You might want to check out some of Bob Rogers' SHARE talks on "what you do when you're a CPU"..
Certainly the storage efficiency (i.e., the instruction is shorter) of BCTR is a plus in its favor. However, BCTR(G) is not superscalar (someone else will have to explain the full ramifications of this). And the fact that the machine needs to treat BCTR(G) as a branch until it can get to the point where it determines that the R2 field is 0 means that the machine's instruction decode logic must decode the BCTR(G) by itself. "By itself" is never a good thing for modern machines That is about as far as my knowledge takes me. I'm sure others can help take this further if needed. Peter Relson z/OS Core Technology Design ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html