RE: [PATCH] cpu/hotplug: Allow the CPU in CPU_UP_PREPARE state to be brought up again.

["Longpeng (Mike, Cloud Infrastructure Service Product Dept.)" <longpeng2@xxxxxxxxxx>]

> -----Original Message-----
> From: Valentin Schneider [mailto:valentin.schneider@xxxxxxx]
> Sent: Wednesday, November 24, 2021 2:14 AM
> To: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>; Longpeng (Mike, Cloud
> Infrastructure Service Product Dept.) <longpeng2@xxxxxxxxxx>
> Cc: linux-kernel@xxxxxxxxxxxxxxx; Gonglei (Arei) <arei.gonglei@xxxxxxxxxx>;
> x86@xxxxxxxxxx; xen-devel@xxxxxxxxxxxxxxxxxxxx; Peter Zijlstra
> <peterz@xxxxxxxxxxxxx>; Ingo Molnar <mingo@xxxxxxxxxx>; Boris Ostrovsky
> <boris.ostrovsky@xxxxxxxxxx>; Juergen Gross <jgross@xxxxxxxx>; Stefano
> Stabellini <sstabellini@xxxxxxxxxx>; Thomas Gleixner <tglx@xxxxxxxxxxxxx>;
> Ingo Molnar <mingo@xxxxxxxxxx>; Borislav Petkov <bp@xxxxxxxxx>; Dave Hansen
> <dave.hansen@xxxxxxxxxxxxxxx>; H. Peter Anvin <hpa@xxxxxxxxx>
> Subject: Re: [PATCH] cpu/hotplug: Allow the CPU in CPU_UP_PREPARE state to be
> brought up again.
> 
> On 22/11/21 16:47, Sebastian Andrzej Siewior wrote:
> > From: "Longpeng(Mike)" <longpeng2@xxxxxxxxxx>
> >
> > A CPU will not show up in virtualized environment which includes an
> > Enclave. The VM splits its resources into a primary VM and a Enclave
> > VM. While the Enclave is active, the hypervisor will ignore all requests
> > to bring up a CPU and this CPU will remain in CPU_UP_PREPARE state.
> > The kernel will wait up to ten seconds for CPU to show up
> > (do_boot_cpu()) and then rollback the hotplug state back to
> > CPUHP_OFFLINE leaving the CPU state in CPU_UP_PREPARE. The CPU state is
> > set back to CPUHP_TEARDOWN_CPU during the CPU_POST_DEAD stage.
> >
> 
> For my own education, this is talking about *host* CPU hotplug, right?
> 

It's about the *guest* CPU hotplug.

1. Users in Primary VM:
Split out vcpuX (offline from Primary VM) for Enclave VM

2. Hypervisor:
Set a flag for vcpuX, all requests from Primary VM to bring up vcpuX
will be ignore.

3. Users in Primary VM:
echo 1 > .../vcpuX/online would fail and leave the CPU state of vcpuX
in CPU_UP_PREPARE.

4. Users in Primary VM terminate the Enclave VM:
Hypervisor should clear the flag (set in step 2) of vcpuX, so the vcpuX
can continue to receive requests.

5. Users in Primary VM:
Try to online the vcpuX again (expect success), but it's always failed.


> > After the Enclave VM terminates, the primary VM can bring up the CPU
> > again.
> >
> > Allow to bring up the CPU if it is in the CPU_UP_PREPARE state.
> >
> > [bigeasy: Rewrite commit description.]
> >
> > Signed-off-by: Longpeng(Mike) <longpeng2@xxxxxxxxxx>
> > Signed-off-by: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>
> > Link: https://lore.kernel.org/r/20210901051143.2752-1-longpeng2@xxxxxxxxxx
> > ---
> >
> > For XEN: this changes the behaviour as it allows to invoke
> > cpu_initialize_context() again should it have have earlier. I *think*
> > this is okay and would to bring up the CPU again should the memory
> > allocation in cpu_initialize_context() fail.
> 
> Virt stuff notwithstanding, that looks OK to me.
> Reviewed-by: Valentin Schneider <valentin.schneider@xxxxxxx>
> 
> That said, AFAICT only powerpc makes actual use of the state being set to
> CPU_UP_PREPARE, it looks to be needless bookkeeping for everyone else (and
> there's archs that seem happy using only CPU_DEAD / CPU_POST_DEAD).
> 
> >
> >  kernel/smpboot.c | 7 +++++++
> >  1 file changed, 7 insertions(+)
> >
> > diff --git a/kernel/smpboot.c b/kernel/smpboot.c
> > index f6bc0bc8a2aab..34958d7fe2c1c 100644
> > --- a/kernel/smpboot.c
> > +++ b/kernel/smpboot.c
> > @@ -392,6 +392,13 @@ int cpu_check_up_prepare(int cpu)
> >                */
> >               return -EAGAIN;
> >
> > +   case CPU_UP_PREPARE:
> > +           /*
> > +            * Timeout while waiting for the CPU to show up. Allow to try
> > +            * again later.
> > +            */
> > +           return 0;
> > +
> >       default:
> >
> >               /* Should not happen.  Famous last words. */
> > --
> > 2.33.1