*
* Copyright (C) Linus Torvalds, 2005
*/
-#include <stdarg.h>
#include "cache.h"
struct cache_entry **active_cache = NULL;
+static time_t index_file_timestamp;
unsigned int active_nr = 0, active_alloc = 0, active_cache_changed = 0;
-int ce_match_stat(struct cache_entry *ce, struct stat *st)
+/*
+ * This only updates the "non-critical" parts of the directory
+ * cache, ie the parts that aren't tracked by GIT, and only used
+ * to validate the cache.
+ */
+void fill_stat_cache_info(struct cache_entry *ce, struct stat *st)
+{
+ ce->ce_ctime.sec = htonl(st->st_ctime);
+ ce->ce_mtime.sec = htonl(st->st_mtime);
+#ifdef USE_NSEC
+ ce->ce_ctime.nsec = htonl(st->st_ctim.tv_nsec);
+ ce->ce_mtime.nsec = htonl(st->st_mtim.tv_nsec);
+#endif
+ ce->ce_dev = htonl(st->st_dev);
+ ce->ce_ino = htonl(st->st_ino);
+ ce->ce_uid = htonl(st->st_uid);
+ ce->ce_gid = htonl(st->st_gid);
+ ce->ce_size = htonl(st->st_size);
+
+ if (assume_unchanged)
+ ce->ce_flags |= htons(CE_VALID);
+}
+
+static int ce_compare_data(struct cache_entry *ce, struct stat *st)
+{
+ int match = -1;
+ int fd = open(ce->name, O_RDONLY);
+
+ if (fd >= 0) {
+ unsigned char sha1[20];
+ if (!index_fd(sha1, fd, st, 0, NULL))
+ match = memcmp(sha1, ce->sha1, 20);
+ close(fd);
+ }
+ return match;
+}
+
+static int ce_compare_link(struct cache_entry *ce, unsigned long expected_size)
+{
+ int match = -1;
+ char *target;
+ void *buffer;
+ unsigned long size;
+ char type[10];
+ int len;
+
+ target = xmalloc(expected_size);
+ len = readlink(ce->name, target, expected_size);
+ if (len != expected_size) {
+ free(target);
+ return -1;
+ }
+ buffer = read_sha1_file(ce->sha1, type, &size);
+ if (!buffer) {
+ free(target);
+ return -1;
+ }
+ if (size == expected_size)
+ match = memcmp(buffer, target, size);
+ free(buffer);
+ free(target);
+ return match;
+}
+
+static int ce_modified_check_fs(struct cache_entry *ce, struct stat *st)
+{
+ switch (st->st_mode & S_IFMT) {
+ case S_IFREG:
+ if (ce_compare_data(ce, st))
+ return DATA_CHANGED;
+ break;
+ case S_IFLNK:
+ if (ce_compare_link(ce, st->st_size))
+ return DATA_CHANGED;
+ break;
+ default:
+ return TYPE_CHANGED;
+ }
+ return 0;
+}
+
+static int ce_match_stat_basic(struct cache_entry *ce, struct stat *st)
{
unsigned int changed = 0;
switch (ntohl(ce->ce_mode) & S_IFMT) {
case S_IFREG:
changed |= !S_ISREG(st->st_mode) ? TYPE_CHANGED : 0;
- /* We consider only the owner x bit to be relevant for "mode changes" */
- if (0100 & (ntohl(ce->ce_mode) ^ st->st_mode))
+ /* We consider only the owner x bit to be relevant for
+ * "mode changes"
+ */
+ if (trust_executable_bit &&
+ (0100 & (ntohl(ce->ce_mode) ^ st->st_mode)))
changed |= MODE_CHANGED;
break;
case S_IFLNK:
if (ce->ce_ctime.sec != htonl(st->st_ctime))
changed |= CTIME_CHANGED;
-#ifdef NSEC
+#ifdef USE_NSEC
/*
* nsec seems unreliable - not all filesystems support it, so
* as long as it is in the inode cache you get right nsec
if (ce->ce_uid != htonl(st->st_uid) ||
ce->ce_gid != htonl(st->st_gid))
changed |= OWNER_CHANGED;
- if (ce->ce_dev != htonl(st->st_dev) ||
- ce->ce_ino != htonl(st->st_ino))
+ if (ce->ce_ino != htonl(st->st_ino))
changed |= INODE_CHANGED;
+
+#ifdef USE_STDEV
+ /*
+ * st_dev breaks on network filesystems where different
+ * clients will have different views of what "device"
+ * the filesystem is on
+ */
+ if (ce->ce_dev != htonl(st->st_dev))
+ changed |= INODE_CHANGED;
+#endif
+
if (ce->ce_size != htonl(st->st_size))
changed |= DATA_CHANGED;
+
return changed;
}
+int ce_match_stat(struct cache_entry *ce, struct stat *st, int ignore_valid)
+{
+ unsigned int changed;
+
+ /*
+ * If it's marked as always valid in the index, it's
+ * valid whatever the checked-out copy says.
