PDA

View Full Version : Re: IRQL_NOT_LESS_OR_EQUAL


J.Clarke
July 5th 03, 12:11 AM
On Sat, 5 Jul 2003 11:07:39 +1000
"Blaedmon" > wrote:

> conflicting device driver irq's are triggered usually by the obvious -
> vying device driver irq's ;) Non-Certified drivers such as graphics
> drivers are the main culprits, in my experience. You could check your
> device irq's listing and see what irq's are being shared - as a rule
> for creative soundcards DONT LET THEM SHARE ANYTHING.

With 2K/XP you don't have this luxury. The OS is going to share
interrupts whether you like it or not. Whoever came up with the notion
that there should be no manual overrides needs to be brought to Osama's
attention, but that doesn't resolve the problem.

One thing you can do, however, is find out which slots in the machine
are hard-wired to the same interrupt line and make sure that you don't
have high-traffic devices in both slots.

It may be necessary to play musical boards until you get them in a
sequence that causes Windows to assign interrupts in a manner that they
are all happy with.

> Another reason for this error is overheating, as youve probably
> surmised. Check, also, your chipset - see if its hot on the skin to
> touch (shouldnt be). Look at also your graphics card fan - is it
> working? ;) Your cpu fans, by default, need attention as AMD chips
> (duron) run a lot hotter than pentium.
>
> "Terry Pinnell" > wrote in message
> ...
> > I've been getting increasingly frequent crashes of my Windows XP
> > Home PC in recent months. Complete and abrupt shut down, with the
> > cryptic, daunting Error 'Stop 0x0000000A or IRQL_NOT_LESS_OR_EQUAL'.
> > I can't pin down what causes it, because there are always several
> > major programs running at the time, including MSIE6, Agent, TextPad,
> > Snagit, etc, apart from the OS itself. But, FWIW, I'm always online
> > at the time.
> >
> > I'm wondering if this might be a graphics card issue?
> >
> > CPU = AMD Athlon XP1800+ with 512MB PC2100 DDR memory, m/b = ASUS
> > A7A266-E,System Chipset = M1647 ALiMAGiK 1 AGP System Controller,
> > graphics = 64MB NVidia Geforce2 MX, sound = Sound Blaster 5.1 Audio
> > with Dolby surround and Creative Four Point Surround (FPS1600),
> > modem= Conexant CXT1035 - HCF 56k v90 (internal), printer = Lexmark
> > Z23, scanner = Umax AstraSlim 3400 USB.
> >
> > --
> > Terry, West Sussex, UK
>
>


--
--
--John
Reply to jclarke at ae tee tee global dot net
(was jclarke at eye bee em dot net)

Blaedmon
July 5th 03, 02:07 AM
conflicting device driver irq's are triggered usually by the obvious - vying
device driver irq's ;) Non-Certified drivers such as graphics drivers are
the main culprits, in my experience. You could check your device irq's
listing and see what irq's are being shared - as a rule for creative
soundcards DONT LET THEM SHARE ANYTHING.
Another reason for this error is overheating, as youve probably surmised.
Check, also, your chipset - see if its hot on the skin to touch (shouldnt
be). Look at also your graphics card fan - is it working? ;) Your cpu fans,
by default, need attention as AMD chips (duron) run a lot hotter than
pentium.

"Terry Pinnell" > wrote in message
...
> I've been getting increasingly frequent crashes of my Windows XP Home
> PC in recent months. Complete and abrupt shut down, with the cryptic,
> daunting Error 'Stop 0x0000000A or IRQL_NOT_LESS_OR_EQUAL'. I can't
> pin down what causes it, because there are always several major
> programs running at the time, including MSIE6, Agent, TextPad, Snagit,
> etc, apart from the OS itself. But, FWIW, I'm always online at the
> time.
>
> I'm wondering if this might be a graphics card issue?
>
> CPU = AMD Athlon XP1800+ with 512MB PC2100 DDR memory, m/b = ASUS
> A7A266-E,System Chipset = M1647 ALiMAGiK 1 AGP System Controller,
> graphics = 64MB NVidia Geforce2 MX, sound = Sound Blaster 5.1 Audio
> with Dolby surround and Creative Four Point Surround (FPS1600), modem
> = Conexant CXT1035 - HCF 56k v90 (internal), printer = Lexmark Z23,
> scanner = Umax AstraSlim 3400 USB.
>
> --
> Terry, West Sussex, UK

