From owner-freebsd-drivers@FreeBSD.ORG Mon Feb 13 17:15:43 2012 Return-Path: Delivered-To: freebsd-drivers@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id E17E5106566B for ; Mon, 13 Feb 2012 17:15:43 +0000 (UTC) (envelope-from jhb@freebsd.org) Received: from cyrus.watson.org (cyrus.watson.org [65.122.17.42]) by mx1.freebsd.org (Postfix) with ESMTP id B9B9C8FC15 for ; Mon, 13 Feb 2012 17:15:43 +0000 (UTC) Received: from bigwig.baldwin.cx (bigwig.baldwin.cx [96.47.65.170]) by cyrus.watson.org (Postfix) with ESMTPSA id 729EF46B0A; Mon, 13 Feb 2012 12:15:43 -0500 (EST) Received: from jhbbsd.localnet (unknown [209.249.190.124]) by bigwig.baldwin.cx (Postfix) with ESMTPSA id 089DAB922; Mon, 13 Feb 2012 12:15:43 -0500 (EST) From: John Baldwin To: freebsd-drivers@freebsd.org Date: Mon, 13 Feb 2012 11:56:25 -0500 User-Agent: KMail/1.13.5 (FreeBSD/8.2-CBSD-20110714-p10; KDE/4.5.5; amd64; ; ) References: In-Reply-To: MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Message-Id: <201202131156.25554.jhb@freebsd.org> X-Greylist: Sender succeeded SMTP AUTH, not delayed by milter-greylist-4.2.7 (bigwig.baldwin.cx); Mon, 13 Feb 2012 12:15:43 -0500 (EST) Cc: Bret Ketchum Subject: Re: MSI-X support X-BeenThere: freebsd-drivers@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Writing device drivers for FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 13 Feb 2012 17:15:44 -0000 On Friday, February 10, 2012 8:17:49 am Bret Ketchum wrote: > I've an HBA which supports MSI-X but uses the same BAR for its memory > mapped registers, the MSI-X table and PBA. They're just at different > offsets, similar to a QLogic 8Gb ISP2532. I've perused the source and it > looks like 8.1 doesn't support this flavor of MSI-X configuration out of > the box. Posting here to see if I'm just blind. 8.1 should support this just fine. All MSI-X requires is that the driver have allocated and activated the relevant BAR before calling pci_alloc_msix(). This makes the PCI bus code simpler for exactly the case you mention since the bus doesn't have to deal with trying to allocate the same BAR twice, etc. -- John Baldwin