symlinks.c

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