From owner-freebsd-current Thu Sep 24 20:39:50 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id UAA18627 for freebsd-current-outgoing; Thu, 24 Sep 1998 20:39:50 -0700 (PDT) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from owl.org (owl.org [198.206.215.91]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id UAA18616; Thu, 24 Sep 1998 20:39:43 -0700 (PDT) (envelope-from cgull@owl.org) Received: (from cgull@localhost) by owl.org (8.8.8/8.8.8/cgull) id XAA13284; Thu, 24 Sep 1998 23:38:59 -0400 (EDT) (envelope-from cgull) Message-ID: <19980924233859.21276@owl.org> Date: Thu, 24 Sep 1998 23:38:59 -0400 From: john hood To: Bruce Evans , gibbs@plutotech.com, shimon@simon-shapiro.org Cc: current@FreeBSD.ORG, eivind@yes.no, mcdougall@ameritech.net, sos@FreeBSD.ORG Subject: Re: options DPT_LOST_IRQ References: <199809242301.JAA03399@godzilla.zeta.org.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.89.1 In-Reply-To: <199809242301.JAA03399@godzilla.zeta.org.au>; from Bruce Evans on Fri, Sep 25, 1998 at 09:01:33AM +1000 Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Fri, Sep 25, 1998 at 09:01:33AM +1000, Bruce Evans wrote: > >b. In the case of a cache hit, the DPT will complete operations and > > generate interrupts less than a microsecond apart. The (now removed, > > soon to be added again) measure_performance option demonstrates this > > clearly. Such bursts can have up to 64 interrupts per burst. The > > FreeBSD interrupt code has known holes in it, that will cause such > > closely spaced interrupts to be lost. This can also cause the system to > > hang. > > I don't know of any holes. If a device raises and lowers its irq in less > than a microsecond, then it comes close to violating best-case PIC timing. > In any case, ix86's can not process an interrupt in less than about 5 > i/o times (perhaps 2.5-6 usec) in the best case. If a device raises > and lowers its 64 times in < 64 usec, then at best the handler would see > about 64/2.5 separate interrupts. This is with a generous allocation of > 1 i/o time for device-specific interrupt handling. This isn't true for systems/software using APICs, no? A quick perusal of the code seems to show that they use memory-mapped I/O, less of it than a PIC, and IIRC, it's not across an ISA-speed bus. --jh -- Mr. Belliveau said, "the difference was the wise, John Hood, cgull intelligent look on the face of the cow." He was @ *so* right. --Ofer Inbar owl.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message