mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Pierre Gondois <pierre.gondois@arm.com>
To: Yicong Yang <yangyicong@huawei.com>
Cc: linuxppc-dev@lists.ozlabs.org, bp@alien8.de,
	dave.hansen@linux.intel.com, mingo@redhat.com,
	linux-arm-kernel@lists.infradead.org, mpe@ellerman.id.au,
	peterz@infradead.org, tglx@linutronix.de, sudeep.holla@arm.com,
	will@kernel.org, catalin.marinas@arm.com, x86@kernel.org,
	linux-kernel@vger.kernel.org, dietmar.eggemann@arm.com,
	gregkh@linuxfoundation.org, rafael@kernel.org,
	jonathan.cameron@huawei.com, prime.zeng@hisilicon.com,
	linuxarm@huawei.com, yangyicong@hisilicon.com, xuwei5@huawei.com,
	guohanjun@huawei.com
Subject: Re: [PATCH v5 3/4] arm64: topology: Support SMT control on ACPI based system
Date: Tue, 27 Aug 2024 17:40:54 +0200	[thread overview]
Message-ID: <a998c723-7451-439a-9c88-7c8b5c1b890b@arm.com> (raw)
In-Reply-To: <20240806085320.63514-4-yangyicong@huawei.com>

Hello Yicong,
IIUC we are looking for the maximum number of threads a CPU can have
in the platform. But is is actually possible to have a platform with
CPUs not having the same number of threads ?

For instance kernel/cpu.c::cpu_smt_num_threads_valid() will check
that the number of threads is either 1 or cpu_smt_max_threads (as
CONFIG_SMT_NUM_THREADS_DYNAMIC is not enabled for arm64). However
a (hypothetical) platform with:
- CPU0: 2 threads
- CPU1: 4 threads
should not be able to set the number of threads to 2, as
1 < 2 < cpu_smt_max_threads (cf. cpu_smt_num_threads_valid()).

So if there is an assumption that all the CPUs have the same number of
threads, it should be sufficient to count the number of threads for one
CPU right ?

In the other (unlikely) case (i.e. CPUs can have various number of threads),
I think we should either:
- use your method and check that all the CPUs have the same number of threads
- get the number of threads for one CPU and check that all the CPUs are
    identical using the ACPI_PPTT_ACPI_IDENTICAL tag
- have a per-cpu cpu_smt_max_threads value ?

Same comment for the DT patch. If there is an assumption that all CPUs have
the same number of threads, then update_smt_num_threads() could only be called
once I suppose,

Regards,
Pierre


On 8/6/24 10:53, Yicong Yang wrote:
> From: Yicong Yang <yangyicong@hisilicon.com>
> 
> For ACPI we'll build the topology from PPTT and we cannot directly
> get the SMT number of each core. Instead using a temporary xarray
> to record the SMT number of each core when building the topology
> and we can know the largest SMT number in the system. Then we can
> enable the support of SMT control.
> 
> Signed-off-by: Yicong Yang <yangyicong@hisilicon.com>
> ---
>   arch/arm64/kernel/topology.c | 24 ++++++++++++++++++++++++
>   1 file changed, 24 insertions(+)
> 
> diff --git a/arch/arm64/kernel/topology.c b/arch/arm64/kernel/topology.c
> index 1a2c72f3e7f8..f72e1e55b05e 100644
> --- a/arch/arm64/kernel/topology.c
> +++ b/arch/arm64/kernel/topology.c
> @@ -15,8 +15,10 @@
>   #include <linux/arch_topology.h>
>   #include <linux/cacheinfo.h>
>   #include <linux/cpufreq.h>
> +#include <linux/cpu_smt.h>
>   #include <linux/init.h>
>   #include <linux/percpu.h>
> +#include <linux/xarray.h>
>   
>   #include <asm/cpu.h>
>   #include <asm/cputype.h>
> @@ -43,11 +45,16 @@ static bool __init acpi_cpu_is_threaded(int cpu)
>    */
>   int __init parse_acpi_topology(void)
>   {
> +	int thread_num, max_smt_thread_num = 1;
> +	struct xarray core_threads;
>   	int cpu, topology_id;
> +	void *entry;
>   
>   	if (acpi_disabled)
>   		return 0;
>   
> +	xa_init(&core_threads);
> +
>   	for_each_possible_cpu(cpu) {
>   		topology_id = find_acpi_cpu_topology(cpu, 0);
>   		if (topology_id < 0)
> @@ -57,6 +64,20 @@ int __init parse_acpi_topology(void)
>   			cpu_topology[cpu].thread_id = topology_id;
>   			topology_id = find_acpi_cpu_topology(cpu, 1);
>   			cpu_topology[cpu].core_id   = topology_id;
> +
> +			entry = xa_load(&core_threads, topology_id);
> +			if (!entry) {
> +				xa_store(&core_threads, topology_id,
> +					 xa_mk_value(1), GFP_KERNEL);
> +			} else {
> +				thread_num = xa_to_value(entry);
> +				thread_num++;
> +				xa_store(&core_threads, topology_id,
> +					 xa_mk_value(thread_num), GFP_KERNEL);
> +
> +				if (thread_num > max_smt_thread_num)
> +					max_smt_thread_num = thread_num;
> +			}
>   		} else {
>   			cpu_topology[cpu].thread_id  = -1;
>   			cpu_topology[cpu].core_id    = topology_id;
> @@ -67,6 +88,9 @@ int __init parse_acpi_topology(void)
>   		cpu_topology[cpu].package_id = topology_id;
>   	}
>   
> +	cpu_smt_set_num_threads(max_smt_thread_num, max_smt_thread_num);
> +
> +	xa_destroy(&core_threads);
>   	return 0;
>   }
>   #endif

