Jump to content











Photo
- - - - -

disk read error, how to restore MBR and BCD (if this is it)

crc mbr bcd windows 7

  • Please log in to reply
37 replies to this topic

#1 matrix.rebooted

matrix.rebooted

    Newbie

  • Members
  • 26 posts
  •  
    France

Posted 04 September 2020 - 08:45 AM


Hello,

short story:

Disk have 3 partitions, all bootable, first is windows 7. After nvidia drivers (chipset+graphics) uninstall, get disk read error and partitions are now flagged as RAW but disk is ok and files can be access via gparted, parted magic, testdisk, what you want except windows (neither other win7 disk nor dvd). Problem is to restore disk's boot without formatting, have backup of MBR and BCD but how to restore? I have limited knowledge, tried to follow some howto but I'm a bit lost.

long story:

The computer is a Compaq, a bit outdated but still usefull, with Nvidia plateform (GeForce 6150SE nForce 430).
The hard drive has 3 partitions, all bootable: I kept original Vista to update bios if needed, as the HP recovery partition, in case, and added windows 7.

Recently some new event appeared (Event id: 129 Reset to device, \Device\RaidPort0, was issued) and after some readings about this nvidia 'bug' I decided to uninstall the drivers (chipset and graphic), and since then I get disk read error on boot and looking at the disk shows partitions are now flaged as RAW. Trying to access with win7 repair disk it says CRC error (cyclic redunduncy).

Drive is not dead and I can access files! (Plus, HDTune find the drive is healthy) But my problem is to restore the drive with it's mbr, boot sector and bcd, and, I have backup of them too! but how? (when install was ok, I made backup of MBR (512o) and PBR (8ko) using MBRbackup and bootice, and BCD file using easyBCD.

From what I have read, chkdsk could help but is to use in a last attempt since it may rewrite mbr and/or bcd.. Anyway, for now it fails on RAW partitions.

In fact, I think I would need to know first what is causing problem (mbr,bcq,bs?). I think the PBR is still untouched since mbrbackup shows the same as the backup I made before and other tools like gparted, parted magic, paragon, correctly find the partitions and access the files too. But for the MBR, Testdisk shows a file full of zeros while not the case with MBRbackup. Looking at the files with parted magic, it seems BCD file has gone!.. As for the boot sector, I don't know.

Also, there is (too) many sites showing some bootrec, bootedit, bootbcd, bootsect, diskpart commands... often different and/or in different order, I lost my self. Besides, if I boot with another drive (win7 too), the system is C: and defectuous partitions of the other drive become e: and f: making myself even more lost when trying to use these commands.

Hopefully, some times ago I made a multiboot usb flash drive with some tools (system rescue, ubuntu repair disk, hiren, malekal live cd, win7 repair disk, dart 8.1, even win98 bootdisk and dos6.22 ;) and some other tools and light linux distros like antix, deuterium); while it was interessting to build it, it would be even more usefull to know how to use it now! too many tools here too :)

I also tried with win7 CD, nothing more, except trying to re-install it on the win7 partition, it says unable to write to partition (kind of hopefully! since I don't want to format)

Should I rebuild some basic mbr,bcd just to access disk in a first time, then restore the backed up version, but this might modify partition table, or should I rewrite mbr with dd, dsfo, but how, from where? same with bcd, any try with a windows doesn't allow to access disk, so I cant use tools like easybcd. .

So I come to ask you so I don't mess the drive more, tell me if any other info or files are needed,

Thank you



#2 wimb

wimb

    Platinum Member

  • Developer
  • 3756 posts
  • Interests:Boot and Install from USB
  •  
    Netherlands

Posted 04 September 2020 - 02:16 PM

short story:

Disk have 3 partitions, all bootable, first is windows 7. After nvidia drivers (chipset+graphics) uninstall, get disk read error and partitions are now flagged as RAW but disk is ok and files can be access via gparted, parted magic, testdisk, what you want except windows (neither other win7 disk nor dvd). Problem is to restore disk's boot without formatting, have backup of MBR and BCD but how to restore? I have limited knowledge, tried to follow some howto but I'm a bit lost.
 

