Fix progress token is already registered crash

After we started reporting progress when running cargo check during
loading, it is possible to crash the client with two identical progress
tokens.

This points to a deeper issue: we might be running several cargo checks
concurrently, which doesn't make sense.

This commit linearizes all workspace fetches, making sure no updates are
lost.

As an additional touch, it also normalizes progress & result reporting,
to make sure they stand in sync.

1 files changed, 25 insertions, 0 deletions

diff --git a/crates/rust-analyzer/src/op_queue.rs b/crates/rust-analyzer/src/op_queue.rs
new file mode 100644
index 000000000..51d66f4b3
--- /dev/null
+++ b/crates/rust-analyzer/src/op_queue.rs
@@ -0,0 +1,25 @@
+//! Bookkeeping to make sure only one long-running operation is executed.
+
+#[derive(Default)]
+pub(crate) struct OpQueue {
+    op_scheduled: bool,
+    op_in_progress: bool,
+}
+
+impl OpQueue {
+    pub(crate) fn request_op(&mut self) {
+        self.op_scheduled = true;
+    }
+    pub(crate) fn should_start_op(&mut self) -> bool {
+        if !self.op_in_progress && self.op_scheduled {
+            self.op_in_progress = true;
+            self.op_scheduled = false;
+            return true;
+        }
+        false
+    }
+    pub(crate) fn op_completed(&mut self) {
+        assert!(self.op_in_progress);
+        self.op_in_progress = false;
+    }
+}