[Lldb-commits] [lldb] [lldb] Support arbitrary precision integer registers (PR #166363)

[Matej Košík via lldb-commits]

================
@@ -218,7 +218,7 @@ class RegisterValue {
   }
 
   void SetUInt128(llvm::APInt uint) {
-    m_type = eTypeUInt128;
+    m_type = eTypeUIntBig;
     m_scalar = std::move(uint);
   }
----------------
sedymrak wrote:

I am not sure which of the two possibilties:
- `eTypeUInt`
- `eTypeUIntAP`
is better.

However, before answering the above question, we might consider clarifying another problem:

"Do we need `RegisterValue::m_type` (and `RegisterValue::Type`) at all?"

If we removed `RegisterValue::m_type`, wouldn't we still be able to recover all the relevant information from the primary source of information? (I guess `RegisterValue::m_scalar`)?

https://github.com/llvm/llvm-project/pull/166363