gatecat[m] has joined #prjcombine
<gatecat[m]> A somewhat untrustworthy source (who does have a good connection with various Lattice people but has not been correct about everything) claimed that the dedicated I2C routing was missing a level shifter and could damage the device if used
<Wanda[cis]> oh cute
<Wanda[cis]> except it doesn't make all that much sense
<Wanda[cis]> the default state if you don't instantiate the controller is that all inputs to the controller come from dedicated pins
<Wanda[cis]> so if it was dangerous, that'd mean that empty bitstream would be capable of damaging the device
Micko[m] has joined #prjcombine
<Micko[m]> Wanda: if you need some design that actually do work with SPI and I2C on iCE40 UP5K I did something long time ago https://github.com/mmicko/mikrobus-upduino/tree/master/src/picosoc and it is integrated with picosoc note it was still arachne-pnr used
<Wanda[cis]> hm
<Wanda[cis]> might come useful at some point
<Wanda[cis]> ... should obtain an actual devboard first tbh
<Wanda[cis]> the only ice40 I have is in a glasgow
<Micko[m]> ip_wrapper.v is most important one. this was upduino and my custom adapter
<Micko[m]> not any official lattice board
<Wanda[cis]> got the cursed thing to work
<Wanda[cis]> there's another condition in bitmap that disables dedicated pin routing for SB_I2C and SB_SPI specifically, while allowing all other hard IP with dedicated pins to use them
<Wanda[cis]> (note: there is no other hard IP with dedicated pins)
<Wanda[cis]> so they really went and disabled all the logic post-release, huh.
<Wanda[cis]> <Wanda[cis]> "the default state if you don't..." <- also I'd like to retract that, turns out the mux is 3-way
<Wanda[cis]> fabric, dedicated pin, nothing?
<Wanda[cis]> (also there must be a third place where they disabled the dedicated pin logic, somewhere in the router, but having extra routes in the bitstream isn't a problem for my purposes, so I won't bother fixing it)
kiboneu has joined #prjcombine
<Wanda[cis]> <Wanda[cis]> "also I'd like to retract that..." <- or not, I confused two different muxes
<Wanda[cis]> ... I should just hw-test it, worst case an UP5K goes up in smoke
<_whitenotifier> [prjunnamed/prjcombine] wanda-phi pushed 1 commit to main [+0/-0/±1] https://github.com/prjunnamed/prjcombine/compare/f00024f3c283...8933edb7aa40
<_whitenotifier> [prjunnamed/prjcombine] wanda-phi 8933edb - sdf: misc fixes.