e2fsck: speed up runs when using tdb for large atypical filesystems

Optimize how the tdb library so that running with [scratch_files] in
/etc/e2fsck.conf is more efficient.  Use a better hash function,
supplied by Rogier Wolff, and supply an estimate of the size of the
hash table to tdb_open instead of using the default (which is way too
small in most cases).  Also, disable the tdb locking and fsync calls,
since it's not necessary for our use in this case (which is
essentially as cheap swap space; the tdb files do not contain
persistent data.)

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>

diff --git a/e2fsck/dirinfo.c b/e2fsck/dirinfo.c
index e98795cfb75687882a7b5c10f94facc118716596..c81777a50ad494cb949a741ee3723d95b136a18a 100644 (file)
--- a/e2fsck/dirinfo.c
+++ b/e2fsck/dirinfo.c
@@ -67,7 +67,11 @@ static void setup_tdb(e2fsck_t ctx, ext2_ino_t num_dirs)
                db->tdb = NULL;
                return;
        }
-       db->tdb = tdb_open(db->tdb_fn, 0, TDB_CLEAR_IF_FIRST,
+
+       if (num_dirs < 99991)
+               num_dirs = 99991; /* largest 5 digit prime */
+
+       db->tdb = tdb_open(db->tdb_fn, num_dirs, TDB_NOLOCK | TDB_NOSYNC,
                           O_RDWR | O_CREAT | O_TRUNC, 0600);
        close(fd);
 }

diff --git a/lib/ext2fs/icount.c b/lib/ext2fs/icount.c
index 1a18b3777b7f2107f9ae2e0cea547e41414aff93..5d64ac4323363d174a45c72f79435c8f0b9e5c8a 100644 (file)
--- a/lib/ext2fs/icount.c
+++ b/lib/ext2fs/icount.c
@@ -180,6 +180,7 @@ errcode_t ext2fs_create_icount_tdb(ext2_filsys fs, char *tdb_dir,
        ext2_icount_t   icount;
        errcode_t       retval;
        char            *fn, uuid[40];
+       ext2_ino_t      num_inodes;
        int             fd;
 
        retval = alloc_icount(fs, flags,  &icount);
@@ -193,8 +194,18 @@ errcode_t ext2fs_create_icount_tdb(ext2_filsys fs, char *tdb_dir,
        sprintf(fn, "%s/%s-icount-XXXXXX", tdb_dir, uuid);
        fd = mkstemp(fn);
 
+       /*
+        * This is an overestimate of the size that we will need; the
+        * ideal value is the number of used inodes with a count
+        * greater than 1.  OTOH the times when we really need this is
+        * with the backup programs that use lots of hard links, in
+        * which case the number of inodes in use approaches the ideal
+        * value.
+        */
+       num_inodes = fs->super->s_inodes_count - fs->super->s_free_inodes_count;
+
        icount->tdb_fn = fn;
-       icount->tdb = tdb_open(fn, 0, TDB_CLEAR_IF_FIRST,
+       icount->tdb = tdb_open(fn, num_inodes, TDB_NOLOCK | TDB_NOSYNC,
                               O_RDWR | O_CREAT | O_TRUNC, 0600);
        if (icount->tdb) {
                close(fd);

diff --git a/lib/ext2fs/tdb.c b/lib/ext2fs/tdb.c
index 79534eb457c00f919f8fc9baf6fcf9378129cf86..83c834e577dd850aeca10cd16243de7bf95c1076 100644 (file)
--- a/lib/ext2fs/tdb.c
+++ b/lib/ext2fs/tdb.c
@@ -3711,17 +3711,17 @@ void tdb_enable_seqnum(struct tdb_context *tdb)
 static struct tdb_context *tdbs = NULL;
 
 
-/* This is based on the hash algorithm from gdbm */
+/* This is from a hash algorithm suggested by Rogier Wolff */
 static unsigned int default_tdb_hash(TDB_DATA *key)
 {
        u32 value;      /* Used to compute the hash value.  */
        u32   i;        /* Used to cycle through random values. */
 
        /* Set the initial value from the key size. */
-       for (value = 0x238F13AF * key->dsize, i=0; i < key->dsize; i++)
-               value = (value + (key->dptr[i] << (i*5 % 24)));
+       for (value = 0, i=0; i < key->dsize; i++)
+               value = value * 256 + key->dptr[i] + (value >> 24) * 241;
 
-       return (1103515243 * value + 12345);
+       return value;
 }