| 1 | #define DISABLE_SIGN_COMPARE_WARNINGS |
| 2 | |
| 3 | #include "git-compat-util.h" |
| 4 | #include "gettext.h" |
| 5 | #include "setup.h" |
| 6 | #include "symlinks.h" |
| 7 | |
| 8 | static int threaded_check_leading_path(struct cache_def *cache, const char *name, |
| 9 | int len, int warn_on_lstat_err); |
| 10 | static int threaded_has_dirs_only_path(struct cache_def *cache, const char *name, int len, int prefix_len); |
| 11 | |
| 12 | /* |
| 13 | * Returns the length (on a path component basis) of the longest |
| 14 | * common prefix match of 'name_a' and 'name_b'. |
| 15 | */ |
| 16 | static int longest_path_match(const char *name_a, int len_a, |
| 17 | const char *name_b, int len_b, |
| 18 | int *previous_slash) |
| 19 | { |
| 20 | int max_len, match_len = 0, match_len_prev = 0, i = 0; |
| 21 | |
| 22 | max_len = len_a < len_b ? len_a : len_b; |
| 23 | while (i < max_len && name_a[i] == name_b[i]) { |
| 24 | if (name_a[i] == '/') { |
| 25 | match_len_prev = match_len; |
| 26 | match_len = i; |
| 27 | } |
| 28 | i++; |
| 29 | } |
| 30 | /* |
| 31 | * Is 'name_b' a substring of 'name_a', the other way around, |
| 32 | * or is 'name_a' and 'name_b' the exact same string? |
| 33 | */ |
| 34 | if (i >= max_len && ((len_a > len_b && name_a[len_b] == '/') || |
| 35 | (len_a < len_b && name_b[len_a] == '/') || |
| 36 | (len_a == len_b))) { |
| 37 | match_len_prev = match_len; |
| 38 | match_len = i; |
| 39 | } |
| 40 | *previous_slash = match_len_prev; |
| 41 | return match_len; |
| 42 | } |
| 43 | |
| 44 | static struct cache_def default_cache = CACHE_DEF_INIT; |
| 45 | |
| 46 | static inline void reset_lstat_cache(struct cache_def *cache) |
| 47 | { |
| 48 | strbuf_reset(&cache->path); |
| 49 | cache->flags = 0; |
| 50 | /* |
| 51 | * The track_flags and prefix_len_stat_func members is only |
| 52 | * set by the safeguard rule inside lstat_cache() |
| 53 | */ |
| 54 | } |
| 55 | |
| 56 | #define FL_DIR (1 << 0) |
| 57 | #define FL_NOENT (1 << 1) |
| 58 | #define FL_SYMLINK (1 << 2) |
| 59 | #define FL_LSTATERR (1 << 3) |
| 60 | #define FL_ERR (1 << 4) |
| 61 | #define FL_FULLPATH (1 << 5) |
| 62 | |
| 63 | /* |
| 64 | * Check if name 'name' of length 'len' has a symlink leading |
| 65 | * component, or if the directory exists and is real, or not. |
| 66 | * |
| 67 | * To speed up the check, some information is allowed to be cached. |
| 68 | * This can be indicated by the 'track_flags' argument, which also can |
| 69 | * be used to indicate that we should check the full path. |
| 70 | * |
| 71 | * The 'prefix_len_stat_func' parameter can be used to set the length |
| 72 | * of the prefix, where the cache should use the stat() function |
| 73 | * instead of the lstat() function to test each path component. |
| 74 | */ |
| 75 | static int lstat_cache_matchlen(struct cache_def *cache, |
| 76 | const char *name, int len, |
| 77 | unsigned int *ret_flags, unsigned int track_flags, |
| 78 | int prefix_len_stat_func) |
| 79 | { |
| 80 | int match_len, last_slash, last_slash_dir, previous_slash; |
| 81 | unsigned int save_flags; |
| 82 | int ret, saved_errno = 0; |
| 83 | struct stat st; |
| 84 | |
| 85 | if (cache->track_flags != track_flags || |
| 86 | cache->prefix_len_stat_func != prefix_len_stat_func) { |
| 87 | /* |
