odd behaviour from an old pc any suggestions?

Hello, I am trying to do a Xp installation from a flashdrive onto an old desktop. The odd thing is: either my usb is not recognised or the text setup ends right before loading the eula conformation.

It seems to have problems with partitions larger then 2GB in size and boots most loaders well from an 2GB fat16 partition.

This is an asus a7v8x-mx motherboard we arte talking about.

Can anyone tell me something about the odd behaviour? Why wont it boot from flashdrives with partitions larger than 2G?

Hello, I am trying to do a Xp installation from a flashdrive onto an old desktop. The odd thing is: either my usb is not recognised or the text setup ends right before loading the eula conformation.

 

It seems to have problems with partitions larger then 2GB in size and boots most loaders well from an 2GB fat16 partition.

 

This is an asus a7v8x-mx motherboard we arte talking about.

 

Can anyone tell me something about the odd behaviour? Why wont it boot from flashdrives with partitions larger than 2G?

What behaviour?  You haven't shared any observations.

There is a diagnostic MBR available that will tell you what geometry BIOS uses for a USB disk, if you can get that far.

You can also try using PLoP.

I thought it was odd behaviour that it only accepts partitions on my flash drive smaller than 2GB. And the textsetup crashing is odd as well, but I am now checking if this could just be caused by my setup files. Right now it crashed in virtualbox, so perhaps my ISO file is no longer working (it did before but was on a hard disk for a while, is is possible part of it became corrupted).

I was wondering if it is a known thing, for motherboards this old to only accept USB drives formtatted in a certain way?
This is not known to me.

...what geometry BIOS uses for a USB disk....

I dont really understand what you are saying here, are you confirming that the BIOS may only accept USB drives formatted in a certain way?

I thought is was a bug or somehting.

Edited by Fiction, 10 January 2013 - 10:48 AM.

I thought it was odd behaviour that it only accepts partitions on my flash drive smaller than 2GB.

"Odd behaviour" and "accepts" aren't observations.  Nobody knows what those mean, except you.

And the textsetup crashing is odd as well, but I am now checking if this could just be caused by my setup files. Right now it crashed in virtualbox, so perhaps my ISO file is no longer working (it did before but was on a hard disk for a while, is is possible part of it became corrupted).

"Crashing" isn't an observation.  Nobody knows what that crash looks like, except you.

I was wondering if it is a known thing, for motherboards this old to only accept USB drives formtatted in a certain way?
This is not known to me.

 

I dont really understand what you are saying here, are you confirming that the BIOS may only accept USB drives formatted in a certain way?

 

I thought is was a bug or somehting.

The BIOS' USB-boot support decides on a cylinders/heads/sectors geometry for the USB storage device.  If this geometry does not result in the ability to address the entire storage device, then the storage device is "truncated."

There is special diagnostic MBR code (courtesy of "TJ", who submitted it to Syslinux over 3 years ago) which will tell you the geometry that BIOS decided on.

In order to use this, you must backup your USB storage device's current MBR to a backup file, then overwrite the storage device's first 440 bytes of the MBR with this diagnostic MBR code, then boot the computer.  Then you will be able to report observations of its output.  Then you can restore the previous MBR from your backup file.

I am sorry, you are completely right. I should be way more specific.

First: I fixed the textsetup crash, so we do no longer need to worry about that, or it's details.

Second,

I think you already know what I was trying to ask: how should I format my flashdrive, for it to be recognised? How should I test this?

You partially awnsered this question, but my vocabulary (geometry=hardware specs & how the device is formated?) is not that good I'm afraid. I'm not completely sure if I understand what this sentence means:

The BIOS' USB-boot support decides on a cylinders/heads/sectors geometry for the USB storage device.  If this geometry does not result in the ability to address the entire storage device, then the storage device is "truncated."

I'm not sure what program to use to write this mbr to my flashdrive. (Or to backup the old one.) How important is backing up the old one? Can't I just use a empty flashdrive and reformat it at the end?

Once I get this working, I'll be able to tell how I should format my flashdrive for it to be recognized?

I realize I am sort of a noob around here and I apologize for the time wasted because of this.

Edited by Fiction, 10 January 2013 - 10:57 PM.

First: I fixed the textsetup crash, so we do no longer need to worry about that, or it's details.

Good news!

I think you already know what I was trying to ask: how should I format my flashdrive, for it to be recognised? How should I test this?

...

I'm not sure what program to use to write this mbr to my flashdrive. (Or to backup the old one.) How important is backing up the old one? Can't I just use a empty flashdrive and reformat it at the end?

You can use TinyHexer or DD for Windows or various other, so-called "hex-editors".  I expect that you will gain some valuable experience by using these, as they very often come in handy.

