epics/pvAccess
0
0
Fork 0
Code Issues Pull Requests Projects Wiki Activity

server monitor cleanup

Browse Source 
ServerMonitorRequesterImpl::getChannelMonitor()
seems like no possibility of deadlock atm.
This commit is contained in:
Michael Davidsaver
2018-03-22 13:34:55 -07:00
parent 4fcbdff580
commit e9cc8298a6
1 changed files with 6 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
src/server/responseHandlers.cpp
View File

@@ -1985,7 +1985,7 @@ void ServerMonitorRequesterImpl::monitorConnect(const Status& status, Monitor::s
{
Lock guard(_mutex);
_status = status;
_channelMonitor = monitor; //TODO inconsistent locking for _channelMonitor
_channelMonitor = monitor;
_structure = structure;
}
TransportSender::shared_pointer thisSender = shared_from_this();
@@ -2053,7 +2053,7 @@ void ServerMonitorRequesterImpl::destroy()
Monitor::shared_pointer ServerMonitorRequesterImpl::getChannelMonitor()
{
//Lock guard(_mutex);
Lock guard(_mutex);
return _channelMonitor;
}
@@ -2082,14 +2082,14 @@ void ServerMonitorRequesterImpl::send(ByteBuffer* buffer, TransportSendControl*
}
else
{
Monitor::shared_pointer monitor = _channelMonitor;
Monitor::shared_pointer monitor(_channelMonitor);
if (!monitor)
return;
// TODO asCheck ?
MonitorElement::shared_pointer element = monitor->poll();
if (element.get())
MonitorElement::Ref element(monitor);
if (element)
{
control->startMessage((int8)CMD_MONITOR, sizeof(int32)/sizeof(int8) + 1);
buffer->putInt(_ioid);
@@ -2106,7 +2106,7 @@ void ServerMonitorRequesterImpl::send(ByteBuffer* buffer, TransportSendControl*
element->overrunBitSet->serialize(buffer, control);
}
monitor->release(element);
element.reset();
// TODO if we try to proces several monitors at once, then fairness suffers
// TODO compbine several monitors into one message (reduces payload)