Re: [PATCH RFC V6 00/24] Support of Virtual CPU Hotplug-like Feature for ARMv8+ Arch

[Gavin Shan <gshan@redhat.com>]

Hi Salil,

On 10/1/25 11:01 AM, salil.mehta@opnsrc.net wrote:
> From: Salil Mehta <salil.mehta@huawei.com>
> 
> [!] Sending again: It looks like mails sent from my official ID are being held
> somewhere. Hence, I am using my other email address. Sorry for any inconvenience
> this may have caused.
> 

[...]

> 
> ============================
> (X) ORGANIZATION OF PATCHES
> ============================
> 
>   [Patch 1-2, 22-23] New HMP/QMP interface ('device_set') related changes
>      (*) New ('DeviceState::admin_power_state') property; Enabled/Disabled States and handling
>      (*) New Qemu CLI parameter ('-smp CPUS, disabled=N') handling
>      (*) Logic to find the existing object not part of the QOM
>   [Patch 3-5, 10] logic required during machine init.
>      (*) Some validation checks.
>      (*) Introduces core-id,socket-id,cluster-id property and some util functions required later.
>      (*) Logic to setup lazy realization of the QOM vCPUs
>      (*) Logic to pre-create vCPUs in the KVM host kernel.
>   [Patch 6-7, 8-9] logic required to size the GICv3 State
>      (*) GIC initialization pre-sized with possible vCPUs.
>      (*) Introduction of the GICv3 CPU Interface `accessibility` property & accessors
>      (*) Refactoring to make KVM & TCG 'GICv3CPUState' initialization common.
>      (*) Changes in GICv3 post/pre-load function for migration
>   [Patch 11,14-16,19] logic related to ACPI at machine init time.
>      (*) ACPI CPU OSPM interface for ACPI _STA.Enable/Disable handling
>      (*) ACPI GED framework to cater to CPU DeviceCheck/Eject Events.
>      (*) ACPI DSDT, MADT changes.
>   [Patch 12-13, 17] Qdev, Virt Machine, PowerState Handler Changes
>      (*) Changes to introduce 'PowerStateHandler' and its abstract interface.
>      (*) Qdev changes to handle the administrative enabling/disabling of device
>      (*) Virt Machine implementation of 'PowerStateHandler' Hooks
>      (*) vCPU thread user-space parking and unparking logic.
>   [Patch 18,20-21,24] Misc.
>      (*) Handling of SMCC Hypercall Exits by KVM to Qemu for PSCI.
>      (*) Mitigation to avoid using 'pause_all_vcpus' during ICC_CTLR_EL1 reset.
>      (*) Mitigation when TCG 'TB Code Cache' is found saturated
> 

[...]

> 
> ================
> (XII) Change Log
> ================
> 

Here seems missed the changelog from RFCv5 -> RFCv6?

> RFC V4 -> RFC V5:
> -----------------
> 1. Dropped "[PATCH RFC V4 19/33] target/arm: Force ARM vCPU *present* status ACPI *persistent*"
>     - Seperated the architecture agnostic ACPI changes required to support vCPU Hotplug
>       Link: https://lore.kernel.org/qemu-devel/20241014192205.253479-1-salil.mehta@huawei.com/#t
> 2. Dropped "[PATCH RFC V4 02/33] cpu-common: Add common CPU utility for possible vCPUs"
>     - Dropped qemu{present,enabled}_cpu() APIs. Commented by Gavin (Redhat), Miguel(Oracle), Igor(Redhat)
> 3. Added "Reviewed-by: Miguel Luis <miguel.luis@oracle.com>" to [PATCH RFC V4 01/33]
> 3. Dropped the `CPUState::disabled` flag and introduced `GICv3State::num_smp_cpus` flag
>     - All `GICv3CPUState' between [num_smp_cpus,num_cpus) are marked as 'inaccessible` during gicv3_common_realize()
>     - qemu_enabled_cpu() not required - removed!
>     - removed usage of `CPUState::disabled` from virt.c and hw/cpu64.c
> 4. Removed virt_cpu_properties() and introduced property `mp-affinity` get accessor
> 5. Dropped "[PATCH RFC V4 12/33] arm/virt: Create GED device before *disabled* vCPU Objects are destroyed"
> 

[...]

It maybe known issue, but there are a bunch of failing qtests, listed as below.

# make -j 60 check-qtest
   :
Summary of Failures:

10/29 qemu:qtest+qtest-aarch64 / qtest-aarch64/xlnx-versal-trng-test     ERROR            0.48s   killed by signal 6 SIGABRT
11/29 qemu:qtest+qtest-aarch64 / qtest-aarch64/xlnx-canfd-test           ERROR            0.49s   killed by signal 6 SIGABRT
12/29 qemu:qtest+qtest-aarch64 / qtest-aarch64/ast2700-gpio-test         ERROR            0.47s   killed by signal 6 SIGABRT
13/29 qemu:qtest+qtest-aarch64 / qtest-aarch64/ast2700-hace-test         ERROR            0.48s   killed by signal 6 SIGABRT
14/29 qemu:qtest+qtest-aarch64 / qtest-aarch64/ast2700-smc-test          ERROR            0.48s   killed by signal 6 SIGABRT
23/29 qemu:qtest+qtest-aarch64 / qtest-aarch64/test-hmp                  ERROR            3.14s   killed by signal 6 SIGABRT
26/29 qemu:qtest+qtest-aarch64 / qtest-aarch64/bios-tables-test          ERROR            8.83s   killed by signal 6 SIGABRT
27/29 qemu:qtest+qtest-aarch64 / qtest-aarch64/qom-test                  ERROR           12.75s   killed by signal 6 SIGABRT
28/29 qemu:qtest+qtest-aarch64 / qtest-aarch64/qos-test                  ERROR           16.32s   killed by signal 6 SIGABRT

Ok:                20
Fail:              9

Thanks,
Gavin