h_ro has quit [Ping timeout: 260 seconds]
* cr1901
wants a New Lattice Smell car tree
cr1901 has quit [Read error: Connection reset by peer]
cr1901 has joined #prjcombine
h_ro has joined #prjcombine
<
Wanda[cis]>
ohhh there's a new Vivado and it has spartanuplus devices
<
Wanda[cis]>
I wonder how much it's going to break my geometry extraction code
<
Wanda[cis]>
oh only up to su35p; no XP5IO for me
h_ro has quit [Ping timeout: 276 seconds]
h_ro has joined #prjcombine
<
Wanda[cis]>
as for "platform management controler", apparently the configuration center grew an AXI bus
<
Wanda[cis]>
(with the FPGA fabric being the requester)
<
Wanda[cis]>
it has also been renamed from "CONFIG_SITE" to "CSEC_SITE"
<
Wanda[cis]>
so there's a good chance this is some "secure element" bullshit going on?
<
Wanda[cis]>
no other new internal pins on that thing, just an AXI bus.
<
Wanda[cis]>
also "Lassen" is this a codename for this thing
<
Wanda[cis]>
huh. extra-large HDIO banks with 42 pins instead of 24.
<
Wanda[cis]>
I miscounted, 84 pins
<
Wanda[cis]>
... and one of the extra-large HDIO banks even owns a CMT? huh.
<
Wanda[cis]>
crazy shit
<
Wanda[cis]>
like catgirls and doggirls living together
<
Wanda[cis]>
also surprising absolutely noone, su35p/su25p/su10p are the same die
h_ro has quit [Ping timeout: 245 seconds]
h_ro has joined #prjcombine
<
Wanda[cis]>
... okay the banks do have 42 pins, there are just twice as many of them per clock region
<
Wanda[cis]>
oh. and it uses HDIO as the config bank.
<
Wanda[cis]>
how obscene.
<
Wanda[cis]>
ohh also the configuration interface is different?
<
Wanda[cis]>
it seems they removed parallel flash configuration
<
Wanda[cis]>
and added some new octal SPI mode ... which happens to have ECC? wtf is this
<
Wanda[cis]>
okay I have beaten the geometry extractor into submission
<
Wanda[cis]>
... let's see if I broke anything else
<
jn>
at this point, i wouldn't be surprised if the SPI contoller also had DDR
<
jn>
i've seen it in flash datasheets
<
Wanda[cis]>
yeah me neither
<
Wanda[cis]>
but it's hard to know from just a pin list
h_ro has quit [Ping timeout: 276 seconds]
<
cr1901>
Can't wait for hexadec-SPI