Check iExit when trying to connect to objects #200
Conversation
konglobemeralt
requested review from
LukasWikander,
alfaro01,
sepast and
viktorjo
There was a problem hiding this comment.
This change does not fix the problem of ObjectControl being stuck in "Was not able to connect to object, retry in 3 sec...". (for a test, try e.g. not starting a VO on the specified IP address and then pressing connect, then shutting down the GUC and ctrl-C the server)
konglobemeralt
changed the title
There was a problem hiding this comment.
I think that a few changes should be applied before approving. I agree with the statement that we should do something more than reporting, however I would like it if we could have the ability of canceling the connection to objects after this pull-request is done.
server/src/objectcontrol.c
Outdated
| @@ -872,7 +872,7 @@ void objectcontrol_task(TimeType * GPSTime, GSDType * GSD, LOG_LEVEL logLevel) { | |||
|
|
|||
| } | |||
|
|
|||
| } while (iResult < 0 && DisconnectU8 == 0); | |||
| } while (iExit == 0 && iResult < 0 && DisconnectU8 == 0); | |||
There was a problem hiding this comment.
In this do-while loop I think that the procedure should a bit different. I think that the for-loop on line 814 should be included inside of this, so that in each iteration of the do-while, the program checks all objects and reports which could not be connected. If this is implemented together with the removal of the sleep function, then the user will be able to at least disconnect from all objects.
There was a problem hiding this comment.
Do you mean we should have it the other way around? Right now the for-loop is inside of the do-while
There was a problem hiding this comment.
I didn't see that the for-loop was that large. We cannot change this properly now, but we have to change that in the future.
There was a problem hiding this comment.
have you added a new task? Can add this otherwise.
There was a problem hiding this comment.
I have it on my todo list :) I will do it.
|
This pull request seems to be stuck in format checking on jenkins, for some reason it fails. Could be worth investigating. |
@viktorjo This is an issue we have on every branch either I or Jesper pushes. I think Lukas has some settings configured to his Indent version which Jenkins relies on in order to accept the format. Me and Lukas looked at this yesterday and hopefully we can fix this soon. |
Big change in connection loop in OC