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>
18 #include <sys/queue.h>
20 #include <stdio.h>
21 #include <stdlib.h>
22 #include <stdint.h>
23 #include <string.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 32
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_PACKED)
81 size += strlen(obj->path_packfile);
83 if ((obj->flags & GOT_OBJ_FLAG_DELTIFIED) == 0)
84 return size;
86 SIMPLEQ_FOREACH(delta, &obj->deltas.entries, entry) {
87 if (SIZE_MAX - (sizeof(*delta) + delta->delta_len) < size)
88 return SIZE_MAX;
89 size += sizeof(*delta) + delta->delta_len;
90 }
92 return size;
93 }
95 size_t
96 get_size_tree(struct got_tree_object *tree)
97 {
98 size_t size = sizeof(*tree);
99 struct got_tree_entry *te;
101 SIMPLEQ_FOREACH(te, &tree->entries.head, entry) {
102 size += sizeof(*te);
103 size += strlen(te->name);
104 size += sizeof(*te->id);
107 return size;
110 size_t
111 get_size_commit(struct got_commit_object *commit)
113 size_t size = sizeof(*commit);
114 struct got_object_qid *qid;
116 size += sizeof(*commit->tree_id);
117 size += strlen(commit->author);
118 size += strlen(commit->committer);
119 size += strlen(commit->logmsg);
121 SIMPLEQ_FOREACH(qid, &commit->parent_ids, entry)
122 size += sizeof(*qid) + sizeof(*qid->id);
124 return size;
127 size_t
128 get_size_tag(struct got_tag_object *tag)
130 size_t size = sizeof(*tag);
132 size += strlen(tag->tag);
133 size += strlen(tag->tagger);
134 size += strlen(tag->tagmsg);
136 return size;
139 const struct got_error *
140 got_object_cache_add(struct got_object_cache *cache, struct got_object_id *id, void *item)
142 const struct got_error *err = NULL;
143 struct got_object_cache_entry *ce;
144 int nelem;
145 size_t size;
147 switch (cache->type) {
148 case GOT_OBJECT_CACHE_TYPE_OBJ:
149 size = get_size_obj((struct got_object *)item);
150 break;
151 case GOT_OBJECT_CACHE_TYPE_TREE:
152 size = get_size_tree((struct got_tree_object *)item);
153 break;
154 case GOT_OBJECT_CACHE_TYPE_COMMIT:
155 size = get_size_commit((struct got_commit_object *)item);
156 break;
157 case GOT_OBJECT_CACHE_TYPE_TAG:
158 size = get_size_tag((struct got_tag_object *)item);
159 break;
162 if (size > GOT_OBJECT_CACHE_MAX_ELEM_SIZE) {
163 #ifdef GOT_OBJ_CACHE_DEBUG
164 char *id_str;
165 if (got_object_id_str(&id_str, id) != NULL)
166 return got_error_from_errno("got_object_id_str");
167 fprintf(stderr, "%s: not caching ", getprogname());
168 switch (cache->type) {
169 case GOT_OBJECT_CACHE_TYPE_OBJ:
170 fprintf(stderr, "object");
171 break;
172 case GOT_OBJECT_CACHE_TYPE_TREE:
173 fprintf(stderr, "tree");
174 break;
175 case GOT_OBJECT_CACHE_TYPE_COMMIT:
176 fprintf(stderr, "commit");
177 break;
178 case GOT_OBJECT_CACHE_TYPE_TAG:
179 fprintf(stderr, "tag");
180 break;
182 fprintf(stderr, " %s (%zd bytes; %zd MB)\n", id_str, size,
183 size/1024/1024);
184 free(id_str);
185 #endif
186 cache->cache_toolarge++;
187 return NULL;
190 nelem = got_object_idset_num_elements(cache->idset);
191 if (nelem >= cache->size) {
192 err = got_object_idset_remove((void **)&ce,
193 cache->idset, NULL);
194 if (err)
195 return err;
196 switch (cache->type) {
197 case GOT_OBJECT_CACHE_TYPE_OBJ:
198 got_object_close(ce->data.obj);
199 break;
200 case GOT_OBJECT_CACHE_TYPE_TREE:
201 got_object_tree_close(ce->data.tree);
202 break;
203 case GOT_OBJECT_CACHE_TYPE_COMMIT:
204 got_object_commit_close(ce->data.commit);
205 break;
206 case GOT_OBJECT_CACHE_TYPE_TAG:
207 got_object_tag_close(ce->data.tag);
208 break;
210 free(ce);
211 cache->cache_evict++;
214 ce = malloc(sizeof(*ce));
215 if (ce == NULL)
