Date: Wed, 05 Sep 2012 09:33:03 +0300 From: Andriy Gapon <avg@FreeBSD.org> To: Andrey Zonov <zont@FreeBSD.org> Cc: Edward Tomasz Napierala <trasz@FreeBSD.org>, freebsd-arch@FreeBSD.org Subject: Re: [patch] unprivileged mlock(2) Message-ID: <5046F21F.2080603@FreeBSD.org> In-Reply-To: <50463EEB.60207@FreeBSD.org> References: <503DD433.2030108@FreeBSD.org> <201208290906.q7T96C9j032802@gw.catspoiler.org> <20120829092318.GW33100@deviant.kiev.zoral.com.ua> <503F2D24.8050103@FreeBSD.org> <503F476E.1010505@FreeBSD.org> <50463610.6070805@FreeBSD.org> <50463EEB.60207@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
on 04/09/2012 20:48 Andrey Zonov said the following: > On 9/4/12 9:10 PM, Andriy Gapon wrote: >> on 30/08/2012 13:58 Andrey Zonov said the following: [snip] >>> [2] http://people.freebsd.org/~zont/racct.patch >> >> This patch looks correct. >> > > There is no need for this patch as I mentioned earlier. racct_set() > doesn't add any additional locking here. I thought that hiding the racct call behind RACCT was a worthy target on its own. >> And it also makes me wonder why kern/kern_racct.c is marked as standard while >> all(?) uses of racct API are placed under RACCT option. > > Not all. I think some code was not easy to put under RACCT. But perhaps it should still have been a goal for this optional feature. Unfortunately, Edward hasn't replied yet. >> Ditto for kern_rctl.c/RCTL. >> I think that excluding these file if the options are not used would help to catch >> cases where the API is used unconditionally and it would also help to reduce >> kernel sizes a tiny bit too. -- Andriy Gapon
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?5046F21F.2080603>