> 	Dan, just because you disagree with the approach

I do not disagree with the approach. Yes, it is not what I originally
suggested, but it's a perfectly reasonable approach to fixing the
problem. The changes I suggested making were all still within the
framework you designed.

> and it does not do what you want

What I want is for gcc to not generate code with unaligned accesses on
the PowerPC, in the same way that it does not not generate code with
unaligned accesses on any other platform where that would cause a
kernel trap.

Your patch does not fix that problem, and also causes the
implementation to disagree with the documentation, in the way I
mentioned.

I am sorry for wanting gcc to not have these problems.

> You are misunderstanding how GCC and the PowerPC architecture
> operate, and just being obstinate.

Ha ha.

-- Dan