23 replies to this topic

[Doodoo]

Hi everyone,

I've just treated myself with a new 32 Gb USB stick, and so I headed straight away to get the latest releases of RMPrepUSB and BootIce to format it... but i've got very strange results (and behaviours when booting !)

First I tried RMPrepUSB 2.1.648
Formated the stick in HDD mode (2 partitions), all NTFS.

After formatting, here is what "Drive Info" gives me:

Disk Signature bytes (1B8h-1BBh) = CA 5C CF 6E



Partition 1   SIZE=30531.531MiB   Type: 07 NTFS  *ACTIVE*

START POS   = CYL:0 HD:32 SEC:33  END POS = CYL:1023 HD:254 SEC:63

START (LBA) = 2,048 (00000800)	  SIZE (LBA) = 62,528,576 (03BA1C40)



Partition 2   SIZE=0.031MiB   Type: 21 Hidden(rsvd)	  

START POS   = CYL:1023 HD:254 SEC:63  END POS = CYL:1023 HD:254 SEC:63

START (LBA) = 62,530,624 (03BA2440)	  SIZE (LBA) = 63 (0000003F)



Partition 3   SIZE=0MiB   Type: 00		

START POS   = CYL:0 HD:0 SEC:0  END POS = CYL:0 HD:0 SEC:0

START (LBA) = 0 (00000000)	  SIZE (LBA) = 0 (00000000)



Partition 4   SIZE=0MiB   Type: 00		

START POS   = CYL:0 HD:0 SEC:0  END POS = CYL:0 HD:0 SEC:0

START (LBA) = 0 (00000000)	  SIZE (LBA) = 0 (00000000)



P1   Start=2,048 (1,048,576 bytes) End=62,530,623 (32,015,678,976 bytes)

P2   Start=62,530,624 (32,015,679,488 bytes) End=62,530,686 (32,015,711,232 bytes)





Drive 1 SanDisk Ultra Backup is 32,015,679,488 bytes long (29.817GiB)

F/W Rev.=1.20  Serial No.=20043514 [ 32 30 30 34 33 35 31 34 ]

What looks really wierd is that the second partition ends beyond the disk boundary... Is this intentional ? Or am I missing something ?

Next I tried BootIce 2012-05-06
When I do "Parts Manage", "Reformat USB Stick", "Next" (no matter which disk mode I chose), I always get an error message "Invalid Number" ... Can anyone reproduce this ?

Next I tried BootIce 0.9-2011-0330
Same procedure, configured a first NTFS partition of 30533 Mb, and a second FAT16 partition of 0 Mb.
Managed to get past the buggy interface... Eventually made it to the actual formating.

Now here is what DriveInfo gives me:

Disk Signature bytes (1B8h-1BBh) = CA 5C CF 6E



Partition 1   SIZE=1866.898MiB   Type: 07 NTFS  *ACTIVE*

START POS   = CYL:0 HD:1 SEC:1  END POS = CYL:237 HD:254 SEC:63

START (LBA) = 63 (0000003F)	  SIZE (LBA) = 3,823,407 (003A572F)



Partition 2   SIZE=28665.603MiB   Type: 0C FAT32LBA (2047GB max)	  

START POS   = CYL:238 HD:0 SEC:1  END POS = CYL:1023 HD:254 SEC:63

START (LBA) = 3,823,470 (003A576E)	  SIZE (LBA) = 58,707,154 (037FCCD2)



Partition 3   SIZE=0MiB   Type: 00		

START POS   = CYL:0 HD:0 SEC:0  END POS = CYL:0 HD:0 SEC:0

START (LBA) = 0 (00000000)	  SIZE (LBA) = 0 (00000000)



Partition 4   SIZE=0MiB   Type: 00		

START POS   = CYL:0 HD:0 SEC:0  END POS = CYL:0 HD:0 SEC:0

START (LBA) = 0 (00000000)	  SIZE (LBA) = 0 (00000000)



P1   Start=63 (32,256 bytes) End=3,823,469 (1,957,616,128 bytes)

P2   Start=3,823,470 (1,957,616,640 bytes) End=62,530,623 (32,015,678,976 bytes)





Drive 1 SanDisk Ultra Backup is 32,015,679,488 bytes long (29.817GiB)

F/W Rev.=1.20  Serial No.=20043514 [ 32 30 30 34 33 35 31 34 ]

Conclusion: first partition is NTFS and of size 1.9 Gb, second partition is FAT32 and 28 Gb.... Not quite what I asked !

