Modified a cifinput rule in the magic tech file for reading
thick-oxide p-diff that looks for lack of N+ implant but not
specifically for the presence of P+ implant.  This avoids losing
the diffusion completely on read-in when the P+ implant is in
another cell.  This hack avoids issues when reading in the
insanely disorganized layout for the I/O OVT GPIO cell, and
should not cause issues with reading any valid layout.

VERSION

diff --git a/VERSION b/VERSION
index 4a6ed31..452226a 100644
--- a/VERSION
+++ b/VERSION
@@ -1 +1 @@
-1.0.497
+1.0.498

sky130/magic/sky130.tech

diff --git a/sky130/magic/sky130.tech b/sky130/magic/sky130.tech
index 851ffec..8ca96e3 100644
--- a/sky130/magic/sky130.tech
+++ b/sky130/magic/sky130.tech
@@ -2603,6 +2603,7 @@
  and HVI,hvcheck
  labels DIFF
 
+ # Note: Not checking PSDM specifically, just lack of NSDM
  templayer mvpdiffarea DIFF,DIFFTXT,DIFFPIN,barediff
  and-not POLY
  and NWELL,nwelcheck
@@ -2610,7 +2611,6 @@
  and HVI,hvcheck
  and-not DIODE
  and-not DIFFRES
- and PSDM
  copyup mvpdifcheck
  labels DIFF
  labels DIFFPIN port
@@ -3471,12 +3471,13 @@
  or mvpdcnowell
  labels CONT
 
+ # Note: Not checking PSDM specifically, just lack of NSDM
  templayer mvpscbase CONT
  or barecont
  and LI
  or barelicont
  and DIFF,TAP
- and PSDM
+ and-not NSDM
  and-not NWELL,nwelcheck
  and-not mvpfetexpand
  and HVI,hvcheck