 

That is quite a story .... Uninstall of nvidia graphics drivers will not change MBR or PBR or remove BCD as you suggested / observed ....

In any case it was unwise to uninstall chipset drivers ....

 

What Tool do you use now to get info from the disk ? Can you give Screenshots ?

Can you boot from USB with Win10XPE to investigate the drive ? 

 

After booting from USB with Win10XPE or other PE you might use Tiny_Hexer_MBR and make a Screenshot of your MBR (= sector 0) of your Physicaldrive

 

MBR_Work_x64.exe > TinyHexer > Disk > Open drive > Physicaldrive0 (or your disk number)

 

Please give us screenshot as e.g. made with Portable FSCapture94 that normally would look like

 

[attachment=17343:TinyHexer_2020-09-04_155635.jpg]



#3 matrix.rebooted

matrix.rebooted

    Newbie

  • Members
  • 26 posts
  •  
    France

Posted 04 September 2020 - 04:36 PM

That is quite a story ....

:P

So, for now booting a windows 7 32bit hard drive with problematic drive as second disk (both sata, in fact just reverted boot order with my storage drive that is also bootable!).

I have no win10 drive or live usb tool or PE that has any link with windows 10 B) , besides all system I use are still x86/32bits and bios compliant. But directly running mpth_small.exe worked, here it is; the 3 partitions of the drive seen from the currently booted windows are on Physical Drive1:

part1: E:(old vista)
rvn1ACj.png

 

part2: F:(my win7)
K9xmcIq.png

 

part3: G:(old recovery)
mY3PFu3.png
for more info or access drive content, I could use system rescue or ubuntu boot repair disk, booted from live usb, they embed gparted or parted magic. But win7PE or win7RE doesn't boot from the usb if they detect the defective drive, don't know why.

NB: just observed that partition 3 (factory_image) doesn't seem to have been impacted, I can access it from windows, but it is neither the system nor the boot partition.

 

Here is a report of the PBR made when all was working:

Partition table for \\.\PhysicalDrive0
#  Boot  StartCyl  StartHead  StartSect  EndCyl  EndHead  EndSect  RelSect    TotSect    Type
0  Yes   1         1          0          1023    254      63       63         146785842  NTFS
1  No    1023      254        63         1023    254      63       146785905  453659535  NTFS
2  No    1023      254        63         1023    254      63       600445440  24691905   NTFS
3  No    0         0          0          0       0        0        0          0          Unknown file system


         # : Partition number
      Boot : Is this partition marked as bootable?
  StartCyl : Start cylinder
 StartHead : Start head
 StartSect : Start sector
    EndCyl : End cylinder
   EndHead : End head
   EndSect : End sector
   RelSect : Number of sectors between MBR and first sector of the partition
   TotSect : Total number of sectors in partition
      Type : Type of partition

and here the report from easybcd (before, too)

(sorry, in french :huh:)

Il y a 2 entrées dans le chargeur d'amorçage de Windows

Entrée par défaut : Windows 7
Compte à rebours : 5 seconde(s)
Disque d'amorçage : D:\

Entrée #1
Nom : Windows 7
Identifiant BCD : {current}
Disque : C:\
Chemin du chargeur d'amorçage : \Windows\system32\winload.exe

Entrée #2
Nom : Microsoft Windows Vista
Identifiant BCD : {cd8d3ced-7502-11dd-89d7-0021971778b6}
Disque : D:\
Chemin du chargeur d'amorçage : \Windows\system32\winload.exe

détails:

Windows Boot Manager
--------------------
identifier              {9dea862c-5cdd-4e70-acc1-f32b344d4795}
device                  partition=D:
description             Windows Boot Manager
locale                  fr-FR
inherit                 {7ea2e1ac-2e61-4728-aaa3-896d9d0a9f0e}
default                 {f9f4d741-bcfa-11ea-a902-c1e0b09a5b56}
resumeobject            {f9f4d740-bcfa-11ea-a902-c1e0b09a5b56}
displayorder            {f9f4d741-bcfa-11ea-a902-c1e0b09a5b56}
                        {cd8d3ced-7502-11dd-89d7-0021971778b6}
toolsdisplayorder       {b2721d73-1db4-4c62-bf78-c548a880142d}
timeout                 5

