From: Igor Mammedov <imammedo@redhat.com>
To: "Marc-André Lureau" <marcandre.lureau@gmail.com>
Cc: "Stefano Stabellini" <sstabellini@kernel.org>,
"Eduardo Habkost" <ehabkost@redhat.com>,
QEMU <qemu-devel@nongnu.org>,
xen-devel@lists.xenproject.org,
"Anthony Perard" <anthony.perard@citrix.com>,
"Paolo Bonzini" <pbonzini@redhat.com>,
"Andreas Färber" <afaerber@suse.de>
Subject: Re: [Qemu-devel] [PATCH for-3.2 v5 07/19] hw: apply accel compat properties without touching globals
Date: Thu, 13 Dec 2018 13:06:54 +0100 [thread overview]
Message-ID: <20181213130654.1eb72eb1@redhat.com> (raw)
In-Reply-To: <CAJ+F1CJ55pcq6OZdnSBVNGQwRSE=gQD0swPeHw_Roj2qVOnxSA@mail.gmail.com>
On Wed, 12 Dec 2018 16:00:06 +0400
Marc-André Lureau <marcandre.lureau@gmail.com> wrote:
> Hi
> On Mon, Dec 10, 2018 at 8:55 PM Igor Mammedov <imammedo@redhat.com> wrote:
> >
> > On Mon, 10 Dec 2018 17:45:22 +0100
> > Igor Mammedov <imammedo@redhat.com> wrote:
> >
> > > On Tue, 4 Dec 2018 18:20:11 +0400
> > > Marc-André Lureau <marcandre.lureau@redhat.com> wrote:
> > >
> > > > Instead of registering compat properties as globals, let's keep them
> > > > in their own array, to avoid mixing with user globals.
> > > >
> > > > Introduce object_apply_global_props() function, to apply compatibility
> > > > properties from a GPtrArray.
> > > >
> > > > Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
> > > > ---
> > > > include/hw/qdev-core.h | 10 ++++++++++
> > > > include/qom/object.h | 3 +++
> > > > include/sysemu/accel.h | 4 +---
> > > > accel/accel.c | 12 ------------
> > > > hw/core/qdev.c | 9 +++++++++
> > > > hw/xen/xen-common.c | 9 ++++++---
> > > > qom/object.c | 25 +++++++++++++++++++++++++
> > > > vl.c | 1 -
> > > > 8 files changed, 54 insertions(+), 19 deletions(-)
> > > >
> > > [...]
> > > > diff --git a/hw/xen/xen-common.c b/hw/xen/xen-common.c
> > > > index 6ec14c73ca..4532aa8632 100644
> > > > --- a/hw/xen/xen-common.c
> > > > +++ b/hw/xen/xen-common.c
> > > > @@ -174,18 +174,21 @@ static GlobalProperty xen_compat_props[] = {
> > > > .driver = "migration",
> > > > .property = "send-section-footer",
> > > > .value = "off",
> > > > - },
> > > > - { /* end of list */ },
> > > > + }
> > > > };
> > > >
> > > > static void xen_accel_class_init(ObjectClass *oc, void *data)
> > > > {
> > > > AccelClass *ac = ACCEL_CLASS(oc);
> > > > +
> > > > ac->name = "Xen";
> > > > ac->init_machine = xen_init;
> > > > ac->setup_post = xen_setup_post;
> > > > ac->allowed = &xen_allowed;
> > > > - ac->global_props = xen_compat_props;
> > > > + ac->compat_props = g_ptr_array_new();
> > > where is matching free for that?
> > can we at least annotate it somehow so that valgrind won't complain about this leak?
>
> If you check my commits on qemu, you should see that I do care (too
> much?) about leaks :)
>
> In this case though, I don't see valgrind or asan complaining, I guess
> it's still a reachable reference.
> Do you think a FIXME comment would be helpful?
I've looked at other cases were we leak, and well we leak a lot so it's
probably futile exercise. But the comment won't hurt and will work as remainder.
> (/me wish we had a proper object system, GObject, but that ship as
> long sailed..)
