]> git.ipfire.org Git - thirdparty/git.git/blobdiff - read-cache.c
Merge branch 'js/empty-index-fixes'
[thirdparty/git.git] / read-cache.c
index d0a3c9082b18bae56041f39b155f08991a440da2..27703e144658da858f627a24236b046f84b370bb 100644 (file)
@@ -3,24 +3,41 @@
  *
  * Copyright (C) Linus Torvalds, 2005
  */
-#include "cache.h"
+#include "git-compat-util.h"
+#include "alloc.h"
+#include "bulk-checkin.h"
 #include "config.h"
+#include "date.h"
 #include "diff.h"
 #include "diffcore.h"
+#include "hex.h"
 #include "tempfile.h"
 #include "lockfile.h"
 #include "cache-tree.h"
 #include "refs.h"
 #include "dir.h"
-#include "object-store.h"
+#include "object-file.h"
+#include "object-store-ll.h"
+#include "oid-array.h"
 #include "tree.h"
 #include "commit.h"
 #include "blob.h"
+#include "environment.h"
+#include "gettext.h"
+#include "mem-pool.h"
+#include "name-hash.h"
+#include "object-name.h"
+#include "path.h"
+#include "preload-index.h"
+#include "read-cache.h"
 #include "resolve-undo.h"
+#include "revision.h"
 #include "run-command.h"
 #include "strbuf.h"
+#include "trace2.h"
 #include "varint.h"
 #include "split-index.h"
+#include "symlinks.h"
 #include "utf8.h"
 #include "fsmonitor.h"
 #include "thread-utils.h"
@@ -29,6 +46,7 @@
 #include "csum-file.h"
 #include "promisor-remote.h"
 #include "hook.h"
+#include "wrapper.h"
 
 /* Mask for the name length in ce_flags in the on-disk index */
 
@@ -163,61 +181,6 @@ void rename_index_entry_at(struct index_state *istate, int nr, const char *new_n
                add_index_entry(istate, new_entry, ADD_CACHE_OK_TO_ADD|ADD_CACHE_OK_TO_REPLACE);
 }
 
-void fill_stat_data(struct stat_data *sd, struct stat *st)
-{
-       sd->sd_ctime.sec = (unsigned int)st->st_ctime;
-       sd->sd_mtime.sec = (unsigned int)st->st_mtime;
-       sd->sd_ctime.nsec = ST_CTIME_NSEC(*st);
-       sd->sd_mtime.nsec = ST_MTIME_NSEC(*st);
-       sd->sd_dev = st->st_dev;
-       sd->sd_ino = st->st_ino;
-       sd->sd_uid = st->st_uid;
-       sd->sd_gid = st->st_gid;
-       sd->sd_size = st->st_size;
-}
-
-int match_stat_data(const struct stat_data *sd, struct stat *st)
-{
-       int changed = 0;
-
-       if (sd->sd_mtime.sec != (unsigned int)st->st_mtime)
-               changed |= MTIME_CHANGED;
-       if (trust_ctime && check_stat &&
-           sd->sd_ctime.sec != (unsigned int)st->st_ctime)
-               changed |= CTIME_CHANGED;
-
-#ifdef USE_NSEC
-       if (check_stat && sd->sd_mtime.nsec != ST_MTIME_NSEC(*st))
-               changed |= MTIME_CHANGED;
-       if (trust_ctime && check_stat &&
-           sd->sd_ctime.nsec != ST_CTIME_NSEC(*st))
-               changed |= CTIME_CHANGED;
-#endif
-
-       if (check_stat) {
-               if (sd->sd_uid != (unsigned int) st->st_uid ||
-                       sd->sd_gid != (unsigned int) st->st_gid)
-                       changed |= OWNER_CHANGED;
-               if (sd->sd_ino != (unsigned int) 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 (check_stat && sd->sd_dev != (unsigned int) st->st_dev)
-                       changed |= INODE_CHANGED;
-#endif
-
-       if (sd->sd_size != (unsigned int) st->st_size)
-               changed |= DATA_CHANGED;
-
-       return changed;
-}
-
 /*
  * This only updates the "non-critical" parts of the directory
  * cache, ie the parts that aren't tracked by GIT, and only used
@@ -263,7 +226,7 @@ static int ce_compare_link(const struct cache_entry *ce, size_t expected_size)
        if (strbuf_readlink(&sb, ce->name, expected_size))
                return -1;
 
