W dniu 31.10.2015 o 02:15, Rob Herring pisze: > On Fri, Oct 30, 2015 at 2:23 AM, Krzysztof Kozlowski > <k.kozlowski@xxxxxxxxxxx> wrote: >> On 30.10.2015 15:58, Pavel Fedin wrote: >>> Hello! >>> >>>>> Add documentation for new subnode properties, allowing bank configuration. >>>>> Based on u-boot implementation, but heavily reworked. >>>> >>>> Please, carefully look at: >>>> Documentation/devicetree/bindings/net/gpmc-eth.txt >>>> Documentation/devicetree/bindings/bus/ti-gpmc.txt >>> >>> Thank you very much. Indeed, this looks very similar. By the way, should i document smsc over sromc in the same manner, writing >>> devicetree/bindings/net/sromc-eth.txt? >>> >>> This is a short reply for now, i'll make longer one (or just a new version) after studying these existing bindings and trying to >>> apply them. >> >> Existing SROMC bindings document is small so one document for everything >> should be sufficient. This can be always split if new type of devices >> will be using SROMC (BTW, do you know of any other devices using SROMC >> on Exynos?). >> >>> >>> Pankaj: >>> >>>>> +&sromc { >>>>> + pinctrl-names = "default"; >>>>> + pinctrl-0 = <&srom_ctl>, <&srom_ebi>; >>>>> + >>>>> + ethernet@07000000 { >>>>> + compatible = "smsc,lan9115"; >>>>> + reg = <0x07000000 0x10000>; >>>>> + phy-mode = "mii"; >>>>> + interrupt-parent = <&gpx0>; >>>>> + interrupts = <5 8>; >>>>> + reg-io-width = <2>; >>>>> + smsc,irq-push-pull; >>>>> + smsc,force-internal-phy; >>>>> + >>>>> + samsung,srom-bank = <3>; >>>>> + samsung,srom-data-width = <2>; >>>>> + samsung,srom-timing = <1 9 12 1 9 1 1>; >>>> >>>> I think this is not correct. We can't change binding of "smsc,lan9115" >>>> which is already documented here [1]. These samsung specific srom >>>> properties should be in srom node or its subnode, but not in this way. >>> >>> So, if you look at gpmc-eth.txt, you'll see that this approach is perfectly valid (this is a reply to another msg, just don't want >>> to post one more single-line reply). >> >> >> Yes, the binding of smsc,lan9115 is not changed. >> >> Putting srom properties in separate bank node would be good also but >> then some mapping (connection) between ethernet and bank should be added >> probably... > > Is this to get some data needed by the ethernet from the ROM? We have > the nvmem binding to support that. No, this does not act like nvram driver. It is rather a memory controller. The same as existing TI GPMC: Documentation/devicetree/bindings/net/gpmc-eth.txt although configuration is different so gpmc driver cannot be reused. Best regards, Krzysztof -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html