symlinks.c

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