[jaclaz]

DiracDeBroglie, on 10 July 2012 - 09:22 AM, said:

Well, if you can interpred as a human being directly HEX code, great for you! But that is not the case with me, I have to stick to something that readable by average human beings.
j

NO, I cannot , (or not so easily , compare with http://reboot.pro/2246/#entry14607 )

but Daniel B. Sedory (aka the Starman) seemingly can and he is so nice as to publish his results.

jaclaz

[DiracDeBroglie]

Tripredacus, on 10 July 2012 - 08:02 AM, said:

Well since your system isn't untouched, the fact your settings are incorrect is because of one of these two issues:

1. You fiddled with the BCD.
2. The OEM uses some non-standard way to run recovery.
3. There is always something else possible.

You can see all of Reagentc switches with /?

While it is normal to see the winre.wim in C:\Windows\System32\Recovery, the one to actually use in booting is typically in the recovery partition. So I'd guess that your manufacturer is either using a 3rd party tool, or some in-house method of handling recovery, rather than using the official Microsoft guidance. For example, even your reagentc /info data is lacking...

Windows RE staged should be 1 not 0
Setup enabled should be 1 not 0
WinRE.WIM directory should have a path
Setup Files should have a path

That path is the same as the one in Recovery Environment.

But since you are back to working, I wouldn't worry about going about and fixing (breaking) it now. It may be by design as not everyone does it the same way.

Thanks for the info. I'll probably install soon a clean install of Win7HP which I will download from one the MS associate servers. I got an ASUS notebook K93SV and the way ASUS works is very sloppy; lots of bloatware and other rubbish. I'll get rid of it by a clean install. To boot from the Recovery/Re-Install partition I need to tap F9 at boot time.

BTW, the info which can be seen on the screen in Avanced Boot Options (ABO), is that also located in the BCD store?

Do I interpret it correct that first Windows Boot Manager loads its boot options (Win7, WinRE, Windows Setup, ...WinXP, ...Memtest, ...) and then the ABO? Or is it different somehow?

Thanks,
j

[Tripredacus]

DiracDeBroglie, on 10 July 2012 - 09:39 AM, said:

To boot from the Recovery/Re-Install partition I need to tap F9 at boot time.

BTW, the info which can be seen on the screen in Avanced Boot Options (ABO), is that also located in the BCD store?

Do I interpret it correct that first Windows Boot Manager loads its boot options (Win7, WinRE, Windows Setup, ...WinXP, ...Memtest, ...) and then the ABO? Or is it different somehow?

Yeah the F9... this means they are using something "custom" likely because they entered into agreements with their software provider and didn't ditch them when the easy option came out with Vista or 7. This is why HP still used SoftThinks recovery with Vista, even though MS gave OEMs the tools to do it themselves with minimal development time. Windows 7 was even easier (aka no Visual Studio required). So with the F9, that is in the MBR which can be intercepted ahead of time before booting on your normally active (OS) partition. The basic function of that F9 is to take the OS partition, un-active it, and activate the recovery partition, then allow the PC to boot "normally". Of course then your recovery partition reverts these settings, which allows the next reboot to boot to the OS again.

As far as the Boot Menu in Windows, I am not entirely sure where that is stored. I know that it *can* be configured (adding options) since the Vista recovery method required a script to be used to add that "Recovery" option in the menu. However, Windows 7 natively has an ability to know to display that option if a valid recovery partition is present and/or certain settings are set in BCD.

I think (no clue really) that Bootmgr works likely similar to the F9 hand-off idea I posted before. Without any interaction, its going to run winload.exe (boot Windows), but if you interrupt that by using the F8 key, it boots to the menu. I don't believe that any of those options are preloaded, but when one is executed it knows what to do. I would presume a pre-load would delay boot time, and besides its not actually possible to execute more than one of those options at once.

[DiracDeBroglie]

Tripredacus, on 10 July 2012 - 10:56 AM, said:

I think (no clue really) that Bootmgr works likely similar to the F9 hand-off idea I posted before. Without any interaction, its going to run winload.exe (boot Windows), but if you interrupt that by using the F8 key, it boots to the menu. I don't believe that any of those options are preloaded, but when one is executed it knows what to do. I would presume a pre-load would delay boot time, and besides its not actually possible to execute more than one of those options at once.

Just did a test with an EasyBCD created bootable partition, with 2 options in the Window Boot Manager's (WBM) menu: WinRE (WinRE.WIM) and Windows Setup (which basically calls a BOOT.WIM with a WinPE in it). When the menu pops up for about 10 seconds or so, I tap F8, which makes the WinRE start loading first, and that takes some time because that is a WIM image. After the loading is finished the ABO (Save Mode, Save Mode with Networking, Save Mode with Command Prompt, ...) pops up. After proceeding with the ABO WinRE boots.