216 return got_error_from_errno("malloc");
217 memcpy(&ce->id, id, sizeof(ce->id));
218 switch (cache->type) {
219 case GOT_OBJECT_CACHE_TYPE_OBJ:
220 ce->data.obj = (struct got_object *)item;
221 break;
222 case GOT_OBJECT_CACHE_TYPE_TREE:
223 ce->data.tree = (struct got_tree_object *)item;
224 break;
225 case GOT_OBJECT_CACHE_TYPE_COMMIT:
226 ce->data.commit = (struct got_commit_object *)item;
227 break;
228 case GOT_OBJECT_CACHE_TYPE_TAG:
229 ce->data.tag = (struct got_tag_object *)item;
230 break;
233 err = got_object_idset_add(cache->idset, id, ce);
234 if (err) {
235 if (err->code == GOT_ERR_OBJ_EXISTS) {
236 free(ce);
237 err = NULL;
240 return err;
243 void *
244 got_object_cache_get(struct got_object_cache *cache, struct got_object_id *id)
246 struct got_object_cache_entry *ce;
248 cache->cache_searches++;
249 ce = got_object_idset_get(cache->idset, id);
250 if (ce) {
251 cache->cache_hit++;
252 switch (cache->type) {
253 case GOT_OBJECT_CACHE_TYPE_OBJ:
254 return ce->data.obj;
255 case GOT_OBJECT_CACHE_TYPE_TREE:
256 return ce->data.tree;
257 case GOT_OBJECT_CACHE_TYPE_COMMIT:
258 return ce->data.commit;
259 case GOT_OBJECT_CACHE_TYPE_TAG:
260 return ce->data.tag;
264 cache->cache_miss++;
265 return NULL;
268 #ifdef GOT_OBJ_CACHE_DEBUG
269 static void
270 print_cache_stats(struct got_object_cache *cache, const char *name)
272 fprintf(stderr, "%s: %s cache: %d elements, %d searches, %d hits, "
273 "%d missed, %d evicted, %d too large\n", getprogname(), name,
274 got_object_idset_num_elements(cache->idset),
275 cache->cache_searches, cache->cache_hit,
276 cache->cache_miss, cache->cache_evict, cache->cache_toolarge);
279 const struct got_error *
280 check_refcount(struct got_object_id *id, void *data, void *arg)
282 struct got_object_cache *cache = arg;
283 struct got_object_cache_entry *ce = data;
284 struct got_object *obj;
285 struct got_tree_object *tree;
286 struct got_commit_object *commit;
287 struct got_tag_object *tag;
288 char *id_str;
290 if (got_object_id_str(&id_str, id) != NULL)
291 return NULL;
293 switch (cache->type) {
294 case GOT_OBJECT_CACHE_TYPE_OBJ:
295 obj = ce->data.obj;
296 if (obj->refcnt == 1)
297 break;
298 fprintf(stderr, "object %s has %d unclaimed references\n",
299 id_str, obj->refcnt - 1);
300 break;
301 case GOT_OBJECT_CACHE_TYPE_TREE:
302 tree = ce->data.tree;
303 if (tree->refcnt == 1)
304 break;
305 fprintf(stderr, "tree %s has %d unclaimed references\n",
306 id_str, tree->refcnt - 1);
307 break;
308 case GOT_OBJECT_CACHE_TYPE_COMMIT:
309 commit = ce->data.commit;
310 if (commit->refcnt == 1)
311 break;
312 fprintf(stderr, "commit %s has %d unclaimed references\n",
313 id_str, commit->refcnt - 1);
314 break;
315 case GOT_OBJECT_CACHE_TYPE_TAG:
316 tag = ce->data.tag;
317 if (tag->refcnt == 1)
318 break;
319 fprintf(stderr, "tag %s has %d unclaimed references\n",
320 id_str, tag->refcnt - 1);
321 break;
323 free(id_str);
324 return NULL;
326 #endif
328 void
329 got_object_cache_close(struct got_object_cache *cache)
331 #ifdef GOT_OBJ_CACHE_DEBUG
332 switch (cache->type) {
333 case GOT_OBJECT_CACHE_TYPE_OBJ:
334 print_cache_stats(cache, "object");
335 break;
336 case GOT_OBJECT_CACHE_TYPE_TREE:
337 print_cache_stats(cache, "tree");
338 break;
339 case GOT_OBJECT_CACHE_TYPE_COMMIT:
340 print_cache_stats(cache, "commit");
341 break;
342 case GOT_OBJECT_CACHE_TYPE_TAG:
343 print_cache_stats(cache, "tag");
344 break;
347 got_object_idset_for_each(cache->idset, check_refcount, cache);
348 #endif
350 if (cache->idset) {
351 got_object_idset_free(cache->idset);
352 cache->idset = NULL;
354 cache->size = 0;