[PATCH] ARM IAS: handle errors more appropriately

Renato Golin renato.golin at linaro.org
Sat Dec 28 03:42:21 PST 2013


  Hi Saleem,

  Again, the change makes sense to me, but I'm not sure this was intentional or not. Jim, Tim could know more, I've added them as reviewers.

  cheers,
  --renato


================
Comment at: lib/Target/ARM/AsmParser/ARMAsmParser.cpp:8111
@@ -8101,1 +8110,3 @@
+    return false;
+  }
 
----------------
This is redundant, but it's also more future-proof (if someone adds code in between).

================
Comment at: lib/Target/ARM/AsmParser/ARMAsmParser.cpp:8212
@@ -8189,3 +8211,3 @@
     Note(FnStartLoc, "previous .fnstart starts here");
-    return true;
+    return false;
   }
----------------
This one looks like an intentional True, which hints that the others were too. I don't remember why they all return true, it could be just left over from an early refactor that didn't go deep enough.


http://llvm-reviews.chandlerc.com/D2483



More information about the llvm-commits mailing list