Re: [Clax86list] Re: instruction rescheduler?
- From: "Charles A. Crayne" <spamtrap@xxxxxxxxxx>
- Date: Fri, 6 May 2005 06:22:16 +0000 (UTC)
On Fri, 6 May 2005 03:18:53 +0000 (UTC)
"randyhyde@xxxxxxxxxxxxx" <spamtrap@xxxxxxxxxx> wrote:
:Maybe your statement is true on the Alpha. It is not true on the x86.
"The one byte NOP, xchg EAX,EAX, has special hardware support. Although it
still consumes a __op and its accompanying resources, the dependence upon
the old value of EAX is removed. Therefore, this __op can be executed at the
earliest possible opportunity, reducing the number of outstanding
instructions. This is the lowest cost NOP possible."
-- Intel Pentium 4 & Xeon Optimization Guide
.
- Follow-Ups:
- Re: [Clax86list] Re: instruction rescheduler?
- From: randyhyde@xxxxxxxxxxxxx
- Re: [Clax86list] Re: instruction rescheduler?
- References:
- instruction rescheduler?
- From: Brandon J. Van Every
- instruction rescheduler?
- Prev by Date: Re: Storing/Saving a BMP in ASM
- Next by Date: Re: new public domain code available
- Previous by thread: Re: instruction rescheduler?
- Next by thread: Re: [Clax86list] Re: instruction rescheduler?
- Index(es):
Relevant Pages
|