Open gadfort opened 3 weeks ago
An additional case which is very similar:
Old placement:
New placement:
Place: sc_issue_tiny_rocket_job0_ihp130_sg13g2_stdcell_place0_20241029-124122.tar.gz
Route: sc_issue_tiny_rocket_job0_ihp130_sg13g2_stdcell_route0_20241029-124053.tar.gz
@gadfort it would be helpful to have the macro placement step as well
First testcase: sc_issue_ethmac_job0_ihp130_sg13g2_stdcell_floorplan0_20241029-131630.tar.gz
Second testcase: sc_issue_tiny_rocket_job0_ihp130_sg13g2_stdcell_floorplan0_20241029-131607.tar.gz
@maliberty here you go
Describe the bug
When enabling rtlmp over the previous macro placer, the resultant design is not able to be routed.
Relates to proposed removal of previous macro placer: https://github.com/The-OpenROAD-Project/OpenROAD/pull/6047
Expected Behavior
Design is routable
Environment
To Reproduce
Placement: sc_issue_ethmac_job0_ihp130_sg13g2_stdcell_place0_20241029-121603.tar.gz
Routing: sc_issue_ethmac_job0_ihp130_sg13g2_stdcell_route0_20241029-120252.tar.gz
Relevant log output
No response
Screenshots
place_macro placement:
rtlmp placement:
congresion markers:
Additional Context
No response