alarm / setitimer question and serial record communication

Joel Sherrill <joel@OARcorp.com> joel.sherrill at OARcorp.com
Wed Sep 22 15:22:43 UTC 2004


Peter Dufault wrote:
> 
> On Sep 22, 2004, at 9:12 AM, Joel Sherrill <joel at OARcorp.com> wrote:
> 
>> My answer in cases like this is always to fix the RTEMS implementation
>> to match the expected standard behavior.   Provide test case and
>> documentation updates.
> 
> 
> I agree, but I'm surprised "alarm()" doesn't seem to interrupt the 
> thread.  Am I missing something or is that how it is?

Right now, that's how it is.

The read() isn't implemented using an interruptible blocking operation.
Classic semaphores/mutexes by default have non-interruptible blocking.
Given their use in key parts of the system infrastructure, an option
might be to add an per-instance attribute making the blocking on that
particular one interruptible.



> Peter
> 
>>
> Peter Dufault
> HD Associates, Inc.
> 


-- 
Joel Sherrill, Ph.D.             Director of Research & Development
joel at OARcorp.com                 On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
    Support Available             (256) 722-9985




More information about the users mailing list