-       buffer = read_object_file(&ce->oid, &type, &size);
+       buffer = repo_read_object_file(the_repository, &ce->oid, &type, &size);
        if (buffer) {
                if (size == sb.len)
                        match = memcmp(buffer, sb.buf, size);
@@ -488,87 +451,30 @@ int ie_modified(struct index_state *istate,
        return 0;
 }
 
-int base_name_compare(const char *name1, size_t len1, int mode1,
-                     const char *name2, size_t len2, int mode2)
+static int cache_name_stage_compare(const char *name1, int len1, int stage1,
+                                   const char *name2, int len2, int stage2)
 {
-       unsigned char c1, c2;
-       size_t 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;
-}
-
-/*
- * df_name_compare() is identical to base_name_compare(), except it
- * compares conflicting directory/file entries as equal. Note that
- * while a directory name compares as equal to a regular file, they
- * then individually compare _differently_ to a filename that has
- * a dot after the basename (because '\0' < '.' < '/').
- *
- * This is used by routines that want to traverse the git namespace
- * but then handle conflicting entries together when possible.
- */
-int df_name_compare(const char *name1, size_t len1, int mode1,
-                   const char *name2, size_t len2, int mode2)
-{
-       unsigned char c1, c2;
-       size_t len = len1 < len2 ? len1 : len2;
-       int cmp;
-
-       cmp = memcmp(name1, name2, len);
+       cmp = name_compare(name1, len1, name2, len2);
        if (cmp)
                return cmp;
-       /* Directories and files compare equal (same length, same name) */
-       if (len1 == len2)
-               return 0;
-       c1 = name1[len];
-       if (!c1 && S_ISDIR(mode1))
-               c1 = '/';
-       c2 = name2[len];
-       if (!c2 && S_ISDIR(mode2))
-               c2 = '/';
-       if (c1 == '/' && !c2)
-               return 0;
-       if (c2 == '/' && !c1)
-               return 0;
-       return c1 - c2;
-}
 
-int name_compare(const char *name1, size_t len1, const char *name2, size_t len2)
-{
-       size_t min_len = (len1 < len2) ? len1 : len2;
-       int cmp = memcmp(name1, name2, min_len);
-       if (cmp)
-               return cmp;
-       if (len1 < len2)
+       if (stage1 < stage2)
                return -1;
-       if (len1 > len2)
+       if (stage1 > stage2)
                return 1;
        return 0;
 }
 
-int cache_name_stage_compare(const char *name1, int len1, int stage1, const char *name2, int len2, int stage2)
+int cmp_cache_name_compare(const void *a_, const void *b_)
 {
-       int cmp;
-
-       cmp = name_compare(name1, len1, name2, len2);
-       if (cmp)
-               return cmp;
+       const struct cache_entry *ce1, *ce2;
 
-       if (stage1 < stage2)
-               return -1;
-       if (stage1 > stage2)
-               return 1;
-       return 0;
+       ce1 = *((const struct cache_entry **)a_);
+       ce2 = *((const struct cache_entry **)b_);
+       return cache_name_stage_compare(ce1->name, ce1->ce_namelen, ce_stage(ce1),
+                                 ce2->name, ce2->ce_namelen, ce_stage(ce2));
 }
 
 static int index_name_stage_pos(struct index_state *istate,
@@ -2631,7 +2537,7 @@ int repo_index_has_changes(struct repository *repo,
 
        if (tree)
                cmp = tree->object.oid;
-       if (tree || !get_oid_tree("HEAD", &cmp)) {
+       if (tree || !repo_get_oid_tree(repo, "HEAD", &cmp)) {
                struct diff_options opt;
 
                repo_diff_setup(repo, &opt);
@@ -2904,6 +2810,16 @@ static int record_ieot(void)
        return !git_config_get_index_threads(&val) && val != 1;
 }
 