Windows Boot Loader
-------------------
identifier              {f9f4d741-bcfa-11ea-a902-c1e0b09a5b56}
device                  partition=C:
path                    \Windows\system32\winload.exe
description             Windows 7
locale                  fr-FR
inherit                 {6efb52bf-1766-41db-a6b3-0ee5eff72bd7}
recoverysequence        {f9f4d742-bcfa-11ea-a902-c1e0b09a5b56}
recoveryenabled         Yes
osdevice                partition=C:
systemroot              \Windows
resumeobject            {f9f4d740-bcfa-11ea-a902-c1e0b09a5b56}
nx                      OptIn

Windows Boot Loader
-------------------
identifier              {cd8d3ced-7502-11dd-89d7-0021971778b6}
device                  partition=D:
path                    \Windows\system32\winload.exe
description             Microsoft Windows Vista
locale                  fr-FR
inherit                 {6efb52bf-1766-41db-a6b3-0ee5eff72bd7}
recoverysequence        {572bcd55-ffa7-11d9-aae2-0007e994107d}
recoveryenabled         Yes
osdevice                partition=D:
systemroot              \Windows
resumeobject            {cd8d3cee-7502-11dd-89d7-0021971778b6}
nx                      OptIn
sos                     No

 

 

 

Uninstall of nvidia graphics drivers will not change MBR or PBR or remove BCD as you suggested / observed ....

In any case it was unwise to uninstall chipset drivers ....

just uninstalled and directly reboot, not doing anything else; but sure, wasn't thinking, just fed up with the freezes and not finding a solution while apparently this bug has more than 10 years :angry:, well at least didn't find.



#4 wimb

wimb

    Platinum Member

  • Developer
  • 3756 posts
  • Interests:Boot and Install from USB
  •  
    Netherlands

Posted 04 September 2020 - 05:02 PM

Thanks for the Screenshots and all info.

 

These are screenshots of the PBR of your three partitions.

Strange is that your Win7 drive has an NTLDR Bootsector that is not valid to boot Win7 but valid to boot XP.

If the second partition is set active then you cannot boot Win7 and get a disk error.

 

It depends on how you organised dual boot of Vista and Win7.

Can you describe how you think you arranged to boot with Windows 7 .

 

I asked you for a Screenshot of the MBR in your case Physicaldrive1

You have got to use R-mouse menu to Run as Administrator mpth_small.exe

Can you give that Screenshot.



#5 matrix.rebooted

matrix.rebooted

    Newbie

  • Members
  • 26 posts
  •  
    France

Posted 04 September 2020 - 05:48 PM

sorry for the screens, hope it is this one

 

PIT0Ylh.png

 

What I think I've done:

 - first, original drive came with 2 partitions (vista and factory_image after) 
 - second, I guess it was with minitool partition wizard, I added a partition to the left, reduced the one of vista and increased the first one newly created 
 - then, inserted win7 dvd and told it to install on first partition 

When done, it automatically created a dualboot menu (since 7 came after vista) and with 7 as default, that's it.

Not sure which is system partition and boot partition, but I didnt modify default install.

 



#6 Wonko the Sane

Wonko the Sane

    The Finder

  • Advanced user
  • 16066 posts
  • Location:The Outside of the Asylum (gate is closed)
  •  
    Italy

Posted 04 September 2020 - 06:49 PM

@matrix.rebooted

Better than the screenshot, you should post a copy of the needed sectors.

(otherwise we would need to re-type the relevant hex values reading them from your screenshots, wasting a lot of time and risking to make some typos).

 

Easier tool to do so is HDhacker:

http://dimio.altervista.org/eng/

 

You need to make a copy of one sector of:
MBR (Physical drive) <- you will have two Physical Drives available when booted from another disk, Physicaldrive 0 and PhysicalDrive 1, make a copy of both, calling them MBR_0.dat and MBR_1.dat.

