroft if (ckp) {
512 1.1 mycroft fs->lfs_nactive = 0;
513 1.8 perseant /* If we *know* everything's on disk, write both sbs */
514 1.33 perseant /* XXX should wait for this one */
515 1.21 chs if (sync)
516 1.83 dholland lfs_writesuper(fs, lfs_sb_getsboff(fs, fs->lfs_activesb));
517 1.83 dholland lfs_writesuper(fs, lfs_sb_getsboff(fs, 1 - fs->lfs_activesb));
518 1.48 perseant if (!(fs->lfs_ivnode->v_mount->mnt_iflag & IMNT_UNMOUNT)) {
519 1.35 perseant lfs_auto_segclean(fs);
520 1.48 perseant /* If sync, we can clean the remainder too */
521 1.48 perseant if (sync)
522 1.48 perseant lfs_auto_segclean(fs);
523 1.48 perseant }
524 1.8 perseant fs->lfs_activesb = 1 - fs->lfs_activesb;
525 1.89 maya
526 1.51 perseant LFS_ENTER_LOG("segunlock_ckp", __FILE__, __LINE__, 0, 0, curproc->p_pid);
527 1.89 maya
528 1.72 ad mutex_enter(&lfs_lock);
529 1.26 perseant --fs->lfs_seglock;
530 1.26 perseant fs->lfs_lockpid = 0;
531 1.58 perseant fs->lfs_locklwp = 0;
532 1.72 ad mutex_exit(&lfs_lock);
533 1.26 perseant wakeup(&fs->lfs_seglock);
534 1.1 mycroft }
535 1.27 perseant /* Reenable fragment size changes */
536 1.66 ad rw_exit(&fs->lfs_fraglock);
537 1.40 perseant if (do_unmark_dirop)
538 1.40 perseant lfs_unmark_dirop(fs);
539 1.1 mycroft } else if (fs->lfs_seglock == 0) {
540 1.72 ad mutex_exit(&lfs_lock);
541 1.1 mycroft panic ("Seglock not held");
542 1.1 mycroft } else {
543 1.1 mycroft --fs->lfs_seglock;
544 1.72 ad mutex_exit(&lfs_lock);
545 1.1 mycroft }
546 1.41 yamt }
547 1.41 yamt
548 1.41 yamt /*
549 1.69 perseant * Drain dirops and start writer.
550 1.69 perseant *
551 1.69 perseant * No simple_locks are held when we enter and none are held when we return.
552 1.41 yamt */
553 1.41 yamt int
554 1.41 yamt lfs_writer_enter(struct lfs *fs, const char *wmesg)
555 1.41 yamt {
556 1.41 yamt int error = 0;
557 1.41 yamt
558 1.51 perseant ASSERT_MAYBE_SEGLOCK(fs);
559 1.72 ad mutex_enter(&lfs_lock);
560 1.41 yamt
561 1.41 yamt /* disallow dirops during flush */
562 1.41 yamt fs->lfs_writer++;
563 1.41 yamt
564 1.41 yamt while (fs->lfs_dirops > 0) {
565 1.49 perry ++fs->lfs_diropwait;
566 1.72 ad error = mtsleep(&fs->lfs_writer, PRIBIO+1, wmesg, 0,
567 1.72 ad &lfs_lock);
568 1.49 perry --fs->lfs_diropwait;
569 1.41 yamt }
570 1.41 yamt
571 1.41 yamt if (error)
572 1.41 yamt fs->lfs_writer--;
573 1.41 yamt
574 1.72 ad mutex_exit(&lfs_lock);
575 1.41 yamt
576 1.41 yamt return error;
577 1.41 yamt }
578 1.41 yamt
579 1.41 yamt void
580 1.41 yamt lfs_writer_leave(struct lfs *fs)
581 1.41 yamt {
582 1.67 thorpej bool dowakeup;
583 1.41 yamt
584 1.51 perseant ASSERT_MAYBE_SEGLOCK(fs);
585 1.72 ad mutex_enter(&lfs_lock);
586 1.41 yamt dowakeup = !(--fs->lfs_writer);
587 1.41 yamt if (dowakeup)
588 1.87 maya cv_broadcast(&fs->lfs_diropscv);
589 1.88 maya mutex_exit(&lfs_lock);
590 1.1 mycroft }
591 1.57 perseant
592 1.57 perseant /*
593 1.57 perseant * Unlock, wait for the cleaner, then relock to where we were before.
