Cheat Engine

[killer+bee]

DBVM not booting.

Get the message:

"This tool originated at
www.cheatengine.org
Remove the disk"

At this point the system freezes. alt-ctrl-del ineffective. requires hard reset.

I burned the image to cd and verified the write. Obviously it boots to a point.

My system:

P5N32-E SLI Plus motherboard.
E6600 dual core duo
4 gig ram
Vista 64 ultimate
sata hd

Perhaps dbvm is for vista 32 and not 64. I've not read this but it's my first guess.

[Dark Byte]

no, it works on 64-bit, thats what I tested it on.
Is your first hard disk the bootable hard disk ?

also, try using the floppy disk image. Thats the one I always test with (and at least then the cd won't get picked up as first hard disk if you have a weird bios)

but next version i'll add some extra info to it.

[killer+bee]

[Dark Byte]

a usb stick won't work with the img no. When loaded it's handled as a harddisk and dbvm current'y doesn't support booting up from hard disk yet. (it loads the first sector of the harddisk into the virtual machine and starts executing from there)

question: You say you get the message to remove the disk. Do you also get the text saying : "Virtual machine manager loaded" ?

[Dark Byte]

hmm, if you feel like helping/debugging.
http://cheatengine.org/dbvm04.rar

it's got some extra information on the display. (I assume you don't have a serial port/cable attached to another comp. Else I could even give the debug version with complete menu system)

[killer+bee]

[Dark Byte]

sigh, no real floppy disk, no real serial port... I'm amazed it has a keyboard. Hmm, actually, it wouldn't surprise me if it's a usb keyboard...

anyhow, in this new version you should see as last message "Virtual machine configuration successfull. Launching..."

if it freezes before that it's not related to the boot sequence but stuff like configuring the VM, or detecting if your system has intel-vt

also, I don't think it's grub thats causing a problem, when I tested dbvm on linux I had to mess with the launch parameters in the grub menu to test stuff. (what better way to test for bugs than an open source os)

[killer+bee]

I may try the floppy idea again. After all it should not matter that vista doesn't see the usb floppy. Once the dbvm is in memory it shouldn't matter that the floppy isn't recognized by vista. BTW, I built this box by hand and would appreciate it if it wasn't looked down on because I didn't see a need for legacy ports or a 1.44 floppy device. I'm only trying to help here. Not that it matters but it does have ps2.

CD boot of dbvm 0.4 on vista 64 box:

Starting the virtual machine
Sending other CPU cores into VMX wait mode
Wait done, other cores are ready to be handled by vm
Calling startvmx for main core
System check successfull. INTEL-VT is supported
Checks successfull. Going to call vmxon

Then system freeze.

edit: should = should not

Further edit since I can't double post:

I was able to configure my bios to recognize the floppy drive. I booted from floppy and the results were identical to booting from CD. Same debug messages and same freeze point as above.

The floppy is preferable for this since I can overwrite. I don't have any cd-rw's currently and starting to get low on my cd-r's.

I hope this is helpful.

[Dark Byte]

yes, sorry about that remark. It's just that usb devices make stuff a lot harder. (E.g I prefer just sending data to a specific port, but with usb that's not possible. E.g usb serial cable. without the proper configure software I can't use it)

anyhow, I doubt the floppy version works, seeing vmxon fails to run. Which is exactly where I didn't expect it so not very detailed.

It could be due to having 4GB ram and it somehow messed up memory allocation, or the VirtualToPhysical routine is messing up.

I'll make a version with even more debug info to the display

[killer+bee]

I have the ability to take avi video and compress with xvid if you need to scroll out a lot of output.

I made a short video of the previous boot to check resolution and it turned out well. 1.8mb with audio removed. It would have been less but I included the post messages and I kept the resolution fairly high.

Also I can easily degrade the ram in increments of 1gig if that would be useful in troubleshooting.

Not sure if you caught my post edit from before but I was able to get the usb floppy working with the bios.

[Dark Byte]

yes, I've read that, floppy's are the best. (fast and reusable)
also, I uploaded a new dbvm at http://cheatengine.org/dbvm04b.rar
it should now give a bit more info I hope. If it's really vmxon and not one of the other 2 functions inbetween it'll be hard to debug

also, I have 2GB ram and tested with that (in case you feel like taking 2GB out)

[killer+bee]

I formatted the floppy (not necessary but makes me feel good), downloaded dbvm .04b and dd'ed vmdisk.img to floppy.

Booted vista 64 with 4gig of system ram and got:

Starting the virtual machine
Sending other CPU cores into VMX wait mode
Wait done, other cores are ready to be handled by vm
Calling startvmx for main core
System check successfull. INTEL-VT is supported
Checks successfull. Going to call vmxon

Then system freeze.

This is no change from before.

Booted vista 64 with 2gig of system ram:

Same messages as above. Same freeze.

Not sure if it matters but the motherboard uses the nvidia chipset (nforce 650i) not the intel chipset for all i/o such as sata drives, ram, ethernet, etc. Running bios version Asus p5n32-e sli plus 901 (latest).

I double checked and I do have VT set to enabled in bios.

[Dark Byte]

you sure the write to disk was successful?
because what you explain is a pure crash on vmxon

also, the mainboard shouldn't matter much, besides the bios option to allow VT, but as you said, it's on

try this one, it'll contain a few checks for NULL pointers before vmxon is called
http://cheatengine.org/dbvm04c.rar

it also contains a check for unexpected interrupts like pagefault and general protection errors.

[killer+bee]

[Dark Byte]

I really expected a general protection error here

when do you get that INTERRUPT FIRED:Pagefault message ?
right after "Checks successfull. going to call vmxon", or even before that ?

And is the vmxon_region NULL or not ? (thats part of the new messages)