Jump to content











Photo
- - - - -

truecrypt bootloader question


  • Please log in to reply
53 replies to this topic

#26 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 02 September 2009 - 08:34 AM

I cant find it.. each link you drop has 10 or more other links.

how do I use dencorso's compiled method?


You run it? :whistling:

Remember:

Life is "trying things to see if they work"


Four easy steps:
  • get HDhacker (or use a disk editor or whatever tool you are familiar with):
    http://dimio.altervista.org/eng/
  • extract/save the bootsector from your drive (first sector of Logicaldrive), to say, C:\mysector.bin
  • run
    killchs c:\mysector.bin
  • restore the patched file to your drive

jaclaz

#27 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 02 September 2009 - 02:09 PM

and then those 10 other links have at least 10 more links. and many links have 10+ pages some of which Ive read from page 1 to end several times allready. its intresting, sometimes I end up reading threads that have nothing todo with the problem at hand but in reading them I find other areas of intrest that I feel compelled to research.

this makes rabbit holes look like walking in the park! for right now I really want to get TC booting USB in this notebook so I can move on to the other stuff.


I will give it a shot as soon as I can and report back success or failure again.

thanks

#28 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 02 September 2009 - 02:43 PM

this makes rabbit holes look like walking in the park!

Murphy's Laws apply on real rabbit holes:
http://www.wwmag.net/catchrabit.htm

Let alone the virtual rabbit holes on the internet!

You need a looong stick, and patience.....

:whistling:

jaclaz

#29 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 03 September 2009 - 01:30 AM

no go/ fail.

first I quick formatted the USB HDD NTFS single partition with XP then I ran MBRWiz and saved the MBR then I ran HDHacker exactly as Jaclaz suggested and then ran Killchs then HDHacker again then rechecked to verify Killchs changes then I booted internal HDD with fresh XPsp2 and only USBoot installed and ran phase-III to copy OS to USB HDD then I booted USB HDD successfully and installed TC

then rebooted and failed TC PBA with same blinking cursor after password entered! MAN.


so then I decided to try Killchs on the USB flashdrive I have that successfully boots encrypted OS in my desktop but not in the notebook.

I get the message "Signature not fount. Not an NTFS/FAT32 or already patched"

and viewing the bootsector with HDHacker shows nothing recognizable, just lots of strange characters. I thought TC didnt encrypt the boot sector but it must do something there because it doesnt ressemble any other boot sectors I have.


anyone have suggestions? I dont know what to try next. I'll reread this thread again too. if I cant find a next step I think I may try to learn Grub4dos and see if I can manage to help it boot somehow but I have no experience with Grub4dos nor do I know if thats the best next step for solving this. I still have a feeling that TC is not finding the MBR it needs when attempting to boot in the notebook but I dont know how to test that.

hmmm.. what next?

edit: BTW I really appreciate all the helpful suggestions so far! last week I didnt know much about the bootsector or MBR at least now I can view and save them! Thanks

#30 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 03 September 2009 - 01:44 PM

Let see if we can go a little further. :)

The MBR is the first sector of the whole disk (Physicaldrive).

The bootsector is the first sector of the partition (Logicaldrive).

In the MBR you should check that the CHS and LBA data is balanced, use beeblebrox and my spreadsheet as reference:
http://www.boot-land...?showtopic=2959
http://www.boot-land...?...c=2959&st=9

(this is likely to be part of the problem of why a motherboard is booting or not)

The bootsector seems to be OK, it should be patched only once.

It is possible that the problem with the laptop is due to it's "peculiar" way to sense devices, is it a Lenovo, by any chance?

Many Lenovo's use a geometry of n x 240 x 63 instead of the usual n x 255 x63.

Also, you may want to try (please read SHOULD try) using one of the utilities, RMPREPUSB or USB_XP_Setup.exe
http://www.boot-land...?...ic=7739&hl=
http://www.boot-land...?showtopic=5306
that create a second (unused) partition.
This last method has been found to be needed in order to recognise a USB stick as "HDD-like" devices buy a few "bad" BIOSes, though AFAICR they were only desktop and not notebooks, it is possible that your peculiar notebook behaves like that.