| 88 | * As a safeguard rule we clear the cache if the |
| 89 | * values of track_flags and/or prefix_len_stat_func |
| 90 | * does not match with the last supplied values. |
| 91 | */ |
| 92 | reset_lstat_cache(cache); |
| 93 | cache->track_flags = track_flags; |
| 94 | cache->prefix_len_stat_func = prefix_len_stat_func; |
| 95 | match_len = last_slash = 0; |
| 96 | } else { |
| 97 | /* |
| 98 | * Check to see if we have a match from the cache for |
| 99 | * the 2 "excluding" path types. |
| 100 | */ |
| 101 | match_len = last_slash = |
| 102 | longest_path_match(name, len, cache->path.buf, |
| 103 | cache->path.len, &previous_slash); |
| 104 | *ret_flags = cache->flags & track_flags & (FL_NOENT|FL_SYMLINK); |
| 105 | |
| 106 | if (!(track_flags & FL_FULLPATH) && match_len == len) |
| 107 | match_len = last_slash = previous_slash; |
| 108 | |
| 109 | if (*ret_flags && match_len == cache->path.len) |
| 110 | return match_len; |
| 111 | /* |
| 112 | * If we now have match_len > 0, we would know that |
| 113 | * the matched part will always be a directory. |
| 114 | * |
| 115 | * Also, if we are tracking directories and 'name' is |
| 116 | * a substring of the cache on a path component basis, |
| 117 | * we can return immediately. |
| 118 | */ |
| 119 | *ret_flags = track_flags & FL_DIR; |
| 120 | if (*ret_flags && len == match_len) |
| 121 | return match_len; |
| 122 | } |
| 123 | |
| 124 | /* |
| 125 | * Okay, no match from the cache so far, so now we have to |
| 126 | * check the rest of the path components. |
| 127 | */ |
| 128 | *ret_flags = FL_DIR; |
| 129 | last_slash_dir = last_slash; |
| 130 | if (len > cache->path.len) |
| 131 | strbuf_grow(&cache->path, len - cache->path.len); |
| 132 | while (match_len < len) { |
| 133 | do { |
| 134 | cache->path.buf[match_len] = name[match_len]; |
| 135 | match_len++; |
| 136 | } while (match_len < len && name[match_len] != '/'); |
| 137 | if (match_len >= len && !(track_flags & FL_FULLPATH)) |
| 138 | break; |
| 139 | last_slash = match_len; |
| 140 | cache->path.buf[last_slash] = '\0'; |
| 141 | |
| 142 | if (last_slash <= prefix_len_stat_func) |
| 143 | ret = stat(cache->path.buf, &st); |
| 144 | else |
| 145 | ret = lstat(cache->path.buf, &st); |
| 146 | |
| 147 | if (ret) { |
| 148 | *ret_flags = FL_LSTATERR; |
| 149 | saved_errno = errno; |
| 150 | if (errno == ENOENT) |
| 151 | *ret_flags |= FL_NOENT; |
| 152 | } else if (S_ISDIR(st.st_mode)) { |
| 153 | last_slash_dir = last_slash; |
| 154 | continue; |
| 155 | } else if (S_ISLNK(st.st_mode)) { |
| 156 | *ret_flags = FL_SYMLINK; |
| 157 | } else { |
| 158 | *ret_flags = FL_ERR; |
| 159 | } |
| 160 | break; |
| 161 | } |
| 162 | |
| 163 | /* |
| 164 | * At the end update the cache. Note that max 3 different |
| 165 | * path types, FL_NOENT, FL_SYMLINK and FL_DIR, can be cached |
| 166 | * for the moment! |
| 167 | */ |
| 168 | save_flags = *ret_flags & track_flags & (FL_NOENT|FL_SYMLINK); |
| 169 | if (save_flags && last_slash > 0) { |
| 170 | cache->path.buf[last_slash] = '\0'; |
| 171 | cache->path.len = last_slash; |
| 172 | cache->flags = save_flags; |
| 173 | } else if ((track_flags & FL_DIR) && last_slash_dir > 0) { |
| 174 | /* |
| 175 | * We have a separate test for the directory case, |
| 176 | * since it could be that we have found a symlink or a |
