[lldb] Fix -Waggressive-loop-optimizations warning
authorVedant Kumar <vsk@apple.com>
Mon, 7 Nov 2016 02:39:37 +0000 (02:39 +0000)
committerVedant Kumar <vsk@apple.com>
Mon, 7 Nov 2016 02:39:37 +0000 (02:39 +0000)
We shouldn't access past the end of an array, even if we think that the
layout of the struct containing the array is always what we expect. The
compiler is free to optimize away the stores as undefined behavior, and
in fact, GCC 6.2.1 claims it will do exactly this.

llvm-svn: 286093

lldb/source/Plugins/ObjectFile/Mach-O/ObjectFileMachO.cpp

index 04bed4e..45d3b12 100644 (file)
@@ -667,8 +667,12 @@ public:
         // x0-x29 + fp + lr + sp + pc (== 33 64-bit registers) plus cpsr (1
         // 32-bit register)
         if (count >= (33 * 2) + 1) {
-          for (uint32_t i = 0; i < 33; ++i)
+          for (uint32_t i = 0; i < 29; ++i)
             gpr.x[i] = data.GetU64(&offset);
+          gpr.fp = data.GetU64(&offset);
+          gpr.lr = data.GetU64(&offset);
+          gpr.sp = data.GetU64(&offset);
+          gpr.pc = data.GetU64(&offset);
           gpr.cpsr = data.GetU32(&offset);
           SetError(GPRRegSet, Read, 0);
         }