107 replies to this topic

[popov]

@Wonko

 

I have removed laptop battery and running just off charger now, so whenever I need to reboot I simply unplug power cable. When testing last thing I had like 6 blackouts in "r-" mode, then I plugged "rw" and it worked, so again I did "r-" and it blacked out, finally after that "r-" worked. Those blackouts are happening only on "r-". Also if I remeber correctly, besides blackouts this laptop freezes at manual boot device selection when "r-" is connected.

 

 

What about MBR2P63S ???

 

 

Results identical to MBR1P63S on all devices.

 

 

Now results of MBR1P32S

 

DV2718US(both in r- and rw mode):

 

80

0000 FE02 0020 0343

LHDD0 00

SHDD0 00

 

P5GC-MX(rw):

 

80

0000 7F02 003F 0373

LHDD0 00

SHDD0 00

 

P5GC-MX(r-):

 

 

00

0000 FF01 0000 03FF

L1

SFDD0 00

 

A7N8X(rw):

 

80

0000 7F03 003F 036F

LHDD0 00

SHDD0 00

 

A7N8X(r-):

 

skips

 

 

Results of MBR2P32S were exactly sme as MBR1P32S.

Edited by popov, 01 April 2013 - 03:42 PM.

[steve6375]

re. 'blackouts' - I get this often on laptops when booting some linux distros. After a shutdown/hang/halt, I can switch them off and on again (hold down power button) and there is no POST/display. To fix I have to remove battery and unplug AC, then re-connect.

 

It seems your BIOS is determined to treat the USB as a floppy in r-mode - but why? Even a floppy is rw?

 

The A7N8X also is treating the USB as an FDD - so the USB drive is responding differently in r mode which both system BIOSes are detecting (but not QEMU or the other system).

 

Can you boot to grub4dos in r mode on the P5GC ?

[popov]

MBR0P.BIN results

 

DV2718US(both in r- and rw mode):

 

 

80

0000 FE02 0020 0343

LHDD0 00

SHDD0 00

 

P5GC-MX(rw):

 

80

0000 7F02 003F 0373

LHDD0 00

SHDD0 00

 

P5GC-MX(r-):

 

00

0000 FF01 0000 03FF

L1

SFDD0 00

 

A7N8X(rw):

 

80

0000 7F03 003F 036F

LHDD0 00

SHDD0 00

 

A7N8X(r-):

 

skips

 

 

 

re. 'blackouts' - I get this often on laptops when booting some linux distros. After a shutdown/hang/halt, I can switch them off and on again (hold down power button) and there is no POST/display. To fix I have to remove battery and unplug AC, then re-connect.

 

 

During test it blacked out only once when drive was in "rw" mode, but were very frequent for "r-". But of course you are correct, because I had other flash drives and it sometimes randomly blacks out on them too. Still it doesn't explain that I had like 10 in a row while testing one of those files.

 

 

It seems your BIOS is determined to treat the USB as a floppy in r-mode - but why? Even a floppy is rw?

 

The A7N8X also is treating the USB as an FDD - so the USB drive is responding differently in r mode which both system BIOSes are detecting (but not QEMU or the other system).

 

Can you boot to grub4dos in r mode on the P5GC ?

 

Afaik A7N8X had problems from the very first launch, it didn't support USB booting for quite time until ASUS finally made some driver update. It could be some defect with their mobos, not sure, but the way they fixed it was making it emulate USB-HDD with USB-FDD, so in fact it was using USB-FDD driver to read USB-HDD disks. If it's true, then probably it's hardcoded in BIOS and nothing can be done about it. Unconfimed.

 

P5GC-MX is cheap limited mobo that also does not offer anything near real experience. I never saw poorer BIOS before, literary there's nothing inside. There is no way of choosing USB-HDD/FDD/ZIP/etc simply, because there is just one single USB option.

 

Sadly I don't have access to P5GC-MX right now, but I can test it on MSI 870A-G54, which full recent mobo with latest drivers. Just don't have much time atm. Cheers.

Edited by popov, 01 April 2013 - 05:00 PM.

[popov]

MSI 870A-G54 tests complete

 

MBR0P.BIN, MBR1P32S, MBR2P32S

 

"rw":

80

0000 7F02 003F 0373

LHDD00 00

SHDD00 00

 

"r-":

00

0000 7F01 003F 0373

L1

SHDD00 00

 

MBR1P63S.BIN, MBR2P63S

 

"rw":

80

0000 FE02 003F 03BD

LHDD00 00

SHDD00 00

 

"r-":

00

0000 FE01 003F 03BD

L1

SHDD00 00

 

USB options in BIOS for MSI mobo were actually very limited.

Edited by popov, 01 April 2013 - 06:35 PM.

