Re: Kernel Single-user Memory usage breakdown results and tools posted in the NetBSD WIKI

["Juha Keski-Saari" <juha.keski-saari%teleca.com@localhost>]

I tested the basic allocation pattern we profiled from the kernel
bootup with both the malloc() and kmem_alloc(), and got exactly the
same results in terms of physical pages required. The difference
between the allocators may be negligible in this scenarion because in
the single-user boot there are about 160 allocation requests and only 3
calls to free(), so the case is not particularly hard to manage.

If there was a more complex use case to test, there might be more
interesting results.


Juha Keski-Saari