Had my first BSOD in quite a few years (Win 10 v2004)

So had a Blue Screen for the first time in quite a number of years, so
long ago as a matter of fact that I had likely cleared out the previous
one just to save disk space since it was so old. The only reason it's
interesting is because this is the first BSOD under v2004, after
upgrading to that version a few days ago. I don't know if they are
related, or whether I'm going to even get another one soon.

I ran the dump through Bluescreenview and found the following:

Bug Check Code 0x0000012b
Parameter 1 ffffffff`c00002c4
Parameter 2 00000000`000006b1
Parameter 3 00000214`0fb02090
Parameter 4 ffffc701`37afa000
Caused By Driver ntoskrnl.exe

According to the following Microsoft webpage, this is caused by some
sort of memory corruption, "FAULTY_HARDWARE_CORRUPTED_PAGE". There are
only two reasons why this error happens: "Memory Manager Page Not Zero
Error", or "Compressed Store Manager Error". After looking at the
parameters it seems to have been caused by the latter rather than the
former.

https://docs.microsoft.com/en-us/win...corrupted-page

Beyond that, I'm just going to run a RAM diagnostic next time I reboot
it. And I'll wait to see if it happens again in short order, or not.

Yousuf Khan

Yousuf Khan wrote:

... this is the first BSOD under v2004, after upgrading to that
version a few days ago.

Did you force the update by downloading from Microsoft's catalog site,
or did you let the WU client retrieve the update?

https://docs.microsoft.com/en-us/win...indows-10-2004

After reading that, I'm putting off the 2004 build until they resolve a
few more of the "investigating" issues, like LSASS, Bluetooth, and audio
drivers. Some I don't care about, like aksfridge.sys and aksdf.sys
(neither of which are currently on my system) which is for Aladdin
security dongles that I don't use, nVidia drivers since I have an AMD
video card, variable refresh rate on Intel iGPU since I use a video card
instead of the graphics in the CPU (hell, I can't even use Intel's
update tool to get a new onboard video driver while the video card is in
use), and I don't have anything Thunderbolt.

I ran the dump through Bluescreenview and found the following:

Bug Check Code 0x0000012b
Parameter 1 ffffffff`c00002c4
Parameter 2 00000000`000006b1
Parameter 3 00000214`0fb02090
Parameter 4 ffffc701`37afa000
Caused By Driver ntoskrnl.exe

I thought Nirsoft's Bluescreenview showed a list or chain of
dependencies on a crash.

Didn't v2004 have some microcode overrides to address further
vulnerabilities in Intel CPUs? Those add libs to Windows that alter how
it issues instructions to the CPU, so they are software overrides since
obviously you cannot burn in new firmware code into the CPU.

Beyond that, I'm just going to run a RAM diagnostic next time I reboot
it. And I'll wait to see if it happens again in short order, or not.

memtest is not a stress or load test on memory. It just sequentially
writes and reads from memory. I've had memtest pass all memory as okay,
but still got race conditions in Windows when the memory got stressed.
You can start with memtest, but follow with memory stress testing (i.e.,
test under load) by using Prime95 or HeavyLoad.

On 6/25/2020 1:13 PM, VanguardLH wrote:
Yousuf Khan wrote:

... this is the first BSOD under v2004, after upgrading to that
version a few days ago.

Did you force the update by downloading from Microsoft's catalog site,
or did you let the WU client retrieve the update?

I let WU present it to me.

I ran the dump through Bluescreenview and found the following:

Bug Check Code 0x0000012b
Parameter 1 ffffffff`c00002c4
Parameter 2 00000000`000006b1
Parameter 3 00000214`0fb02090
Parameter 4 ffffc701`37afa000
Caused By Driver ntoskrnl.exe

I thought Nirsoft's Bluescreenview showed a list or chain of
dependencies on a crash.

Yeah it does, but that's too long and I didn't post it here. It's not
likely relevant anyways unless it happens again.

Didn't v2004 have some microcode overrides to address further
vulnerabilities in Intel CPUs? Those add libs to Windows that alter how
it issues instructions to the CPU, so they are software overrides since
obviously you cannot burn in new firmware code into the CPU.

Yeah, possibly, I don't know, it's not relevant to me, since I'm using
an AMD FX-8300.

Beyond that, I'm just going to run a RAM diagnostic next time I reboot
it. And I'll wait to see if it happens again in short order, or not.

memtest is not a stress or load test on memory. It just sequentially
writes and reads from memory. I've had memtest pass all memory as okay,
but still got race conditions in Windows when the memory got stressed.
You can start with memtest, but follow with memory stress testing (i.e.,
test under load) by using Prime95 or HeavyLoad.

Yeah, but it's fast, and I really don't think this is anything but a
random occurrence. I'll do the better memory tests if it happens again.

Yousuf Khan

On 25/06/2020 17:49, Yousuf Khan wrote:
So had a Blue Screen for the first time in quite a number of years,


It could be something to do with PIA.Â* Pakistan International Airline
has suspended most of its pilots because they were not qualified to fly
planes.

