From cb0f9931ad0ab8407f58db4cf4a3d7c4004947d5 Mon Sep 17 00:00:00 2001
From: Dan <46821332+nsadeveloper789@users.noreply.github.com>
Date: Tue, 6 Apr 2021 16:20:21 -0400
Subject: [PATCH] GP-0: Relaxing the Largest sub-register mapper test.

---
 .../debug/mapping/LargestSubDebuggerRegisterMapperTest.java | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/Ghidra/Debug/Debugger/src/test/java/ghidra/app/plugin/core/debug/mapping/LargestSubDebuggerRegisterMapperTest.java b/Ghidra/Debug/Debugger/src/test/java/ghidra/app/plugin/core/debug/mapping/LargestSubDebuggerRegisterMapperTest.java
index 0c9bfa855c..20e53d5bab 100644
--- a/Ghidra/Debug/Debugger/src/test/java/ghidra/app/plugin/core/debug/mapping/LargestSubDebuggerRegisterMapperTest.java
+++ b/Ghidra/Debug/Debugger/src/test/java/ghidra/app/plugin/core/debug/mapping/LargestSubDebuggerRegisterMapperTest.java
@@ -265,7 +265,11 @@ public class LargestSubDebuggerRegisterMapperTest extends AbstractGhidraHeadedDe
 
 		rv = waitForValue(() -> rm.targetToTrace("eax", genBytes4()));
 		assertEquals(new RegisterValue(lRAX, new BigInteger("0000000089abcdef", 16)), rv);
-		assertNull(rm.targetToTrace("rax", genBytes8())); // Should no longer understand this one
+		/**
+		 * Should this be kept? I favoring a more accepting model makes sense, but I worry about
+		 * erroneous cases we might not be catching, by re-generating missing registers on the fly.
+		 */
+		// assertNull(rm.targetToTrace("rax", genBytes8())); // Should no longer understand this one
 
 		// This might be quite kick, back to 64-bit
 		mb.testProcess1.regs.addRegistersFromLanguage(getSLEIGH_X86_64_LANGUAGE(), r -> true);