Re: [PATCH v2] cpu/hotplug: Fix NULL kobject warning in cpuhp_smt_enable()

[Catalin Marinas <catalin.marinas@arm.com>]

On Mon, Apr 27, 2026 at 10:35:07AM +0800, Jinjie Ruan wrote:
> On arm64, when booting with `maxcpus` greater than the number of present
> CPUs (e.g., QEMU -smp cpus=4,maxcpus=8), some CPUs are marked as 'present'
> but have not yet been registered via register_cpu(). Consequently,
> the per-cpu device objects for these CPUs are not yet initialized.
[...]
> Fix this by:
> 
> 1. When booting with ACPI, checking the ACPI_MADT_ENABLED flag in the GICC
>    entry before calling set_cpu_present() during SMP initialization.
> 
> 2. Properly managing the present mask in acpi_map_cpu() and
>    acpi_unmap_cpu() to support actual CPU hotplug events, This aligns with
>    other architectures like x86 and LoongArch.

I had a chat with James earlier and IIUC the decision was to mark all
CPUs present and the GIC must be fully initialised. But digging through
the GICv3 code, I don't see it depending on cpu_present_mask but rather
on the "always on" MADT GICR description. So I think it should be safe
as long as we don't rely on the GICC gicr_base_address. But we should
update Documentation/arch/arm64/cpu-hotplug.rst to no longer state that
all online-capable vCPUs are marked as present by the kernel.

(or maybe I misunderstood all this)

-- 
Catalin