qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Thomas Huth <thuth@redhat.com>
To: "Philippe Mathieu-Daudé" <philmd@linaro.org>, qemu-devel@nongnu.org
Cc: "Peter Maydell" <peter.maydell@linaro.org>,
	"Eduardo Habkost" <eduardo@habkost.net>,
	"Cédric Le Goater" <clg@redhat.com>,
	"Paolo Bonzini" <pbonzini@redhat.com>,
	"Markus Armbruster" <armbru@redhat.com>,
	qemu-arm@nongnu.org,
	"Pierrick Bouvier" <pierrick.bouvier@linaro.org>
Subject: Re: [PATCH] hw/arm/bcm283x: Initialize CPU objects in SoC common DeviceRealize()
Date: Wed, 29 Oct 2025 09:01:06 +0100	[thread overview]
Message-ID: <0a951fde-2ea9-4707-b80a-26d33475d35f@redhat.com> (raw)
In-Reply-To: <20251029075257.63635-1-philmd@linaro.org>

On 29/10/2025 08.52, Philippe Mathieu-Daudé wrote:
> QOM .instance_init() handler can not fail. QDev DeviceRealize
> can.
> 
> The device-introspect QTest enumerates all QDev types and
> instantiate each one, without realizing it, then introspects
> the instance properties.
> 
> When switching to a single QEMU binary, all QDev types are
> available in the binary, but only a filtered subset might be
> available, depending on which previous target the binary is
> trying to mimic.
> 
> In particular with the Raspi machines, the TYPE_RASPI4B_MACHINE
> and ARM_CPU_TYPE_NAME("cortex-a72") will be built in the
> qemu-system-arm binary, while not available (because filtered
> as being 64-bit, for the qemu-system-aarch64 binary).
> 
> However the TYPE_BCM2838 SoC is not filtered out, and will
> abort when being initialized, because the "cortex-a72" CPU type
> is filtered out, leading to device-introspect failure:
> 
>    1/1 qemu:qtest+qtest-arm / qtest-arm/device-introspect-test        ERROR            2.46s   killed by signal 6 SIGABRT
>    stderr:
>    unknown type 'cortex-a72-arm-cpu'
>    Broken pipe
>    ../../tests/qtest/libqtest.c:199: kill_qemu() tried to terminate QEMU process but encountered exit status 1 (expected 0)
>    (test program exited with status code -6)
>    TAP parsing error: Too few tests run (expected 167, got 5)
> 
> In order to avoid that, move the CPU *initialization* in the
> SoC DeviceRealize handler, so the SoC initialization won't
> fail, while realization still will.
> 
> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
> ---
>   hw/arm/bcm2836.c | 10 +++++-----
>   1 file changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/hw/arm/bcm2836.c b/hw/arm/bcm2836.c
> index cd61ba15054..6e4066f137d 100644
> --- a/hw/arm/bcm2836.c
> +++ b/hw/arm/bcm2836.c
> @@ -25,12 +25,7 @@ static void bcm283x_base_init(Object *obj)
>   {
>       BCM283XBaseState *s = BCM283X_BASE(obj);
>       BCM283XBaseClass *bc = BCM283X_BASE_GET_CLASS(obj);
> -    int n;
>   
> -    for (n = 0; n < bc->core_count; n++) {
> -        object_initialize_child(obj, "cpu[*]", &s->cpu[n].core,
> -                                bc->cpu_type);
> -    }
>       if (bc->core_count > 1) {
>           qdev_property_add_static(DEVICE(obj), &bcm2836_enabled_cores_property);
>           qdev_prop_set_uint32(DEVICE(obj), "enabled-cpus", bc->core_count);
> @@ -65,6 +60,11 @@ bool bcm283x_common_realize(DeviceState *dev, BCMSocPeripheralBaseState *ps,
>       BCM283XBaseClass *bc = BCM283X_BASE_GET_CLASS(dev);
>       Object *obj;
>   
> +    for (int n = 0; n < bc->core_count; n++) {
> +        object_initialize_child(OBJECT(dev), "cpu[*]", &s->cpu[n].core,
> +                                bc->cpu_type);
> +    }
> +
>       /* common peripherals from bcm2835 */
>   
>       obj = object_property_get_link(OBJECT(dev), "ram", &error_abort);

Reviewed-by: Thomas Huth <thuth@redhat.com>



  reply	other threads:[~2025-10-29  8:02 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-10-29  7:52 [PATCH] hw/arm/bcm283x: Initialize CPU objects in SoC common DeviceRealize() Philippe Mathieu-Daudé
2025-10-29  8:01 ` Thomas Huth [this message]
2025-10-29  8:01 ` Manos Pitsidianakis
2025-10-29 14:14 ` Peter Maydell
2025-10-30  9:54   ` Thomas Huth

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=0a951fde-2ea9-4707-b80a-26d33475d35f@redhat.com \
    --to=thuth@redhat.com \
    --cc=armbru@redhat.com \
    --cc=clg@redhat.com \
    --cc=eduardo@habkost.net \
    --cc=pbonzini@redhat.com \
    --cc=peter.maydell@linaro.org \
    --cc=philmd@linaro.org \
    --cc=pierrick.bouvier@linaro.org \
    --cc=qemu-arm@nongnu.org \
    --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).