sunxi-6.6: NextThing C.H.I.P (PocketChip) community support

[TheSnowfield]

Description

Community support for PocketChip SD (MMC2 breakout)

How Has This Been Tested?

./compile.sh BOARD=pocketchip-sd BRANCH=current
Burn the image into the SD card, normally boot.

Work And Tested

• NAND device MMC2 breakout
• RGB display
• TCA8418 keyboard
• WLAN

Not Tested

• Touch Screen

More

I need a few days to wait for the new MMC2 breakout board because I used the wrong SD slot package of the PCB, causing the SD card install direction to conflict with an inductor which made it hard to install the SD card.
The current test board looks like this:

It uses Allwinner R8(A13) pin mux, to convert NAND to MMC2, avoiding the NAND boot problem.

[The-going]

@TheSnowfield Please add exactly the same changes to the series.armbian file.
Can you show the difference between the new boot script file and the one that was used as the basis?
As output the diff -u file1 file2 command

[TheSnowfield]

@TheSnowfield Please add exactly the same changes to the series.armbian file. Can you show the difference between the new boot script file and the one that was used as the basis? As output the diff -u file1 file2 command

Sure, I only added two lines that set the default boot from the MMC2. If we have a better solution...

--- boot-sunxi.cmd      2025-01-04 02:52:47.576181416 +0800
+++ boot-sunxi-pocketchip-sd.cmd        2024-12-31 19:50:17.685512643 +0800
@@ -17,6 +17,10 @@
 setenv rootdev "/dev/mmcblk${devnum}p1"
 setenv earlycon "off"
 
+# Pocketchip SD only boot from mmc2
+setenv devtype "mmc"
+setenv mmc_bootdev "1"
+
 # Remember the default u-boot fdtfile
 setenv deffdt_file ${fdtfile}

[The-going]

It seems to me that these changes that you have made are enough for the device to be downloadable.

 CONFIG_SYS_I2C_MVTWSI=y
 CONFIG_SYS_I2C_SLAVE=0x7f
 CONFIG_SYS_I2C_SPEED=400000
-# CONFIG_MMC is not set
+CONFIG_MMC=y
+CONFIG_MMC_SUNXI_SLOT_EXTRA=2
 CONFIG_AXP_ALDO3_VOLT=3300
 CONFIG_AXP_ALDO4_VOLT=3300

But why doesn't this happen in your case with a standard script?

[TheSnowfield]

It seems to me that these changes that you have made are enough for the device to be downloadable.

 CONFIG_SYS_I2C_MVTWSI=y
 CONFIG_SYS_I2C_SLAVE=0x7f
 CONFIG_SYS_I2C_SPEED=400000
-# CONFIG_MMC is not set
+CONFIG_MMC=y
+CONFIG_MMC_SUNXI_SLOT_EXTRA=2
 CONFIG_AXP_ALDO3_VOLT=3300
 CONFIG_AXP_ALDO4_VOLT=3300

But why doesn't this happen in your case with a standard script?

May I miss something? From my memory, I was always testing the u-boot via sunxi FEL rather than writing to the SD card... I know is not necessary to modify the boot script, but I'm confused either.
But anyway, I will figure it out later. I will test the boot with the standard boot script again...

[The-going]

I will test the boot with the standard boot script again...

Try adding the debugging code directly to the boot script and compiling it to understand which variables u-boot has already set in the environment.
As echo "devnum=$devnum"

If the standard script crashes for some loading scenario, we have to fix it.

[TheSnowfield]

I will test the boot with the standard boot script again...

Try adding the debugging code directly to the boot script and compiling it to understand which variables u-boot has already set in the environment. As echo "devnum=$devnum"

If the standard script crashes for some loading scenario, we have to fix it.

Okay, I confirmed that was my mistake, I will remove the boot script changes.
Thanks for your help!

[igorpecovnik]

There was a huge merge on kernel configs, so more adjustments are needed. Sorry.

[TheSnowfield]

There was a huge merge on kernel configs, so more adjustments are needed. Sorry.

Woa.. But never mind, Im gonna to rebase these changes..