Re: [lwip-users] robustness question

[Leon Woestenberg]

Hello Stéphane,

Stéphane Dalton wrote:

> (that I got from the DEVEL branch of the CVS) but I'd like to get users
> feedback about the robustness of lwip and its performance. Has anyone of
>  
*Robustness*: the degree to which a system or component can function 
correctly in the
presence of invalid inputs or stressful environment conditions.

lwIP is sprinkled with assertions to detect invalid input during 
development time, as
well as run-time checks against the major programming errors.

There are quite numerous checks against invalidly sized packets etc.

I cannot measure robustness for you. (Yes I can, but that would have to 
be on the basis
of a 30-page document describing what your test vectors are :-)

We are using it in industrial applications. We never had crashes due to 
lwIP, at least no
end user reports that indicate such. Knock on wood.

> you even benchmark the stack or compared it with commercial product? I've
>  
I think lwIP strongly competes with some of the TCP/IP stacks targetted 
towards 16-bit processors.
You just have to dive into the code and function headers to learn to 
interface with lwIP, and do not
count on support, although people have been quite lucky the last few 
months getting support from the
lwIP mailing lists.

Regards,

Leon.