We use Magic for DRC (

sky130hvl

PDK) on this macro block (gds attached), which is a part of the LDO generator in OpenFASoC. The resulting report (attached) contains

nwell

-

dnwell

minimum spacing errors at the highlighted locations in the attached image. The scripts used for running DRC are here and here. The distance between the mentioned coordinates is

~ 5um

but the report says they are

< 4.5um

? Am I reading this wrong? Also the points reported (highlighted with a ruler measurement in the attached image) seem to be at the edge of the same region. In the attached image, dnwell is red, nwell is green and the other layers have been disabled. The sky130 rule:

nwell.7: Min spacing between nwell and deep nwell on separate nets

@Tim Edwards can you please take a look? Thanks! cc @mehdi

The error is between the deep nwell edge with the corner at coordinate (76.51um, 48.385um) and the unrelated nwell with corner coordinate (74.045um, 48.17um), which is less than 2 1/2 microns distant.

Ah, are the coordinates in the report incorrect?

The coordinates are correct. They are marking the portion of the deep nwell that is too close to the opposing nwell, which is how magic marks DRC errors (which may be different from how other tools mark the same DRC error). The coordinates represent the area occupied by deep nwell within a larger search area extending from the opposing nwell edge, which might be what you were expecting; that search area is roughly (74.020, 43.680) to (78.560, 48.410).

Oh I understand now,thanks!