a subcell: A further check is done to see if there are any devices
that interact with the substrate in the shielded area in the same
subcell. If not, then this is a trivial case where a different
substrate area exists but does not impact the subcell and should be
ignored. This change prevents some extraction errors where a substrate
area can get isolated because there was an unrelated substrate area
nearby. Also: Corrected parsing of "dsubckt" types when reading .ext
files; this applies to any device where the optional substrate is
omitted. There are apparently two conflicting methods when a device
doesn't have a substrate: (1) Output "None" for the substrate, or
(2) don't output anything. The 2nd method can be detected by counting
the number of arguments on the line in the .ext file. The extflat/
EFread.c code handles both methods. ResReadExtFile() now does too,
but probably code should be changed to just use method (1) always.
Also: Prevented the "tile not visited" error message in "extresist"
from printing duplicate messages for every device associated with the
same tile. The message can appear when a labeled net has unconnected
regions, and "extract do unique" is not used.
381714e2d5