gc: --prune prunes unreferenced objects.
[git/git.git] / builtin-gc.c
1 /*
2 * git gc builtin command
3 *
4 * Cleanup unreachable files and optimize the repository.
5 *
6 * Copyright (c) 2007 James Bowes
7 *
8 * Based on git-gc.sh, which is
9 *
10 * Copyright (c) 2006 Shawn O. Pearce
11 */
12
13 #include "builtin.h"
14 #include "cache.h"
15 #include "parse-options.h"
16 #include "run-command.h"
17
18 #define FAILED_RUN "failed to run %s"
19
20 static const char * const builtin_gc_usage[] = {
21 "git-gc [options]",
22 NULL
23 };
24
25 static int pack_refs = 1;
26 static int aggressive_window = -1;
27 static int gc_auto_threshold = 6700;
28 static int gc_auto_pack_limit = 20;
29
30 #define MAX_ADD 10
31 static const char *argv_pack_refs[] = {"pack-refs", "--all", "--prune", NULL};
32 static const char *argv_reflog[] = {"reflog", "expire", "--all", NULL};
33 static const char *argv_repack[MAX_ADD] = {"repack", "-d", "-l", NULL};
34 static const char *argv_prune[] = {"prune", NULL};
35 static const char *argv_rerere[] = {"rerere", "gc", NULL};
36
37 static int gc_config(const char *var, const char *value)
38 {
39 if (!strcmp(var, "gc.packrefs")) {
40 if (!strcmp(value, "notbare"))
41 pack_refs = -1;
42 else
43 pack_refs = git_config_bool(var, value);
44 return 0;
45 }
46 if (!strcmp(var, "gc.aggressivewindow")) {
47 aggressive_window = git_config_int(var, value);
48 return 0;
49 }
50 if (!strcmp(var, "gc.auto")) {
51 gc_auto_threshold = git_config_int(var, value);
52 return 0;
53 }
54 if (!strcmp(var, "gc.autopacklimit")) {
55 gc_auto_pack_limit = git_config_int(var, value);
56 return 0;
57 }
58 return git_default_config(var, value);
59 }
60
61 static void append_option(const char **cmd, const char *opt, int max_length)
62 {
63 int i;
64
65 for (i = 0; cmd[i]; i++)
66 ;
67
68 if (i + 2 >= max_length)
69 die("Too many options specified");
70 cmd[i++] = opt;
71 cmd[i] = NULL;
72 }
73
74 static int too_many_loose_objects(void)
75 {
76 /*
77 * Quickly check if a "gc" is needed, by estimating how
78 * many loose objects there are. Because SHA-1 is evenly
79 * distributed, we can check only one and get a reasonable
80 * estimate.
81 */
82 char path[PATH_MAX];
83 const char *objdir = get_object_directory();
84 DIR *dir;
85 struct dirent *ent;
86 int auto_threshold;
87 int num_loose = 0;
88 int needed = 0;
89
90 if (gc_auto_threshold <= 0)
91 return 0;
92
93 if (sizeof(path) <= snprintf(path, sizeof(path), "%s/17", objdir)) {
94 warning("insanely long object directory %.*s", 50, objdir);
95 return 0;
96 }
97 dir = opendir(path);
98 if (!dir)
99 return 0;
100
101 auto_threshold = (gc_auto_threshold + 255) / 256;
102 while ((ent = readdir(dir)) != NULL) {
103 if (strspn(ent->d_name, "0123456789abcdef") != 38 ||
104 ent->d_name[38] != '\0')
105 continue;
106 if (++num_loose > auto_threshold) {
107 needed = 1;
108 break;
109 }
110 }
111 closedir(dir);
112 return needed;
113 }
114
115 static int too_many_packs(void)
116 {
117 struct packed_git *p;
118 int cnt;
119
120 if (gc_auto_pack_limit <= 0)
121 return 0;
122
123 prepare_packed_git();
124 for (cnt = 0, p = packed_git; p; p = p->next) {
125 char path[PATH_MAX];
126 size_t len;
127 int keep;
128
129 if (!p->pack_local)
130 continue;
131 len = strlen(p->pack_name);
132 if (PATH_MAX <= len + 1)
133 continue; /* oops, give up */
134 memcpy(path, p->pack_name, len-5);
135 memcpy(path + len - 5, ".keep", 6);
136 keep = access(p->pack_name, F_OK) && (errno == ENOENT);
137 if (keep)
138 continue;
139 /*
140 * Perhaps check the size of the pack and count only
141 * very small ones here?
