From e60910dec84c22eaac2bc9927f6d190d3d4f85eb Mon Sep 17 00:00:00 2001 From: "benjamin.franksen" <benjamin.franksen@helmholtz-berlin.de> Date: Thu, 10 Oct 2013 15:20:34 +0000 Subject: [PATCH] seq: removed an out-dated comment The first removed paragraph was just wrong. The second one proposed alternatives which I no longer regard as being superior. Connection loss is not supposed to happen regularly during normal operation, so the small overhead of waking all state sets is negligable in practice. --- src/seq/seq_ca.c | 12 +----------- 1 file changed, 1 insertion(+), 11 deletions(-) diff --git a/src/seq/seq_ca.c b/src/seq/seq_ca.c index b03b6cb8..8572c6ba 100644 --- a/src/seq/seq_ca.c +++ b/src/seq/seq_ca.c @@ -494,16 +494,6 @@ void seq_conn_handler(int connected, void *arg) Why each one? Because pvConnectCount and pvMonitorCount should act like monitored anonymous channels. Any state set might be using these functions inside a when-condition and it is expected - that such conditions get checked whenever these counts change. - - Another reason is the pvConnected built-in: a state set with a - when(pvConnected(var)) should be able to make progress - if the channel is now connected. - - TODO: This is really a crude solution. It would be better to post - special events reserved for pvConnectCount and pvMonitorCount - (if we could assume safe mode is always on we'd just turn them - into anonymous PVs), and to post the regular PV event for the - variable that has connected (or disconnected). */ + that such conditions get checked whenever these counts change. */ ss_wakeup(sp, 0); } -- GitLab