Date: Mon, 28 Nov 2005 22:20:52 +0300 From: Gleb Smirnoff <glebius@FreeBSD.org> To: Luigi Rizzo <rizzo@icir.org> Cc: Vsevolod Lobko <seva@ip.net.ua>, rwatson@FreeBSD.org, Ruslan Ermilov <ru@FreeBSD.org>, net@FreeBSD.org Subject: Re: parallelizing ipfw table Message-ID: <20051128192052.GA25711@cell.sick.ru> In-Reply-To: <20051128083236.A65831@xorpc.icir.org> References: <20051127005943.GR25711@cell.sick.ru> <20051127135529.GF25711@cell.sick.ru> <20051127194545.GA76200@ip.net.ua> <20051127195914.GI25711@cell.sick.ru> <20051128062732.GA58778@ip.net.ua> <20051128161934.GY25711@cell.sick.ru> <20051128083236.A65831@xorpc.icir.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Nov 28, 2005 at 08:32:36AM -0800, Luigi Rizzo wrote: L> Given that you (Gleb) are putting in some new features, I would L> suggest a variant to your tablearg thing, i.e. L> L> + a 'setvar index value' which can be put as an always-true L> option within a rule e.g. to store partial evaluation results. L> 'index' could be a small integer e.g. 0-9, value a 32-bit L> value which is either a constant or a packet field (src-ip...) L> or your tablearg. L> + allow a 'varN' arguments everywhere you can have a rule or pipe or L> queue number. L> L> This way you can easily implement your proposal, and a lot more. L> One should remember that variables are not meant to be staved L> with the packet's state (e.g. when a packet goes back and forth to L> dummynet) but other than that i think it is a useful feature and L> a simple one to implement. I agree. The introduction of local variable "tablearg" is the most unpleasant place of my patch. I will implement your proposal as soon as we have a second think that needs to be stored during one rule evaluation. -- Totus tuus, Glebius. GLEBIUS-RIPN GLEB-RIPE
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20051128192052.GA25711>