04:28 <chewitt> @benoitm974 I updated my 5.11.y patchset with https://github.com/superna9999/linux/commit/d858863c8a1f0634c513b868357644942366d57f which restricts the phase change to g12/sm1 devices and dropped the quirk from the H96 dts

04:52 <chewitt> @pmjdebruijn I've not seen low power warnings .. old power cell perhaps?

13:06 <benoitm974> chewitt: thanks for the update, for the CMD53 patch revert any test I could do to confirm and who shoudl I communicate this regression if any ?

13:07 <chewitt> I reverted the scatter patch from my branch

13:07 <chewitt> best to give feedback on the mailing list

13:08 <chewitt> if you (like me) are not subscribed, you can download the "message" from patchwork, import it to a mail client, then reply to the message

13:08 <chewitt> then all the references are intact and your comments will show up as a proper reply.. not just a new message

13:09 <chewitt> hopefully that makes sense?

13:10 <benoitm974> Ok one other thing I've tried to go give it a try with the H9 X3 chinese box, the box seems very similar to X-air-2G, i've manager to get the wifi/bluetooth to work, yet the internal phy for ethernet is not working, the definition of internal phy is from G12 common, has anyone manage to have a internal phy ethernet box working with the mainline dtb ?

13:29 <benoitm974> For the phase discussion it's interesting to see Baylibre already had a clock phase tuning in a 4.13 drivers that got droped over time (: https://github.com/kinmax/linux-4.13.9/blob/d38cadf9678859e28f8e4b9732d275b0dbab112a/drivers/mmc/host/meson-gx-mmc.c

22:13 <benoitm974> For reference I extracted the dts from android of the cheap 2Gb H9 X3 tv box in case here: https://pastebin.com/kQnJpB4m