BSOD 0xc000021a - Cannot boot Win 10 at all

axe0 · 14 Jul 2016

Good :)

The situation is a bit unfortunate as we cannot check the SMART of the drive

gaztech · 15 Jul 2016

axe0 said:

Good :)

The situation is a bit unfortunate as we cannot check the SMART of the drive

Hi,
I did run WD Diagnostics against the drive when I tool it out and put it in a drive caddy. The diagnostics came up clean and came up with a PASS on SMART.

So, I though it was about time I wrote down a chronological list of what I've tried. I was making copious notes all the way through the process. Here goes:

1) Tried booting to safe mode. - Not possible. All variants will not load up and I get error c000021a. Normal mode of course throws the same error.

2) Booting into troubleshooting mode does work after 3 boot attempts as does booting from a recovery USB stick.

3) Running sfc /scannow results in an error: "Windows Resource Protection could not perform the requested operation", although before it threw iup the error, it said "Verification 100% complete". Huh?

4) Running sfc /scannow /offbootdir=C:\ /offwindir=d:/windows results in the same error as above (takes much longer to process though).

5) After identifying that the WindowsTrustedInstaller service was not running, I ran "net start WindowsTrustedInstaller" to start the service and re-ran items 3 and 4 above. Same errors were noted on both.

6) I found the 64bit devcon utility in the WDK from MS and put this on my USB stick. I thought I would try to disable my graphics driver as this was suspect. After finding the correct ID using devcon I ran the tool to disable the driver. I then ran all the above 1,2,3,4 again - same results. Then re-anabled the driver. It wasn't helping.

7) Decided to try some DSIM commands to see if a rebuild was possible. Tried :

DISM /ONLINE /CLEANUP-IMAGE /RESTOREHEALTH

This resulted in:
Error 50: DISM does not support serrvicing Windows PE with the /online option

I've come across this before in recovery mode. There is a fix... I deleted the MiniNT registry key at:
HKLM\SYSTEM\CurrentControlSet\Control\

Running DISM again it worked and it ran to 20% then stopped with another error:
Error: 1
Incorrect function
The DISM log file can be found at x:\Windows\Logs\DISM\dism.log

Log is attached.

So, tried:

DISM /ONLINE /CLEANUP-IMAGE /RESTOREHEALTH /Source:WIM:f:\sources\install.wim:1 /limitaccess

(F: was my DVDV drive with a recovery DVD in it)

This time the progress bar went to 100% but gave me the "Error 1: Incorrect function" too.

Hope the flow of work here is methodical and correct.

As you can see, I've checked loads of stuff. I may have missed some things out or misinterpreted something but after doing all this you would have thought something positive would have come out of it - but nothing!

I'm concerned that none of the repair utilities run due to Windows Resource Protection getting in the way. Surely recovery is just that - recovery! How come a "recovery" utility requires extra permissions to do the repairs that it was designed to do? If it needs permission that you can't envoke then it can't actually be a "recovery" utilty can it? I just don't get this. MS has some work to do on these tools.

I don't know where to go with this now. Aaargh!

Attachment 90225

axe0 · 15 Jul 2016

Could you run HDTune on your hard drive in the caddy?
I'd like to see the SMART attributes and there values, make a screenshot of the health tab in HDTune and post it.

gaztech · 15 Jul 2016

Hi @axe0,
Tried HDTune and it doesn't seem to want to read the SMART info.

Odd, until I checked this and it seems that HDTune doesn't read WD drives very well in this area. There are many posts about it.

I Tried WD Data Lifeguard and also Crystal Disk Info. Both read the information and it seems ok to me (maybe you know better...). I have included the screenshots here for you to take a look at:

Attachment 90238 Attachment 90239

What do you think?

axe0 · 15 Jul 2016

What I think is not important.

The reallocated sector count is very high, it isn't that high with a hard drive that certainly works properly.
I cannot say if the drive is the cause, but it does impact on the problem.

Ztruker · 15 Jul 2016

3 dumps like this:

Code:

BugCheck 124, {0, ffffe001f83f3028, f2000000, 5}
Probably caused by : GenuineIntel

1 each of the following:

Code:

BugCheck 12E, {ffffe001fdf1f300, ffffe001fdf76600, ffffe001ffbe4000, 20000}
Probably caused by : partmgr.sys ( partmgr!PmWrite+1c8 )

Code:

BugCheck 9F, {4, 12c, ffffe000e1a4f040, fffff8027326aaf0}
Implicit thread is now ffffe000`e1a4f040
Probably caused by : BthEnum.sys

Looking at the dump info, I see you have 3 of 4 memory slots populated:

Code:

