Re: [PATCH 3/5] hw/core: Remove uses of QERR_DEVICE_NO_HOTPLUG

[Damien Hedde]

On 11/19/21 09:20, Markus Armbruster wrote:
Philippe Mathieu-Daudé <phi...@redhat.com> writes:

QERR_DEVICE_NO_HOTPLUG definition is obsolete since 2015 (commit
4629ed1e989, "qerror: Finally unused, clean up"). Replace the two
uses and drop the definition.

Signed-off-by: Philippe Mathieu-Daudé <phi...@redhat.com>
---
  include/qapi/qmp/qerror.h | 3 ---
  hw/core/qdev.c            | 3 ++-
  softmmu/qdev-monitor.c    | 2 +-
  3 files changed, 3 insertions(+), 5 deletions(-)

diff --git a/include/qapi/qmp/qerror.h b/include/qapi/qmp/qerror.h
index 596fce0c54e..f49ae01cdb0 100644
--- a/include/qapi/qmp/qerror.h
+++ b/include/qapi/qmp/qerror.h
@@ -26,9 +26,6 @@
  #define QERR_DEVICE_IN_USE \
      "Device '%s' is in use"
  
-#define QERR_DEVICE_NO_HOTPLUG \
-    "Device '%s' does not support hotplugging"
-
  #define QERR_FEATURE_DISABLED \
      "The feature '%s' is not enabled"
  
diff --git a/hw/core/qdev.c b/hw/core/qdev.c
index 7f06403752d..14375861c36 100644
--- a/hw/core/qdev.c
+++ b/hw/core/qdev.c
@@ -734,7 +734,8 @@ static void device_set_realized(Object *obj, bool value, 
Error **errp)
      static int unattached_count;
  
      if (dev->hotplugged && !dc->hotpluggable) {
-        error_setg(errp, QERR_DEVICE_NO_HOTPLUG, object_get_typename(obj));
+        error_setg(errp, "Device '%s' does not support hotplugging",
+                   object_get_typename(obj));
          return;
      }
  
diff --git a/softmmu/qdev-monitor.c b/softmmu/qdev-monitor.c
index 4851de51a5c..35a885a6623 100644
--- a/softmmu/qdev-monitor.c
+++ b/softmmu/qdev-monitor.c
@@ -896,7 +896,7 @@ void qdev_unplug(DeviceState *dev, Error **errp)
      }
  
      if (!dc->hotpluggable) {
-        error_setg(errp, QERR_DEVICE_NO_HOTPLUG,
+        error_setg(errp, "Device '%s' does not support hotplugging",
                     object_get_typename(OBJECT(dev)));
          return;
      }

When the same error is detected in multiple places, I like to ask myself
whether the code calls for a refactoring.  But I can't see a useful one
here.


Since the realize check will catch more case (devices created 
recursively by a command or simply by C code), we could remove the 
softmmu/qdev-monitor.c check and rely on realize catching that just after.
In theory failing during realize should have the same result as "not 
trying to create the device in the first place".

--
Damien