[ros-kernel] Different Hardware Architectures

KJK::Hyperion noog at libero.it
Fri Apr 23 22:38:38 CEST 2004


At 19.37 23/04/2004, you wrote:
>ld should require very little change,

relocations are probably going to be the only feature to reimplement

>as AFAIK PE is PE is... no matter what the target CPU.

small correction: 64 bit architectures use a slightly different format 
(PE+), not backward compatible

>Anyone know if it's possible to cross-compile from any host CPU 
>architecture to x86 Win32,

GCC can even be compiled as a cross compiler on a system that is neither 
the host nor the target of the compiler to be built, so I'd guess it's possible

>Theoretically; mnyes (not a kbd slip). Practically, I don't see this ever 
>happening. I believe the effort to make that happen could be one to three 
>orders of magnitudes larger than make the tools emit and parse COFF and PE.

not to mention that "introspection" (the program parsing its own executable 
at run-time) is fully supported and documented on Windows. "Win32 <=> PE 
executables", pretty much

>- Do the Power CPU's support endianness selection (on startup), or am I 
>confusing them with MIPS?

they must have, since Windows NT for PowerPC existed, *and* the X-Box 2 
(which has an NT-based kernel) is rumored to be a PowerPC (and Microsoft 
never claimed that Windows can run on a big endian architecture) 



More information about the Ros-kernel mailing list