The alternative is try using fbinst, it is "guaranteed" to boot on "any" PC, but it's up to you if it is "compatible" with TC :), something that I personally doubt. ;)

jaclaz

#31 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 03 September 2009 - 07:25 PM

see, thats the thing.. (sent hardware details in PM) it boots XP on several USBs without TC.

I have a 4GB bootable USB flash with TC system encryption installed that boots in my desktop but not the notebook but all my recent set ups have been on a 40GB USB HDD and I ran accross a post that mentioned the larger USB drive may cause problems (it boots except when TC PBA is installed) its faster than flash is why I use it for R&D

I dont know how to image the encrypted 4GB stick yet or I would save it and use it for further R&D


I'll keep following your suggestions, you have been the most help so far, we havnt solved this issue but Im learning more in the proccess. I do hope we can solve it. there will be no lack of effort on my part but my time for experimenting is limited.

I have considered "online's" approach:

http://www.boot-land...?showtopic=5315

but I since I have a USBoot prepaired stick I should be able to shorten the steps involved quite a bit. (hopefully)

also "mr_'s" tecnique on wiki with Grub4dos is an option to look at but I would really rather just have single boot USB directly. its something really simple here Im sure. we will be laughing when we find it.

I really do want to experiment with Grub4dos, preferably after this is solved.

now do I check the CHS/LBA balance with TC or without?

#32 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 03 September 2009 - 11:45 PM

I dont know how to image the encrypted 4GB stick yet or I would save it and use it for further R&D

Easy (please read as complex :)):
use dsfo. (just read the readme.txt inside the DSFOK package)

Complex (please read easy :)):
use any of the several imageing utilities with gui and "advanced" features like compression and the like:
suggested, among the several available ones, for which an almost hopefully complete list is here:
http://www.msfn.org/...re-t100299.html

Clonedisk:
http://www.boot-land...?showtopic=8480

Odin:
http://www.boot-land...?showtopic=5871

Silinux dd for winddows:
http://www.boot-land...?showtopic=7786


now do I check the CHS/LBA balance with TC or without?

The partition entries (please read MBR DATA) should be the same no matter which MBR CODE ("normal" 2K/XP, grub4dos, TC DOS/win9x, Linux) is used.

:)

jaclaz

#33 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 04 September 2009 - 02:01 PM

Jaclaz, are you sure you are not overlooking the fact that this motherboard and bios boot USB fine including many different USB and with different OS's?

its only upon TC PBA that it fails to boot with a blinking cursor after password is entered but the same TC encrypted XP stick boots fine in other machine. this problem is reproduced on otherr sticks. Im just asking to be sure.

in referance to the blinking cursor at boot, I keep reading it is because the MBR is not found. is this correct? ifso, why would TC booloader have trouble on one machine and not another?


also, Ive read that encrypted systems can only be imaged in RAW format, is this correct?

thanks, going to use beeblebrox now..

edit: any suggestions for a small freware to view .xls files?

#34 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 04 September 2009 - 06:45 PM

Jaclaz, are you sure you are not overlooking the fact that this motherboard and bios boot USB fine including many different USB and with different OS's?

its only upon TC PBA that it fails to boot with a blinking cursor after password is entered but the same TC encrypted XP stick boots fine in other machine. this problem is reproduced on otherr sticks. Im just asking to be sure.

Yep. :D
You are describing perfectly a situation in which the TC PBA (whatever is a PBA ;)) has proved to be incompatible with that particular machine BIOS only, or, to be more accurate the combination of the TC PBA, of the BIOS and of the methods with which you partition/format your sticks produces a non-booting situation in that particular case only.
The point is whether this proved incompatibility is due to a "bug" in either the TC or in the BIOS (for which we can do nothing about) or to something else (DATA that we can try to correct/change in the way the stick is partitioned/formatted)

in referance to the blinking cursor at boot, I keep reading it is because the MBR is not found. is this correct? ifso, why would TC booloader have trouble on one machine and not another?

