19:56 <DavidHeidelberg> digetx Hello, what approach would you prefer? I plan to convert some Documentation TXT files to the YAML (since I started playing with it and don't want to waste all the yaml suffering). Should I send txt -> yaml converts to your tree and you'll send it together w/ other stuff to upstream or send it to you and then send it to upstream myself?

20:54 <DavidHeidelberg> why https://github.com/grate-driver/linux/blob/master/arch/arm/boot/dts/tegra20.dtsi#L863 has phy_type, since phy_type should be only defined in https://github.com/grate-driver/linux/blob/master/arch/arm/boot/dts/tegra20.dtsi#L880

20:55 <DavidHeidelberg> I'm trying convert bindings yaml and it screams at me ;-)

20:56 <DavidHeidelberg> also `nvidia,has-legacy-mode` is defined in tegra20-usb-phy, but not in tegra ehci

21:13 <jenneron[m]> btw tegra30 dsi is present in DT, but not in driver as well

21:14 <DavidHeidelberg> yeah, I did fix

21:14 <DavidHeidelberg> https://github.com/okias/linux/commit/d90b70cfca80c75d9fc2a9f5ba71ba13329b56fb

21:15 <DavidHeidelberg> could look something like this

21:21 <DavidHeidelberg> since phy has all required atributes

23:34 <DavidHeidelberg> so much rewrite and so little visible difference.. meh. but at least I noticed that some attributes are duplicated and DTS and most likely, when someone writes new DTS for device they can mistake it and fill this stuff into different node..

23:34 <DavidHeidelberg> the patch can be partly wrong https://github.com/okias/linux/commit/cafc4c188077069626e0d9397b97efdf00659969

23:36 <DavidHeidelberg> I guess the dr_mode part is wrong and dr_mode should probably go into tegra-ehci binding