Thank you lobo! You helped me to find my oversight

In facts there is a difference between the two installation regarding the size of the boot partition, but coarsely checked I didn't find differences.
SD card and eMMC have no big differences. The only problem is the first 4Mbytes which contains the functionality to let access the eMMC as UMS. When that is damaged, then here comes the problem that we're talking about.
(11-03-2018, 12:41 PM)Radioman Wrote: How do I find what the "kernel_name" is?
The kernel_name is the definition that is given to a device by the kernel. So for pluggable devices you may find a notification on the bottom of kernel debug message, just do after plugged in
So for certain distro, such device will be named sd
XY, which
X stands for the device number and
Y stands for the partition number. See
this
I didn't state any of those names, because it vary from a PC and the TB itself, which they are named mmcblk
Np
P, where
N is the device number and
P the partition number. Also because using
dd for windows the naming convention is very much different.
There's also a trick to write the u-boot file. We must write it on no partition, but at the device beginning, so there won't be the partition definition to issue to shell command. In particular the file is place at 64 sectors away from the begin.
These distinctions are necessary, because depend where one will do the operations. If doing on a PC will be mostly as per the former statement, whereas doing directly on the tinker board, would be the latter.