Hi Chunfeng, On Mon, Jun 25, 2018 at 8:37 AM Chunfeng Yun <chunfeng.yun@xxxxxxxxxxxx> wrote: > > On Sun, 2018-06-24 at 20:00 +0200, Martin Blumenstingl wrote: > > Hello Chunfeng, > > > > thank you for the patch! > > > > On Fri, Jun 22, 2018 at 8:33 AM Chunfeng Yun <chunfeng.yun@xxxxxxxxxxxx> wrote: > > > > > > 1. use IS_ERR() but not IS_ERR_OR_NULL() because devm_of_phy_get_by_index() > > > never return NULL value; > > ACK - good catch! > > > > > 2. devm_of_phy_get_by_index() should not fail for a valid index > > I have learned that the PHY framework can return -ENODEV if the PHY is: > > 1. supposed to be handled by the legacy USB PHY framework > > 2. the PHY node is disabled in devicetree > A little confused, we can avoid this error by not using a disabled or > "usb-nop-xceiv" phy in host node. I think that would be the mid or long term goal > If ignore error of -ENODEV, we will also skip an error case which we > don't want it happen. as far as I understand the problem is that there are .dts files out there which specify "usb-nop-xceiv" in the "phys" property, see [0] (there might be even some more out-of-tree cases) so I think the -ENODEV check can only be removed once nothing uses "usb-nop-xceiv" anymore Regards Martin [0] https://patchwork.kernel.org/patch/10160181/ -- 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