Usually the blinking cursor has nothing to do with the MBR not found, but with the bootsector not found, or with something else not found or loaded improperly.

You see, the MBR is ALWAYS found, as it is ALWAYS first sector of hard disk and ALL BIOSes on ALL PC's boot by loading the first sector of (first) hard disk.
Once the MBR code is loaded, it looks for the MBR data, and due to a combination of factors, it may wrongly "understand" the data and instead of loading the bootsector, load something else, OR (and this is probably the case with the blinking cursor) it is the bootsector CODE that has problems in interpreting EITHER:
  • the MBR DATA
  • the bootsector DATA
  • other DATA coming forom the BIOS



edit: any suggestions for a small freware to view .xls files?

NOT VIEWERS ONLY (a simple viewer won't be of much help for using my spreadsheets):

Not Freeware, but affordable, the best app around (well, maybe not the best, but the most portable and small):
Spread32:
http://www.byedesign.co.uk/

It is fully functional with a little nag when you open and close a file, try it, once you will have tested you'll probably find that those would be the best $20 bucks you ever spent on a software.

Alternatives (freeware) listed here:
http://www.boot-land...?...ic=8898&hl=

Another Freeware VIEWER ONLY (only smart if you ALREADY have .NET ;) installed):
http://bytescout.com...d_freeware.html
http://bytescout.com/pad/xlsviewer.zip

:)

jaclaz

#35 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 05 September 2009 - 04:42 AM

OK, I just wanted to make sure we are on the same page with my problem here and again I can not thank you enough for sticking with me on this. youre helping me tremendously regardless if we solve this or not.

PBA = Pre Boot Authentication

I think we can rule out that TC PBA is incompatable with this particular machines BIOS because I can install TC on it directly on the internal HDD and no problems. the problem only occurs with this machine when booting USB. does that tell you it is most likely something we can solve?

I will have a look at your spread sheets asap and try to compare with my CHS/LBA Balance as you suggest allthough Im not sure what to look for however I should have something more to go by. thanks

#36 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 05 September 2009 - 07:26 AM

I think we can rule out that TC PBA is incompatable with this particular machines BIOS because I can install TC on it directly on the internal HDD and no problems. the problem only occurs with this machine when booting USB. does that tell you it is most likely something we can solve?


There is no way to know.

It is possible that the BIOS routines that access the internal HD are slightly different from those used on the USB, as an example - but please DO NOT take this as "truth", just a guess, it is possible that the BIOS expects the mass storage device connected to the USB BUS to be conforming to a given HD barrier or geometry, for the sake of reasoning, let's assume the 512 Mb CHS one:
http://www.pcguide.c...izeMB504-c.html

But the fact that a "normally" formatted USB stick works allright on this machine should mean that it is likely to be the TC code that needs some particular settings.

jaclaz

#37 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 05 September 2009 - 03:06 PM

Im looking at the USB HDD in Beeblebrox and I have loaded your CHS_LBA_v2.xls in Spread32. nice work Jaclaz! that looks like it was a quite a task to create.

Do I enter the data from Beeblebrox into your spreadsheet or am I only comparing info and ifso is there anything specifically I should start with?

" But the fact that a "normally" formatted USB stick works allright on this machine should mean that it is likely to be the TC code that needs some particular settings


it has been a widely reported issue on the TC forums, Im starting to think that this may indeed be the case. it may be time for a different tecnique. Im looking into online's guide using Grub4dos and a second partition. I am anxious to learn about Grub4dos anyway however I would like to have your recomendation on this matter before procceding.

eitherway, I would like to learn how to use your spread sheet and Beeblebrox for future referance if nothing else.

#38 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 05 September 2009 - 04:38 PM

eitherway, I would like to learn how to use your spread sheet and Beeblebrox for future referance if nothing else.


