Hi again Harald,

Harald Welte wrote:

>>With this patch, if an expectation is not found for the conntrack, it 
>>will look for an helper and go out, this way we save three "if's". By 
>>other hand, if an expectation is found we save one "if".
>>    
>>
>
>yes, thanks for the patch.  It definitely makes the code more readable.
>If it really makes a difference in performance - I don't know.
>
>The compiler will certainly not recompute the value of 'expected', and
>since it also knows that expected is not written to, it could actually
>optimize here.  Whether gcc is smart enough to do so, I don't know.. one
>would have to look at the generated assembly code.
>
>  
>
well, I'm not a compiler geek so I don't know either. But it's true,
it's more readable :-). So maybe "optimization" wasn't the correct subject.

>>I consider that the general case is that an expectation is not found, 
>>isn't it? because most connection tracked by the system don't need a 
>>helper, so they not need to handle expectations.
>>    
>>
>
>yes.  Maybe it's even worth using 'if (unlikely(expected))' in that
>case.
>
>  
>
I didn't modify the patch, well I think it's ok anyway.

>>BTW, can we consider that if there's no helper for a given tuple, it 
>>won't have expectations? because while writing the patch i had more 
>>ideas, maybe this could be optimized a bit more. Well, I'm not sure...
>>    
>>
>
>No, this is an invalid assumption. 
>  
>
ok, thanks!

>>hope that i'm not missing anything! if so, please let me know.
>>    
>>
>
>looks fine to me.  would you mind submitting a patch syncing 2.4.x to
>your changes?  Thanks.
>  
>
I attached the patch for the 2.4.x tree.

Pablo