Author: abartlet
Date: 2005-07-27 00:25:17 +0000 (Wed, 27 Jul 2005)
New Revision: 8792

WebSVN: http://websvn.samba.org/cgi-bin/viewcvs.cgi?view=rev&root=samba&rev=8792

Log:
Clarify comments

Modified:
   branches/SAMBA_4_0/source/torture/auth/pac.c


Changeset:
Modified: branches/SAMBA_4_0/source/torture/auth/pac.c
===================================================================
--- branches/SAMBA_4_0/source/torture/auth/pac.c        2005-07-27 00:24:47 UTC 
(rev 8791)
+++ branches/SAMBA_4_0/source/torture/auth/pac.c        2005-07-27 00:25:17 UTC 
(rev 8792)
@@ -215,6 +215,9 @@
                return False;
        }
 
+       /* The krbtgt key in use when the above PAC was generated.
+        * This is an arcfour-hmac-md5 key, extracted with our 'net
+        * samdump' tool. */
        krbtgt_bytes = smbpasswd_gethexpwd(mem_ctx, 
"B286757148AF7FD252C53603A150B7E7");
        if (!krbtgt_bytes) {
                DEBUG(0, ("Could not interpret krbtgt key"));
@@ -296,10 +299,9 @@
 
        dump_data(10,validate_blob.data,validate_blob.length);
 
-       /* all we can check is the length of the buffers,
-        * to check that the alignment and padding is ok,
-        * we can't compare the bytes, because we use a different algorithm
-        * to create the pointer values
+       /* compare both the length and the data bytes after a
+        * pull/push cycle.  This ensures we use the exact same
+        * pointer, padding etc algorithms as win2k3.
         */
        if (tmp_blob.length != validate_blob.length) {
                DEBUG(0, ("PAC push failed: orignial buffer length[%u] != 
created buffer length[%u]\n",

Reply via email to