> Talking about compiling modules, could someone confirm any performance > improvement if a kernel code is statically linked over compiling it as > module ??
No measurement data available with us. But I'd expect the differences to be zero once the module is loaded. At the most, there might be the (immeasurable?) differences of a pointer-dereferenced function call in the case of loadable modules, versus statically linked function call if it's included at compile time. This difference will probably add the overheads of one data fetch, not even one machine instruction (I'm assuming Intel, therefore CISC), and the data may occasionally be in the CPU cache, thus making the difference immeasurable if you are looking at overall kernel performance. But this is all theoretical analysis and conjecture. If you want confirmation by measured tests, we haven't done any. :) Shuvam _______________________________________________ ilugd mailing list [EMAIL PROTECTED] http://frodo.hserus.net/mailman/listinfo/ilugd
