yipee!

The spoolss AddJob function has an [in,out] buffer not an [in] buffer
(despite the comment in the code to the contrary). Also, we must fail
this function - not just blindly reply "no problem" as AddJob should
always fail on non-local printers.

This fixes a bug where the "print test page" failed about half the
time. I suspect it will also fix a bunch of other intermittent spoolss
bugs where the client (incorrectly) tries to use the AddJob printing
interface.
(This used to be commit 14e534a8907c34b53e00a63756efd71903ff9432)

1 files changed, 2 insertions, 0 deletions

diff --git a/source3/include/rpc_spoolss.h b/source3/include/rpc_spoolss.h
index 1e0a53d9e0..cb6b29cab3 100755
--- a/source3/include/rpc_spoolss.h
+++ b/source3/include/rpc_spoolss.h
@@ -958,6 +958,8 @@ SPOOL_Q_ADDJOB;
 
 typedef struct spool_r_addjob
 {
+	NEW_BUFFER *buffer;
+	uint32 needed;
 	uint32 status;
 }
 SPOOL_R_ADDJOB;