Avoid using 'lstat()' to figure out directories
[git/git.git] / symlinks.c
CommitLineData
f859c846
JH
1#include "cache.h"
2
92604b46
KB
3/*
4 * Returns the length (on a path component basis) of the longest
148bc06b 5 * common prefix match of 'name_a' and 'name_b'.
92604b46 6 */
148bc06b
KB
7static int longest_path_match(const char *name_a, int len_a,
8 const char *name_b, int len_b,
9 int *previous_slash)
c40641b7 10{
aeabab5c 11 int max_len, match_len = 0, match_len_prev = 0, i = 0;
92604b46 12
148bc06b
KB
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] == '/') {
aeabab5c 16 match_len_prev = match_len;
92604b46 17 match_len = i;
aeabab5c 18 }
92604b46
KB
19 i++;
20 }
60b458b7 21 /*
148bc06b
KB
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?
60b458b7 24 */
148bc06b
KB
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))) {
aeabab5c 28 match_len_prev = match_len;
60b458b7 29 match_len = i;
aeabab5c
KB
30 }
31 *previous_slash = match_len_prev;
92604b46 32 return match_len;
c40641b7
LT
33}
34
148bc06b
KB
35static struct cache_def {
36 char path[PATH_MAX + 1];
37 int len;
38 int flags;
39 int track_flags;
40 int prefix_len_stat_func;
41} cache;
42
60b458b7 43static inline void reset_lstat_cache(void)
c40641b7 44{
92604b46
KB
45 cache.path[0] = '\0';
46 cache.len = 0;
47 cache.flags = 0;
60b458b7
KB
48 /*
49 * The track_flags and prefix_len_stat_func members is only
50 * set by the safeguard rule inside lstat_cache()
51 */
c40641b7
LT
52}
53
92604b46 54#define FL_DIR (1 << 0)
09c93066
KB
55#define FL_NOENT (1 << 1)
56#define FL_SYMLINK (1 << 2)
57#define FL_LSTATERR (1 << 3)
58#define FL_ERR (1 << 4)
bad4a54f 59#define FL_FULLPATH (1 << 5)
92604b46
KB
60
61/*
62 * Check if name 'name' of length 'len' has a symlink leading
09c93066 63 * component, or if the directory exists and is real, or not.
92604b46
KB
64 *
65 * To speed up the check, some information is allowed to be cached.
bad4a54f
KB
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.
92604b46 72 */
57199892 73static int lstat_cache(const char *name, int len,
bad4a54f 74 int track_flags, int prefix_len_stat_func)
c40641b7 75{
aeabab5c 76 int match_len, last_slash, last_slash_dir, previous_slash;
bad4a54f 77 int match_flags, ret_flags, save_flags, max_len, ret;
c40641b7 78 struct stat st;
f859c846 79
bad4a54f
KB
80 if (cache.track_flags != track_flags ||
81 cache.prefix_len_stat_func != prefix_len_stat_func) {
09c93066 82 /*
60b458b7
KB
83 * As a safeguard rule we clear the cache if the
84 * values of track_flags and/or prefix_len_stat_func
85 * does not match with the last supplied values.
09c93066 86 */
60b458b7
KB
87 reset_lstat_cache();
88 cache.track_flags = track_flags;
89 cache.prefix_len_stat_func = prefix_len_stat_func;
09c93066
KB
90 match_len = last_slash = 0;
91 } else {
92 /*
93 * Check to see if we have a match from the cache for
94 * the 2 "excluding" path types.
95 */
aeabab5c 96 match_len = last_slash =
148bc06b
KB
97 longest_path_match(name, len, cache.path, cache.len,
98 &previous_slash);
09c93066
KB
99 match_flags = cache.flags & track_flags & (FL_NOENT|FL_SYMLINK);
100 if (match_flags && match_len == cache.len)
101 return match_flags;
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 match_flags = track_flags & FL_DIR;
111 if (match_flags && len == match_len)
112 return match_flags;
113 }
c40641b7 114
92604b46
KB
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] != '/');
bad4a54f 127 if (match_len >= max_len && !(track_flags & FL_FULLPATH))
92604b46
KB
128 break;
129 last_slash = match_len;
130 cache.path[last_slash] = '\0';
f859c846 131
bad4a54f
KB
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) {
92604b46 138 ret_flags = FL_LSTATERR;
09c93066
KB
139 if (errno == ENOENT)
140 ret_flags |= FL_NOENT;
92604b46
KB
141 } else if (S_ISDIR(st.st_mode)) {
142 last_slash_dir = last_slash;
c40641b7 143 continue;
92604b46
KB
144 } else if (S_ISLNK(st.st_mode)) {
145 ret_flags = FL_SYMLINK;
146 } else {
147 ret_flags = FL_ERR;
f859c846 148 }
c40641b7 149 break;
f859c846 150 }
92604b46
KB
151
152 /*
09c93066
KB
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!
