@@ -316,7 +316,12 @@ void do_dx_hash(int argc, char *argv[], int sci_idx EXT2FS_ATTR((unused)),
int hash_flags = 0;
const struct ext2fs_nls_table *encoding = NULL;
- hash_seed[0] = hash_seed[1] = hash_seed[2] = hash_seed[3] = 0;
+ hash_seed[0] = current_fs->super->s_hash_seed[0];
+ hash_seed[1] = current_fs->super->s_hash_seed[1];
+ hash_seed[2] = current_fs->super->s_hash_seed[2];
+ hash_seed[3] = current_fs->super->s_hash_seed[3];
+
+ hash_version = current_fs->super->s_def_hash_version;
reset_getopt();
while ((c = getopt(argc, argv, "h:s:ce:")) != EOF) {
debugfs hash command computes the hash for the given filename. It takes hash_seed and hash_version (i.e hash algorithm) as arguments. User has to refer to the superblock to get these values used by the filesystem. If the arguments are not given then debugfs computes hash assuming both hash_seed and hash_version are zeros. In most of the cases this assumption will be different from the actual hash_seed and hash_version used by the filesystem. In general user will be in need of hash computed from hash_seed and hash_version of the filesystem. So, instead of assuming hash_seed and hash_version as zero when the arguments are not provided, read these directly from the superblock to simplify the task of user. Example: Before:- debugfs: hash -s 524e5394-e2a3-43fa-b192-79720b1fe3e1 -h half_md4 file1 Hash of file1 is 0x4a8d8c94 (minor 0x17a37f43) After improvement:- debugfs: hash file1 Hash of file1 is 0x4a8d8c94 (minor 0x17a37f43) Signed-off-by: Srivathsa Dara <srivathsa.d.dara@oracle.com> --- debugfs/htree.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)