[Lldb-commits] [PATCH] Refactor POSIXThread::GetRegisterContext and use i386 register context

[Todd Fiala]

  I'm in the process of testing the change now.  Mostly looks good, but there are a few key missing break statements that need to be added.  In the test I'm running, I've added those commented above.  I'll attach my final diff after my 32-bit build is done and I test it out.


================
Comment at: source/Plugins/Process/POSIX/POSIXThread.cpp:162
@@ -162,11 +161,3 @@
                 }
-
-                if (reg_interface)
-                {
-                    RegisterContextPOSIXProcessMonitor_mips64 *reg_ctx = new RegisterContextPOSIXProcessMonitor_mips64(*this, 0, reg_interface);
-                    m_posix_thread = reg_ctx;
-                    m_reg_context_sp.reset(reg_ctx);
-                }
-                break;
             }
 
----------------
We lost an essential break; (out of the case ArchSpec::eCore_mips64) here.

================
Comment at: source/Plugins/Process/POSIX/POSIXThread.cpp:180
@@ -190,1 +179,3 @@
+            }
 
+            case ArchSpec::eCore_x86_64_x86_64:
----------------
We should have a break out of the x86 cases here.

================
Comment at: source/Plugins/Process/POSIX/POSIXThread.cpp:193
@@ -196,3 +192,3 @@
+                        break;
                 }
             }
----------------
We should have a break out of the x86_64 case here.

================
Comment at: source/Plugins/Process/POSIX/POSIXThread.cpp:208
@@ +207,3 @@
+                m_reg_context_sp.reset(reg_ctx);
+            }
+            case ArchSpec::eCore_x86_32_i386:
----------------
We need a break out of this case here.

================
Comment at: source/Plugins/Process/POSIX/POSIXThread.cpp:217
@@ -200,1 +216,3 @@
+                m_reg_context_sp.reset(reg_ctx);
+            }
             default:
----------------
We need a break out of these cases here.


http://llvm-reviews.chandlerc.com/D2765