From: Anthony Liguori <aliguori@us.ibm.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: aliguori@linux.vnet.ibm.com, andreas.faerber@web.de,
qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [PATCH 2/4] qdev: add children before qdev_init
Date: Tue, 27 Mar 2012 16:08:51 -0500 [thread overview]
Message-ID: <4F722C63.4040203@us.ibm.com> (raw)
In-Reply-To: <1332866328-25443-3-git-send-email-pbonzini@redhat.com>
On 03/27/2012 11:38 AM, Paolo Bonzini wrote:
> We want the composition tree to to be in order by the time we call
> qdev_init, so that a single set of the toplevel realize property can
> propagate all the way down the composition tree.
>
> This is not the case so far. Unfortunately, this is incompatible
> with calling qdev_init in the constructor wrappers for devices,
> so for now we need to unattach some devices that are created through
> those wrappers. This will be fixed by removing qdev_init and instead
> setting the toplevel realize property after machine init.
>
> Signed-off-by: Paolo Bonzini<pbonzini@redhat.com>
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Regards,
Anthony Liguori
> ---
> hw/pc_piix.c | 18 +-----------------
> hw/piix_pci.c | 3 +--
> hw/ppc_prep.c | 2 +-
> hw/qdev-monitor.c | 8 ++++----
> 4 files changed, 7 insertions(+), 24 deletions(-)
>
> diff --git a/hw/pc_piix.c b/hw/pc_piix.c
> index 3f99f9a..747c40f 100644
> --- a/hw/pc_piix.c
> +++ b/hw/pc_piix.c
> @@ -146,7 +146,6 @@ static void pc_init1(MemoryRegion *system_memory,
> MemoryRegion *ram_memory;
> MemoryRegion *pci_memory;
> MemoryRegion *rom_memory;
> - DeviceState *dev;
>
> pc_cpus_init(cpu_model);
>
> @@ -224,11 +223,7 @@ static void pc_init1(MemoryRegion *system_memory,
>
> pc_register_ferr_irq(gsi[13]);
>
> - dev = pc_vga_init(isa_bus, pci_enabled ? pci_bus : NULL);
> - if (dev) {
> - object_property_add_child(object_get_root(), "vga", OBJECT(dev), NULL);
> - }
> -
> + pc_vga_init(isa_bus, pci_enabled ? pci_bus : NULL);
> if (xen_enabled()) {
> pci_create_simple(pci_bus, -1, "xen-platform");
> }
> @@ -255,17 +250,6 @@ static void pc_init1(MemoryRegion *system_memory,
> }
> idebus[0] = qdev_get_child_bus(&dev->qdev, "ide.0");
> idebus[1] = qdev_get_child_bus(&dev->qdev, "ide.1");
> -
> - /* FIXME there's some major spaghetti here. Somehow we create the
> - * devices on the PIIX before we actually create it. We create the
> - * PIIX3 deep in the recess of the i440fx creation too and then lose
> - * the DeviceState.
> - *
> - * For now, let's "fix" this by making judicious use of paths. This
> - * is not generally the right way to do this.