+enum write_extensions {
+       WRITE_NO_EXTENSION =              0,
+       WRITE_SPLIT_INDEX_EXTENSION =     1<<0,
+       WRITE_CACHE_TREE_EXTENSION =      1<<1,
+       WRITE_RESOLVE_UNDO_EXTENSION =    1<<2,
+       WRITE_UNTRACKED_CACHE_EXTENSION = 1<<3,
+       WRITE_FSMONITOR_EXTENSION =       1<<4,
+};
+#define WRITE_ALL_EXTENSIONS ((enum write_extensions)-1)
+
 /*
  * On success, `tempfile` is closed. If it is the temporary file
  * of a `struct lock_file`, we will therefore effectively perform
@@ -2912,7 +2828,7 @@ static int record_ieot(void)
  * rely on it.
  */
 static int do_write_index(struct index_state *istate, struct tempfile *tempfile,
-                         int strip_extensions, unsigned flags)
+                         enum write_extensions write_extensions, unsigned flags)
 {
        uint64_t start = getnanotime();
        struct hashfile *f;
@@ -2950,7 +2866,7 @@ static int do_write_index(struct index_state *istate, struct tempfile *tempfile,
        }
 
        if (!istate->version)
-               istate->version = get_index_format_default(the_repository);
+               istate->version = get_index_format_default(r);
 
        /* demote version 3 to version 2 when the latter suffices */
        if (istate->version == 3 || istate->version == 2)
@@ -3085,8 +3001,8 @@ static int do_write_index(struct index_state *istate, struct tempfile *tempfile,
                        return -1;
        }
 
-       if (!strip_extensions && istate->split_index &&
-           !is_null_oid(&istate->split_index->base_oid)) {
+       if (write_extensions & WRITE_SPLIT_INDEX_EXTENSION &&
+           istate->split_index) {
                struct strbuf sb = STRBUF_INIT;
 
                if (istate->sparse_index)
@@ -3100,7 +3016,8 @@ static int do_write_index(struct index_state *istate, struct tempfile *tempfile,
                if (err)
                        return -1;
        }
-       if (!strip_extensions && !drop_cache_tree && istate->cache_tree) {
+       if (write_extensions & WRITE_CACHE_TREE_EXTENSION &&
+           !drop_cache_tree && istate->cache_tree) {
                struct strbuf sb = STRBUF_INIT;
 
                cache_tree_write(&sb, istate->cache_tree);
@@ -3110,7 +3027,8 @@ static int do_write_index(struct index_state *istate, struct tempfile *tempfile,
                if (err)
                        return -1;
        }
-       if (!strip_extensions && istate->resolve_undo) {
+       if (write_extensions & WRITE_RESOLVE_UNDO_EXTENSION &&
+           istate->resolve_undo) {
                struct strbuf sb = STRBUF_INIT;
 
                resolve_undo_write(&sb, istate->resolve_undo);
@@ -3121,7 +3039,8 @@ static int do_write_index(struct index_state *istate, struct tempfile *tempfile,
                if (err)
                        return -1;
        }
-       if (!strip_extensions && istate->untracked) {
+       if (write_extensions & WRITE_UNTRACKED_CACHE_EXTENSION &&
+           istate->untracked) {
                struct strbuf sb = STRBUF_INIT;
 
                write_untracked_extension(&sb, istate->untracked);
@@ -3132,7 +3051,8 @@ static int do_write_index(struct index_state *istate, struct tempfile *tempfile,
                if (err)
                        return -1;
        }
-       if (!strip_extensions && istate->fsmonitor_last_update) {
+       if (write_extensions & WRITE_FSMONITOR_EXTENSION &&
+           istate->fsmonitor_last_update) {
                struct strbuf sb = STRBUF_INIT;
 
                write_fsmonitor_extension(&sb, istate);
@@ -3206,8 +3126,10 @@ static int commit_locked_index(struct lock_file *lk)
                return commit_lock_file(lk);
 }
 
-static int do_write_locked_index(struct index_state *istate, struct lock_file *lock,
-                                unsigned flags)
+static int do_write_locked_index(struct index_state *istate,
+                                struct lock_file *lock,
+                                unsigned flags,
+                                enum write_extensions write_extensions)
 {
        int ret;
        int was_full = istate->sparse_index == INDEX_EXPANDED;
@@ -3225,7 +3147,7 @@ static int do_write_locked_index(struct index_state *istate, struct lock_file *l
         */
        trace2_region_enter_printf("index", "do_write_index", the_repository,
                                   "%s", get_lock_file_path(lock));
-       ret = do_write_index(istate, lock->tempfile, 0, flags);
+       ret = do_write_index(istate, lock->tempfile, write_extensions, flags);
        trace2_region_leave_printf("index", "do_write_index", the_repository,
                                   "%s", get_lock_file_path(lock));
 
