From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from moutng.kundenserver.de (moutng.kundenserver.de [212.227.126.183]) by ozlabs.org (Postfix) with ESMTP id 5FD92DE0C2 for ; Tue, 22 Apr 2008 15:17:42 +1000 (EST) From: Stefan Roese To: benh@kernel.crashing.org Subject: Re: [PATCH v2] [POWERPC] 4xx: Add endpoint support to 4xx PCIe driver Date: Tue, 22 Apr 2008 07:17:30 +0200 References: <1208789670-28909-1-git-send-email-sr@denx.de> <1208812591.9640.45.camel@pasglop> In-Reply-To: <1208812591.9640.45.camel@pasglop> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-15" Message-Id: <200804220717.30546.sr@denx.de> Cc: linuxppc-dev@ozlabs.org List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Monday 21 April 2008, Benjamin Herrenschmidt wrote: > On Mon, 2008-04-21 at 16:54 +0200, Stefan Roese wrote: > > -------------- > > - As suggested by Benjamin Herrenschmidt, don't determine endpoint mode > > by looking at the already configured PCIe port state, but evaluate > > the device_type property instead. This makes it possible for boards > > without a previously run PCIe configuration (e.g. U-Boot) to configure > > a port as endpoint as well. > > > > - Don't map the big external config space upon endpoint configuration. > > > > - Introduce "vendor-id" and "device-id" properties to be written to > > the host bridge. > > and class-code too ? Sure, could be done. But this is not really endpoint/root-complex related. So I suggest to add this later with an additional patch. > Appart from that, I'll have a quick look at the patch later today but at > first it looks good. Thanks. Best regards, Stefan