If I, on the other hand, change the boot order, that is, Windows Setup (default) and WinRE, I have the same procudere about the ABO menu after which I can boot into the Recovery partition's BOOT.WIM.

It is not quit clear where exactly the ABO menu is located (bootmgr.exe or bcd), but it is sure that the appearance on the screen can only happen if WinLoad.exe has been executed and/or the Win Kernel loaded.

j

[DiracDeBroglie]

jaclaz, on 09 July 2012 - 04:31 AM, said:

The HP hardware normally comes with a "specific", "proprietary" MBR CODE, that allows to press F11 to get to the Recovery partition.
If you change the MBR code (as an example by installing *any* MBR based bootmanager and or simply run a MBRFIX or on Windows 7 a bootsect with the /mbr switch) that original MBR code is gone for good and restoring it (unless you have a copy/backup of it) is/will be - to say the least a nightmare.
jaclaz

Are there any bootmangers, bootloader, bootselectors, (like Grub, LILO, ... ) that can be put in the MBR and that become active only when tapping a particular F-key at boot time, after which the MBR bootselector performs a particular task like making a hidden recovery partition bootable, then effectively boot into it and after that turn the ID of all the altered partitions back to what is was before?? Or is this kind of property only for proprietary bootloaders of the OEMs (Asus, Dell, HP, ..)?

j

This post has been edited by DiracDeBroglie: 10 July 2012 - 02:01 PM

[Tripredacus]

You can do your own boot options using Terabyte Unlimited's MBR program. You can see some example here where I created my own recovery partition.
http://www.msfn.org/...very-partition/

[DiracDeBroglie]

Tripredacus, on 10 July 2012 - 03:03 PM, said:

You can do your own boot options using Terabyte Unlimited's MBR program. You can see some example here where I created my own recovery partition.
http://www.msfn.org/...very-partition/

There are several MBR programs of TU, could you provide me with the exact link?
Thanks, j

[jaclaz]

DiracDeBroglie, on 11 July 2012 - 02:27 AM, said:

There are several MBR programs of TU, could you provide me with the exact link?
Thanks, j

This is the one you want MBR Utility, file mbr.zip:
http://www.msfn.org/...very-partition/
http://www.msfn.org/...post__p__769544

OR, forget about it and get OS-BS rectius mbldr instead:
http://reboot.pro/334/
http://www.msfn.org/...d/page__st__105

BTW this info is in the thread you were ALREADY pointed to:
http://www.msfn.org/...d-not-be-found/
(there are reasons why a link is given )

jaclaz

[DiracDeBroglie]

jaclaz, on 11 July 2012 - 03:05 AM, said:

OR, forget about it and get OS-BS rectius mbldr instead:
http://reboot.pro/334/
http://www.msfn.org/...d/page__st__105
jaclaz

I checked out MBLdr and that could be a good alternative to replace proprietary boot selectors, only kicking into action by tapping a particular Function key.

The link you gave me was in the context of someone who could not see the Recovery partition anymore. In emergency cases one can apply an even simplier solution for booting off the Recovery/Re-Install partition without having to resort to any MBR fiddling in case the MBR bootselector is lost. Burn a GPartEd LiveCD, boot from that, make the Recovery partition unhidden by unticking the checkbox hidden, and tick the checkbox Boot. Restarting the computer boots right away into the Recovery partition.

Even more simple, run the repair disc, CD/DVD, (and this any normal human being with a normal iq should have), go into the cmd console, then:
X:\diskpart
diskpart>list disk ; select disk 0 (if your recov par is on disk 0) ; list partition ; select partition [number of recov par] ; detail part ; set id=0C override (in case it was 1C, hidden fat32x) ; detail part (again) ; Active ; detail part (again)
Now the Recovery partition is type 0C, and active (will boot).
Quit the repair disc, restart computer and it'll boot into the Recov part.

PowerQuest's Partition Table Editor can also change the hidden fat32x (1C) ID to unhidden (0C). Reboot, mark Recovery partition as active in Disk Management and reboot again. Computer boots right away into the Recovery partition.

Booting into a Recovery/Re-install partition is *never* a problem, at least if you haven't been fiddling with the bootsector of it (as in my case).

I still need more time to go into all this (and into so many other posts and links) but I am off for 2 weeks not having access to any test machine. I can still read the forums now and then, though.

