Blob


1 /*
2 * Copyright (c) 2018 Stefan Sperling <stsp@openbsd.org>
3 *
4 * Permission to use, copy, modify, and distribute this software for any
5 * purpose with or without fee is hereby granted, provided that the above
6 * copyright notice and this permission notice appear in all copies.
7 *
8 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
9 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
10 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
11 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
12 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
13 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
14 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
15 */
17 #include <sys/time.h>
19 #include <stdio.h>
20 #include <stdlib.h>
21 #include <stdint.h>
22 #include <string.h>
23 #include <limits.h>
24 #include <sha1.h>
25 #include <zlib.h>
27 #include "got_error.h"
28 #include "got_object.h"
30 #include "got_lib_delta.h"
31 #include "got_lib_inflate.h"
32 #include "got_lib_object.h"
33 #include "got_lib_object_idset.h"
34 #include "got_lib_object_cache.h"
36 /*
37 * XXX This should be reworked to track cache size and usage in bytes,
38 * rather than tracking N elements capped to a maximum element size.
39 */
40 #define GOT_OBJECT_CACHE_SIZE_OBJ 256
41 #define GOT_OBJECT_CACHE_SIZE_TREE 256
42 #define GOT_OBJECT_CACHE_SIZE_COMMIT 64
43 #define GOT_OBJECT_CACHE_SIZE_TAG 2048
44 #define GOT_OBJECT_CACHE_MAX_ELEM_SIZE 1048576 /* 1 MB */
46 const struct got_error *
47 got_object_cache_init(struct got_object_cache *cache,
48 enum got_object_cache_type type)
49 {
50 memset(cache, 0, sizeof(*cache));
52 cache->idset = got_object_idset_alloc();
53 if (cache->idset == NULL)
54 return got_error_from_errno("got_object_idset_alloc");
56 cache->type = type;
57 switch (type) {
58 case GOT_OBJECT_CACHE_TYPE_OBJ:
59 cache->size = GOT_OBJECT_CACHE_SIZE_OBJ;
60 break;
61 case GOT_OBJECT_CACHE_TYPE_TREE:
62 cache->size = GOT_OBJECT_CACHE_SIZE_TREE;
63 break;
64 case GOT_OBJECT_CACHE_TYPE_COMMIT:
65 cache->size = GOT_OBJECT_CACHE_SIZE_COMMIT;
66 break;
67 case GOT_OBJECT_CACHE_TYPE_TAG:
68 cache->size = GOT_OBJECT_CACHE_SIZE_TAG;
69 break;
70 }
71 return NULL;
72 }
74 size_t
75 get_size_obj(struct got_object *obj)
76 {
77 size_t size = sizeof(*obj);
78 struct got_delta *delta;
80 if ((obj->flags & GOT_OBJ_FLAG_DELTIFIED) == 0)
81 return size;
83 STAILQ_FOREACH(delta, &obj->deltas.entries, entry) {
84 if (SIZE_MAX - sizeof(*delta) < size)
85 return SIZE_MAX;
86 size += sizeof(*delta);
87 }
89 return size;
90 }
92 size_t
93 get_size_tree(struct got_tree_object *tree)
94 {
95 size_t size = sizeof(*tree);
97 size += sizeof(struct got_tree_entry) * tree->nentries;
98 return size;
99 }
101 size_t
102 get_size_commit(struct got_commit_object *commit)
104 size_t size = sizeof(*commit);
105 struct got_object_qid *qid;
107 size += sizeof(*commit->tree_id);
108 size += strlen(commit->author);
109 size += strlen(commit->committer);
110 size += strlen(commit->logmsg);
