1 /*
2  * Copyright (c) 2019, 2020 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  */
16 
17 #include <sys/queue.h>
18 
19 #include <stdlib.h>
20 #include <string.h>
21 #include <stdint.h>
22 #include <stdio.h>
23 #include <zlib.h>
24 #include <limits.h>
25 #include <time.h>
26 #include <errno.h>
27 
28 #include "got_compat.h"
29 
30 #include "got_object.h"
31 #include "got_error.h"
32 
33 #include "got_lib_delta.h"
34 #include "got_lib_inflate.h"
35 #include "got_lib_object.h"
36 #include "got_lib_delta_cache.h"
37 
38 #ifndef nitems
39 #define nitems(_a) (sizeof(_a) / sizeof((_a)[0]))
40 #endif
41 
42 #define GOT_DELTA_CACHE_MIN_BUCKETS	64
43 #define GOT_DELTA_CACHE_MAX_BUCKETS	2048
44 #define GOT_DELTA_CACHE_MAX_CHAIN	2
45 #define GOT_DELTA_CACHE_MAX_DELTA_SIZE	1024
46 
47 struct got_cached_delta {
48 	off_t offset;
49 	uint8_t *data;
50 	size_t len;
51 };
52 
53 struct got_delta_cache_head {
54 	struct got_cached_delta entries[GOT_DELTA_CACHE_MAX_CHAIN];
55 	unsigned int nchain;
56 };
57 
58 struct got_delta_cache {
59 	struct got_delta_cache_head *buckets;
60 	unsigned int nbuckets;
61 	unsigned int totelem;
62 	int cache_search;
63 	int cache_hit;
64 	int cache_miss;
65 	int cache_evict;
66 	int cache_toolarge;
67 	int cache_maxtoolarge;
68 	unsigned int flags;
69 #define GOT_DELTA_CACHE_F_NOMEM	0x01
70 	SIPHASH_KEY key;
71 };
72 
73 const struct got_error *
74 got_delta_cache_alloc(struct got_delta_cache **new)
75 {
76 	const struct got_error *err;
77 	struct got_delta_cache *cache;
78 
79 	*new = NULL;
80 
81 	cache = calloc(1, sizeof(*cache));
82 	if (cache == NULL)
83 		return got_error_from_errno("calloc");
84 
85 	cache->buckets = calloc(GOT_DELTA_CACHE_MIN_BUCKETS,
86 	    sizeof(cache->buckets[0]));
87 	if (cache->buckets == NULL) {
88 		err = got_error_from_errno("calloc");
89 		free(cache);
90 		return err;
91 	}
92 	cache->nbuckets = GOT_DELTA_CACHE_MIN_BUCKETS;
93 
94 	arc4random_buf(&cache->key, sizeof(cache->key));
95 	*new = cache;
96 	return NULL;
97 }
98 
99 void
100 got_delta_cache_free(struct got_delta_cache *cache)
101 {
102 	struct got_cached_delta *delta;
103 	unsigned int i;
104 
105 #ifdef GOT_DELTA_CACHE_DEBUG
106 	fprintf(stderr, "%s: delta cache: %u elements, %d searches, %d hits, "
107 	    "%d missed, %d evicted, %d too large (max %d)\n", getprogname(),
108 	    cache->totelem, cache->cache_search, cache->cache_hit,
109 	    cache->cache_miss, cache->cache_evict, cache->cache_toolarge,
110 	    cache->cache_maxtoolarge);
111 #endif
112 	for (i = 0; i < cache->nbuckets; i++) {
113 		struct got_delta_cache_head *head;
114 		int j;
115 		head = &cache->buckets[i];
116 		for (j = 0; j < head->nchain; j++) {
117 			delta = &head->entries[j];
118 			free(delta->data);
119 		}
120 	}
121 	free(cache->buckets);
122 	free(cache);
123 }
124 
125 static uint64_t
126 delta_cache_hash(struct got_delta_cache *cache, off_t delta_offset)
127 {
128 	return SipHash24(&cache->key, &delta_offset, sizeof(delta_offset));
129 }
130 
131 #ifndef GOT_NO_DELTA_CACHE
132 static const struct got_error *
133 delta_cache_resize(struct got_delta_cache *cache, unsigned int nbuckets)
134 {
135 	struct got_delta_cache_head *buckets;
136 	size_t i;
137 
138 	buckets = calloc(nbuckets, sizeof(buckets[0]));
139 	if (buckets == NULL) {
140 		if (errno != ENOMEM)
141 			return got_error_from_errno("calloc");
142 		/* Proceed with our current amount of hash buckets. */
143 		cache->flags |= GOT_DELTA_CACHE_F_NOMEM;
144 		return NULL;
145 	}
146 
147 	arc4random_buf(&cache->key, sizeof(cache->key));
148 
149 	for (i = 0; i < cache->nbuckets; i++) {
