From owner-freebsd-current Fri Mar 17 12:28:41 1995 Return-Path: current-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id MAA09143 for current-outgoing; Fri, 17 Mar 1995 12:28:41 -0800 Received: from estienne.cs.berkeley.edu (estienne.CS.Berkeley.EDU [128.32.42.147]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id MAA09136 for ; Fri, 17 Mar 1995 12:28:36 -0800 Received: from localhost (localhost [127.0.0.1]) by estienne.cs.berkeley.edu (8.6.9/8.6.9) with SMTP id MAA16591; Fri, 17 Mar 1995 12:27:49 -0800 Message-Id: <199503172027.MAA16591@estienne.cs.berkeley.edu> X-Authentication-Warning: estienne.cs.berkeley.edu: Host localhost didn't use HELO protocol To: se@MI.Uni-Koeln.DE (Stefan Esser) cc: current@FreeBSD.org Subject: Re: Sharing interrupts with PCI devices? In-reply-to: Your message of "Fri, 17 Mar 1995 10:54:34 +0100." <199503170954.AA26274@FileServ2.MI.Uni-Koeln.DE> Date: Fri, 17 Mar 1995 12:27:49 -0800 From: "Justin T. Gibbs" Sender: current-owner@FreeBSD.org Precedence: bulk >On Mar 16, 18:49, "Justin T. Gibbs" wrote: >} Subject: Sharing interrupts with PCI devices? >} I saw some commits go by about sharing interrupts under PCI. The >} PCI motherboard I have has an aic7870 on it, and when a 2940 is >} also in the machine, the bios assigns them the same interrupt. >} The probe appears to run fine for the first card, but the second >} card hangs sortly after initialization. I haven't found a way to >} force them onto separate interrupts in the BIOS or elsewhere, so >} it looks like I'm stuck (the onboard controller cannot be disabled >} :-(). Does the PCI code already contain logic to pass the proper >} unit number down to the driver if two cards use the same interrupt? > >Shared interrupts have been working >for some time on our local systems, >and the code will be put into the >CVS repository Real Soon Now (TM). > >PCI interrupts can be shared, but ISA >interrupts will have to be assigned >uniquely as before. > Why can't we chain them? I thought that the PCI spec allowed for PCI devices to share ISA irqs. > >The PCI code generally finds the IRQ >line the PCI interrupt is routed to >by a PCI BIOS call, or (as the current >BSD code does) in a PCI configuration >space register. I take it that both methods return the same result. >The NCR driver doesn't install an >interrupt handler if the IRQ found >this way is '0'. It works in a polled >mode instead. There is nothing to >prevent you from making the driver >configuration depend on the IRQ being >specified. So it's in fact quite easy >to selectively disable the aic7870 in >the above mentioned case ... So how do I enforce an irq? Can I just specify it in the kernel config file? My irq comes out as 10 for both the onboard controller and the separate 294x. > > >Regards, STefan > >-- > Stefan Esser Internet: > Zentrum fuer Paralleles Rechnen Tel: +49 221 4706019 > Universitaet zu Koeln FAX: +49 221 4705160 > Weyertal 80 > 50931 Koeln -- Justin T. Gibbs ============================================== TCS Instructional Group - Programmer/Analyst 1 Cory | Po | Danube | Volga | Parker | Torus ==============================================