strongsaxophone has quit [Ping timeout: 240 seconds]
strongsaxophone has joined #yosys
citypw has quit [Ping timeout: 248 seconds]
develonepi3 has joined #yosys
<develonepi3>
Hi all: Updated yosys yesterday. I have 2 systems with yosys installed. The first was built with clang Yosys 0.9+1706 (git sha1 34d2fbd, clang 3.8.0-2ubuntu4 -fPIC -Os), which is the one that fails creating my catzip.blif. The version created with gcc (Yosys 0.9+1706 (git sha1 34d2fbd2, gcc 9.2.0 -fPIC -Os) works okay.
emeb has quit [Quit: Leaving.]
indy has quit [Ping timeout: 240 seconds]
dys has quit [Ping timeout: 265 seconds]
rohitksingh has joined #yosys
m4ssi has quit [Remote host closed the connection]
fsasm has quit [Ping timeout: 260 seconds]
az0re has joined #yosys
rohitksingh has quit [Ping timeout: 265 seconds]
rohitksingh has joined #yosys
rohitksingh has quit [Ping timeout: 265 seconds]
kraiskil has quit [Ping timeout: 268 seconds]
rjo has quit [*.net *.split]
_florent_ has quit [*.net *.split]
esden has quit [*.net *.split]
bubble_buster has quit [*.net *.split]
emily has quit [*.net *.split]
flokli has quit [*.net *.split]
ZipCPU has quit [*.net *.split]
kristianpaul has quit [*.net *.split]
rjo has joined #yosys
ZipCPU has joined #yosys
bubble_buster has joined #yosys
esden has joined #yosys
emily has joined #yosys
_florent_ has joined #yosys
flokli has joined #yosys
kristianpaul has joined #yosys
anticw has quit [Remote host closed the connection]
anticw has joined #yosys
<develonepi3>
Hi all: I went back to Yosys 0.9+1706 (git sha1 0488492, clang 3.8.0-2ubuntu4 -fPIC -Os) on clang build and the catzip.blif was created.
Jybz has joined #yosys
fsasm has joined #yosys
kraiskil has joined #yosys
strongsaxophone has quit [Ping timeout: 268 seconds]
kraiskil has quit [Ping timeout: 265 seconds]
Jybz has quit [Quit: Konversation terminated!]
rohitksingh has joined #yosys
<Xiretza>
34d2fbd2 "Add opt_lut_ins pass" seems to have broken nextpnr-xilinx routing. looking at the diff between the json dumps, inputs that are driven by a LUT that now got optimized away entirely have a wire ID of 0 - does that make sense?
<daveshah>
That's a constant 0 - this is a nextpnr-xilinx issue definitely, I need to have more of a look
<Xiretza>
router2 starts off normal, but then reproducibly gets stuck at overused=147 for 20+ iterations
<daveshah>
Yeah, I think the problem is to do with mux input legalisation
indy has joined #yosys
kraiskil has joined #yosys
fsasm has quit [Ping timeout: 268 seconds]
rohitksingh has quit [Ping timeout: 265 seconds]
<daveshah>
In the end it seems to be actually a side effect of opt_lut_ins, that because a LUT in a mux structure could now be smaller than a LUT6, it would be possible in some cases to pack another LUT5 in - but in some cases this causes an output legality issue
<daveshah>
About to push what is hopefully a fix
<daveshah>
Xiretza: ok, fix pushed to xilinx-upstream, seems to route for me at least now