PBR (logical drive) <- you will have several volumes available, C:, D:, E:, F: G:, make a copy of each of them, calling them PBR_C.dat, PBR_D.dat, etc.

 

This way we will have also the data for the second hard disk (the one you are now booting to) which are not actually needed, but this way we avoid confusion with shifted physicaldrive numbers and drive letters. 

 

Then put all the .dat files into a .zip archive and either attach the zip to your post or upload the .zip to a free file sharing site and post a link to it.

 

:duff:

Wonko



#7 matrix.rebooted

matrix.rebooted

    Newbie

  • Members
  • 26 posts
  •  
    France

Posted 04 September 2020 - 08:01 PM

done http://www.mediafire...PBR_all.7z/file

 

0 is current healthy win7home,
1 is faulty with win7pro in dualboot

as for PBR, hdhacker was unable to read it for faulty partitions on 2nd drive (F: and G:, says 'error') so I used tiny hexer and saved as .dat, hope it is the same format

 


Edited by matrix.rebooted, 04 September 2020 - 08:02 PM.


#8 wimb

wimb

    Platinum Member

  • Developer
  • 3756 posts
  • Interests:Boot and Install from USB
  •  
    Netherlands

Posted 04 September 2020 - 08:33 PM

In your healthy Win7 can you give a Screenshot of Windows Disk Management showing disks and partitions.

 

The first and the second entry of the partition table seem to be OK.

The first partition is set active (vista booting with bootmgr)



#9 matrix.rebooted

matrix.rebooted

    Newbie

  • Members
  • 26 posts
  •  
    France

Posted 04 September 2020 - 08:48 PM

@wimb

forgot to mention, I added this screenshot in the above archive! :)

 

just for you

LTMCXsd.png


Edited by matrix.rebooted, 04 September 2020 - 08:52 PM.


#10 wimb

wimb

    Platinum Member

  • Developer
  • 3756 posts
  • Interests:Boot and Install from USB
  •  
    Netherlands

Posted 05 September 2020 - 06:36 AM

MBR code + partition table OK - 1st partition is set active and PBR OK for booting with bootmgr

The 3 PBR seem to be OK, but Win7 PBR has NTLDR bootsector OK for booting XP.

Win7  can boot with bootmgr and Boot\BCD located on 1st partition (=vista)

 

Conclusion: Corrupt NTFS FileSystem of partition 1 and 2

 

Repair:

1. In Administrator Command Window try to use chkdsk to repair NTFS FileSystem Or use TestDisk

2. Try Convert RAW to NTFS EaseUs  Or DMDE Or AOMEIAOMEI

3. Otherwise Format and Apply Backup WIM File (probably not available) Or Re-Install Vista and Windows 7



#11 matrix.rebooted

matrix.rebooted

    Newbie

  • Members
  • 26 posts
  •  
    France

Posted 05 September 2020 - 07:09 AM

ok, thank you

 

I'll let you know, I hope first step will work, I would like not to format given that, for example, using parted magic, all disk and partitions content are fully accessible and usable.

 

Could you give me a little more details about the procedure using testdisk?



#12 Wonko the Sane

Wonko the Sane

    The Finder

  • Advanced user
  • 16066 posts
  • Location:The Outside of the Asylum (gate is closed)
  •  
    Italy

Posted 05 September 2020 - 07:10 AM

Repair: Format and Apply Backup WIM File (probably not available) Or Re-Install Vista and Windows 7

You must be using a strange dictionary, what is it , the Attila the Hun edition? :w00t:

(where "repair" means "totally obliterate"? ;))

 

@matrix.rebooted

As wimb said the MBR is just fine and the PBR's  are also seemingly fine (though they have "odd" offsets to the $MFT and $MFTmirr, but this may be perfectly normal and connected to your prior fiddling with resizing/moving/creating partitions with minitool).

It is more strange that Hdhacker can access E: (which it might not, as it is seen as RAW) and cannot access G: (which should be fine).

 

The diagnosis/conclusion by wimb is correct :) (but NOT the suggested remedy :ph34r: ), those two filesystems are somehow corrupted.(the MBR, PBR's and \boot\BCD are not the issue or at least not the main one), but they might well be repairable (by CHKDSK)

 

You could (should) get DMDE:

https://dmde.com/download.html

and have a look at the volumes with it.

