Jump to content

How to install Windows from USB- WinSetupFromUSB with GUI


Recommended Posts

You are very attentive, thank you for that. I already composed the following post, you were ahead of me:

I just discovered that, even when formatted with NTFS, the partition on the stick holding the installation source shouldn't be over 2GB. At 900MB, no Overrides, PQI acts flawlessly, "Unknown" still is assigned D:, wonder how come they act so differently.

My only problem now is to put an extended partition on the PQI holding a logical partition (in order to avoid a mess in driveletter assignment) which in turn holds all the stuff needed to copy over during installation. Matter of finding out and changing scripts, I hope.

Rest the question why "Unknown" doesn't confirm to the migrate stuff. For the moment I'm more curious than desperate, what may change when I don't succeed in putting a second partition on the PQI.

And answer your post, after installing XP with the "Unknown":

The PQI, which caused the BSOD, I had the same suspicion about, so I reformatted it with the HP program and ran different test, amongst them chkdsk, no problems. Being larger than 2GB turned out to be the culprit.

The installation finished, I attach the exported .reg from the new installation, more accurate than from "any" computer, I think.

Mounted.7z

Link to comment
Share on other sites


You are very attentive, thank you for that. I already composed the following post, you were ahead of me:

I just discovered that, even when formatted with NTFS, the partition on the stick holding the installation source shouldn't be over 2GB. At 900MB, no Overrides, PQI acts flawlessly, "Unknown" still is assigned D:, wonder how come they act so differently.

My only problem now is to put an extended partition on the PQI holding a logical partition (in order to avoid a mess in driveletter assignment) which in turn holds all the stuff needed to copy over during installation. Matter of finding out and changing scripts, I hope.

Rest the question why "Unknown" doesn't confirm to the migrate stuff. For the moment I'm more curious than desperate, what may change when I don't succeed in putting a second partition on the PQI.

And answer your post, after installing XP with the "Unknown":

The PQI, which caused the BSOD, I had the same suspicion about, so I reformatted it with the HP program and ran different test, amongst them chkdsk, no problems. Being larger than 2GB turned out to be the culprit.

The installation finished, I attach the exported .reg from the new installation, more accurate than from "any" computer, I think.

Mounted.7z

It's weird this 2GB barrier...Could it be the flash drive is faulty or with fake size? Or BIOS? Have you tried on other machines, the same USB stick?

As for the extended partition- in this case you'd need to use a filter driver such as hitachi microdrive filter driver or dummydisk in order Windows to see it as fixed disk and be able to show and use other partitions on it. Or to turn the removable bit on the USB stick controller, making it fixed drive.

As for the drive letters- what drive letter is the USB stick in question on this machine?

Link to comment
Share on other sites

Weird it is.. Both the flash drives were checked with RMUSBPrep for true size. They act the same on two different computers.

The driveletter is D: in the Mounted.7z, persistent after replugging.

As for the extended partition, this turned out not to be necessary after all, because a second primary partition wasn't seen at setup so no drive letter interfered. I'm thinking of integrating the Hitachi filter driver (cfa) and check for the correct drive letter in the script after install. The only hassle is editing the .inf for every individual flash drive. So changing the removable bit could be a better idea, but then the extra drive letter will show up during setup and may cause problems. I'll try it out. Unless you have a brainwave...

Link to comment
Share on other sites

You could use cfadisk with a generic PNPID, search cdob's topics with keyword cfadisk. You could also use dummydisk.sys, it does not require PNPID and does work during Text mode.

As for the drive letter of the unknown USB stick- I guess it doesn't have a serial number, let's confirm:

http://www.nirsoft.net/utils/usb_devices_view.html

Does it show a serial number? Does this stick get a new drive letter when plugged into another USB port on the same computer?

If that's the case..hmm, changing USB port might also mean a new ParentID Prefix/MountedDevices key, perhaps also a new drive letter, look at section USB serial number:

http://www.uwe-sieber.de/usbtrouble_e.html

Would have to figure out how Vista/7 deal with these cases too. No such USB stick around to play with, but interesting issue to be studied further.

Thanks cdob for pointing the right track.

Link to comment
Share on other sites

Being larger than 2GB turned out to be the culprit.

That's quite strange. I suspect a BIOS limitation.

Which motherboard do you use?

Your sitck is greater 2GB. Which storage size do you use now?

I'm thinking of integrating the Hitachi filter driver (cfa) and check for the correct drive letter in the script after install.

Post #1024 contains a dummydisk example

Integrate drivers to windows source files first.

Link to comment
Share on other sites

@ilko: "Unkwown" reports no serial number. The computer has not been set to ignore it. So apparently this is what breaks assigning the drive letter to "U" and beyond.

Aside: It may be desirable to add one or two letters, as some MSCR's contain three devices and can't easily be switched off (notebooks).

Any flash drive gets D: assigned, unless a second one is plugged simultaneously, then the next available letter is used. Replugging this second one after a restart makes it return to it's assigned letter.

Thanks for pointing to Uwe Sieber's site. I'm using his USBDLM for years now.

@cdob: This testing happens on a MS-7207V2, my garbage bin, although on a brand new ASUS-based notebook my confusion started. I must agree, not very trustworthy BOISes, but the MS-7207 has been eating everything so far. I used 900MB for storage, enough for the XP-source and some initial stuff like DP's.

Both: I didn't go into the cfadisk matter for being too tired, tried the dummydisk way, which seemed pretty straightforward. But it didn't work, even put Dmmydisk.sys in Dosnet.inf, doublechecked the typing, but it never showed up in the installation. So I took the crowbar and initiated the first script with the commands as in attachement. It does the job, the second partition shows in seconds and I can go on.

