On Fri, 8 Apr 2011 07:35:09 -0400, Peter Relson <[email protected]> wrote:
>[...] >An interesting, but almost inevitable, phenomenon is that if you started >with two programs, one in asssembler, one in a HLL, both highly optimized, >over time, as that program is modified, the assembler one usually gets >worse, and the HLL one gets better. > >That's because the HLL one gets to benefit from advances in compiler >optimization technology upon recompilation. >And the assembler one suffers from the "what register is it safe to use" >and "can I safely rely on having access to this control sructure here" >paradigms that many assembler modifications must go through. The answer is >often not "the register that is best for the overall optimization of the >module" and "yes, so I do not have to reload as much". The cost of >analyzing/optimizing the whole program again is not usually justifiable >when making a modification. This has the ring of truth. Personally, i like programming in a high- level language (C++ being more high-level than C, e.g.), but the dumps are harder to analyze. Then, shops i've been in want to use Dignus instead of the IBM compiler, mainly because of LE. But now y'all have Metal. Does it put enough annotations anywhere to allow some tool other than IPCS to analyze dumps at the source-level? Thanks, Justin R. Bendich 1-512-241-7367 > >Peter Relson >z/OS Core Technology Design
