Re: Justifying Optimization

Joachim Durchholz <joachim_d@gmx.de> schreibt:


> It turned out that a callback had been declared with a wrong
> signature, causing stack corruption with production code and running
> well and fine in debug mode.


>The point is: I spent an entire week debugging at assembly level to
>fix a bug that was entirely irrelevant to the logic of the
>application.


Did you never ask for better development tools, or programming
languages, which reveal such bugs on the very first compilation of the
code?


It may be annoying to old assembly hackers, which now are forced to
use C and continue writing their code in the "good old way", when
their new compiler will disallow the use of many of their bad
habits. But a good (rigid) programming language and compiler is the
cheapest "ghost buster", since preventing bugs can replace as many
weeks of hunting bugs as you like. Also the optimization of such code
is safer to accomplish, because the source code contains many hints
about what and how things can be optimized, in contrast to weakly
typed code, where only an intensive global analysis can reveal side
effects and other practices, which can make optimized code behave
differently from non optimized code.


DoDi

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