Use initial-exec tls for libunwind's recursion flag

If we don't do it, then reading variable calls to __tls_get_addr, which
uses malloc on first call. initial-exec makes dynamic linker reserve tls
offset for recusion flag early and thus avoid unsafe calls to malloc.

This fixes issue #786.

src/base/basictypes.h

@@ -192,6 +192,12 @@ struct CompileAssert {
 # define ATTRIBUTE_UNUSED
 #endif
 
+#if defined(HAVE___ATTRIBUTE__) && defined(HAVE_TLS)
+#define ATTR_INITIAL_EXEC __attribute__ ((tls_model ("initial-exec")))
+#else
+#define ATTR_INITIAL_EXEC
+#endif
+
 #define COMPILE_ASSERT(expr, msg)                               \
   typedef CompileAssert<(bool(expr))> msg[bool(expr) ? 1 : -1] ATTRIBUTE_UNUSED
 

src/stacktrace_libunwind-inl.h

@@ -47,6 +47,8 @@ extern "C" {
 #include <libunwind.h>
 }
 #include "gperftools/stacktrace.h"
+
+#include "base/basictypes.h"
 #include "base/logging.h"
 
 // Sometimes, we can try to get a stack trace from within a stack
@@ -56,7 +58,7 @@ extern "C" {
 // recursive request, we'd end up with infinite recursion or deadlock.
 // Luckily, it's safe to ignore those subsequent traces.  In such
 // cases, we return 0 to indicate the situation.
-static __thread int recursive;
+static __thread int recursive ATTR_INITIAL_EXEC;
 
 #if defined(TCMALLOC_ENABLE_UNWIND_FROM_UCONTEXT) && (defined(__i386__) || defined(__x86_64__)) && defined(__GNU_LIBRARY__)
 #define BASE_STACKTRACE_UNW_CONTEXT_IS_UCONTEXT 1

src/thread_cache.h

@@ -259,12 +259,6 @@ class ThreadCache {
   // Since we don't really use dlopen in google code -- and using dlopen
   // on a malloc replacement is asking for trouble in any case -- that's
   // a good tradeoff for us.
-#ifdef HAVE___ATTRIBUTE__
-#define ATTR_INITIAL_EXEC __attribute__ ((tls_model ("initial-exec")))
-#else
-#define ATTR_INITIAL_EXEC
-#endif
-
 #ifdef HAVE_TLS
   struct ThreadLocalData {
     ThreadCache* heap;