[cfe-commits] r96545 - in /cfe/trunk: lib/Sema/SemaStmt.cpp test/Sema/switch.c

[Douglas Gregor]

Author: dgregor
Date: Wed Feb 17 18:56:01 2010
New Revision: 96545

URL: http://llvm.org/viewvc/llvm-project?rev=96545&view=rev
Log:
Don't diagnose overflow in case statements when the conversion is a
signed<->unsigned conversion with the same bit width. Fixes
<rdar://problem/7658121>.

Modified:
    cfe/trunk/lib/Sema/SemaStmt.cpp
    cfe/trunk/test/Sema/switch.c

Modified: cfe/trunk/lib/Sema/SemaStmt.cpp
URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaStmt.cpp?rev=96545&r1=96544&r2=96545&view=diff

==============================================================================
--- cfe/trunk/lib/Sema/SemaStmt.cpp (original)
+++ cfe/trunk/lib/Sema/SemaStmt.cpp Wed Feb 17 18:56:01 2010
@@ -340,11 +340,11 @@
   } else if (NewSign != Val.isSigned()) {
     // Convert the sign to match the sign of the condition.  This can cause
     // overflow as well: unsigned(INTMIN)
+    // We don't diagnose this overflow, because it is implementation-defined 
+    // behavior.
+    // FIXME: Introduce a second, default-ignored warning for this case?
     llvm::APSInt OldVal(Val);
     Val.setIsSigned(NewSign);
-
-    if (Val.isNegative())  // Sign bit changes meaning.
-      Diag(Loc, DiagID) << OldVal.toString(10) << Val.toString(10);
   }
 }
 

Modified: cfe/trunk/test/Sema/switch.c
URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Sema/switch.c?rev=96545&r1=96544&r2=96545&view=diff

==============================================================================
--- cfe/trunk/test/Sema/switch.c (original)
+++ cfe/trunk/test/Sema/switch.c Wed Feb 17 18:56:01 2010
@@ -240,3 +240,17 @@
   }
   return -1;
 }
+
+// <rdar://problem/7658121>
+enum {
+  EC0 = 0xFFFF0000,
+  EC1 = 0xFFFF0001,
+};
+
+int test14(int a) {
+  switch(a) {
+  case EC0: return 0;
+  case EC1: return 1;
+  }
+  return 0;
+}