From 58d03e437a95da78f612f833a249376aa202b502 Mon Sep 17 00:00:00 2001
From: Kartik Agaram <vc@akkartik.com>
Date: Tue, 9 Jul 2019 18:02:44 -0700
Subject: [PATCH] bug: null address messing up instruction decode

Leads to a bad error message.
---
 subx/014indirect_addressing.cc | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/subx/014indirect_addressing.cc b/subx/014indirect_addressing.cc
index dd00c614..3767c46d 100644
--- a/subx/014indirect_addressing.cc
+++ b/subx/014indirect_addressing.cc
@@ -874,7 +874,7 @@ void test_add_r32_to_mem_at_r32_plus_disp8() {
 }
 
 :(before "End Mod Special-cases(addr)")
-case 1:  // indirect + disp8 addressing
+case 1: {  // indirect + disp8 addressing
   switch (rm) {
   default:
     addr = Reg[rm].u;
@@ -882,11 +882,16 @@ case 1:  // indirect + disp8 addressing
     break;
   // End Mod 1 Special-cases(addr)
   }
+  int8_t displacement = static_cast<int8_t>(next());
   if (addr > 0) {
-    addr += static_cast<int8_t>(next());
+    addr += displacement;
     trace(Callstack_depth+1, "run") << "effective address is 0x" << HEXWORD << addr << " (after adding disp8)" << end();
   }
+  else {
+    trace(Callstack_depth+1, "run") << "null address; skipping displacement" << end();
+  }
   break;
+}
 
 :(code)
 void test_add_r32_to_mem_at_r32_plus_negative_disp8() {