qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Peter Maydell <peter.maydell@linaro.org>
To: qemu-devel@nongnu.org
Cc: qemu-s390x@nongnu.org, "Thomas Huth" <thuth@redhat.com>,
	"Richard Henderson" <richard.henderson@linaro.org>,
	"Philippe Mathieu-Daudé" <philmd@linaro.org>,
	"David Hildenbrand" <david@redhat.com>,
	"Ilya Leoshkevich" <iii@linux.ibm.com>,
	"Cornelia Huck" <cohuck@redhat.com>,
	"Halil Pasic" <pasic@linux.ibm.com>,
	"Eric Farman" <farman@linux.ibm.com>,
	"Christian Borntraeger" <borntraeger@linux.ibm.com>,
	"Nico Boehr" <nrb@linux.ibm.com>,
	"Nina Schoetterl-Glausch" <nsg@linux.ibm.com>
Subject: [PATCH v2 10/11] hw/core/qdev: Simplify legacy_reset handling
Date: Fri, 30 Aug 2024 15:58:11 +0100	[thread overview]
Message-ID: <20240830145812.1967042-11-peter.maydell@linaro.org> (raw)
In-Reply-To: <20240830145812.1967042-1-peter.maydell@linaro.org>

Now that all devices which still implement a the legacy reset method
register it via device_class_legacy_reset(), we can simplify the
handling of these devices.  Instead of using the complex
Resettable::get_transitional_function machinery, we register a hold
phase method which invokes the DeviceClass::legacy_reset method.

This will allow us to remove all the get_transitional_function
handling from resettable.c.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
 hw/core/qdev.c | 55 ++++++++++++++++++++------------------------------
 1 file changed, 22 insertions(+), 33 deletions(-)

diff --git a/hw/core/qdev.c b/hw/core/qdev.c
index 9af0ed3e1b7..db36f54d914 100644
--- a/hw/core/qdev.c
+++ b/hw/core/qdev.c
@@ -747,35 +747,6 @@ device_vmstate_if_get_id(VMStateIf *obj)
     return qdev_get_dev_path(dev);
 }
 
-static void device_transitional_reset(Object *obj)
-{
-    DeviceClass *dc = DEVICE_GET_CLASS(obj);
-
-    /*
-     * Device still using DeviceClass legacy_reset method. This doesn't
-     * reset children. device_get_transitional_reset() checked that
-     * this isn't NULL.
-     */
-    dc->legacy_reset(DEVICE(obj));
-}
-
-/**
- * device_get_transitional_reset:
- * check if the device's class is ready for multi-phase
- */
-static ResettableTrFunction device_get_transitional_reset(Object *obj)
-{
-    DeviceClass *dc = DEVICE_GET_CLASS(obj);
-    if (dc->legacy_reset) {
-        /*
-         * dc->reset has been overridden by a subclass,
-         * the device is not ready for multi phase yet.
-         */
-        return device_transitional_reset;
-    }
-    return NULL;
-}
-
 static void device_class_init(ObjectClass *class, void *data)
 {
     DeviceClass *dc = DEVICE_CLASS(class);
@@ -800,12 +771,9 @@ static void device_class_init(ObjectClass *class, void *data)
      * A NULL legacy_reset implies a three-phase reset device. Devices can
      * only be reset using three-phase aware mechanisms, but we still support
      * for transitional purposes leaf classes which set the old legacy_reset
-     * method via device_class_set_legacy_reset(). If they do so, then
-     * device_get_transitional_reset() will notice and arrange for the
-     * DeviceClass::legacy_reset() method to be called during the hold phase.
+     * method via device_class_set_legacy_reset().
      */
     dc->legacy_reset = NULL;
-    rc->get_transitional_function = device_get_transitional_reset;
 
     object_class_property_add_bool(class, "realized",
                                    device_get_realized, device_set_realized);
@@ -817,8 +785,29 @@ static void device_class_init(ObjectClass *class, void *data)
                                    offsetof(DeviceState, parent_bus), NULL, 0);
 }
 
