Issue 1596: Runner throws null ref exception when new line after EOF is missing (#1687)

* Issue 1596: runner throws nullref exception when writting env var

* Adding tests for missing new line after EOF marker

* Changing newline to new line

.github/workflows/build.yml

@@ -50,13 +50,6 @@ jobs:
         ${{ matrix.devScript }} layout Release ${{ matrix.runtime }}
       working-directory: src
 
-    # Run tests
-    - name: L0
-      run: |
-        ${{ matrix.devScript }} test
-      working-directory: src
-      if: matrix.runtime != 'linux-arm64' && matrix.runtime != 'linux-arm'
-
     # Check runtime/externals hash
     - name: Compute/Compare runtime and externals Hash
       shell: bash
@@ -80,6 +73,13 @@ jobs:
         DOTNET_RUNTIME_HASH: ${{hashFiles('**/_layout_trims/runtime/**/*')}}
         EXTERNALS_HASH: ${{hashFiles('**/_layout_trims/externals/**/*')}}
 
+    # Run tests
+    - name: L0
+      run: |
+        ${{ matrix.devScript }} test
+      working-directory: src
+      if: matrix.runtime != 'linux-arm64' && matrix.runtime != 'linux-arm'
+
     # Create runner package tar.gz/zip
     - name: Package Release
       if: github.event_name != 'pull_request'

src/Misc/contentHash/externals/linux-arm

@@ -1 +1 @@
-6ca4a0e1c50b7079ead05321dcf5835c1c25f23dc632add8c1c4667d416d103e
+6ed30a2c1ee403a610d63e82bb230b9ba846a9c25cec9e4ea8672fb6ed4e1a51

src/Misc/contentHash/externals/linux-arm64

@@ -1 +1 @@
-b5951dc607d782d9c7571a7224e940eb0975bb23c54ff25c7afdbf959a417081
+711c30c51ec52c9b7a9a2eb399d6ab2ab5ee1dc72de11879f2f36f919f163d78

src/Misc/contentHash/externals/linux-x64

@@ -1 +1 @@
-af819e92011cc9cbca90e8299f9f7651f2cf6bf45b42920f9a4ca22795486147
+a49479ca4b4988a06c097e8d22c51fd08a11c13f40807366236213d0e008cf6a

src/Misc/contentHash/externals/osx-x64

@@ -1 +1 @@
-aa0e6bf4bfaabf48c962ea3b262dca042629ab332005f73d282faec908847036
+8e97df75230b843462a9b4c578ccec604ee4b4a1066120c85b04374317fa372b

src/Misc/contentHash/externals/win-x64

@@ -1 +1 @@
-40328cff2b8229f9b578f32739183bd8f6aab481c21dadc052b09f1c7e8e4665
+f75a671e5a188c76680739689aa75331a2c09d483dce9c80023518c48fd67a18

src/Misc/expressionFunc/hashFiles/src/hashFiles.ts

@@ -53,3 +53,11 @@ async function run(): Promise<void> {
 }
 
 run()
+  .then(out => {
+    console.log(out)
+    process.exit(0)
+  })
+  .catch(err => {
+    console.error(err)
+    process.exit(1)
+  })

src/Misc/externals.sh

@@ -3,7 +3,7 @@ PACKAGERUNTIME=$1
 PRECACHE=$2
 
 NODE_URL=https://nodejs.org/dist
-NODE12_VERSION="12.13.1"
+NODE12_VERSION="12.22.7"
 NODE16_VERSION="16.13.0"
 
 get_abs_path() {
@@ -143,7 +143,7 @@ fi
 # Download the external tools for Linux PACKAGERUNTIMEs.
 if [[ "$PACKAGERUNTIME" == "linux-x64" ]]; then
     acquireExternalTool "$NODE_URL/v${NODE12_VERSION}/node-v${NODE12_VERSION}-linux-x64.tar.gz" node12 fix_nested_dir
-    acquireExternalTool "https://vstsagenttools.blob.core.windows.net/tools/nodejs/${NODE12_VERSION}/alpine/x64/node-${NODE12_VERSION}-alpine-x64.tar.gz" node12_alpine
+    acquireExternalTool "https://vstsagenttools.blob.core.windows.net/tools/nodejs/${NODE12_VERSION}/alpine/x64/node-v${NODE12_VERSION}-alpine-x64.tar.gz" node12_alpine
     acquireExternalTool "$NODE_URL/v${NODE16_VERSION}/node-v${NODE16_VERSION}-linux-x64.tar.gz" node16 fix_nested_dir
     acquireExternalTool "https://vstsagenttools.blob.core.windows.net/tools/nodejs/${NODE16_VERSION}/alpine/x64/node-v${NODE16_VERSION}-alpine-x64.tar.gz" node16_alpine
 fi

src/Misc/layoutbin/RunnerService.js

@@ -3,94 +3,136 @@
 // Licensed under the MIT license. See LICENSE file in the project root for full license information.
 
