symlinks.con commit lstat_cache: optionally return match_len (4856ff2)
   1#include "cache.h"
   2
   3/*
   4 * Returns the length (on a path component basis) of the longest
   5 * common prefix match of 'name_a' and 'name_b'.
   6 */
   7static int longest_path_match(const char *name_a, int len_a,
   8                              const char *name_b, int len_b,
   9                              int *previous_slash)
  10{
  11        int max_len, match_len = 0, match_len_prev = 0, i = 0;
  12
  13        max_len = len_a < len_b ? len_a : len_b;
  14        while (i < max_len && name_a[i] == name_b[i]) {
  15                if (name_a[i] == '/') {
  16                        match_len_prev = match_len;
  17                        match_len = i;
  18                }
  19                i++;
  20        }
  21        /*
  22         * Is 'name_b' a substring of 'name_a', the other way around,
  23         * or is 'name_a' and 'name_b' the exact same string?
  24         */
  25        if (i >= max_len && ((len_a > len_b && name_a[len_b] == '/') ||
  26                             (len_a < len_b && name_b[len_a] == '/') ||
  27                             (len_a == len_b))) {
  28                match_len_prev = match_len;
  29                match_len = i;
  30        }
  31        *previous_slash = match_len_prev;
  32        return match_len;
  33}
  34
  35static struct cache_def default_cache;
  36
  37static inline void reset_lstat_cache(struct cache_def *cache)
  38{
  39        cache->path[0] = '\0';
  40        cache->len = 0;
  41        cache->flags = 0;
  42        /*
  43         * The track_flags and prefix_len_stat_func members is only
  44         * set by the safeguard rule inside lstat_cache()
  45         */
  46}
  47
  48#define FL_DIR      (1 << 0)
  49#define FL_NOENT    (1 << 1)
  50#define FL_SYMLINK  (1 << 2)
  51#define FL_LSTATERR (1 << 3)
  52#define FL_ERR      (1 << 4)
  53#define FL_FULLPATH (1 << 5)
  54
  55/*
  56 * Check if name 'name' of length 'len' has a symlink leading
  57 * component, or if the directory exists and is real, or not.
  58 *
  59 * To speed up the check, some information is allowed to be cached.
  60 * This can be indicated by the 'track_flags' argument, which also can
  61 * be used to indicate that we should check the full path.
  62 *
  63 * The 'prefix_len_stat_func' parameter can be used to set the length
  64 * of the prefix, where the cache should use the stat() function
  65 * instead of the lstat() function to test each path component.
  66 */
  67static int lstat_cache_matchlen(struct cache_def *cache,
  68                                const char *name, int len,
  69                                int *ret_flags, int track_flags,
  70                                int prefix_len_stat_func)
  71{
  72        int match_len, last_slash, last_slash_dir, previous_slash;
  73        int save_flags, max_len, ret;
  74        struct stat st;
  75
  76        if (cache->track_flags != track_flags ||
  77            cache->prefix_len_stat_func != prefix_len_stat_func) {
  78                /*
  79                 * As a safeguard rule we clear the cache if the
  80                 * values of track_flags and/or prefix_len_stat_func
  81                 * does not match with the last supplied values.
  82                 */
  83                reset_lstat_cache(cache);
  84                cache->track_flags = track_flags;
  85                cache->prefix_len_stat_func = prefix_len_stat_func;
  86                match_len = last_slash = 0;
  87        } else {
  88                /*
  89                 * Check to see if we have a match from the cache for
  90                 * the 2 "excluding" path types.
  91                 */
  92                match_len = last_slash =
  93                        longest_path_match(name, len, cache->path, cache->len,
  94                                           &previous_slash);
  95                *ret_flags = cache->flags & track_flags & (FL_NOENT|FL_SYMLINK);
  96
  97                if (!(track_flags & FL_FULLPATH) && match_len == len)
  98                        match_len = last_slash = previous_slash;
  99
 100                if (*ret_flags && match_len == cache->len)
 101                        return match_len;
 102                /*
 103                 * If we now have match_len > 0, we would know that
 104                 * the matched part will always be a directory.
 105                 *
 106                 * Also, if we are tracking directories and 'name' is
 107                 * a substring of the cache on a path component basis,
 108                 * we can return immediately.
 109                 */
 110                *ret_flags = track_flags & FL_DIR;
 111                if (*ret_flags && len == match_len)
 112                        return match_len;
 113        }
 114
 115        /*
 116         * Okay, no match from the cache so far, so now we have to
 117         * check the rest of the path components.
 118         */
 119        *ret_flags = FL_DIR;
 120        last_slash_dir = last_slash;
 121        max_len = len < PATH_MAX ? len : PATH_MAX;
 122        while (match_len < max_len) {
 123                do {
 124                        cache->path[match_len] = name[match_len];
 125                        match_len++;
 126                } while (match_len < max_len && name[match_len] != '/');
 127                if (match_len >= max_len && !(track_flags & FL_FULLPATH))
 128                        break;
 129                last_slash = match_len;
 130                cache->path[last_slash] = '\0';
 131
 132                if (last_slash <= prefix_len_stat_func)
 133                        ret = stat(cache->path, &st);
 134                else
 135                        ret = lstat(cache->path, &st);
 136
 137                if (ret) {
 138                        *ret_flags = FL_LSTATERR;
 139                        if (errno == ENOENT)
 140                                *ret_flags |= FL_NOENT;
 141                } else if (S_ISDIR(st.st_mode)) {
 142                        last_slash_dir = last_slash;
 143                        continue;
 144                } else if (S_ISLNK(st.st_mode)) {
 145                        *ret_flags = FL_SYMLINK;
 146                } else {
 147                        *ret_flags = FL_ERR;
 148                }
 149                break;
 150        }
 151
 152        /*
 153         * At the end update the cache.  Note that max 3 different
 154         * path types, FL_NOENT, FL_SYMLINK and FL_DIR, can be cached
 155         * for the moment!