[Memory Device (Type 17) - Length 28 - Handle 0037h]
  Physical Memory Array Handle  0035h
  Memory Error Info Handle      [Not Provided]
  Total Width                   72 bits
  Data Width                    64 bits
  Size                          2048MB
  Form Factor                   09h - DIMM
  Device Set                    [None]
  Device Locator                DIMM0
  Bank Locator                  BANK0
  Memory Type                   12h - DDR
  Type Detail                   0080h - Synchronous
  Speed                         1066MHz
  Manufacturer                  Manufacturer0
  Serial Number                        
  Asset Tag Number                         a 
  Part Number                   PartNum0
[Memory Device (Type 17) - Length 28 - Handle 0039h]
  Physical Memory Array Handle  0035h
  Memory Error Info Handle      [Not Provided]
  Total Width                   [Unknown]
  Data Width                    [Unknown]
  Size                          [Not Populated]
  Form Factor                   09h - DIMM
  Device Set                    [None]
  Device Locator                DIMM1
  Bank Locator                  BANK1
  Memory Type                   02h - Unknown
  Type Detail                   0004h - Unknown
  Speed                         0MHz
  Manufacturer                  Manufacturer1
  Serial Number                        
  Asset Tag Number                          
  Part Number                   PartNum1
[Memory Device (Type 17) - Length 28 - Handle 003bh]
  Physical Memory Array Handle  0035h
  Memory Error Info Handle      [Not Provided]
  Total Width                   72 bits
  Data Width                    64 bits
  Size                          2048MB
  Form Factor                   09h - DIMM
  Device Set                    [None]
  Device Locator                DIMM2
  Bank Locator                  BANK2
  Memory Type                   12h - DDR
  Type Detail                   0080h - Synchronous
  Speed                         1066MHz
  Manufacturer                  Manufacturer2
  Serial Number                        
  Asset Tag Number                          
  Part Number                   PartNum2
[Memory Device (Type 17) - Length 28 - Handle 003dh]
  Physical Memory Array Handle  0035h
  Memory Error Info Handle      [Not Provided]
  Total Width                   72 bits
  Data Width                    64 bits
  Size                          2048MB
  Form Factor                   09h - DIMM
  Device Set                    [None]
  Device Locator                DIMM3
  Bank Locator                  BANK3
  Memory Type                   12h - DDR
  Type Detail                   0080h - Synchronous
  Speed                         1066MHz
  Manufacturer                  Manufacturer3
  Serial Number                        
  Asset Tag Number                          
  Part Number                   PartNum3

That doesn't seem like a normal setup. I'd suggest removing the memory module in SLOT3, leaving 2GB in SLOT0 and SLOT2, just to see how that affects things. Is the memory all the same, not mixed?

I don't know what the dump Bugcheck codes indicate other than that they jump around a bit and none are conclusive in pointing to the actual cause of the problem.

See what @axe0 says but it might be time to run Driver Verifier and CPU and GPU stress tests. Wait for @axe0 response though.

axe0 · 16 Jul 2016

About the dumps I already responded, check a (few) page(s) back :)

gaztech · 16 Jul 2016

Ztruker said:

That doesn't seem like a normal setup. I'd suggest removing the memory module in SLOT3, leaving 2GB in SLOT0 and SLOT2, just to see how that affects things. Is the memory all the same, not mixed?

I don't know what the dump Bugcheck codes indicate other than that they jump around a bit and none are conclusive in pointing to the actual cause of the problem.

See what @axe0 says but it might be time to run Driver Verifier and CPU and GPU stress tests. Wait for @axe0 response though.

You're right. it isn't a normal setup. I originally purchased 4 * 2GB modules from a friend of mine as they were the recommended type for my mobo. However, he damaged one taking it out of his machine (don't ask!) so I only have 3. However, my mobo user manual clearly states that this is an acceptable configuration with this make and model of RAM so I put it in.

That was about 2 years ago and all has been ok. It is of course possible that WIn10 baulks at this config so for the moment I will continue any tests with just two sticks in place.

I will be receiving a new hard disk in a few days to I'm going to start by cloning my "faulty" drive (not sure if it is or not) and seeing if I can boot from the new cloned device. I don't even know if Windows will complain about this but I'm going to try it anyway.

Failing that, I might see if I can get a new/used motherboard of exactly the same type and try that. Do you know if that breaks Windows hardware rules with the software if what you're replacing is the same?

Sure, I "could" just nuke the disk and re-install, but I'm an engineer and I simply cannot just do that without finding out what is causing this error. I won't be the first to get it and won't be the last. I need to do all I can to offer up any solutions here as everyone has been very helpful. Maybe I won't find the problem but I'm going to give it all I can.

Do Microsoft read these forums? If not, they should!

axe0 · 16 Jul 2016

It is not a good idea to clone the hard drive, because if the root cause is not the hard drive itself (what we do not know) you just take the problem onto the next one.

Changing the motherboard causes activation problems, unless you have a retail key which under normal circumstances need to be deactivated first. Changing the motherboard is seen as a new system by Microsoft, that is why it causes activation problems, in many cases a phone activation is required to activate the key again because in many cases the key is not retail or it wasn't deactivated first for whatever reason. Besides, Microsoft doesn't activate a key on a new motherboard without a reason like a motherboard failure or something.

Do Microsoft read these forums? If not, they should!

We do not know, I don't know what activities are logged, but I do not think the kind of activities that make it possible to find out if Microsoft reads this forum are logged. Note that this is a wild estimated guess based on nothing.

gaztech · 16 Jul 2016

axe0 said:

It is not a good idea to clone the hard drive, because if the root cause is not the hard drive itself (what we do not know) you just take the problem onto the next one.

Yes, I agree with you. However, if it's a new drive, I clone it and it works then we can all rest easy. If the issue is exactly the same then we've created another "unknown". I know this is clutching at straws here but I really do want to find out what the hell has happened if only to try to prevent another incident. If the clone doesn't work, I can always reinstall a new system onto it afterwards anyway. I just think it's a logical step to see what the outcome is.

I have a retail key for Windows 7 Pro, which is what I upgraded this system from. I guess I could in that case get the key deactivated/reactivated again. I suspect I'll have to do that again anyway if I try to reinstall onto a new disk. We'll have to wait and see.

If only I could do an inplace upgrade from the command line then i could be sure that at least all the necessary system files were fine.

Or, if there was a way to turn off Windows Resource Protection temporarily whilst sfc could do its stuff, that might fix it. Who knows...?

I'm suspecting hardware now but even that is difficult to faultfind if Windows is going to barf at the sight of different (but identical) hardware, like the mobo.

Is there a way to boot from the drive in a caddy so that I can eliminate the mobo disk controller? I tried it and Windows wasn't having any of it.