c++ - Is fastcall really faster? -


Is the Fastcoll Calling Conference really more fast than other calling conventions, such as cdecl? Is there a benchmark that shows how the performance of the conference affects the performance?

It depends on the platform. For Xenon PowerPC, for example, due to a load-hit-store problem with the data passing on the stack, this can be an order of difference of magnitude. I have given time to end time on approximately 45 cycles of the overhead of a cdecl function, while compared to ~ 4 for fastcall . X86 (Intel and AMD), the effect can be very low, because the registers are all shaded and renamed.

The answer is in fact that you need to benchmark yourself on the specific platform that you care about. / P>


Comments

Popular posts from this blog

sql - dynamically varied number of conditions in the 'where' statement using LINQ -

asp.net mvc - Dynamically Generated Ajax.BeginForm -

Debug on symbian -