programmer optimizations?

Hi,


We are having small debate here as to the use of some of the
code optimizations by the programmer himself.


e.g.
replacing (n / 4)
by (n >> 2)


replacing if ((i < 0) || (j < 0) || (k < 0))
                by if ((i | j | k) < 0)


as you can well imagine, the list is endless


Since this produces somewhat non portable code, the view is that
such optimizations should be discouraged.


What I want to know is your view on using such "cryptic" optimizations
and whether any of the commercial compilers attempt to do such "target
machine" specific optimizations. The compiler we have here does not
seem to do it.(I am using the compiler XLC which comes with AIX on
IBM RS/6000 machines)


Ciao,
Afsal
[One reason that such optimizations are discouraged is that they're often
wrong. Turning n/4 into n>>2 is only correct if n is unsigned. For the tests
vs. or's, in the POWER it may well be faster to test and use the condition
registers than to fetch all three values each time and store results. I find
that my time is usually much better spent looking for algorithmic improvements,
changing a linear O(n) table search to a binary O(log n) or hashed O(1)
version. -John]
--

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