If you have a spare USB storage device whose data can be completely lost/erased, then yes, I'd advise using that spare and simply writing the diagnostic MBR code to it.  If you are thinking about changing the MBR code for a USB storage device with valuable data on it, you must backup that MBR, first.  Otherwise, you could live to regret it.  Also: Make sure that backup is not on the USB storage device, itself!  It should be backed up elsewhere.

Once I get this working, I'll be able to tell how I should format my flashdrive for it to be recognized?

Possibly.  But if it turns out that your BIOS has a maximum C/H/S that truncates the device, then that BIOS has a limitation and you might have better results by using PLoP.

I realize I am sort of a noob around here and I apologize for the time wasted because of this.

"C/H/S geometry" are three numbers which basically state how large the device is.  They would be most useful for reporting here, at this point.  Then someone can multiply them and tell you how large BIOS thinks the USB storage device is.

I am using windows7 on this pc. How do I load the USB drive MBR in the hex editor? There obviously is no 'file' to select. Or is this a functionality of the editor themselves? If not I'd rather use the XVI32 hex editor, since I'm more familiar with it.

*edit*

I realize being a windows user makes me look even more of a noob

Is it possible to open an MBR as if it's a file by default in linux?

Edited by Fiction, 10 January 2013 - 11:59 PM.

[quote name="Fiction" post="165570" timestamp="1357814295"for motherboards this old to only accept USB drives formtatted in a certain way?Yes, that's it.

Historically there are no USB flash hardware at BIOS level.
Historically there are Floppy, hard disks at different sizes, ZIP and what's so ever.
There had been several size limitation at hard disks so far, the same goes for USB flash booting.
Historically a old BIOS may map USB flash to another historic hardware.

Imagine back then there exist 2 GB flash drives only.
A historic BIOS is builded and testeded for 2 GB flash drives.
The manufacturer can't test a 4 GB drive, because there exist no 4 GB drive back then.
A 4 GB drive could not be tested back then.

To summerize, simplified approach: use a 2 GB flash drive at a historic BIOS

Sorry for the layout, forum software is a nightmare.

I am using windows7 on this pc. How do I load the USB drive MBR in the hex editor? There obviously is no 'file' to select. Or is this a functionality of the editor themselves? If not I'd rather use the XVI32 hex editor, since I'm more familiar with it.

TinyHexer contains a File menu option to work with a disk.  XVI32 does not.  You can also use wxHexEditor.

I realize being a windows user makes me look even more of a noob

Is it possible to open an MBR as if it's a file by default in linux?

If you're using Linux, you can use bvi.

I am trying this, using TinyHexer. So far no luck. It wont boot the thing: it tells me 'Disk boot failure, insert system disk and press enter' or it boots onto my hdd depending on the bios boot settings. It simply wont boot from the usb.

You told me to overwrite the first 440 bytes, but this file only contains 437 bytes. What should I do with the next 3? I tried filling with zeros, or leaving them alone, neither worked.

Or should I do this, after lowering the partition size to 2GB ? This has a 7GB 'Raw' partition (according to windows diskmanager) when I'm done with tinyhexer.

I must say: I really like the way you are explaining this to me so far I am learning a lot.

Something is not clear to me. What I understand from your first post is that initially the USB drive booted correctly, and only then at a certain point of the XP installation, it failed (to read some EULA file or something of that sort). But I'm not sure I understood correctly.

So now it doesn't boot at all? Or, was it ever successfully booting?

Have you tried looking in the BIOS itself to see how it is recognized? It may be listed in the booting devices, whether as HDD (as any other HDD you may have) or as some USB device. If the BIOS recognizes it (I mean, not by trying to boot the device, but by being listed somewhere in your BIOS, as a potential bootable device for example), then look at the size the BIOS informs about it. Is it the full size (or at least approximately the full size)?

If you can see your USB drive from Win7, can't you use RMPrepUSB to clean it, test it, add a boot code, partition it and format it? Have you at least tried to do it?

Also, maybe I missed it but, how did you actually format your USB device in the first place? How did you initially attempt to make it bootable?

O, the flashdrive does boot, but only when the partition on the flash drive is smaller then 2G. That's what I wanted to know: how should I've seen that.

@ Sha0 please see my previous post.

I am trying this, using TinyHexer. So far no luck. It wont boot the thing: it tells me 'Disk boot failure, insert system disk and press enter' or it boots onto my hdd depending on the bios boot settings. It simply wont boot from the usb.

 

You told me to overwrite the first 440 bytes, but this file only contains 437 bytes. What should I do with the next 3? I tried filling with zeros, or leaving them alone, neither worked.

 

Or should I do this, after lowering the partition size to 2GB ? This has a 7GB 'Raw' partition (according to windows diskmanager) when I'm done with tinyhexer.

