Re: (Once again) Branch likely instructions (MIPS R4000)

"Vijay Krishnan" <vijaykrishnan@acmet.com>
23 Sep 2003 13:30:26 -0400

          From comp.compilers
Related articles
(Once again) Branch likely instructions (MIPS R4000) girishvgNOSPAM@yahoo.com (Girish Gulawani) (2003-09-22)
Re: (Once again) Branch likely instructions (MIPS R4000) vijaykrishnan@acmet.com (Vijay Krishnan) (2003-09-23)
Re: (Once again) Branch likely instructions (MIPS R4000) fa6nxvc102@sneakemail.com (Tim Olson) (2003-09-23)
| List of all articles for this month |
From: "Vijay Krishnan" <vijaykrishnan@acmet.com>
Newsgroups: comp.compilers
Date: 23 Sep 2003 13:30:26 -0400
Organization: Compilers Central
References: 03-09-068
Keywords: architecture
Posted-Date: 23 Sep 2003 13:30:25 EDT

Branch likely instructions are helpful, when the branch is actually
taken, as the name suggests. The use will however depend on the
number of times. you expect the condition to be satisfied.


That will result in more optimal code and faster code in execution
time as well. as the instruction would have been updated into the
pipeline.


It really depends on your code. if your aim is only code size
reduction, than whatever the case maybe.you may want to use them
liberally , as the code may easily fill in the delay slot.




Hope this helps.
Vijay


----- Original Message -----
From: Girish Gulawani
Sent: Tuesday, September 23, 2003 9:12 AM
Subject: (Once again) Branch likely instructions (MIPS R4000)


How Much Am I To Lose If I Don'T Use Branch-Likely Instructions?
My Target Is Toshiba (R4k) Tx49. The Software Is Quite Loopy & Mathematical.

Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.