[Libre-soc-dev] daily kan-ban update 02oct2020

Luke Kenneth Casson Leighton lkcl at lkcl.net
Fri Oct 2 20:54:15 BST 2020

today continuing with the pinmux program getting it to generate
coordinated information for the soc, litex, *and* coriolis2 ioring.

in particular, staf has asked for the entire pinset to go via JTAG IO
testing.  this involves creating two identical sets of IO pins, each
the opposite direction of the other, which are connected to JTAG IO
for either "passthru" under normal operation and "under JTAG control"
for testing.

then those two pairs go out through litex, which does absolutely
nothing to them other than to wire one set to the actual IOpads and
the other set to the peripherals (UART, SDRAM, I2C).

then finally once handed to coriolis2 the IO pads get created
depending on the exact same spec, in, out, enable.

it means having 100+ Pins all identified by function, type, name and
quantity, in *four* different formats.

to try this manually would be insane.

so the pinmux program is what generates the information from a central
"spec" so that, as usual, transcription errors are avoided.

eventually we will have an actual pinmux, with CSRs n everyfink.  just
not right now, there is too much else to do


More information about the Libre-soc-dev mailing list