The abridged justification is as follows:

-   The deadlock seems to occur if flag initialization happens to occur while a sample is being created.
    -   Each sample has its own mutex that is locked when a new sample is registered, i.e. created for the first time.
    -   The flag implicitly creates a global sampler object which locks `graveyard_`'s mutex.
-   Usually, in `PushDead`, the `graveyard` is locked before the sample, hence triggering deadlock detection.
-   This lock order can never be recreated since this code is executed exactly once per sample object, and the sample object cannot be accessed until after the method returns.
-   It should therefore be safe to ignore any locking order condition that may occur during sample creation.

PiperOrigin-RevId: 493901903
Change-Id: I094abca82c1a8a82ac392383c72469d68eef09c4

1 files changed, 8 insertions, 0 deletions

diff --git a/absl/profiling/internal/sample_recorder.h b/absl/profiling/internal/sample_recorder.h
index ef1489b1..371f6c47 100644
--- a/absl/profiling/internal/sample_recorder.h
+++ b/absl/profiling/internal/sample_recorder.h
@@ -199,6 +199,14 @@ T* SampleRecorder<T>::Register(Targs&&... args) {
     sample = new T();
     {
       absl::MutexLock sample_lock(&sample->init_mu);
+      // If flag initialization happens to occur (perhaps in another thread)
+      // while in this block, it will lock `graveyard_` which is usually always
+      // locked before any sample. This will appear as a lock inversion.
+      // However, this code is run exactly once per sample, and this sample
+      // cannot be accessed until after it is returned from this method.  This
+      // means that this lock state can never be recreated, so we can safely
+      // inform the deadlock detector to ignore it.
+      sample->init_mu.ForgetDeadlockInfo();
       sample->PrepareForSampling(std::forward<Targs>(args)...);
     }
     PushNew(sample);