If you are seeing "7 GB raw", then there does not appear to be a partition table.  Are you using your spare USB storage device, that doesn't have any files?  If so, I'd suggest using all zeroes for anything that didn't come from the diagnostic MBR, except the last two bytes of the whole sector: Make them 0x55 0xAA.  A disk isn't bootable without those two bytes.  If you're using your spare and it was never partitioned, then that could be why they're not there.

Hey, cool.  "You're" versus "your" and "they're" versus "there".

Hi Sha0,

Noticed I was opening the 'volume' (assigned F:\\) and not '\\.\PHYSICALDRIVE1' (first sector 0), like I was supposed to. I tried opening his guy, filling access space with 0-s except the last 55aa. No luck yet. I'll try some more variations (not filling it with 0-s for example) but right now it's still being ignored completely when booting.

"You're" versus "your" and "they're" versus "there".

Yeah, sorry, I'm not a native speaker. I tend to make a lot of grammar mistakes when im not paying enough attention to it.

*edit*

I think I did something wrong while saving (I think it needs to be dismounted). Anyway, I got it working, I'll post the needed values in a few minutes.

Edited by Fiction, 12 January 2013 - 05:28 PM.

Ok, sorry about the double post, but apparently I used up al my available 'edits'.

Filling all remaining bytes up with 0-s was a bad idea. It only seemed to work when I left the other bits alone. Filling it up with 0-s removed all partitions from the device and (apparently) made it unable to boot.

Right now It does something at last.

My display tells me:

DIAG L D80h C3BAh HFFh S3Fh P? 0????????h M????h E??h

Not sure if this is what was supposed to happen, with all the question marks and all. But this is my first noticeable response out of it.

Since some of the 'words' start with C H and S it may be more coherent then I initially thought.

So, how should I proceed now?

Well, I tried many things to get my windows xp installation usb bootable. I don't really remember all of the tools, methods and formats I've tried.

I do know that the first thing to become boot-able was a 2GB FAT16 partition, 'filled' using novicorp wintoflash.

My fist attempts where possibly on a usb drive of a size around 7.3 GiB, so this might have been, 'truncated' was it?, by default. However I do remember trying a 3.7 GiB usb afterwards (I think). It is all kind of hazy, I'm afraid.

By the way, most of the programs did either the same thing wintoflash does, or takes the 'YUMI - Multiboot USB Creator' approach. However the later wont work because the ISO in question is larger then my RAM memory (a whooping 512 MB).

All in all, to find the real culprit I would need to do most of my own 'experiments' again. But I tried a lot and don't really know where to start. Suggestions? Perhaps you know 'experiments' that are way more useful then the once I did.

By the way, out of curiosity, how do you know there are 512 bytes-per-sector? Is this always the case? What is'the INT 0x13 drive number' and why is it usefull to know it is 0x80? What do all the question marks after the P,0,M and E imply?

I was commenting on the sentence I had just used before, not criticizing your English.  Sorry for the misunderstanding.

I had a feeling you were not criticizing. I just did not understand what you meant by it. I do now however lol

Well, I tried many things to get my windows xp installation usb bootable. I don't really remember all of the tools, methods and formats I've tried.

 

I do know that the first thing to become boot-able was a 2GB FAT16 partition, 'filled' using novicorp wintoflash.

Sorry, I'm not familiar with wintoflash.

My fist attempts where possibly on a usb drive of a size around 7.3 GiB, so this might have been, 'truncated' was it?, by default.

I disagree.  The mbr-diag.bin output you shared shows that the whole device can be addressed by INT 0x13, so I do not think it would have been truncated.  That was the main point of my last post.

However I do remember trying a 3.7 GiB usb afterwards (I think). It is all kind of hazy, I'm afraid.

 

By the way, most of the programs did either the same thing wintoflash does, or takes the 'YUMI - Multiboot USB Creator' approach. However the later wont work because the ISO in question is larger then my RAM memory (a whooping 512 MB).

I'm not familiar with that YUMI, either.

All in all, to find the real culprit I would need to do most of my own 'experiments' again. But I tried a lot and don't really know where to start. Suggestions? Perhaps you know 'experiments' that are way more useful then the once I did.

I'd suggest using the HP USB Disk Storage Format Tool, which is described a bit under section #3 at this page.

By the way, out of curiosity, how do you know there are 512 bytes-per-sector? Is this always the case?

It is pretty common for INT 0x13 hard disks to have 512-byte sectors. It's not guaranteed, but anything else would be remarkable.

What is'the INT 0x13 drive number' and why is it usefull to know it is 0x80?

Some BIOSes support booting a USB storage device as a "superfloppy" (over-sized floppy), where there is no partition table.  Some BIOSes only support this mode.  The INT 0x13 drive number would be 0x00 for such a situation.  Numbers >= 0x80 are non-floppy devices.

What do all the question marks after the P,0,M and E imply?

That your partition table was empty at the time you booted.

