Two BSODs in a row, system failing?

Discussion in 'General Hardware' started by AsiJu, Jul 25, 2022.

  1. AsiJu

    AsiJu Ancient Guru

    Messages:
    8,097
    Likes Received:
    2,858
    GPU:
    MSI RX6800XT G.XT.
    So I got 2 successive BSODs all of a sudden.
    First one said PAGE_FAULT_IN_NONPAGED_AREA and second that triggered immediately after a reboot was IRQL_NOT_LESS_OR_EQUAL.

    The BSOD screens were also misaligned and garbled from the bottom.

    Never had 2 BSODs in succession like that, what do you reckon is the likely culprit?

    I've had CTDs too lately.
     
  2. Horus-Anhur

    Horus-Anhur Ancient Guru

    Messages:
    4,178
    Likes Received:
    4,874
    GPU:
    RTX 2070 Super
    Were you doing some heavy workload on the PC at the time?
     
  3. AsiJu

    AsiJu Ancient Guru

    Messages:
    8,097
    Likes Received:
    2,858
    GPU:
    MSI RX6800XT G.XT.
    browsing Guru forums so pretty heavy :D

    (in other words no, had just finished a gaming session though)
     
  4. Horus-Anhur

    Horus-Anhur Ancient Guru

    Messages:
    4,178
    Likes Received:
    4,874
    GPU:
    RTX 2070 Super
    My first bet is memory.
    Try running TestMem5 v0.12 with Anta@Extreme

    Also run sfc scannow
     

  5. wavetrex

    wavetrex Maha Guru

    Messages:
    1,487
    Likes Received:
    1,166
    GPU:
    Zotac GTX1080 AMP!
    +1 for memory

    While my profile says 32GB, unfortunately I only use 24 now. It seems one of my sticks went bad in the last 2 years since I bought it... sadly, the warranty is at the other end of the continent, so not bothering with it for one memory stick.

    It started with random "glitches" (missing objects in games or other random errors), or software freezing or crashing without warning... then it moved to BSOD and downright system freezing.
    At that moment I knew something has gone bad, my computer has been stable until that point.

    RAM was the first I checked (by simply removing one stick at a time)... and it was the cause. Very stable again... just with a bit of "brains" missing...
     
  6. rflair

    rflair Don Coleus Staff Member

    Messages:
    4,377
    Likes Received:
    893
    GPU:
    5700XT
  7. sykozis

    sykozis Ancient Guru

    Messages:
    22,211
    Likes Received:
    1,313
    GPU:
    Asus RX6700XT
    First, see if it occurs again. If it does, remove any overclock and see if it continues to happen. If it does, test the memory first.
     
  8. AsiJu

    AsiJu Ancient Guru

    Messages:
    8,097
    Likes Received:
    2,858
    GPU:
    MSI RX6800XT G.XT.
    Thanks for the replies! Also I understand the OP was skimp on information but I wanted to see what the ideas are so I can give more info.
    Also 99 % certain this is some hardware problem at this point.

    I recently tested RAM with TestMem5 v0.12 with Anta@Extreme and it passed, also passed 10 consecutive runs of windows memory diagnostic tool.
    The memory was recently bought, 32 GB of Gskill (full specs n profile).

    No overclocks on system of any kind. Well RAM is at XMP but I've tweaked voltages a little as per suggestions in Ryzen thread for stability.
    I can ofc try to run at very default but the CTDs are quite random and those BSODs were a new thing.

    PSU is also fairly new but ofc it could be flawed.

    Sfc scans have come up clean mostly or corruption has been repairable. Also I'm pretty sure a system crash can cause corruption in itself.

    Could this be due to a motherboard problem? I'm still using first gen Ryzen mobo, it's 5 years old and could have had some hardware kinks to begin with.
    I have had a random cold boot problem from day one where a boot device is just not found. Simple reboot solves it but something is off.
    This issue has persisted across all BIOS versions, Windows reinstalls and main drives even.

    All other internal parts have been replaced at least once during the mobo's lifespan (CPU, GPU, RAM, PSU, bought NVME as main drive to replace SATA SSD)
     
    Last edited: Jul 26, 2022
  9. Undying

    Undying Ancient Guru

    Messages:
    19,494
    Likes Received:
    7,828
    GPU:
    RTX 2080S AMP
    Seeing you recently bought the new memory and problems started not long after i would still go with ram instability. Try downclocking it one level down and see what happens use the same timings. Also try the old ram kit if you still have it.
     
    sykozis likes this.
  10. Freeman

    Freeman Master Guru

    Messages:
    393
    Likes Received:
    68
    GPU:
    UHD 750
    This is RAM related.
    Some driver accessing stuff on memory and it don't have access rights.

    Third party software? Driver? Access rights?

    If there's no problem, I wouldn't tweak Ryzen system RAM.
     

    Attached Files:

    Last edited: Jul 26, 2022

  11. Horus-Anhur

    Horus-Anhur Ancient Guru

    Messages:
    4,178
    Likes Received:
    4,874
    GPU:
    RTX 2070 Super
    A few years ago, I had a similar problem.
    I would get random crashes, most of them would be in the desktop, just using the browser. Rarely had crashes while pushing the system.
    Turned out to be the power supply.
     
