(flush_write): Limit filenames

of the members that straddle multivolume archive
boundary to 100 characters.

diff --git a/src/buffer.c b/src/buffer.c
index c1a88a3ee325245047c7dd62fce9c7adb19c4826..d7c6844c4fd9da9eb8873d94991334f6ad4397ce 100644 (file)
--- a/src/buffer.c
+++ b/src/buffer.c
@@ -510,13 +510,18 @@ flush_write (void)
       if (volume_label_option)
        record_start++;
 
+      if (strlen (real_s_name) > NAME_FIELD_SIZE)
+       FATAL_ERROR ((0, 0,
+                     _("%s: file name too long to be stored in a GNU multivolume header"),
+                     quotearg_colon (real_s_name)));
+      
       memset (record_start, 0, BLOCKSIZE);
 
       /* FIXME: Michael P Urban writes: [a long name file] is being written
         when a new volume rolls around [...]  Looks like the wrong value is
         being preserved in real_s_name, though.  */
 
-      strcpy (record_start->header.name, real_s_name);
+      strncpy (record_start->header.name, real_s_name, NAME_FIELD_SIZE);
       record_start->header.typeflag = GNUTYPE_MULTIVOL;
       OFF_TO_CHARS (real_s_sizeleft, record_start->header.size);
       OFF_TO_CHARS (real_s_totsize - real_s_sizeleft,
@@ -763,7 +768,7 @@ flush_read (void)
        {
          uintmax_t s1, s2;
          if (cursor->header.typeflag != GNUTYPE_MULTIVOL
-             || strcmp (cursor->header.name, real_s_name))
+             || strncmp (cursor->header.name, real_s_name, NAME_FIELD_SIZE))
            {
              WARN ((0, 0, _("%s is not continued on this volume"),
                     quote (real_s_name)));