rough estimate of kadmin addprinc performance?

[Russ Allbery]

Nico Williams <nico at cryptonector.com> writes:

> I'm starting to think that we should use mmap() only for reading the
> log, but for writing we should pwrite() and then fsync() when we're
> done.  This can probably be done with relatively little violence to the
> code.

You will then not see those changes in your mmap on many platforms unless
you do extra work to force a sync, and even that doesn't always work.  We
have a bunch of configure glue in INN to try to detect all the various
problems and work around them, but it's a relatively hard problem.

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>