Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 21 May 1998 07:23:02 -0500
From:      Jonathan Lemon <jlemon@americantv.com>
To:        Chuck Robey <chuckr@glue.umd.edu>
Cc:        Soren Kristensen <soekris@alameda.net>, hackers@FreeBSD.ORG
Subject:   Re: Original PC and talk
Message-ID:  <19980521072302.21950@right.PCS>
In-Reply-To: <Pine.BSF.3.96.980520233946.312a-100000@localhost>; from Chuck Robey on May 05, 1998 at 11:47:17PM -0400
References:  <3563A210.31CF@alameda.net> <Pine.BSF.3.96.980520233946.312a-100000@localhost>

next in thread | previous in thread | raw e-mail | index | archive | help
On May 05, 1998 at 11:47:17PM -0400, Chuck Robey wrote:
> All those tricks (the same ones) are why processors like the DEC Alpha
> are so hot.  Things like register renaming don't give you much
> improvement if you're talking about such a tiny humber of registers to
> begin with (referring to the X86 here).

Huh?

Register renaming (from the architecture's point of view) only refers
to the internal on-chip registers (reservation stations), not the
externally visible registers (from the compiler's point of view).

There can be a lot of internal registers.  These are marshalled and
committed (or squashed) by the reorder buffer at the end of execution
(assuming a relatively modern chip here).

This reordering does give a performance boost, even if there are only
a few architecturally visible registers, as in the case of x86.
--
Jonathan

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19980521072302.21950>