Closed mikesinouye closed 3 days ago
clang-tidy review says "All clean, LGTM! :+1:"
@mikesinouye Thanks for the contribution! Could you update the PPL unit tests as well? They're failing in our CI, which prevents the PR from merging.
getObjName is wrong, I suppose you meant getName. However getId is relied on elsewhere so I see no reason to be different here.
clang-tidy review says "All clean, LGTM! :+1:"
We are observing non-determinism in the output of the
place_pins
step across multiple runs of OpenROAD with fixed initial conditions. The root cause seems to be iteration over a set ofodb::dbBTerm*
.With this change, we observe deterministic pin placement. Using
dbBTerm.getId()
for the comparison was considered but we suggest usinggetObjNames()
to avoid being affected by upstream allocation order.