[libc-commits] [libc] [libc][pthreads] pthread_cond_{destroy|init|signal|wait} (PR #88583)

[Nick Desaulniers via libc-commits]

================
@@ -0,0 +1,37 @@
+//===-- Linux implementation of the pthread_cond_wait function ------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "pthread_cond_wait.h"
+
+#include "include/llvm-libc-macros/generic-error-number-macros.h" // EINVAL
+#include "src/__support/common.h"
+#include "src/threads/linux/CndVar.h"
+
+#include <pthread.h> // pthread_cond_t, pthread_mutex_t
+// TODO: https://github.com/llvm/llvm-project/issues/88580
+#include <threads.h> // thrd_success
+
+namespace LIBC_NAMESPACE {
+
+LLVM_LIBC_FUNCTION(int, pthread_cond_wait,
+                   (pthread_cond_t * cond, pthread_mutex_t *mutex)) {
+  if (!cond || !mutex)
+    return EINVAL;
+
+  auto *C = reinterpret_cast<CndVar *>(cond);
+  auto *M = reinterpret_cast<Mutex *>(mutex);
+  int ret = C->wait(M);
----------------
nickdesaulniers wrote:

So I guess one big issue with re-using CndVar is that CndVar::wait is implemented in terms of FUTEX_SYSCALL_ID, which is very much going to be Linux-only. Same with CndVar::notify_one and broadcast.

Though perhaps that's a problem to solve for #88580.

https://github.com/llvm/llvm-project/pull/88583