forums.ps2dev.org

sambro · Joined: 29 Sep 2005 Posts: 7

So I was doing some Gu tests today ... Must have messed up somewhere because I crashed with a bus error (data), using PSPlink.

Anyway, I didn't think much of it, then something caught my eye ....

groepaz · Joined: 01 Sep 2005 Posts: 305

they are using that all over the place, eg have a look at the very end of the memory where initial stackframes are located :)
_________________
http://www.hitmen-console.org
http://hitmen.c02.at/files/yapspd/

digihoe · Joined: 14 May 2005 Posts: 108

I remember a sat reciver firmware that contained DEADCAFE... It turned out that CAFE was a bootloader password one had to enter (rs232 interface + keypressing on the reciver to start the bootloader) to get into bootloader mode and be able to flash a dead reciver (dead as badly flashed)... Wonder if SONY would leave the same clue...

Best regards!

chp · Joined: 23 Jun 2004 Posts: 313

Dream on, 0xDEADBEEF (together with 0xBAADFOOD, 0xBD.. to 0xFD.. (win32 crt) and others) are very common memory fill codes used for debugging purposes.
_________________
GE Dominator

Raphael · Joined: 17 Jan 2006 Posts: 646 Location: Germany

Actually this is something often used by memory managers, to mark
beginning and ends of allocated memory regions for debug purposes. So
whenever the MM is requested to allocate x bytes, it actually allocates
x + 2*y, fills the first and last y bytes with a easily recognizeable
bits (0xDEADBEEF, 0xBAADF00D, etc.) and returns a pointer to x+y to the
caller.
Now whenever the memory is about to be freed or the application crashes,
it prints out these first and last y bytes of the used memory region,
and if these have different bit patterns, then you know that somewhere
in the program you access unallocated memory.
In your case you seem to have accessed such memory also, which probably
causes the crash, though I'm not sure if this memory marking is done by
the PSP, the SDK or PSPLink.

TyRaNiD · Joined: 18 Jan 2004 Posts: 918

Well in this situation 0xDEADBEEF is actually the default register value the kernel sets when a new thread is created, some of the registers have not been used during the running of the application and that is why some of them are set to that value.