First, I like the documentation and the demo project for the grisp2 very much!
I would like to install an erlang application on the eMMC memory of a grisp2 and could not find
documentation about it.
My current approach would be to deploy to an SDcard, then boot into a busybox shell and copy
the release over to the internal memory. Unfortunately, the cp command of busybox doesn’t do recursive.
Hm, I might write a function in my application, which copies itself to the eMMC…
According to the specs on kickstarter, the eMMC should be 4GB, plenty enough for my projects.
Bonus question: the file structure of the internal memory is different from a deployment to the SDcard.
How should I transform an SDcard deployment so that it runs from the internal memory?
barebox@GRiSP2:/mnt ls mmc1.0
. … DEMO demo.zImage erl_inetrc
grisp.ini grisp_demo loader
If you are talking about setting up the eMMC manually, I would prepare the image as described in https://github.com/grisp/grisp2-rtems-toolchain/blob/master/README.md eMMC Image section, put it on the SD card and copy it to the device with the cp command in the boot loader console. Regarding the differences between the SD and eMMC, on the eMMC there is supposed to be two system partitions for staged software update. In the readme only the fist one is actually defined, but the space for the second system partition should be reserved. I think that if you look at the image in the toolchain repo, you would see the partitions are defined even though they are truncated. This is so it is smaller but you can still start a RTEMS shell and format it later on.
Thanks a lot, the link looks very good, haven’t read that before.
The background is: I will explore grisp2 as platform for our projects, and from a previous
project I know that our customers are hesitant to use SD cards (That ruled out a raspberry,
but enabled a beaglebone black).