Greetings All,
I found a partial solution after thinka about what "Maciej" pointed out
about holding the interrupt to long or short.
I turns out that this is becoming more of a hardware issue with a partial
software solution. Hardware designers are ignoring compliance with
history. Some PCI card ASIC designers are allowing the interrupt to hold
longer th get back some the arbitration time. Since many if the Intel
systems break the PCI arbitration-bit by design, everyone is fighting for
time on the BUS
I had to sit down and really think this one out and make some phone calls
to get close to verification, but no one will acknowledge this as truth or
fiction.
The reality is that the host-drive pair on older systems need the an
expiry reset to give the ISR more breathing room.
The problem is that this does not hold true in all cases :-(
Ingo, I apologize for pointing the finger at you, but I will need to work
with you and Martin to resolve the ones that do not allow for this simple
fix.
Cheers,
Andre Hedrick
The Linux ATA/IDE guy
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Wed May 31 2000 - 21:00:25 EST