Date: Sun, 8 Aug 2010 14:35:53 +0200 From: Attilio Rao <attilio@freebsd.org> To: Ed Maste <emaste@freebsd.org> Cc: svn-src-projects@freebsd.org, src-committers@freebsd.org Subject: Re: svn commit: r211029 - projects/sv/sys/net Message-ID: <AANLkTinmYTxAESx-yg2yzvH6aLP7LQOBV4YrP4CO4Aw4@mail.gmail.com> In-Reply-To: <20100808114725.GA34359@sandvine.com> References: <201008071739.o77HdM2Y009458@svn.freebsd.org> <20100808114725.GA34359@sandvine.com>
index | next in thread | previous in thread | raw e-mail
2010/8/8 Ed Maste <emaste@freebsd.org>: > On Sat, Aug 07, 2010 at 05:39:22PM +0000, Attilio Rao wrote: > >> Author: attilio >> Date: Sat Aug 7 17:39:22 2010 >> New Revision: 211029 >> URL: http://svn.freebsd.org/changeset/base/211029 >> >> Log: >> Add a tunable for nd_enable. >> As long as TUNABLE_INT is discouraged, however, switch the type int -> long >> and adjust accordingly the sysctl stubs. > > This doesn't really make sense as far as I can tell - and particularly > for sysctls, using a long can introduce 32-bit compat issues. I'd > prefer that this remain an int. > > What benefit does using a long provide in this case? Did you see the recent thread on hackers@ about it? If we choice to go with a direction on TUNABLES_* I'd prefer to follow it now. Attilio -- Peace can only be achieved by understanding - A. Einsteinhelp
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?AANLkTinmYTxAESx-yg2yzvH6aLP7LQOBV4YrP4CO4Aw4>