 

Besides more advanced features you should use it to check the "indicators":

https://dmde.com/man...partitions.html

and report, this might give us a hint on what the issue might be.

 

Then, if needed, you should be able to access the contents of the disk with it and save/backup important/relevant files.

 

The actual normally suggested procedure (if you can afford a spare disk) is to make a full, "forensic sound" image of the whole disk (so you will need a disk slightly larger than that one).

 

Then (once you have the image, and so you can go back anytyme) you can try running CHKDSK.

 

Forget whatever you have read here and there about CHKDSK usage, you run it on each volume THREE times:

1) CHKDSK <driveletter> 

it will say something about having been run in read only mode and that there is the need to run it with the /F parameter (but actually it already made a few fixes)

then:
2) CHKDSK /F <driveletter>

hopefully it will take quite a bit of time and fix the filesystem

at this point (if it worked) the volumes should be not RAW anymore and you can run again CHKDSK <driveletter> that should not report particular issues

then

3) CHKDSK /R <driveletter> even if everything seem fine after the /F has run, it should be run, and surely it must be run if the /F did not work

 

:duff:

Wonko



#13 matrix.rebooted

matrix.rebooted

    Newbie

  • Members
  • 26 posts
  •  
    France

Posted 05 September 2020 - 07:43 AM

@Wonko the Sane

 

thank you sir for the supplement, I'll try all this as soon as possible

 

 

 

It is more strange that Hdhacker can access E: (which it might not, as it is seen as RAW) and cannot access G: (which should be fine).

yes (or no!) hdhacker can not access E: neither; for E: and F: I used tiny hexer

 

 

 

you will need a disk slightly larger than that one

which I haven't but files were already backed up, their should be no problem to format if needed but what I am interested in is to restore the disk boot configuration and menu, not necessarily the personal content; and indeed, I will surely try to 'fiddle' something before that ;)

 

 

 

(where "repair" means "totally obliterate"? ;))

If I remember correctly, I think I've added an option 'deban' to boot&nuke hard drive by pressing a key in my multiboot usb drive, that should be a solution too :lol:


Edited by matrix.rebooted, 05 September 2020 - 07:44 AM.


#14 wimb

wimb

    Platinum Member

  • Developer
  • 3756 posts
  • Interests:Boot and Install from USB
  •  
    Netherlands

Posted 05 September 2020 - 02:06 PM

Easier tool to do so is HDhacker:

http://dimio.altervista.org/eng/

 

 

Much easier is to use MBR_Backup using as Administrator batchprogram MBR_Backup.cmd Or use AutoIt program MBR_Backup_x64.exe

 

Automatically Make Backup of all Bootsectors of all Local Harddisks Fixed + Removable

More Info: https://msfn.org/boa...189-mbr-backup/



#15 matrix.rebooted

matrix.rebooted

    Newbie

  • Members
  • 26 posts
  •  
    France

Posted 06 September 2020 - 09:48 PM

hi,

have done some homework ^_^ but basically nothing new since I don't really know what to search for

 

 

MBR_Backup.cmd

Here it is, in this archive and with some screenshots and reports from DMDE

 

reboot.pro_reports.7z

 

chkdsk

Unfortunately, it keeps saying 'unavailable for RAW partitions'
Hiren's miniXP, win7PESE, win7RE all hang at boot and RE via win7 dvd gives same error message
Using right-clic/properties on a drive in windows explorer, it says something like 'drive locked, cant access'
Eventually, if you think it's not too risky, I could unplug the drive, boot a windows iso, then hotplug the data cable of the sata drive...
 

