#416 DRC shows spurious MnFil.a1 and TMnFil.a violations Issue created by smunaut So I was trying Magic fill algorithm and then running the result through DRC and there was a lot of violations reported.

Rule M1Fil.a1: 55502 error(s)
Rule M2Fil.a1: 81040 error(s)
Rule M3Fil.a1: 179170 error(s)
Rule M4Fil.a1: 197511 error(s)
Rule M5Fil.a1: 182144 error(s)
Rule TM1Fil.a: 1735 error(s)
Rule TM2Fil.a: 66799 error(s)

However when loading the resulting DRC and inspecting it, the violations appear to be spurious. Here's an example for

M1Fil.a1

: [Image](https://private-user-images.githubusercontent.com/129013/428283633-b10949b0-7698-4c37-ad3d-bb9fd68a4da0.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3NDQ3MTkwMzUsIm5iZiI6MTc0NDcxODczNSwicGF0aCI6Ii8xMjkwMTMvNDI4MjgzNjMzLWIxMDk0OWIwLTc2OTgtNGMzNy1hZDNkLWJiOWZkNjhhNGRhMC5wbmc_WC1BbXotQWxnb3JpdGhtPUFXUzQtSE1BQy1TSEEyNTYmWC1BbXotQ3JlZGVudGlhbD1BS0lBVkNPRFlMU0E1M1BRSzRaQSUyRjIwMjUwNDE1JTJGdXMtZWFzdC0xJTJGczMlMkZhd3M0X3JlcXVlc3QmWC1BbXotRGF0ZT0yMDI1MDQxNVQxMjA1MzVaJlgtQW16LUV4cGlyZXM9MzAwJlgtQW16LVNpZ25hdHVyZT0wMmEyZDI3NDAzNjYwOTUxMjQyMjUyNWI0ZjU2ZjY4YWRkNmI2OGU4NWJhNzY4N2I0M2M3Y2U4MmQ4MDA2MTk0JlgtQW16LVNpZ25lZEhlYWRlcnM9aG9zdCJ9.z-UI3fizBiK6WAXzS6_QPNZ0mjN-N1vx8jxz0rUHOi8) The edge pair don't belong to the same fill polygon at all ... and those polygons are far enough appart and are wider than 1u. Attached a reproducer. It's only the fill pattern itself without the design, but it's enough to see the issue. There is also a minimized lydrc file with just the

M1Fil.a1

rule to run faster but you can also use the maximal DRC to reproduce. ( Sorry for the .tar.xz inside a .zip but I needed xz to be under file limit and github only accepts zip ... ) drc_bug.zip (cc @RTimothyEdwards ) IHP-GmbH/IHP-Open-PDK