92604b46 156 */
09c93066 157 save_flags = ret_flags & track_flags & (FL_NOENT|FL_SYMLINK);
bad4a54f 158 if (save_flags && last_slash > 0 && last_slash <= PATH_MAX) {
92604b46
KB
159 cache.path[last_slash] = '\0';
160 cache.len = last_slash;
161 cache.flags = save_flags;
60b458b7 162 } else if ((track_flags & FL_DIR) &&
bad4a54f 163 last_slash_dir > 0 && last_slash_dir <= PATH_MAX) {
92604b46
KB
164 /*
165 * We have a separate test for the directory case,
09c93066
KB
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.
92604b46
KB
170 *
171 * But if we are allowed to track real directories, we
172 * can still cache the path components before the last
09c93066 173 * one (the found symlink or non-existing component).
92604b46
KB
174 */
175 cache.path[last_slash_dir] = '\0';
176 cache.len = last_slash_dir;
177 cache.flags = FL_DIR;
178 } else {
60b458b7 179 reset_lstat_cache();
92604b46
KB
180 }
181 return ret_flags;
182}
183
aeabab5c
KB
184/*
185 * Invalidate the given 'name' from the cache, if 'name' matches
186 * completely with the cache.
187 */
57199892 188void invalidate_lstat_cache(const char *name, int len)
aeabab5c
KB
189{
190 int match_len, previous_slash;
191
148bc06b
KB
192 match_len = longest_path_match(name, len, cache.path, cache.len,
193 &previous_slash);
aeabab5c
KB
194 if (len == match_len) {
195 if ((cache.track_flags & FL_DIR) && previous_slash > 0) {
196 cache.path[previous_slash] = '\0';
197 cache.len = previous_slash;
198 cache.flags = FL_DIR;
cb319c36 199 } else {
60b458b7 200 reset_lstat_cache();
cb319c36 201 }
aeabab5c
KB
202 }
203}
204
bda6eb0d
KB
205/*
206 * Completely clear the contents of the cache
207 */
208void clear_lstat_cache(void)
209{
60b458b7 210 reset_lstat_cache();
bda6eb0d
KB
211}
212
bad4a54f
KB
213#define USE_ONLY_LSTAT 0
214
92604b46
KB
215/*
216 * Return non-zero if path 'name' has a leading symlink component
217 */
57199892 218int has_symlink_leading_path(const char *name, int len)
92604b46 219{
57199892 220 return lstat_cache(name, len,
bad4a54f 221 FL_SYMLINK|FL_DIR, USE_ONLY_LSTAT) &
92604b46 222 FL_SYMLINK;
f859c846 223}
09c93066
KB
224
225/*
226 * Return non-zero if path 'name' has a leading symlink component or
227 * if some leading path component does not exists.
228 */
57199892 229int has_symlink_or_noent_leading_path(const char *name, int len)
09c93066 230{
57199892 231 return lstat_cache(name, len,
bad4a54f 232 FL_SYMLINK|FL_NOENT|FL_DIR, USE_ONLY_LSTAT) &
09c93066
KB
233 (FL_SYMLINK|FL_NOENT);
234}
bad4a54f
KB
235
236/*
237 * Return non-zero if all path components of 'name' exists as a
238 * directory. If prefix_len > 0, we will test with the stat()
239 * function instead of the lstat() function for a prefix length of
240 * 'prefix_len', thus we then allow for symlinks in the prefix part as
241 * long as those points to real existing directories.
242 */
57199892 243int has_dirs_only_path(const char *name, int len, int prefix_len)
bad4a54f 244{
57199892 245 return lstat_cache(name, len,
bad4a54f
KB
246 FL_DIR|FL_FULLPATH, prefix_len) &
247 FL_DIR;
248}
78478927
KB
249
250static struct removal_def {
251 char path[PATH_MAX];
252 int len;
253} removal;
254
255static void do_remove_scheduled_dirs(int new_len)
256{
257 while (removal.len > new_len) {
258 removal.path[removal.len] = '\0';
259 if (rmdir(removal.path))
260 break;
261 do {
262 removal.len--;
263 } while (removal.len > new_len &&
264 removal.path[removal.len] != '/');
265 }
266 removal.len = new_len;
78478927
KB
267}
268
269void schedule_dir_for_removal(const char *name, int len)
270{
271 int match_len, last_slash, i, previous_slash;
272
273 match_len = last_slash = i =
274 longest_path_match(name, len, removal.path, removal.len,
275 &previous_slash);
276 /* Find last slash inside 'name' */
277 while (i < len) {
278 if (name[i] == '/')
279 last_slash = i;
280 i++;
281 }
282
283 /*
284 * If we are about to go down the directory tree, we check if
285 * we must first go upwards the tree, such that we then can
286 * remove possible empty directories as we go upwards.
287 */
288 if (match_len < last_slash && match_len < removal.len)
289 do_remove_scheduled_dirs(match_len);
290 /*
291 * If we go deeper down the directory tree, we only need to
292 * save the new path components as we go down.
293 */
294 if (match_len < last_slash) {
295 memcpy(&removal.path[match_len], &name[match_len],
296 last_slash - match_len);
297 removal.len = last_slash;
298 }
78478927
KB
299}
300
301void remove_scheduled_dirs(void)
302{
303 do_remove_scheduled_dirs(0);
78478927 304}