>>>>> "Edgar" == Edgar F Hilton <[EMAIL PROTECTED]> writes:
Edgar> Question: Does anybody have a rough idea of how long is a
Edgar> "common" bus delay introduced by reading or writing from
Edgar> either ISA or PCI cards?
Edgar> I am experiencing a 5 micro-second delay every time I perform
Edgar> a read or write from any of my ISA cards. I am now trying to
Edgar> move my code into using PCI cards, and would like to get a
Edgar> handle of how big these delays would be with the PCI cards.
5 microseconds sounds plausible for ISA.
PCI should be significantly faster.
The specifics will depend a great deal on your system design. Bus
bridges add latency. Other cards that are using DMA add latency.
Something to keep in mind: the notion of a defined and bounded worst
case latency seems to be a foreign concept to most bus designers.
Best case throughput is easy. Best case latency they can probably
tell you. "Typical" latency is going to be a stretch. Worst case,
forget it.
Unfortunately, you can't easily measure worst case latency either.
paul
--- [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/
