Experimental Physics and
Industrial Control System

If multiple threads block on the same event then the current
implementation (posix) may spuriously wake up more than
one thread as a result of epicsEventSignal().

This is because the underlying pthread_cond_signal()/pthread_cond_wait()
explicitly have these same semantics.

case (and yield the same semantics as the RTEMS and vxWorks
variants).

-- Till

*** base-3-14-12/src/libCom/osi/os/posix/osdEvent.c.orig	2011-01-28 19:49:39.345588580 -0600
--- base-3-14-12/src/libCom/osi/os/posix/osdEvent.c	2011-01-28 19:51:58.901443789 -0600
***************
*** 133,139 ****
      status = mutexLock(&pevent->mutex);
      checkStatusQuit(status,"pthread_mutex_lock","epicsEventWait");
      /*no need for while since caller must be prepared for no work*/
!     if(!pevent->isFull) {
          status = condWait(&pevent->cond,&pevent->mutex);
          checkStatusQuit(status,"pthread_cond_wait","epicsEventWait");
      }
--- 133,139 ----
      status = mutexLock(&pevent->mutex);
      checkStatusQuit(status,"pthread_mutex_lock","epicsEventWait");
      /*no need for while since caller must be prepared for no work*/
!     while (!pevent->isFull) {
          status = condWait(&pevent->cond,&pevent->mutex);
          checkStatusQuit(status,"pthread_cond_wait","epicsEventWait");
      }
***************
*** 151,157 ****
  
      status = mutexLock(&pevent->mutex);
      checkStatusQuit(status,"pthread_mutex_lock","epicsEventWaitWithTimeout");
!     if(!pevent->isFull) {
          convertDoubleToWakeTime(timeout,&wakeTime);
          status = condTimedwait(
              &pevent->cond,&pevent->mutex,&wakeTime);
--- 151,157 ----
  
      status = mutexLock(&pevent->mutex);
      checkStatusQuit(status,"pthread_mutex_lock","epicsEventWaitWithTimeout");
!     while (!pevent->isFull && 0 == status) {
          convertDoubleToWakeTime(timeout,&wakeTime);
          status = condTimedwait(
              &pevent->cond,&pevent->mutex,&wakeTime);

Replies:

RE: epicsEvent (posix implementation) bug ? james.rowland

Re: epicsEvent (posix implementation) bug ? Andrew Johnson

Navigate by Date:

Prev: Re: epicsMessageQueue Linux PREEMPT_RT Till Straumann

Next: Motor driver for Parker Hannifin ACR series controllers? Mark Rivers

Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  <2011>  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 

Navigate by Thread:

Prev: RE: ca_create_subscription element count Jeff Hill

Next: RE: epicsEvent (posix implementation) bug ? james.rowland

Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  <2011>  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024