#avr Logs

May 08 2020

12:53 AM rue_mohr: I said microchip would screw it up
06:37 AM Smidge204: The virtual port thing is actually not that bad, it's the inconsistency that's a facepalmer :/
06:41 AM Smidge204: I'm thinking I can get super cheeky with my pin usage; If I implement my own SPI I think I can get away with some shit
06:42 AM LeoNerd: Virtual port?
06:42 AM Smidge204: Yeah
06:42 AM LeoNerd: Oh those bits. Yeah, they're mostly for legacy-ish support.. but bitwise instructions too
06:43 AM LeoNerd: You don't really need the bitwise instructions though since the various *SET and *CLR registers are easy enough for even multiple bits at once
06:43 AM Smidge204: The PORT registers are not inside bit-addressible memory space, so they added VPORT registers that are which map to them
06:44 AM LeoNerd: Yeah, those ones
06:44 AM cehteh: well in one way thats cool, but in another way microchip remove the simplicity from the the architecture, which was one of its main advantages
06:44 AM LeoNerd: I don't use them much, but I did make use of it in my WS2812 driver, just to make it easier to port from the older chips
06:45 AM Smidge204: mm
06:45 AM LeoNerd: I don't particularly mind them. They can be handy but easily ignored if you don't need them
06:45 AM LeoNerd: It's not like they get in the way of anything,... as compared some of the "missing features" of the new chips
06:46 AM LeoNerd: Most notably: lack of fCPU from external crystal
06:46 AM Smidge204: I don't mind them either, I just wonder why they didn't extend that to the rest of the port config registers
06:46 AM LeoNerd: Actually so far I think that's the only thing I really miss
06:46 AM Smidge204: (I can guess why, of course)
06:46 AM LeoNerd: Because the rest of the port config registers don't really need fast bit slicing
06:47 AM LeoNerd: They're really just boot-time setup
06:48 AM Smidge204: Anyway here me out; I noticed on the device I yoinked the SPI flash chip off of that they combined MOSI and MISO to a single pin and a 100Ohm resistor. I also think I can cheese it by combining the chip select pin for the SPI flash and the data pin for the I2C
06:49 AM LeoNerd: Ahyes, half-duplex SPI. I've got a colour OLED display which uses similar
06:49 AM Smidge204: As long as the I2C clock line is stable, flipping the data line won't trigger a start condition, and as long as there's no SPI clock or data flipping the CS line shouldn't have any ill effect either
10:06 AM Dustpuppy: hi
10:11 AM nohit: hi
10:23 AM matrixstorm: hi
10:56 AM rue_mohr: low
10:57 AM matrixstorm: ?
11:01 AM nohit: low
11:10 AM gaston53: Is there anybody here working on SAMA5D27 microchip board ?
11:22 AM matrixstorm: well gaston53 - this is the avr channel
11:22 AM matrixstorm: better luck @arm
11:22 AM matrixstorm: ?
11:24 AM gaston53: matrixstorm i will check arm channel
01:29 PM Emil: matrixstorm: well, this is an avr channel, I'd say The channel has moved elsewhere
02:01 PM Smidge204: You would, filthy splitters :)
02:03 PM Emil: Smidge204: Come over, consider joining #avrs
02:10 PM Smidge204: Nah. Not much worse than someone who camps in a channel they don't like trying to siphon off users.
02:12 PM Smidge204: I don't know how long that nonsense has been going on but it's been a thing for at least a year. Everyone's moved on but you at this point
11:27 PM day_ is now known as day