+ */
+ if (!ignore_valid && (ce->ce_flags & htons(CE_VALID)))
+ return 0;
+
+ changed = ce_match_stat_basic(ce, st);
+
+ /*
+ * Within 1 second of this sequence:
+ * echo xyzzy >file && git-update-index --add file
+ * running this command:
+ * echo frotz >file
+ * would give a falsely clean cache entry. The mtime and
+ * length match the cache, and other stat fields do not change.
+ *
+ * We could detect this at update-index time (the cache entry
+ * being registered/updated records the same time as "now")
+ * and delay the return from git-update-index, but that would
+ * effectively mean we can make at most one commit per second,
+ * which is not acceptable. Instead, we check cache entries
+ * whose mtime are the same as the index file timestamp more
+ * carefully than others.
+ */
+ if (!changed &&
+ index_file_timestamp &&
+ index_file_timestamp <= ntohl(ce->ce_mtime.sec))
+ changed |= ce_modified_check_fs(ce, st);
+
+ return changed;
+}
+
+int ce_modified(struct cache_entry *ce, struct stat *st, int really)
+{
+ int changed, changed_fs;
+ changed = ce_match_stat(ce, st, really);
+ if (!changed)
+ return 0;
+ /*
+ * If the mode or type has changed, there's no point in trying
+ * to refresh the entry - it's not going to match
+ */
+ if (changed & (MODE_CHANGED | TYPE_CHANGED))
+ return changed;
+
+ /* Immediately after read-tree or update-index --cacheinfo,
+ * the length field is zero. For other cases the ce_size
+ * should match the SHA1 recorded in the index entry.
+ */
+ if ((changed & DATA_CHANGED) && ce->ce_size != htonl(0))
+ return changed;
+
+ changed_fs = ce_modified_check_fs(ce, st);
+ if (changed_fs)
+ return changed | changed_fs;
+ return 0;
+}
+
+int base_name_compare(const char *name1, int len1, int mode1,
+ const char *name2, int len2, int mode2)
+{
+ unsigned char c1, c2;
+ int len = len1 < len2 ? len1 : len2;
+ int cmp;
+
+ cmp = memcmp(name1, name2, len);
+ if (cmp)
+ return cmp;
+ c1 = name1[len];
+ c2 = name2[len];
+ if (!c1 && S_ISDIR(mode1))
+ c1 = '/';
+ if (!c2 && S_ISDIR(mode2))
+ c2 = '/';
+ return (c1 < c2) ? -1 : (c1 > c2) ? 1 : 0;
+}
+
int cache_name_compare(const char *name1, int flags1, const char *name2, int flags2)
{
int len1 = flags1 & CE_NAMEMASK;
return -1;
if (len1 > len2)
return 1;
+
+ /* Compare stages */
+ flags1 &= CE_STAGEMASK;
+ flags2 &= CE_STAGEMASK;
+
if (flags1 < flags2)
return -1;
if (flags1 > flags2)
while (last > first) {
int next = (last + first) >> 1;
struct cache_entry *ce = active_cache[next];
- int cmp = cache_name_compare(name, namelen, ce->name, htons(ce->ce_flags));
+ int cmp = cache_name_compare(name, namelen, ce->name, ntohs(ce->ce_flags));
if (!cmp)
return next;
if (cmp < 0) {
}
/* Remove entry, return true if there are more entries to go.. */
-int remove_entry_at(int pos)
+int remove_cache_entry_at(int pos)
{
active_cache_changed = 1;
active_nr--;
return 1;
}
-int remove_file_from_cache(char *path)
+int remove_file_from_cache(const char *path)
{
int pos = cache_name_pos(path, strlen(path));
if (pos < 0)
pos = -pos-1;
while (pos < active_nr && !strcmp(active_cache[pos]->name, path))
- remove_entry_at(pos);
+ remove_cache_entry_at(pos);
return 0;
}
-int same_name(struct cache_entry *a, struct cache_entry *b)
+int ce_same_name(struct cache_entry *a, struct cache_entry *b)
{
int len = ce_namelen(a);
return ce_namelen(b) == len && !memcmp(a->name, b->name, len);
}
-/* We may be in a situation where we already have path/file and path
- * is being added, or we already have path and path/file is being
- * added. Either one would result in a nonsense tree that has path
- * twice when git-write-tree tries to write it out. Prevent it.
