[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RTEMS ISR question - management of the ET PSR-bit on ERC32
- Date: Sat, 12 Mar 2011 13:09:34 +0100
- From: aconchillo at ice.csic.es (Aleix Conchillo Flaqué)
- Subject: RTEMS ISR question - management of the ET PSR-bit on ERC32
On 03/10/2011 06:03 PM, Matthew McBraida wrote:
> I have a question about interrupt handling with RTEMS/ERC32:
> I understand from TSC691E Users manual issue I, sections 3.7.17 / 3.8.5
> / 184.108.40.206 that: a trap causes the processor to clear ET; a synchronous
> trap when ET=0 results in an error condition that forces a reset; and
> that RETT automatically sets ET back to 1.
> Am I right to believe that the RTEMS generic ISR ?_ISR_Handler()? sets
> ET to 1 just before calling the user-trap-handler, and then clears it
> again just after it returns?
> I guess that the end result of all this is that normally for most of the
> processing time of a trap handler, a synch trap occurring will not
> result in a processor error...
> (I refer to the TSC691E Users manual instead of the TSC695F manual due
> to the far greater detail given in the former).
Right, traps are enabled before calling the user trap handler and
redisabled again after it. I just had the 4.6.6 code available, but you
can see this on lines 615 and 645, respectively.