>
> >
> > >
> > > > +
> > > > + compat_props_add(ac->compat_props,
> > > > + xen_compat_props, G_N_ELEMENTS(xen_compat_props));
> > > > }
> > > >
> > > > #define TYPE_XEN_ACCEL ACCEL_CLASS_NAME("xen")
> > > > diff --git a/qom/object.c b/qom/object.c
> > > > index 17921c0a71..dbdab0aead 100644
> > > > --- a/qom/object.c
> > > > +++ b/qom/object.c
> > > > @@ -370,6 +370,31 @@ static void object_post_init_with_type(Object *obj, TypeImpl *ti)
> > > > }
> > > > }
> > > >
> > > > +void object_apply_global_props(Object *obj, const GPtrArray *props, Error **errp)
> > > > +{
> > > > + Error *err = NULL;
> > > > + int i;
> > > > +
> > > > + if (!props) {
> > > > + return;
> > > > + }
> > > > +
> > > > + for (i = 0; i < props->len; i++) {
> > > > + GlobalProperty *p = g_ptr_array_index(props, i);
> > > > +
> > > > + if (object_dynamic_cast(obj, p->driver) == NULL) {
> > > > + continue;
> > > > + }
> > > > + p->used = true;
> > > > + object_property_parse(obj, p->value, p->property, &err);
> > > > + if (err != NULL) {
> > > > + error_prepend(&err, "can't apply global %s.%s=%s: ",
> > > > + p->driver, p->property, p->value);
> > > > + error_propagate(errp, err);
> > > > + }
> > > > + }
> > > > +}
> > > > +
> > > > static void object_initialize_with_type(void *data, size_t size, TypeImpl *type)
> > > > {
> > > > Object *obj = data;
> > > > diff --git a/vl.c b/vl.c
> > > > index a5ae5f23d2..88ba658572 100644
> > > > --- a/vl.c
> > > > +++ b/vl.c
> > > > @@ -2968,7 +2968,6 @@ static void user_register_global_props(void)
> > > > */
> > > > static void register_global_properties(MachineState *ms)
> > > > {
> > > > - accel_register_compat_props(ms->accelerator);
> > > > machine_register_compat_props(ms);
> > > > user_register_global_props();
> > > > }
> > >
> > >
> >
> >
>
>
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel
WARNING: multiple messages have this Message-ID (diff)
From: Igor Mammedov <imammedo@redhat.com>
To: "Marc-André Lureau" <marcandre.lureau@gmail.com>
Cc: "Stefano Stabellini" <sstabellini@kernel.org>,
"Eduardo Habkost" <ehabkost@redhat.com>,
QEMU <qemu-devel@nongnu.org>,
"Paolo Bonzini" <pbonzini@redhat.com>,
"Anthony Perard" <anthony.perard@citrix.com>,
xen-devel@lists.xenproject.org,
"Andreas Färber" <afaerber@suse.de>
Subject: Re: [Qemu-devel] [PATCH for-3.2 v5 07/19] hw: apply accel compat properties without touching globals
Date: Thu, 13 Dec 2018 13:06:54 +0100 [thread overview]
Message-ID: <20181213130654.1eb72eb1@redhat.com> (raw)
In-Reply-To: <CAJ+F1CJ55pcq6OZdnSBVNGQwRSE=gQD0swPeHw_Roj2qVOnxSA@mail.gmail.com>
On Wed, 12 Dec 2018 16:00:06 +0400
Marc-André Lureau <marcandre.lureau@gmail.com> wrote:
> Hi
> On Mon, Dec 10, 2018 at 8:55 PM Igor Mammedov <imammedo@redhat.com> wrote:
> >
> > On Mon, 10 Dec 2018 17:45:22 +0100
> > Igor Mammedov <imammedo@redhat.com> wrote:
> >
> > > On Tue, 4 Dec 2018 18:20:11 +0400
> > > Marc-André Lureau <marcandre.lureau@redhat.com> wrote:
> > >
> > > > Instead of registering compat properties as globals, let's keep them
> > > > in their own array, to avoid mixing with user globals.
> > > >
> > > > Introduce object_apply_global_props() function, to apply compatibility
> > > > properties from a GPtrArray.
> > > >
> > > > Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
> > > > ---
> > > > include/hw/qdev-core.h | 10 ++++++++++
> > > > include/qom/object.h | 3 +++
> > > > include/sysemu/accel.h | 4 +---
> > > > accel/accel.c | 12 ------------
> > > > hw/core/qdev.c | 9 +++++++++
> > > > hw/xen/xen-common.c | 9 ++++++---
> > > > qom/object.c | 25 +++++++++++++++++++++++++
> > > > vl.c | 1 -
> > > > 8 files changed, 54 insertions(+), 19 deletions(-)
> > > >
> > > [...]
