New
#1
Multiple BSOD Stop Codes, System File Corruption, RAM Issues
So, I have a few different issues stacked on top of each other. Faulty RAM, old drivers, and system file corruption. BSODs directly related to what seems like all three.
I've been lurking on tenforums for years, but these issues are out of my experience so I need help despite being a power user moderately familiar with PC maintenance.
Sorry about the long post, I'm trying to get all the relevant information set out. I've formatted it to make it easier to read.
Why haven't I fixed these earlier? I've also been dealing with health issues that had to come first. I didn't want to touch the PC issues by myself while stressed out everything else and the machine otherwise operates normally.
I'm in Australia, but I am at my computer almost 24/7, most active between 8am - 12am GMT / 7pm - 11am AEDT / 3am - 7pm EST.
Files all attached at end of post because it's a bit long.
There seem to be two or three layers of problems:
- the RAM's faulty, or going bad, or something is wrong with the page file (?),
- my system files have corruption, but fixable corruption,
- outdated drivers (the classic problem, I know, I know),
- my SSD is nearly full (~50GB free of 932GB) and doesn't have much room to breathe.
Am I correct that the system file corruption needs to be fixed first? What are the first steps?
I've used WinDBG to analyse some of the dump files but am not sure how to interpret the results.
System file corruption:
- sfc /scannow fails with message "Windows Resource Protection could not perform the requested operation.".
- Dism /Online /Cleanup-Image /RestoreHealth fails at 3.8% with message "Error: 14098 The component store has been corrupted."
(both are in elevated/admin CMD)
I understand this can be fixed with a Windows USB ISO but don't know how to proceed.
Both DISM and SFC have succeeded in the past, following previous BSODs. CHKDSK has successfully completed a couple times in 2022, but not for a few months.
SSD health was good last time I checked (about 88% in CrystalDiskInfo).
I ran Windows Memory Diagnostic and it found hardware errors. Pretty telling even without doing MemTest86, but I don't know if I can get replacements. I can't afford new sticks and have not yet contacted the store I purchased from about a replacement because I only ran the WMD test in late December. I may be out of warranty and ineligible for replacements.
It's possible the RAM has been bad since the beginning but I didn't notice because I used to only leave the PC running for 8-12 hours at a time.
I know this makes a difference but don't fully understand why: the PC has been plugged in and powered for over a year. It's plugged into a UPS/sine wave battery backup and has not been unplugged from that. The last time the UPS was fully drained and unplugged was during bad storms in mid 2022. The PC has been shut down/turned off many times.
Windows
Windows 10 version is Education (I got my key from TAFE, a university), 21H2 build 19044.1889.
I have a pending major update, I believe it's 22H2. The update has been prompting to install since Nov/Dec but I've been continuously pausing it because I don't know if I should install it before fixing the system files.
Other various updates (security, driver, etc) have installed without an issue, and very often right after a BSOD.
Should I install the update? Will that repair the system files?
Drivers, versions, overclocking, etc
BIOS has never been updated. Version: 7C02v30 (11/07/2019), states E7C02AMS.350 in BIOS interface. I've read that if it's stable, you should leave it alone, is it a good idea to update in this case? Later versions have greater memory compatibility (relevant because my model of RAM is not on the QVL, but a very similar set is).
GPU driver is 516.94 (9th August 2022). I've hesitated to update because I've read about a lot of instability for this card (1660S, Turing) on newer drivers, and for all of 2022 the recommended driver was 511.79 (14th Feb 2022). Windows updated the driver automatically in Aug-Oct 2022 and I have had nvlddmkm.sys show up in some BSOD dump files, pointing to the NVIDIA driver in those cases. That's consistent with what I've read because nvlddmkm has apparently been a consistent problem for a decent chunk of people, but have not rolled back the driver because Windows will just update it again.
CPU is not overclocked, but had 'Game Boost' setting enabled, which does some sort of overclocking, turned on until about 23rd Dec. I turned it off when I realised what it did.
RAM has A-XMP enabled but never had apparent stability issues until 2022. A-XMP is a recommended overclock for the model of RAM. I assume I should go turn this off shortly.
BSODs
I've been getting a variety of BSODs since about December 2021/January 2022 (first recorded one is 12/01/2022 but looking at my chat records I had one before it). They happen several times a month, sometimes more often (there are instances of two within a few hours of each other).
