There is a deadlock that becomes more likely when we block to print all stack threads, so we are reverting SERVER-71520 in SERVER-90777. Once we resolve the deadlock, we should recommit SERVER-71520 so that we can get important diagnostic information back.
- depends on
-
SERVER-90775 libunwind deadlocks when called from signal handler while dl_iterate_phdr is running
-
- Closed
-
- is related to
-
SERVER-90777 Revert SERVER-71520
-
- Closed
-
-
SERVER-99670 should invariant() that the signal processing thread is started in printAllThreadStacksBlocking() to avoid a deadlock
-
- Closed
-
- related to
-
-
- Closed
-
-
-
- Closed
-