Experimental Physics and
Industrial Control System

Andrew Johnson via Core-talk <[email protected]> · Tue, 02 Jul 2019 23:18:09 -0000

Review: Approve

> The change should only result in transient resource leaks 

Oops, sorry - you're right, I assumed that things would break if epicsThreadMustJoin() was a no-op, but looking more closely I guess they shouldn't, sorry!  I have not looked in detail at the shutdown logic changes in dbEvent.c, I trust you to have tested that reasonably.

I have significantly adjusted and expanded the Release Notes again.

I have changed the EPICS_THREAD_CAN_JOIN macro, it is now only defined when a working epicsThreadMustJoin() is available, so external code can use a simple #ifdef to determine that the API is available and works on this Base version & target.

> ... EPICS_THREAD_OPTS_INIT ...

Done. External code can now use #ifdef EPICS_THREAD_OPTS_INIT to determine whether the epicsThreadCreateOpt() API is available on this Base version.

I just added another test to epicsThreadTest.cpp to check that a join actually delays the caller; the empty epicsThreadMustJoin() as compiled for VxWorks < 6.9 will return immediately. This check should detect that and fail on VxWorks 6.8 say, but is marked as ToDo when that's expected.

I'm happy with this now, others may wish to take a final look given my recent updates.
-- 
https://code.launchpad.net/~epics-core/epics-base/+git/Com/+merge/361379
Your team EPICS Core Developers is subscribed to branch epics-base:7.0.

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System