150 		unsigned int j;
151 		for (j = 0; j < cache->buckets[i].nchain; j++) {
152 			struct got_delta_cache_head *head;
153 			struct got_cached_delta *delta;
154 			uint64_t idx;
155 			delta = &cache->buckets[i].entries[j];
156 			idx = delta_cache_hash(cache, delta->offset) % nbuckets;
157 			head = &buckets[idx];
158 			if (head->nchain < nitems(head->entries)) {
159 				struct got_cached_delta *new_delta;
160 				new_delta = &head->entries[head->nchain];
161 				memcpy(new_delta, delta, sizeof(*new_delta));
162 				head->nchain++;
163 			} else {
164 				free(delta->data);
165 				cache->totelem--;
166 			}
167 		}
168 	}
169 
170 	free(cache->buckets);
171 	cache->buckets = buckets;
172 	cache->nbuckets = nbuckets;
173 	return NULL;
174 }
175 
176 static const struct got_error *
177 delta_cache_grow(struct got_delta_cache *cache)
178 {
179 	unsigned int nbuckets;
180 
181 	if ((cache->flags & GOT_DELTA_CACHE_F_NOMEM) ||
182 	    cache->nbuckets == GOT_DELTA_CACHE_MAX_BUCKETS)
183 		return NULL;
184 
185 	if (cache->nbuckets >= GOT_DELTA_CACHE_MAX_BUCKETS / 2)
186 		nbuckets = GOT_DELTA_CACHE_MAX_BUCKETS;
187 	else
188 		nbuckets = cache->nbuckets * 2;
189 
190 	return delta_cache_resize(cache, nbuckets);
191 }
192 #endif
193 
194 const struct got_error *
195 got_delta_cache_add(struct got_delta_cache *cache,
196     off_t delta_data_offset, uint8_t *delta_data, size_t delta_len)
197 {
198 #ifdef GOT_NO_DELTA_CACHE
199 	return got_error(GOT_ERR_NO_SPACE);
200 #else
201 	const struct got_error *err = NULL;
202 	struct got_cached_delta *delta;
203 	struct got_delta_cache_head *head;
204 	uint64_t idx;
205 
206 	if (delta_len > GOT_DELTA_CACHE_MAX_DELTA_SIZE) {
207 		cache->cache_toolarge++;
208 		if (delta_len > cache->cache_maxtoolarge)
209 			cache->cache_maxtoolarge = delta_len;
210 		return got_error(GOT_ERR_NO_SPACE);
211 	}
212 
213 	if (cache->nbuckets * 3 < cache->totelem * 4) {
214 		err = delta_cache_grow(cache);
215 		if (err)
216 			return err;
217 	}
218 
219 	idx = delta_cache_hash(cache, delta_data_offset) % cache->nbuckets;
220 	head = &cache->buckets[idx];
221 	if (head->nchain >= nitems(head->entries)) {
222 		delta = &head->entries[head->nchain - 1];
223 		free(delta->data);
224 		memset(delta, 0, sizeof(*delta));
225 		head->nchain--;
226 		cache->totelem--;
227 		cache->cache_evict++;
228 	}
229 
230 	delta = &head->entries[head->nchain];
231 	delta->offset = delta_data_offset;
232 	delta->data = delta_data;
233 	delta->len = delta_len;
234 	head->nchain++;
235 	cache->totelem++;
236 
237 	return NULL;
238 #endif
239 }
240 
241 void
242 got_delta_cache_get(uint8_t **delta_data, size_t *delta_len,
243     struct got_delta_cache *cache, off_t delta_data_offset)
244 {
245 	uint64_t idx;
246 	struct got_delta_cache_head *head;
247 	struct got_cached_delta *delta;
248 	int i;
249 
250 	idx = delta_cache_hash(cache, delta_data_offset) % cache->nbuckets;
251 	head = &cache->buckets[idx];
252 
253 	cache->cache_search++;
254 	*delta_data = NULL;
255 	*delta_len = 0;
256 	for (i = 0; i < head->nchain; i++) {
257 		delta = &head->entries[i];
258 		if (delta->offset != delta_data_offset)
259 			continue;
260 		cache->cache_hit++;
261 		if (i > 0) {
262 			struct got_cached_delta tmp;
263 			memcpy(&tmp, &head->entries[0], sizeof(tmp));
264 			memcpy(&head->entries[0], &head->entries[i],
265 			    sizeof(head->entries[0]));
266 			memcpy(&head->entries[i], &tmp,
267 			    sizeof(head->entries[i]));
268 			delta = &head->entries[0];
269 		}
270 		*delta_data = delta->data;
271 		*delta_len = delta->len;
272 		return;
273 	}
274 
275 	cache->cache_miss++;
276 }