3 replies to this topic

[_Melk_]

Hi,

I've been using Alcohol Free (home) for a few years now. I recently change my pc to my signature.

After the new installation (including Alcohol 52%), everything was peachy .

---

The problem started after I started to use sleep to turn off my pc.
IF and only IF Firefox is opened when I go to sleep, when I return the sptd.sys or scsiport.sys (explained below), will use 80% of my core0 just for interrupts.

To recreate, open Firefox with Alcohol 52% installed with at last 1 virtual drive and mount any image.
Make sure ff is opened (you can see the window in the desktop).
Go to sleep.
Resume and watch the interrupts go crazy.

How I realized that is was the sptd.sys or scsiport.sys;

When using wither the Performance Viewer after a trace (xperf -on latency....), I first saw that the driver was unknown. (Same thing with LatencyMon.exe).
After much reading, I was told that Deamontools or Alcohol might be the cause, so I *just* disable the sptd.sys to see what would happen.
After disabling sptd.sys, when I ran LatencyMon, the same results where shown, BUT the driver was scsiport.sys instead of unknown now.

By then, I still wasn't sure if it was Alcohol or another device that might be using this driver and since I couldn't find out, I uninstalled Alcohol 52% and the problem is now gone.

Retested it multiple times with the same results (now working from resume from S3 state -- no issues at all).

If you need me to make more tests or if you have any questions, please feel free to let me know. I have plenty of time to test this so shoot ^^

Regards,

M

Edited by _Melk_, 02 April 2011 - 09:47 PM.

[_Melk_]

Hmmm,

After a lot more testing, this problem occurs with or **without** SPTD & Alcohol installed.

So, my apologies for thinking it was caused by Alcohol 52% I'm glad I can reinstall it now ^^

I'll just have to figure out why scsiport.sys gets crazy when I wake up from suspend (and I don't want to use hibernate). Worst comes to worst, I'll just NOT use suspend and revert to normal boot procedure.

Again, sorry for implying the bug came from Alcohol and/or SPTD.

Regards,

M

[_Melk_]

Final update (I hope)..

Turns out if was one if my SATA6G controller that was causing all the fuss. Renaming scsiport.sys made it so that those 2 controllers weren't working properly.

SO, I deactivated them in the BIOS since I don't use them anyways (Using the Intel controller instead, don't ask me why there's THREE controller for sata on my MB :/).
After deactivation, I put back scsiport.sys, reinstalled Alcohol 52% & the new version of SPTDInst (1.78) and voila! All working as intented. Tested 20 suspend and resume, and no more high cpu interrupts.

Posting this for those who might be interested. I haven't checked *which* of the two controller has the issue yet (their drivers that is). If anyone wants to know, please let me know here, otherwise, this thread can be closed.

The controllers are JMicron JMB362 & Marvell SE9128. Again, I'll test those some other time (been at this for the last 6 hours ON A SATURDAY! ouch) if anyone is interested.

Thanks for listening

Regards,

M

Edited by _Melk_, 03 April 2011 - 01:43 AM.

[Charalambos]

Thanks for the updates and it is good that you found a way to solve your problem.