Blaedmon
July 5th 03, 06:33 AM
actually, u can override automatic IRQ steering within windows 2000, it
simply involves loading your pc 'computer device driver' as a standard pc.
You can do this automatically during installation (pressing f7 or 6) when
installation asks for scsi drivers, etc, also. It'll then allow manual irq
steering. I had to do this myself a few months ago - and I agree with you -
NOTHING should be non-configurable. Especially something so important
towards syetem stability as irq steering. But for the lay-user, such methods
around madness they'll never know. :/



"J.Clarke" > wrote in message
...
> On Sat, 5 Jul 2003 11:07:39 +1000
> "Blaedmon" > wrote:
>
> > conflicting device driver irq's are triggered usually by the obvious -
> > vying device driver irq's ;) Non-Certified drivers such as graphics
> > drivers are the main culprits, in my experience. You could check your
> > device irq's listing and see what irq's are being shared - as a rule
> > for creative soundcards DONT LET THEM SHARE ANYTHING.
>
> With 2K/XP you don't have this luxury. The OS is going to share
> interrupts whether you like it or not. Whoever came up with the notion
> that there should be no manual overrides needs to be brought to Osama's
> attention, but that doesn't resolve the problem.
>
> One thing you can do, however, is find out which slots in the machine
> are hard-wired to the same interrupt line and make sure that you don't
> have high-traffic devices in both slots.
>
> It may be necessary to play musical boards until you get them in a
> sequence that causes Windows to assign interrupts in a manner that they
> are all happy with.
>
> > Another reason for this error is overheating, as youve probably
> > surmised. Check, also, your chipset - see if its hot on the skin to
> > touch (shouldnt be). Look at also your graphics card fan - is it
> > working? ;) Your cpu fans, by default, need attention as AMD chips
> > (duron) run a lot hotter than pentium.
> >
> > "Terry Pinnell" > wrote in message
> > ...
> > > I've been getting increasingly frequent crashes of my Windows XP
> > > Home PC in recent months. Complete and abrupt shut down, with the
> > > cryptic, daunting Error 'Stop 0x0000000A or IRQL_NOT_LESS_OR_EQUAL'.
> > > I can't pin down what causes it, because there are always several
> > > major programs running at the time, including MSIE6, Agent, TextPad,
> > > Snagit, etc, apart from the OS itself. But, FWIW, I'm always online
> > > at the time.
> > >
> > > I'm wondering if this might be a graphics card issue?
> > >
> > > CPU = AMD Athlon XP1800+ with 512MB PC2100 DDR memory, m/b = ASUS
> > > A7A266-E,System Chipset = M1647 ALiMAGiK 1 AGP System Controller,
> > > graphics = 64MB NVidia Geforce2 MX, sound = Sound Blaster 5.1 Audio
> > > with Dolby surround and Creative Four Point Surround (FPS1600),
> > > modem= Conexant CXT1035 - HCF 56k v90 (internal), printer = Lexmark
> > > Z23, scanner = Umax AstraSlim 3400 USB.
> > >
> > > --
> > > Terry, West Sussex, UK
> >
> >
>
>
> --
> --
> --John
> Reply to jclarke at ae tee tee global dot net
> (was jclarke at eye bee em dot net)

Terry Pinnell
July 5th 03, 08:31 AM
"Blaedmon" > wrote:

>conflicting device driver irq's are triggered usually by the obvious - vying
>device driver irq's ;) Non-Certified drivers such as graphics drivers are
>the main culprits, in my experience. You could check your device irq's
>listing and see what irq's are being shared - as a rule for creative
>soundcards DONT LET THEM SHARE ANYTHING.
>Another reason for this error is overheating, as youve probably surmised.
>Check, also, your chipset - see if its hot on the skin to touch (shouldnt
>be). Look at also your graphics card fan - is it working? ;) Your cpu fans,
>by default, need attention as AMD chips (duron) run a lot hotter than
>pentium.

Thanks. Will take a look at the fans and clean up etc.

On the IRQ sharing issue, this sounds a bit heavy for my level of
know-how, and I'm also not clear how in practice I'd do that check. I
have of course gone through all devices and they appear to be 'working
correctly'...