> > > > diff --git a/hw/xen/xen-common.c b/hw/xen/xen-common.c
> > > > index 6ec14c73ca..4532aa8632 100644
> > > > --- a/hw/xen/xen-common.c
> > > > +++ b/hw/xen/xen-common.c
> > > > @@ -174,18 +174,21 @@ static GlobalProperty xen_compat_props[] = {
> > > > .driver = "migration",
> > > > .property = "send-section-footer",
> > > > .value = "off",
> > > > - },
> > > > - { /* end of list */ },
> > > > + }
> > > > };
> > > >
> > > > static void xen_accel_class_init(ObjectClass *oc, void *data)
> > > > {
> > > > AccelClass *ac = ACCEL_CLASS(oc);
> > > > +
> > > > ac->name = "Xen";
> > > > ac->init_machine = xen_init;
> > > > ac->setup_post = xen_setup_post;
> > > > ac->allowed = &xen_allowed;
> > > > - ac->global_props = xen_compat_props;
> > > > + ac->compat_props = g_ptr_array_new();
> > > where is matching free for that?
> > can we at least annotate it somehow so that valgrind won't complain about this leak?
>
> If you check my commits on qemu, you should see that I do care (too
> much?) about leaks :)
>
> In this case though, I don't see valgrind or asan complaining, I guess
> it's still a reachable reference.
> Do you think a FIXME comment would be helpful?
I've looked at other cases were we leak, and well we leak a lot so it's
probably futile exercise. But the comment won't hurt and will work as remainder.
> (/me wish we had a proper object system, GObject, but that ship as
> long sailed..)
>
> >
> > >
> > > > +
> > > > + compat_props_add(ac->compat_props,
> > > > + xen_compat_props, G_N_ELEMENTS(xen_compat_props));
> > > > }
> > > >
> > > > #define TYPE_XEN_ACCEL ACCEL_CLASS_NAME("xen")
> > > > diff --git a/qom/object.c b/qom/object.c
> > > > index 17921c0a71..dbdab0aead 100644
> > > > --- a/qom/object.c
> > > > +++ b/qom/object.c
> > > > @@ -370,6 +370,31 @@ static void object_post_init_with_type(Object *obj, TypeImpl *ti)
> > > > }
> > > > }
> > > >
> > > > +void object_apply_global_props(Object *obj, const GPtrArray *props, Error **errp)
> > > > +{
> > > > + Error *err = NULL;
> > > > + int i;
> > > > +
> > > > + if (!props) {
> > > > + return;
> > > > + }
> > > > +
> > > > + for (i = 0; i < props->len; i++) {
> > > > + GlobalProperty *p = g_ptr_array_index(props, i);
> > > > +
> > > > + if (object_dynamic_cast(obj, p->driver) == NULL) {
> > > > + continue;
> > > > + }
> > > > + p->used = true;
> > > > + object_property_parse(obj, p->value, p->property, &err);
> > > > + if (err != NULL) {
> > > > + error_prepend(&err, "can't apply global %s.%s=%s: ",
> > > > + p->driver, p->property, p->value);
> > > > + error_propagate(errp, err);
> > > > + }
> > > > + }
> > > > +}
> > > > +
> > > > static void object_initialize_with_type(void *data, size_t size, TypeImpl *type)
> > > > {
> > > > Object *obj = data;
> > > > diff --git a/vl.c b/vl.c
> > > > index a5ae5f23d2..88ba658572 100644
> > > > --- a/vl.c
> > > > +++ b/vl.c
> > > > @@ -2968,7 +2968,6 @@ static void user_register_global_props(void)
> > > > */
> > > > static void register_global_properties(MachineState *ms)
> > > > {
> > > > - accel_register_compat_props(ms->accelerator);
> > > > machine_register_compat_props(ms);
> > > > user_register_global_props();
> > > > }
> > >
> > >
> >
> >
>
>
next prev parent reply other threads:[~2018-12-13 12:06 UTC|newest]
Thread overview: 65+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-12-04 14:20 [Qemu-devel] [PATCH for-3.2 v5 00/19] Generalize machine compatibility properties Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 01/19] tests: qdev_prop_check_globals() doesn't return "all_used" Marc-André Lureau
2018-12-04 14:20 ` [Qemu-arm] [PATCH for-3.2 v5 02/19] qom: make interface types abstract Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] " Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 03/19] qom: make user_creatable_complete() specific to UserCreatable Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 04/19] accel: register global_props like machine globals Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 05/19] qdev: move qdev_prop_register_global_list() to tests Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 06/19] qom: remove unimplemented class_finalize Marc-André Lureau
