cvs commit: src/sys/kern subr_param.c sys_pipe.c src/sys/sys pipe.h

[Poul-Henning Kamp]

In message <20030710153734.GA1101 at technokratis.com>, Bosko Milekic writes:

>  With the recent bug reporting regarding kmem_map being too small
>  coming in, I'd have to say that I agree with that assertion.  Address
>  space has become more expensive than memory, at least on ia32, and
>  it's gotten to the point where we need to re-evaluate some of our
>  autosizing and max size constants (for example, see VM_KMEM_SIZE_MAX).

I have set my cross-hair on struct buf (including sbufs) and struct
bio in this respect, they will not require permanent KVA when I'm done
with them, but the time-line is 6-current not 5-stable.  We may be
able do some low-hanging fruit for 5-stable if we are sufficiently
motivated.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.