 var childProcess = require("child_process");
-var path = require("path")
+var path = require("path");
+const { exit } = require("process");
 
-var supported = ['linux', 'darwin']
+var supported = ["linux", "darwin"];
 
 if (supported.indexOf(process.platform) == -1) {
-    console.log('Unsupported platform: ' + process.platform);
-    console.log('Supported platforms are: ' + supported.toString());
+  console.log("Unsupported platform: " + process.platform);
+  console.log("Supported platforms are: " + supported.toString());
   process.exit(1);
 }
 
 var stopping = false;
 var listener = null;
 
+var exitServiceAfterNFailures = Number(
+  process.env.GITHUB_ACTIONS_SERVICE_EXIT_AFTER_N_FAILURES
+);
+
+if (exitServiceAfterNFailures <= 0) {
+  exitServiceAfterNFailures = NaN;
+}
+
+var consecutiveFailureCount = 0;
+
+var gracefulShutdown = function (code) {
+  console.log("Shutting down runner listener");
+  stopping = true;
+  if (listener) {
+    console.log("Sending SIGINT to runner listener to stop");
+    listener.kill("SIGINT");
+
+    console.log("Sending SIGKILL to runner listener");
+    setTimeout(() => listener.kill("SIGKILL"), 30000).unref();
+  }
+};
+
 var runService = function () {
-    var listenerExePath = path.join(__dirname, '../bin/Runner.Listener');
+  var listenerExePath = path.join(__dirname, "../bin/Runner.Listener");
   var interactive = process.argv[2] === "interactive";
 
   if (!stopping) {
     try {
       if (interactive) {
-                console.log('Starting Runner listener interactively');
-                listener = childProcess.spawn(listenerExePath, ['run'], { env: process.env });
+        console.log("Starting Runner listener interactively");
+        listener = childProcess.spawn(listenerExePath, ["run"], {
+          env: process.env,
+        });
       } else {
-                console.log('Starting Runner listener with startup type: service');
-                listener = childProcess.spawn(listenerExePath, ['run', '--startuptype', 'service'], { env: process.env });
+        console.log("Starting Runner listener with startup type: service");
+        listener = childProcess.spawn(
+          listenerExePath,
+          ["run", "--startuptype", "service"],
+          { env: process.env }
+        );
       }
 
       console.log(`Started listener process, pid: ${listener.pid}`);
 
-            listener.stdout.on('data', (data) => {
-                process.stdout.write(data.toString('utf8'));
+      listener.stdout.on("data", (data) => {
+        if (data.toString("utf8").includes("Listening for Jobs")) {
+          consecutiveFailureCount = 0;
+        }
+        process.stdout.write(data.toString("utf8"));
       });
 
-            listener.stderr.on('data', (data) => {
-                process.stdout.write(data.toString('utf8'));
+      listener.stderr.on("data", (data) => {
+        process.stdout.write(data.toString("utf8"));
       });
 
       listener.on("error", (err) => {
         console.log(`Runner listener fail to start with error ${err.message}`);
       });
 
-            listener.on('close', (code) => {
+      listener.on("close", (code) => {
         console.log(`Runner listener exited with error code ${code}`);
 
         if (code === 0) {
-                    console.log('Runner listener exit with 0 return code, stop the service, no retry needed.');
+          console.log(
+            "Runner listener exit with 0 return code, stop the service, no retry needed."
+          );
           stopping = true;
         } else if (code === 1) {
-                    console.log('Runner listener exit with terminated error, stop the service, no retry needed.');
+          console.log(
+            "Runner listener exit with terminated error, stop the service, no retry needed."
+          );
           stopping = true;
         } else if (code === 2) {
-                    console.log('Runner listener exit with retryable error, re-launch runner in 5 seconds.');
-                } else if (code === 3) {
-                    console.log('Runner listener exit because of updating, re-launch runner in 5 seconds.');
+          console.log(
+            "Runner listener exit with retryable error, re-launch runner in 5 seconds."
+          );
+          consecutiveFailureCount = 0;
+        } else if (code === 3 || code === 4) {
+          console.log(
+            "Runner listener exit because of updating, re-launch runner in 5 seconds."
+          );
+          consecutiveFailureCount = 0;
         } else {
-                    console.log('Runner listener exit with undefined return code, re-launch runner in 5 seconds.');
+          var messagePrefix = "Runner listener exit with undefined return code";
+          consecutiveFailureCount++;
+          if (
+            !isNaN(exitServiceAfterNFailures) &&
+            consecutiveFailureCount >= exitServiceAfterNFailures
+          ) {
+            console.error(
+              `${messagePrefix}, exiting service after ${consecutiveFailureCount} consecutive failures`
+            );
+            gracefulShutdown(5);
+            return;
+          } else {
+            console.log(`${messagePrefix}, re-launch runner in 5 seconds.`);
+          }
         }
 
