From: Markus Armbruster <armbru@redhat.com>
To: Bibo Mao <maobibo@loongson.cn>
Cc: Song Gao <gaosong@loongson.cn>,
Jiaxun Yang <jiaxun.yang@flygoat.com>,
qemu-devel@nongnu.org, Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [PATCH v4 4/4] target/loongarch: Set dest error with error_abort in virt_cpu_irq_init
Date: Wed, 19 Mar 2025 07:09:42 +0100 [thread overview]
Message-ID: <877c4l4km1.fsf@pond.sub.org> (raw)
In-Reply-To: <20250319020847.1511759-5-maobibo@loongson.cn> (Bibo Mao's message of "Wed, 19 Mar 2025 10:08:47 +0800")
Bibo Mao <maobibo@loongson.cn> writes:
> In function virt_cpu_irq_init(), there is notification with ipi and extioi
> interrupt controller for cpu creation. Local variable with error type is
> used, however there is no check with its return value.
Good catch.
When the first call fails, we pass non-null @err to the second call,
which is wrong. If that one also fails, it'll likely trip
error_setv()'s assertion.
> Here set dest error object with error_abort, rather than local variable, so
> application will abort to run if there is error.
Why is failure impossible there?
If failure is impossible, the code before the patch is harmlessly wrong.
If failure is possible, the code before the patch has a crash bug, and
the patch makes it crash harder, i.e. when either call fails instead of
when both fail.
> Fixes: 50ebc3fc47fe (hw/intc/loongarch_ipi: Notify ipi object when cpu is plugged)
> Fixes: 087a23a87c57 (hw/intc/loongarch_extioi: Use cpu plug notification)
> Signed-off-by: Bibo Mao <maobibo@loongson.cn>
> ---
> hw/loongarch/virt.c | 7 ++++---
> 1 file changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/hw/loongarch/virt.c b/hw/loongarch/virt.c
> index a9fab39dd8..f10a4704ab 100644
> --- a/hw/loongarch/virt.c
> +++ b/hw/loongarch/virt.c
> @@ -327,7 +327,6 @@ static void virt_cpu_irq_init(LoongArchVirtMachineState *lvms)
> MachineClass *mc = MACHINE_GET_CLASS(ms);
> const CPUArchIdList *possible_cpus;
> CPUState *cs;
> - Error *err = NULL;
>
> /* cpu nodes */
> possible_cpus = mc->possible_cpu_arch_ids(ms);
> @@ -337,8 +336,10 @@ static void virt_cpu_irq_init(LoongArchVirtMachineState *lvms)
> continue;
> }
>
> - hotplug_handler_plug(HOTPLUG_HANDLER(lvms->ipi), DEVICE(cs), &err);
> - hotplug_handler_plug(HOTPLUG_HANDLER(lvms->extioi), DEVICE(cs), &err);
> + hotplug_handler_plug(HOTPLUG_HANDLER(lvms->ipi), DEVICE(cs),
> + &error_abort);
> + hotplug_handler_plug(HOTPLUG_HANDLER(lvms->extioi), DEVICE(cs),
> + &error_abort);
> }
> }
next prev parent reply other threads:[~2025-03-19 6:11 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-19 2:08 [PATCH v4 0/4] target/loongarch: Solve some issues reported from coccinelle Bibo Mao
2025-03-19 2:08 ` [PATCH v4 1/4] target/loongarch: Fix error handling of KVM feature checks Bibo Mao
2025-03-19 2:08 ` [PATCH v4 2/4] hw/loongarch/virt: Remove unnecessary NULL pointer Bibo Mao
2025-03-19 6:50 ` Markus Armbruster
2025-03-19 8:14 ` bibo mao
2025-03-19 8:43 ` Markus Armbruster
2025-03-19 2:08 ` [PATCH v4 3/4] target/loongarch: Remove unnecessary temporary variable assignment Bibo Mao
2025-03-19 2:08 ` [PATCH v4 4/4] target/loongarch: Set dest error with error_abort in virt_cpu_irq_init Bibo Mao
2025-03-19 6:09 ` Markus Armbruster [this message]
2025-03-19 7:58 ` bibo mao
2025-03-19 8:52 ` Markus Armbruster
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=877c4l4km1.fsf@pond.sub.org \
--to=armbru@redhat.com \
--cc=gaosong@loongson.cn \
--cc=jiaxun.yang@flygoat.com \
--cc=maobibo@loongson.cn \
--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 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.