02:52 <chewitt> benoitm974 "diff -Naurp original.dts changed.dts" would be easier to read, or use "git diff path/to/file"

02:53 <chewitt> benoitm974 could you share a real name and email address for "Tested-by" on the H96 dts

06:55 <benoitm974> Chewitt: I can't just did between 2 file since the dts were all reworked completely between 5.3 and 5.11... the Ethernet definition used to be in the gt12a file initially and got moved to the g12 common file to share between all g12 a/b model.

08:49 <benoitm974> OK so I've talked it down to those two lines :

08:49 <benoitm974> https://www.irccloud.com/pastebin/vn5J28Qs/

08:50 <benoitm974> This prevent the internal MAC IRQ / definition on the SM1 H9 X3 box

08:52 <benoitm974> https://github.com/torvalds/linux/commit/f3362f0c18174a1f334a419ab7d567a36bd1b3f3

10:04 <narmstrong> hmm, yeah, it may reset the MDIO multiplexer driver

14:21 <benoitm974> narmstrong: I'm wondering if when removed it work because the vendor u-boot already initialized it and it wouldn't work with mainline u-boot ? But anyway currently it's broken even with this morning 5.11-rc4, it does work with external phy on a SM1 with 1GB external but does not with internal PHY SM1.

14:22 <benoitm974> I've also check the DWMAC drive, there were limited changes around this, so it seems really to only due to DTS definition, I've also seen a comment from the mailing list someone asking why we should reset the phy since they were using the dts without reset for month with no issues...

14:25 <benoitm974> Chewitt for name/email you can reference the same as this dtb: https://github.com/torvalds/linux/blob/master/arch/arm/boot/dts/armada-xp-lenovo-ix4-300d.dts