Ken Teh wrote:
>
> Seems to me it's perfectly normal to lose interrupts if the time between
> adjacent interrupts is longer than the write. You have to reenable
> interrupts before returning from the interrupt handler. And if you
> received another one before reenabling, then it gets lost. Interrupts are
> not queued.
Well, the handler starts with re-enabling interrupts, reads the data and
puts it in the buffer. At last the handler checks wether it received new
data. If it has received new data, it also means it has been working
with a combination of old and new data since there is no hardware
buffer.
If the handler is invoked without delay (or little of course) there's
less change of the data beiing overwritten by new data. (No handshaking
or something)
Perhaps I havn't expressed myself correct, I don't care the interrupt
gets lost but it isn't acceptable that the data is overwritten before it
is completely written to the buffer.
Hope I've expressed myself more correctly now.
Greetings Arnaud
-- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
---
For more information on Real-Time Linux see:
http://www.rtlinux.org/rtlinux/
