Fix run_for blocking indefinitely with outstanding work

run_task() had no timeout parameter — the reactor always blocked
forever or polled instantly, ignoring the timeout from do_one().
Propagate the timeout to each backend's I/O syscall (epoll_wait,
select, kevent) and return after a timed reactor poll so
run_one_until can recheck the deadline.

Author: sgerbino

include/boost/corosio/native/detail/epoll/epoll_scheduler.hpp

@@ -117,7 +117,8 @@ class BOOST_COROSIO_DECL epoll_scheduler final : public reactor_scheduler_base
 
 private:
     void
-    run_task(std::unique_lock<std::mutex>& lock, context_type* ctx) override;
+    run_task(std::unique_lock<std::mutex>& lock, context_type* ctx,
+        long timeout_us) override;
     void interrupt_reactor() const override;
     void update_timerfd() const;
 
@@ -294,9 +295,16 @@ epoll_scheduler::update_timerfd() const
 }
 
 inline void
-epoll_scheduler::run_task(std::unique_lock<std::mutex>& lock, context_type* ctx)
+epoll_scheduler::run_task(
+    std::unique_lock<std::mutex>& lock, context_type* ctx, long timeout_us)
 {
-    int timeout_ms = task_interrupted_ ? 0 : -1;
+    int timeout_ms;
+    if (task_interrupted_)
+        timeout_ms = 0;
+    else if (timeout_us < 0)
+        timeout_ms = -1;
+    else
+        timeout_ms = static_cast<int>((timeout_us + 999) / 1000);
 
     if (lock.owns_lock())
         lock.unlock();

include/boost/corosio/native/detail/kqueue/kqueue_scheduler.hpp

@@ -139,7 +139,8 @@ class BOOST_COROSIO_DECL kqueue_scheduler final : public reactor_scheduler_base
 
 private:
     void
-    run_task(std::unique_lock<std::mutex>& lock, context_type* ctx) override;
+    run_task(std::unique_lock<std::mutex>& lock, context_type* ctx,
+        long timeout_us) override;
     void interrupt_reactor() const override;
     long calculate_timeout(long requested_timeout_us) const;
 
@@ -285,9 +286,10 @@ kqueue_scheduler::calculate_timeout(long requested_timeout_us) const
 
 inline void
 kqueue_scheduler::run_task(
-    std::unique_lock<std::mutex>& lock, context_type* ctx)
+    std::unique_lock<std::mutex>& lock, context_type* ctx, long timeout_us)
 {
-    long effective_timeout_us = task_interrupted_ ? 0 : calculate_timeout(-1);
+    long effective_timeout_us =
+        task_interrupted_ ? 0 : calculate_timeout(timeout_us);
 
     if (lock.owns_lock())
         lock.unlock();

include/boost/corosio/native/detail/reactor/reactor_scheduler.hpp

@@ -279,7 +279,8 @@ class reactor_scheduler_base
 
     /// Run the platform-specific reactor poll.
     virtual void
-    run_task(std::unique_lock<std::mutex>& lock, context_type* ctx) = 0;
+    run_task(std::unique_lock<std::mutex>& lock, context_type* ctx,
+        long timeout_us) = 0;
 
     /// Wake a blocked reactor (e.g. write to eventfd or pipe).
     virtual void interrupt_reactor() const = 0;
@@ -775,15 +776,16 @@ reactor_scheduler_base::do_one(
                 return 0;
             }
 
-            task_interrupted_ = more_handlers || timeout_us == 0;
+            long task_timeout_us = more_handlers ? 0 : timeout_us;
+            task_interrupted_ = task_timeout_us == 0;
             task_running_.store(true, std::memory_order_release);
 
             if (more_handlers)
                 unlock_and_signal_one(lock);
 
             try
             {
-                run_task(lock, ctx);
+                run_task(lock, ctx, task_timeout_us);
             }
             catch (...)
             {
@@ -793,6 +795,8 @@ reactor_scheduler_base::do_one(
 
             task_running_.store(false, std::memory_order_relaxed);
             completed_ops_.push(&task_op_);
+            if (timeout_us > 0)
+                return 0;
             continue;
         }
 

include/boost/corosio/native/detail/select/select_scheduler.hpp

@@ -127,7 +127,8 @@ class BOOST_COROSIO_DECL select_scheduler final : public reactor_scheduler_base
 
 private:
     void
-    run_task(std::unique_lock<std::mutex>& lock, context_type* ctx) override;
+    run_task(std::unique_lock<std::mutex>& lock, context_type* ctx,
+        long timeout_us) override;
     void interrupt_reactor() const override;
     long calculate_timeout(long requested_timeout_us) const;
 
@@ -302,9 +303,10 @@ select_scheduler::calculate_timeout(long requested_timeout_us) const
 
 inline void
 select_scheduler::run_task(
-    std::unique_lock<std::mutex>& lock, context_type* ctx)
+    std::unique_lock<std::mutex>& lock, context_type* ctx, long timeout_us)
 {
-    long effective_timeout_us = task_interrupted_ ? 0 : calculate_timeout(-1);
+    long effective_timeout_us =
+        task_interrupted_ ? 0 : calculate_timeout(timeout_us);
 
     // Snapshot registered descriptors while holding lock.
     // Record which fds need write monitoring to avoid a hot loop:

test/unit/io_context.cpp

@@ -463,6 +463,50 @@ struct io_context_test
         BOOST_TEST(counter == 1);
     }
 
+    void testRunForWithOutstandingWork()
+    {
+        io_context ioc;
+        auto ex = ioc.get_executor();
+
+        // Simulate persistent outstanding work (like a listening acceptor)
+        ex.on_work_started();
+
+        auto start    = std::chrono::steady_clock::now();
+        std::size_t n = ioc.run_for(std::chrono::milliseconds(200));
+        auto elapsed  = std::chrono::steady_clock::now() - start;
+
+        auto ms = std::chrono::duration_cast<std::chrono::milliseconds>(elapsed)
+                      .count();
+
+        // Must return after ~200ms, not block forever
+        BOOST_TEST(n == 0);
+        BOOST_TEST(ms >= 150);
+        BOOST_TEST(ms < 1000);
+
+        ex.on_work_finished();
+    }
+
+    void testRunOneForWithOutstandingWork()
+    {
+        io_context ioc;
+        auto ex = ioc.get_executor();
+
+        ex.on_work_started();
+
+        auto start    = std::chrono::steady_clock::now();
+        std::size_t n = ioc.run_one_for(std::chrono::milliseconds(200));
+        auto elapsed  = std::chrono::steady_clock::now() - start;
+
+        auto ms = std::chrono::duration_cast<std::chrono::milliseconds>(elapsed)
+                      .count();
+
+        BOOST_TEST(n == 0);
+        BOOST_TEST(ms >= 150);
+        BOOST_TEST(ms < 1000);
+
+        ex.on_work_finished();
+    }
+
     void testExecutorRunningInThisThread()
     {
         io_context ioc;
@@ -610,6 +654,8 @@ struct io_context_test
         testRunOneFor();
         testRunOneUntil();
         testRunFor();
+        testRunForWithOutstandingWork();
+        testRunOneForWithOutstandingWork();
         testExecutorRunningInThisThread();
         testMultithreaded();
         testMultithreadedStress();