<vup>
the 432 there is actually $NUM_FUNCTION_BLOCK * 108
<vup>
and for the xc9572xl which this was written for NUM_FUNCTION_BLOCK is four
<vup>
but with the xc95288xl this is 16
<vup>
but replacing the 432 with 16 * 108 actually yields the same usercode_low=90 and usercode_high=105 as with the other devices
<d1b2>
<brainstorm> Wohoo, thanks @vup! So @whitequark hunch was right with "it's quite possible that USERCODE is always at 90,105"
<vup>
fwiw this gives you the jed words for a device with `nfb` function blocks: `[nfb * 108 * row + 8 * nfb * ((8 * col + 6) // 8) for row in range(6,8) for col in range(8)]`
<_whitenotifier-1>
[glasgow] brainstorm commented on pull request #285: XC95288XL support - https://git.io/JPWRT
<vup>
so you could add support for the xc95144xl aswell, (whose `word_with` would be `64`)
<vup>
also fyi im @rroohhh :)
<d1b2>
<brainstorm> Argh, I was @-ing the wrong person over github? Sorry :_/
<vup>
also I think bitstream_words is wrong, it should be 1620 as all the others
<d1b2>
<brainstorm> Ah, good one, I guess that should be a separate PR though.
<d1b2>
<brainstorm> Alright, let me test that out...
<d1b2>
<brainstorm> % ls -alh cpld_144_with_1620.bit ../../../../cpld_144_foo.bit -rw-r--r-- 1 rvalls staff 101K Oct 25 18:39 ../../../../cpld_144_foo.bit <--- bitstream_words=6420 -rw-r--r-- 1 rvalls staff 26K Oct 29 23:21 cpld_144_with_1620.bit <--- @vup suggestion
<d1b2>
<brainstorm> Alright, I'll commit that, you definitely know your way much more through the guts of this thing 😉