- *
- * If ok-to-replace is specified, we remove the conflicting entries
- * from the cache so the caller should recompute the insert position.
- * When this happens, we return non-zero.
+int ce_path_match(const struct cache_entry *ce, const char **pathspec)
+{
+ const char *match, *name;
+ int len;
+
+ if (!pathspec)
+ return 1;
+
+ len = ce_namelen(ce);
+ name = ce->name;
+ while ((match = *pathspec++) != NULL) {
+ int matchlen = strlen(match);
+ if (matchlen > len)
+ continue;
+ if (memcmp(name, match, matchlen))
+ continue;
+ if (matchlen && name[matchlen-1] == '/')
+ return 1;
+ if (name[matchlen] == '/' || !name[matchlen])
+ return 1;
+ if (!matchlen)
+ return 1;
+ }
+ return 0;
+}
+
+/*
+ * Do we have another file that has the beginning components being a
+ * proper superset of the name we're trying to add?
*/
-static int check_file_directory_conflict(const struct cache_entry *ce,
- int ok_to_replace)
+static int has_file_name(const struct cache_entry *ce, int pos, int ok_to_replace)
{
- int pos, replaced = 0;
- const char *path = ce->name;
- int namelen = strlen(path);
+ int retval = 0;
+ int len = ce_namelen(ce);
int stage = ce_stage(ce);
- char *pathbuf = xmalloc(namelen + 1);
- char *cp;
+ const char *name = ce->name;
- memcpy(pathbuf, path, namelen + 1);
+ while (pos < active_nr) {
+ struct cache_entry *p = active_cache[pos++];
- /*
- * We are inserting path/file. Do they have path registered at
- * the same stage? We need to do this for all the levels of our
- * subpath.
- */
- cp = pathbuf;
- while (1) {
- char *ep = strchr(cp, '/');
- if (!ep)
+ if (len >= ce_namelen(p))
break;
- *ep = 0; /* first cut it at slash */
- pos = cache_name_pos(pathbuf,
- htons(create_ce_flags(ep-cp, stage)));
- if (0 <= pos) {
- /* Our leading path component is registered as a file,
- * and we are trying to make it a directory. This is
- * bad.
- */
- if (!ok_to_replace) {
- free(pathbuf);
- return -1;
- }
- fprintf(stderr, "removing file '%s' to replace it with a directory to create '%s'.\n", pathbuf, path);
- remove_entry_at(pos);
- replaced = 1;
- }
- *ep = '/'; /* then restore it and go downwards */
- cp = ep + 1;
+ if (memcmp(name, p->name, len))
+ break;
+ if (ce_stage(p) != stage)
+ continue;
+ if (p->name[len] != '/')
+ continue;
+ retval = -1;
+ if (!ok_to_replace)
+ break;
+ remove_cache_entry_at(--pos);
}
- free(pathbuf);
+ return retval;
+}
- /* Do we have an entry in the cache that makes our path a prefix
- * of it? That is, are we creating a file where they already expect
- * a directory there?
- */
- pos = cache_name_pos(path,
- htons(create_ce_flags(namelen, stage)));
+/*
+ * Do we have another file with a pathname that is a proper
+ * subset of the name we're trying to add?
+ */
+static int has_dir_name(const struct cache_entry *ce, int pos, int ok_to_replace)
+{
+ int retval = 0;
+ int stage = ce_stage(ce);
+ const char *name = ce->name;
+ const char *slash = name + ce_namelen(ce);
- /* (0 <= pos) cannot happen because add_cache_entry()
- * should have taken care of that case.
- */
- pos = -pos-1;
+ for (;;) {
+ int len;
- /* pos would point at an existing entry that would come immediately
- * after our path. It could be the same as our path in higher stage,
- * or different path but in a lower stage.
- *
- * E.g. when we are inserting path at stage 2,
- *
- * 1 path
- * pos-> 3 path
- * 2 path/file1
- * 3 path/file1
- * 2 path/file2
- * 2 patho
- *
- * We need to examine pos, ignore it because it is at different
- * stage, examine next to find the path/file at stage 2, and
- * complain. We need to do this until we are not the leading
- * path of an existing entry anymore.
