Re: [PATCH v16 3/8] cpu/hotplug: Add dynamic parallel bringup states before CPUHP_BRINGUP_CPU

From: David Woodhouse <dwmw2@infradead.org>
To: Thomas Gleixner <tglx@linutronix.de>,
	Usama Arif <usama.arif@bytedance.com>,
	kim.phillips@amd.com, brgerst@gmail.com
Cc: piotrgorski@cachyos.org, oleksandr@natalenko.name,
	arjan@linux.intel.com, mingo@redhat.com, bp@alien8.de,
	dave.hansen@linux.intel.com, hpa@zytor.com, x86@kernel.org,
	pbonzini@redhat.com, paulmck@kernel.org,
	linux-kernel@vger.kernel.org, kvm@vger.kernel.org,
	rcu@vger.kernel.org, mimoja@mimoja.de, hewenliang4@huawei.com,
	thomas.lendacky@amd.com, seanjc@google.com,
	pmenzel@molgen.mpg.de, fam.zheng@bytedance.com,
	punit.agrawal@bytedance.com, simon.evans@bytedance.com,
	liangma@liangbit.com, gpiccoli@igalia.com
Subject: Re: [PATCH v16 3/8] cpu/hotplug: Add dynamic parallel bringup states before CPUHP_BRINGUP_CPU
Date: Thu, 23 Mar 2023 23:12:21 +0000	[thread overview]
Message-ID: <8dff6ae5ffaebfbcc55a01c04420fd478070b830.camel@infradead.org> (raw)
In-Reply-To: <871qlf83wj.ffs@tglx>

[-- Attachment #1: Type: text/plain, Size: 3329 bytes --]

On Fri, 2023-03-24 at 00:05 +0100, Thomas Gleixner wrote:
> Still the rest can be simplified as below.

...

> --- a/kernel/cpu.c
> +++ b/kernel/cpu.c
> @@ -1504,13 +1504,45 @@ int bringup_hibernate_cpu(unsigned int s
>  
>  void bringup_nonboot_cpus(unsigned int setup_max_cpus)
>  {
> -       unsigned int cpu;
> +       unsigned int cpu, n = 1;
>  
> +       /*
> +        * On architectures which have setup the CPUHP_BP_PARALLEL_STARTUP
> +        * state, this invokes all BP prepare states and the parallel
> +        * startup state sends the startup IPI to each of the to be onlined
> +        * APs. This avoids waiting for each AP to respond to the startup
> +        * IPI in CPUHP_BRINGUP_CPU. The APs proceed through the low level
> +        * bringup code and then wait for the control CPU to release them
> +        * one by one for the final onlining procedure in the loop below.
> +        *
> +        * For architectures which do not support parallel bringup all
> +        * states are fully serialized in the loop below.
> +        */
> +       if (!cpuhp_step_empty(true, CPUHP_BP_PARALLEL_STARTUP) {

I'll take using cpuhp_step_empty().

> +                       for_each_present_cpu(cpu) {
> +                               if (n++ >= setup_max_cpus)
> +                                       break;
> +                               cpu_up(cpu, CPUHP_BP_PARALLEL_STARTUP);
> +                       }
> +       }
> +
> +       /* Do the per CPU serialized bringup to ONLINE state */
>         for_each_present_cpu(cpu) {
>                 if (num_online_cpus() >= setup_max_cpus)
>                         break;
> -               if (!cpu_online(cpu))
> -                       cpu_up(cpu, CPUHP_ONLINE);
> +
> +               if (!cpu_online(cpu)) {
> +                       struct cpuhp_cpu_state *st = per_cpu_ptr(&cpuhp_state, cpu);
> +                       int ret = cpu_up(cpu, CPUHP_ONLINE);
> +
> +                       /*
> +                        * Due to the above preparation loop a failed online attempt
> +                        * might have only rolled back to CPUHP_BP_PARALLEL_STARTUP. Do the
> +                        * remaining cleanups. NOOP for the non parallel case.
> +                        */
> +                       if (ret && can_rollback_cpu(st))
> +                               WARN_ON(cpuhp_invoke_callback_range(false, cpu, st, CPUHP_OFFLINE));
> +               }

And I'll take doing this bit unconditionally (it's basically a no-op if
they already got rolled all the way back to CPUHP_OFFLINE, right?).

But the additional complexity of having multiple steps is fairly
minimal, and I'm already planning to *use* another one even in x86, as
discussed.

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 5965 bytes --]