Well, basically you enter in the spreadsheet (on the appropriate sheet, "PTtables" obviously ;))
  • the geometry of the drive
  • optionally the number of hidden sectors
then enter EITHER:
  • the CHS values (values you can actually see in a partition table DATA)
  • the LBA values
  • the Sectors Before/NumSectors (values you can actually see in a partition table DATA)

The calculated data (one of the other two sets of values you have NOT entered) should be same as what you can find in partition data.

If any cell on PTtables has a RED BACKGROUND, it is very probable that a problem exists with the data entered.

The MBR_to_PT and PT_to_MBR are useful if you want to "play" with a hexeditor or just experiment.

All other sheets are just a "split" of the 2 sheets above, to (hopefully) explain how the DATA is stored in a partition table.

jaclaz

#39 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 05 September 2009 - 06:37 PM

therere are several cells with red background in CHS_LBA_v2.xls before I enter any data.

I entered the geometry and then imediately noticed that the "NumSectors" in Beeblebrox does not match the "Tot Sectors" in CHS_LBA_v2.xls

NumSectors = 78135296
Tot Sectors = 78,140,160

I dont know if this means anything but seems worthy to mention.

next, when I try to enter the "Partition Table Values" that are listed in Beeblebrox, starting with the first box next to "0-" is "type" =07 but when I try to enter that value in CHS_LBA_v2.xls I get a "Cell is protected. Unprotect sheet first." message. and the same with the next value "Boot" =80


I am able to enter the data for BCyl, BHd, BSec, ECyl, EHD, ESec

should I "Unprotect the sheet first" to continue?

#40 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 05 September 2009 - 07:12 PM

therere are several cells with red background in CHS_LBA_v2.xls before I enter any data.

Yes, that's "by design" as the Microsoft guys would put it.

I entered the geometry and then imediately noticed that the "NumSectors" in Beeblebrox does not match the "Tot Sectors" in CHS_LBA_v2.xls

NumSectors = 78135296
Tot Sectors = 78,140,160


I dont know if this means anything but seems worthy to mention.


It means that you have unbalanced CHS/LBA geometry, please post all the values you see in beeblebrox.

next, when I try to enter the "Partition Table Values" that are listed in Beeblebrox, starting with the first box next to "0-" is "type" =07 but when I try to enter that value in CHS_LBA_v2.xls I get a "Cell is protected. Unprotect sheet first." message. and the same with the next value "Boot" =80
I am able to enter the data for BCyl, BHd, BSec, ECyl, EHD, ESec

My bad ;), I must have forgotten to "uncheck" those cells (the "Type" and "Boot" ones)

should I "Unprotect the sheet first" to continue?

Yep ;), unprotect it, select and unselect for "protection" the cells D10...E13, D16...E19, D22...E25 and re-protect the worksheet (the protection is NOT needed it is only there to avoid inadvertently inserting data overwriting a formula.

:)

jaclaz

#41 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 05 September 2009 - 07:39 PM

It means that you have unbalanced CHS/LBA geometry, please post all the values you see in beeblebrox.


so mabey it is a DATA issue here (fingers crossed) here is what I have in Beeblebrox:



Sector 0 -MBR
Size in Megabytes : 38,154.00MB
Total Sectors on Disk : 976768064

Disk Geometry : Cyls 4864, Heads 255, Sectors 63


Partition Table Values

Type = 07
Boot = 80
BCyl = 0
BHd = 32
BSec = 33
ECyl = 1023
EHD = 254
ESec = 63
StartSector = 2048
NumSectors = 78135296

File System : IFS (NTFS/HPFS)
Size (MB) : 38152.0
Absolute Start : 2048


I really hope we can figure this out, Ive briefly looked at other possible methods to encrypt OS on USB and I think most of them would take me months to complete and still the possibility that the issue remains and not to mention new ones may be created in the proccess.

the only one Ive found that may be worth the gamble is NK!'s hybrid set-up on the CD fourm

http://www.911cd.net...showtopic=21445

Im still reading up on that one but it seems less time consuming and involved than the others mainly due to incorporating USBoot.