CU later folks, j

[jaclaz]

DiracDeBroglie, on 11 July 2012 - 11:34 AM, said:

I checked out MBLdr and that could be a good alternative to replace proprietary boot selectors, only kicking into action by tapping a particular Function key.
....
PowerQuest's Partition Table Editor can also change the hidden fat32x (1C) ID to unhidden (0C). Reboot, mark Recovery partition as active in Disk Management and reboot again. Computer boots right away into the Recovery partition.

Well, I got you exactly where I wanted you to be , also grub4dos has this feature (and many others), see here (be warned it is advanced-advanced ):
http://reboot.pro/7138/
(and take into account that since the above thread, quite a few new tricks can be done with latish releases of grub4dos )

DiracDeBroglie, on 11 July 2012 - 11:34 AM, said:

I still need more time to go into all this (and into so many other posts and links) but I am off for 2 weeks not having access to any test machine. I can still read the forums now and then, though.

See you when you are back .

jaclaz

[pointertovoid]

Please forgive the elementary question but...
I see clearly the advantages of Ntfs over Fat32 and imagine FatX as a Fat64.
What improvement could bring FatX over Ntfs?

[jaclaz]

pointertovoid, on 20 July 2012 - 03:20 AM, said:

Please forgive the elementary question but...
I see clearly the advantages of Ntfs over Fat32 and imagine FatX as a Fat64.
What improvement could bring FatX over Ntfs?

I don't see any senceful question, elementary or not.
What the heck is "FatX"?

jaclaz

[Tripredacus]

FAT32X? Wikipedia is very elementary about it.

Quote

The latter type (partition 0x0C) is also named FAT32X in order to indicate usage of LBA disk access instead of CHS. On such partitions, some CHS-related geometry entries in the EBPB record, namely the number of sectors per track and the number of heads, may contain no or misleading values and should not be used.

[jaclaz]

Tripredacus, on 20 July 2012 - 08:52 AM, said:

FAT32X? Wikipedia is very elementary about it.

Quote

The latter type (partition 0x0C) is also named FAT32X in order to indicate usage of LBA disk access instead of CHS. On such partitions, some CHS-related geometry entries in the EBPB record, namely the number of sectors per track and the number of heads, may contain no or misleading values and should not be used.

NOT only elementary, also much VAGUE if not plainly WRONG.

There is NO DIFFERENCE whatsoever in the BPB between a bootsector created on the SAME device by the SAME OS by any "proper" tool, on partitions marked with ID 0B or 0C,
http://www.win.tue.n...on_types-1.html
this is still connected to the PLAINLY WRONG idea that a Partition ID is anything more than a Partition ID, see:
http://homepage.ntlw...ystem-type.html
and - if we are talking of actually booting from that partition - the H and S correct values are VITAL on Windows NT systems, because of some "botched" or "relic of good ol' CHS days" in the actual bootsector CODE (and this applies to NTFS too).
See:
http://www.911cd.net...ic=21702&st=129

The whole idea of calling a FAT32 "FAT32X" is "botched", there is only a FAT32 filesystem, that can be accessed through BOTH CHS and LBA if below the CHS limit or only through LBA if over it (both in size and position on disk).

But the question by pointertovoid was about "FatX", which is a particular filesystem for the Xbox:
http://en.wikipedia....tion_Table#FATX
that has clearly nothing to do with FAT32 (and with this thread) and FAT64 does not simply exist, it is the WRONG name given sometimes to exFAT:
http://en.wikipedia.org/wiki/ExFAT

jaclaz

[pointertovoid]

My fault, I had mistaken it for exfat, sorry.

[jaclaz]

pointertovoid, on 22 July 2012 - 03:32 AM, said:

My fault, I had mistaken it for exfat, sorry.

Then the advantage could be that exFAT is seemingly faster than NTFS:
http://reboot.pro/17255/
http://thessdreview....ftware/1834.htm

Though of course there is nothing in the above except for the sentence:

Quote

(exFAT in my benchmarks has produced the fastest results)

and as we discussed here, the reliability of benchmarks and their actual being representative of "real life usage" has to be evaluated carefully.
It would "make sense", in the sense that theoretically a FAT32 should be faster than NTFS (it needs to read less data to get to the file you want at least until the number of files grows to very high numbers), but, as seen here:
http://www.msfn.org/...d-on-usb-stick/
at least in some applications it has been somehow "dumbed down" by the good guys at MS.
See also the very detailed:
http://www.uwe-siebe...usbstick_e.html
exFAT, being more similar to FAT32 and being also been "pushed" by the same MS guys, could have been optimized, see also:
http://www.ntfs.com/ntfs_vs_fat.htm
It is only slightly faster on USB sticks along results of this test (also already mentioned in the given thread), but only in a few of the many benchmarks:
http://blog.testfrea...-ntfs-vs-exfat/

jaclaz

This post has been edited by jaclaz: 22 July 2012 - 04:37 AM

[DiracDeBroglie]

jaclaz, on 09 July 2012 - 12:08 PM, said:

