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 <limits.h>
25 #include <sha1.h>
26 #include <zlib.h>
28 #include "got_error.h"
29 #include "got_object.h"
31 #include "got_lib_delta.h"
32 #include "got_lib_inflate.h"
33 #include "got_lib_object.h"
34 #include "got_lib_object_idset.h"
35 #include "got_lib_object_cache.h"
37 /*
38 * XXX This should be reworked to track cache size and usage in bytes,
39 * rather than tracking N elements capped to a maximum element size.
40 */
41 #define GOT_OBJECT_CACHE_SIZE_OBJ 256
42 #define GOT_OBJECT_CACHE_SIZE_TREE 256
43 #define GOT_OBJECT_CACHE_SIZE_COMMIT 64
44 #define GOT_OBJECT_CACHE_SIZE_TAG 2048
45 #define GOT_OBJECT_CACHE_MAX_ELEM_SIZE 1048576 /* 1 MB */
47 const struct got_error *
48 got_object_cache_init(struct got_object_cache *cache,
49 enum got_object_cache_type type)
50 {
51 memset(cache, 0, sizeof(*cache));
53 cache->idset = got_object_idset_alloc();
54 if (cache->idset == NULL)
55 return got_error_from_errno("got_object_idset_alloc");
57 cache->type = type;
58 switch (type) {
59 case GOT_OBJECT_CACHE_TYPE_OBJ:
60 cache->size = GOT_OBJECT_CACHE_SIZE_OBJ;
61 break;
62 case GOT_OBJECT_CACHE_TYPE_TREE:
63 cache->size = GOT_OBJECT_CACHE_SIZE_TREE;
64 break;
65 case GOT_OBJECT_CACHE_TYPE_COMMIT:
66 cache->size = GOT_OBJECT_CACHE_SIZE_COMMIT;
67 break;
68 case GOT_OBJECT_CACHE_TYPE_TAG:
69 cache->size = GOT_OBJECT_CACHE_SIZE_TAG;
70 break;
71 }
72 return NULL;
73 }
75 size_t
76 get_size_obj(struct got_object *obj)
77 {
78 size_t size = sizeof(*obj);
79 struct got_delta *delta;
81 if ((obj->flags & GOT_OBJ_FLAG_DELTIFIED) == 0)
82 return size;
84 STAILQ_FOREACH(delta, &obj->deltas.entries, entry) {
85 if (SIZE_MAX - sizeof(*delta) < size)
86 return SIZE_MAX;
87 size += sizeof(*delta);
88 }
90 return size;
91 }
93 size_t
94 get_size_tree(struct got_tree_object *tree)
95 {
96 size_t size = sizeof(*tree);
98 size += sizeof(struct got_tree_entry) * tree->nentries;
99 return size;
102 size_t
103 get_size_commit(struct got_commit_object *commit)
105 size_t size = sizeof(*commit);
106 struct got_object_qid *qid;
108 size += sizeof(*commit->tree_id);
109 size += strlen(commit->author);
110 size += strlen(commit->committer);
111 size += strlen(commit->logmsg);
113 STAILQ_FOREACH(qid, &commit->parent_ids, entry)
114 size += sizeof(*qid) + sizeof(*qid->id);
116 return size;
119 size_t
120 get_size_tag(struct got_tag_object *tag)
122 size_t size = sizeof(*tag);
124 size += strlen(tag->tag);
125 size += strlen(tag->tagger);
126 size += strlen(tag->tagmsg);
128 return size;
131 const struct got_error *
132 got_object_cache_add(struct got_object_cache *cache, struct got_object_id *id, void *item)
134 const struct got_error *err = NULL;
135 struct got_object_cache_entry *ce;
136 int nelem;
137 size_t size;
139 switch (cache->type) {
140 case GOT_OBJECT_CACHE_TYPE_OBJ:
141 size = get_size_obj((struct got_object *)item);
142 break;
143 case GOT_OBJECT_CACHE_TYPE_TREE:
144 size = get_size_tree((struct got_tree_object *)item);
145 break;
146 case GOT_OBJECT_CACHE_TYPE_COMMIT:
147 size = get_size_commit((struct got_commit_object *)item);
148 break;
149 case GOT_OBJECT_CACHE_TYPE_TAG:
150 size = get_size_tag((struct got_tag_object *)item);
151 break;
152 default:
153 return got_error(GOT_ERR_OBJ_TYPE);
156 if (size > GOT_OBJECT_CACHE_MAX_ELEM_SIZE) {
157 #ifdef GOT_OBJ_CACHE_DEBUG
158 char *id_str;
159 if (got_object_id_str(&id_str, id) != NULL)
160 return got_error_from_errno("got_object_id_str");
161 fprintf(stderr, "%s: not caching ", getprogname());
162 switch (cache->type) {
163 case GOT_OBJECT_CACHE_TYPE_OBJ:
164 fprintf(stderr, "object");
165 break;
166 case GOT_OBJECT_CACHE_TYPE_TREE:
167 fprintf(stderr, "tree");
168 break;
169 case GOT_OBJECT_CACHE_TYPE_COMMIT:
170 fprintf(stderr, "commit");
171 break;
172 case GOT_OBJECT_CACHE_TYPE_TAG:
173 fprintf(stderr, "tag");
174 break;
176 fprintf(stderr, " %s (%zd bytes; %zd MB)\n", id_str, size,
177 size/1024/1024);
178 free(id_str);
179 #endif
180 cache->cache_toolarge++;
181 return got_error(GOT_ERR_OBJ_TOO_LARGE);
184 nelem = got_object_idset_num_elements(cache->idset);
