4 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License version 2 only,
8 * as published by the Free Software Foundation.
10 * This program is distributed in the hope that it will be useful, but
11 * WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * General Public License version 2 for more details (a copy is included
14 * in the LICENSE file that accompanied this code).
16 * You should have received a copy of the GNU General Public License
17 * version 2 along with this program; If not, see
18 * http://www.gnu.org/licenses/gpl-2.0.html
23 * Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved.
24 * Use is subject to license terms.
26 * Copyright (c) 2012, 2014, Intel Corporation.
29 * This file is part of Lustre, http://www.lustre.org/
30 * Lustre is a trademark of Sun Microsystems, Inc.
32 * lustre/fld/fld_cache.c
34 * FLD (Fids Location Database)
36 * Author: Pravin Shelar <pravin.shelar@sun.com>
37 * Author: Yury Umanets <umka@clusterfs.com>
40 #define DEBUG_SUBSYSTEM S_FLD
42 #include <libcfs/libcfs.h>
43 #include <linux/module.h>
44 #include <linux/math64.h>
45 #include <obd_support.h>
46 #include <lustre_fld.h>
47 #include "fld_internal.h"
52 struct fld_cache *fld_cache_init(const char *name,
53 int cache_size, int cache_threshold)
55 struct fld_cache *cache;
58 LASSERT(name != NULL);
59 LASSERT(cache_threshold < cache_size);
63 RETURN(ERR_PTR(-ENOMEM));
65 INIT_LIST_HEAD(&cache->fci_entries_head);
66 INIT_LIST_HEAD(&cache->fci_lru);
68 cache->fci_cache_count = 0;
69 rwlock_init(&cache->fci_lock);
71 strlcpy(cache->fci_name, name,
72 sizeof(cache->fci_name));
74 cache->fci_cache_size = cache_size;
75 cache->fci_threshold = cache_threshold;
77 /* Init fld cache info. */
78 memset(&cache->fci_stat, 0, sizeof(cache->fci_stat));
80 CDEBUG(D_INFO, "%s: FLD cache - Size: %d, Threshold: %d\n",
81 cache->fci_name, cache_size, cache_threshold);
89 void fld_cache_fini(struct fld_cache *cache)
93 LASSERT(cache != NULL);
94 fld_cache_flush(cache);
96 CDEBUG(D_INFO, "FLD cache statistics (%s):\n", cache->fci_name);
97 CDEBUG(D_INFO, " Total reqs: %llu\n", cache->fci_stat.fst_count);
98 CDEBUG(D_INFO, " Cache reqs: %llu\n", cache->fci_stat.fst_cache);
99 CDEBUG(D_INFO, " Cache hits: %u%%\n",
100 pct(cache->fci_stat.fst_cache, cache->fci_stat.fst_count));
108 * delete given node from list.
110 void fld_cache_entry_delete(struct fld_cache *cache,
111 struct fld_cache_entry *node)
113 list_del(&node->fce_list);
114 list_del(&node->fce_lru);
115 cache->fci_cache_count--;
120 * fix list by checking new entry with NEXT entry in order.
122 static void fld_fix_new_list(struct fld_cache *cache)
124 struct fld_cache_entry *f_curr;
125 struct fld_cache_entry *f_next;
126 struct lu_seq_range *c_range;
127 struct lu_seq_range *n_range;
128 struct list_head *head = &cache->fci_entries_head;
133 list_for_each_entry_safe(f_curr, f_next, head, fce_list) {
134 c_range = &f_curr->fce_range;
135 n_range = &f_next->fce_range;
137 LASSERT(lu_seq_range_is_sane(c_range));
138 if (&f_next->fce_list == head)
141 if (c_range->lsr_flags != n_range->lsr_flags)
144 LASSERTF(c_range->lsr_start <= n_range->lsr_start,
145 "cur lsr_start "DRANGE" next lsr_start "DRANGE"\n",
146 PRANGE(c_range), PRANGE(n_range));
148 /* check merge possibility with next range */
149 if (c_range->lsr_end == n_range->lsr_start) {
150 if (c_range->lsr_index != n_range->lsr_index)
152 n_range->lsr_start = c_range->lsr_start;
153 fld_cache_entry_delete(cache, f_curr);
157 /* check if current range overlaps with next range. */
158 if (n_range->lsr_start < c_range->lsr_end) {
159 if (c_range->lsr_index == n_range->lsr_index) {
160 n_range->lsr_start = c_range->lsr_start;
161 n_range->lsr_end = max(c_range->lsr_end,
163 fld_cache_entry_delete(cache, f_curr);
165 if (n_range->lsr_end <= c_range->lsr_end) {
167 fld_cache_entry_delete(cache, f_curr);
169 n_range->lsr_start = c_range->lsr_end;
172 /* we could have overlap over next
173 * range too. better restart. */
177 /* kill duplicates */
178 if (c_range->lsr_start == n_range->lsr_start &&
179 c_range->lsr_end == n_range->lsr_end)
180 fld_cache_entry_delete(cache, f_curr);
187 * add node to fld cache
189 static inline void fld_cache_entry_add(struct fld_cache *cache,
190 struct fld_cache_entry *f_new,
191 struct list_head *pos)
193 list_add(&f_new->fce_list, pos);
194 list_add(&f_new->fce_lru, &cache->fci_lru);
196 cache->fci_cache_count++;
197 fld_fix_new_list(cache);
201 * Check if cache needs to be shrunk. If so - do it.