EaseUs, AOMEI

It seems to be mainly a backup and format solution with a lot of trials softs... backup is not needed, format can be done with free tools

 

 

 

TestDisk

(don't know how to repair filesystem with testdisk and didn't find a lot of details but it seems to imply formatting too.)
Some screens and log in the archive, it finds 2 of the 3 partitions (not win7) but when trying to show files it says 'filesystem is damaged'
Says boot sector is ok for part2 (win7) but bad for part1 and 3! (backup boot sector is zero everywhere)

 

dmde

Screenshots in the archive. When starting, it scans the drive and find CRC error in LBA 75 and the first of partition 2 (146785905), each times I choosed 'ignore'. Don't know what is LBA 75, first of partition 1 is 63.
Once finished, I can open the volumes and access the content.
When trying to build virtual FS, it says error in MFT
The indicators for each partitions are 'EBCF', volume parameters are in the screens too.

 

 



#16 Wonko the Sane

Wonko the Sane

    The Finder

  • Advanced user
  • 16066 posts
  • Location:The Outside of the Asylum (gate is closed)
  •  
    Italy

Posted 07 September 2020 - 07:28 AM

Ok.

 

Good :). (which means bad :().

 

TESTDISK won't repair that filesystem, the most it can do it is to fix some address errors (which for the moment seem fine) and restore the first few records of the $MFT by copying over it the $MFTMirr.

 

Hot-plugging the disk won't be of any help.

 

The backup boot sector being zero is not an issue, this can be replaced/rebuilt manually if needed (only for the recorf the backup bootsector is actually only one sector, the first sector of $Boot, which is the whole bootsector on NTFS and is 16 sectors in size.  

 

LBA means Large Logical Block Address, it is one of the two methods (the other being CHS or Cylinder Head Sector) used to address a sector on disk.

 

LBA 75 simply means "the sector that is at offset 75" or "the 76th sector on disk" (before it there are sectors 0-74), the MBR or first absolute sector on disk is LBA0.

 

In your case, since your first partition starts at LBA63, it is one of the sectors part of the $Boot of first partition, but I don't think - at first sight - that it can be the single cause of all these issues.

 

I'll have a look at the DMDE screenshots and - if I find something - I will provide instructions for further tests.

 

:duff:

Wonko


Edited by Wonko the Sane, 07 September 2020 - 08:44 AM.


#17 wimb

wimb

    Platinum Member

  • Developer
  • 3756 posts
  • Interests:Boot and Install from USB
  •  
    Netherlands

Posted 07 September 2020 - 08:26 AM

@martrix.rebooted

 

Thanks for the detailed info. 

It is strange that two drives in one operation of uninstall Win7 chipset drivers get Corrupt NTFS FileSystem.

Somehow I think it must be reversible also simply, but until now I have got no idea how to do so ....

 

It means that Format and Re-Install  remains as the best option, also because the data on these drives have already been backupped by you .....

 

 

LBA means Large Block Address, it is one of the two methods (the other being CHS or Cylinder Head Sector) used to address a sector on disk.

 

 

LBA means Logical Block Address   ;)



#18 Wonko the Sane

Wonko the Sane

    The Finder

  • Advanced user
  • 16066 posts
  • Location:The Outside of the Asylum (gate is closed)
  •  
    Italy

Posted 07 September 2020 - 08:37 AM

OK, I checked the screenshots and files.

The actual filesytem(s) seem just fine.

The issue is on the actual disk. :ph34r:

The two errors 23 you had (CRC error) mean that those two sectors cannot be read properly.

First error is on LBA 75, that since the volume starts on 63 is at relative offset 75-63=12 i.e. within the (16 sectors sized) of the $Boot on first volume.
Second error is on LBA 146785914, that since the volume starts on 146785905 is at relative offset 146785914-146785905=9 i.e. within the (16 sectors sized) of the $Boot on second volume.

Most probably the Windows mounting (and booting mechanism) accesses the $Boot as a monolithic file, so even if only one sector in it throws a CRC error, the whole $Boot is invalid (as seen by Windows).

This is confirmed by the (non-)files you got a results of the MBR_Backup (and that confirm the HDhacker issues earlier):
\MBR_Backup\HD1_0x1549F232\HD1_146785905.dat
\MBR_Backup\HD1_0x1549F232\HD1_63.dat
are 0 bytes in size.

