"Martin v. Loewis" wrote:
> > It seems like all shared libraries and shared objects on Intel take
> > a serious penalty every time they access non-auto storage.
>
> The performance loss comes from not being able to use ebx, not because
> of the overhead of loading the register. I guess a modern process does
> not even touch main memory for that...
Not if the stack resides in uncacheable pages, although I can't think of
a situation in which someone would WANT that.
At any rate, thank you to all who replied on this thread, it has been
most enlightening. As I'm sure has become clear I am not someone who has
ever touched a code generator/optimizer although I'll definitely be
studying it in more detail.
Thanks,
Scott