 156         */
 157        save_flags = *ret_flags & track_flags & (FL_NOENT|FL_SYMLINK);
 158        if (save_flags && last_slash > 0 && last_slash <= PATH_MAX) {
 159                cache->path[last_slash] = '\0';
 160                cache->len = last_slash;
 161                cache->flags = save_flags;
 162        } else if ((track_flags & FL_DIR) &&
 163                   last_slash_dir > 0 && last_slash_dir <= PATH_MAX) {
 164                /*
 165                 * We have a separate test for the directory case,
 166                 * since it could be that we have found a symlink or a
 167                 * non-existing directory and the track_flags says
 168                 * that we cannot cache this fact, so the cache would
 169                 * then have been left empty in this case.
 170                 *
 171                 * But if we are allowed to track real directories, we
 172                 * can still cache the path components before the last
 173                 * one (the found symlink or non-existing component).
 174                 */
 175                cache->path[last_slash_dir] = '\0';
 176                cache->len = last_slash_dir;
 177                cache->flags = FL_DIR;
 178        } else {
 179                reset_lstat_cache(cache);
 180        }
 181        return match_len;
 182}
 183
 184static int lstat_cache(struct cache_def *cache, const char *name, int len,
 185                       int track_flags, int prefix_len_stat_func)
 186{
 187        int flags;
 188        (void)lstat_cache_matchlen(cache, name, len, &flags, track_flags,
 189                        prefix_len_stat_func);
 190        return flags;
 191}
 192
 193#define USE_ONLY_LSTAT  0
 194
 195/*
 196 * Return non-zero if path 'name' has a leading symlink component
 197 */
 198int threaded_has_symlink_leading_path(struct cache_def *cache, const char *name, int len)
 199{
 200        return lstat_cache(cache, name, len, FL_SYMLINK|FL_DIR, USE_ONLY_LSTAT) & FL_SYMLINK;
 201}
 202
 203/*
 204 * Return non-zero if path 'name' has a leading symlink component
 205 */
 206int has_symlink_leading_path(const char *name, int len)
 207{
 208        return threaded_has_symlink_leading_path(&default_cache, name, len);
 209}
 210
 211/*
 212 * Return non-zero if path 'name' has a leading symlink component or
 213 * if some leading path component does not exists.
 214 */
 215int has_symlink_or_noent_leading_path(const char *name, int len)
 216{
 217        struct cache_def *cache = &default_cache;       /* FIXME */
 218        return lstat_cache(cache, name, len,
 219                           FL_SYMLINK|FL_NOENT|FL_DIR, USE_ONLY_LSTAT) &
 220                (FL_SYMLINK|FL_NOENT);
 221}
 222
 223/*
 224 * Return non-zero if all path components of 'name' exists as a
 225 * directory.  If prefix_len > 0, we will test with the stat()
 226 * function instead of the lstat() function for a prefix length of
 227 * 'prefix_len', thus we then allow for symlinks in the prefix part as
 228 * long as those points to real existing directories.
 229 */
 230int has_dirs_only_path(const char *name, int len, int prefix_len)
 231{
 232        struct cache_def *cache = &default_cache;       /* FIXME */
 233        return lstat_cache(cache, name, len,
 234                           FL_DIR|FL_FULLPATH, prefix_len) &
 235                FL_DIR;
 236}
 237
 238static struct removal_def {
 239        char path[PATH_MAX];
 240        int len;
 241} removal;
 242
 243static void do_remove_scheduled_dirs(int new_len)
 244{
 245        while (removal.len > new_len) {
 246                removal.path[removal.len] = '\0';
 247                if (rmdir(removal.path))
 248                        break;
 249                do {
 250                        removal.len--;
 251                } while (removal.len > new_len &&
 252                         removal.path[removal.len] != '/');
 253        }
 254        removal.len = new_len;
 255}
 256
 257void schedule_dir_for_removal(const char *name, int len)
 258{
 259        int match_len, last_slash, i, previous_slash;
 260
 261        match_len = last_slash = i =
 262                longest_path_match(name, len, removal.path, removal.len,
 263                                   &previous_slash);
 264        /* Find last slash inside 'name' */
 265        while (i < len) {
 266                if (name[i] == '/')
 267                        last_slash = i;
 268                i++;
 269        }
 270
 271        /*
 272         * If we are about to go down the directory tree, we check if
 273         * we must first go upwards the tree, such that we then can
 274         * remove possible empty directories as we go upwards.
 275         */
 276        if (match_len < last_slash && match_len < removal.len)
 277                do_remove_scheduled_dirs(match_len);
 278        /*
 279         * If we go deeper down the directory tree, we only need to
 280         * save the new path components as we go down.
 281         */
 282        if (match_len < last_slash) {
 283                memcpy(&removal.path[match_len], &name[match_len],
 284                       last_slash - match_len);
 285                removal.len = last_slash;
 286        }
 287}
 288
 289void remove_scheduled_dirs(void)
 290{
 291        do_remove_scheduled_dirs(0);
 292}