librpc idl: netlogon netr_identity_info logon_id to 64 bit
[sfrench/samba-autobuild/.git] / librpc / idl / netlogon.idl
index 22f86b9207668a5e505ba8ee13bddb410a5f0116..b4474f7ec496dd1d5b221e79a00ba6bd54de7802 100644 (file)
@@ -133,11 +133,35 @@ interface netlogon
                MSV1_0_SUBAUTHENTICATION_DLL_EX         = 0x00100000
        } netr_LogonParameterControl;
 
+       /* Summary of the of the Query and Response from Microsoft on
+        * the usage of logon_id in netr_IdendityInfo
+        *
+        * [REG:119013019612095] [MS-NRPC]: NETLOGON_LOGON_IDENTITY_INFO: Does
+        * the Reserved field have LogonId meaning?
+        *
+        * Questions:
+        *   In NetrLogonSamLogonEx does the Reserved field
+        *   (of NETLOGON_LOGON_IDENTITY_INFO) have LogonId meaning?
+        *
+        *   What is a valid LogonID, and does have any audit usage?
+        *
+        *   Samba is sending a constant "deadbeef" in hex and would like to
+        *   understand any usage of this field.
+        *
+        * Response:
+        *   The NRPC spec is accurate in defining the field as Reserved, and
+        *   without protocol significance. In the header file in our source
+        *   code, it is defined as LogonId and commented as such, but it’s
+        *   effectively not used. This is probably why the API structure has
+        *   that field name. It may have been intended as such but it’s not
+        *    used.
+        *
+        * Samba now sends a random value in this field.
+        */
        typedef struct {
                lsa_String  domain_name;
                netr_LogonParameterControl parameter_control; /* see MSV1_0_* */
-               uint32      logon_id_low;
-               uint32      logon_id_high;
+               udlong logon_id;
                lsa_String  account_name;
                lsa_String  workstation;
        } netr_IdentityInfo;