RE: [lwip-devel] memp_sanity() insane

[Zschocke, Florian]

address@hidden wrote on Thursday, February 26, 2004 12:13 AM:

> I don't think it should be necessary to disable all asserts
> to get rid of
> a cpu hog like this - isn't that what LWIP_DEBUG was for anyway?

Acually in the copy we use we have mem_sanity enclosed in
#if defined(LWIP_DEBUG) && ! defined(LWIP_NOASSERT)
But it could be that we put that in because we, too, noticed that it was
hogging all our CPU resources while running in debug mode. Maybe it should
get it's own define, since it is something that does have an severe impact
on performance and is only needed when one suspects problems in the memory
management.

Florian