Hi DJ,
This is OK, but note that it prevents some operations like:
__far int i;
foo()
{
i ++;
}
from being implemented with a minimum set of opcodes. This might be
particularly troublesome for volatile far things.
Right - it is something I will have to look into.
Cheers
Nick
This is OK, but note that it prevents some operations like:
__far int i;
foo()
{
i ++;
}
from being implemented with a minimum set of opcodes. This might be
particularly troublesome for volatile far things.
Hi DJ,
It turns out that the optimization in rl78_force_nonfar_3 to allow
some special cases to be kept in far pointers does not always work.
The test case included with this patch will trigger ICEs if the
optimization is allowed to persist.
So, may I check this patch in please ?
Cheer
