From 60f5dd5b29e701b3ab601addd3cdefb508acc7d1 Mon Sep 17 00:00:00 2001 From: Alexander Barton Date: Fri, 21 May 2010 23:44:20 +0200 Subject: [PATCH] Update comments: subprocesses not only can be resolver processes --- src/ngircd/conn.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/src/ngircd/conn.c b/src/ngircd/conn.c index 91ff7796..dff9336f 100644 --- a/src/ngircd/conn.c +++ b/src/ngircd/conn.c @@ -756,7 +756,8 @@ Conn_Handler(void) continue; /* TLS/SSL layer needs to write data; deal with this first */ #endif if (Proc_InProgress(&My_Connections[i].proc_stat)) { - /* Wait for completion of resolver sub-process ... */ + /* Wait for completion of forked subprocess + * and ignore the socket in the meantime ... */ io_event_del(My_Connections[i].sock, IO_WANTREAD); continue; @@ -772,6 +773,7 @@ Conn_Handler(void) IO_WANTREAD); continue; } + io_event_add(My_Connections[i].sock, IO_WANTREAD); } @@ -1073,7 +1075,7 @@ Conn_Close( CONN_ID Idx, const char *LogMsg, const char *FwdMsg, bool InformClie in_k, out_k); } - /* cancel running resolver */ + /* Kill possibly running subprocess */ if (Proc_InProgress(&My_Connections[Idx].proc_stat)) Proc_Kill(&My_Connections[Idx].proc_stat);