[steve6375]

At least the geometry is right even when it treats it as a FDD. It does look like the USB drive is reporting itself as an FDD.

Presumably in Windows, it appears as an HDD when in r- mode (after disconnecting and reconnecting)? USBDeview. Dev Man, etc?

[popov]

USB Device Tree Viewer produces same output for both "r-" and "rw", the only minor idfference is "Kernel Name", but that is dynamically assigned each time I reconnect drive.

 

    =========================== USB Port3 ===========================

Connection Status        : Device is connected
Port Chain               : 3-3
PortAttributes           : 0x00000002 (Shared USB2)

      ======================== USB Device ========================

        +++++++++++++++++ Device Information ++++++++++++++++++
Device Description       : USB Mass Storage Device
Device ID                : USB\VID_3538&PID_0901\08AD330000005236
Driver KeyName           : {36FC9E60-C465-11CF-8056-444553540000}\0009
Driver                   : D:\WINDOWS\system32\DRIVERS\USBSTOR.SYS (Version: 5.1.2600.5512  Date: 2008-04-14)
Legacy BusType           : PNPBus
Class                    : USB
Service                  : USBSTOR
Enumerator               : USB
Location Info            : USB Flash Drive
Manufacturer Info        : Compatible USB storage device
Capabilities             : Removable, UniqueID
Address                  : 3
Problem Code             : 0
Power State              : D0 (supported: D0, D3, wake from D0, wake from D3)
Child Device 1           : Disk drive
 Device ID               : USBSTOR\DISK&VEN_GENERIC&PROD_USB_FLASH_DISK&REV_1100\08AD330000005236&0
 Class                   : DiskDrive
  Volume                 : \\?\Volume{ec879900-8e7a-11e2-8dfd-00dc32be41ab}\
  Kernel Name            : \Device\Harddisk1\DP(1)0-0+14
  Mountpoint             : H:\

        ---------------- Connection Information ---------------
Connection Index         : 0x03
Connection Status        : 0x01 (DeviceConnected)
Current Config Value     : 0x01
Device Address           : 0x01
Is Hub                   : 0x00 (no)
Number Of Open Pipes     : 0x02 (2)
Device Bus Speed         : 0x02 (High-Speed)
Pipe0ScheduleOffset      : 0x00 (0)
Pipe1ScheduleOffset      : 0x00 (0)

        ------------- Device Qualifier Descriptor -------------
bLength                  : 0x12 (18 bytes)
bDescriptorType          : 0x01 (Device Qualifier Descriptor)
bcdUSB                   : 0x200 (USB Version 2.00)
bDeviceClass             : 0x00 (defined by the interface descriptors)
bDeviceSubClass          : 0x00
bDeviceProtocol          : 0x00
bMaxPacketSize0          : 0x40 (64 bytes)
idVendor                 : 0x3538
idProduct                : 0x0901
bcdDevice                : 0x1100
iManufacturer            : 0x01
 Language 0x0409         : "PQI"
iProduct                 : 0x02
 Language 0x0409         : "USB Flash Drive"
iSerialNumber            : 0x03
 Language 0x0409         : "08AD330000005236"
bNumConfigurations       : 0x01

        -------------- Configuration Descriptor ---------------
bLength                  : 0x09 (9 bytes)
bDescriptorType          : 0x02 (Configuration Descriptor)
wTotalLength             : 0x0020 (32 bytes)
bNumInterfaces           : 0x01
bConfigurationValue      : 0x01
iConfiguration           : 0x00
bmAttributes             : 0x80 (Bus Powered)
MaxPower                 : 0xFA (500 mA)

        ---------------- Interface Descriptor -----------------
bLength                  : 0x09 (9 bytes)
bDescriptorType          : 0x04 (Interface Descriptor)
bInterfaceNumber         : 0x00
bAlternateSetting        : 0x00
bNumEndpoints            : 0x02
bInterfaceClass          : 0x08 (Mass Storage)
bInterfaceSubClass       : 0x06 (SCSI transparent command set)
bInterfaceProtocol       : 0x50 (Bulk­Only Transport)
iInterface               : 0x00

        ----------------- Endpoint Descriptor -----------------
bLength                  : 0x07 (7 bytes)
bDescriptorType          : 0x05 (Endpoint Descriptor)
bEndpointAddress         : 0x81 (Direction=IN  EndpointID=1)
bmAttributes             : 0x02 (TransferType=Bulk)
wMaxPacketSize           : 0x200 (max 512 bytes)
bInterval                : 0xFF (at most 1 NAK each 255 microframes)