I disagree.  The mbr-diag.bin output you shared shows that the whole device can be addressed by INT 0x13, so I do not think it would have been truncated.  That was the main point of my last post.

I am sorry I thought you meant that the 7.308 GiB what the BIOS's limit, but now I think you meant there is NO limit at all.

That is about 7.308 GiB.  That means that the BIOS appears to be fully capable of addressing the entire device.

The value we get out of this is 7.308 because this is (aprox) the cappacity of my usb device: it can access the entire drive. Do I understand correctly?

What is we'd get this value, while using a usb device with a capacity of 20GB? Then there would have been a limitation, wouldn't there?

I'd suggest using the HP USB Disk Storage Format Tool

I used the HP USB format tool before, but never succeeded in creating a boot-able msdos drive before.

Odd enough, today it succeeded on first try , I used a FAT32 parition with a size of 7.3 GiB. (according to windows that is).

I guess now I could just copy the xp installation over and start winnt.exe in dos?

Perhaps the 2G limit was caused by the programs I used or something.

What if for some weird reason I would like the device to be formatted in FAT16? (not that I do, just curiosity)

The HP tool tells me that 'the volume is to large'. Even when I change the size of the partition, it will still try to format all of the space.

I am sorry I thought you meant that the 7.308 GiB what the BIOS's limit, but now I think you meant there is NO limit at all.

The output of mbr-diag.bin showed how many sectors could be addressed, and it matched your earlier post about the size of the device, so there is no smaller limit involved; the BIOS can handle at least this size of USB storage device, as you've proved.

The value we get out of this is 7.308 because this is (aprox) the cappacity of my usb device: it can access the entire drive. Do I understand correctly?

Yes.

What is we'd get this value, while using a usb device with a capacity of 20GB? Then there would have been a limitation, wouldn't there?

We cannot know the BIOS limitation, if any, unless it is documented or unless experiment reveals it.

I used the HP USB format tool before, but never succeeded in creating a boot-able msdos drive before.

Odd enough, today it succeeded on first try , I used a FAT32 parition with a size of 7.3 GiB. (according to windows that is).

I guess now I could just copy the xp installation over and start winnt.exe in dos?

Not really.  The Windows XP installation process does not expect the installation files to be on a USB flash storage device.  It might not load drivers for USB storage, so the USB device might be inaccessible by the installation process.  You can modify the installation files to modify this behaviour; fairly old news, but you'd need to read all about it.

Perhaps the 2G limit was caused by the programs I used or something.

Possibly!

What if for some weird reason I would like the device to be formatted in FAT16? (not that I do, just curiosity)

The HP tool tells me that 'the volume is to large'. Even when I change the size of the partition, it will still try to format all of the space.

The limit of a FAT16 volume is documented by several online resources and can be researched on the Internet by using the right keywords.

The limit of a FAT16 volume is documented by several online resources and can be researched on the Internet by using the right keywords.

I know that, it is limited to 2GiB. However, what I was trying to ask: how would I make a bootable ms-dos usb device in FAT16? Since the HP tool will not doe it, because the partition needs to be smaller than the full capacity of the device.

Like I said: this is a question only out of curiosity, FAT32 works a treat.

Not really.  The Windows XP installation process does not expect the installation files to be on a USB flash storage device.  It might not load drivers for USB storage, so the USB device might be inaccessible by the installation process.  You can modify the installation files to modify this behaviour; fairly old news, but you'd need to read all about it.

That's a pity. I tried googleing this but nothing came up. Any good pages on how to do this? I thought Nlite was capable of doing something like this, but I can't seem to find it in the program, perhaps I'm mistaking.

The HP format tool has been since several years "deprecated" because it has a series of limitations (some may be features, some are simply "wrong").

There are more suitable tools (and some "better" ones) but more than that the process of booting from USB (and installing XP from or to an USB device) has been detailed in - I presume - ALL possible ways.

A way to have "total" control on a partition (no matter if on USB or not) is to "manually" create the partition table for it.

Then the FORMAT command in XP will do the rest (provided that the values are "sensible").

This may be handy:

http://reboot.pro/to...atch-001-alpha/

read this also:
http://reboot.pro/to...ease-001-alpha/

Everything (or *almost everything*) about installing an XP from a USB device is here:

http://www.msfn.org/...ndows-from-usb/

several different approaches, some fully automated, some "manual" some going into the uttermost specific details.

Wonko

Thank you Wonko, I'll read into this. If I run into new questions, I'll ask them in a new threat. Since the issue I started this threat with has already been solved by Sha0.

Thanks allot, both of you, especially Sha0, I learned a lot!

my idea is that your bios could be the problem.. aslo the fast that that boards used usb pins to the front of your computer. problems with usb ports on that board was comman and gave techs lots of headaches.

here is a place where you can download bios upgrade http://support.asus....&p=1&ft=4&os=17

caution is the rule whne doing bios reflash.