> - */
> - object_property_add_child(object_resolve_path("/i440fx/piix3", NULL),
> - "rtc", (Object *)rtc_state, NULL);
> } else {
> for(i = 0; i< MAX_IDE_BUS; i++) {
> ISADevice *dev;
> diff --git a/hw/piix_pci.c b/hw/piix_pci.c
> index e0268fe..9017565 100644
> --- a/hw/piix_pci.c
> +++ b/hw/piix_pci.c
> @@ -276,8 +276,8 @@ static PCIBus *i440fx_common_init(const char *device_name,
> b = pci_bus_new(&s->busdev.qdev, NULL, pci_address_space,
> address_space_io, 0);
> s->bus = b;
> - qdev_init_nofail(dev);
> object_property_add_child(object_get_root(), "i440fx", OBJECT(dev), NULL);
> + qdev_init_nofail(dev);
>
> d = pci_create_simple(b, 0, device_name);
> *pi440fx_state = DO_UPCAST(PCII440FXState, dev, d);
> @@ -316,7 +316,6 @@ static PCIBus *i440fx_common_init(const char *device_name,
> pci_bus_irqs(b, piix3_set_irq, pci_slot_get_pirq, piix3,
> PIIX_NUM_PIRQS);
> }
> - object_property_add_child(OBJECT(dev), "piix3", OBJECT(piix3), NULL);
> piix3->pic = pic;
> *isa_bus = DO_UPCAST(ISABus, qbus,
> qdev_get_child_bus(&piix3->dev.qdev, "isa.0"));
> diff --git a/hw/ppc_prep.c b/hw/ppc_prep.c
> index 06d589d..86c9336 100644
> --- a/hw/ppc_prep.c
> +++ b/hw/ppc_prep.c
> @@ -615,8 +615,8 @@ static void ppc_prep_init (ram_addr_t ram_size,
> sys = sysbus_from_qdev(dev);
> pcihost = DO_UPCAST(PCIHostState, busdev, sys);
> pcihost->address_space = get_system_memory();
> - qdev_init_nofail(dev);
> object_property_add_child(object_get_root(), "raven", OBJECT(dev), NULL);
> + qdev_init_nofail(dev);
> pci_bus = (PCIBus *)qdev_get_child_bus(dev, "pci.0");
> if (pci_bus == NULL) {
> fprintf(stderr, "Couldn't create PCI host controller.\n");
> diff --git a/hw/qdev-monitor.c b/hw/qdev-monitor.c
> index 2e82962..031cb83 100644
> --- a/hw/qdev-monitor.c
> +++ b/hw/qdev-monitor.c
> @@ -458,10 +458,6 @@ DeviceState *qdev_device_add(QemuOpts *opts)
> qdev_free(qdev);
> return NULL;
> }
> - if (qdev_init(qdev)< 0) {
> - qerror_report(QERR_DEVICE_INIT_FAILED, driver);
> - return NULL;
> - }
> if (qdev->id) {
> object_property_add_child(qdev_get_peripheral(), qdev->id,
> OBJECT(qdev), NULL);
> @@ -472,6 +468,10 @@ DeviceState *qdev_device_add(QemuOpts *opts)
> OBJECT(qdev), NULL);
> g_free(name);
> }
> + if (qdev_init(qdev)< 0) {
> + qerror_report(QERR_DEVICE_INIT_FAILED, driver);
> + return NULL;
> + }
> qdev->opts = opts;
> return qdev;
> }
next prev parent reply other threads:[~2012-03-27 21:10 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-03-27 16:38 [Qemu-devel] [PATCH 0/4] qdev: give all devices a canonical QOM path Paolo Bonzini
2012-03-27 16:38 ` [Qemu-devel] [PATCH 1/4] qom: add container_get Paolo Bonzini
2012-03-27 21:07 ` Anthony Liguori
2012-03-27 16:38 ` [Qemu-devel] [PATCH 2/4] qdev: add children before qdev_init Paolo Bonzini
2012-03-27 21:08 ` Anthony Liguori [this message]
2012-03-27 16:38 ` [Qemu-devel] [PATCH 3/4] qdev: give all devices a canonical path Paolo Bonzini
2012-03-27 21:12 ` Anthony Liguori
2012-03-27 16:38 ` [Qemu-devel] [PATCH 4/4] qdev: put all devices under /machine Paolo Bonzini
2012-03-27 21:11 ` Anthony Liguori
2012-03-28 14:01 ` Andreas Färber
2012-03-28 14:34 ` [Qemu-devel] [PATCH v2 " Paolo Bonzini
2012-03-28 15:10 ` Andreas Färber
2012-03-28 15:19 ` Paolo Bonzini
2012-04-05 11:21 ` [Qemu-devel] [PATCH] qom: Refine container_get() to allow using a custom root Andreas Färber
2012-04-05 11:30 ` Paolo Bonzini
2012-03-27 16:42 ` [Qemu-devel] [PATCH 0/4] qdev: give all devices a canonical QOM path Anthony Liguori
2012-04-02 21:21 ` Anthony Liguori
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=4F722C63.4040203@us.ibm.com \
--to=aliguori@us.ibm.com \
--cc=aliguori@linux.vnet.ibm.com \
--cc=andreas.faerber@web.de \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.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 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).