Added "gds polygon subcells true" to the GDS import script for
GF180MCU. This goes along with an observation that the change
to magic's code to correct the original problem with reading
the GF180MCU I/O library causes issues when reading the Sky130
I/O library, and is, in fact, an improperly implemented method.
It happens to work for GF180MCU because all of the non-
Manhattan geometry is in metal and does not go through boolean
operations. Magic has been corrected for Sky130 (and general)
behavior, but this line now needs to be added to the GF180MCU
script to avoid the original problem with reading the I/O
library.
diff --git a/VERSION b/VERSION
index 6746252..a569f17 100644
--- a/VERSION
+++ b/VERSION
@@ -1 +1 @@
-1.0.384
+1.0.385
diff --git a/gf180mcu/custom/scripts/gds_import_io.tcl b/gf180mcu/custom/scripts/gds_import_io.tcl
index 04be814..e4d5a5f 100644
--- a/gf180mcu/custom/scripts/gds_import_io.tcl
+++ b/gf180mcu/custom/scripts/gds_import_io.tcl
@@ -3,4 +3,5 @@
# no longer has the issue of slow reads of complex polygons.
gds ordering on
gds flatten true
+gds polygon subcells true
gds flatglob *_CDNS_*