  parent reply	other threads:[~2024-08-27 15:41 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-08-06  8:53 [PATCH v5 0/4] Support SMT control on arm64 Yicong Yang
2024-08-06  8:53 ` [PATCH v5 1/4] cpu/SMT: Provide a default topology_is_primary_thread() Yicong Yang
2024-08-16 15:54   ` Dietmar Eggemann
2024-08-19  7:25     ` Yicong Yang
2024-08-06  8:53 ` [PATCH v5 2/4] arch_topology: Support SMT control for OF based system Yicong Yang
2024-08-16 15:55   ` Dietmar Eggemann
2024-08-19  7:18     ` Yicong Yang
2024-08-06  8:53 ` [PATCH v5 3/4] arm64: topology: Support SMT control on ACPI " Yicong Yang
2024-08-16 15:55   ` Dietmar Eggemann
2024-08-19  7:03     ` Yicong Yang
2024-08-22  7:19       ` Dietmar Eggemann
2024-08-27 15:40   ` Pierre Gondois [this message]
2024-08-29  7:40     ` Yicong Yang
2024-08-29 12:46       ` Pierre Gondois
2024-08-30  9:35         ` Yicong Yang
2024-09-02  7:43           ` Pierre Gondois
2024-09-03 12:44             ` Yicong Yang
2024-09-05  8:34               ` Pierre Gondois
2024-09-05 12:02                 ` Yicong Yang
2024-09-06  7:06                   ` Morten Rasmussen
2024-09-06  8:36                     ` Yicong Yang
2024-09-12 11:59                       ` Morten Rasmussen
2024-09-12 12:53                         ` Michal Suchánek
2024-08-06  8:53 ` [PATCH v5 4/4] arm64: Kconfig: Enable HOTPLUG_SMT Yicong Yang
2024-08-27 15:40   ` Pierre Gondois
2024-08-29  6:50     ` Yicong Yang

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=a998c723-7451-439a-9c88-7c8b5c1b890b@arm.com \
    --to=pierre.gondois@arm.com \
    --cc=bp@alien8.de \
    --cc=catalin.marinas@arm.com \
    --cc=dave.hansen@linux.intel.com \
    --cc=dietmar.eggemann@arm.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=guohanjun@huawei.com \
    --cc=jonathan.cameron@huawei.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linuxarm@huawei.com \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=mingo@redhat.com \
    --cc=mpe@ellerman.id.au \
    --cc=peterz@infradead.org \
    --cc=prime.zeng@hisilicon.com \
    --cc=rafael@kernel.org \
    --cc=sudeep.holla@arm.com \
    --cc=tglx@linutronix.de \
    --cc=will@kernel.org \
    --cc=x86@kernel.org \
    --cc=xuwei5@huawei.com \
    --cc=yangyicong@hisilicon.com \
    --cc=yangyicong@huawei.com \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®