TABLE OF CONTENTS The first difficulty is exemplified by the second chapter.
Robert French, et al., The Dynamical Hypothesis in Cognitive Science.
Because most code running on Intel EM64T will be derived from 32-bit code initially, it makes sense to examine the challenges of porting code to this new platform. One would expect the primary difficulties to come from changes in the size of pointers, as implied by the term 64-bit, and this difference is certainly one of the central challenges. Similarly, a migration must cope with changes in the sizes of fundamental data types. This chapter discusses the code changes that you must make and a few subtle repercussions that you can expect; the next chapter shows useful ways to find these items and to fix them efficiently
Just a few years ago, when a developer referred to a computer as being 8-bit, 16-bit, or 32-bit, the term had a very specific meaning. It referred to the standard size of a word on the processor architecture. A word is the default operand of an assembly-language instruction. Memory addresses and integers in those days were all words. Hence, when an Intel Pentium III processor was described as being "32-bit," it meant that addresses and integers that were 32 bits wide. Life was simple then.
Today, this terminology is less useful because default operands can have various sizes even on the same processor. The Intel EM64T architecture, for example, is a hybrid when it comes to the sizes of its operands.