| 177 | * non-existing directory and the track_flags says |
| 178 | * that we cannot cache this fact, so the cache would |
| 179 | * then have been left empty in this case. |
| 180 | * |
| 181 | * But if we are allowed to track real directories, we |
| 182 | * can still cache the path components before the last |
| 183 | * one (the found symlink or non-existing component). |
| 184 | */ |
| 185 | cache->path.buf[last_slash_dir] = '\0'; |
| 186 | cache->path.len = last_slash_dir; |
| 187 | cache->flags = FL_DIR; |
| 188 | } else { |
| 189 | reset_lstat_cache(cache); |
| 190 | } |
| 191 | if (saved_errno) |
| 192 | errno = saved_errno; |
| 193 | return match_len; |
| 194 | } |
| 195 | |
| 196 | static unsigned int lstat_cache(struct cache_def *cache, const char *name, int len, |
| 197 | unsigned int track_flags, int prefix_len_stat_func) |
| 198 | { |
| 199 | unsigned int flags; |
| 200 | (void)lstat_cache_matchlen(cache, name, len, &flags, track_flags, |
| 201 | prefix_len_stat_func); |
| 202 | return flags; |
| 203 | } |
| 204 | |
| 205 | #define USE_ONLY_LSTAT 0 |
| 206 | |
| 207 | /* |
| 208 | * Return non-zero if path 'name' has a leading symlink component |
| 209 | */ |
| 210 | int threaded_has_symlink_leading_path(struct cache_def *cache, const char *name, int len) |
| 211 | { |
| 212 | return lstat_cache(cache, name, len, FL_SYMLINK|FL_DIR, USE_ONLY_LSTAT) & FL_SYMLINK; |
| 213 | } |
| 214 | |
| 215 | int has_symlink_leading_path(const char *name, int len) |
| 216 | { |
| 217 | return threaded_has_symlink_leading_path(&default_cache, name, len); |
| 218 | } |
| 219 | |
| 220 | int check_leading_path(const char *name, int len, int warn_on_lstat_err) |
| 221 | { |
| 222 | return threaded_check_leading_path(&default_cache, name, len, |
| 223 | warn_on_lstat_err); |
| 224 | } |
| 225 | |
| 226 | /* |
| 227 | * Return zero if some leading path component of 'name' does not exist. |
| 228 | * |
| 229 | * Return -1 if leading path exists and is a directory. |
| 230 | * |
| 231 | * Return the length of a leading component if it either exists but it's not a |
| 232 | * directory, or if we were unable to lstat() it. If warn_on_lstat_err is true, |
| 233 | * also emit a warning for this error. |
| 234 | */ |
| 235 | static int threaded_check_leading_path(struct cache_def *cache, const char *name, |
| 236 | int len, int warn_on_lstat_err) |
| 237 | { |
| 238 | unsigned int flags; |
| 239 | int match_len = lstat_cache_matchlen(cache, name, len, &flags, |
| 240 | FL_SYMLINK|FL_NOENT|FL_DIR, USE_ONLY_LSTAT); |
| 241 | int saved_errno = errno; |
| 242 | |
| 243 | if (flags & FL_NOENT) |
| 244 | return 0; |
| 245 | else if (flags & FL_DIR) |
| 246 | return -1; |
| 247 | if (warn_on_lstat_err && (flags & FL_LSTATERR)) { |
| 248 | char *path = xmemdupz(name, match_len); |
| 249 | errno = saved_errno; |
| 250 | warning_errno(_("failed to lstat '%s'"), path); |
| 251 | free(path); |
| 252 | } |
| 253 | return match_len; |
| 254 | } |
| 255 | |
| 256 | int has_dirs_only_path(const char *name, int len, int prefix_len) |
| 257 | { |
| 258 | return threaded_has_dirs_only_path(&default_cache, name, len, prefix_len); |
| 259 | } |
| 260 | |
| 261 | /* |
| 262 | * Return non-zero if all path components of 'name' exists as a |
| 263 | * directory. If prefix_len > 0, we will test with the stat() |
