Re: [tree-ssa] Out of SSA status and issues

[Michael S. Zick <mszick@goquest.com>]

On Tuesday 13 May 2003 01:50 pm, Geoff Keating wrote:
> kenner@vlsi1.ultra.nyu.edu (Richard Kenner) writes:
> >     Consider instead that '*p' is just the name given a register's
> >     contents.  Neither 'i' nor '9' are processor internal.  Both are
> >     external to the cpu.  The memory reference for 'i' is source code
> >     implicit, the memory reference for '*p' is source code explicit.
> >
> > True, but given caching effects, it's hard to see how *p could be less
> > expensive than 'i'.
>
> Given the code
>
> i = *p;
> // operations that don't use 'i' or change '*p'
> foo(i);
>
> and this is the only use of 'i', surely this is always as fast or faster
> than
>
> foo(*p);
>
> due to reducing register pressure.
I think this sub-thread has developed a split viewpoint.
I (and I think RK) was talking about optimization at the hardware
independent level.
The above seems to be in the area of hardware dependent
optimizations.
Translation: You folks have lost me on this one, I'll go back
to lurking.

Mike