@@ -3254,7 +3176,7 @@ static int write_split_index(struct index_state *istate,
 {
        int ret;
        prepare_to_write_split_index(istate);
-       ret = do_write_locked_index(istate, lock, flags);
+       ret = do_write_locked_index(istate, lock, flags, WRITE_ALL_EXTENSIONS);
        finish_writing_split_index(istate);
        return ret;
 }
@@ -3329,7 +3251,7 @@ static int write_shared_index(struct index_state *istate,
 
        trace2_region_enter_printf("index", "shared/do_write_index",
                                   the_repository, "%s", get_tempfile_path(*temp));
-       ret = do_write_index(si->base, *temp, 1, flags);
+       ret = do_write_index(si->base, *temp, WRITE_NO_EXTENSION, flags);
        trace2_region_leave_printf("index", "shared/do_write_index",
                                   the_repository, "%s", get_tempfile_path(*temp));
 
@@ -3406,9 +3328,8 @@ int write_locked_index(struct index_state *istate, struct lock_file *lock,
        if ((!si && !test_split_index_env) ||
            alternate_index_output ||
            (istate->cache_changed & ~EXTMASK)) {
-               if (si)
-                       oidclr(&si->base_oid);
-               ret = do_write_locked_index(istate, lock, flags);
+               ret = do_write_locked_index(istate, lock, flags,
+                                           ~WRITE_SPLIT_INDEX_EXTENSION);
                goto out;
        }
 
@@ -3434,8 +3355,8 @@ int write_locked_index(struct index_state *istate, struct lock_file *lock,
                /* Same initial permissions as the main .git/index file */
                temp = mks_tempfile_sm(git_path("sharedindex_XXXXXX"), 0, 0666);
                if (!temp) {
-                       oidclr(&si->base_oid);
-                       ret = do_write_locked_index(istate, lock, flags);
+                       ret = do_write_locked_index(istate, lock, flags,
+                                                   ~WRITE_SPLIT_INDEX_EXTENSION);
                        goto out;
                }
                ret = write_shared_index(istate, &temp, flags);
@@ -3556,7 +3477,8 @@ void *read_blob_data_from_index(struct index_state *istate,
        }
        if (pos < 0)
                return NULL;
-       data = read_object_file(&istate->cache[pos]->oid, &type, &sz);
+       data = repo_read_object_file(the_repository, &istate->cache[pos]->oid,
+                                    &type, &sz);
        if (!data || type != OBJ_BLOB) {
                free(data);
                return NULL;
@@ -3566,35 +3488,6 @@ void *read_blob_data_from_index(struct index_state *istate,
        return data;
 }
 
