03:33 <chewitt> etrig: same effect is no change (not working) ?

04:03 <etrig> chewitt: yes, the same failure to load u-boot on reboot with same console messages

07:59 <narmstrong> hmmmmmmmmmmmmmmmmmmmmmm

07:59 <jbrunet> etrig: the OC2 has 3 regulators when it comes to SDcard. 2 working in tandem (U12 and U13) to provide 3.3v or 1.8v for the signals and also a 3.3v (U16) to provide power supply to the card

08:00 <jbrunet> If you can, check the output voltage of these regulators when stuck, a basic voltmeter should be enough

08:01 <jbrunet> if you don't see 3.3v on both the signal and supply, it's not good

08:21 <jbrunet> also when signals are transitioning from 1.8v back to 3.3v (switching from U13 to U13), the SDcard must be power cycled (turning off then on U16). a scope will be needed to check that)

16:50 <lvrp16> narmstrong: thanks for the patch!

17:37 <etrig> jbrunet: looks like it's still 1.8v on U12/U13 after a reboot

17:38 <etrig> triple checking that the patch was properly applied

17:51 <jbrunet> the problem is that those regulators are GPIO controlled and custom to each

17:51 <jbrunet> board

17:52 <jbrunet> but the ROM code (the piece of SW in the SoC looking for a bootloader) is the same on every SoC ...

17:53 <jbrunet> So if the GPIOs are not reset on a warn reset, the regulators maybe in a state that does not allow the ROM code to properly the SDCard

17:58 <etrig> I just remembered that I still have a couple regulator-always-on entries that I added for &tf_io and &tflash_vdd

17:58 <etrig> maybe I should remove those

18:03 <etrig> looks like I already removed that

18:23 <etrig> on a cold boot I see 3.3v on U12/U13 then it lowers to 1.8v while linux boots, as expected, but then stays 1.8v after a reboot

18:30 <etrig> maybe a weird question, but can one actually use $(git apply) to apply a patch outside of a git repository?

18:30 <xdarklight> you probably want "git am"

18:33 <etrig> I ask because I'm using archlinuxarm and just added the mmc_sd_shutdown patch to the linux-aarch64 PKGBUILD

18:33 <etrig> https://github.com/archlinuxarm/PKGBUILDs/blob/master/core/linux-aarch64/PKGBUILD

18:34 <etrig> now I'm not sure if those other patches are really being applied either

18:34 <xdarklight> ah hang on, a patch outside of a git repository

18:36 <xdarklight> I'm not sure about that either

18:37 <etrig> % git apply -v ../0099-arm64-s905-uhs-1-sd-card.patch

18:37 <etrig> Skipped patch 'drivers/mmc/core/sd.c'.

19:33 <jbrunet> etrig: the u12/u13 is always by design. They are both fed VDDIO_AO3v3 which is always on, then one regulator is on when the other is off. Which one is controlled by the GPIO. the output cannot be off

19:34 <jbrunet> u16 MUST not be always on. It used to power cycle the sdcard. This is critical to get out of the UHS modes