09:39 <pepijndevos> hmmm dilema...

09:39 <pepijndevos> The GW1N-1 has 6 clock pins, but there are 8 clock spines to fuzz

09:40 <pepijndevos> {'A6': 'SPINER0', 'B6': 'SPINER1', 'C6': 'SPINER2', 'D6': 'SPINER3', 'A7': 'SPINER4', 'B7': 'SPINER5'} {'A4': 'SPINEL0', 'B4': 'SPINEL1', 'C4': 'SPINEL2', 'D4': 'SPINEL3', 'A5': 'SPINEL4', 'B5': 'SPINEL5'}

09:40 <pepijndevos> Spot the pattern...

09:40 <pepijndevos> I'm tempted to just hardcode that part and complete the pattern for the last two bits.

09:41 <pepijndevos> I already have the fuzzer, but as of now that would only get you the 6 spines the vendor likes to use for the 6 clock pins I have

09:56 <pepijndevos> {'W830': 'IOB10A', 'CLK1': 'IOL5A', 'CLK2': 'IOL7A', 'LSR0': 'IOR5A', 'CLK0': 'IOB11A', 'LSR1': 'IOR7A'}

09:56 <pepijndevos> The sources are a lot more erratic though...

10:16 <omnitechnomancer> can the clock spines be used to route global resets or clock enables as well?

10:17 <pepijndevos> I think so... lemme see

10:20 <pepijndevos> In [15]: [dest for dest, srcs in db.grid[2][2].pips.items() if "GB00" in srcs]

10:20 <pepijndevos> Out[15]: ['CLK0', 'CLK1', 'CLK2', 'EW10', 'EW20']

10:20 <pepijndevos> huh....

10:23 <pepijndevos> There is a lot to learn... I should try what the vendor tools do when you route a global net to a reset or enable

10:23 <pepijndevos> So EW10 are weird things that go in two directions.

10:23 <pepijndevos> (east and west)

10:23 <pepijndevos> And the CE0 does take LB01 as an input, and I have no idea what that is.

10:43 <omnitechnomancer> weird

12:53 <pepijndevos> My guess is Global Branch and Local Branch, but what a local branch is remains to be determined

12:54 <pepijndevos> I have yet to find anything that *drives* a LB

12:55 <pepijndevos> What if.... EW10 silently drive them...

15:22 <pepijndevos> waaaaaa

15:23 <pepijndevos> I tried to apply my fuzzer to GW1NR-9 and it's... different

15:23 <pepijndevos> GW1N-1 has 2 center tiles for left and right... GW1NR-9 has THREE!!! HUH???