- */
+ for (;;) {
+ if (*--slash == '/')
+ break;
+ if (slash <= ce->name)
+ return retval;
+ }
+ len = slash - name;
- while (pos < active_nr) {
- struct cache_entry *other = active_cache[pos];
- if (strncmp(other->name, path, namelen))
- break; /* it is not our "subdirectory" anymore */
- if ((ce_stage(other) == stage) &&
- other->name[namelen] == '/') {
- if (!ok_to_replace)
- return -1;
- fprintf(stderr, "removing file '%s' under '%s' to be replaced with a file\n", other->name, path);
- remove_entry_at(pos);
- replaced = 1;
- continue; /* cycle without updating pos */
+ pos = cache_name_pos(name, ntohs(create_ce_flags(len, stage)));
+ if (pos >= 0) {
+ retval = -1;
+ if (ok_to_replace)
+ break;
+ remove_cache_entry_at(pos);
+ continue;
+ }
+
+ /*
+ * Trivial optimization: if we find an entry that
+ * already matches the sub-directory, then we know
+ * we're ok, and we can exit.
+ */
+ pos = -pos-1;
+ while (pos < active_nr) {
+ struct cache_entry *p = active_cache[pos];
+ if ((ce_namelen(p) <= len) ||
+ (p->name[len] != '/') ||
+ memcmp(p->name, name, len))
+ break; /* not our subdirectory */
+ if (ce_stage(p) == stage)
+ /* p is at the same stage as our entry, and
+ * is a subdirectory of what we are looking
+ * at, so we cannot have conflicts at our
+ * level or anything shorter.
+ */
+ return retval;
+ pos++;
}
- pos++;
}
- return replaced;
+ return retval;
+}
+
+/* We may be in a situation where we already have path/file and path
+ * is being added, or we already have path and path/file is being
+ * added. Either one would result in a nonsense tree that has path
+ * twice when git-write-tree tries to write it out. Prevent it.
+ *
+ * If ok-to-replace is specified, we remove the conflicting entries
+ * from the cache so the caller should recompute the insert position.
+ * When this happens, we return non-zero.
+ */
+static int check_file_directory_conflict(const struct cache_entry *ce, int pos, int ok_to_replace)
+{
+ /*
+ * We check if the path is a sub-path of a subsequent pathname
+ * first, since removing those will not change the position
+ * in the array
+ */
+ int retval = has_file_name(ce, pos, ok_to_replace);
+ /*
+ * Then check if the path might have a clashing sub-directory
+ * before it.
+ */
+ return retval + has_dir_name(ce, pos, ok_to_replace);
}
int add_cache_entry(struct cache_entry *ce, int option)
int pos;
int ok_to_add = option & ADD_CACHE_OK_TO_ADD;
int ok_to_replace = option & ADD_CACHE_OK_TO_REPLACE;
- pos = cache_name_pos(ce->name, htons(ce->ce_flags));
+ int skip_df_check = option & ADD_CACHE_SKIP_DFCHECK;
+
+ pos = cache_name_pos(ce->name, ntohs(ce->ce_flags));
- /* existing match? Just replace it */
+ /* existing match? Just replace it. */
if (pos >= 0) {
active_cache_changed = 1;
active_cache[pos] = ce;
* will always replace all non-merged entries..
