diff options
Diffstat (limited to 'source/rpc_server/srv_pipe_hnd.c')
| -rw-r--r-- | source/rpc_server/srv_pipe_hnd.c | 16 |
1 files changed, 12 insertions, 4 deletions
diff --git a/source/rpc_server/srv_pipe_hnd.c b/source/rpc_server/srv_pipe_hnd.c index 65725bc2baa..61b608c0e82 100644 --- a/source/rpc_server/srv_pipe_hnd.c +++ b/source/rpc_server/srv_pipe_hnd.c @@ -40,10 +40,13 @@ static int pipes_open; * writers more than I hate the Windows spooler service driver * writers. This gets around a combination of bugs in the spooler * and the HP 8500 PCL driver that causes a spooler spin. JRA. + * + * bumped up from 20 -> 64 after viewing traffic from WordPerfect + * 2002 running on NT 4.- SP6 */ #ifndef MAX_OPEN_SPOOLSS_PIPES -#define MAX_OPEN_SPOOLSS_PIPES 20 +#define MAX_OPEN_SPOOLSS_PIPES 64 #endif static int current_spoolss_pipes_open; @@ -267,6 +270,9 @@ pipes_struct *open_rpc_pipe_p(char *pipe_name, if (is_spoolss_pipe) current_spoolss_pipes_open++; + /* Ensure the connection isn't idled whilst this pipe is open. */ + conn->num_files_open++; + return chain_p; } @@ -497,13 +503,13 @@ authentication failed. Denying the request.\n", p->name)); } /* - * Check the data length doesn't go over the 10Mb limit. + * Check the data length doesn't go over the 15Mb limit. * increased after observing a bug in the Windows NT 4.0 SP6a * spoolsv.exe when the response to a GETPRINTERDRIVER2 RPC * will not fit in the initial buffer of size 0x1068 --jerry 22/01/2002 */ - if(prs_data_size(&p->in_data.data) + data_len > 15*1024*1024) { + if(prs_offset(&p->in_data.data) + data_len > 15*1024*1024) { DEBUG(0,("process_request_pdu: rpc data buffer too large (%u) + (%u)\n", (unsigned int)prs_data_size(&p->in_data.data), (unsigned int)data_len )); set_incoming_fault(p); @@ -957,7 +963,9 @@ BOOL close_rpc_pipe_hnd(pipes_struct *p, connection_struct *conn) ZERO_STRUCTP(p); SAFE_FREE(p); - + + conn->num_files_open--; + return True; } |