qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Anthony Liguori <anthony@codemonkey.ws>
To: Liu Ping Fan <qemulist@gmail.com>, qemu-devel@nongnu.org
Cc: Paolo Bonzini <pbonzini@redhat.com>,
	Liu Ping Fan <pingfank@linux.vnet.ibm.com>,
	Avi Kivity <avi@redhat.com>
Subject: Re: [Qemu-devel] [PATCH 10/10] qdev: fix create in place obj's life cycle problem
Date: Mon, 27 Aug 2012 08:19:55 -0500	[thread overview]
Message-ID: <874nnoh3pg.fsf@codemonkey.ws> (raw)
In-Reply-To: <1345801763-24227-11-git-send-email-qemulist@gmail.com>

Liu Ping Fan <qemulist@gmail.com> writes:

> From: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>
> Scene:
>   obja lies in objA, when objA's ref->0, it will be freed,
> but at that time obja can still be in use.
>
> The real example is:
> typedef struct PCIIDEState {
>     PCIDevice dev;
>     IDEBus bus[2]; --> create in place
>     .....
> }
>
> When without big lock protection for mmio-dispatch, we will hold
> obj's refcnt. So memory_region_init_io() will replace the third para
> "void *opaque" with "Object *obj".
> With this patch, we can protect PCIIDEState from disappearing during
> mmio-dispatch hold the IDEBus->ref.
>
> And the ref circle has been broken when calling qdev_delete_subtree().
>
> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>

I think this is solving the wrong problem.  There are many, many
dependencies a device may have on other devices.  Memory allocation
isn't the only one.

The problem is that we want to make sure that a device doesn't "go away"
while an MMIO dispatch is happening.  This is easy to solve without
touching referencing counting.

The device will hold a lock while the MMIO is being dispatched.  The
delete path simply needs to acquire that same lock.  This will ensure
that a delete operation cannot finish while MMIO is still in flight.

Regarding deleting a device, not all devices are capable of being
deleted and specifically, devices that are composed within the memory of
another device cannot be directly deleted (they can only be deleted
as part of their parent's destruction).

Regards,

Anthony Liguori

> ---
>  hw/qdev.c |    2 ++
>  hw/qdev.h |    1 +
>  2 files changed, 3 insertions(+), 0 deletions(-)
>
> diff --git a/hw/qdev.c b/hw/qdev.c
> index e2339a1..b09ebbf 100644
> --- a/hw/qdev.c
> +++ b/hw/qdev.c
> @@ -510,6 +510,8 @@ void qbus_create_inplace(BusState *bus, const char *typename,
>  {
>      object_initialize(bus, typename);
>  
> +    bus->overlap = parent;
> +    object_ref(OBJECT(bus->overlap));
>      bus->parent = parent;
>      bus->name = name ? g_strdup(name) : NULL;
>      qbus_realize(bus);
> diff --git a/hw/qdev.h b/hw/qdev.h
> index 182cfa5..9bc5783 100644
> --- a/hw/qdev.h
> +++ b/hw/qdev.h
> @@ -117,6 +117,7 @@ struct BusState {
>      int allow_hotplug;
>      bool qom_allocated;
>      bool glib_allocated;
> +    DeviceState *overlap;
>      int max_index;
>      QTAILQ_HEAD(ChildrenHead, BusChild) children;
>      QLIST_ENTRY(BusState) sibling;
> -- 
> 1.7.4.4

  parent reply	other threads:[~2012-08-27 13:20 UTC|newest]

