[OpenBIOS] [commit] r1078 - trunk/openbios-devel/arch/ppc/qemu

Segher Boessenkool segher at kernel.crashing.org
Fri Jan 11 18:47:51 CET 2013


>> 0 > here 5 44332211 filll  ok
>> 0 > here 8 dump
>> fff503c0  44 33 22 11 44 00 00 00  00 00 00 00 00 00 00 00  
>> D3".D...........
>
> This output in incorrect. This is what Open Firmware outputs:
> ff94d75c:  44 33 22 11 44 33 22 11

That is debatable; you could say that Apple's implementation is
buggy, instead.  You can also say that the API is badly designed,
this problem is wholly avoidable.

>> 0 > here 8 44332211 filll  ok
>> 0 > here 8 dump
>> fff503c0  44 33 22 11 44 33 22 11  00 00 00 00 00 00 00 00  
>> D3".D3"........
>
> This output is correct.
>
> Segher is wrong.

No, I'm not.  I *said* that Apple OF writes too many bytes.

I don't think that emulating this behaviour is useful.  I don't think
that Mark's do-the-tail-by-bytes is useful either.  As I said, bad API;
we should just consider any length not a multiple of four as resulting
in undefined behaviour.


Segher




More information about the OpenBIOS mailing list