From mboxrd@z Thu Jan 1 00:00:00 1970 From: Qing He Subject: Re: [RFC] irq rate limit Date: Thu, 10 Sep 2009 17:17:38 +0800 Message-ID: <20090910091738.GD25462@ub-qhe2> References: <20090909084825.GA25462@ub-qhe2> <20090909141219.GC9418@phenom.dumpdata.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <20090909141219.GC9418@phenom.dumpdata.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: Konrad Rzeszutek Wilk Cc: "xen-devel@lists.xensource.com" List-Id: xen-devel@lists.xenproject.org On Wed, 2009-09-09 at 22:12 +0800, Konrad Rzeszutek Wilk wrote: > On Wed, Sep 09, 2009 at 04:48:25PM +0800, Qing He wrote: > > This is a poc patch of guest irq ratelimit. The main motivation is > > to ensure Xen's responsiveness during an irq storm, that caused by > > improper hardware or drivers. If one interrupt is fired at a high > > What kind of hardware exhibits this? I happened to see this on some network cards using MSI, everything just hung up. And to some extent, the driver of the device can control this. If assigned devices of one of the guests behave this way, at least other guests should still be operable. Thanks, Qing