*/
if (pos < active_nr && ce_stage(ce) == 0) {
- while (same_name(active_cache[pos], ce)) {
+ while (ce_same_name(active_cache[pos], ce)) {
ok_to_add = 1;
- if (!remove_entry_at(pos))
+ if (!remove_cache_entry_at(pos))
break;
}
}
if (!ok_to_add)
return -1;
- if (check_file_directory_conflict(ce, ok_to_replace)) {
+ if (!skip_df_check &&
+ check_file_directory_conflict(ce, pos, ok_to_replace)) {
if (!ok_to_replace)
return -1;
- pos = cache_name_pos(ce->name, htons(ce->ce_flags));
+ pos = cache_name_pos(ce->name, ntohs(ce->ce_flags));
pos = -pos-1;
}
errno = EBUSY;
if (active_cache)
- return error("more than one cachefile");
+ return active_nr;
+
errno = ENOENT;
+ index_file_timestamp = 0;
fd = open(get_index_file(), O_RDONLY);
- if (fd < 0)
- return (errno == ENOENT) ? 0 : error("open failed");
+ if (fd < 0) {
+ if (errno == ENOENT)
+ return 0;
+ die("index file open failed (%s)", strerror(errno));
+ }
size = 0; // avoid gcc warning
- map = (void *)-1;
+ map = MAP_FAILED;
if (!fstat(fd, &st)) {
size = st.st_size;
errno = EINVAL;
map = mmap(NULL, size, PROT_READ | PROT_WRITE, MAP_PRIVATE, fd, 0);
}
close(fd);
- if (-1 == (int)(long)map)
- return error("mmap failed");
+ if (map == MAP_FAILED)
+ die("index file mmap failed (%s)", strerror(errno));
hdr = map;
if (verify_hdr(hdr, size) < 0)
offset = offset + ce_size(ce);
active_cache[i] = ce;
}
+ index_file_timestamp = st.st_mtime;
return active_nr;
unmap:
munmap(map, size);
errno = EINVAL;
- return error("verify header failed");
+ die("index file corrupt");
}
#define WRITE_BUFFER_SIZE 8192
-static char write_buffer[WRITE_BUFFER_SIZE];
+static unsigned char write_buffer[WRITE_BUFFER_SIZE];
static unsigned long write_buffer_len;
static int ce_write(SHA_CTX *context, int fd, void *data, unsigned int len)
SHA1_Update(context, write_buffer, left);
}
+ /* Flush first if not enough space for SHA1 signature */
+ if (left + 20 > WRITE_BUFFER_SIZE) {
+ if (write(fd, write_buffer, left) != left)
+ return -1;
+ left = 0;
+ }
+
/* Append the SHA1 signature at the end */
SHA1_Final(write_buffer + left, context);
left += 20;
return 0;
}
+static void ce_smudge_racily_clean_entry(struct cache_entry *ce)
+{
+ /*
+ * The only thing we care about in this function is to smudge the
+ * falsely clean entry due to touch-update-touch race, so we leave
+ * everything else as they are. We are called for entries whose
+ * ce_mtime match the index file mtime.
+ */
+ struct stat st;
+
+ if (lstat(ce->name, &st) < 0)
+ return;
+ if (ce_match_stat_basic(ce, &st))
+ return;
+ if (ce_modified_check_fs(ce, &st)) {
+ /* This is "racily clean"; smudge it. Note that this
+ * is a tricky code. At first glance, it may appear
+ * that it can break with this sequence:
+ *
+ * $ echo xyzzy >frotz
+ * $ git-update-index --add frotz
+ * $ : >frotz
+ * $ sleep 3
+ * $ echo filfre >nitfol
+ * $ git-update-index --add nitfol
+ *
+ * but it does not. Whe the second update-index runs,
+ * it notices that the entry "frotz" has the same timestamp
+ * as index, and if we were to smudge it by resetting its
+ * size to zero here, then the object name recorded
+ * in index is the 6-byte file but the cached stat information
+ * becomes zero --- which would then match what we would
+ * obtain from the filesystem next time we stat("frotz").
+ *
+ * However, the second update-index, before calling
+ * this function, notices that the cached size is 6
+ * bytes and what is on the filesystem is an empty
+ * file, and never calls us, so the cached size information
+ * for "frotz" stays 6 which does not match the filesystem.
+ */
+ ce->ce_size = htonl(0);
+ }
+}
+
int write_cache(int newfd, struct cache_entry **cache, int entries)
{
SHA_CTX c;
struct cache_header hdr;
- int i;
+ int i, removed;
+
+ for (i = removed = 0; i < entries; i++)
+ if (!cache[i]->ce_mode)
+ removed++;
hdr.hdr_signature = htonl(CACHE_SIGNATURE);
hdr.hdr_version = htonl(2);
- hdr.hdr_entries = htonl(entries);
+ hdr.hdr_entries = htonl(entries - removed);
SHA1_Init(&c);
if (ce_write(&c, newfd, &hdr, sizeof(hdr)) < 0)
for (i = 0; i < entries; i++) {
struct cache_entry *ce = cache[i];
+ if (!ce->ce_mode)
+ continue;
+ if (index_file_timestamp &&
+ index_file_timestamp <= ntohl(ce->ce_mtime.sec))
+ ce_smudge_racily_clean_entry(ce);
if (ce_write(&c, newfd, ce, ce_size(ce)) < 0)
return -1;
}