x61s hardware interrupts

[richarddd]

Running sysinternals process explorer on my x61s I see that the CPU is spending about 1/3 of its time in Hardware Interrupts. It works fine on boot, then the hardware interrupts start after a few hours and the system keeps running hot until the next reboot.

According to the process explorer docs: On Windows NT-based systems Process Explorer shows two artificial processes: Interrupts and DPCs. These processes reflect the amount of time the system spends servicing hardware interrupts and Deferred Procedure Calls (DPCs), respectively. High CPU consumption by these activities can indicate a hardware problem or device driver bug. To see the total number of interrupts and DPCs executed since the system booted add the Context Switch column. Another sometimes useful metric is the number of interrupts and DPCs generated per refresh interval, which you see when you add the CSwitches Delta column

Event viewer doesn't show anything of interest.


Any idea how to figure out if there is a hardware problem and how to deal with it?

[Harryc]

Sure, turn off/disable one hardware item at a time in BIOS and see what happens. Start with the wireless card.

[richarddd]

Any faster method? The machine usually runs for a good number of hours before the problem starts, so this could be a very long process.

Do you mean disable in bios or in device manager (I'm running XP)?

It would be nice of there were some log or software which shows which hardware is interupting

[Harryc]

Do you mean disable in bios or in device manager (I'm running XP)?
(

I think I said disable in BIOS

[richarddd]

I turned off wifi in BIOS. It ran fine for about 10 hours until I launched lightroom and photoshop, at which point interrupts started in earnest. I then rebooted and turned wifi back on. I launched LR and PS a bit later, without ill effect.

What next?

[Harryc]

Look into running Kenrate and isolating it to a module, and then run it's "z module-name” option to isolate further. Obviously it would have to be failing at the time to be effective -
http://www.nynaeve.net/?p=45

[richarddd]

I'll try that and report back

thanks

[richarddd]

My system ran fine from yesterday's post until a short while ago.

Using kernrate, the problem appears to be with
RtlIpv6StringToAddressA

Googling that turned up a few thinkpad users who reported problems similar to mine. For example
http://paranoidmike.blogspot.com/2010/0 ... r-bad.html

The solution may be to update the network driver. I have 2007 versions of the network and wireless drivers. Lenovo shows 2010 versions. I'll try installing those.

[Harryc]

Awesome ... nice detective work. If the network driver fixes it please let us know.

[richarddd]

I installed the network and wireless drivers. So far it works, but given that the problem takes a while to manifest, I'll wait a day or two to see if it continues.

Thank you for your help on this. Much appreciated.

[richarddd]

Alas, that didn't solve the problem.

Another possibility cause I've read about is the harddrive reverting to PIO mode rather than DMA. It appears that the drive is neither, which doesn't makes sense to me, as it seems to indicate that there is no device attached.



EDIT: I reinstalled the Intel matrix storage manager and updated the BIOS. Too early to tell if it had any affect on the hardware interrupts. It didn't do anything for the DMA issue.

I also discovered that suspending and restarting the machine stops the hardware interrupt problem for a while. This is much quicker than restarting.

[Harryc]

An X61 has an SATA controller. There are no settings for DMA/PIO mode.

[richarddd]

That would do it.

Which brings us back to trying to figure out why I'm getting these hardware interrupt storms. Any other ideas?

[nparag]

I got the exact same problem on my X61.
I noticed something today:
All was fine after a clean reboot while connected with a power cord. I opened Process Explorer and looked at the numbers. Then I plugged out the power cord leaving the laptop on battery only, and suddenly the interrupts were flying to 30%.
I thought it may be related to power management.
I know there is no process which is behind this since I suspended/killed every single process and service in my system.
Any ideas anyone?

[richarddd]

I never found a good answer for my machine. When I get a storm of hardware interrupts, I put it into suspend mode, then restart. This makes the problem go away for a while. With luck someone will have a better solution.

[nparag]

Did you install an iSCSI initiator?

[richarddd]

I didn't install one

[GpsPasSion]

I opened Process Explorer and looked at the numbers. Then I plugged out the power cord leaving the laptop on battery only, and suddenly the interrupts were flying to 30%.
I thought it may be related to power management

Same exact problem, hard to understand what's going on. Unfortunately the suspend method doesn't work, at least standard suspend, will try hibernating.

Some x61 users fixed the problem with older video drivers here http://forum.thinkpads.com/viewtopic.ph ... 16#p623616 but it seems they were having the problem at all times.

EDIT - yes hibernation in/out does fix it so I guess I'll live with that for now instead of messing with the drivers. The strange thing is that I pulled out the chord and the interrupts didn't come back...oops spoke too quickly, they are back at 30% !

EDIT 2 - ok looks like we have a fix courtesy of one FredSnow here (got there by searching for RtlIpv6StringToAddressA that was apparently causing my problem per the tools above)

Finally, I noticed that I could initiate the storm by switching to battery. I suspected the Lenovo Power Manager, but I didn't want to remove it. I noticed a setting that will stop the disk drive when the machine is idle. I set it to "never" for battery use and the problem disappeared.

I suspect there might be a problem with the Lenovo Power Manager device driver. If your interrupt storm seems battery-related, try disabling the Power Manager's ability to stop the disk on idle.

[GpsPasSion]

So that "no idle" fix didn't help after all on my X61T...but reverting to an older 7ld140ww.exe video driver did, case closed.

[michael55]

My thanks to the research you all have done. Hopefully, one of your solutions will help me. My X61 has the same problem. For whatever reason, I assumed the high level of hardware interrupts was hard drive related, but it survived the replacement of the spinning platter disk with an SSD.

[richarddd]

So that "no idle" fix didn't help after all on my X61T...but reverting to an older 7ld140ww.exe video driver did, case closed.

Worked for me