It remains "queer", I mean that a bad sector is developed on a disk AND it is not intercepted by the hard disk controller AND it belongs to a $Boot is already rare or extremely improbable but that TWO sectors go bad AND BOTH are not fixed by the controller AND BOTH belong to (two different) $Boot files is something that would be good to feed to the Heart of Gold Improbability Drive for a nice round trip around the universe!  :w00t:

Anyway there are THREE possible way outs:
1) Make an image of the whole disk and restore it to a new disk (and then correct the bad sectors if needed)
2) Run the Manufacturer tool to diagnose (and hopefully repair) the disk
3) Try overwriting the sectors forcibly (if possible at all) with 00's (both sector 9 and 12 of a $Boot file are all 00's anyway)

The "safer" way would be #1, #3 is also safe (at the most it won't work) the #2 is theoretically safe but - only for the record - since it implies scanning the whole disk IF there is an ongoing development of bad sectors due to heat, head crash, etc. more bad sectors may develop.

Since you have the data backed up, excluded #1 for lack of means, I would go for #2 which is also the "official" way, usually it simply works, a minimal reference:
https://msfn.org/boa...comment=1186299

@wimb
Sure :), my bad :blush:  corrected.

 


It means that Format and Re-Install  remains as the best option, also because the data on these drives have already been backupped by you ....

 

I see you are still using that dictionary where "best option" is always "annihilate and destroy" :whistling:

 

:duff:
Wonko



#19 matrix.rebooted

matrix.rebooted

    Newbie

  • Members
  • 26 posts
  •  
    France

Posted 07 September 2020 - 08:59 AM

@wonko the sane

is something that would be good to feed to the Heart of Gold Improbability Drive for a nice round trip around the universe!


Lucky me!

Maybe this controller error is related to the nvidia bug (nvstor2), which is also happening with the second I'm currently using (since same motherboard, same drives!) but without freezes for now, so I let as is for now...

2) WDC tool is now running, reading the linked page ;)
cant access SMART data (as CrystalDisk too)
quick test says found bad sector that could be repair..
extended test on the run

3) how could I force write zeros in theses sectors?

@wimb

It is strange that two drives in one operation of uninstall chipset drivers get Corrupt NTFS FileSystem.

in fact they are 2 partitions of the same drive.

thank you both,

more to come in the next episode :)

just as a side note, I found back the BCD, indeed it was in the 1st partition (system and boot), was not looking in the right place ;)

 


Edited by matrix.rebooted, 07 September 2020 - 09:00 AM.


#20 wimb

wimb

    Platinum Member

  • Developer
  • 3756 posts
  • Interests:Boot and Install from USB
  •  
    Netherlands

Posted 07 September 2020 - 09:15 AM

Since MBR_Backup.cmd cannot make a backup of the PBR of Drive E: and F:

I suggest to recreate the PBR of these drives by using in your working Win7 :

 

Open Admin Command Window and give the following code to repair the PBR of Drive E: and F:

bootsect /nt60 E:
bootsect /nt60 F:

If it fails then add /force option

bootsect /nt60 E: /force
bootsect /nt60 F: /force

This is a safe operation that recreates the PBR Boot Code, that is probably corrupt in your case (LBA 75 has error is inside PBR bootcode)

 

[attachment=17348:BootSect_2020-09-07_112743.jpg]



#21 matrix.rebooted

matrix.rebooted

    Newbie

  • Members
  • 26 posts
  •  
    France

Posted 07 September 2020 - 09:26 AM

ok, should I try to repair bad sector with the Western Digital tool before or not?

 

I thought the PBRs saved with tiny hexer at the beginning, were ok?

 

eventually, since HP's factory_image (3rd part.) seems intact, I could use it to restore, then re-partition, then reinstall win7, will see after



#22 wimb

wimb

    Platinum Member

  • Developer
  • 3756 posts
  • Interests:Boot and Install from USB
  •  
    Netherlands

Posted 07 September 2020 - 09:34 AM

ok, should I try to repair bad sector with the Western Digital tool before or not?

 

I thought the PBRs saved with tiny hexer at the beginning, were ok?

