Force call the fallback handler.

Always check to see if the fallback handler has been called and is
not trying to dump a specific thread.

Bug: 69110957

Test: Verified on a system where the prctl value changes, that before the
Test: change it dumps multiple tombstones, and after the change it
Test: works as expected.
Test: Ran debuggerd unit tests.
Test: Dumped process using debuggerd -b <PID> and debuggerd <PID>.
Change-Id: Id98bbe96cced9335f7c3e17088bb4ab2ad2e7a64

1 files changed, 6 insertions, 1 deletions

diff --git a/debuggerd/handler/debuggerd_handler.cpp b/debuggerd/handler/debuggerd_handler.cpp
index d6b6d58d8..bd202ff14 100644
--- a/debuggerd/handler/debuggerd_handler.cpp
+++ b/debuggerd/handler/debuggerd_handler.cpp
@@ -429,7 +429,12 @@ static void debuggerd_signal_handler(int signal_number, siginfo_t* info, void* c
     abort_message = g_callbacks.get_abort_message();
   }
 
-  if (prctl(PR_GET_NO_NEW_PRIVS, 0, 0, 0, 0) == 1) {
+  // If sival_int is ~0, it means that the fallback handler has been called
+  // once before and this function is being called again to dump the stack
+  // of a specific thread. It is possible that the prctl call might return 1,
+  // then return 0 in subsequent calls, so check the sival_int to determine if
+  // the fallback handler should be called first.
+  if (info->si_value.sival_int == ~0 || prctl(PR_GET_NO_NEW_PRIVS, 0, 0, 0, 0) == 1) {
     // This check might be racy if another thread sets NO_NEW_PRIVS, but this should be unlikely,
     // you can only set NO_NEW_PRIVS to 1, and the effect should be at worst a single missing
     // ANR trace.