Fix an out-of-bounds error in RegisterContextDarwin_arm64

[labath]

Summary

gcc diagnoses this as "array subscript 63 is above array bounds of
'RegisterContextDarwin_arm64::VReg [32]'".

The correct fix seems to be subtracting the fpu register base index, but
I have no way of verifying that this actually works.

[labath]

Jason, what do you make of this? I believe the fix is pretty straight-forward, but since this does actually change behaviour, I think it would be good if you had a quick look at it first.

[clayborg]

lgtm

[jasonmolenda]

Sorry for not weighing in earlier, yes this is good.

[labath]

Thank you.

source/Plugins/Process/Utility/RegisterContextDarwin_arm64.cpp

bool RegisterContextDarwin_arm64::ReadRegister(const RegisterInfo *reg_info,
   case fpu_v24:
   case fpu_v25:
   case fpu_v26:
   case fpu_v27:
   case fpu_v28:
   case fpu_v29:
   case fpu_v30:
   case fpu_v31:
-    value.SetBytes(fpu.v[reg].bytes.buffer, reg_info->byte_size,
+    value.SetBytes(fpu.v[reg - fpu_v0].bytes.buffer, reg_info->byte_size,
                    endian::InlHostByteOrder());
     break;
 
   case fpu_s0:
   case fpu_s1:
   case fpu_s2:
   case fpu_s3:
   case fpu_s4: