RE: [ECOS] Timer tick length

[Gary Thomas <gthomas at redhat dot com>]

On 04-Apr-00 Grant Edwards wrote:
> 
>> > If I go with that configuration when I run the ping_test, I get
>> > "recvfrom: Operation timed out" messages quite quickly -- it
>> > takes less than 2 seconds for all 16 timeouts. Does the TCP/IP
>> > stack assume that a clock tick is 10ms?  (On my system it's
>> > 1ms.)
>>
>> Yup.  We [currently] don't have a good way to tell the length
>> of a tick.
> 
> The problem of application code not knowing how long a "tick"
> is seems to keep popping up.  In most other RTOS systems I've
> used, there was a constant or a pair of macros defined in a
> header that allowed application code to convert between ticks
> and "real" time (seconds or milliseconds).  This allowed one to
> write code that didn't break when the length of a clock tick
> was changed.
> 
> I think the eCos really needs something like that.  I've had to
> change the tick value back and fourth several times to
> accomodate different application code.  It's currently at 10ms
> to allow the TCP/IP stack to work, but I'll have to come up
> with a way to get it back to 1ms so that other portions of my
> application will work correctly and yet keep TCP/IP working.
> 
> Any thoughts on providing a way to let application code know
> how long a tick is?
> 

The information _is_ there, it's just not pretty.

Here's how to get at it (culled from ecc/kernel/XXX/tests/tm_basic.cxx):

  #include <pkgconf/kernel.h>

  long rtc_resolution[] = CYGNUM_KERNEL_COUNTERS_RTC_RESOLUTION;
  long ns_per_system_clock;

  // To convert system clock 'ticks' to nanoseconds (ns)
  ns_per_system_clock = 1000000/rtc_resolution[1];
  ns = (ns_per_system_clock * (long long)ticks) / CYGNUM_KERNEL_COUNTERS_RTC_PERIOD;

Does this suffice?