this is a symptom reported form time to time (IRQ time high with no apparant
reason).

it would be nice to get to the bottom of it.


"Ilya Zakharevich" wrote in message
news:csc6jd$emr$1@agate.berkeley.edu...
>
> During the last week the lazy-time CPU load on my system jumped from
> 0% to 13%. According to DosPerfSysCall(), 97% of these is spent at
> interrupt time.[*]
>
> The only change I remember is install of AMOUSE. Can it be the
> reason? Any way to it?
>
> Thanks,
> Ilya
>
>
>[*] Checked by something like
>
> perl -wle "$|=1; while (1) {@n=OS2:erfSysCall;print(($n[3]-$o[3])/
($n[0]-$o[0])); @o=@n;sleep 1}"

[A complimentary Cc of this posting was sent to
eric w
], who wrote in article <35bgk1F4j3poqU1@individual.net>:
> > During the last week the lazy-time CPU load on my system jumped from
> > 0% to 13%. According to DosPerfSysCall(), 97% of these is spent at
> > interrupt time.[*]
> >
> > The only change I remember is install of AMOUSE. Can it be the
> reason? Any way to fix it?

> this is a symptom reported form time to time (IRQ time high with no apparant
> reason).

Is it reported in connection to AMOUSE, or unrelated?

Meanwhile my system runs again at load 0%. No substantial change to
configuration was made.

I have no reason to think that this is related to install of AMOUSE
any more...

Thanks,
Ilya

seems to me this is a buggy device driver or kernelbug but it is very
elusive.
seems to come & go; I read about someone reporting this problem, but then it
goes as quickly as it came.

....eric

"Ilya Zakharevich" wrote in message
news:ct1n5d$2l4r$1@agate.berkeley.edu...
> [A complimentary Cc of this posting was sent to
> eric w
> ], who wrote in article <35bgk1F4j3poqU1@individual.net>:
> > > During the last week the lazy-time CPU load on my system jumped from
> > > 0% to 13%. According to DosPerfSysCall(), 97% of these is spent at
> > > interrupt time.[*]
> > >
> > > The only change I remember is install of AMOUSE. Can it be the
> > reason? Any way to fix it?
>
> > this is a symptom reported form time to time (IRQ time high with no
apparant
> > reason).
>
> Is it reported in connection to AMOUSE, or unrelated?
>
> Meanwhile my system runs again at load 0%. No substantial change to
> configuration was made.
>
> I have no reason to think that this is related to install of AMOUSE
> any more...
>
> Thanks,
> Ilya

On Mon, 24 Jan 2005 19:54:10 UTC, "eric w" wrote:

> seems to me this is a buggy device driver or kernelbug but it is very
> elusive.
> seems to come & go; I read about someone reporting this problem, but then it
> goes as quickly as it came.
>

Grab IRQMon from hobbes. It is a graphical monitor of the interrupts
being used. It won't tell you what's wrong, but, it will tell you
which IRQ is chewing up the CPU.


--
David Forrester
davidfor at internode dot on dot net
http://www.os2world.com/djfos2

[A complimentary Cc of this posting was sent to
David Forrester
], who wrote in article :

> Grab IRQMon from hobbes. It is a graphical monitor of the interrupts
> being used. It won't tell you what's wrong, but, it will tell you
> which IRQ is chewing up the CPU.

Pity that it came too late. But maybe this IRQ flood will return, and
I will use your advice.

Thanks,
Ilya

[A complimentary Cc of this posting was sent to
eric w
], who wrote in article <35l233F4o02c0U1@individual.net>:
> seems to me this is a buggy device driver or kernelbug but it is very
> elusive.
> seems to come & go; I read about someone reporting this problem, but then it
> goes as quickly as it came.

Meanwhile the IRQ flood suddently reappeared. IRQMON shows it is
IRQ=5. Changing from PAR1284 to PRINT01 did not help. Other stuff on
IRQ=5 is USB and Ethernet...

Puzzled,
Ilya

my first move would be to dedicate an IRQ to my NIC. I have always found
IRQ sharing on OS/2 to be like Russian Roulette!

"Ilya Zakharevich" wrote in message
news:d2k3ir$2pt8$1@agate.berkeley.edu...
> [A complimentary Cc of this posting was sent to
> eric w
> ], who wrote in article <35l233F4o02c0U1@individual.net>:
> > seems to me this is a buggy device driver or kernelbug but it is very
> > elusive.
> > seems to come & go; I read about someone reporting this problem, but
then it
> > goes as quickly as it came.
>
> Meanwhile the IRQ flood suddently reappeared. IRQMON shows it is
> IRQ=5. Changing from PAR1284 to PRINT01 did not help. Other stuff on
> IRQ=5 is USB and Ethernet...
>
> Puzzled,
> Ilya

[A complimentary Cc of this posting was sent to
eric w
], who wrote in article :
> my first move would be to dedicate an IRQ to my NIC. I have always found
> IRQ sharing on OS/2 to be like Russian Roulette!

Sure. But note that the current setup *works*. ;-) While all the
others I tried before do not. ;-(

Ilya