From f036077547b86438d35f20cb63c4c7f6153cc1be Mon Sep 17 00:00:00 2001
From: Richard Hodges <hodges.r@gmail.com>
Date: Tue, 7 Jul 2020 12:31:45 +0200
Subject: [PATCH] Fix race in http-crawl

---
 CHANGELOG.md                             |  4 ++
 example/http/client/crawl/http_crawl.cpp | 62 +++++++++++-------------
 2 files changed, 31 insertions(+), 35 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 16211ef5..2ac3ed44 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,3 +1,7 @@
+* Fix race in http-crawl example.
+
+--------------------------------------------------------------------------------
+
 Version 298:
 
 * Support BOOST_ASIO_NO_TS_EXECUTORS.
diff --git a/example/http/client/crawl/http_crawl.cpp b/example/http/client/crawl/http_crawl.cpp
index 5a621d14..da925207 100644
--- a/example/http/client/crawl/http_crawl.cpp
+++ b/example/http/client/crawl/http_crawl.cpp
@@ -46,7 +46,6 @@ using tcp = net::ip::tcp;               // from <boost/asio/ip/tcp.hpp>
 // This structure aggregates statistics on all the sites
 class crawl_report
 {
-    net::io_context& ioc_;
     net::strand<
         net::io_context::executor_type> strand_;
     std::atomic<std::size_t> index_;
@@ -55,8 +54,7 @@ class crawl_report
 
 public:
     crawl_report(net::io_context& ioc)
-        : ioc_(ioc)
-        , strand_(ioc_.get_executor())
+        : strand_(ioc.get_executor())
         , index_(0)
         , hosts_(urls_large_data())
     {
@@ -150,6 +148,7 @@ class worker : public std::enable_shared_from_this<worker>
     };
 
     crawl_report& report_;
+    net::strand<net::io_context::executor_type> ex_;
     tcp::resolver resolver_;
     beast::tcp_stream stream_;
     beast::flat_buffer buffer_; // (Must persist between reads)
@@ -164,8 +163,9 @@ public:
         crawl_report& report,
         net::io_context& ioc)
         : report_(report)
-        , resolver_(net::make_strand(ioc))
-        , stream_(net::make_strand(ioc))
+        , ex_(net::make_strand(ioc.get_executor()))
+        , resolver_(ex_)
+        , stream_(ex_)
     {
         // Set up the common fields of the request
         req_.version(11);
@@ -347,21 +347,14 @@ int main(int argc, char* argv[])
         std::cerr <<
             "Usage: http-crawl <threads>\n" <<
             "Example:\n" <<
-            "    http-crawl 100 1\n";
+            "    http-crawl 100\n";
         return EXIT_FAILURE;
     }
     auto const threads = std::max<int>(1, std::atoi(argv[1]));
 
-    // The io_context is required for all I/O
+    // The io_context is used to aggregate the statistics
     net::io_context ioc;
 
-    // Building a tracked executor ensures that the underlying context's
-    // run() function will not return until the tracked executor is destroyed
-    net::any_io_executor work =
-        net::require(
-            ioc.get_executor(),
-            net::execution::outstanding_work.tracked);
-
     // The report holds the aggregated statistics
     crawl_report report{ioc};
 
@@ -371,17 +364,26 @@ int main(int argc, char* argv[])
     std::vector<std::thread> workers;
     workers.reserve(threads + 1);
     for(int i = 0; i < threads; ++i)
+    {
+        // Each worker will eventually add some data to the aggregated
+        // report. Outstanding work is tracked in each worker to
+        // represent the forthcoming delivery of this data by that
+        // worker.
+        auto reporting_work = net::require(
+            ioc.get_executor(),
+            net::execution::outstanding_work.tracked);
+
         workers.emplace_back(
-        [&report]
-        {
-            // We use a separate io_context for each worker because
-            // the asio resolver simulates asynchronous operation using
-            // a dedicated worker thread per io_context, and we want to
-            // do a lot of name resolutions in parallel.
-            net::io_context ioc{1};
-            std::make_shared<worker>(report, ioc)->run();
-            ioc.run();
-        });
+            [&report, reporting_work] {
+                // We use a separate io_context for each worker because
+                // the asio resolver simulates asynchronous operation using
+                // a dedicated worker thread per io_context, and we want to
+                // do a lot of name resolutions in parallel.
+                net::io_context ioc;
+                std::make_shared<worker>(report, ioc)->run();
+                ioc.run();
+            });
+    }
 
     // Add another thread to run the main io_context which
     // is used to aggregate the statistics
@@ -393,17 +395,7 @@ int main(int argc, char* argv[])
 
     // Now block until all threads exit
     for(std::size_t i = 0; i < workers.size(); ++i)
-    {
-        auto& thread = workers[i];
-
-        // If this is the last thread, reset the
-        // work object so that it can return from run.
-        if(i == workers.size() - 1)
-            work = {};
-
-        // Wait for the thread to exit
-        thread.join();
-    }
+        workers[i].join();
 
     std::cout <<
         "Elapsed time:    " << chrono::duration_cast<chrono::seconds>(t.elapsed()).count() << " seconds\n";