|
|
Magic Mailing List |
|
From: R. Timothy Edwards (tim AT stravinsky DOT jhuapl.edu) Date: Sat Apr 12 2003 - 19:20:59 EDT
To any who may care: Over the past week, I dug into my own code to fix a couple of long- standing but subtle errors. One was the non-Manhattan geometry handling of the cif output "bloat-or" function. CIF output should now be correct for non-Manhattan geometry in all cases, including non-Manhattan geometry on transistor gates (this applies to revision 35, which I will post on Monday, so don't rush out to get it). The other was a partially-implemented "widespacing" rule to cover DRC rules where spacing is different when measured to a layer with a width greater than some specified value. This usually applies only to metal routing. A few weeks ago while demonstrating the new features in magic 7.2, I realized that the "widespacing" rule can be thought of as two coupled rules. The first looks in one direction from an edge and determines whether the layer meets the width requirement. If it does, then the second rule, which looks in the other direction from the edge for spacing requirements, is "triggered". This works perfectly (as far as I can tell) on the widespacing rule, but it seems to me that the concept of the "triggered rule" is probably more general and has other uses besides widespacing. If you can think of any good uses to which it can be put, please let me know. Also, by request from Philippe, revision 35 has a "delete" function for cells, invoked by the command "cellname delete <cell>". I also stuck in a "cellname rename <cell> <newname>" function, which differs from "save" in that it changes a cell name without writing anything to disk. ---Tim
|
|
|
|