         if (!stopping) {
           setTimeout(runService, 5000);
         }
       });
-
     } catch (ex) {
       console.log(ex);
     }
   }
-}
+};
 
 runService();
-console.log('Started running service');
+console.log("Started running service");
 
-var gracefulShutdown = function (code) {
-    console.log('Shutting down runner listener');
-    stopping = true;
-    if (listener) {
-        console.log('Sending SIGINT to runner listener to stop');
-        listener.kill('SIGINT');
-
-        console.log('Sending SIGKILL to runner listener');
-        setTimeout(() => listener.kill('SIGKILL'), 30000).unref();
-    }
-}
-
-process.on('SIGINT', () => {
+process.on("SIGINT", () => {
   gracefulShutdown(0);
 });
 
-process.on('SIGTERM', () => {
+process.on("SIGTERM", () => {
   gracefulShutdown(0);
 });

src/Misc/layoutbin/hashFiles/index.js

@@ -1621,7 +1621,15 @@ function run() {
         }
     });
 }
-run();
+run()
+    .then(out => {
+    console.log(out);
+    process.exit(0);
+})
+    .catch(err => {
+    console.error(err);
+    process.exit(1);
+});
 
 
 /***/ }),

src/Runner.Common/HostContext.cs

@@ -216,6 +216,8 @@ namespace GitHub.Runner.Common
                 _userAgents.Add(new ProductInfoHeaderValue("RunnerId", runnerSettings.AgentId.ToString(CultureInfo.InvariantCulture)));
                 _userAgents.Add(new ProductInfoHeaderValue("GroupId", runnerSettings.PoolId.ToString(CultureInfo.InvariantCulture)));
             }
+
+            _userAgents.Add(new ProductInfoHeaderValue("CommitSHA", BuildConstants.Source.CommitHash));
         }
 
         public string GetDirectory(WellKnownDirectory directory)

src/Runner.Listener/JobDispatcher.cs

@@ -285,7 +285,7 @@ namespace GitHub.Runner.Listener
                     {
                         // at this point, the job execution might encounter some dead lock and even not able to be cancelled.
                         // no need to localize the exception string should never happen.
-                        throw new InvalidOperationException($"Job dispatch process for {jobDispatch.JobId} has encountered unexpected error, the dispatch task is not able to be canceled within 45 seconds.");
+                        throw new InvalidOperationException($"Job dispatch process for {jobDispatch.JobId} has encountered unexpected error, the dispatch task is not able to be cancelled within 45 seconds.");
                     }
                 }
                 else
@@ -363,7 +363,7 @@ namespace GitHub.Runner.Listener
                     Trace.Info($"Start renew job request {requestId} for job {message.JobId}.");
                     Task renewJobRequest = RenewJobRequestAsync(_poolId, requestId, lockToken, orchestrationId, firstJobRequestRenewed, lockRenewalTokenSource.Token);
 
-                    // wait till first renew succeed or job request is canceled
+                    // wait till first renew succeed or job request is cancelled
                     // not even start worker if the first renew fail
                     await Task.WhenAny(firstJobRequestRenewed.Task, renewJobRequest, Task.Delay(-1, jobRequestCancellationToken));
 
@@ -704,7 +704,7 @@ namespace GitHub.Runner.Listener
                 {
                     // OperationCanceledException may caused by http timeout or _lockRenewalTokenSource.Cance();
                     // Stop renew only on cancellation token fired.
-                    Trace.Info($"job renew has been canceled, stop renew job request {requestId}.");
+                    Trace.Info($"job renew has been cancelled, stop renew job request {requestId}.");
                     return;
                 }
                 catch (Exception ex)
@@ -762,7 +762,7 @@ namespace GitHub.Runner.Listener
                         }
                         catch (OperationCanceledException) when (token.IsCancellationRequested)
                         {
-                            Trace.Info($"job renew has been canceled, stop renew job request {requestId}.");
+                            Trace.Info($"job renew has been cancelled, stop renew job request {requestId}.");
                         }
                     }
                     else

