[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

gettimeofday seconds rollover problem?



> Ultimately, however, the OS' sync. primitives do disable thread- 
> dispatching and/or
> interrupts so introducing a memory barrier there should be good  
> enough.
>
> IMO, the memory barrier is fine for thread-dispatch but I'm still a  
> little bit sceptical
> about introducing it to rtems_interrupt_disable/enable (especially  
> on a RISC machine).

Can you try a couple tests and confirm your suspicions?

Your suggestion to add the barriers to the _ISR_Enable/Disable/Flash  
macros seems like a very simple and robust solution (if your  
suspicions aren't confirmed....)

>
> T.
>
>>
>>>
>>> Nobody has addressed my point # 1: adding a memory barrier
>>> to rtems_interrupt_disable/enable does not address thread- 
>>> dispatch- disable
>>> protected code. If we're going to add this safe-guard, it should
>>> be introduced to thread_dispatch_disable/enable, too.
>>

-- 
Eric Norum <norume at aps.anl.gov>
Advanced Photon Source
Argonne National Laboratory
(630) 252-4793