        ----------------- Endpoint Descriptor -----------------
bLength                  : 0x07 (7 bytes)
bDescriptorType          : 0x05 (Endpoint Descriptor)
bEndpointAddress         : 0x02 (Direction=OUT  EndpointID=2)
bmAttributes             : 0x02 (TransferType=Bulk)
wMaxPacketSize           : 0x200 (max 512 bytes)
bInterval                : 0xFF (at most 1 NAK each 255 microframes)
 

[popov]

I have noticed one thing, not that it matters but on some systems(didn't try on all) when I insert stick as "rw" and then, after initial detection phase completes(ie. I see first signs of booting), I can move switch back to "r-", and everything works normally just like it would be in "r-". Another thing, is that I've tried to format stick with RMPrepUSB as USB-FDD and then I have set USB-FDD as only USB booting device, and in "rw" everything was working but "r-" was still the same. I guess that I'll try fbinst, just now don't have time for that.

[popov]

I don't see how can I use FBINST by myself, seems that lots of stuff can be broken with it. There was also bootlace option as I described before.

 

Anyway I found out good website with download section, their tools might be worth of checking: http://code.google.c.../downloads/list

 

Also I am wondering on another thing. Maybe I should use something like PLOP on small CD or another USB stick and just chain boot USB driver and then "r-" USB stick...

Edited by popov, 03 April 2013 - 10:23 AM.

[Wonko the Sane]

Unless you choose the "wrong" device, the only device affected (temporarily) will be the USB stick.

The given instructions:

http://reboot.pro/to...ws-7-and-vista/

seem to me clear enough.

 

The site you found is the repository of Ilko_t (for the WinsetupFromUSB with GUI) that has some additional files, let it alone and try the given link above, for the moment.

 

Wonko

[popov]

I did what the guy posted in frist post, but it didn't change anything. Now when I tried to repeat step again, this command:

 

 

fbinst (hd1) format --fat32 --force

 

produced this error:

 

Fbinst: error: can't lock disk

 

Update: resolved error above, it seems that fbinst doesn't always unload itself from memory when finished. Once corresponding process is closed, everything works fine.

 

Formatted with fbinst as zip/chs(just like the guy from #6):

 

http://reboot.pro/to...ws-7-and-vista/

 

...no difference.

 

Is there any way of using plopboot straight from USB?

Edited by popov, 03 April 2013 - 12:12 PM.

[Wonko the Sane]

In case, you may want to try the atttached which should be the "automated" installer by Maanu (for which the actual link is invalid).

It is a "compiled SFX", you can open the thingy in 7-zip and extract/verify the contents.

 

Wonko

Attached Files

•  USBoot-2.zip   240.66KB   324 downloads

[Wonko the Sane]

Formatted with fbinst as zip/chs(just like the guy from #6):

 

http://reboot.pro/to...ws-7-and-vista/

WHY/how?

 

Just simply try doing this:

 

fbinst --list ( use the proper hd number from this command)

fbinst (hd#) format --fat32 --force ENTER
fbinst (hd#) add grldr grldr ENTER
fbinst (hd#) add-menu fb.cfg fb.txt ENTER

 

and nothing else.

 

It will either work or not and if you could provide a descrtiption of WHAT happens both with RW and R, it may help.

 

Wonko

[popov]

It doesn't work, I did that manually in fbinst. In "r-" one mobo skips, second hangs with just flashing underscore(very common for that mobo), while "rw" is always fine. Same results for both #1 and #6(I don't know what's with plop).

[Wonko the Sane]

Well, we have pretty much aended available options.

The only thing remaining to be tried would be the triple MBR thingy, but fbinst should have worked like ti (and even better).

It seems like that stick besides R/RW selection "hides" or "changes" something.

Till now the results seemed to imply that it behaves not completely unlike the switch that some ZIP disk drives had, that simply hid MBR+hidden sectors. but I wonder what is actually hidden by that switch, could it be that it switches NOT "from HD-like to FD-like" but rather "from HD-like to CD-like"?

 

If this is the case, you could try a hybrid boot .iso  

http://reboot.pro/to...os-isohybrided/

 

Wonko

[popov]

I will definitely try this option, but now I have messed one of computers HDD's with some dumb chineese tools and I am trying to recover what was there. It seems that tool messed filesystems on both HDD and USB Flash Stick(wtf?), I have reformatted USB but HDD is real problem, as there were many usable files and well working Windows 7 os. It seems that tool - Lvyanan 1JF9z isn't only boot manager tool, but also hidden(?) partition recovery tool, that somehow got activated on both HDD and Flash.

[MedEvil]

Now that we're through, throwing everything and the kitchen sink at the problem. How about we try some fault isolation?

The test are done on 3 different computers, from 3 different manufacturers, yet yield the same results.
The chance that all 3 BIOS have the exact same problem is highly unlikely.
So the problem has to lie with the USB-Stick.

