id = 650
title = "Monitor device_add triggers deadlock when calling drain_call_rcu on QEMU >= 6.0.0"
state = "closed"
created_at = "2021-09-30T12:31:08.233Z"
closed_at = "2022-01-25T09:06:57.879Z"
labels = ["Regression", "Stable::to backport", "kind::Bug"]
url = "https://gitlab.com/qemu-project/qemu/-/issues/650"
host-os = "Debian bullseye/sid"
host-arch = "amd64"
qemu-version = "6.1.50 (v6.1.0-936-gba0fa56bc0)"
guest-os = "n/a"
guest-arch = "n/a"
description = """It hangs"""
reproduce = """1. Run the QEMU:
   ```
   ./qemu-system-mips64 -nographic
   ```
2. Enter into the QEMU monitor: press ctrl-a c
3. Execute command `device_add` without arguments:
```
(qemu) device_add
```
4. It hangs so bad that only `kill -9` helps"""
additional = """I didn't test versions between 4.2.0 and 6.0.0, but I can confirm that 6.0.0, 6.1.0 and the latest master pull have this bug, while version 4.2.0 doesn't have it.

I've tracked the problem and found this.

1. Command `device_add` calls function `drain_call_rcu`. `drain_call_rcu` waits indefinitely for drain_complete_event.
2. Function `cpu_exec` in accel/tcg/cpu-exec.c calls `rcu_read_lock` but does not call `rcu_read_unlock()`. `cpu_exec` just spins in its inner loop.
3. Function `call_rcu_thread` hanged in calling the `synchronize_rcu` which calls `wait_for_readers`.

If I execute `stop` command in QEMU monitor before calling `device_add` command, no hang happen."""