#42 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 06 September 2009 - 10:29 AM

Values:
07-80-0-32-33-1023-254-63-2048-78135296

Don't make much sense :D, they seem to me like a drive deriving from an image exported with a MBR from a .img with Winimage (that may cause the "33") later re-accessed/fiddled with with diskpart under Vista WITHOUT the registry Fix (which may cause the "2048"):
http://www.911cd.net...showtopic=21186

I refuse to believe that RMPREPUSB created those values. ;)

A CHS 4864/255/63 should have the "virtual" values:
07-80-0-1-1-4863-254-63-63-78140097
Since we don't know whether it really has a "full" 4864th cylinder, let's take one less:
07-80-0-1-1-4862-254-63-63-78124032

which translates to (ACTUAL VALUES):
07-80-0-1-1-1023-254-63-63-78124032

Let's start from scratch, OK? ;)

  • 0 out the MBR (disk editor or dsfo)
  • Apply a standard MBR (MBRFIX or similar)
  • Write the above "actual" values with beeblebrox
  • disconnect the drive
  • re-connect the drive
  • Format the drive through disk management

jaclaz

#43 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 06 September 2009 - 04:47 PM

Values:
07-80-0-32-33-1023-254-63-2048-78135296

Don't make much sense w00t.gif, they seem to me like a drive deriving from an image exported with a MBR from a .img with Winimage (that may cause the "33") later re-accessed/fiddled with with diskpart under Vista WITHOUT the registry Fix (which may cause the "2048")

I refuse to believe that RMPREPUSB created those values.


novice experiments.. I think that after I tried RMPREPUSB without success I zeroed it again then initialized and formated with Vista that was from a DriveImageXML Image.

as of now I have followed the six steps in your last post exactly and it is now formating (not quick format) NTFS.

I will make a back-up MBR now before procceding then I copy USBoot XPsp2 OS onto it and then retry TC. will there be no need for HPtool or other?


I want study more about balancing and your CHS_LBA_v2.xls, I want to learn how to check and balance any MBR so first I can check all my exsisting drives but also be able to check/balance other ones. is there more documentation available for this?

edit: also, quick question: after formatting NTFS I have allways unchecked "Allow Indexing Service to index this disk for fast file searching" is this acceptable practice?

#44 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 06 September 2009 - 05:35 PM

will there be no need for HPtool or other?

Let's see if it works "as is".
The HP tool would make definitely an "unbalanced" CHS/LBA (which since you are beyond 8Gb/Cylinder 1023 doesn't matter) but will also not respect Cylinder boundaries, which may be a problem.
The "good" thing about the HP tool is the MBR hack, but it probably won't be needed, and one can change those few bytes manually with a Disk editor or using one of my batches.


edit: also, quick question: after formatting NTFS I have allways unchecked "Allow Indexing Service to index this disk for fast file searching" is this acceptable practice?

Sure :), it is actually (at least by me ;)) RECOMMENDED, it is one of the stupidest things ever added by MS:
in order to find quickly something (that you search for once in a lifetime), you have the system behaving sluggishly because of the indexing service :)

If you want to find something quickly forget bult-in "Search" and use one of these:
http://www.boot-land...?...c=6848&st=2
http://www.boot-land...?showtopic=2898

jaclaz

#45 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 06 September 2009 - 06:51 PM

It does indeed boot without other tools but this is a USB HDD however installing TC PBA Fails again.

what should I do from here to get past this unknown issue?

we know TC'd XP boots on this notebook (Intel) directly, we know it boots from USB directly in another machine (AMD)

we dont know (yet) why TC does not boot USB in this machine.

we also know I cant get to the bottom of this myself so thanks for all the lessons so far Master Jaclaz!

#46 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 07 September 2009 - 02:59 AM

now this is intresting..

