Fixed bug found by JF where if the amount of data to return in the prs_struct

was small enough to pass the SMBtrans max data test we weren't setting the
"more data needed" error. I was fogetting the RPC_HEADER_LEN.... I now
check the correct length (current pdu staging area length). More printer
spool enum stuff works now. This bug does not affect TNG.
Jeremy.
(This used to be commit 1c9d2c016a78b0ed11dc68adeabb709903444837)

diff --git a/source3/smbd/ipc.c b/source3/smbd/ipc.c
index fc05811ccfe03e9ede783897978144fdea06f44b..271b0e253b61305b2029190238b88376c49a0c9d 100644 (file)
--- a/source3/smbd/ipc.c
+++ b/source3/smbd/ipc.c
@@ -3215,7 +3215,7 @@ static BOOL api_rpc_trans_reply(char *outbuf, pipes_struct *p)
                return False;
        }
 
-       send_trans_reply(outbuf, NULL, 0, rdata, data_len, (int)prs_offset(&p->out_data.rdata) > data_len);
+       send_trans_reply(outbuf, NULL, 0, rdata, data_len, p->out_data.current_pdu_len > data_len);
 
        free(rdata);
        return True;