142 */
143 cnt++;
144 }
145 return gc_auto_pack_limit <= cnt;
146 }
147
148 static int need_to_gc(void)
149 {
150 /*
151 * Setting gc.auto and gc.autopacklimit to 0 or negative can
152 * disable the automatic gc.
153 */
154 if (gc_auto_threshold <= 0 && gc_auto_pack_limit <= 0)
155 return 0;
156
157 /*
158 * If there are too many loose objects, but not too many
159 * packs, we run "repack -d -l". If there are too many packs,
160 * we run "repack -A -d -l". Otherwise we tell the caller
161 * there is no need.
162 */
163 if (too_many_packs())
164 append_option(argv_repack, "-A", MAX_ADD);
165 else if (!too_many_loose_objects())
166 return 0;
167 return 1;
168 }
169
170 int cmd_gc(int argc, const char **argv, const char *prefix)
171 {
172 int prune = 0;
173 int aggressive = 0;
174 int auto_gc = 0;
175 char buf[80];
176
177 struct option builtin_gc_options[] = {
178 OPT_BOOLEAN(0, "prune", &prune, "prune unreferenced objects"),
179 OPT_BOOLEAN(0, "aggressive", &aggressive, "be more thorough (increased runtime)"),
180 OPT_BOOLEAN(0, "auto", &auto_gc, "enable auto-gc mode"),
181 OPT_END()
182 };
183
184 git_config(gc_config);
185
186 if (pack_refs < 0)
187 pack_refs = !is_bare_repository();
188
189 argc = parse_options(argc, argv, builtin_gc_options, builtin_gc_usage, 0);
190 if (argc > 0)
191 usage_with_options(builtin_gc_usage, builtin_gc_options);
192
193 if (aggressive) {
194 append_option(argv_repack, "-f", MAX_ADD);
195 if (aggressive_window > 0) {
196 sprintf(buf, "--window=%d", aggressive_window);
197 append_option(argv_repack, buf, MAX_ADD);
198 }
199 }
200
201 if (auto_gc) {
202 /*
203 * Auto-gc should be least intrusive as possible.
204 */
205 prune = 0;
206 if (!need_to_gc())
207 return 0;
208 fprintf(stderr, "Packing your repository for optimum "
209 "performance. You may also\n"
210 "run \"git gc\" manually. See "
211 "\"git help gc\" for more information.\n");
212 } else {
213 /*
214 * Use safer (for shared repos) "-A" option to
215 * repack when not pruning. Auto-gc makes its
216 * own decision.
217 */
218 if (prune)
219 append_option(argv_repack, "-a", MAX_ADD);
220 else
221 append_option(argv_repack, "-A", MAX_ADD);
222 }
223
224 if (pack_refs && run_command_v_opt(argv_pack_refs, RUN_GIT_CMD))
225 return error(FAILED_RUN, argv_pack_refs[0]);
226
227 if (run_command_v_opt(argv_reflog, RUN_GIT_CMD))
228 return error(FAILED_RUN, argv_reflog[0]);
229
230 if (run_command_v_opt(argv_repack, RUN_GIT_CMD))
231 return error(FAILED_RUN, argv_repack[0]);
232
233 if (prune && run_command_v_opt(argv_prune, RUN_GIT_CMD))
234 return error(FAILED_RUN, argv_prune[0]);
235
236 if (run_command_v_opt(argv_rerere, RUN_GIT_CMD))
237 return error(FAILED_RUN, argv_rerere[0]);
238
239 if (auto_gc && too_many_loose_objects())
240 warning("There are too many unreachable loose objects; "
241 "run 'git prune' to remove them.");
242
243 return 0;
244 }