% Interrupt Time: How Much Is Too Much?

[John Saunders]

The last time I wrote a device driver for a PC, Windows 1.0 had just been
released, so I need some opinions:

I'm tracking down a problem with my laptop computer, which has become very
slow lately. The primary symptom (to the eye of someone who has written
drivers for VMS and for PCs) is the fact of high interrupt time. I presume
this means what it has always meant - time spent in an ISR.

I'm constantly seeing more than 50% interrupt time. Is this too much? Common
sense says "yes!", but I'd like to be sure before I try having the disk
replaced.

BTW, the reproducer I'm using is just "xcopy c:\windows\system32\*.*
c:\temp". That gives me peaks over 95%.

Thanks for any help,
John Saunders

P.S. I found RATTV3, and it's showing atapi.sys with many ISR buckets over 1
millisecond. That does not sound good.

P.P.S. I wish I could just see some simple counters like number of
interrupts, errors, and/or spurious interrupts. If such counts are kept for
this laptop, they're fairly well hidden. If someone knows where I should go
to see such counts, it would really help me justify the replacement disk.

 

[John Saunders]

"John Saunders" <john.saunders at trizetto.com> wrote in message
....

I'm constantly seeing more than 50% interrupt time. Is this too much?
Common sense says "yes!", but I'd like to be sure before I try having the
disk replaced.

BTW, the reproducer I'm using is just "xcopy c:\windows\system32\*.*
c:\temp". That gives me peaks over 95%.

The answer was in KB817472: IDE ATA and ATAPI disks use PIO mode after
multiple time-out or CRC errors occur
(http://support.microsoft.com/kb/817472/en-us). There were indeed an unusual
number of timeouts after resuming from standby, and the driver was in fact
using PIO mode. Uninstalling the driver and letting Windows re-install on
restart solved the problem. I'm back on Ultra DMA 5 and I have a laptop
again and not a heater.

John