112 STAILQ_FOREACH(qid, &commit->parent_ids, entry)
113 size += sizeof(*qid) + sizeof(*qid->id);
115 return size;
118 size_t
119 get_size_tag(struct got_tag_object *tag)
121 size_t size = sizeof(*tag);
123 size += strlen(tag->tag);
124 size += strlen(tag->tagger);
125 size += strlen(tag->tagmsg);
127 return size;
130 const struct got_error *
131 got_object_cache_add(struct got_object_cache *cache, struct got_object_id *id, void *item)
133 const struct got_error *err = NULL;
134 struct got_object_cache_entry *ce;
135 int nelem;
136 size_t size;
138 switch (cache->type) {
139 case GOT_OBJECT_CACHE_TYPE_OBJ:
140 size = get_size_obj((struct got_object *)item);
141 break;
142 case GOT_OBJECT_CACHE_TYPE_TREE:
143 size = get_size_tree((struct got_tree_object *)item);
144 break;
145 case GOT_OBJECT_CACHE_TYPE_COMMIT:
146 size = get_size_commit((struct got_commit_object *)item);
147 break;
148 case GOT_OBJECT_CACHE_TYPE_TAG:
149 size = get_size_tag((struct got_tag_object *)item);
150 break;
151 default:
152 return got_error(GOT_ERR_OBJ_TYPE);
155 if (size > GOT_OBJECT_CACHE_MAX_ELEM_SIZE) {
156 #ifdef GOT_OBJ_CACHE_DEBUG
157 char *id_str;
158 if (got_object_id_str(&id_str, id) != NULL)
159 return got_error_from_errno("got_object_id_str");
160 fprintf(stderr, "%s: not caching ", getprogname());
161 switch (cache->type) {
162 case GOT_OBJECT_CACHE_TYPE_OBJ:
163 fprintf(stderr, "object");
164 break;
165 case GOT_OBJECT_CACHE_TYPE_TREE:
166 fprintf(stderr, "tree");
167 break;
168 case GOT_OBJECT_CACHE_TYPE_COMMIT:
169 fprintf(stderr, "commit");
170 break;
171 case GOT_OBJECT_CACHE_TYPE_TAG:
172 fprintf(stderr, "tag");
173 break;
175 fprintf(stderr, " %s (%zd bytes; %zd MB)\n", id_str, size,
176 size/1024/1024);
177 free(id_str);
178 #endif
179 cache->cache_toolarge++;
180 return got_error(GOT_ERR_OBJ_TOO_LARGE);
183 nelem = got_object_idset_num_elements(cache->idset);
184 if (nelem >= cache->size) {
185 err = got_object_idset_remove((void **)&ce,
186 cache->idset, NULL);
187 if (err)
188 return err;
189 switch (cache->type) {
190 case GOT_OBJECT_CACHE_TYPE_OBJ:
191 got_object_close(ce->data.obj);
192 break;
193 case GOT_OBJECT_CACHE_TYPE_TREE:
194 got_object_tree_close(ce->data.tree);
195 break;
196 case GOT_OBJECT_CACHE_TYPE_COMMIT:
197 got_object_commit_close(ce->data.commit);
198 break;
199 case GOT_OBJECT_CACHE_TYPE_TAG:
200 got_object_tag_close(ce->data.tag);
201 break;
203 free(ce);
204 cache->cache_evict++;
207 ce = malloc(sizeof(*ce));
208 if (ce == NULL)
209 return got_error_from_errno("malloc");
210 memcpy(&ce->id, id, sizeof(ce->id));
211 switch (cache->type) {
212 case GOT_OBJECT_CACHE_TYPE_OBJ:
213 ce->data.obj = (struct got_object *)item;
214 break;
215 case GOT_OBJECT_CACHE_TYPE_TREE:
216 ce->data.tree = (struct got_tree_object *)item;
217 break;
218 case GOT_OBJECT_CACHE_TYPE_COMMIT:
219 ce->data.commit = (struct got_commit_object *)item;
220 break;
221 case GOT_OBJECT_CACHE_TYPE_TAG:
222 ce->data.tag = (struct got_tag_object *)item;
223 break;