--
Terry, West Sussex, UK

J.Clarke
July 5th 03, 09:05 AM
On Sat, 05 Jul 2003 08:31:35 +0100
Terry Pinnell > wrote:

> "Blaedmon" > wrote:
>
> >conflicting device driver irq's are triggered usually by the obvious
> >- vying device driver irq's ;) Non-Certified drivers such as graphics
> >drivers are the main culprits, in my experience. You could check your
> >device irq's listing and see what irq's are being shared - as a rule
> >for creative soundcards DONT LET THEM SHARE ANYTHING.
> >Another reason for this error is overheating, as youve probably
> >surmised. Check, also, your chipset - see if its hot on the skin to
> >touch (shouldnt be). Look at also your graphics card fan - is it
> >working? ;) Your cpu fans, by default, need attention as AMD chips
> >(duron) run a lot hotter than pentium.
>
> Thanks. Will take a look at the fans and clean up etc.
>
> On the IRQ sharing issue, this sounds a bit heavy for my level of
> know-how, and I'm also not clear how in practice I'd do that check. I
> have of course gone through all devices and they appear to be 'working
> correctly'...

Conceptually it's really simple. Most devices built into the
motherboard or plugging into AGP or PCI slots have a mechanism to alert
the processor that they need attention. This mechanism is called an
"interrupt".

In the first PCs interrupts couldn't be shared--every
device that used them had switches or jumpers or a configuration program
that was used to set it to particular interrupts and it was up to the
system integrator to assign them correctly. From the technician's
viewpoint this was lovely as you had complete and absolute control over
interrupt assignments.

After a while, as more and more junk got added to the boards, it was
found that there were not enough interrupts available in the original PC
design to go around. This was addressed in the 32-bit buses (EISA and
PCI anyway--I don't recall how Microchannel and VLB dealt with the
issue) that were developed by allowing two devices to use the same
interrupt. This still worked out all right, as when two devices didn't
work and play well together you just put them on different interrupts
and shared interrupts between two other devices that were happier
working together.

The problem with all this was that it depended on a technician setting
things up properly. The Powers That Be became aware that the users were
not screwing things up sufficiently well, so they created an Artificial
Stupidity called "Plug and Pray" to do that job for them, and built it
into their latest operating systems. With Plug and Pray the operating
system makes all such decisions, and while it usually does reasonably
well, it occasionally guesses wrong--it has no master database which
suggests that one combination will work and another will not and does
not appear to be bright enough to record which interrupt combination it
used and to not do that again if the system crashes.

So there are two problems--devices that don't like to share interrupts
and an operating system that makes them try to do so.

Dealing with this is not beyond you. Just go into Computer Management
and pick "IRQ" off the list and it will show you what the assignments
are in your machine. Odds are that if you're running XP or 2K it will
show all devices on the same interrupt--those operating systems try to
dynamically assign interrupts instead of picking a set and sticking with
them, thus giving the Artificial Stupidity enough leverage to even break
a system that has been working reliably.

As Blaedmon quite rightly pointed out, you _can_ give yourself manual
overrides by reinstalling using "Standard PC" instead of letting the OS
default to ACPI. If you do that then you can go into Device Manager and
for each device specify what interrupt is to be used. Otherwise, you
need to just move boards around until you find a combination that is
happy--moving the boards causes Windows to assign the interrupts in a
different sequence.

But that's not the end of the story. For some reason many motherboard
manufacturers hard-wire two slots to the same interrupt. If so, that
should be mentioned in the manual for the motherboard
somewhere--unfortunately you may have to read the thing cover-to-cover
to find the information. If that is the case, then if possible avoid
using both of those slots. If you can't avoid it the best option is to
put something that doesn't need an interrupt in one of them or failing
that put low-traffic devices in them--your sound board and disk
controller for example should not be in that pair.


>
> --
> Terry, West Sussex, UK
>


--
--
--John
Reply to jclarke at ae tee tee global dot net
(was jclarke at eye bee em dot net)

Terry Pinnell
July 8th 03, 08:37 AM
(k-man) wrote:

