By RickRamsey-Oracle on Feb 24, 2014
... rewrite your code. Better yet, write code that avoids this problem in the first place. The problem can occur when an application wants to load a value that it has just stored in memory. Read After Write (RAW) operations are common, so most chips are designed with hardware that makes that happen fast. But in some cases, you can write code that stumps the hardware. And so it stalls.
And you tumble to earth in horror, screaming for your life and clawing at the controls.
And you smack into the a pile of rocks. Or, to the horror of young mothers in minivans, the freeway during rush-hour traffic. Or worse, the middle of the ocean, so that if you somehow survive the impact, you drown. And nobody finds your body. And your loved ones can never move on.
Unless you're wearing a parachute. Like the one we just published from Darryl Gove.
by Darryl Gove
Darryl explains exactly how a processor can stall from a bad RAW operation, and the common situations that cause this problem. Then he shows you how to identify, fix, and avoid writing that kind of code. Examples included. Help your loved ones move on. Read Darryl's article.
About the Author
Darryl Gove is a senior principal software engineer in the Oracle Solaris Studio team, working on optimizing applications and benchmarks for current and future processors. He is also the author of the books Multicore Application Programming, Solaris Application Programming, and The Developer's Edge.
Read Darryl Gove's blog on blogs.oracle.com/d.
Picture of radial engine taken by Rick Ramsey at Bay Area Aerospace Museum