185 if (nelem >= cache->size) {
186 err = got_object_idset_remove((void **)&ce,
187 cache->idset, NULL);
188 if (err)
189 return err;
190 switch (cache->type) {
191 case GOT_OBJECT_CACHE_TYPE_OBJ:
192 got_object_close(ce->data.obj);
193 break;
194 case GOT_OBJECT_CACHE_TYPE_TREE:
195 got_object_tree_close(ce->data.tree);
196 break;
197 case GOT_OBJECT_CACHE_TYPE_COMMIT:
198 got_object_commit_close(ce->data.commit);
199 break;
200 case GOT_OBJECT_CACHE_TYPE_TAG:
201 got_object_tag_close(ce->data.tag);
202 break;
204 free(ce);
205 cache->cache_evict++;
208 ce = malloc(sizeof(*ce));
209 if (ce == NULL)
210 return got_error_from_errno("malloc");
211 memcpy(&ce->id, id, sizeof(ce->id));
212 switch (cache->type) {
213 case GOT_OBJECT_CACHE_TYPE_OBJ:
214 ce->data.obj = (struct got_object *)item;
215 break;
216 case GOT_OBJECT_CACHE_TYPE_TREE:
217 ce->data.tree = (struct got_tree_object *)item;
218 break;
219 case GOT_OBJECT_CACHE_TYPE_COMMIT:
220 ce->data.commit = (struct got_commit_object *)item;
221 break;
222 case GOT_OBJECT_CACHE_TYPE_TAG:
223 ce->data.tag = (struct got_tag_object *)item;
224 break;
227 err = got_object_idset_add(cache->idset, id, ce);
228 if (err)
229 free(ce);
230 return err;
233 void *
234 got_object_cache_get(struct got_object_cache *cache, struct got_object_id *id)
236 struct got_object_cache_entry *ce;
238 cache->cache_searches++;
239 ce = got_object_idset_get(cache->idset, id);
240 if (ce) {
241 cache->cache_hit++;
242 switch (cache->type) {
243 case GOT_OBJECT_CACHE_TYPE_OBJ:
244 return ce->data.obj;
245 case GOT_OBJECT_CACHE_TYPE_TREE:
246 return ce->data.tree;
247 case GOT_OBJECT_CACHE_TYPE_COMMIT:
248 return ce->data.commit;
249 case GOT_OBJECT_CACHE_TYPE_TAG:
250 return ce->data.tag;
254 cache->cache_miss++;
255 return NULL;
258 #ifdef GOT_OBJ_CACHE_DEBUG
259 static void
260 print_cache_stats(struct got_object_cache *cache, const char *name)
262 fprintf(stderr, "%s: %s cache: %d elements, %d searches, %d hits, "
263 "%d missed, %d evicted, %d too large\n", getprogname(), name,
264 got_object_idset_num_elements(cache->idset),
265 cache->cache_searches, cache->cache_hit,
266 cache->cache_miss, cache->cache_evict, cache->cache_toolarge);
269 const struct got_error *
270 check_refcount(struct got_object_id *id, void *data, void *arg)
272 struct got_object_cache *cache = arg;
273 struct got_object_cache_entry *ce = data;
274 struct got_object *obj;
275 struct got_tree_object *tree;
276 struct got_commit_object *commit;
277 struct got_tag_object *tag;
278 char *id_str;
280 if (got_object_id_str(&id_str, id) != NULL)
281 return NULL;
283 switch (cache->type) {
284 case GOT_OBJECT_CACHE_TYPE_OBJ:
285 obj = ce->data.obj;
286 if (obj->refcnt == 1)
287 break;
288 fprintf(stderr, "object %s has %d unclaimed references\n",
289 id_str, obj->refcnt - 1);
290 break;
291 case GOT_OBJECT_CACHE_TYPE_TREE:
292 tree = ce->data.tree;
293 if (tree->refcnt == 1)
294 break;
295 fprintf(stderr, "tree %s has %d unclaimed references\n",
296 id_str, tree->refcnt - 1);
297 break;
298 case GOT_OBJECT_CACHE_TYPE_COMMIT:
299 commit = ce->data.commit;
300 if (commit->refcnt == 1)
301 break;
302 fprintf(stderr, "commit %s has %d unclaimed references\n",
303 id_str, commit->refcnt - 1);
304 break;
305 case GOT_OBJECT_CACHE_TYPE_TAG:
306 tag = ce->data.tag;
307 if (tag->refcnt == 1)
308 break;
309 fprintf(stderr, "tag %s has %d unclaimed references\n",
310 id_str, tag->refcnt - 1);
311 break;
313 free(id_str);
314 return NULL;
316 #endif
318 void
319 got_object_cache_close(struct got_object_cache *cache)
321 #ifdef GOT_OBJ_CACHE_DEBUG
322 switch (cache->type) {
323 case GOT_OBJECT_CACHE_TYPE_OBJ:
324 print_cache_stats(cache, "object");
325 break;
326 case GOT_OBJECT_CACHE_TYPE_TREE:
327 print_cache_stats(cache, "tree");
328 break;
329 case GOT_OBJECT_CACHE_TYPE_COMMIT:
330 print_cache_stats(cache, "commit");
331 break;
332 case GOT_OBJECT_CACHE_TYPE_TAG:
333 print_cache_stats(cache, "tag");
334 break;
337 got_object_idset_for_each(cache->idset, check_refcount, cache);
338 #endif
340 if (cache->idset) {
341 got_object_idset_free(cache->idset);
342 cache->idset = NULL;
344 cache->size = 0;