The Pakistan Civil Aviation Authority carried out an investigation
that found that 40% of pilots in the country held fake pilot licenses.
In response to the findings, Pakistan International Airlines (PIA)
suspended 150 of its 426 pilots pending further investigation.

https://www.aerotime.aero/clement.charpentreau/25272-after-alarming-report-pia-takes-action-on-fake-pilot-licenses-PK8303-crash

Their only qualification was to recruit suicide bombers like you, Mr Khan.


--
With over 1.2 billion devices now running Windows 10, customer
satisfaction is higher than any previous version of windows.

Yousuf Khan wrote:

VanguardLH wrote:

Didn't v2004 have some microcode overrides to address further
vulnerabilities in Intel CPUs? Those add libs to Windows that alter
how it issues instructions to the CPU, so they are software
overrides since obviously you cannot burn in new firmware code into
the CPU.

Yeah, possibly, I don't know, it's not relevant to me, since I'm using
an AMD FX-8300.

AMD isn't immune to microcode vulnerabilities.

https://techcommunity.microsoft.com/...s/ba-p/1000845
https://www.amd.com/en/corporate/product-security

The Intel microcode override patches are delivered via WU. I don't know
if it's the same for AMD, or if they expect you to use a separate update
checker tool.

https://support.microsoft.com/en-us/...windows-8-base

I can't tell from that if WU will present the AMD microcode update, or
if you're expected to somehow divine the update and do the "Fix It"
yourself. Maybe the microcode patches were available separately before
the v2004 upgrade. I don't how they might roll in prior updates into a
build upgrade.

https://www.askwoody.com/2020/kb-449...amd-stupidity/

While that mentions AMD, users get these microcode patches even for CPUs
they don't have. The libs get installed for all CPUs, and on boot the
OS figures out which microcode override lib to use based on the CPU it
detects is currently in the hardware setup.

VanguardLH wrote:
Yousuf Khan wrote:

VanguardLH wrote:

Didn't v2004 have some microcode overrides to address further
vulnerabilities in Intel CPUs? Those add libs to Windows that alter
how it issues instructions to the CPU, so they are software
overrides since obviously you cannot burn in new firmware code into
the CPU.
Yeah, possibly, I don't know, it's not relevant to me, since I'm using
an AMD FX-8300.

AMD isn't immune to microcode vulnerabilities.

https://techcommunity.microsoft.com/...s/ba-p/1000845
https://www.amd.com/en/corporate/product-security

The Intel microcode override patches are delivered via WU. I don't know
if it's the same for AMD, or if they expect you to use a separate update
checker tool.

https://support.microsoft.com/en-us/...windows-8-base

I can't tell from that if WU will present the AMD microcode update, or
if you're expected to somehow divine the update and do the "Fix It"
yourself. Maybe the microcode patches were available separately before
the v2004 upgrade. I don't how they might roll in prior updates into a
build upgrade.

https://www.askwoody.com/2020/kb-449...amd-stupidity/

While that mentions AMD, users get these microcode patches even for CPUs
they don't have. The libs get installed for all CPUs, and on boot the
OS figures out which microcode override lib to use based on the CPU it
detects is currently in the hardware setup.

You need a "CPU Version" utility.

The "version number" is the microcode patch number.

A CPU leaves the factory with a version of zero or one
(whatever the origin of that number is).

The BIOS has microcode in it, including a bug patch for
bugs detected around the time of CPU release. Maybe
the version is 5 or so.

If the OS microcode loader loads an update, it might be 23.

Using a CPU Version utility, you can check the number,
then compare to the number for processor family you're using.

Linux released the latest patches, so their OS has
Spectre/Meltdown, such as it is.

I think Microsoft held back the Intel ones, once the Intel
ones were a "quality nightmare". The status quo of accepting
and publishing the Intel ones without questioning them, has
stopped since Spectre/Meltdown. But, there *should* be
a web page on microsoft.com with a package for that, for
people who need it (HIPPA or PCI, those kind of people).
It's more likely a server is patched for those, than
a lot of desktops.

Fixing this, if it needs fixing, will require some
research. It's not a maintenance/free-lunch SaaS thing.
You will need your archeology tools to trace the
twists and turns.

Paul

On 6/25/2020 4:55 PM, VanguardLH wrote:
Yousuf Khan wrote:
Yeah, possibly, I don't know, it's not relevant to me, since I'm using
an AMD FX-8300.

AMD isn't immune to microcode vulnerabilities.

https://techcommunity.microsoft.com/...s/ba-p/1000845
https://www.amd.com/en/corporate/product-security

The Intel microcode override patches are delivered via WU. I don't know
if it's the same for AMD, or if they expect you to use a separate update
checker tool.

Yes, but AMD are vulnerable to fewer of them, and regardless, they
probably need a BIOS upgrade to work as well. Owing to the age of my
processor, there will be absolutely no more BIOS updates for me, and
therefore likely not any patches in the Windows drivers either. AMD will
update Ryzen processors, but not FX.

Besides, this is going in a completely irrelevant direction. Those
microcode updates and stuff were sent out ages ago, in previous security
patch updates. They wouldn't wait until full feature updates to roll
those out.

Yousuf Khan