We should not use hardcoded partition numbers

[lmbarros]

Some of our images have a different partition layout, so those hardcoded p1 and p2 in the script will fail with, e.g., RockPi flasher images.

Ideally, we should be using https://github.com/balena-io-modules/balena-image-fs to get the embedded image (instead of implementing a different solution here).

[jellyfish-bot]

[lmbarros] This issue has attached support thread https://jel.ly.fish/2c2897d3-0659-4045-b8e4-6c4e8d1c82f1

[20k-ultra]

Is this issue related to #11 ?

[lmbarros]

Hi @20k-ultra. I am not super familiar with these tools, but yeah, sure looks like it is!

[jellyfish-bot]

[pdcastro] This issue has attached support thread https://jel.ly.fish/0cd5268c-16dc-4c66-a2d8-560a4790dbd1

[pdcastro]

Workaround for the Rock Pi 4B balenaOS flasher image partition layout:

After cloning this repo, edit the balena-image-flasher-unwrap script and replace all occurrences of "p1" and "p2" with "p4" and "p5". There should be 4 lines to be changed, as per following diff:

-mount "${balena_image_flasher_loop_dev}p1" "$balena_image_flasher_boot_mnt" > /dev/null 2>&1
+mount "${balena_image_flasher_loop_dev}p4" "$balena_image_flasher_boot_mnt" > /dev/null 2>&1

-mount "${balena_image_flasher_loop_dev}p2" "$balena_image_flasher_root_mnt" > /dev/null 2>&1
+mount "${balena_image_flasher_loop_dev}p5" "$balena_image_flasher_root_mnt" > /dev/null 2>&1

-mount "${balena_image_loop_dev}p1" "$balena_image_boot_mnt"
+mount "${balena_image_loop_dev}p4" "$balena_image_boot_mnt"

-mount "${balena_image_loop_dev}p2" "$balena_image_root_mnt"
+mount "${balena_image_loop_dev}p5" "$balena_image_root_mnt"