Hi Pavel,
Hi!
/home/pavel/wagabuibui/u-boot/board/ebv/socfpga/Makefile: No such file or directory make[2]: *** No rule to make target `/home/pavel/wagabuibui/u-boot/board/ebv/socfpga/Makefile'. Stop.
I feared as much, so thats why I asked ;)
...and I don't think we want to do board/{altera,ebv} symlink. Are there any other options? Or is "altera" in the boards file simply acceptable?
This is a problem that will turn up in the future even more, so I propose to solve it correctly now. It will not be long before we want
Well, OTOH it is orthogonal problem to the "board name is shared between socrates and altera" and "config is shared between altera and virtual target". And this patch is going to go stale rather quickly.
I admit, I do not understand that fully.
to have our own configuration for our MCV module and this will certainly be sold by DENX. I think we need an infrastructure to allow for boards sold by arbitrary manufacturers all using the Altera chip.
The situation as such is not uncommon, so maybe you can follow examples from different CPUs? I.e. how is the imx6 handled on the different base boards?
The examples I seen were different: there different board vendors actually needed different code.
AFAICT, one solution would be to put "-" in that column, and do "git mv board/altera/ board/socfpga/".
Putting "-" in the vendor column just doesn't feel right. How about using a minimal board C file for socrates under ebv/socrates that only implements checkboard and shares the rest?
But if we decide to go that way, it should really be separate patch.
I still like to see a solution that scales to things we already know will happen ;) Looking at the original patch, with this in mind even the #define ALTERA_BOARD_NAME doesn't look right any longer.
Thanks Detlev