src/Runner.Listener/SelfUpdater.cs

@@ -504,7 +504,7 @@ namespace GitHub.Runner.Listener
                     }
                     catch (OperationCanceledException) when (token.IsCancellationRequested)
                     {
-                        Trace.Info($"Runner download has been canceled.");
+                        Trace.Info($"Runner download has been cancelled.");
                         throw;
                     }
                     catch (Exception ex)

src/Runner.Plugins/Repository/v1.0/GitSourceProvider.cs

@@ -166,7 +166,7 @@ namespace GitHub.Runner.Plugins.Repository.v1_0
             }
             else
             {
-                // delete the index.lock file left by previous canceled build or any operation cause git.exe crash last time.
+                // delete the index.lock file left by previous cancelled build or any operation cause git.exe crash last time.
                 string lockFile = Path.Combine(targetPath, ".git\\index.lock");
                 if (File.Exists(lockFile))
                 {
@@ -181,7 +181,7 @@ namespace GitHub.Runner.Plugins.Repository.v1_0
                     }
                 }
 
-                // delete the shallow.lock file left by previous canceled build or any operation cause git.exe crash last time.		
+                // delete the shallow.lock file left by previous cancelled build or any operation cause git.exe crash last time.		
                 string shallowLockFile = Path.Combine(targetPath, ".git\\shallow.lock");
                 if (File.Exists(shallowLockFile))
                 {

src/Runner.Plugins/Repository/v1.1/GitSourceProvider.cs

@@ -150,7 +150,7 @@ namespace GitHub.Runner.Plugins.Repository.v1_1
             }
             else
             {
-                // delete the index.lock file left by previous canceled build or any operation cause git.exe crash last time.
+                // delete the index.lock file left by previous cancelled build or any operation cause git.exe crash last time.
                 string lockFile = Path.Combine(targetPath, ".git\\index.lock");
                 if (File.Exists(lockFile))
                 {
@@ -165,7 +165,7 @@ namespace GitHub.Runner.Plugins.Repository.v1_1
                     }
                 }
 
-                // delete the shallow.lock file left by previous canceled build or any operation cause git.exe crash last time.		
+                // delete the shallow.lock file left by previous cancelled build or any operation cause git.exe crash last time.		
                 string shallowLockFile = Path.Combine(targetPath, ".git\\shallow.lock");
                 if (File.Exists(shallowLockFile))
                 {

src/Runner.Sdk/Util/IOUtil.cs

@@ -108,7 +108,7 @@ namespace GitHub.Runner.Sdk
             }
 
             // Create a new token source for the parallel query. The parallel query should be
-            // canceled after the first error is encountered. Otherwise the number of exceptions
+            // cancelled after the first error is encountered. Otherwise the number of exceptions
             // could get out of control for a large directory with access denied on every file.
             using (var tokenSource = CancellationTokenSource.CreateLinkedTokenSource(cancellationToken))
             {

src/Runner.Worker/ActionManager.cs

@@ -654,7 +654,7 @@ namespace GitHub.Runner.Worker
                     actionDownloadInfos = await jobServer.ResolveActionDownloadInfoAsync(executionContext.Global.Plan.ScopeIdentifier, executionContext.Global.Plan.PlanType, executionContext.Global.Plan.PlanId, executionContext.Root.Id, new WebApi.ActionReferenceList { Actions = actionReferences }, executionContext.CancellationToken);
                     break;
                 }
-                catch (Exception ex) when (!executionContext.CancellationToken.IsCancellationRequested) // Do not retry if the run is canceled.
+                catch (Exception ex) when (!executionContext.CancellationToken.IsCancellationRequested) // Do not retry if the run is cancelled.
                 {
                     // UnresolvableActionDownloadInfoException is a 422 client error, don't retry
                     // Some possible cases are:

src/Runner.Worker/FileCommandManager.cs

@@ -181,6 +181,10 @@ namespace GitHub.Runner.Worker
                                 {
                                     throw new Exception($"Invalid environment variable value. Matching delimiter not found '{delimiter}'");
                                 }
+                                if (newline == null)
+                                {
+                                    throw new Exception($"Invalid environment variable value. EOF marker missing new line.");
+                                }
                                 endIndex = index - newline.Length;
                                 tempLine = ReadLine(text, ref index, out newline);
                             }