  12. Truder

    Truder Ancient Guru

    Messages:
    2,003
    Likes Received:
    997
    GPU:
    RX 6700XT Nitro+
    Get your RAM replaced under warranty - you've been having problems with games crashing and such, the problem will only get worse as time goes on and you're at risk of having anything written from ram to storage be corrupted.
     
  13. Espionage724

    Espionage724 Master Guru

    Messages:
    615
    Likes Received:
    274
    GPU:
    EVGA RTX 3060 XC
    I have a 2700X, and I can't run my RAM at XMP (4x8GB); XMP is 3600 but it fails stability testing and is only stable up to 3200. I believe this is a Zen+ limitation though. Also in my case, I had to manually bump up SOC voltage from the default auto 0.8V to 1.1V, and set the RAM to 1.4V. Basically, if I wanted anything faster than 2133, my motherboard needs specified voltages. I also RMA'd two of my sticks as they failed memtest even at 2133 (was probably just 1 stick but I didn't test single sticks and G.SKILL replaced both sticks since they were a kit)

    I used HCI memtest for 800% coverage overnight one time, and shortly after 900% I got an error. This was when I was still trying to find that 3200 stability spot.

    If I had your hardware right now, I'd be at 2133 and do an overnight stability test. Then play some games. You have an AMD GPU, and in my experience it's really easy to cause CTD or instability with unstable RAM settings when playing Vulkan games (like Dota 2, or anything through DXVK).

    Do a hard CMOS reset; power the system off, set the CMOS reset jumper, pull the CMOS battery, turn the system on, pull the plug out, and let it sit for a few minutes.

    Some firmware keep settings around through software CMOS resets, and I've noticed it be an issue with my X470 board when messing with rBAR and other settings where things just feel "off" until I do the hard CMOS reset.
     
    Last edited: Jul 26, 2022
  14. AsiJu

    AsiJu Ancient Guru

    Messages:
    8,097
    Likes Received:
    2,858
    GPU:
    MSI RX6800XT G.XT.
    Thanks all for input again, running RAM at stock 2133 now to see if it makes a difference.

    If it does I'll RMA the kit. That said I did have CTDs with old memories too.

    PSU is ofc possible too but it's very new, bought little over a year ago. But that's my next target I guess, I do have an old 750 W spare I can swap in. Worse unit though (multirail) but should suffice for troubleshooting purposes.

    If neither helps then I'll just swap the mobo I guess, been actually thinking of that before due to the persistent cold boot bug.
    Btw @Espionage724 I've done CMOS hard reset too a few times, made no difference. I'm fairly sure it's a problem with this particular mobo.

    Also yeah I have Zen 2 and memory passes tests at 3600. I did get error in windows diagnostic tool once but that was resolved by enabling geardown mode and doing the mentioned voltage adjustments (basically slightly higher VDDP and VDDG).
    The error happened after just abt 15 minutes when it did, with adjustments the system passed overnight run (10 passes).
    In addition, like said CTDs happened with old memory set too, just not as much (so I shrugged it off as software bug).

    My theory is that there is a problem either with the motherboard or with the IMC, veering more towards the former.
    It just about worked with old 16 GB 3000 SR set but now that I have 32 GB 3600 DR set it gets pushed harder and fails.

    Now I'm off to the system busting task of downloading and installing Office...
     
    Last edited: Jul 26, 2022
  15. AsiJu

    AsiJu Ancient Guru

    Messages:
    8,097
    Likes Received:
    2,858
    GPU:
    MSI RX6800XT G.XT.
    As it happens Doom Eternal (Vulkan) is the prime crasher to desktop. Had quite a few in Elden Ring too which is DX12 so also low-level API.
     

  16. AsiJu

    AsiJu Ancient Guru

    Messages:
    8,097
    Likes Received:
    2,858
    GPU:
    MSI RX6800XT G.XT.
    Also: I do still have my old R5 1600 but it can't really be used to test CPU swap properly because I'm quite sure system will not boot with first gen Ryzen and 32 GB DR RAM at 3600...

    unless the problems persist with RAM at default I suppose.
    I guess first gen Ryzen could boot with RAM at default.
     
  17. Freeman

    Freeman Master Guru

    Messages:
    393
    Likes Received:
    68
    GPU:
    UHD 750
    Not only hardware, might be some weird software / driver as well or software settings.
     
  18. AsiJu

    AsiJu Ancient Guru

    Messages:
    8,097
    Likes Received:
    2,858
    GPU:
    MSI RX6800XT G.XT.
    yeah but I've already spent hours troubleshooting software and honestly the latest development, 2 BSODs just like that while doing practically nothing, is pretty severe.

    Or even if it would be ultimately software it's prolly such an obscure bug no developer will look into it and I'm left with replacing hardware all the same.

    Anyway testing with RAM at default now, will take weeks to verify really since crashes are so random.
    Keep all ideas coming in the meantime, I appreciate.
     
  19. Astyanax

    Astyanax Ancient Guru

    Messages:
    13,658
    Likes Received:
    5,444
    GPU:
    GTX 1080ti
    passing memtest doesn't mean memory is ok, it means move on to stronger stress tests.
     
  20. -Tj-

    -Tj- Ancient Guru

    Messages:
    17,553
    Likes Received:
    2,183
    GPU:
    Zotac GTX980Ti OC
    Try to raise some cpu memory controller voltages related to memory, probably something on the edge of stability. And it gets more exposed when memory heats up because of gpu.


    OR if its happening idle, check again c-states etc, idk what are those on amd.
     

Share This Page