Make recursive mutexes faster

Recursive mutexes now go as fast as normal mutexes. The tradeoff is they
are no longer safe to use in signal handlers. However you can still have
signal safe mutexes if you set your mutex to both recursive and pshared.
You can also make functions that use recursive mutexes signal safe using
sigprocmask to ensure recursion doesn't happen due to any signal handler

The impact of this change is that, on Windows, many functions which edit
the file descriptor table rely on recursive mutexes, e.g. open(). If you
develop your app so it uses pread() and pwrite() then your app should go
very fast when performing a heavily multithreaded and contended workload

For example, when scaling to 40+ cores, *NSYNC mutexes can go as much as
1000x faster (in CPU time) than the naive recursive lock implementation.
Now recursive will use *NSYNC under the hood when it's possible to do so

test/libc/intrin/pthread_mutex_lock_test.c

@@ -95,7 +95,6 @@ TEST(pthread_mutex_lock, recursive) {
   }
   ASSERT_EQ(0, pthread_mutex_lock(&lock));
   ASSERT_EQ(0, pthread_mutex_unlock(&lock));
-  ASSERT_EQ(0, pthread_mutex_unlock(&lock));
   ASSERT_EQ(0, pthread_mutex_destroy(&lock));
 }
 

test/libc/thread/footek_test.c

@@ -1,4 +1,4 @@
-#define USE        POSIX
+#define USE        POSIX_RECURSIVE
 #define ITERATIONS 100000
 #define THREADS    30
 

test/libc/thread/pthread_cancel_deferred_cond_test.c

@@ -2,10 +2,11 @@
 #include <pthread.h>
 #include <stdlib.h>
 #include <unistd.h>
+#include "libc/stdio/stdio.h"
 
 int got_cleanup;
-pthread_cond_t cv = PTHREAD_COND_INITIALIZER;
-pthread_mutex_t mu = PTHREAD_MUTEX_INITIALIZER;
+pthread_cond_t cv;
+pthread_mutex_t mu;
 
 void cleanup(void* arg) {
   got_cleanup = 1;
@@ -23,6 +24,12 @@ void* worker(void* arg) {
 int main(int argc, char* argv[]) {
   void* rc;
   pthread_t th;
+  pthread_mutexattr_t at;
+  pthread_mutexattr_init(&at);
+  pthread_mutexattr_settype(&at, PTHREAD_MUTEX_NORMAL);
+  pthread_mutex_init(&mu, &at);
+  pthread_mutexattr_destroy(&at);
+  pthread_cond_init(&cv, 0);
   if (pthread_create(&th, 0, worker, 0))
     return 2;
   if (pthread_cancel(th))
@@ -37,4 +44,6 @@ int main(int argc, char* argv[]) {
     return 7;
   if (pthread_mutex_unlock(&mu))
     return 8;
+  pthread_mutex_destroy(&mu);
+  pthread_cond_destroy(&cv);
 }

test/libc/thread/pthread_cancel_test.c

@@ -40,6 +40,12 @@ atomic_int gotcleanup;
 
 void SetUpOnce(void) {
   testlib_enable_tmp_setup_teardown();
+  pthread_mutexattr_t at;
+  pthread_mutexattr_init(&at);
+  pthread_mutexattr_settype(&at, PTHREAD_MUTEX_NORMAL);
+  pthread_mutex_init(&mu, &at);
+  pthread_mutexattr_destroy(&at);
+  pthread_cond_init(&cv, 0);
 }
 
 void SetUp(void) {