This is the mail archive of the ecos-discuss@sources.redhat.com mailing list for the eCos project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: at91eb40a Timer Counter


"Transferencia de Archivos" <transfers@phaber.com> writes:

> >eCos currently uses TC0 to implement its main system timer and TC2 to
> >implement a microsecond delay timer. You could probably disable the
> >latter, but disabling the system timer would only work if you have no
> >code that needs timing services of any sort. This is fairly rare, even
> > for very simple programs.  -- 
> >Nick Garnett                    eCos Kernel Architect
> >http://www.ecoscentric.com      The eCos and RedBoot experts
> 
> Thank you all, who answered my last message, and now I have other
> questions In fact I really need to get the maximum free
> timer-counters of the board, since I need a pretty fast measurement
> rate of signals.(PWM) Anyone knows, wich features are compromised
> disabling the microsecond timer?

In theory you can transfer the microsecond timer over to using the
same timer as the system clock. However, you will have to rewrite the
hal_delay_us() function in the HAL. An example of how to write the new
routine can be found in the i386 HAL.

> So the "clock ticks" I would expect
> from ecos in this case would be maximum the MCK/2 that TC0 supports?

I'm not sure what you are asking here. It is a bad idea to run the
system clock at too high a rate, since you would then swamp the CPU
with interrupts. 

> And a final question, since the Interrupt model that eCos uses as
> the documentation says, is based on the ARM model, should I expect
> that the interrupt model work would be transparent from ecos
> code?(On the AT91EB40a).
> 

eCos has its own interrupt architecture. This invoves decoding all
interrupt sources into their own ISR vectors by using the interrupt
controller. However, you can also access the underlying hardware
interrupt vectors if you really want to. However this is not
recommended.


-- 
Nick Garnett                    eCos Kernel Architect
http://www.ecoscentric.com      The eCos and RedBoot experts


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]