19.10.2012 09:57, Sylvain Munaut пишет:
He's talking about this structure :
struct mframe_sched_item { const struct tdma_sched_item *sched_set; uint16_t modulo; uint16_t frame_nr; uint16_t flags; };
Yes, exactly. Pardon for being unclear.
However I don't see the gain of space as really that much of an advantage vs the potential time lost looking for bugs when playing with non-standard multiframes.
I agree. Btw, is this non-standard multiframes experimentation available in some public repo? Would love to have a look at the code.
The fact that we are currently almost full memory is just because we still use the compal loader. We should just deprecate that and oompletely and only compile the chain loader and loader with that constrained memory layout.
Indeed - I've got error about rssi.compalram.elf file which is too big to be used without chainloading anyway.
Maybe we can make this configurable via something like ./configure --enable-compalram-rssi ./configure --enable-chainload-rssi etc.