Why has it got to be so damn hard to format a USB stick

[wimb]

May be you can try BOOT_USB.exe - Format USB-stick e.g. NTFS with FAT32 2nd partition

http://www.911cd.net...showtopic=24424

Also USB_Part_Flip.exe - Change which primary partition of USB-stick is visible in Windows

[Doodoo]

May be you can try BOOT_USB.exe - Format USB-stick e.g. NTFS with FAT32 2nd partition

Thanks for the tip !
Not sure if that was caused by BootIce screwing the partition table in the first place... but BOOT_USB always seems to leave the partition table unchanged, I can't get the 2 partitions I want, even when I check "2nd partition", and no matter the size....

[Doodoo]

At least BOOT_USB seems to be getting the partition table right (with only one partition though) !!
(62,524,917 + 63) / 63 / 255 = 3892 which is the reported number of cylinders by all tools !

Disk Signature bytes (1B8h-1BBh) = CA 5C CF 6E



Partition 1   SIZE=30529.745MiB   Type: 07 NTFS  *ACTIVE*

START POS   = CYL:0 HD:1 SEC:1  END POS = CYL:1023 HD:254 SEC:63

START (LBA) = 63 (0000003F)	  SIZE (LBA) = 62,524,917 (03BA0DF5)



Partition 2   SIZE=0MiB   Type: 00		

START POS   = CYL:0 HD:0 SEC:0  END POS = CYL:0 HD:0 SEC:0

START (LBA) = 0 (00000000)	  SIZE (LBA) = 0 (00000000)



Partition 3   SIZE=0MiB   Type: 00		

START POS   = CYL:0 HD:0 SEC:0  END POS = CYL:0 HD:0 SEC:0

START (LBA) = 0 (00000000)	  SIZE (LBA) = 0 (00000000)



Partition 4   SIZE=0MiB   Type: 00		

START POS   = CYL:0 HD:0 SEC:0  END POS = CYL:0 HD:0 SEC:0

START (LBA) = 0 (00000000)	  SIZE (LBA) = 0 (00000000)



P1   Start=63 (32,256 bytes) End=62,524,979 (32,012,789,248 bytes)





Drive 1 SanDisk Ultra Backup is 32,015,679,488 bytes long (29.817GiB)

F/W Rev.=1.20  Serial No.=20043514 [ 32 30 30 34 33 35 31 34 ]

[wimb]

What Windows version are you using ?
Where can I download Drive Info ?

Normally 2nd partition of USB-stick is invisible in Windows.
You can use USB_Part_Flip.exe to make 2nd instead of 1st partition visible
and then apply Format to format the second partition.

TinyHexer mpth_small.exe can be used from MULTIPStartProgs to view the MBR
Strange that the 2nd partion is not there in your case.
Here with 16 GB Corsair and using XP OS then everything is OK

[steve6375]

The 2nd partition made by RMPrepUSB is a dummy partition. It only exists to fool any BIOS into treating the USB Flash drive as a hard disk rather than a ZIP drive (which by definition only has one partition). Don't worry that the 2nd partition has 'bad' parameters, no filesystem exists on the 2nd partition and it is not used.

[Doodoo]

What Windows version are you using ?
Where can I download Drive Info ?

I'm using a plain and simple XP. "DriveInfo" is just one button in the RMPrepUsb GUI.

Strange that the 2nd partion is not there in your case.

In my experiment, if I start with two partitions (before formatting with USB_BOOT), I end up with the same two (size unchanged).
If I start with one partition, I still end-up with one. I'll try again to confirm !

[Doodoo]

Don't worry that the 2nd partition has 'bad' parameters, no filesystem exists on the 2nd partition and it is not used.

I understand that the main purpose of the 2nd partition is to fool the BIOS.... but still.... having an unvalid partition table does not feel right to me (well perhaps the partition table can be considered valid, but it's not consistent with the disk geometry)

As soon as you start using other tools on a stick prepared in such a way (CloneDisk, BootIce, etc, etc...), you never know how they are going to behave, and you can expect all sorts of wierd and wonderful things. My impressoin is that it wouldn't hurt to make the partition tiny (but still within the disk limits) so that other tools don't complain !

[steve6375]

If you don't want the 2nd partition, then don't tick the Boot as HDD (2 PTNS) tick box. ( didn't want to waste disk space and to be 'legal' I would have to waste 1 cylinder's worth of space).

