I have been given some SDF that has been generated for min/max
write_sdf ../SDF/func.sdf -precision 6 -max_view func_worst -min_view func_best
When I look in the SDF file, I see:
(IOPATH RN Q () (0.445556::1.320084))
(IOPATH RN QN (0.282808::0.854712) ())
(IOPATH (posedge CK) Q (0.555220::1.656612) (0.407100::1.234008))
(IOPATH (posedge CK) QN (0.245180::0.769608) (0.303232::0.970056))
(WIDTH (negedge RN) (::0.482700))
(WIDTH (posedge CK) (::0.356600))
(WIDTH (negedge CK) (::0.502200))
(PERIOD (posedge CK) (::1.000000))
(PERIOD (negedge CK) (::1.000000))
(SETUPHOLD (posedge D) (posedge CK) (::0.351500) (-0.094300::))
(SETUPHOLD (negedge D) (posedge CK) (::1.047800) (-0.297200::))
(RECREM (posedge RN) (posedge CK) (::0.381600) (-0.049500::))
(SETUPHOLD (posedge SE) (posedge CK) (::1.057100) (-0.088100::))
(SETUPHOLD (negedge SE) (posedge CK) (::0.964600) (-0.146400::))
(SETUPHOLD (posedge SI) (posedge CK) (::0.389800) (-0.107100::))
(SETUPHOLD (negedge SI) (posedge CK) (::1.007200) (-0.292000::))
i.e. The MIN/MAX triplets are present for the IOPATHS, but for SETUPHOLD I see only max for setup and min for hold.
I think that this is giving problems in post layout gate level simulation (ncsim) where I use TOOL_CONTROL and +maxdelays. In this case, when I run a max simulation, no hold value will be annotated, so I see timing issues (since actual hold time is negative).
I hope someone understands what I'm talking about ;-)
Can anyone provide a solution.