https://open-source-silicon.dev logo
#sky130
Title
# sky130
s

Sajjad Ahmed

06/07/2021, 4:11 PM
i am having issue with synthesized netlist when using the latest version of sky130A pdk. there is a cell which has broken connections. an can be seen in snaps. why is it so?? the version used in MPW-1 shuttle doesn't have this issue. the cell is "sky130_fd_sc_hc__dfxtp" and its input "D" is not connected in new version but it is fine in the old one. should we replace it from old version or there is update expected. in current version a wire "D__delayed" is assigned to the cell but it doesn't have any connection with the actual input. @User @User
t

Tim Edwards

06/07/2021, 4:47 PM
This is a known issue, is documented in the issue tracker. It has not changed since MPW-one but should have something to do with whether a verilog simulation declares FUNCTIONAL or not.
s

Sajjad Ahmed

06/07/2021, 4:52 PM
@Tim Edwards i didn't get your point. should i replace the cell with older one which is looking fine. or is there any other solution???
t

Tim Edwards

06/07/2021, 4:58 PM
Your so-called "old version" is the FUNCTIONAL verilog, while your so-called "new version" is the BEHAVIORAL verilog. If you have a "define FUNCTIONAL" somewhere in the verilog, it will use the functional view. The behavioral view has the known issue.
s

Sajjad Ahmed

06/07/2021, 5:06 PM
@Rameen Anwar @Wajeh ul hasan @Aireen Amir Jalal have we defined FUNCTIONAL somewhere in our flow??
@Tim Edwards i have reviewed the file and yes we need to define FUNCTIONAL. thank you!!