[Doodoo]

didn't want to waste disk space and to be 'legal' I would have to waste 1 cylinder's worth of space.

I still think it would be nice, perhaps as an extra option, to accept that 1 cylinder will be wasted to make the 2nd partition entirely legal

[Doodoo]

Any feedback on BootIce (2012-05-06) ?
Does "Parts Manage", "Reformat USB Stick" work for anyone ?

[Doodoo]

Don't worry that the 2nd partition has 'bad' parameters, no filesystem exists on the 2nd partition and it is not used.

By the way, even the first partition may (I think) cause problems, as it is not aligned on a cylinder boundary:

P1   Start=2,048 (1,048,576 bytes) End=62,530,623 (32,015,678,976 bytes)

(62,530,623 + 1) is not divisible by 63, nor 255, let alone 63*255....

[Doodoo]

In my experiment, if I start with two partitions (before formatting with USB_BOOT), I end up with the same two (size unchanged).
If I start with one partition, I still end-up with one. I'll try again to confirm !

My bad, wimb ! Your tool seems to be working fine for me now...
I'm only human and perhaps I made a mistake.... or something wierd happened because I used BootIce / RMPrepUSB in the first place, which did something unusual that your tool didn't like...

Anyway, here's whet i'm getting now after using USB_BOOT:

Disk Signature bytes (1B8h-1BBh) = 1C 96 1C 96



Partition 1   SIZE=30521.9MiB   Type: 07 NTFS  *ACTIVE*

START POS   = CYL:0 HD:1 SEC:1  END POS = CYL:1023 HD:254 SEC:63

START (LBA) = 63 (0000003F)	  SIZE (LBA) = 62,508,852 (03B9CF34)



Partition 2   SIZE=7.844MiB   Type: 0C FAT32LBA (2047GB max)	  

START POS   = CYL:1023 HD:0 SEC:1  END POS = CYL:1023 HD:254 SEC:63

START (LBA) = 62,508,915 (03B9CF73)	  SIZE (LBA) = 16,065 (00003EC1)



Partition 3   SIZE=0MiB   Type: 00		

START POS   = CYL:0 HD:0 SEC:0  END POS = CYL:0 HD:0 SEC:0

START (LBA) = 0 (00000000)	  SIZE (LBA) = 0 (00000000)



Partition 4   SIZE=0MiB   Type: 00		

START POS   = CYL:0 HD:0 SEC:0  END POS = CYL:0 HD:0 SEC:0

START (LBA) = 0 (00000000)	  SIZE (LBA) = 0 (00000000)



P1   Start=63 (32,256 bytes) End=62,508,914 (32,004,563,968 bytes)

P2   Start=62,508,915 (32,004,564,480 bytes) End=62,524,979 (32,012,789,248 bytes)





Drive 1 SanDisk Ultra Backup is 32,015,679,488 bytes long (29.817GiB)

F/W Rev.=1.20  Serial No.=20043514 [ 32 30 30 34 33 35 31 34 ]

Also, both partitions are nicely aligned on a cylinder boundary
(62,508,914 + 1) / 63/255 = 3891
(62,524,979 + 1) / 63/255 = 3892

[wimb]

Also, both partitions are nicely aligned on a cylinder boundary
(62,508,914 + 1) / 63/255 = 3891
(62,524,979 + 1) / 63/255 = 3892

Thanks for testing again and showing good result
with two partitions ending on cylinder boundary.

[Doodoo]

I'm only human and perhaps I made a mistake....

I now realise that what I probably did, is to press the "GO" button, which installs the Grub4Dos MBR, and various other things... but does not reformat the drive

[steve6375]

If the partition is bigger than 1024*255*63 then the max CHS parameters are used in the partition table. END POS = CYL:1023 HD:254 SEC:63
Old BIOSes and old OSes only need the CHS partition table parameters to end at the end of a whole cylinder, AFAIK no BIOS or OS has any requirement for the newer LBA partition table entries of Start LBA and Size to end on a whole cylinder.
Therefore, as long as the partition size is larger than 1024*255*63, RMPrepUSB does not create partitions that end at the end of a whole cylinder as there is no need for this.
If however the partition is smaller than 8GB then the end position will be the end of a whole cylinder.

[wimb]

I now realise that what I probably did, is to press the "GO" button, which installs the Grub4Dos MBR, and various other things... but does not reformat the drive

Yes, pressing GO instead of Format Stick button will be the explanation.