2018-12-04 14:20 ` [PATCH for-3.2 v5 07/19] hw: apply accel compat properties without touching globals Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] " Marc-André Lureau
2018-12-10 16:45 ` Igor Mammedov
2018-12-10 16:45 ` [Qemu-devel] " Igor Mammedov
2018-12-10 16:54 ` Igor Mammedov
2018-12-10 16:54 ` Igor Mammedov
2018-12-12 12:00 ` Marc-André Lureau
2018-12-12 12:00 ` Marc-André Lureau
2018-12-13 12:06 ` Igor Mammedov [this message]
2018-12-13 12:06 ` Igor Mammedov
2018-12-11 13:14 ` Igor Mammedov
2018-12-11 13:14 ` Igor Mammedov
2018-12-04 14:20 ` [Qemu-arm] [PATCH for-3.2 v5 08/19] hw: apply machine " Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] " Marc-André Lureau
2018-12-10 17:31 ` [Qemu-arm] " Eduardo Habkost
2018-12-10 17:31 ` Eduardo Habkost
2018-12-11 12:07 ` Marc-André Lureau
2018-12-11 12:07 ` Marc-André Lureau
2018-12-11 14:02 ` [Qemu-arm] " Eduardo Habkost
2018-12-11 14:02 ` Eduardo Habkost
2018-12-11 14:23 ` [Qemu-arm] " Eduardo Habkost
2018-12-11 14:23 ` Eduardo Habkost
2018-12-11 14:30 ` Marc-André Lureau
2018-12-11 14:30 ` Marc-André Lureau
2018-12-11 15:52 ` [Qemu-arm] " Igor Mammedov
2018-12-11 15:52 ` Igor Mammedov
2018-12-11 17:43 ` [Qemu-arm] " Eduardo Habkost
2018-12-11 17:43 ` Eduardo Habkost
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 09/19] hw: remove SET_MACHINE_COMPAT Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 10/19] qdev: make a separate helper function to apply compat properties Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 11/19] qdev: all globals are now user-provided Marc-André Lureau
2018-12-10 17:00 ` Igor Mammedov
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 12/19] qdev-props: convert global_props to GPtrArray Marc-André Lureau
2018-12-10 17:05 ` Igor Mammedov
2018-12-11 12:12 ` Marc-André Lureau
2018-12-11 13:03 ` Igor Mammedov
2018-12-11 13:04 ` Igor Mammedov
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 13/19] qdev-props: remove errp from GlobalProperty Marc-André Lureau
2018-12-10 17:20 ` Igor Mammedov
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 14/19] qdev-props: call object_apply_global_props() Marc-André Lureau
2018-12-10 17:28 ` Igor Mammedov
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 15/19] qom: add object_class_get_class_data() Marc-André Lureau
2018-12-11 16:02 ` Igor Mammedov
2018-12-12 18:48 ` Marc-André Lureau
2018-12-04 14:20 ` [Qemu-arm] [PATCH for-3.2 v5 16/19] RFC: arm: replace instance_post_init() Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] " Marc-André Lureau
2018-12-11 13:43 ` [Qemu-arm] " Igor Mammedov
2018-12-11 13:43 ` Igor Mammedov
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 17/19] hw/i386: add pc-i440fx-4.0 & pc-q35-4.0 Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 18/19] hw/arm/virt: add virt-4.0 machine type Marc-André Lureau
2018-12-04 14:20 ` Marc-André Lureau
2018-12-04 14:20 ` [Qemu-devel] [PATCH for-3.2 v5 19/19] hostmem: use object id for memory region name with >= 4.0 Marc-André Lureau
2018-12-04 14:22 ` [Qemu-devel] [PATCH for-3.2 v5 00/19] Generalize machine compatibility properties Marc-André Lureau
2018-12-10 17:07 ` Eduardo Habkost
2018-12-10 17:31 ` Igor Mammedov
2018-12-10 17:39 ` Eduardo Habkost
2018-12-11 15:11 ` Igor Mammedov
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=20181213130654.1eb72eb1@redhat.com \
--to=imammedo@redhat.com \
--cc=afaerber@suse.de \
--cc=anthony.perard@citrix.com \
--cc=ehabkost@redhat.com \
--cc=marcandre.lureau@gmail.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=sstabellini@kernel.org \
--cc=xen-devel@lists.xenproject.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.