Experimental Physics and
Industrial Control System

Kate Feng <[email protected]> · Thu, 16 Sep 2004 10:43:46 -0400

Till Straumann wrote:

> Folks.
> the caRepeater thread calls 'exit()' which is not really
> what you want to do from the ca server (I guess)
> -- T.
>

Many thanks to Till Straumann for tracking  down this bug.
A while ago, I posted the following message to tech-talk:

> The IOC hung at iocInit() after we tried to add the eighth
> motor controller cards in the crate running RTEMS-mvme2307.
> Further tests showed that,  as a wokaround,one clock tick of delay is
> needed between the end of the scaler_init_record()
> and before it entered the oms_init_record().  At this point, it seems
> that the bug  might not be  even RTEMS related.
> rtems_task_wake_after(0) works fine.

After applying this patch to my  3.14.6 EPICS base,  I do not need the
workaround of one clock tick delay in the  drvOms58.cc motor driver.
Indeed, the bug is not RTEMS related.

Thanks,
Kate

>
>   ------------------------------------------------------------------------
> *** base-3.14.6/src/ca/repeater.cpp.orig        Mon Feb  2 22:32:00 2004
> --- base-3.14.6/src/ca/repeater.cpp     Wed Sep 15 18:31:58 2004
> ***************
> *** 484,490 ****
>   /*
>    *  ca_repeater ()
>    */
> ! void ca_repeater ()
>   {
>       tsFreeList < repeaterClient, 0x20 > freeList;
>       int size;
> --- 484,490 ----
>   /*
>    *  ca_repeater ()
>    */
> ! static int ca_repeater_func ()
>   {
>       tsFreeList < repeaterClient, 0x20 > freeList;
>       int size;
> ***************
> *** 506,512 ****
>           if ( SOCKERRNO == SOCK_EADDRINUSE ) {
>               osiSockRelease ();
>               debugPrintf ( ( "CA Repeater: exiting because a repeater is already running\n" ) );
> !             exit (0);
>           }
>           char sockErrBuf[64];
>           epicsSocketConvertErrnoToString (
> --- 506,512 ----
>           if ( SOCKERRNO == SOCK_EADDRINUSE ) {
>               osiSockRelease ();
>               debugPrintf ( ( "CA Repeater: exiting because a repeater is already running\n" ) );
> !             return (0);
>           }
>           char sockErrBuf[64];
>           epicsSocketConvertErrnoToString (
> ***************
> *** 515,521 ****
>               __FILE__, sockErrBuf );
>           osiSockRelease ();
>           delete [] pBuf;
> !         exit(0);
>       }
>
>       debugPrintf ( ( "CA Repeater: Attached and initialized\n" ) );
> --- 515,521 ----
>               __FILE__, sockErrBuf );
>           osiSockRelease ();
>           delete [] pBuf;
> !         return(0);
>       }
>
>       debugPrintf ( ( "CA Repeater: Attached and initialized\n" ) );
> ***************
> *** 574,579 ****
> --- 574,586 ----
>
>           fanOut ( from, pMsg, size, freeList );
>       }
> +     return -1;
> + }
> +
> + void ca_repeater(void)
> + {
> +       if ( 0 == ca_repeater_func() )
> +               exit(0);
>   }
>
>   /*
> ***************
> *** 582,588 ****
>   extern "C" void caRepeaterThread ( void * /* pDummy */ )
>   {
>       taskwdInsert ( epicsThreadGetIdSelf(), NULL, NULL );
> !     ca_repeater ();
>   }
>
>
> --- 589,596 ----
>   extern "C" void caRepeaterThread ( void * /* pDummy */ )
>   {
>       taskwdInsert ( epicsThreadGetIdSelf(), NULL, NULL );
> !     if ( 0 == ca_repeater_func () )
> !         taskwdRemove( epicsThreadGetIdSelf() );
>   }
>
>

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System