Is this a new problem with existing applications? (in other words: you used to be able to do what you are trying to do, but now - suddenly - you have a problem where none had been before)?
Is this new problem with new software?
Have you recently changed any hardware or other software?
It appears that the issue happens with the game "Battlefield"
Given that your dump is pointing to hardware, my first guess is that you may have a hardware incompatibility with the game.
1. Please check your hardware to make certain your system should run the game.
2. Please make sure your graphics drivers are updated.
3. Sometimes this kind of issue shows up with "crack"/warez software. I hope that is not the case here.
There is quite new computer. I've bought it about 2 months ago.
The configuration is AMD Phenom2 x2 550 + 4GB ram 1333 + asrock mainboard (NVidia 630a) + Geforce 8600GS + Win XP SP 2
But at first everything went fine. BSODs began about week ago. I didn't change the hardware in the meantime. I think it is not connected with the Battlefield game as BSODs are randomly comming with different software - sometimes just after reboot - without any application except system ones. Also I installed Battlefield 3 days ago, after first few BSODs.
I've made some memory tests today with Memtest86 but all tests was passed without errors. I also updated hardware drivers (chipset, lan, gfx, sound). I changed Firefox to Chrome. But still nothing.
You will find attached a few other minidumps from all random BSODs today.
I'm running low on ideas what to fix and more I fix I feel like BSODs are more frequent
Please note that (unless you start getting errors reported) you will need to let this run a long time.
A good way of doing this is to let it run over night. In each 'pass" there are a series of eight memory tests that are run, each more complicated than the previous. You need to let it run a minimum of 7 passes, unless you start getting errors. If you have errors reported, you may want to re-run the test with only one stick of ram installed, checked each stick until you isolate those that are faulty. That way you can limit your replacement to those that actually need it.
Let us know the results. Faulty memory occassionally/rarely will pass Memtest.
But if it fails to pass 7 passes of Memtest, it is bad.
If your memory comes up without errors, then we can check other things.
Mini082310-01.dmp - BugCheck 1A,
This memory management error is usually hardware related.
This checklist is a helpful approach to troubleshooting some specific Stop messages, such as this ...
1. Examine the “System” and “Application” logs in Event Viewer for other recent errors that might give further clues. To do this, launch EventVwr.msc from a Run box; or open “Administrative Tools” in the Control Panel then launch Event Viewer.
2. If you’ve recently added new hardware, remove it and retest.
3. Run hardware diagnostics supplied by the manufacturer.
4. Make sure device drivers and system BIOS are up-to-date.
5. However, if you’ve installed new drivers just before the problem appeared, try rolling them back to the older ones.
6. Open the box and make sure all hardware is correctly installed, well seated, and solidly connected.
7. Confirm that all of your hardware is on the Hardware Compatibility List. If some of it isn’t, then pay particular attention to the non-HCL hardware in your troubleshooting.
8. Check for viruses.
9. Investigate recently added software.
10. Examine (and try disabling) BIOS memory options such as caching or shadowing.
You other dumps are typically due to faulty drivers / software.
(Though 0xC2 and 0xA may occasionally indicate failing hardware - since you have one dmp pointing to memory, I am suggesting you start there.)
Are you running any cracks or warez?
(Please answer this question directly.)
Thanks for help and checklist.
I'll make night memcheck and get back with the answer tomorrow.
About the checklist. I've investigated no 1 without any trace.
2-9 are more or less checked already.
No 10 is what I'll look into in a moment. I wonder if it could be something with dual channel access to memory. It was set by default but memory I have (Goodram) was not a dual-channel-set. Another suspicion is with the memory speed. Its vendor speed is 667Mhz (DDR-1333) but my mainboard on auto-setup set them at 400Mhz. I put them to 667 by manual settings but maybe it is something with mainboard/processor?
It seems I've found a solution!
It was probably memory settings in BIOS.
As I wrote before my memory was on auto setup with frequency on manual set.
I had it manualy set at 667Mhz (DDR3 1333) but auto-setup of timings was still wrong. The vendor setting should be 9-9-9-25-34 and it was auto-set at something like 6-4-7-20-30 (or something like that, didn't write it down).
When I set it to numbers recomended for this modules (9-9-9-25-34 on 1,5V) it is working 1,5 hour on high performance without any glitch.
I could probably lower it to 8-8-8-24 on 1,6V but safety first so no overclocking for a moment
Glad to hear (no warez) and glad to hear you found the solution!
It occurs to me that Memtest might report bad memory
when the real issue is simply what you found, the bios settings were incorrect for the modules in use!
(note to self... remember this for future instances)
The question, for me, remains why did it take so long to show up.
But at first everything went fine. BSODs began about week ago. I didn't change the hardware in the meantime.
Well, why it took 2 months to realy show up.
Probably light usage - only system + internet (FireFox) 1-2 hours daily didn't put the memory into enough stress. When I've got a break from work 2 weeks ago I started to install new software like games to play during vacations (Battlefield) and the memory was used much more and the possibility to cause BSOD was much stronger (high quality video, lots of net usage etc.) - my 2-core Phenom was in full swing and that caused much memory stress. Also 5+ hours a day of usage made a difference.
That is my guess.
What I learnt is that one should not trust auto-setup of memory modules in BIOS - one should always check the numbers, even if it looks like it works ok.