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 = 50;
  29 static const char *prune_expire = "2.weeks.ago";
  30
  31 #define MAX_ADD 10
  32 static const char *argv_pack_refs[] = {"pack-refs", "--all", "--prune", NULL};
  33 static const char *argv_reflog[] = {"reflog", "expire", "--all", NULL};
  34 static const char *argv_repack[MAX_ADD] = {"repack", "-d", "-l", NULL};
  35 static const char *argv_prune[] = {"prune", "--expire", NULL, NULL};
  36 static const char *argv_rerere[] = {"rerere", "gc", NULL};
  37
  38 static int gc_config(const char *var, const char *value, void *cb)
  39 {
  40         if (!strcmp(var, "gc.packrefs")) {
  41                 if (value && !strcmp(value, "notbare"))
  42                         pack_refs = -1;
  43                 else
  44                         pack_refs = git_config_bool(var, value);
  45                 return 0;
  46         }
  47         if (!strcmp(var, "gc.aggressivewindow")) {
  48                 aggressive_window = git_config_int(var, value);
  49                 return 0;
  50         }
  51         if (!strcmp(var, "gc.auto")) {
  52                 gc_auto_threshold = git_config_int(var, value);
  53                 return 0;
  54         }
  55         if (!strcmp(var, "gc.autopacklimit")) {
  56                 gc_auto_pack_limit = git_config_int(var, value);
  57                 return 0;
  58         }
  59         if (!strcmp(var, "gc.pruneexpire")) {
  60                 if (value && strcmp(value, "now")) {
  61                         unsigned long now = approxidate("now");
  62                         if (approxidate(value) >= now)
  63                                 return error("Invalid %s: '%s'", var, value);
  64                 }
  65                 return git_config_string(&prune_expire, var, value);
  66         }
  67         return git_default_config(var, value, cb);
  68 }
  69
  70 static void append_option(const char **cmd, const char *opt, int max_length)
  71 {
  72         int i;
  73
  74         for (i = 0; cmd[i]; i++)
  75                 ;
  76
  77         if (i + 2 >= max_length)
  78                 die("Too many options specified");
  79         cmd[i++] = opt;
  80         cmd[i] = NULL;
  81 }
  82
  83 static int too_many_loose_objects(void)
  84 {
  85         /*
  86          * Quickly check if a "gc" is needed, by estimating how
  87          * many loose objects there are.  Because SHA-1 is evenly
  88          * distributed, we can check only one and get a reasonable
  89          * estimate.
  90          */
  91         char path[PATH_MAX];
  92         const char *objdir = get_object_directory();
  93         DIR *dir;
  94         struct dirent *ent;
  95         int auto_threshold;
  96         int num_loose = 0;
  97         int needed = 0;
  98
  99         if (gc_auto_threshold <= 0)
 100                 return 0;
 101
 102         if (sizeof(path) <= snprintf(path, sizeof(path), "%s/17", objdir)) {
 103                 warning("insanely long object directory %.*s", 50, objdir);
 104                 return 0;
 105         }
 106         dir = opendir(path);
 107         if (!dir)
 108                 return 0;
 109
 110         auto_threshold = (gc_auto_threshold + 255) / 256;
 111         while ((ent = readdir(dir)) != NULL) {
 112                 if (strspn(ent->d_name, "0123456789abcdef") != 38 ||
 113                     ent->d_name[38] != '\0')
 114                         continue;
 115                 if (++num_loose > auto_threshold) {
 116                         needed = 1;
 117                         break;
 118                 }
 119         }
 120         closedir(dir);
 121         return needed;
 122 }
 123
 124 static int too_many_packs(void)
 125 {
 126         struct packed_git *p;
 127         int cnt;
 128
 129         if (gc_auto_pack_limit <= 0)
 130                 return 0;
 131
 132         prepare_packed_git();
 133         for (cnt = 0, p = packed_git; p; p = p->next) {
 134                 if (!p->pack_local)
 135                         continue;
 136                 if (p->pack_keep)
 137                         continue;
 138                 /*
 139                  * Perhaps check the size of the pack and count only
 140                  * very small ones here?