The raw images created with and without write protection are identical. So The WP switch does not hide anything.

The only things left as problem source are:
- the bootloader / manager has issues with a write protected device (pretty unlikely since it works for others)
- the wp switch switches the device to become something else

Unless someone has a better idea, i would suggest to disable all boot media in BIOS and then always only enable 1 after the other to see what the BIOS 'thinks' of the USB-Stick with and without WP.

[popov]

Well, it might be funny for you but it's not for me. I was using Lvyanan program because it started to post some results, as it menu appeared both in "rw" and "r-" mode. The point is, I didn't even click anything and it somehow jumped on my hard drive, messed it and then messed my flash stick same way. Now I just hope that partition table ain't wiped, and that I will be able to get some stuff back, if not full OS then maybe just some word documents that I had there. Anyway I made new thread for that and even mailed few people who made that prog, but I don't really expect any results.

[MedEvil]

A messed-up partition table is no big thing. there are plenty of programs, which can scan a HDD to find partitions and fix the table.

If that's all, that got broken, you're very lucky, as it is the easiest to fix.

[Wonko the Sane]

The raw images created with and without write protection are identical. So The WP switch does not hide anything.

The only things left as problem source are:
- the bootloader / manager has issues with a write protected device (pretty unlikely since it works for others)
- the wp switch switches the device to become something else

Yep , that's exactly what I posted right before you (though you expressed it much better )

 

Wonko

[popov]

I've tried fresh GRUB on empty stick. A7N8X(the one that never detects stick with its own driver issues - I wrote earlier about this) has 4 options related to USB: USB-HDD, USB-ZIP, USB-FDD and USB-CDROM. The only thing that booted normally was "rw" stick in USB-HDD. Everything else displayed standard "DISK BOOT FAILURE, INSERT SYSTEM DISK AND PRESS ENTER." for both stick modes.

 

Just for fun I ran CrystalDisk Mark on few of my pendrives, and PQI drive with switch had reads: Seq and 512 reads at around 20MB/s, 4K and 4KQD32 at around 5MB/s, but writes were like: Seq - 4MB/s, 512 - 0,67MB/s, 4K - 0.005MB/sec and 4KQD32 0.01MB/s.

 

So I compared to my other 2.0 sticks and those write speeds were horrible. For example for stick that had similar read results: 17MB/s // 5MB/s, writes were 6MB/s // 2MB/s // 0.75MB/s // 0.08MB/s.

 

 

Edited by popov, 05 April 2013 - 02:03 AM.

[popov]

Hey, I am back. I'd like to try a bit different approach, ie. creating boot manager that could run from CD/other USB stick/etc, which would have its own USB drivers, completely independant of ones used by mobos, to see if it boots this way, and what are readings. Just wondering wht options I do have, I know that PLop is good, but are there any other "out of box" solutions, and what are other options?

[popov]

Also, is there any good wiping tool for USB? Just to be sure that tests are performed on clean sticks.

[popov]

So I have tested few things - obviously chain booting via plop worked well.

But after that I've tried few other boot loaders(direct boot) on write protected stick and I had rather interesting results. Wee gave me "Urr! Wee" message, which meant that stick was actually detected and something was loader, just not what I need. Then I tried UltraISO boot manager which has various boot and write options with ability to hide boot paritions and this program actually depenadant on mode can display some errors and shows sings that it starts reading MBR but is unable to read filesystem. At least it looks that way, but this is still progress compared to what I had before. One of specified errors is Q0.

Edited by popov, 10 April 2013 - 02:22 PM.

[MedEvil]

As far as i know, all bootmanagers except Plop use the BIOS and it's drivers to boot the selected OS. Hence they all should fail as well.
STill, if you happen to get out of any, what exactly goes wrong, you might have a chance to fix it after all.

[popov]

UltraISO boot manager is somehow able to read something from MBR even when stick is write protected. From what it shows, it cannot get to filesystem/boot sector, no matter what is there(tried probably all filesystems and common loaders). It also has some option to chose "write method" with few USB-HDD and USB-ZIP possibilities, and dependant on choice, ability to hide boot partition, also with various options. I've tried some of those options without any success, but I didn't try them all. Would any of those matter at all?

 

As for boot loaders, some time ago, when many motherboards didn't have USB/(other) boot option yet, people were building their own loaders to make it possible. I guess that those progs are unavailable by now, but I am pretty sure that people use different techniques to make it work. Wikipedia has small list of boot loaders

 

http://en.wikipedia....of_boot_loaders

 

I might be checking some of them now, if there is any other lists with more of them, then I'll be pleased to get it.

Edited by popov, 11 April 2013 - 05:22 AM.