topic S32G2 bsp 43.0 eMMC boot in S32G

S32G2 bsp 43.0 eMMC boot

chansookang — Wed, 16 Jul 2025 00:07:33 GMT

Hello, i tried yocto linux bsp 43.0 build and get these images

i successfully boot by using SD card but i want to boot in eMMC.

By using flashtool, i want to know which image file i should use and how to write image on emmc. (my board is S32G274ardb2 rev2.1)

i tried this but i dont have that fip.s32 file

Re: S32G2 bsp 43.0 eMMC boot

chenyin_h — Wed, 16 Jul 2025 06:31:01 GMT

Hello, @chansookang

Thanks for your post.

From BSP43, the original fip.s32 is divided into two images, one is the bl2_w_dtb.s32, the other one is the fip.bin.

I suggest referencing the method 2 in chapter 4.1.3 of BSP43.0 UM for details of updating the images for eMMC.

Chenyin

Re: S32G2 bsp 43.0 eMMC boot

chansookang — Wed, 16 Jul 2025 07:15:38 GMT

Hello thanks for your reply

i already tried by looking at 4.1.3 method 2, successfully write in eMMC from SD card

but i want to setup by using flash tool.

i wrote bl2_w_dtb.s32 at 0x0 and fip.bin at 0x80000/0x40000 with flash tool but error message occurs below

NOTICE: Reset status: Power-On Reset
PANIC at PC : 0x000000000000003fition table

Re: S32G2 bsp 43.0 eMMC boot

chenyin_h — Thu, 17 Jul 2025 03:15:07 GMT

Hello, @chansookang

Thanks for your reply.

From my experience, I suggest directly following the method 2 for flashing your eMMC, I understand that you have done it.

For using Flash tool, it is a little complicated. Firstly, change the default bl2_w_dtb.s32-sdcard, derived only the data from 0x1000 to the end and save it, say derived.bl2_w_dtb.s32-sdcard.

Then use the flash tool to flash the newly generated image with the offset 0x1000 to have a try.

Chenyin

Re: S32G2 bsp 43.0 eMMC boot

chansookang — Tue, 22 Jul 2025 06:30:28 GMT

Thank you !

Re: S32G2 bsp 43.0 eMMC boot

A_VIJAYA7 — Thu, 18 Sep 2025 13:31:17 GMT

Hi Chenyin,
We are planning to proceed with Method 1, which involves eMMC flashing without using QuadSPI.
Could you please share the documentation or guidance on how NXP validated this method?

Regards,
Vijaya