Fix crash in PropagateRemoteMkdir

Fixes: #9170

changelog/unreleased/9170

@@ -0,0 +1,6 @@
+Bugfix: Crash when handling locked files
+
+We fixed a crash that could occur when trying to add a locked folder to the databse.
+
+
+https://github.com/owncloud/client/issues/9170

src/libsync/propagateremotemkdir.cpp

@@ -172,12 +172,13 @@ void PropagateRemoteMkdir::success()
             done(SyncFileItem::FatalError, result->error());
         });
 
-        watcher->setFuture(QtConcurrent::run([p = propagator(), itemCopy] {
+        // make sure job still exists, the concurrent run might still be ongoing after the job was killed
+        watcher->setFuture(QtConcurrent::run([job = QPointer<PropagateRemoteMkdir>(this), itemCopy] {
             // Try to update the meta data with a 30s timeout
             const auto start = std::chrono::steady_clock::now();
-            while ((std::chrono::steady_clock::now() - start) < UpdateMetaDataRetyTimeOut) {
+            while (job && (std::chrono::steady_clock::now() - start) < UpdateMetaDataRetyTimeOut) {
                 QThread::sleep(1);
-                auto result = std::make_shared<resultType>(p->updateMetadata(itemCopy));
+                auto result = std::make_shared<resultType>(job->propagator()->updateMetadata(itemCopy));
                 if (*result) {
                     if (result->get() == Vfs::ConvertToPlaceholderResult::Locked) {
                         continue;