Receiving the titular error, need help diagnosing what if anything is dying. Here's whats happened so far and my attempts to diagnose/fix the problem. Apologies for the short novel but I've reached the end of what I can do! So the other day I was started getting the windows is out of memory error whilst playing BlackMesa and whilst my gf was using my the system for some Indesign work. I checked and I remember a ages ago I disabled the page file to free up space on my SSD. So I go to re-enable it and reboot and then go back into BlackMesa, now my troubles truly begin. The game locks up and I figure its just random game bug but the system refuses to alt-tab and doesn't respond to ctrl-alt-del. It then full on BSODs with error msg KERNEL_DATA_INPAGE_ERROR I google the stop code and follow various instructions from other forums to check for malware, rootkits etc, so I rKill, malwarebytes etc. doesn't make any difference. System keeps crashing, sometimes with BSOD, other times with just horrible artifacting/flickering. Further research suggests that my particular code is to do with when something in the pagefile is read back to RAM (and fails). So its either the RAM or the SSD that is malfunctioning. Progress I've run the HD tune health check, no bad blocks, and chkdsk /f and I get no errors, so the SSD seems guilt free. I've taken the ram out and given them the n64 treatment, replaced them and am currently running memtest86. After 1hr40 I have 17 errors, not good, so this points to the RAM being the culprit. However I am skeptical that it is the RAM that is causing the crashing as; 1. The crashing started happening just after re-enabling the PF (making me suspect the SSD more) 2. The system has been rock solid at the current bios settings for about a year, I've not touched them for ages (I reccon if I'd run memtest before all this started I might have had the same number of errors, maybe, idk whether you can have a seemingly stable system for that long with that kind of error rate.) It would e an incredible coincidence that the RAM starts to die at the exact time I re-enable the PF. ...so the SSD is to blame perhaps? However it seems unlikely to me that an M4 that is less than a year old is dying. So questions, 1. Have you guys encountered this error with the culprit being an SSD? 2. Will these kind of memtest results make a system definitely unstable, or could the RAM have been like this for ages and me not notice in day to day usage? 3. Could it be something else entirely? Mobo sata controller dying?! Any wisdom on the subject greatly appreciated! PS the rig is the "main rig" in the sig running 7x64, the RAM is not OC'd just CPU, temps are good There are no minidumps to upload, however I took photos of some of the BSODs and the memtest http://imgur.com/a/2N5ih
Errors in memtest, it is a fault with memory. I would test each stick in slot one to determine which stick is the cause and replace as needed. If you are still sure it is the ssd for some reason, I would run the secure erase command on it to get it like factory fresh and reinstall OS.
Thanks for the reply. Are you 100% it means the memory is damaged? Can memtest errors not ever be caused by a dodgy OC? Like I said it seems fishy that the memory would suddenly decide to cause BSODs when the only system change I have made is re-enabling the page file
One stick gets through 2 passes successfully, the other fails on the first run, I guess it was just a weird coincidence. I'll run the system with one stick for a while and see if that stops the crashing. Thanks + rep
Page file error often points to memory in most cases from my experience. As you see a stick is failing testing so it was just that the particular game was doing something memory intensive that made the problem evident. I have had similar thing in the past where it seem fine until you do one thing that cause stress in a particular way then bam, error and find out a bad component after testing and specifically looking for a problem when it seemed fine otherwise.
Well I carried on with the one "good" stick of ram last night, and after a while BlackMesa crashed again, to black screen, so I rebooted and just left the machine on overnight at desktop and in the morning it had crashed again in the same way, so although one of the sticks was giving errors, it wasn't the cause of the crashing. I'll try disabling the PF again and leave the system running today whilst I'm at work. I've ordered 2x4GB from ebuyer, I've been wanting to upgrade that for a while anyway, but it doesn't seem like its going to fix the problem. I'll report back tonight when I get home
You shouldn't disable page file. It is important to have one for correct system operation. If you are concerned with disk thrashing on your ssd you can increase the file cache in ram by entering in CMD fsutil behavior set memoryusage 2 Or move the pagefile to hdd if you wish instead though it is obviously much slower and is not as good performance as having page on ssd.
I want to have the PF enabled, I just wanted to disable it again to see if the problem went away, and it hasn't, or rather now I have a different problem, I'm not sure, now its crashing, sometimes just horribly, and just now with the error A process or thread crucial to system operation has unexpectedly exited or been terminated. I think I'm going to image the SSD to another HD and boot from that and see if the problem remains, if it does then its not the SSD, either there's something fkd with my windows install, or the HD controller perhaps...? anyways I've got no time tonight, I'll have a go tomorrow and see if I can isolate the issue
Progress at last! Cloned the SSD image to a caviar black, all issues gone. Same sata port/cable. Now I remember a while age there was a firmware update for the M4, I'll flash it, image back to the SSD and see how we go!
MISSION ACOMPLISHED That did the trick. I was still at firmware 0001, flashed to 040H and cloned back, all issues gone. Very very odd that the M4 should decide to take a detour via shitsville all of a sudden, but there we are, case closed BTW that memory tweak fsutil behavior set memoryusage 2does absolutely nothing, the string is there purely for compatibility reasons http://www.tweakhound.com/2009/12/11/ntfsmemoryusage2-tweak/
Ah don't worry, I'm not blaming you man, just had no idea that the previous firmwares had that big of an issue. For obvious reasons its not well advertised.
I got stung by crucial with the M4. I sold around 300 of them in various sizes then after a few months people started shouting at me Spent ages trawling through invoices tracking down M4 owners so I could update the firmware (0309 at the time). But I still use them even though they are getting old now (Bit like me really).