202 * Remove one entry in list and so on until cache is shrunk enough.
204 static int fld_cache_shrink(struct fld_cache *cache)
206 struct fld_cache_entry *flde;
207 struct list_head *curr;
211 LASSERT(cache != NULL);
213 if (cache->fci_cache_count < cache->fci_cache_size)
216 curr = cache->fci_lru.prev;
218 while (cache->fci_cache_count + cache->fci_threshold >
219 cache->fci_cache_size && curr != &cache->fci_lru) {
221 flde = list_entry(curr, struct fld_cache_entry, fce_lru);
223 fld_cache_entry_delete(cache, flde);
227 CDEBUG(D_INFO, "%s: FLD cache - Shrunk by "
228 "%d entries\n", cache->fci_name, num);
234 * kill all fld cache entries.
236 void fld_cache_flush(struct fld_cache *cache)
240 write_lock(&cache->fci_lock);
241 cache->fci_cache_size = 0;
242 fld_cache_shrink(cache);
243 write_unlock(&cache->fci_lock);
249 * punch hole in existing range. divide this range and add new
253 static void fld_cache_punch_hole(struct fld_cache *cache,
254 struct fld_cache_entry *f_curr,
255 struct fld_cache_entry *f_new)
257 const struct lu_seq_range *range = &f_new->fce_range;
258 const u64 new_start = range->lsr_start;
259 const u64 new_end = range->lsr_end;
260 struct fld_cache_entry *fldt;
263 OBD_ALLOC_GFP(fldt, sizeof *fldt, GFP_ATOMIC);
267 /* overlap is not allowed, so dont mess up list. */
270 /* break f_curr RANGE into three RANGES:
271 * f_curr, f_new , fldt
277 fldt->fce_range.lsr_start = new_end;
278 fldt->fce_range.lsr_end = f_curr->fce_range.lsr_end;
279 fldt->fce_range.lsr_index = f_curr->fce_range.lsr_index;
282 f_curr->fce_range.lsr_end = new_start;
284 /* add these two entries to list */
285 fld_cache_entry_add(cache, f_new, &f_curr->fce_list);
286 fld_cache_entry_add(cache, fldt, &f_new->fce_list);
288 /* no need to fixup */
293 * handle range overlap in fld cache.
295 static void fld_cache_overlap_handle(struct fld_cache *cache,
296 struct fld_cache_entry *f_curr,
297 struct fld_cache_entry *f_new)
299 const struct lu_seq_range *range = &f_new->fce_range;
300 const u64 new_start = range->lsr_start;
301 const u64 new_end = range->lsr_end;
302 const u32 mdt = range->lsr_index;
304 /* this is overlap case, these case are checking overlapping with
305 * prev range only. fixup will handle overlaping with next range. */
307 if (f_curr->fce_range.lsr_index == mdt) {
308 f_curr->fce_range.lsr_start = min(f_curr->fce_range.lsr_start,
311 f_curr->fce_range.lsr_end = max(f_curr->fce_range.lsr_end,
315 fld_fix_new_list(cache);
317 } else if (new_start <= f_curr->fce_range.lsr_start &&
318 f_curr->fce_range.lsr_end <= new_end) {
319 /* case 1: new range completely overshadowed existing range.
320 * e.g. whole range migrated. update fld cache entry */
322 f_curr->fce_range = *range;
324 fld_fix_new_list(cache);
326 } else if (f_curr->fce_range.lsr_start < new_start &&
327 new_end < f_curr->fce_range.lsr_end) {
328 /* case 2: new range fit within existing range. */
330 fld_cache_punch_hole(cache, f_curr, f_new);
332 } else if (new_end <= f_curr->fce_range.lsr_end) {
334 * [new_start [c_start new_end) c_end)
337 LASSERT(new_start <= f_curr->fce_range.lsr_start);
339 f_curr->fce_range.lsr_start = new_end;
340 fld_cache_entry_add(cache, f_new, f_curr->fce_list.prev);
342 } else if (f_curr->fce_range.lsr_start <= new_start) {
344 * [c_start [new_start c_end) new_end)
347 LASSERT(f_curr->fce_range.lsr_end <= new_end);
349 f_curr->fce_range.lsr_end = new_start;
350 fld_cache_entry_add(cache, f_new, &f_curr->fce_list);
352 CERROR("NEW range ="DRANGE" curr = "DRANGE"\n",
353 PRANGE(range),PRANGE(&f_curr->fce_range));
356 struct fld_cache_entry
357 *fld_cache_entry_create(const struct lu_seq_range *range)
359 struct fld_cache_entry *f_new;
361 LASSERT(lu_seq_range_is_sane(range));
363 OBD_ALLOC_PTR(f_new);
365 RETURN(ERR_PTR(-ENOMEM));
367 f_new->fce_range = *range;
372 * Insert FLD entry in FLD cache.
