From mboxrd@z Thu Jan 1 00:00:00 1970 From: denis@eukrea.com (Denis Carikli) Date: Thu, 13 Mar 2014 10:18:31 +0100 Subject: [PATCH v3][ 5/9] ARM: dts: imx25.dtsi: Fix USB support. In-Reply-To: References: <1394618518-13803-1-git-send-email-denis@eukrea.com> <1394618518-13803-5-git-send-email-denis@eukrea.com> Message-ID: <532177E7.5070208@eukrea.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On 03/12/2014 12:08 PM, Fabio Estevam wrote: > Hi Denis, Hi, > As you add me in the From field, you also need to add: > > Signed-off-by: Fabio Estevam above your > Signed-off-by line. Thanks. >> + usbphy { >> + #address-cells = <1>; >> + #size-cells = <0>; >> + compatible = "simple-bus"; > > I made this comment earlier: why do we place usbphy0/1 under simple-bus? The official ePAPR 1.1 standard talks about the system on a chip's internal I/O bus. So, I wonder if, in general, it makes sense to group together, with a simple-bus compatible, potentially different usb phy, which are connected to potentially different usb controllers. Still if I remove it from the usbphy node, I get the following messages more than once in dmesg: > ci_hdrc ci_hdrc.0: no usb2 phy configured > platform ci_hdrc.0: Driver ci_hdrc requests probe deferral > ci_hdrc ci_hdrc.1: no usb2 phy configured > platform ci_hdrc.1: Driver ci_hdrc requests probe deferral With at the end lsusb printing nothing. > This is not documented in the the bindings. I don't think that the simple-bus has to be added to Documentation/devicetree/bindings/usb/usb-nop-xceiv.txt because that file only talks about what became usbphy's subnodes. Denis.