Print

[Cornholio]

Won't work.

I could probably put together a version that does, if you need it.

[BenJeremy]

QUOTE (Cornholio @ Jan 9 2003, 05:27 AM)

Won't work. I could probably put together a version that does, if you need it.

I think that's what undead had in mind.

For those out there with the experience and tools, building a "dual boot" BIOS with the 1.03 Debug BIOS should be a snap. All it would require is to remove the RC4 and re-mate it with the same exploit used in the D.6 and 4974 BIOSes, right?

[BenJeremy]

"Cherry?"

Is this leading to an elaborate "Like a virgin" joke or something  

I sure wish Team Xecutor would release their BIOS enhancement tools.... but in reality, that probably won't happen anytime soon, given the nature of their work (financially, that is).

They've been very quiet on this front.... It would be nice to see a unified BIOS, with Debug features built in as well as a bunch of other bells and whistles I've stated before.

[tommyhl]

QUOTE (Cherry @ Jan 9 2003, 02:38 PM)

QUOTE (BenJeremy @ Jan 9 2003, 12:58 PM) For those out there with the experience and tools, building a "dual boot" BIOS with the 1.03 Debug BIOS should be a snap. All it would require is to remove the RC4 and re-mate it with the same exploit used in the D.6 and 4974 BIOSes, right? Pretty much. I wouldn't say it was "a snap", but it's not too much work.

What's the extra work apart from the repacking? (Just interested.)

A 1.1 debug BIOS would be great. I was considering looking into it, but I've not even got so far as to work out how to work out how feasible it might be

[undead]

QUOTE

I wouldn't say it was "a snap", but it's not too much work

It is pretty much a snap..  I have working code that does just this !!  the only problem is i cant seem to make the damn Complex Debug Bios work with that exploit ??

I can unpack any old Evox 2.5ColorBlue BIOS with the old RC4, Re-Pack it with the new exploit and works 100%

I even re-packed it as a 512kb and still works..  although the Complex Debug hangs at Boot after the flubber ??

maybe there's some extra stuff in the 2BL that needs to be done ??

for the most part i was able to convert ANY old Bios into the new "Dual" Bios format except for the DAMMN Debug versions...



Is there anything "Special" about the Debug version that i should keep in mind ??

[BenJeremy]

QUOTE (undead @ Jan 10 2003, 12:39 AM)

QUOTE I wouldn't say it was "a snap", but it's not too much work It is pretty much a snap..  I have working code that does just this !!  the only problem is i cant seem to make the damn Complex Debug Bios work with that exploit ?? I can unpack any old Evox 2.5ColorBlue BIOS with the old RC4, Re-Pack it with the new exploit and works 100% I even re-packed it as a 512kb and still works..  although the Complex Debug hangs at Boot after the flubber ?? maybe there's some extra stuff in the 2BL that needs to be done ?? for the most part i was able to convert ANY old Bios into the new "Dual" Bios format except for the DAMMN Debug versions... Is there anything "Special" about the Debug version that i should keep in mind ??

My guess is the packer is assuming a 256k BIOS, and putting a boot block smack dab in the middle of the code, instead of where it belongs.

[Cherry]

QUOTE (undead @ Jan 10 2003, 06:44 AM)

It is pretty much a snap.. I have working code that does just this !!

Well, yes, if you've already written the code you need it's a snap It's getting all the little bits and pieces you need in order thats a pain. Not difficult by any means, but there are a fair few things you need to keep track of.

QUOTE (undead @ Jan 10 2003, 06:44 AM)

ITs probably just some stupid thing that i'm missing.. maybe the kernel entry point is hard coded to retail version in the boot loader or something?

Can't be that - if you've got the anim running then the kernel is already running. One thing it does do at that point is copy the data seg from the BIOS into RAM. If it's copying the wrong data it will probably go boom. Are you 100% sure the data is in the right place ?

