Re: [boost] [MSM] Comparison with ad-hoc FSM implementation

Subject: Re: [boost] [MSM] Comparison with ad-hoc FSM implementation
From: Steven Watanabe (watanabesj_at_[hidden])
Date: 2009-12-10 08:50:43

AMDG

Edouard A. wrote:
>> This is O(1).  Of course a difference is that it has different 
>> entry points for the different events.  This doesn't seem unreasonable 
>> to me, but if you wanted a single entry point you could do something 
>> like:
>>     
>
> I'd like to point out that the implementation of switch is generally pretty
> slow, and we're talking about a code which is generally on the critical
> path.
>
> Switch implies jumping to a computed address (something like jmp
> [eax*8+ebx]... you get the idea), this is very bad for branch prediction and
> for the pipeline of your beloved processor - assuming you're working on
> something close to the IA32 or AMD64.
>   

There are many ways that a compiler can handle a switch statement.
a) sequential if/else
b) binary search
c) jump table

Also, a jump table does not necessarily prevent the processor
from predicting the branch.

> MSM's generated code will most likely contain only fixed jumps which behave
> much, much better, notwithstanding the fact that the template's context will
> help the compiler make better decisions about optimizations, resulting in
> favorable contexts in unconditional code.
>
> If you care about performance, you want those jumps out of your ways.
>   

In Christ,
Steven Watanabe