Since you are using tinyhexer, you may find my little (as always half-@§§ed ) scripts for it, handy.
Here:
http://reboot.pro/8734/
jaclaz

I'm using your 3 scripts, and indeed those are extremely handy. Great added value to Tiny Hexer, those scripts have been very helpful to analyse USB Flash Drives where the MBR is sometimes missing. There is one disadvantage though with Tiny Hexer, there is no userfriendly editor to edit the partition table or any figures like 'sectors before' the PBR. I still use the old PowerQuest's Partition Table Editor (PTEDIT32.EXE), which works just fine. I'm not sure but I think you probably use also another utility for fixing MBRs and PBRs; if I recall correctly it start with HD..... , or something.

johan

[DiracDeBroglie]

jaclaz, on 10 July 2012 - 05:36 AM, said:

You are falling in the "Access is denied" issue with Windows 7, right?
It is one of the "nice" protection features.
See:
http://reboot.pro/12413/
jaclaz

About the error message "System Error." "Code: 5." "Access is denied." when using Tiny Hexer, I tested LockDismount v0.3.0.0 (LD) and it works, at least on non-system drives (as mentioned in the link you gave) like USB Flash Drives. A drawback of LD is that it locks the entire HDD while in fact I only wanted to edit with Tiny Hexer a few sectors of one particular volume or partition on the system HDD. In my case I wanted to test Tiny Hexer by editing just a few empty dummy and meaningless sectors in the recovery partition (which sits on the system HDD). Upon trying to save away the altered test sector in Tiny Hexer I got then the error message. Using LD, LD tried to lock ALL the volumes on my system HDD and showed FAILED for all of the locking attempts -- a pity LD does not work solely on just one particular volume/partition, in stead of on entire HDDs.

I went on to create a few virtual HDDs, or vdisks, in Disk Management and dropped the *.vhd files in a primary partition on the system HDD. To my surprise Tiny Hexer could save away the edited sectors in those vdisks without causing the "Access is denied" error; so no need at all to use LD. An extra check after reboot of the computer revealed the sectors where indeed altered/updated. Although the *.vhd files of the vdisks were located on the system HDD, Tiny Hexer had somehow "direct" disk access to the vdisks.

Annoying thing with my vdisks though is that I have to re-attach them after each reboot.

BTW, I did not get any "Access is denied." error from the old PowerQuest's Partition Table Editor (PTEDIT32.EXE) when saving away an edited PBR on one of the partitions on the system HDD. It looks like Win7 gives direct disk access to some old software, but to some other old software (like Tiny Hexer) it doesn't.

j

[DiracDeBroglie]

jaclaz, on 20 July 2012 - 09:46 AM, said:

There is NO DIFFERENCE whatsoever in the BPB between a bootsector created on the SAME device by the SAME OS by any "proper" tool, on partitions marked with ID 0B or 0C,
http://www.win.tue.n...on_types-1.html
this is still connected to the PLAINLY WRONG idea that a Partition ID is anything more than a Partition ID, see:
http://homepage.ntlw...ystem-type.html
and - if we are talking of actually booting from that partition - the H and S correct values are VITAL on Windows NT systems, because of some "botched" or "relic of good ol' CHS days" in the actual bootsector CODE (and this applies to NTFS too).
See: http://www.911cd.net...ic=21702&st=129
jaclaz

Using DISKPART, I changed the type of my recovery partition from 0C (FAT32X) to 0B (FAT32), made the partition then active and restarted my computer. The computer booted immediately into my recovery partition without any problem. It looks like 0C and 0B are indeed the same.

j

[jaclaz]

DiracDeBroglie, on 06 August 2012 - 12:34 PM, said:

I'm not sure but I think you probably use also another utility for fixing MBRs and PBRs; if I recall correctly it start with HD..... , or something.

Naah, I suggest often Hdhacker, but just as a tool to backup/restore the sectors.
http://dimio.altervista.org/eng/
Normally I use the dsfok tollkit, (command line) to do the same:

and I use my own little spreadsheet to calculate values I later carve by hand with TinyHexer, see:
http://www.msfn.org/...ost__p__1005415

jaclaz