I tend to leave my PC on, no sleep mode or fast startup, so the PC stays awake and just shuts off the monitors. This is because I'm usually actively working on things and frequently fall asleep without shutting down.
I've begun restarting the PC about once every 24-48 hours and this seems to prevent some of the BSODs.
I used to shut it down every night, which might have prevented them through 2020-2021 if the RAM's the problem. I've had the occasional BSOD since building the PC, but they've been more or less frequent over time, e.g recently. A few seem to have been very likely from bad driver or program updates, like in April 2022 when I suddenly got ten BSODs between 1st and 21st April, 5 in the first week and all of them 0x12B and 0x4E.
These BSODs happen when I'm doing regular activities on the PC, and some have happened while I was elsewhere and it was idling (returning to a rebooted PC and a BugCheck log entry). I always have Firefox open, sometimes also Chrome, and generally have another program open, usually MS Word, Excel, Clip Studio Paint (an art program, not relevant to the issues), one of a few games that I play regularly, and sometimes Blender or GIMP.
I disabled Firefox's disk caching out of concern for wear on the SSD (I have Firefox open about 90% of the time), which forces it to cache in the RAM. Should I set this back to default? I'm not sure if it's making the memory-related issues worse, but I'm getting suspicious.
In all but one or two cases the PC has rebooted from the bugcheck looking none the worse for wear. It's only rebooted to a diskcheck once and happily finished that.
Razer Central (for keypad) and Logitech G Hub (for mouse) are always running in the background. Steam isn't running unless a game is.
The software for the lights and etc on the GPU and RAM, Mystic Light and iCue, do not run unless I open them to change the settings. They fight.
Razer keypad and Wacom tablet (purposefully outdated driver because it's stable for my older model tablet) are always plugged in with drivers running.
I never max out my CPU or RAM in day-to-day usage.
I have Afterburner installed but do not normally run it, and only use it for the analytics (temp, etc). Haven't touched any of the overclocking-adjacent settings in that program, I don't overclock aside from the RAM XMP.
List of BSODs/stop codes
BSODS I've gotten:
MEMORY_MANAGEMENT (0x1A)
FAULTY_HARDWARE_CORRUPTED_PAGE) (0x12B) (fairly 'common')
PFN_LIST_CORRUPT (0x4E)
PAGE_FAULT_IN_NONPAGED_AREA (0x50)
BAD_POOL_CALLER (0xC2)
CRITICAL_STRUCTURE_CORRUPTION (0x109) (relatively recent)
SYSTEM_THREAD_EXCEPTION_NOT_HANDLED (0x7E)
REFERENCE_BY_POINTER (0x18)
KERNEL_DATA_INPAGE_ERROR (0x7A)
SYSTEM_SERVICE_EXCEPTION (0x3B)
KERNEL_SECURITY_CHECK_FAILURE (0x139)The last 2 months of BSODs:
3rd Feb (today, pfn_list_corrupt) (happened while I was writing the draft for this post),
1st Feb (yesterday, critical_structure_corruption),
27th Jan (memory_management),
18th Jan (memory_management),
8th Jan (bad_pool_caller),
7th Jan, two (faulty_hardware_corrupted_page followed by memory_management 90 minutes later),
29th Dec (bad_pool_caller),
23rd Dec (system_thread_exception_not_handled),
22nd Dec, two (faulty-hardware_corrupted_page and reference_by_pointer 14hrs later),
15th Dec (kernel_data_inpage_error),
14th Dec (faulty_hardware_corrupted_page),
12th Dec (faulty_hardware_corrupted_page),
8th Dec (memory_management),
6th Dec (faulty_hardware_corrupted_page),
25th Nov (critical_structure_corruption),
23rd Nov (page_fault_in_nonpaged_area).
Files
Folder Link
Latest DISM log (2nd Feb)
Photo of BIOS (23rd Dec)
Photo of Windows Memory Diagnostic (23rd Dec)
Please let me know if you need anything else. The V2 Log Collector seems very thorough.
I have 3 brand new blank 32GB USB sticks for fixing this (e.g for a Windows Installation media/ISO, MemTest86, etc), and also an external HDD I backup personal files to.
Last edited by Nattle; 02 Feb 2023 at 11:54. Reason: Just fixing the line breaks for readability