 

eventually, since HP's factory_image (3rd part.) seems intact, I could use it to restore, then re-partition, then reinstall win7, will see after

 

The code given above by me does not destroy anything, it just repairs the PBR

So I advice you to do this first.

 

With Tiny Hexer you only saved the first sector of PBR, but there are 16 sectors of code that must be repaired probably.

MBR_Backup.cmd will make backup of 16 sectors of the PBR, which apparently fails in case of your RAW Drives E: and F:

 

Als you can have a look with TinyHexer at all 16 sectors of your Drives E: and F: and save in each case the 16 sectors as PBR_E_16.dat and PBR_F_16.dat



#23 Wonko the Sane

Wonko the Sane

    The Finder

  • Advanced user
  • 16066 posts
  • Location:The Outside of the Asylum (gate is closed)
  •  
    Italy

Posted 07 September 2020 - 10:02 AM

ok, should I try to repair bad sector with the Western Digital tool before or not?

Yes, and don't do anything else until the WDC tool has finished.

 

To try forcing the write of a single sector the best tool would be grub4dos dd command (or one of the potentially "dangerous" pseudo-low-level tool like mhdd, hddat or Victoria for Windows), but let's first see if the WDC tool manages to fix the issue/sector.

 

Then (if the WDC tool reports OK) using the bootsect command as wimb suggested is fine :), the tool is normally used to rewrite/correct the CODE re-writing the whole bootsector (i.e. the $Boot file) but the issue right now is at disk/controller level.

 

I would NOT attempt to use it until you can properly read/access that sector (i.e. a re-scan with DMDE must NOT show error 23) because we don't know exactly how the bootsect program works, it's theory of operation is:
1) read existing volume data (DATA)
2) merge these data to an internal bootsector template (CODE)
3) write the whole stuff (DATA+CODE) to the bootsector on disk

depending on the exact way these read and write operation are performed this may:

a. do nothing
b. succeed
c. fail BUT leave DATA intact
d. fail AND corrupt DATA

even if it is #d it should not be a problem, since all meaningful DATA is in the first sector (of which we have the Tiny Hexer backup/copy), still it is not worth the risk until we know if these two stupid sectors are "fixed" by the WDC tool and give not error anymore.

 

:duff:

Wonko



#24 wimb

wimb

    Platinum Member

  • Developer
  • 3756 posts
  • Interests:Boot and Install from USB
  •  
    Netherlands

Posted 07 September 2020 - 10:28 AM

Before anything else:

 

Have a look with TinyHexer and Load 16 sectors of the PBR of your Drives E: and F: and save the 16 sectors as PBR_E_16.dat and PBR_F_16.dat

 

That makes it possible later to use TinyHexer Tools > Compare function and see what has changed on fixing the Errors ....

 

You can use already Tools > Compare to compare PBR_E_16.dat with your HD0_2048.dat PBR file of your NTFS drive that is OK and see what are the differences ....



#25 Wonko the Sane

Wonko the Sane

    The Finder

  • Advanced user
  • 16066 posts
  • Location:The Outside of the Asylum (gate is closed)
  •  
    Italy

Posted 07 September 2020 - 10:48 AM

Before anything else:

 

Have a look with TinyHexer and Load 16 sectors of the PBR of your Drives E: and F: and save in each case the 16 sectors as PBR_E_16.dat and PBR_F_16.dat

 

That makes it possible later to use TinyHexer Tools > Compare function and see what has changed on fixing the Errors ....

 

You can use already Tools > Compare to compare PBR_E_16.dat with your HD0_2048.dat PBR file of your NTFS drive that is OK and see what are the differences ....

If a sector gives CRC error, it gives CRC error and it cannot be read, this happens at a much lower level than the Windows HAL and disk drivers, there is simply no way that Tiny Hexer can read that.

 

And there is no need to compare anything with anything else, a single sector (two times) which is always 00 cannot be read, we don't need to read what we cannot read because we already know that IF we could read it it would be all 00's. (Schroedinger's sector is surely 00's ;) )

 

The other sectors of the $Boot file are not an issue, the bootsect.exe can rebuild them any time, since the DATA is OK (and we already have a backup of that).

 

:duff:

Wonko






1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users