I agree that there's no real reason to muck with those partitions on purpose, but it would be nice to know that if I did accidentally screw it up, I could recover without losing a week or two of development time returning the board. My basic problem is my rootfs is likely to exceed 2G (I thought the board shipped with 4G).
The board appears to be equipped with 2G (rather than the advertised 4G). The part number of the eMMC is "SDIN8DE2-16G". I couldn't find a datasheet for that exact part number, but since the output of "df -h /" is:
Filesystem Size Used Avail Use% Mounted on
/dev/mmcblk0p25 2.0G 1.6G 387M 81% /
I then assume that the "16G" in the part number implies bits; meaning 2G bytes.
I am likely to need to install packages that will cause this 2G to overflow during development. I was hoping to just boot off uSD to support a larger rootfs. If I can't do that, then I need an alternative. Here were some of my thoughts:
1. change the kernel command line to use "rootfs=/dev/mmcblk1xxx"
2. use FS overlay to just expand rootfs to the uSD card (don't know much about this yet)
3. remount things somehow at startup…
To experiment here, I am more likely to accidentally screw up the partitions.
See my problem? Any thoughts?
Thanks,
Ed