summaryrefslogtreecommitdiff
path: root/source4/librpc/rpc/dcerpc_tcp.c
diff options
context:
space:
mode:
authorAndrew Tridgell <tridge@samba.org>2004-01-20 06:07:09 +0000
committerAndrew Tridgell <tridge@samba.org>2004-01-20 06:07:09 +0000
commit8ae5b50a6e787767bc00d42533ca29d8fb136e2e (patch)
treed94e73574fb50c951921f7632fd32d9069962cc9 /source4/librpc/rpc/dcerpc_tcp.c
parent7a4da9654e30ea96b326448c3e9111c2a5604f58 (diff)
downloadsamba-8ae5b50a6e787767bc00d42533ca29d8fb136e2e.tar.gz
samba-8ae5b50a6e787767bc00d42533ca29d8fb136e2e.tar.bz2
samba-8ae5b50a6e787767bc00d42533ca29d8fb136e2e.zip
added code to the RPC-SPOOLSS test that demonstrates that policy
handles are not shared between open dcerpc connections, even when those connections are on the same SMB socket. I have tested this with w2k3, w2k and NT4. It seems that policy handles have a strict scope of the dcerpc connection on which they were opened. I realise that this goes against existing folk-law in the team, but it seems that the previous testing (I'm not sure who did this?) was wrong. Perhaps clients do send us policy handles from other connections, but if they do then the correct thing to do is to fail the operation with a dcerpc fault. I suspect that failing it with exactly the right dcerpc fault code is important. (This used to be commit 2ed24d29bafd9055d5782acdd595cd0f378a651a)
Diffstat (limited to 'source4/librpc/rpc/dcerpc_tcp.c')
-rw-r--r--source4/librpc/rpc/dcerpc_tcp.c1
1 files changed, 1 insertions, 0 deletions
diff --git a/source4/librpc/rpc/dcerpc_tcp.c b/source4/librpc/rpc/dcerpc_tcp.c
index b577260033..afdf243071 100644
--- a/source4/librpc/rpc/dcerpc_tcp.c
+++ b/source4/librpc/rpc/dcerpc_tcp.c
@@ -183,6 +183,7 @@ NTSTATUS dcerpc_pipe_open_tcp(struct dcerpc_pipe **p,
/*
fill in the transport methods
*/
+ (*p)->transport.transport = NCACN_IP_TCP;
(*p)->transport.private = NULL;
(*p)->transport.full_request = tcp_full_request;
(*p)->transport.secondary_request = tcp_secondary_request;