-void stat_validity_clear(struct stat_validity *sv)
-{
-       FREE_AND_NULL(sv->sd);
-}
-
-int stat_validity_check(struct stat_validity *sv, const char *path)
-{
-       struct stat st;
-
-       if (stat(path, &st) < 0)
-               return sv->sd == NULL;
-       if (!sv->sd)
-               return 0;
-       return S_ISREG(st.st_mode) && !match_stat_data(sv->sd, &st);
-}
-
-void stat_validity_update(struct stat_validity *sv, int fd)
-{
-       struct stat st;
-
-       if (fstat(fd, &st) < 0 || !S_ISREG(st.st_mode))
-               stat_validity_clear(sv);
-       else {
-               if (!sv->sd)
-                       CALLOC_ARRAY(sv->sd, 1);
-               fill_stat_data(sv->sd, &st);
-       }
-}
-
 void move_index_extensions(struct index_state *dst, struct index_state *src)
 {
        dst->untracked = src->untracked;
@@ -3838,3 +3731,240 @@ void prefetch_cache_entries(const struct index_state *istate,
                                   to_fetch.oid, to_fetch.nr);
        oid_array_clear(&to_fetch);
 }
+
+static int read_one_entry_opt(struct index_state *istate,
+                             const struct object_id *oid,
+                             struct strbuf *base,
+                             const char *pathname,
+                             unsigned mode, int opt)
+{
+       int len;
+       struct cache_entry *ce;
+
+       if (S_ISDIR(mode))
+               return READ_TREE_RECURSIVE;
+
+       len = strlen(pathname);
+       ce = make_empty_cache_entry(istate, base->len + len);
+
+       ce->ce_mode = create_ce_mode(mode);
+       ce->ce_flags = create_ce_flags(1);
+       ce->ce_namelen = base->len + len;
+       memcpy(ce->name, base->buf, base->len);
+       memcpy(ce->name + base->len, pathname, len+1);
+       oidcpy(&ce->oid, oid);
+       return add_index_entry(istate, ce, opt);
+}
+
+static int read_one_entry(const struct object_id *oid, struct strbuf *base,
+                         const char *pathname, unsigned mode,
+                         void *context)
+{
+       struct index_state *istate = context;
+       return read_one_entry_opt(istate, oid, base, pathname,
+                                 mode,
+                                 ADD_CACHE_OK_TO_ADD|ADD_CACHE_SKIP_DFCHECK);
+}
+
+/*
+ * This is used when the caller knows there is no existing entries at
+ * the stage that will conflict with the entry being added.
+ */
+static int read_one_entry_quick(const struct object_id *oid, struct strbuf *base,
+                               const char *pathname, unsigned mode,
+                               void *context)
+{
+       struct index_state *istate = context;
+       return read_one_entry_opt(istate, oid, base, pathname,
+                                 mode, ADD_CACHE_JUST_APPEND);
+}
+
+/*
+ * Read the tree specified with --with-tree option
+ * (typically, HEAD) into stage #1 and then
+ * squash them down to stage #0.  This is used for
+ * --error-unmatch to list and check the path patterns
+ * that were given from the command line.  We are not
+ * going to write this index out.
+ */
+void overlay_tree_on_index(struct index_state *istate,
+                          const char *tree_name, const char *prefix)
+{
+       struct tree *tree;
+       struct object_id oid;
+       struct pathspec pathspec;
+       struct cache_entry *last_stage0 = NULL;
+       int i;
+       read_tree_fn_t fn = NULL;
+       int err;
+
+       if (repo_get_oid(the_repository, tree_name, &oid))
+               die("tree-ish %s not found.", tree_name);
+       tree = parse_tree_indirect(&oid);
+       if (!tree)
+               die("bad tree-ish %s", tree_name);
+
+       /* Hoist the unmerged entries up to stage #3 to make room */
+       /* TODO: audit for interaction with sparse-index. */
+       ensure_full_index(istate);
+       for (i = 0; i < istate->cache_nr; i++) {
+               struct cache_entry *ce = istate->cache[i];
+               if (!ce_stage(ce))
+                       continue;
+               ce->ce_flags |= CE_STAGEMASK;
+       }
+
+       if (prefix) {
+               static const char *(matchbuf[1]);
+               matchbuf[0] = NULL;
+               parse_pathspec(&pathspec, PATHSPEC_ALL_MAGIC,
+                              PATHSPEC_PREFER_CWD, prefix, matchbuf);
+       } else
+               memset(&pathspec, 0, sizeof(pathspec));
+
+       /*
+        * See if we have cache entry at the stage.  If so,
+        * do it the original slow way, otherwise, append and then
+        * sort at the end.
+        */
+       for (i = 0; !fn && i < istate->cache_nr; i++) {
+               const struct cache_entry *ce = istate->cache[i];
+               if (ce_stage(ce) == 1)
+                       fn = read_one_entry;
+       }
+
+       if (!fn)
+               fn = read_one_entry_quick;
+       err = read_tree(the_repository, tree, &pathspec, fn, istate);
+       clear_pathspec(&pathspec);
+       if (err)
+               die("unable to read tree entries %s", tree_name);