226 err = got_object_idset_add(cache->idset, id, ce);
227 if (err)
228 free(ce);
229 return err;
232 void *
233 got_object_cache_get(struct got_object_cache *cache, struct got_object_id *id)
235 struct got_object_cache_entry *ce;
237 cache->cache_searches++;
238 ce = got_object_idset_get(cache->idset, id);
239 if (ce) {
240 cache->cache_hit++;
241 switch (cache->type) {
242 case GOT_OBJECT_CACHE_TYPE_OBJ:
243 return ce->data.obj;
244 case GOT_OBJECT_CACHE_TYPE_TREE:
245 return ce->data.tree;
246 case GOT_OBJECT_CACHE_TYPE_COMMIT:
247 return ce->data.commit;
248 case GOT_OBJECT_CACHE_TYPE_TAG:
249 return ce->data.tag;
253 cache->cache_miss++;
254 return NULL;
257 #ifdef GOT_OBJ_CACHE_DEBUG
258 static void
259 print_cache_stats(struct got_object_cache *cache, const char *name)
261 fprintf(stderr, "%s: %s cache: %d elements, %d searches, %d hits, "
262 "%d missed, %d evicted, %d too large\n", getprogname(), name,
263 got_object_idset_num_elements(cache->idset),
264 cache->cache_searches, cache->cache_hit,
265 cache->cache_miss, cache->cache_evict, cache->cache_toolarge);
268 const struct got_error *
269 check_refcount(struct got_object_id *id, void *data, void *arg)
271 struct got_object_cache *cache = arg;
272 struct got_object_cache_entry *ce = data;
273 struct got_object *obj;
274 struct got_tree_object *tree;
275 struct got_commit_object *commit;
276 struct got_tag_object *tag;
277 char *id_str;
279 if (got_object_id_str(&id_str, id) != NULL)
280 return NULL;
282 switch (cache->type) {
283 case GOT_OBJECT_CACHE_TYPE_OBJ:
284 obj = ce->data.obj;
285 if (obj->refcnt == 1)
286 break;
287 fprintf(stderr, "object %s has %d unclaimed references\n",
288 id_str, obj->refcnt - 1);
289 break;
290 case GOT_OBJECT_CACHE_TYPE_TREE:
291 tree = ce->data.tree;
292 if (tree->refcnt == 1)
293 break;
294 fprintf(stderr, "tree %s has %d unclaimed references\n",
295 id_str, tree->refcnt - 1);
296 break;
297 case GOT_OBJECT_CACHE_TYPE_COMMIT:
298 commit = ce->data.commit;
299 if (commit->refcnt == 1)
300 break;
301 fprintf(stderr, "commit %s has %d unclaimed references\n",
302 id_str, commit->refcnt - 1);
303 break;
304 case GOT_OBJECT_CACHE_TYPE_TAG:
305 tag = ce->data.tag;
306 if (tag->refcnt == 1)
307 break;
308 fprintf(stderr, "tag %s has %d unclaimed references\n",
309 id_str, tag->refcnt - 1);
310 break;
312 free(id_str);
313 return NULL;
315 #endif
317 void
318 got_object_cache_close(struct got_object_cache *cache)
320 #ifdef GOT_OBJ_CACHE_DEBUG
321 switch (cache->type) {
322 case GOT_OBJECT_CACHE_TYPE_OBJ:
323 print_cache_stats(cache, "object");
324 break;
325 case GOT_OBJECT_CACHE_TYPE_TREE:
326 print_cache_stats(cache, "tree");
327 break;
328 case GOT_OBJECT_CACHE_TYPE_COMMIT:
329 print_cache_stats(cache, "commit");
330 break;
331 case GOT_OBJECT_CACHE_TYPE_TAG:
332 print_cache_stats(cache, "tag");
333 break;
336 got_object_idset_for_each(cache->idset, check_refcount, cache);
337 #endif
339 if (cache->idset) {
340 got_object_idset_free(cache->idset);
341 cache->idset = NULL;
343 cache->size = 0;