I saw that the table that you initially published for BOOT_USB has

Disk Signature bytes (1B8h-1BBh) = CA 5C CF 6E

which is equal as in the tables for Bootice and RMPrepUSB

That could NOT be the result of my Format Stick procedure,
since I generate symmetrical DiskSignature as in the second test of BOOT_USB

Disk Signature bytes (1B8h-1BBh) = 1C 96 1C 96

In my case the MBR is first constructed with MBR BootCode + Signature bytes + Partition Table
and then written to the Stick.

[Doodoo]

Old BIOSes and old OSes only need the CHS partition table parameters to end at the end of a whole cylinder, AFAIK no BIOS or OS has any requirement for the newer LBA partition table entries of Start LBA and Size to end on a whole cylinder.

Fair enough, I'm not an expert at these things.... I may have focused on something that does not really matter !

However, the second, dummy, partition created by RMPrepUSB starts and ends on 1023/254/63 which is kind of the "new" convention to say, don't look for CHS but use LBA instead. To maximise compatibility with "old" bioses, wouldn't it be better to make it start on 1023/0/1 and end on 1023/254/63, which is kind of the "old" convention ?

[steve6375]

To maximise compatibility with "old" bioses, wouldn't it be better to make it start on 1023/0/1 and end on 1023/254/63, which is kind of the "old" convention ?

I didn't know that 1023/0/1 is a convention for use as a start position when the partition starts beyond 8GB? How can that be correct - do you have any reference for that? A disk could have three very small partitions starting near the end of an 8GB disk and they could all start at 1023/0/1 ?? How can that be correct?
AFAIK the convention is to use 1023/254/63 if the sector location is beyond 8GB???

[Doodoo]

Wimb, the tooltip for the checkbox "Update Grub4dos grldr" mentions a "RUN" script... What is this script for ? Do you have a link discussing this ?

[Doodoo]

do you have any reference for that?

Hmmm I can't find anything clear reference so far, but I'll keep looking....
The best I have found so far is an example here: http://www.terabyteu...icle.php?id=288
You can see that the second partition (extended, #1) and third partition (primary, #2) both start on 1023/0/1 and end on 1023/254/63, but have different LBA start and end addresses

[Doodoo]

Hmmm I can't find anything clear reference so far, but I'll keep looking....

There is this one: http://www.win.tue.n...on_types-2.html

Then there is the problem of what to write in (c,h,s) if the numbers do not fit. The main strategies seem to be
(...)
1d. Write (1023,0,1) for the begin CHS of a partition that starts at or past cylinder 1024, and write (1023,255,63) for the end.
1e. Write (1023,0,1) for the begin CHS of a partition that starts at or past cylinder 1024, and write (1023,254,63) for the end.

this one: http://paulski.com/zpages.php?id=2103

Idenifying sectors on a hard drive is done in legacy fashion by CHS (Cylinder/Head/Sector) values and also by LBA (Logical Block Address) values. The last two boxes in each row are the LBA values and on modern systems these are basically the only values that need concern us. In Fig 4 it will be seen that the CHS values in most of the records have the same values; start values of 1023/0/1 and end values of 1023/254/63. These are actually meaningless values but need to be there for legacy reasons even though legacy software wont be able to make use of them

and eventually this one: http://www.msfn.org/...49#entry1001849
(but it's a personal preference whether you regard this as a trusted source or not !)

You have two partitions using for start CHS address 1023/0/1 and ending on 1023/254/63 (which is one of the two conventions to say "don't look for CHS" - the "old" one) and one that use for start CHS address 1023/254/63 and ending on 1023/254/63 (which is the other convention to say "don't look for CHS" -the "new" one).

[wimb]

Wimb, the tooltip for the checkbox "Update Grub4dos grldr" mentions a "RUN" script... What is this script for ? Do you have a link discussing this ?

RUN of sratlf to be used in Grub4dos - http://bbs.wuyou.com...a=page=1&page=1

http://www.msfn.org/...post__p__988191

RUN is to be used in Grub4dos command environment and is a nice enhancement for Grub4dos

[Doodoo]

RUN is to be used in Grub4dos command environment and is a nice enhancement for Grub4dos

This is probably a subject matter of its own for a separate thread.... but personally I can't make any sense of the links you posted... It's just greek to me.
All I manage to understand is that I should type "RUN" at the Grub4Dos command prompt, that's about as far as I go....
If anyone feel they can write a good summary.... please be my guest ! I'm not asking to be spoonfed, but really, this is more than I can digest !