Re: RE: [lwip-devel] UDP - transmission inefficiences

[Kieran Mansley]

On Thu, 2010-01-14 at 07:53 +0100, Simon Goldschmidt wrote:
> Bill Auerbach wrote:
> > If that was the only change, I don't see how that can be.  I can't think
> > of
> > a case where inlining code could be slower.  I could see perhaps a
> > negligible change, but not slower.  Maybe someone else knows the platform
> > and can make more sense of this than I can.
> > 
> > Bill
> > 
> > >Platform: xilinx virtex 5 fpga
> 
> I don't know that platform, but there's always the chance that the
> previous code was better aligned for cache or the new code is slightly
> worse because it uses too many local variables to store all of them in
> registers...
> 
> Anyway, 5% might even be below accuracy of the measurement...

It does highlight the problem of trying to optimise lwIP in this way
though, as improvements for one platform can unintentionally hurt
another.

Ideally inlining would be left up to the compiler as it should be in a
good position to know when it would help and when it would not, but not
all compilers are sophisticated enough to do this well.

Kieran