Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 16 Jan 2007 10:23:17 +0000 (GMT)
From:      Robert Watson <rwatson@FreeBSD.org>
To:        Randall Stewart <rrs@cisco.com>
Cc:        Max Laier <max@love2party.net>, src-committers@FreeBSD.org, cvs-all@FreeBSD.org, cvs-src@FreeBSD.org
Subject:   Re: cvs commit: src/sys/kern kern_subr.c src/sys/sys systm.h src/share/man/man9 Makefile hashinit.9
Message-ID:  <20070116101754.G52843@fledge.watson.org>
In-Reply-To: <45ABD171.1050509@cisco.com>
References:  <200701151506.l0FF6S6D022659@repoman.freebsd.org> <200701151933.39686.max@love2party.net> <45ABD171.1050509@cisco.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 15 Jan 2007, Randall Stewart wrote:

>> and why are we useing new flags, btw?  Can't we just pass down the normal 
>> malloc flags?  I seem to remember that we do that in a couple of places 
>> already.
>
> Well... I was game to either method.. I think the general thought was that 
> there may be OTHER non-memory flags that we direct at hash-init someday.. 
> and having them not locked to memory was a good thing..
>
> Robert might care to comment on this one .. of course it can always be 
> changed to be just the memory flags too :-)

Yes -- this is exactly what I have in mind.  Today, the flags simply indicate 
memory allocation disposition, but I think it's reasonable to assume that like 
several other complex object allocators/initializers, we might want additional 
flags in the future.  I think I prefer the sf_buf_alloc() model (custom flags 
mapped as needed into underlying operations), rather than the soalloc() model 
(reuse some but not all memory flags).

Robert N M Watson
Computer Laboratory
University of Cambridge



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070116101754.G52843>