>MY FATHER HAD THE SAME RECURRING LOOP
>
>IRQL_NOT_LESS_OR_EQUAL
>
>ATTEMPTING TO RE-INSTALL WINXP WAS USELESS
>
>IT TURNED OUT TO BE A OVER HEATED CPU ATHLON 1.2GHZ
>
>HE UPGRADED TO A 25DOLLAR CPU FAN AND THEN TEMP DROPPED
>
>20 DEGREES AND WINXP INSTALLED FINE WITH NO FURTHER
>
>PROBLEMS
>Cheers
>
>
>"J.Clarke" > wrote in message >...
>> On Sat, 05 Jul 2003 08:31:35 +0100
>> Terry Pinnell > wrote:
>>
>> > "Blaedmon" > wrote:
>> >
>> > >conflicting device driver irq's are triggered usually by the obvious
>> > >- vying device driver irq's ;) Non-Certified drivers such as graphics
>> > >drivers are the main culprits, in my experience. You could check your
>> > >device irq's listing and see what irq's are being shared - as a rule
>> > >for creative soundcards DONT LET THEM SHARE ANYTHING.
>> > >Another reason for this error is overheating, as youve probably
>> > >surmised. Check, also, your chipset - see if its hot on the skin to
>> > >touch (shouldnt be). Look at also your graphics card fan - is it
>> > >working? ;) Your cpu fans, by default, need attention as AMD chips
>> > >(duron) run a lot hotter than pentium.
>> >
>> > Thanks. Will take a look at the fans and clean up etc.
>> >
>> > On the IRQ sharing issue, this sounds a bit heavy for my level of
>> > know-how, and I'm also not clear how in practice I'd do that check. I
>> > have of course gone through all devices and they appear to be 'working
>> > correctly'...
>>
>> Conceptually it's really simple. Most devices built into the
>> motherboard or plugging into AGP or PCI slots have a mechanism to alert
>> the processor that they need attention. This mechanism is called an
>> "interrupt".
>>
>> In the first PCs interrupts couldn't be shared--every
>> device that used them had switches or jumpers or a configuration program
>> that was used to set it to particular interrupts and it was up to the
>> system integrator to assign them correctly. From the technician's
>> viewpoint this was lovely as you had complete and absolute control over
>> interrupt assignments.
>>
>> After a while, as more and more junk got added to the boards, it was
>> found that there were not enough interrupts available in the original PC
>> design to go around. This was addressed in the 32-bit buses (EISA and
>> PCI anyway--I don't recall how Microchannel and VLB dealt with the
>> issue) that were developed by allowing two devices to use the same
>> interrupt. This still worked out all right, as when two devices didn't
>> work and play well together you just put them on different interrupts
>> and shared interrupts between two other devices that were happier
>> working together.
>>
>> The problem with all this was that it depended on a technician setting
>> things up properly. The Powers That Be became aware that the users were
>> not screwing things up sufficiently well, so they created an Artificial
>> Stupidity called "Plug and Pray" to do that job for them, and built it
>> into their latest operating systems. With Plug and Pray the operating
>> system makes all such decisions, and while it usually does reasonably
>> well, it occasionally guesses wrong--it has no master database which
>> suggests that one combination will work and another will not and does
>> not appear to be bright enough to record which interrupt combination it
>> used and to not do that again if the system crashes.
>>
>> So there are two problems--devices that don't like to share interrupts
>> and an operating system that makes them try to do so.
>>
>> Dealing with this is not beyond you. Just go into Computer Management
>> and pick "IRQ" off the list and it will show you what the assignments
>> are in your machine. Odds are that if you're running XP or 2K it will
>> show all devices on the same interrupt--those operating systems try to
>> dynamically assign interrupts instead of picking a set and sticking with
>> them, thus giving the Artificial Stupidity enough leverage to even break
>> a system that has been working reliably.
>>
>> As Blaedmon quite rightly pointed out, you _can_ give yourself manual
>> overrides by reinstalling using "Standard PC" instead of letting the OS
>> default to ACPI. If you do that then you can go into Device Manager and
>> for each device specify what interrupt is to be used. Otherwise, you
>> need to just move boards around until you find a combination that is
>> happy--moving the boards causes Windows to assign the interrupts in a
>> different sequence.
>>
>> But that's not the end of the story. For some reason many motherboard
>> manufacturers hard-wire two slots to the same interrupt. If so, that
>> should be mentioned in the manual for the motherboard
>> somewhere--unfortunately you may have to read the thing cover-to-cover
>> to find the information. If that is the case, then if possible avoid
>> using both of those slots. If you can't avoid it the best option is to
>> put something that doesn't need an interrupt in one of them or failing
>> that put low-traffic devices in them--your sound board and disk
>> controller for example should not be in that pair.

Thanks for the on-going advice. Perhaps the most telling comment so
far is the phrase k-man used above: "It turned out to be..."! I wonder
how much trial and error and futile change is typically involved
before a final culprit confidently emerges?! To date the candidates
appear to be:

1. RAM (but overnight tests with Memtest 30 produced no errors)
2. PSU
3. Video driver (Nvidia)
4. Interrupt conflicts
5. Faulty motherboard (ASUS)
6. Unusual software conflict
7. Sound card (Creative)
8. Pagefile settings (fixed? at what sizes? or system-controlled?)
9. CPU overheating
10. Interrupt traffic / board order
11. Keyboard / Mouse drivers
12. Faulty CPU

--
Terry, West Sussex, UK