Re: Printing decimal128 types out of registers

[Thiago Jung Bauermann <bauerman@br.ibm.com>]

On Mon, 2008-01-21 at 18:30 +0100, Mark Kettenis wrote:
> There's nothing wrong with the pseudo-register approach.  In fact it
> is used on many other architectures.  For example sparc64, where pairs
> of single-precision floating point numbers are used for
> double-precision, and four registers are used for quadruple precision.
> And on i386 we have the MMX registers that provide a different view on
> the floating-point registers.  I only think adding the whole tdesc
> machinery around it is a bit too much.

Ah, ok. I had misunderstood.

> If dlN is
> a widely accepted name for the decimal float registers, I think those
> registers should be added.

Actually, it's not a "widely accepted name". dl is the suffix used to
indicate _Decimal128 constants in C, so Luis chose it for the register
names as well.

The Power ISA uses a notation FRXp, where p for "register pair". The
instructions use the number of the first float register in the pair. We
could use $f0p to refer to register pair starting at $f0, but I'm not
sure it would be more intuitive than $dl0. I guess in both cases the
user would need to have previous knowledge or look it up in the manual.
-- 
[]'s
Thiago Jung Bauermann
Software Engineer
IBM Linux Technology Center