+
+       /*
+        * Sort the cache entry -- we need to nuke the cache tree, though.
+        */
+       if (fn == read_one_entry_quick) {
+               cache_tree_free(&istate->cache_tree);
+               QSORT(istate->cache, istate->cache_nr, cmp_cache_name_compare);
+       }
+
+       for (i = 0; i < istate->cache_nr; i++) {
+               struct cache_entry *ce = istate->cache[i];
+               switch (ce_stage(ce)) {
+               case 0:
+                       last_stage0 = ce;
+                       /* fallthru */
+               default:
+                       continue;
+               case 1:
+                       /*
+                        * If there is stage #0 entry for this, we do not
+                        * need to show it.  We use CE_UPDATE bit to mark
+                        * such an entry.
+                        */
+                       if (last_stage0 &&
+                           !strcmp(last_stage0->name, ce->name))
+                               ce->ce_flags |= CE_UPDATE;
+               }
+       }
+}
+
+struct update_callback_data {
+       struct index_state *index;
+       int include_sparse;
+       int flags;
+       int add_errors;
+};
+
+static int fix_unmerged_status(struct diff_filepair *p,
+                              struct update_callback_data *data)
+{
+       if (p->status != DIFF_STATUS_UNMERGED)
+               return p->status;
+       if (!(data->flags & ADD_CACHE_IGNORE_REMOVAL) && !p->two->mode)
+               /*
+                * This is not an explicit add request, and the
+                * path is missing from the working tree (deleted)
+                */
+               return DIFF_STATUS_DELETED;
+       else
+               /*
+                * Either an explicit add request, or path exists
+                * in the working tree.  An attempt to explicitly
+                * add a path that does not exist in the working tree
+                * will be caught as an error by the caller immediately.
+                */
+               return DIFF_STATUS_MODIFIED;
+}
+
+static void update_callback(struct diff_queue_struct *q,
+                           struct diff_options *opt UNUSED, void *cbdata)
+{
+       int i;
+       struct update_callback_data *data = cbdata;
+
+       for (i = 0; i < q->nr; i++) {
+               struct diff_filepair *p = q->queue[i];
+               const char *path = p->one->path;
+
+               if (!data->include_sparse &&
+                   !path_in_sparse_checkout(path, data->index))
+                       continue;
+
+               switch (fix_unmerged_status(p, data)) {
+               default:
+                       die(_("unexpected diff status %c"), p->status);
+               case DIFF_STATUS_MODIFIED:
+               case DIFF_STATUS_TYPE_CHANGED:
+                       if (add_file_to_index(data->index, path, data->flags)) {
+                               if (!(data->flags & ADD_CACHE_IGNORE_ERRORS))
+                                       die(_("updating files failed"));
+                               data->add_errors++;
+                       }
+                       break;
+               case DIFF_STATUS_DELETED:
+                       if (data->flags & ADD_CACHE_IGNORE_REMOVAL)
+                               break;
+                       if (!(data->flags & ADD_CACHE_PRETEND))
+                               remove_file_from_index(data->index, path);
+                       if (data->flags & (ADD_CACHE_PRETEND|ADD_CACHE_VERBOSE))
+                               printf(_("remove '%s'\n"), path);
+                       break;
+               }
+       }
+}
+
+int add_files_to_cache(struct repository *repo, const char *prefix,
+                      const struct pathspec *pathspec, int include_sparse,
+                      int flags)
+{
+       struct update_callback_data data;
+       struct rev_info rev;
+
+       memset(&data, 0, sizeof(data));
+       data.index = repo->index;
+       data.include_sparse = include_sparse;
+       data.flags = flags;
+
+       repo_init_revisions(repo, &rev, prefix);
+       setup_revisions(0, NULL, &rev, NULL);
+       if (pathspec)
+               copy_pathspec(&rev.prune_data, pathspec);
+       rev.diffopt.output_format = DIFF_FORMAT_CALLBACK;
+       rev.diffopt.format_callback = update_callback;
+       rev.diffopt.format_callback_data = &data;
+       rev.diffopt.flags.override_submodule_config = 1;
+       rev.max_count = 0; /* do not compare unmerged paths with stage #2 */
+
+       /*
+        * Use an ODB transaction to optimize adding multiple objects.
+        * This function is invoked from commands other than 'add', which
+        * may not have their own transaction active.
+        */
+       begin_odb_transaction();
+       run_diff_files(&rev, DIFF_RACY_IS_MODIFIED);
+       end_odb_transaction();
+
+       release_revisions(&rev);
+       return !!data.add_errors;
+}