And I got a reasonable amount of sleep last night.
Maybe I can address the ToasterCo backlog?
But the workers are here, about to commence demolition of the old tile floor in the kitchen. They brought a tool: an electric jackhammer, with chisel bit, mounted on a wheeled carriage. The noise level may become rather distracting.
Hm. Perhaps I should fetch my Bluetooth earmuffs/headphones from the barn.
Update: Kitchen activity is very noisy indeed, with occasional quiet spells. I think all the noisy will be done with soon, but today hasn't been terribly productive.
Some progress on ToasterCo, though. I think a Plan is taking shape for dealing with the doggone comm controller chip - and it involves a sort of specialized I/O processor to offload most of the activity from the CPU. Should be enough space in the FPGA to do that. I'm running into annoyances with the controller chip; it has three interrupt outputs, but there seems to be no way to decouple the two aux IRQ pins (RX and TX) from the single all-things-IRQ pin. But... there's a 4-byte interrupt status register, including a 7-bit field that reflects the highest-priority pending action, so that's maybe OK.
And maybe, as I develop a better understanding of how this all works, I'll decide that the AVR core can handle the load after all, and the I/O processor won't be needed. (Less fun, less revenue, but gits 'er done quicker.) The example code I was looking at may turn out to have been wildly overcomplicated.
Comments