From 060c9de07fe6670135f6d577e3a217652c6b3995 Mon Sep 17 00:00:00 2001
From: Philipp Schuster <philipp.schuster@cyberus-technology.de>
Date: Mon, 19 May 2025 11:23:47 +0200
Subject: [PATCH] vmm: introduce nice error messages on exit (CHV and
 ch-remote)

With the foundations of each error type implementing std::error::Error,
we can now nicely walk the `.source()` chain and print an error trace.

This commit introduces improved user-facing error printing when:
- Cloud Hypervisor fails with an error
- ch-remote fails (client error)
- ch-remote fails (remote error)

The additional context is a clear improvement in UX for both users and
developers. In the following example, the new behaviour is shown for
a direct invocation of Cloud Hypervisor leading to a failure. This
looks similar for ch-remote.

```
Old Style
`target/release/cloud-hypervisor --api-socket /tmp/chv2.sock --kernel /etc/bootitems/linux/kernel_minimal/stable.bzImage --cmdline console=ttyS0 --serial tty --console off --disk path=img.raw --initramfs /etc/bootitems/linux/initrd_minimal/default`

Error booting VM: VmBoot(LockingError(BlockError(LockDiskImage(AlreadyLocked)))
```

```
`target/release/cloud-hypervisor --api-socket /tmp/chv2.sock --kernel /etc/bootitems/linux/kernel_minimal/stable.bzImage --cmdline console=ttyS0 --serial tty --console off --disk path=img.raw --initramfs
/etc/bootitems/linux/initrd_minimal/default`

Error: Cloud Hypervisor exited with the following chain of errors:
  0: Error booting VM
  1: The VM could not boot
  2: Error locking disk images: Another instance likely holds a lock
  3: Cannot lock images of all block devices
  4: Failed to get Write lock for disk image: ./img.raw
  5: The file is already locked

Debug Info: VmBoot(VmBoot(LockingError(DiskLockError(LockDiskImage { error: AlreadyLocked, lock_type: Write, path: "./raw_disk.bin" })))
```

Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
---
 src/bin/ch-remote.rs    |  4 ++--
 src/lib.rs              | 27 +++++++++++++++++++++++++++
 src/main.rs             |  4 ++--
 vmm/src/api/http/mod.rs | 13 ++++++++++++-
 4 files changed, 43 insertions(+), 5 deletions(-)
 create mode 100644 src/lib.rs

diff --git a/src/bin/ch-remote.rs b/src/bin/ch-remote.rs
index f0c9b5e09..4ef0b0257 100644
--- a/src/bin/ch-remote.rs
+++ b/src/bin/ch-remote.rs
@@ -1131,8 +1131,8 @@ fn main() {
         }
     };
 
-    if let Err(e) = target_api.do_command(&matches) {
-        eprintln!("Error running command: {e}");
+    if let Err(top_error) = target_api.do_command(&matches) {
+        cloud_hypervisor::cli_print_error_chain(&top_error, "ch-remote");
         process::exit(1)
     };
 }
diff --git a/src/lib.rs b/src/lib.rs
new file mode 100644
index 000000000..ec941ea52
--- /dev/null
+++ b/src/lib.rs
@@ -0,0 +1,27 @@
+// Copyright © 2025 Cyberus Technology GmbH
+//
+// SPDX-License-Identifier: Apache-2.0
+//
+
+/// Prints a chain of errors to the user in a consistent manner.
+/// The user will see a clear chain of errors, followed by debug output
+/// for opening issues.
+pub fn cli_print_error_chain(top_error: &dyn std::error::Error, component: &str) {
+    eprint!("Error: {component} exited with the following ");
+    if top_error.source().is_none() {
+        eprintln!("error:");
+        eprintln!("  {top_error}");
+    } else {
+        eprintln!("chain of errors:");
+        std::iter::successors(Some(top_error), |sub_error| {
+            sub_error.source()
+        })
+        .enumerate()
+        .for_each(|(level, error)| {
+            eprintln!("  {level}: {error}",);
+        });
+    }
+
+    eprintln!();
+    eprintln!("Debug Info: {top_error:?}");
+}
diff --git a/src/main.rs b/src/main.rs
index 0520dcf7c..924cb19c3 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -883,8 +883,8 @@ fn main() {
             path.map(|s| std::fs::remove_file(s).ok());
             0
         }
-        Err(e) => {
-            eprintln!("{e}");
+        Err(top_error) => {
+            cloud_hypervisor::cli_print_error_chain(&top_error, "Cloud Hypervisor");
             1
         }
     };
diff --git a/vmm/src/api/http/mod.rs b/vmm/src/api/http/mod.rs
index 3493a3134..9b2f5d1d8 100644
--- a/vmm/src/api/http/mod.rs
+++ b/vmm/src/api/http/mod.rs
@@ -69,9 +69,20 @@ pub enum HttpError {
 
 const HTTP_ROOT: &str = "/api/v1";
 
+/// Creates the error response's body meant to be sent back to an API client.
+/// The error message contained in the response is supposed to be user-facing,
+/// thus insightful and helpful while balancing technical accuracy and
+/// simplicity.
 pub fn error_response(error: HttpError, status: StatusCode) -> Response {
     let mut response = Response::new(Version::Http11, status);
-    response.set_body(Body::new(format!("{error}")));
+    // We must use debug output here without `#`, as it is currently the only
+    // feasible option to get all relevant error details to the receiver,
+    // i.e., ch-remote, in a balanced form. The Display impl is not guaranteed
+    // to hold all relevant or helpful data.
+    //
+    // TODO: We might print a nice error chain here as well and send it to the
+    // remote, similar to the normal error reporting?
+    response.set_body(Body::new(format!("{error:?}")));
 
     response
 }