This is the mail archive of the
ecos-discuss@sources.redhat.com
mailing list for the eCos project.
CYGACC_CALL_IF_DELAY_US(1000000) fails
- From: "Uwe Kindler" <ukindler at htwm dot de>
- To: <ecos-discuss at sources dot redhat dot com>
- Date: Tue, 25 Nov 2003 15:12:52 +0100
- Subject: [ECOS] CYGACC_CALL_IF_DELAY_US(1000000) fails
Hello,
if I call CYGACC_CALL_IF_DELAY_US(1000000) in order to delay one second,
then this call returns a lot
faster then 1 second. When I debug the code, then I can see, that
HAL_DELAY_US provided by the HAL
is only used if no kernel package is included. HAL_DELAY_US(1000000) works
fine.
If the kernel package is included, the the following calculation takes
place:
// How many ticks total we should wait for.
usec_ticks = usecs*CYGNUM_KERNEL_COUNTERS_RTC_PERIOD;
The RTC period for the EDOSK board is 20625. The board runs with 33 Mhz and
the 16 Bit timer is driven
with a prescaler of 16 - timer clock is 33Mhz/16 = 2062500. We would like to
interrupt 100 times per second
2062500 / 100 = 20625.
If I now delay 1 second, usec_ticks is calculated usec_tick = 1000000 *
20625 - this is larger than int32.
In the reference manual is written:
While most of the time this function is called with very small values,
occasionally
(particularly in some ethernet drivers) it is called with values of several
seconds.
So what can I do - if I use larger timer prescale values in order to lower
the timer period I will never get exactly 100 Hz for the timer clock.
Best Regards, Uwe
--
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss