--- Comment #3 from falk at debian dot org 2009-02-01 10:50 ---
The main problem is that just replacing the code by the below loop won't
necessarily give a speedup. strlen is usually implemented in highly efficient
and platform-specific assembly that treats several bytes at a time. I don
--- Comment #2 from esigra at gmail dot com 2009-01-30 16:12 ---
GCC already understands the semantics of strlen. If one of the operands to "<"
is a constant and the other is strlen, it is optimized (such as "strlen(str) >=
1). It just seems like the case with strlen on both sides is mis
--- Comment #1 from bangerth at gmail dot com 2009-01-30 15:13 ---
Yes, I think this would be an optimizing compiler could potentially perform.
At the same time I think you are expecting too much from the compiler: it
would have to have a semantic understanding of what the strlen functio