| 264 | * function instead of the lstat() function for a prefix length of |
| 265 | * 'prefix_len', thus we then allow for symlinks in the prefix part as |
| 266 | * long as those points to real existing directories. |
| 267 | */ |
| 268 | static int threaded_has_dirs_only_path(struct cache_def *cache, const char *name, int len, int prefix_len) |
| 269 | { |
| 270 | /* |
| 271 | * Note: this function is used by the checkout machinery, which also |
| 272 | * takes care to properly reset the cache when it performs an operation |
| 273 | * that would leave the cache outdated. If this function starts caching |
| 274 | * anything else besides FL_DIR, remember to also invalidate the cache |
| 275 | * when creating or deleting paths that might be in the cache. |
| 276 | */ |
| 277 | return lstat_cache(cache, name, len, |
| 278 | FL_DIR|FL_FULLPATH, prefix_len) & |
| 279 | FL_DIR; |
| 280 | } |
| 281 | |
| 282 | static struct strbuf removal = STRBUF_INIT; |
| 283 | |
| 284 | static void do_remove_scheduled_dirs(int new_len) |
| 285 | { |
| 286 | while (removal.len > new_len) { |
| 287 | removal.buf[removal.len] = '\0'; |
| 288 | if ((startup_info->original_cwd && |
| 289 | !strcmp(removal.buf, startup_info->original_cwd)) || |
| 290 | rmdir(removal.buf)) |
| 291 | break; |
| 292 | do { |
| 293 | removal.len--; |
| 294 | } while (removal.len > new_len && |
| 295 | removal.buf[removal.len] != '/'); |
| 296 | } |
| 297 | removal.len = new_len; |
| 298 | } |
| 299 | |
| 300 | void schedule_dir_for_removal(const char *name, int len) |
| 301 | { |
| 302 | int match_len, last_slash, i, previous_slash; |
| 303 | |
| 304 | if (startup_info->original_cwd && |
| 305 | !strcmp(name, startup_info->original_cwd)) |
| 306 | return; /* Do not remove the current working directory */ |
| 307 | |
| 308 | match_len = last_slash = i = |
| 309 | longest_path_match(name, len, removal.buf, removal.len, |
| 310 | &previous_slash); |
| 311 | /* Find last slash inside 'name' */ |
| 312 | while (i < len) { |
| 313 | if (name[i] == '/') |
| 314 | last_slash = i; |
| 315 | i++; |
| 316 | } |
| 317 | |
| 318 | /* |
| 319 | * If we are about to go down the directory tree, we check if |
| 320 | * we must first go upwards the tree, such that we then can |
| 321 | * remove possible empty directories as we go upwards. |
| 322 | */ |
| 323 | if (match_len < last_slash && match_len < removal.len) |
| 324 | do_remove_scheduled_dirs(match_len); |
| 325 | /* |
| 326 | * If we go deeper down the directory tree, we only need to |
| 327 | * save the new path components as we go down. |
| 328 | */ |
| 329 | if (match_len < last_slash) |
| 330 | strbuf_add(&removal, &name[match_len], last_slash - match_len); |
| 331 | } |
| 332 | |
| 333 | void remove_scheduled_dirs(void) |
| 334 | { |
| 335 | do_remove_scheduled_dirs(0); |
| 336 | } |
| 337 | |
| 338 | void invalidate_lstat_cache(void) |
| 339 | { |
| 340 | reset_lstat_cache(&default_cache); |
| 341 | } |
| 342 | |
| 343 | #undef rmdir |
| 344 | int lstat_cache_aware_rmdir(const char *path) |
| 345 | { |
| 346 | /* Any change in this function must be made also in `mingw_rmdir()` */ |
| 347 | int ret = rmdir(path); |
| 348 | |
| 349 | if (!ret) |
| 350 | invalidate_lstat_cache(); |
| 351 | |
| 352 | return ret; |
| 353 | } |