I should mention that I'm very open to suggestions Andy might have about
any of this. The new approach I'm currently working on with tagged pair
pointers requires changes to both the VM and the compiler, and I'm not
confident that I've made the best choices there. I've made some
preliminary choices for now in order to get something working ASAP, but
I fully expect that the final version will look different based on input
from Andy and others. I also don't know whether the new approach (with
tagged pair pointers) will be preferable to the earlier one (with most
tc7 tags changed to tc11). My goal is to present a couple of working
alternatives, and then we can decide among them.
For now, I went ahead and pushed my new (not yet working) branch
'wip-new-tagging-bis-broken', in case you want to see the details of
this new approach-in-progress. The new branch doesn't yet include the
iflo- or fixrat-enabling patches, which I'll add later.
Regards,
Mark
