From mboxrd@z Thu Jan 1 00:00:00 1970 From: arnd@arndb.de (Arnd Bergmann) Date: Mon, 25 Jun 2012 16:26:08 +0000 Subject: [RFC PATCH v2 1/2] drivers: misc: omap: add a new driver for ocp2scp In-Reply-To: <20120625122449.GA30463@arwen.pp.htv.fi> References: <1340370942-23201-1-git-send-email-kishon@ti.com> <20120625122449.GA30463@arwen.pp.htv.fi> Message-ID: <201206251626.09666.arnd@arndb.de> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Monday 25 June 2012, Felipe Balbi wrote: > > > Can't this live where the scp drivers live? Actually, where is that at? > > > Do we have scp drivers? > > AFAIK, there isn't any driver for scp. But we have a driver for ocp > > and it is present at arch/arm/mach-omap2/omap_l3_noc.c > > I don't think this deserves a directory of its own. Maybe > drivers/platform/arm/omap/ ?? the l3_noc is an OMAP-specific > interconnect and the SCP bus is also an OMAP-specific bus. I don't know > of any other arch/soc who uses the same interconnect IP as OMAP and the > same ocp2scp bridge. That bridge was created by TI for all I know. > > Greg, would drivers/platform/arm/omap/ work for you ? We could also move > the interconnect drivers there. I really don't like the idea of introducing drivers/platform/arm/ because very little of the stuff that one would put in there are actually ARM specific. I have suggested a drivers/bus/ before and people did not see the need back then, and we agreed to continue having a directory for each bus, as we have for the big ones (pci, usb, i2c, spi, ...) and a lot of simple (amba, rapidio, bcma, ...) or obscure (tc, vlynq, nubus, ...) ones. I think we should reconsider the idea of drivers/bus/ with a file per bus in there at least for new buses, but doing a new drivers/scp/ would be ok for me if there is enough opposition against the idea of drivers/bus aggregating different buses. Arnd From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [RFC PATCH v2 1/2] drivers: misc: omap: add a new driver for ocp2scp Date: Mon, 25 Jun 2012 16:26:08 +0000 Message-ID: <201206251626.09666.arnd@arndb.de> References: <1340370942-23201-1-git-send-email-kishon@ti.com> <20120625122449.GA30463@arwen.pp.htv.fi> Mime-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-15" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20120625122449.GA30463@arwen.pp.htv.fi> Sender: linux-doc-owner@vger.kernel.org To: balbi@ti.com Cc: "ABRAHAM, KISHON VIJAY" , Greg KH , grant.likely@secretlab.ca, rob.herring@calxeda.com, rob@landley.net, linux@arm.linux.org.uk, b-cousson@ti.com, rnayak@ti.com, tony@atomide.com, devicetree-discuss@lists.ozlabs.org, linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org List-Id: devicetree@vger.kernel.org On Monday 25 June 2012, Felipe Balbi wrote: > > > Can't this live where the scp drivers live? Actually, where is that at? > > > Do we have scp drivers? > > AFAIK, there isn't any driver for scp. But we have a driver for ocp > > and it is present at arch/arm/mach-omap2/omap_l3_noc.c > > I don't think this deserves a directory of its own. Maybe > drivers/platform/arm/omap/ ?? the l3_noc is an OMAP-specific > interconnect and the SCP bus is also an OMAP-specific bus. I don't know > of any other arch/soc who uses the same interconnect IP as OMAP and the > same ocp2scp bridge. That bridge was created by TI for all I know. > > Greg, would drivers/platform/arm/omap/ work for you ? We could also move > the interconnect drivers there. I really don't like the idea of introducing drivers/platform/arm/ because very little of the stuff that one would put in there are actually ARM specific. I have suggested a drivers/bus/ before and people did not see the need back then, and we agreed to continue having a directory for each bus, as we have for the big ones (pci, usb, i2c, spi, ...) and a lot of simple (amba, rapidio, bcma, ...) or obscure (tc, vlynq, nubus, ...) ones. I think we should reconsider the idea of drivers/bus/ with a file per bus in there at least for new buses, but doing a new drivers/scp/ would be ok for me if there is enough opposition against the idea of drivers/bus aggregating different buses. Arnd