Re: [Bug 1896295] Re: WIN32: IOC scan loops hang

["Johnson, Andrew N. via Core-talk" <core-talk at aps.anl.gov>]

Hi Mark,

We are aiming to have a new release out no later than a week before the Codeathon (which starts March 8th) so the Base work there can be based on that. 

My apologies for the problems on Windows, I might not have been the only core developer who forgot that this was still a real issue. If I remember correctly it takes many days for this problem to appear, so this fix that I just merged could not be confirmed as fixed before then could it?

- Andrew

-- 
Complicity is easy, Simplexity takes real work

> On Feb 20, 2021, at 6:33 PM, Mark Rivers via Core-talk <core-talk at aps.anl.gov> wrote:
> 
> When will there be a release of 7.0.x that includes this fix?  I don't believe there has actually been a 7.0 release that builds and functions without serious bugs on Windows.
> 
> Mark
> 
> 
> -----Original Message-----
> From: Core-talk <core-talk-bounces at aps.anl.gov> On Behalf Of Andrew Johnson via Core-talk
> Sent: Saturday, February 20, 2021 6:18 PM
> To: core-talk at aps.anl.gov
> Subject: [Bug 1896295] Re: WIN32: IOC scan loops hang
> 
> The problem of scan threads hanging was fixed by reversion of the use of monotonic time for timers, but the underlying issue with the monotonic time support on Windows has now been fixed by Freddie's branch.
> 
> ** Changed in: epics-base
>       Status: New => Fix Committed
> 
> --
> You received this bug notification because you are a member of EPICS Core Developers, which is subscribed to EPICS Base.
> Matching subscriptions: epics-core-list-subscription
> https://bugs.launchpad.net/bugs/1896295
> 
> Title:
>  WIN32: IOC scan loops hang
> 
> Status in EPICS Base:
>  Fix Committed
> 
> Bug description:
>  The IOC scan loops hung on all IOCs on several computers, on
>  investigation this looks to be caused by the monotonic clock wrapping.
>  The values of next and now in dbScan.c / periodicTask() are consistent
>  with GetTickCount() wrapping (which happens every 49.7 days), which I
>  believe lead to epicsEventWaitWithTimeout() being called with a large
>  delay. I was initially confused by the debugger saying a delay of 0
>  had been passed to epicsEventWaitWithTimeout(), but I cannot see how
>  this is possible (the minimum delay should be penalty) so I assume the
>  debugger is reporting incorrectly due to the code being in release
>  mode.
> 
>  The epics Monotonic time functions should not normally be using
>  GetTickCount(), however it looks like the logic in the test for
>  whether to use GetPerformanceCounter() is inverted and so
>  GetTickCount() is always used. I will push a patch to fix this, and
>  also to use GetTickCount64() as a fallback instead (which wraps on a
>  much longer timescale).
> 
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/epics-base/+bug/1896295/+subscriptions