I'll try to find time to take a look at this myself. My only work with the debug bios so far has been trying to squeeze the thing into 256K. I recon I can probably manage that, but I spent ages trying various things and basically got bored with it. Maybe I'll take another look.

[defor]

Cornholio/Cherry - I would love you long time if you could pull off converting complex-debug to 1.1 support- I'm currently running 1.03 on a 1.0 box on tsop, and quite pleased with its functionality (although some screenshots i've seen of the standard xdk bootlogo show an entirely different screen- perhaps these are off the old silver units, or just an older bios release)

In terms of modifying bios, are you workign from disassembled versions, or coding from ground up in parts (some teams mention "programmed in c" but this seems impossible without sources in c, unless they refer to subroutines in c that are being linked to by the asm bios) - forgive me, I don't work in machine code and have done my best to stay away from it... (except for that one incident in hacking 68k mouse routines on an old mac)

[undead]

Cherry...

I sent you the Source Code of my DualBios Packer..  it can Pack any Kernel (xboxkernel.exe) into the "DualBios" format thats compatible with V1.0 and V1.1 !!  

The only problem is I think the Bootloader is not passing the Debug Kernel's Entry Point the needed String pointer and other parameters..

so it hangs Just before its supposed to call the XBOXDASH.XBE  


Go check your email...

[fretterh]

QUOTE (Cherry @ Jan 9 2003, 02:38 PM)

QUOTE (BenJeremy @ Jan 9 2003, 12:58 PM) I think that's what undead had in mind. [name changed. don't ask. you will know soon.] Yea, wasn't sure whether he was "just wondering" or actually needed it QUOTE (BenJeremy @ Jan 9 2003, 12:58 PM) For those out there with the experience and tools, building a "dual boot" BIOS with the 1.03 Debug BIOS should be a snap. All it would require is to remove the RC4 and re-mate it with the same exploit used in the D.6 and 4974 BIOSes, right? Pretty much. I wouldn't say it was "a snap", but it's not too much work.

hhh  

[dankydoo]

undead, I think that you have already succeeded.  The debug bios looks for the xshell.xbe or shell.xbe or something like, and not the xboxdash.xbe.  I believe that this is why you are hanging.  Take the files from the debug dash, copy them over BEFORE you flash your new debug bios, then flash this new bios.   I think that you have it.


dankydoo

[Cherry]

[undead]

Hey Cherry !!

The new Debug Kernel BOOTS packed with my Dual BIOS packer !!!  thats a GREAT step forward, Good job !!!

What exactly was being patched in memory with the Complex 2BL ?? maybe we can add it into the DualBios 2BL ?

There is 1 slight hickup still.. when u start a remote debug session, the kernel basically restarts and somehow knows which .xbe to start for the debug session.. the hickup is it hangs right there.. So if i manually reboot the XBOX, the kernel starts up, and Ta-daaa executes the debug session .xbe and hits my breakpoint !!  

the same happens when i end the debug session and the kernel is supposed to start clean with your dash..  it hangs, i manually reboot and it's fine again..

So all-in-all it works ! We have done it !! or at least VERY close..  it can debug just fine, you have to do the "manual" reset though..  

Ideas Cherry ??

l8ter

[Cherry]

QUOTE

The new Debug Kernel BOOTS packed with my Dual BIOS packer !!! thats a GREAT step forward, Good job !!!

QUOTE

maybe we can add it into the DualBios 2BL ?

There's no need to do that, the patch I sent you is a better solution.

The hanging thing - I think I know what that might be. It's SO annoying not being able to test this myself, it would be so much quicker / easier.. Ah well, maybe I'll win one o' those X2Ls

Can you test this on a v1.0 Xbox to see if you get the same behaviour ? (Just to make sure - it could be something to do with the way it forces a reset, which might not work on 1.1s - hopefully, it isn't )

[undead]

QUOTE

Can you test this on a v1.0

I'm using it on a 1.0