Random blue screen since January updates

Topic

New Reply

I’ve been getting random Stop F4 or 7A errors in ntoskrnl.exe and one system freeze (around half a dozen in total) since last week. There seems to be no correlation with user activity, on one occasion I was away from my desk and on others the CPU may have been virtually idle.

I’m running Win7 64 bit on a Lenovo Thinkpad X201 with 4G RAM and a Corsair SSD encrypted with Truecrypt. The only things that have changed in that time frame are:

a. January MS updates
b. I briefly connected a Firefox OS phone by USB that I’ve been lent for evaluation, but didn’t attempt to access it
c. Latest Firefox and Thunderbird updates.

System Restore to the week before any problems appeared hasn’t cured it. (I’ve turned automtic updates off for the moment.) The Corsair utility shows the SSD to be in excellent health, and I ran Spinrite on level 2 to be sure. Last night I ran the Microsoft memory diagnostic (which runs standalone) for 10 hours, running the extended tests. It was half way through its 4th cycle without error by the time I stopped it. Speedfan shows temperatures under 50C.

The Firefox OS phone was my first suspect, followed closely by MS updates, but the system restore would seem to eliminate those, as well as the Firefox and Thunderbird updates, which anyway would hardly be likely to provoke a blue screen.

Despite the clean memory diagnostic, I might try replacing the RAM as pattern-dependent faults are notoriously difficult to diagnose. Otherwise it would seem to come down to CPU or motherboard, which are not so easily eliminated.

Any ideas from lounge members?

Regards – Philip

Reply | Quote

Replies

If you are using 2x2GB RAM sticks, see if you get the random BSODs with just each of the sticks installed.

While I’ve never known how efficient Windows Memory Test is when running while more than one stick is installed, the likes of Memtest86+ is best run on just one stick at a time.

While Memtest86+ can be run with all modules installed, one Lounge member found that it didn’t snag one of the sticks as being bad and because the passes can take a number of hours to run, a quick test is to just use one stick at a time to see if the problem reoccurs.

Intermittent problems are the most difficult to nail down unless you can pinpoint the occurrence to a particular action on your part.

Downloading the chipset/video/audio drivers from Lenovo would be something else to eliminate – else you will need to zip your dump files and post those for those members experienced in BSOD problems to peruse.

Reply | Quote

Thank you Sudo15. Unfortunately it’s a single 4GB stick otherwise my first thought would have been to swap them or try running with only one. Memory testing is a black art which may well catch little more than fairly hard errors. I had a case some years ago of a fault which memtest86 failed to find but GoldMemory showed up reliably! Unless any other good leads come up in the next 24 hours I’ll almost certainly splash out for two new 2GB sticks – and if the problem turns out to be software I may still be able to use one to upgrade to 6GB (if that’s a legal combination).

I’d be keen to post a dump file if anyone would be willing to look at it for me – but is it possible it could contain passwords or private keys? At least a quick strings|grep didn’t show my master password. I’ve tried looking at a couple of dumps with Nirsoft BlueScreenView, which doesn’t actually tell me a lot.

Since I’d had no problems until last week I don’t think it would be a good idea to try updating Lenovo drivers as that would only add another unknown.

Reply | Quote

Why not just get another single 4GB stick and if it proves not to be the RAM then you can just double those up as long as you get the same spec and preferably brand. http://support.lenovo.com/en/documents/pd010141

Does Event Viewer record anything for the BSODs and do they still occur in Safe Mode with Networking or in a clean boot ?

Lenovo has its own set of Diagnostics you could run as a Bootable, but things need to broken before they can be snagged (usually). http://support.lenovo.com/us/en/lenovodiagnosticsolutions

While it could still be a driver glitch if something was being written to the SSD, creating your own restore point and downloading the drivers shouldn’t hurt, although hopefully Lenovo’s Diagnostics will snag something.

Reply | Quote

Hi, Philip.

F4 BSOD’s are frequently indicative of SATA, or recently SSD, errors. Installing the latest firmware can fix some (Crucial M4’s had a ~5200 hour uptime bug that triggered 0xF4’s every hour), otherwise, check the cables and carefully work through the Crucial (it’s generic, can work on many, if not all, SSD’s) ‘fix’: http://forum.crucial.com/t5/Crucial-SSDs/Why-did-my-SSD-quot-disappear-quot-from-my-system/ta-p/65215

Reply | Quote

I have experienced the same type of behavior on two Win 7 laptops since the last MS update. After entering a password to enter an account the screen goes blank. Control+Alt+Del brings up the screen to try various solutions. Task Manager never led me to a problem. I have just been restarting the machine (while bowing reverently to the NW toward MS HQ). Both machines have only MS IE and Chrome browsers. Nothing else has changed on them. No new software. No new hardware.

This suggests to me that the MS updates are the common element.

Reply | Quote

For the record, I’ve concluded that I have a faulty SSD. The Corsair SSD Toolbox shows excellent health under SMART data, but attempting a manual TRIM provokes an error “TRIM failed: layout validation error”.

Memory tests showed no errors, and this was clinched when I replaced my 4GB stick with a borrowed 2GB stick, and continued to get problems. Cloning the SSD to a HD and booting of that instead, I had no BSODs at all, but I’d forgotten how much slower a HD is!

Regards – Philip

Reply | Quote

Glad you found what was wrong and thanks for the update.

I guess it shows just how easily we can be spoilt with more modern hardware

Reply | Quote