search

PabloPL / linux

Linux kernel source tree
Other
17 stars 0 forks source link

Submit minimal galaxys and galaxys4g dts files #4

Closed PabloPL closed 6 years ago

PabloPL commented 6 years ago

Got initial feedback. Preparing fixed version at for-upstream/aries-minimal-v2 branch.

PabloPL commented 6 years ago

@xc-racer99 Did You saw anywhere configuration for max8998? I can't find some max8998 regulators in aries sources.

xc-racer99 commented 6 years ago

As in for LDO6? Have a look at https://github.com/xc-racer99/android_kernel_samsung_aries/blob/aosp-7.1/arch/arm/mach-s5pv210/mach-aries.c#L564 for how its configured for the T959P (used by modem)

ESAFEOUT2, probably has to be deduced from drivers/power/s5pc110_battery.c

LDO10 I have no idea what it should default to... I'll try and read the registers later today and determine what they are.

PabloPL commented 6 years ago

Looking at schematic, it looks like ESAFEOUT1 and ESAFEOUT2 are 5v. LDO6 is not connected to anything (on i9000) and LDO10 is VPLL 1.2V.

@xc-racer99 I've updated that in for-upstream/aries-minimal-v2 branch. Could You check if it's booting on fascinate?

xc-racer99 commented 6 years ago

Nope, not booting for me. I'll try and find out which of the changes is causing the issue.

PabloPL commented 6 years ago

I've removed patches for broken bootloader. Without that, it's not working for me even if using uboot. I would try with this at first. And of course patch for bootloader logs.

Just cherry-pick e22177776f1eb2c98fe5fc1bf68921d5de99957a and edaf7ab6bc034a28febe8b5f847a04bf428b3682.

Edit 1: There is possibility that i'll have unbrickable mod applied to my device.

Edit 2: Sigh, there is something wrong with safeout1, need to do more testing.

xc-racer99 commented 6 years ago

Yeah, its edaf7ab - I'm guessing it's because UART1 isn't initialized by u-boot while UART2 is. I'll do some testing for that.

Can at least verify that this boots now

Edit: For clarification, the first commit (bootloader bug) isn't need to boot from u-boot

PabloPL commented 6 years ago

For me it wasn't able to detect sdcard. Most important difference between v1 and v2 is addition of max8998-pmic regulators and set irq type. Need to do more testing for this. It's sad, that stock kernel don't have settings for all regulators.

xc-racer99 commented 6 years ago

Yep, SD card wasn't detected for me either.

xc-racer99 commented 6 years ago

Ok, based on stock kernel, ENVICHG is always enabled (dumped the registers and worked backwords). LDO10 is inded 1.2V and always enabled

As well, sdcard worked just fine when voltages were removed from the ESAFEOUT regulators... Note that since removing the console parameters from bootparams, I no longer get a serial console - it's just a framebuffer console.

Edit: Doesn't look like voltages are necesary for ESAFEOUT as the only ops for them (in drivers/regulator/max8998.c) are is_enabled, enable, and disable.

xc-racer99 commented 6 years ago

Alright, https://github.com/xc-racer99/linux/commit/d7447026780ed9394f3b25f0a3616b897b770366 is working for me for regulators. https://github.com/xc-racer99/linux/commit/e893b1d233cd799ce453e644cbaee04808fce92f is so that there's still a console on UART2

Feel free to adjust/squish the commits into yours as desired.

PabloPL commented 6 years ago

@xc-racer99 Ok thanks, i'll do it (was bussy for few days) and submit patches again.

PS: Got unbrickable mod applied on my device.

PabloPL commented 6 years ago

Initial patches has been accepted and will be merged soon into mainline. This issue can be closed.

PS: @xc-racer99 Till it get merged into linus tree (aries-minimal-v4 contains only fixed dts patches, without those accepted in v3 version), it's better use aries-minimal-v3 as base branch for all other.

xc-racer99 commented 6 years ago

Great, thanks. I'll work on rebasing the branches I created to v3/v4. I've also gotten the gp2a light/proximity sensor working (with adding the ADC support for the light part) plus the smb380/bma150 accelerometer which I'll push. There's no upstream driver for the magnetic sensor (yas529).

PabloPL commented 6 years ago

Ok, i'm working on submiting changes for touchscreen and fm radio driver (si470x).

PabloPL commented 6 years ago

@xc-racer99 Maybe we could skip for now yas529 sensor and focus now for example on LCD/audio (it would make this device more usefull) ?

xc-racer99 commented 6 years ago

I've been working on audio as well, I have a building driver, but whenever I try to play something, it just hangs. I think there's a clock that I'm not setting up properly, but I'm not sure...

MYEUHD commented 5 years ago

The initial patches have landed in Linus' tree!