panic: pthread_mutex_lock: Invalid argument

[manuelgustavo]

I have 2 servers and 2 clients running on the same binary.

Even without any communication between them, client 1 dials to server 1, and client 2 dials to server 2.
NNG sockets are closed for the clients, followed by NNG sockets being closed for the servers.

If I loop through the execution of the process over and over, e.g.

while ./process ; do :; done

Eventually, in a minute or so, I get:

panic: pthread_mutex_lock: Invalid argument
This message is indicative of a BUG.
Report this at https://github.com/nanomsg/nng/issues
free(): corrupted unsorted chunks

If I don't close the sockets on the server side and leave for the process to clean-up during its teardown, the problem doesn't happen.
The same issue happens if the server stops listening for clients by using nng_listener_close whilst keeping the sockets open.

I'm using clang-14, built under Ubuntu 22.04 using nng v1.10.1.

The call stack (when building with -g -O0)

libc.so.6!__pthread_kill_implementation(int no_tid, int signo, pthread_t threadid) (pthread_kill.c:44)
libc.so.6!__pthread_kill_internal(int signo, pthread_t threadid) (pthread_kill.c:78)
libc.so.6!__GI___pthread_kill(pthread_t threadid, int signo) (pthread_kill.c:89)
libc.so.6!__GI_raise(int sig) (raise.c:26)
libc.so.6!__GI_abort() (abort.c:79)
nni_plat_abort() (/home/secretarium/repo/gateway/submodules/third-party/nng/src/platform/posix/posix_debug.c:23)
nni_panic(const char * fmt) (/home/secretarium/repo/gateway/submodules/third-party/nng/src/core/panic.c:66)
nni_pthread_mutex_lock(pthread_mutex_t * m) (/home/secretarium/repo/gateway/submodules/third-party/nng/src/platform/posix/posix_thread.c:85)
nni_plat_mtx_lock(nni_plat_mtx * mtx) (/home/secretarium/repo/gateway/submodules/third-party/nng/src/platform/posix/posix_thread.c:146)
nni_mtx_lock(nni_mtx * mtx) (/home/secretarium/repo/gateway/submodules/third-party/nng/src/core/thread.c:27)
nni_task_dispatch(nni_task * task) (/home/secretarium/repo/gateway/submodules/third-party/nng/src/core/taskq.c:163)
nni_aio_finish_impl(nni_aio * aio, int rv, size_t count, nni_msg * msg, _Bool sync) (/home/secretarium/repo/gateway/submodules/third-party/nng/src/core/aio.c:454)
nni_aio_finish_error(nni_aio * aio, int result) (/home/secretarium/repo/gateway/submodules/third-party/nng/src/core/aio.c:473)
tls_conn_cb(void * arg) (/home/secretarium/repo/gateway/submodules/third-party/nng/src/supplemental/tls/tls_common.c:138)
nni_taskq_thread(void * self) (/home/secretarium/repo/gateway/submodules/third-party/nng/src/core/taskq.c:47)
nni_thr_wrap(void * arg) (/home/secretarium/repo/gateway/submodules/third-party/nng/src/core/thread.c:94)
nni_plat_thr_main(void * arg) (/home/secretarium/repo/gateway/submodules/third-party/nng/src/platform/posix/posix_thread.c:266)
libc.so.6!start_thread(void * arg) (pthread_create.c:442)
libc.so.6!clone3() (clone3.S:81)

[manuelgustavo]

It also appears to happen only with TLS.