594 1.57 perseant * To be used only at a fairly high level, to address a paucity of free
595 1.57 perseant * segments propagated back from lfs_gop_write().
596 1.57 perseant */
597 1.57 perseant void
598 1.57 perseant lfs_segunlock_relock(struct lfs *fs)
599 1.57 perseant {
600 1.57 perseant int n = fs->lfs_seglock;
601 1.57 perseant u_int16_t seg_flags;
602 1.61 perseant CLEANERINFO *cip;
603 1.61 perseant struct buf *bp;
604 1.57 perseant
605 1.57 perseant if (n == 0)
606 1.57 perseant return;
607 1.57 perseant
608 1.57 perseant /* Write anything we've already gathered to disk */
609 1.57 perseant lfs_writeseg(fs, fs->lfs_sp);
610 1.57 perseant
611 1.61 perseant /* Tell cleaner */
612 1.61 perseant LFS_CLEANERINFO(cip, fs, bp);
613 1.85 dholland lfs_ci_setflags(fs, cip,
614 1.85 dholland lfs_ci_getflags(fs, cip) | LFS_CLEANER_MUST_CLEAN);
615 1.61 perseant LFS_SYNC_CLEANERINFO(cip, fs, bp, 1);
616 1.61 perseant
617 1.57 perseant /* Save segment flags for later */
618 1.57 perseant seg_flags = fs->lfs_sp->seg_flags;
619 1.57 perseant
620 1.57 perseant fs->lfs_sp->seg_flags |= SEGM_PROT; /* Don't unmark dirop nodes */
621 1.57 perseant while(fs->lfs_seglock)
622 1.57 perseant lfs_segunlock(fs);
623 1.57 perseant
624 1.57 perseant /* Wait for the cleaner */
625 1.60 perseant lfs_wakeup_cleaner(fs);
626 1.72 ad mutex_enter(&lfs_lock);
627 1.57 perseant while (LFS_STARVED_FOR_SEGS(fs))
628 1.82 dholland mtsleep(&fs->lfs_availsleep, PRIBIO, "relock", 0,
629 1.72 ad &lfs_lock);
630 1.72 ad mutex_exit(&lfs_lock);
631 1.57 perseant
632 1.57 perseant /* Put the segment lock back the way it was. */
633 1.57 perseant while(n--)
634 1.57 perseant lfs_seglock(fs, seg_flags);
635 1.57 perseant
636 1.61 perseant /* Cleaner can relax now */
637 1.61 perseant LFS_CLEANERINFO(cip, fs, bp);
638 1.85 dholland lfs_ci_setflags(fs, cip,
639 1.85 dholland lfs_ci_getflags(fs, cip) & ~LFS_CLEANER_MUST_CLEAN);
640 1.61 perseant LFS_SYNC_CLEANERINFO(cip, fs, bp, 1);
641 1.61 perseant
642 1.57 perseant return;
643 1.57 perseant }
644 1.60 perseant
645 1.60 perseant /*
646 1.60 perseant * Wake up the cleaner, provided that nowrap is not set.
647 1.60 perseant */
648 1.60 perseant void
649 1.60 perseant lfs_wakeup_cleaner(struct lfs *fs)
650 1.60 perseant {
651 1.60 perseant if (fs->lfs_nowrap > 0)
652 1.60 perseant return;
653 1.60 perseant
654 1.82 dholland wakeup(&fs->lfs_nextsegsleep);
655 1.60 perseant wakeup(&lfs_allclean_wakeup);
656 1.60 perseant }
657