src/Runner.Worker/Handlers/CompositeActionHandler.cs

@@ -299,7 +299,7 @@ namespace GitHub.Runner.Worker.Handlers
                     // Register job cancellation call back only if job cancellation token not been fire before each step run
                     if (!ExecutionContext.Root.CancellationToken.IsCancellationRequested)
                     {
-                        // Test the condition again. The job was canceled after the condition was originally evaluated.
+                        // Test the condition again. The job was cancelled after the condition was originally evaluated.
                         jobCancelRegister = ExecutionContext.Root.CancellationToken.Register(() =>
                         {
                             // Mark job as cancelled
@@ -399,7 +399,7 @@ namespace GitHub.Runner.Worker.Handlers
                         jobCancelRegister = null;
                     }
                 }
-                // Check failed or canceled
+                // Check failed or cancelled
                 if (step.ExecutionContext.Result == TaskResult.Failed || step.ExecutionContext.Result == TaskResult.Canceled)
                 {
                     Trace.Info($"Update job result with current composite step result '{step.ExecutionContext.Result}'.");

src/Runner.Worker/JobRunner.cs

@@ -131,9 +131,9 @@ namespace GitHub.Runner.Worker
                 }
                 catch (OperationCanceledException ex) when (jobContext.CancellationToken.IsCancellationRequested)
                 {
-                    // set the job to canceled
+                    // set the job to cancelled
                     // don't log error issue to job ExecutionContext, since server owns the job level issue
-                    Trace.Error($"Job is canceled during initialize.");
+                    Trace.Error($"Job is cancelled during initialize.");
                     Trace.Error($"Caught exception: {ex}");
                     return await CompleteJobAsync(jobServer, jobContext, message, TaskResult.Canceled);
                 }

src/Runner.Worker/StepsRunner.cs

@@ -130,7 +130,7 @@ namespace GitHub.Runner.Worker
                         // Register job cancellation call back only if job cancellation token not been fire before each step run
                         if (!jobContext.CancellationToken.IsCancellationRequested)
                         {
-                            // Test the condition again. The job was canceled after the condition was originally evaluated.
+                            // Test the condition again. The job was cancelled after the condition was originally evaluated.
                             jobCancelRegister = jobContext.CancellationToken.Register(() =>
                             {
                                 // Mark job as cancelled

src/Test/L0/Listener/SelfUpdaterL0.cs

@@ -774,8 +774,15 @@ namespace GitHub.Runner.Common.Tests.Listener
 
                     var traceFile = Path.GetTempFileName();
                     File.Copy(hc.TraceFileName, traceFile, true);
+                    if (File.ReadAllText(traceFile).Contains("Use trimmed (runtime+externals) package"))
+                    {
                         Assert.Contains("Something wrong with the trimmed runner package, failback to use the full package for runner updates", File.ReadAllText(traceFile));
                     }
+                    else
+                    {
+                        hc.GetTrace().Warning("Skipping the 'TestSelfUpdateAsync_FallbackToFullPackage' test, as the `externals` or `runtime` hashes have been updated");
+                    }
+                }
             }
             finally
             {

src/Test/L0/Worker/SetEnvFileCommandL0.cs

@@ -302,6 +302,50 @@ namespace GitHub.Runner.Common.Tests.Worker
             }
         }
 
+        [Fact]
+        [Trait("Level", "L0")]
+        [Trait("Category", "Worker")]
+        public void SetEnvFileCommand_Heredoc_MissingNewLine()
+        {
+            using (var hostContext = Setup())
+            {
+                var envFile = Path.Combine(_rootDirectory, "heredoc");
+                var content = new List<string>
+                {
+                    "MY_ENV<<EOF",
+                    "line one",
+                    "line two",
+                    "line three",
+                    "EOF",
+                };
+                WriteContent(envFile, content, " ");
+                var ex = Assert.Throws<Exception>(() => _setEnvFileCommand.ProcessCommand(_executionContext.Object, envFile, null));
+                Assert.Contains("Matching delimiter not found", ex.Message);
+            }
+        }
+
+        [Fact]
+        [Trait("Level", "L0")]
+        [Trait("Category", "Worker")]
+        public void SetEnvFileCommand_Heredoc_MissingNewLineMultipleLinesEnv()
+        {
+            using (var hostContext = Setup())
+            {
+                var envFile = Path.Combine(_rootDirectory, "heredoc");
+                var content = new List<string>
+                {
+                    "MY_ENV<<EOF",
+                    @"line one
+                    line two
+                    line three",
+                    "EOF",
+                };
+                WriteContent(envFile, content, " ");
+                var ex = Assert.Throws<Exception>(() => _setEnvFileCommand.ProcessCommand(_executionContext.Object, envFile, null));
+                Assert.Contains("EOF marker missing new line", ex.Message);
+            }
+        }
+
 #if OS_WINDOWS
         [Fact]
         [Trait("Level", "L0")]