+static void do_legacy_reset(Object *obj, ResetType type)
+{
+    DeviceClass *dc = DEVICE_GET_CLASS(obj);
+
+    dc->legacy_reset(DEVICE(obj));
+}
+
 void device_class_set_legacy_reset(DeviceClass *dc, DeviceReset dev_reset)
 {
+    /*
+     * A legacy DeviceClass::reset has identical semantics to the
+     * three-phase "hold" method, with no "enter" or "exit"
+     * behaviour. Classes that use this legacy function must be leaf
+     * classes that do not chain up to their parent class reset.
+     * There is no mechanism for resetting a device that does not
+     * use the three-phase APIs, so the only place which calls
+     * the legacy_reset hook is do_legacy_reset().
+     */
+    ResettableClass *rc = RESETTABLE_CLASS(dc);
+
+    rc->phases.enter = NULL;
+    rc->phases.hold = do_legacy_reset;
+    rc->phases.exit = NULL;
     dc->legacy_reset = dev_reset;
 }
 
-- 
2.34.1



  parent reply	other threads:[~2024-08-30 14:59 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-08-30 14:58 [PATCH v2 00/11] s390: Convert virtio-ccw, cpu to three-phase reset, and followup cleanup Peter Maydell
2024-08-30 14:58 ` [PATCH v2 01/11] hw/s390/ccw-device: Convert to three-phase reset Peter Maydell
2024-09-03 12:33   ` Nina Schoetterl-Glausch
2024-09-03 14:53   ` Philippe Mathieu-Daudé
2024-09-04  6:21   ` Thomas Huth
2024-08-30 14:58 ` [PATCH v2 02/11] hw/s390/virtio-ccw: " Peter Maydell
2024-08-30 14:58 ` [PATCH v2 03/11] target/s390: Convert CPU to Resettable interface Peter Maydell
2024-08-30 14:58 ` [PATCH v2 04/11] hw: Remove device_class_set_parent_reset() Peter Maydell
2024-08-30 14:58 ` [PATCH v2 05/11] target/alpha, hppa: Remove unused parent_reset fields Peter Maydell
2024-09-03 14:42   ` Philippe Mathieu-Daudé
2024-08-30 14:58 ` [PATCH v2 06/11] hw: Define new device_class_set_legacy_reset() Peter Maydell
2024-08-30 14:58 ` [PATCH v2 07/11] hw: Use device_class_set_legacy_reset() instead of opencoding Peter Maydell
2024-08-30 14:58 ` [PATCH v2 08/11] hw: Rename DeviceClass::reset field to legacy_reset Peter Maydell
2024-08-30 14:58 ` [PATCH v2 09/11] hw: Remove device_phases_reset() Peter Maydell
2024-08-30 14:58 ` Peter Maydell [this message]
2024-09-03 14:52   ` [PATCH v2 10/11] hw/core/qdev: Simplify legacy_reset handling Philippe Mathieu-Daudé
2024-08-30 14:58 ` [PATCH v2 11/11] hw/core/resettable: Remove transitional_function machinery Peter Maydell
2024-09-03 14:52   ` Philippe Mathieu-Daudé
2024-09-06 14:38 ` [PATCH v2 00/11] s390: Convert virtio-ccw, cpu to three-phase reset, and followup cleanup Peter Maydell
2024-09-09 13:47   ` Nina Schoetterl-Glausch

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20240830145812.1967042-11-peter.maydell@linaro.org \
    --to=peter.maydell@linaro.org \
    --cc=borntraeger@linux.ibm.com \
    --cc=cohuck@redhat.com \
    --cc=david@redhat.com \
    --cc=farman@linux.ibm.com \
    --cc=iii@linux.ibm.com \
    --cc=nrb@linux.ibm.com \
    --cc=nsg@linux.ibm.com \
    --cc=pasic@linux.ibm.com \
    --cc=philmd@linaro.org \
    --cc=qemu-devel@nongnu.org \
    --cc=qemu-s390x@nongnu.org \
    --cc=richard.henderson@linaro.org \
    --cc=thuth@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).