374 * This function handles all cases of merging and breaking up of
377 int fld_cache_insert_nolock(struct fld_cache *cache,
378 struct fld_cache_entry *f_new)
380 struct fld_cache_entry *f_curr;
381 struct fld_cache_entry *n;
382 struct list_head *head;
383 struct list_head *prev = NULL;
384 const u64 new_start = f_new->fce_range.lsr_start;
385 const u64 new_end = f_new->fce_range.lsr_end;
386 __u32 new_flags = f_new->fce_range.lsr_flags;
390 * Duplicate entries are eliminated in insert op.
391 * So we don't need to search new entry before starting
395 if (!cache->fci_no_shrink)
396 fld_cache_shrink(cache);
398 head = &cache->fci_entries_head;
400 list_for_each_entry_safe(f_curr, n, head, fce_list) {
401 /* add list if next is end of list */
402 if (new_end < f_curr->fce_range.lsr_start ||
403 (new_end == f_curr->fce_range.lsr_start &&
404 new_flags != f_curr->fce_range.lsr_flags))
407 prev = &f_curr->fce_list;
408 /* check if this range is to left of new range. */
409 if (new_start < f_curr->fce_range.lsr_end &&
410 new_flags == f_curr->fce_range.lsr_flags) {
411 fld_cache_overlap_handle(cache, f_curr, f_new);
419 CDEBUG(D_INFO, "insert range "DRANGE"\n", PRANGE(&f_new->fce_range));
420 /* Add new entry to cache and lru list. */
421 fld_cache_entry_add(cache, f_new, prev);
426 int fld_cache_insert(struct fld_cache *cache,
427 const struct lu_seq_range *range)
429 struct fld_cache_entry *flde;
432 flde = fld_cache_entry_create(range);
434 RETURN(PTR_ERR(flde));
436 write_lock(&cache->fci_lock);
437 rc = fld_cache_insert_nolock(cache, flde);
438 write_unlock(&cache->fci_lock);
445 void fld_cache_delete_nolock(struct fld_cache *cache,
446 const struct lu_seq_range *range)
448 struct fld_cache_entry *flde;
449 struct fld_cache_entry *tmp;
450 struct list_head *head;
452 head = &cache->fci_entries_head;
453 list_for_each_entry_safe(flde, tmp, head, fce_list) {
454 /* add list if next is end of list */
455 if (range->lsr_start == flde->fce_range.lsr_start ||
456 (range->lsr_end == flde->fce_range.lsr_end &&
457 range->lsr_flags == flde->fce_range.lsr_flags)) {
458 fld_cache_entry_delete(cache, flde);
465 * Delete FLD entry in FLD cache.
468 void fld_cache_delete(struct fld_cache *cache,
469 const struct lu_seq_range *range)
471 write_lock(&cache->fci_lock);
472 fld_cache_delete_nolock(cache, range);
473 write_unlock(&cache->fci_lock);
476 struct fld_cache_entry *
477 fld_cache_entry_lookup_nolock(struct fld_cache *cache,
478 const struct lu_seq_range *range)
480 struct fld_cache_entry *flde;
481 struct fld_cache_entry *got = NULL;
482 struct list_head *head;
484 head = &cache->fci_entries_head;
485 list_for_each_entry(flde, head, fce_list) {
486 if (range->lsr_start == flde->fce_range.lsr_start ||
487 (range->lsr_end == flde->fce_range.lsr_end &&
488 range->lsr_flags == flde->fce_range.lsr_flags)) {
498 * lookup \a seq sequence for range in fld cache.
500 struct fld_cache_entry *
501 fld_cache_entry_lookup(struct fld_cache *cache,
502 const struct lu_seq_range *range)
504 struct fld_cache_entry *got = NULL;
507 read_lock(&cache->fci_lock);
508 got = fld_cache_entry_lookup_nolock(cache, range);
509 read_unlock(&cache->fci_lock);
515 * lookup \a seq sequence for range in fld cache.
517 int fld_cache_lookup(struct fld_cache *cache,
518 const u64 seq, struct lu_seq_range *range)
520 struct fld_cache_entry *flde;
521 struct fld_cache_entry *prev = NULL;
522 struct list_head *head;
525 read_lock(&cache->fci_lock);
526 head = &cache->fci_entries_head;
528 cache->fci_stat.fst_count++;
529 list_for_each_entry(flde, head, fce_list) {
530 if (flde->fce_range.lsr_start > seq) {
532 *range = prev->fce_range;
537 if (lu_seq_range_within(&flde->fce_range, seq)) {
538 *range = flde->fce_range;
540 cache->fci_stat.fst_cache++;
541 read_unlock(&cache->fci_lock);
545 read_unlock(&cache->fci_lock);