I decided to try DiskCrypt a true opensource system encryption program. it does not have a pre boot test so you have to encrypt the drive without a boot test. it failed with blinking cursor again.

also has no option provided to remove its bootloader. I zeroed out the drive with dsfok then restored MBR with MBRFix and then used disk management to initialized and format quick NTFS

then I took a look at from beeblebrox and noticed that beeblebrox indicates a FAT16B file system. I rechecked in disk management and it indicates NTFS

I go ahead and run USBoot Phase-III to copy the XP onto the USB HDD and use the option to change the computer name which I havnt done before.

it boots OK and actually seems to boot faster!?!

so I install TC again and this time after I enter the password the screen blinks and looks the same "enter password" then I enter the password again and I get a single beep code and this message:

error: memory currupted
-upgrade BIOS
-Use a different motherboard model/brand


very odd. something is going on.. I'd really like to know what even though apparantly I will not be booting a TC'd XP with USB on this machine right away.

I started looking in the BIOS screens to see any options I might have missed and the "flash write protection" option is greyed out.. I cant change it! along with several other greyed out options.

I guess I will contact the manufacturer to see if its possible to change the greyed out parts in the BIOS.

still open for suggestions and willing to try just about anything at this point.


thanks

#47 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 07 September 2009 - 08:19 AM

then I took a look at from beeblebrox and noticed that beeblebrox indicates a FAT16B file system. I rechecked in disk management and it indicates NTFS


I don't believe you, sorry. ;)

Apart from the above, this is how I see you last report:

As soon as you had at least valid MBR data, you did everything in your power (and even more) to fiddle with that stick in "random" or "semi-random" ways. :)

When troubleshooting a problem it is advised to:
  • have a plan (if possible)
  • do things in sequence along the plan (if existing)
  • do things one step at the time

I thought we had already established that the problem might be caused by:

  • MBR DATA
  • MBR CODE
  • Bootsector DATA
  • Bootsector CODE
  • BIOS (and we cannot do anything about it)
  • TC PBA (and we cannot do anything about it)

Out of the 4 things we can experiment with, we only started one of the several tests on first item, and you start going astray, adding variable and overlays to the thing? :)

If this is "pure" booting problem, you need to keep it simple!

DO NOT run the USBoot on it.
DO NOT run any other software.

Start with the normal MBR, with the (hopefully) right MBR DATA.
Add ONLY:
  • NTLDR
  • NTDETECT.COM
  • a BOOT.INI with at least two entries in it

If you can get to the BOOT.INI choices, that means that the thingy boots.
Make a copy of the MBR and bootsector.
Add this TC PBA.
WHAT does it change?
The MBR (compare old with new)?
The Bootsector (compare old with new)?
Does it add any file to the partition?


Side question:
have you tried with a smallish partition (smaller than the 8 Gb limit)?
is it possible that the problem is an incompatibility of some kind with larger partitions?

jaclaz

#48 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 07 September 2009 - 04:26 PM

I don't believe you, sorry. sad.gif


I dont know how it happened, surely it was a mistake on my part but that is what I observed. I did not at any point specify a FAT16b format but at some point that is what Beeblebrox indicated while at the same time disk management indicated NTFS. mabey because I did not disconnect and re-connect, it seems that is required after making changes to MBR on USB Im not sure.

As soon as you had at least valid MBR data, you did everything in your power (and even more) to fiddle with that stick in "random" or "semi-random" ways. w00t.gif


I was anxious but I think I could get back to "at least a valid MBR data" by retracing the steps you have provided.

Start with the normal MBR, with the (hopefully) right MBR DATA.
Add ONLY:

* NTLDR
* NTDETECT.COM
* a BOOT.INI with at least two entries in it


I made it this far from instructions on USBoot for pre-test prior to running USBoot and can report success

Make a copy of the MBR and bootsector.

I will start over, I thought when I saved MBR it included the Bootsector.



Add this TC PBA.
WHAT does it change?


must first have an OS to install TrueCrypt (TC) then its Pre Boot Authentication (PBA)

so in order for me to get to the PBA I have to run USBoot to copy OS to USB then install TC.

