This is possible, but we have no evidence to think this is what
happened.  Moreover, the main thread didn't yet return from
pthread_mutex_unlock when SIGPROF is delivered:

>> #4  0x00000000005c3f27 in backtrace_top_function () at eval.c:4294
>>         pdl = <optimized out>
>> #5  0x000000000063a0da in add_sample (plog=0xcdf060 <cpu>, count=1436) at lisp.h:1192
>> #6  0x0000000000557604 in deliver_process_signal (sig=27, handler=0x63a440 <handle_profiler_signal>) at sysdep.c:1758
>>         old_errno = 11
>>         on_main_thread = true
>> #7  0x00007fbda4812970 in <signal handler called> () at /lib64/libpthread.so.0
>> #8  0x00007fbda481154a in __lll_unlock_wake () at /lib64/libpthread.so.0
>> #9  0x00007fbda480c2e6 in __pthread_mutex_unlock_usercnt () at /lib64/libpthread.so.0
>> >> #10 0x000000000063af2f in release_global_lock () at thread.c:621

Is the global lock already released at this point? are other threads
allowed to run?  What is __lll_unlock_wake about -- doesn't it wake
some other thread?  If it did not yet do so, the other thread couldn't
have been running.

IOW, we need backtrace from all the threads, not just from the main
thread, to draw any definitive conclusions.