 141                  */
 142                 cnt++;
 143         }
 144         return gc_auto_pack_limit <= cnt;
 145 }
 146
 147 static int run_hook(void)
 148 {
 149         const char *argv[2];
 150         struct child_process hook;
 151         int ret;
 152
 153         argv[0] = git_path("hooks/pre-auto-gc");
 154         argv[1] = NULL;
 155
 156         if (access(argv[0], X_OK) < 0)
 157                 return 0;
 158
 159         memset(&hook, 0, sizeof(hook));
 160         hook.argv = argv;
 161         hook.no_stdin = 1;
 162         hook.stdout_to_stderr = 1;
 163
 164         ret = start_command(&hook);
 165         if (ret) {
 166                 warning("Could not spawn %s", argv[0]);
 167                 return ret;
 168         }
 169         ret = finish_command(&hook);
 170         if (ret == -ERR_RUN_COMMAND_WAITPID_SIGNAL)
 171                 warning("%s exited due to uncaught signal", argv[0]);
 172         return ret;
 173 }
 174
 175 static int need_to_gc(void)
 176 {
 177         /*
 178          * Setting gc.auto to 0 or negative can disable the
 179          * automatic gc.
 180          */
 181         if (gc_auto_threshold <= 0)
 182                 return 0;
 183
 184         /*
 185          * If there are too many loose objects, but not too many
 186          * packs, we run "repack -d -l".  If there are too many packs,
 187          * we run "repack -A -d -l".  Otherwise we tell the caller
 188          * there is no need.
 189          */
 190         if (too_many_packs())
 191                 append_option(argv_repack, "-A", MAX_ADD);
 192         else if (!too_many_loose_objects())
 193                 return 0;
 194
 195         if (run_hook())
 196                 return 0;
 197         return 1;
 198 }
 199
 200 int cmd_gc(int argc, const char **argv, const char *prefix)
 201 {
 202         int prune = 0;
 203         int aggressive = 0;
 204         int auto_gc = 0;
 205         int quiet = 0;
 206         char buf[80];
 207
 208         struct option builtin_gc_options[] = {
 209                 OPT_BOOLEAN(0, "prune", &prune, "prune unreferenced objects (deprecated)"),
 210                 OPT_BOOLEAN(0, "aggressive", &aggressive, "be more thorough (increased runtime)"),
 211                 OPT_BOOLEAN(0, "auto", &auto_gc, "enable auto-gc mode"),
 212                 OPT_BOOLEAN('q', "quiet", &quiet, "suppress progress reports"),
 213                 OPT_END()
 214         };
 215
 216         git_config(gc_config, NULL);
 217
 218         if (pack_refs < 0)
 219                 pack_refs = !is_bare_repository();
 220
 221         argc = parse_options(argc, argv, builtin_gc_options, builtin_gc_usage, 0);
 222         if (argc > 0)
 223                 usage_with_options(builtin_gc_usage, builtin_gc_options);
 224
 225         if (aggressive) {
 226                 append_option(argv_repack, "-f", MAX_ADD);
 227                 if (aggressive_window > 0) {
 228                         sprintf(buf, "--window=%d", aggressive_window);
 229                         append_option(argv_repack, buf, MAX_ADD);
 230                 }
 231         }
 232         if (quiet)
 233                 append_option(argv_repack, "-q", MAX_ADD);
 234
 235         if (auto_gc) {
 236                 /*
 237                  * Auto-gc should be least intrusive as possible.
 238                  */
 239                 if (!need_to_gc())
 240                         return 0;
 241                 fprintf(stderr, "Auto packing your repository for optimum "
 242                         "performance. You may also\n"
 243                         "run \"git gc\" manually. See "
 244                         "\"git help gc\" for more information.\n");
 245         } else
 246                 append_option(argv_repack, "-A", MAX_ADD);
 247
 248         if (pack_refs && run_command_v_opt(argv_pack_refs, RUN_GIT_CMD))
 249                 return error(FAILED_RUN, argv_pack_refs[0]);
 250
 251         if (run_command_v_opt(argv_reflog, RUN_GIT_CMD))
 252                 return error(FAILED_RUN, argv_reflog[0]);
 253
 254         if (run_command_v_opt(argv_repack, RUN_GIT_CMD))
 255                 return error(FAILED_RUN, argv_repack[0]);
 256
 257         argv_prune[2] = prune_expire;
 258         if (run_command_v_opt(argv_prune, RUN_GIT_CMD))
 259                 return error(FAILED_RUN, argv_prune[0]);
 260
 261         if (run_command_v_opt(argv_rerere, RUN_GIT_CMD))
 262                 return error(FAILED_RUN, argv_rerere[0]);
 263
 264         if (auto_gc && too_many_loose_objects())
 265                 warning("There are too many unreachable loose objects; "
 266                         "run 'git prune' to remove them.");
 267
 268         return 0;
 269 }