I can compare MBR and Bootsector after adding USBoot OS and then again after adding TC PBA

Side question:
have you tried with a smallish partition (smaller than the 8 Gb limit)?
is it possible that the problem is an incompatibility of some kind with larger partitions?


yes, originally I started with a 4GB flash drive. it boots with TC PBA on my desktop but fails on the notebook. I have left it as-is since it works and because it is flash, experimenting with it slows everything down so I started over with USB HDD. if I procceed with it again I need to learn how to image it (RAW format according to reports since it is encrypted) then re-produce that image on another device to be sure. I am not sure if I can image the 4GB in RAW and reproduce that image on a different size stick. I dont have another 4GB stick only an 8GB


now that I have experimented with DiskCryptor a little bit I have been able to get it further along than the blinking cursor. with DiskCryptor I can get as far as loading the USBoot arcguard then it freezes blue screen with "UNMOUNTABLE_BOOT_DEVICE"

at this point it DiskCryptor seems more compatable than TC

it is also true opensource where as TC according to wiki is not

The TrueCrypt Collective License does not meet the Open Source Definition, and thus has not been approved by the Open Source Initiative. It is considered "non-free" by all the major GNU/Linux distributions (Debian[21], Ubuntu[22], Fedora[23], openSUSE[24], Gentoo[25]). The license is considered "non-free" mainly because of distribution and copyright-liability reasons.[26]

http://en.wikipedia.org/wiki/TrueCrypt

DiskCryptor also seems to have more accessable developers to contact and they apear to be willing to work with these types of issues

http://diskcryptor.n...x.php/Main_Page

for those of us that have compatibility issues with booting TC on USB we may have a better shot at "Whole Drive Encryption" (WDE) booting from USB with DiskCryptor.

Im willing to go in which ever direction you recomend Jaclaz.

#49 Zoso

Zoso

    Silver Member

  • Advanced user
  • 640 posts
  •  
    Isle of Man

Posted 08 September 2009 - 12:12 AM

minor set back.. while I was starting over from scratch again somehow Disk 1 and Disk 2 changed positions in the proccess and so all zeros meant for Disk 2 ended up on Disk 1 and now I know that is irreversable. good thing everything was backed up prior.

however that was my back up that got wiped out so now Im starting over with my back up drive from scratch.

same thing happened on this drive too. I zero'd it out then MBRFix new MBR, disconnect reconnect then used Disk Management to make one primary active and one logical bith NTFS then I use beeblebrox to view and save changes and notice it reports the file system as FAT16B

now to make you a beliver Jaclaz I installed IrFanView and made a screen capture:

http://p1.bikepics.c...775032-full.jpg

Im curious why this was hard to belive and If mabey we should find an alternative to beeblebrox? also, is FAT16B unusual, Ive only heard/read FAT16

anyway, there you have it.. I'll have to get back to the TC issue after I get my backup drive BACK UP.

#50 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 08 September 2009 - 07:49 AM

same thing happened on this drive too. I zero'd it out then MBRFix new MBR, disconnect reconnect then used Disk Management to make one primary active and one logical bith NTFS then I use beeblebrox to view and save changes and notice it reports the file system as FAT16B

now to make you a beliver Jaclaz I installed IrFanView and made a screen capture:

http://p1.bikepics.c...775032-full.jpg

Im curious why this was hard to belive and If mabey we should find an alternative to beeblebrox? also, is FAT16B unusual, Ive only heard/read FAT16

anyway, there you have it.. I'll have to get back to the TC issue after I get my backup drive BACK UP.


Quite obviously one of the two (either MMC or beeblebrox) have been started or accessed the data in different moments, a change has been made and one of the two have not been refreshed, which may also mean, for a USB drive to eject and re-insert it.

Really cannot say what can produce that effect.

What you need an alternative for (in case) is for the Disk Management MMC. :(

Just open the disk in a disk editor and check whether byte at offset 450 reads 06 or 07.

There are two types of FAT16 partition:
  • CHS mapped 06
  • LBA mapped 0E
A partition bigger than 8 Gb (CHS limit) with partition type 06 is simply insane.

For the record, you don't need to install a specific app to do a screen capture, you only need pressing the Print Screen key and paste in a new PAINT image (and if it's the case CROP the relevant part and re-paste into a new PAINT image).

jaclaz




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users