I'm glad I mentioned this problem-to-come, both your responses saved me days of digging into a technique I'not familiar with.

The phenomenon of the 2GB still exists, I could work around it thanks to your great help, but if either of you wants me to check things out I'll be glad to, if it were only out of curiosity.

Off topic: The "Unknown" was bought by a friend of mine in a photoshop in Spain, as medium for emptying her cameracard. (Don't ask.) Being back home she asked me to download the pictures on her computer, on which, as with all my installations, anything about autorun had been switched off. The flashdrive contained an autorun.inf and about the meanest trojan possible, which turned out to be widely spread in Spain by that time. It was easy to make the drive switch owner. Now here is my punishment!

USB.7z

Link to comment
Share on other sites

Unknown seems to have some identity:

Device Name: +[x:]+USB Mass Storage Device(General USB Flash Disk USB Device)

PnP Device ID: VID = 090C PID = 1000

Serial Number: 5&&167ABEC4&&0&&4

Revision: 1100

Device Type: Standard USB device - USB2.0 High-Speed

Chip Vendor: SMI

Chip Part-Number: SM321~SM325

Product Vendor: (N/A)

Product Model: (N/A)

But please, don't give it too much thought. As a matter of fact, I'm on the verge of putting all of the PQI's and this one aside because so much formatting and testing has been done with them that I don't trust them any more for a serious job like this. And it may very well be that with a decent modern controller the 2GB barrier is gone. But nevertheless, I learned a lot, the cfadisk integration will be tried out soon but I probably will stick with my "crowbar" solution as in USB.7z because it can be switched back after the job and the OS will act as normal again.

BTW, if used in a batchfile it is advised to add some ping's between the lines and at the end, it may take some time for the second partition to become available.

Link to comment
Share on other sites

But please, don't give it too much thought. As a matter of fact, I'm on the verge of putting all of the PQI's and this one aside because so much formatting and testing has been done with them that I don't trust them any more for a serious job like this.

Actually, as I see it EXACTLY because of that the "Unknown" SMI stick is a good candidate for tests, as you have *NOTHING* to loose. ;)

Of course if you see the experimenting as "fun". :)

Here:

http://flashboot.ru/index.php?name=Files&op=cat&id=10

Quick "generic" instructions:

http://reboot.pro/13675/page__pid__121989__st__127

jaclaz

Link to comment
Share on other sites

@Robou

Couldn't get this part- does the drive letter change if you plug the same unknown brand USB stick into another USB port on the same computer and no other USB sticks were plugged meanwhile?

What happens if you unplug this one, plug another one and plug back the first one in another port?

Link to comment
Share on other sites

@ Ilko: It is confusing, I agree, the more because I try to make a logical story out of a whole bunch of activities late at night with some booze on the side. But D: is the first available drive letter. So whenever a flashdrive is plugged in it is assigned D:. A second simultaneous one becomes assigned H:, which comes next as available. Now all the drives are unplugged. If I replug SMI, formerly known as "Unknown" (to make things easier!) it always claims D:, even when it has been H: before, whereas any other drive returns to H:. Between every test I cleaned the registry with Devcon and restarted.

@ Jaclaz: Sure it is fun. Unknown territory! It would be more fun if the links would work, Firefox complains the redirection will never complete for the Russian site. But I get the idea and make the bastard act as it should in this context. Only then I'll use it for garbage.

I'm convinced that everything I experience is far from normal. Normal is what I learn from all the posts concerning this subject. But nevertheless, obviously these occurences can happen. I'm also convinced that all your knowledge as published in your posts will eventually get me out of problems. So please, consider this as purely experimental. I feel ashamed by calling all this attention.

Link to comment
Share on other sites

Can you run MkMigrateInf.cmd against the drive letter of that stick, put the resulting c:\migrate.inf in \winsetup\XPpSP..\I386 replacing the existing one and test what drive letter the same USB stick is going to get during text mode. No need to install, just get to the disks/partitions screen.


MkMigratgeInf_c.cmd H: c:\migrate.inf

Replace H: with the drive letter of the USB stick.

I'd love to find out if drive letter of such sticks without serial number can be preserved via migrate.inf.

Link to comment
Share on other sites

So please, consider this as purely experimental. I feel ashamed by calling all this attention.

Yep :thumbup , that's exactly the idea.

But you see the number of volunteers lab rats is so small that I have to cultivate each and every occasion .... ;).

And I do have one word for you:

http://www.imdb.com/title/tt0061722/quotes?qt0282091

Mr. McGuire: I want to say one word to you. Just one word.

Benjamin: Yes, sir.

Mr. McGuire: Are you listening?

Benjamin: Yes, I am.

Mr. McGuire: Plastics.

Benjamin: Just how do you mean that, sir?

pssss, come closer, don't be afraid .... : Opera

jaclaz

Edited by jaclaz
Link to comment
Share on other sites

With the original migrate.inf it is D:

With the fabricated one it is U:

Worthy of your love?

Can you attach the new migrate.inf along with the output of ListUsbDrives.exe -d ?

http://www.uwe-sieber.de/files/listusbdrives.zip

If you have a chance to produce the same way migrate.inf and output of ListUsbDrives.exe on either Vista, Server 2008 or 7 machine it would be great.

I'd like to figure out how exactly ParentID Prefix is calculated when there is no serial number. In case one is curious about the subject:

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...