Thread overview: 82+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-08-24  9:49 [Qemu-devel] [PATCH 0/10] rework on hot unplug Liu Ping Fan
2012-08-24  9:49 ` [Qemu-devel] [PATCH 01/10] qom: add, remove of link property need to ref, unref its target Liu Ping Fan
2012-08-24 14:52   ` Paolo Bonzini
2012-08-24  9:49 ` [Qemu-devel] [PATCH 02/10] qdev: change iterator callback seq Liu Ping Fan
2012-08-24  9:49 ` [Qemu-devel] [PATCH 03/10] qom: export object_property_is_child, object_property_is_link Liu Ping Fan
2012-08-24 14:51   ` Paolo Bonzini
2012-08-25  7:43     ` liu ping fan
2012-08-25  8:04   ` Blue Swirl
2012-08-24  9:49 ` [Qemu-devel] [PATCH 04/10] qdev: introduce new interface to remove composite sub-tree Liu Ping Fan
2012-08-24  9:49 ` [Qemu-devel] [PATCH 05/10] qdev: finalize of qbus, qdev will not the right place to free children Liu Ping Fan
2012-08-24 14:50   ` Paolo Bonzini
2012-08-24  9:49 ` [Qemu-devel] [PATCH 06/10] qom: expose object_property_del_child Liu Ping Fan
2012-08-24 14:44   ` Paolo Bonzini
2012-08-24  9:49 ` [Qemu-devel] [PATCH 07/10] unplug: using new intf qdev_delete_subtree in acpi_piix_eject_slot Liu Ping Fan
2012-08-24 10:24   ` Paolo Bonzini
2012-08-25  7:05     ` liu ping fan
2012-08-24  9:49 ` [Qemu-devel] [PATCH 08/10] qdev: rename qdev_unplug to qdev_unplug_req Liu Ping Fan
2012-08-24 14:48   ` Paolo Bonzini
2012-08-24  9:49 ` [Qemu-devel] [PATCH 09/10] mon: release dev's ref hold by qdev_get_peripheral Liu Ping Fan
2012-08-24  9:49 ` [Qemu-devel] [PATCH 10/10] qdev: fix create in place obj's life cycle problem Liu Ping Fan
2012-08-24 14:42   ` Paolo Bonzini
2012-08-25  7:42     ` liu ping fan
2012-08-27  7:01       ` Paolo Bonzini
2012-08-27  7:47         ` Jan Kiszka
2012-08-27  8:17           ` liu ping fan
2012-08-27  8:27             ` Jan Kiszka
2012-08-27 17:09           ` Avi Kivity
2012-08-27 17:14             ` Jan Kiszka
2012-08-27 18:09               ` Avi Kivity
2012-08-27 18:17                 ` Jan Kiszka
2012-08-27 18:20                   ` Avi Kivity
2012-08-27 18:39                     ` Jan Kiszka
2012-08-27 18:52                       ` Avi Kivity
2012-08-27 19:38                         ` Jan Kiszka
2012-08-27 20:53                           ` Avi Kivity
2012-08-28  1:01                             ` Jan Kiszka
2012-08-29 17:13                               ` Avi Kivity
2012-08-29 17:21                                 ` Jan Kiszka
2012-08-29 17:27                                   ` Avi Kivity
2012-08-29 17:41                                     ` Jan Kiszka
2012-09-03  9:09                                       ` Avi Kivity
2012-08-28  3:09                           ` liu ping fan
2012-08-28  3:38                             ` liu ping fan
2012-08-28  9:42                               ` Jan Kiszka
2012-08-28 10:05                                 ` Paolo Bonzini
2012-08-29 17:23                                 ` Avi Kivity
2012-08-29 17:30                                   ` Jan Kiszka
2012-08-29 17:40                                     ` Avi Kivity
2012-08-29 17:49                                       ` Jan Kiszka
2012-09-01  8:31                                         ` Avi Kivity
2012-09-01  8:57                                           ` Jan Kiszka
2012-09-01  9:30                                             ` Avi Kivity
2012-08-30  5:54                                       ` liu ping fan
2012-08-30  7:08                                         ` Jan Kiszka
2012-08-30  7:47                                           ` liu ping fan
2012-09-01  8:46                                           ` Avi Kivity
2012-09-03  7:44                                             ` liu ping fan
2012-09-03  8:52                                               ` Avi Kivity
2012-09-03 10:06                                                 ` liu ping fan
2012-09-03 10:16                                                   ` Avi Kivity
2012-09-04  2:33                                                     ` liu ping fan
2012-09-04  2:34                                                   ` liu ping fan
2012-09-05  8:19                                                     ` liu ping fan
2012-09-05  9:52                                                       ` Avi Kivity
2012-09-05 10:36                                                         ` Jan Kiszka
2012-09-05 10:53                                                           ` Avi Kivity
2012-09-05 11:11                                                             ` Jan Kiszka
2012-09-05 11:25                                                               ` Avi Kivity
2012-09-05 12:02                                                                 ` Jan Kiszka
2012-09-05 12:17                                                                   ` Avi Kivity
2012-08-27 13:19   ` Anthony Liguori [this message]
2012-08-27 15:02     ` Jan Kiszka
2012-08-27 15:14       ` Anthony Liguori
2012-08-27 15:26         ` Jan Kiszka
2012-08-27 16:24           ` Anthony Liguori
2012-08-27 16:59             ` Jan Kiszka
2012-08-27 18:35             ` Avi Kivity
2012-08-27 19:17               ` Anthony Liguori
2012-08-27 19:22                 ` Jan Kiszka
2012-08-27 20:58                 ` Avi Kivity
2012-08-27 21:34                   ` Paolo Bonzini
2012-08-27 18:27     ` Avi Kivity

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=874nnoh3pg.fsf@codemonkey.ws \
    --to=anthony@codemonkey.ws \
    --cc=avi@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=pingfank@linux.vnet.ibm.com \
    --cc=qemu-devel@nongnu.org \
    --cc=qemulist@gmail.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).