All of lore.kernel.org
 help / color / mirror / Atom feed
From: Vincent Guittot <vincent.guittot@linaro.org>
To: Sudeep Holla <sudeep.holla@arm.com>
Cc: linux-kernel@vger.kernel.org,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Atish Patra <atishp@atishpatra.org>,
	Atish Patra <atishp@rivosinc.com>,
	Dietmar Eggemann <dietmar.eggemann@arm.com>,
	Qing Wang <wangqing@vivo.com>, Rob Herring <robh+dt@kernel.org>,
	"Rafael J . Wysocki" <rafael@kernel.org>,
	Ionela Voinescu <ionela.voinescu@arm.com>,
	Pierre Gondois <pierre.gondois@arm.com>,
	linux-arm-kernel@lists.infradead.org,
	linux-riscv@lists.infradead.org,
	Darren Hart <darren@os.amperecomputing.com>
Subject: Re: [PATCH v5 16/19] arch_topology: Limit span of cpu_clustergroup_mask()
Date: Tue, 28 Jun 2022 12:28:05 +0200	[thread overview]
Message-ID: <CAKfTPtAxNKV=L0DHuBGmBiUww5tDhyAiw-6N3cun11iQLALYHA@mail.gmail.com> (raw)
In-Reply-To: <20220627165047.336669-17-sudeep.holla@arm.com>

On Mon, 27 Jun 2022 at 18:51, Sudeep Holla <sudeep.holla@arm.com> wrote:
>
> From: Ionela Voinescu <ionela.voinescu@arm.com>
>
> Currently the cluster identifier is not set on DT based platforms.
> The reset or default value is -1 for all the CPUs. Once we assign the
> cluster identifier values correctly, the cluster_sibling mask will be
> populated and returned by cpu_clustergroup_mask() to contribute in the
> creation of the CLS scheduling domain level, if SCHED_CLUSTER is
> enabled.
>
> To avoid topologies that will result in questionable or incorrect
> scheduling domains, impose restrictions regarding the span of clusters,
> as presented to scheduling domains building code: cluster_sibling should
> not span more or the same CPUs as cpu_coregroup_mask().
>
> This is needed in order to obtain a strict separation between the MC and
> CLS levels, and maintain the same domains for existing platforms in
> the presence of CONFIG_SCHED_CLUSTER, where the new cluster information
> is redundant and irrelevant for the scheduler.
>
> While previously the scheduling domain builder code would have removed MC
> as redundant and kept CLS if SCHED_CLUSTER was enabled and the
> cpu_coregroup_mask() and cpu_clustergroup_mask() spanned the same CPUs,
> now CLS will be removed and MC kept.
>
> Link: https://lore.kernel.org/r/20220621192034.3332546-18-sudeep.holla@arm.com
> Cc: Darren Hart <darren@os.amperecomputing.com>
> Tested-by: Ionela Voinescu <ionela.voinescu@arm.com>
> Signed-off-by: Ionela Voinescu <ionela.voinescu@arm.com>
> Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
> ---
>  drivers/base/arch_topology.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
>
> diff --git a/drivers/base/arch_topology.c b/drivers/base/arch_topology.c
> index 46fa1b70b02b..277b65cf3306 100644
> --- a/drivers/base/arch_topology.c
> +++ b/drivers/base/arch_topology.c
> @@ -686,6 +686,14 @@ const struct cpumask *cpu_coregroup_mask(int cpu)
>
>  const struct cpumask *cpu_clustergroup_mask(int cpu)
>  {
> +       /*
> +        * Forbid cpu_clustergroup_mask() to span more or the same CPUs as
> +        * cpu_coregroup_mask().
> +        */
> +       if (cpumask_subset(cpu_coregroup_mask(cpu),
> +                          &cpu_topology[cpu].cluster_sibling))
> +               return get_cpu_mask(cpu);

AFAICT, this will change the Altra scheduling topology which will now
have a MC level instead of the CLS but it's probably not a problem as
the flags are the same for now

Acked-by: Vincent Guittot <vincent.guittot@linaro.org>

> +
>         return &cpu_topology[cpu].cluster_sibling;
>  }
>
> --
> 2.36.1
>

WARNING: multiple messages have this Message-ID (diff)
From: Vincent Guittot <vincent.guittot@linaro.org>
To: Sudeep Holla <sudeep.holla@arm.com>
Cc: linux-kernel@vger.kernel.org,
	 Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Atish Patra <atishp@atishpatra.org>,
	 Atish Patra <atishp@rivosinc.com>,
	Dietmar Eggemann <dietmar.eggemann@arm.com>,
	 Qing Wang <wangqing@vivo.com>, Rob Herring <robh+dt@kernel.org>,
	 "Rafael J . Wysocki" <rafael@kernel.org>,
	Ionela Voinescu <ionela.voinescu@arm.com>,
	 Pierre Gondois <pierre.gondois@arm.com>,
	linux-arm-kernel@lists.infradead.org,
	 linux-riscv@lists.infradead.org,
	Darren Hart <darren@os.amperecomputing.com>
Subject: Re: [PATCH v5 16/19] arch_topology: Limit span of cpu_clustergroup_mask()
Date: Tue, 28 Jun 2022 12:28:05 +0200	[thread overview]
Message-ID: <CAKfTPtAxNKV=L0DHuBGmBiUww5tDhyAiw-6N3cun11iQLALYHA@mail.gmail.com> (raw)
In-Reply-To: <20220627165047.336669-17-sudeep.holla@arm.com>

On Mon, 27 Jun 2022 at 18:51, Sudeep Holla <sudeep.holla@arm.com> wrote:
>
> From: Ionela Voinescu <ionela.voinescu@arm.com>
>
> Currently the cluster identifier is not set on DT based platforms.
> The reset or default value is -1 for all the CPUs. Once we assign the
> cluster identifier values correctly, the cluster_sibling mask will be
> populated and returned by cpu_clustergroup_mask() to contribute in the
> creation of the CLS scheduling domain level, if SCHED_CLUSTER is
> enabled.
>
> To avoid topologies that will result in questionable or incorrect
> scheduling domains, impose restrictions regarding the span of clusters,
> as presented to scheduling domains building code: cluster_sibling should
> not span more or the same CPUs as cpu_coregroup_mask().
>
> This is needed in order to obtain a strict separation between the MC and
> CLS levels, and maintain the same domains for existing platforms in
> the presence of CONFIG_SCHED_CLUSTER, where the new cluster information
> is redundant and irrelevant for the scheduler.
>
> While previously the scheduling domain builder code would have removed MC
> as redundant and kept CLS if SCHED_CLUSTER was enabled and the
> cpu_coregroup_mask() and cpu_clustergroup_mask() spanned the same CPUs,
> now CLS will be removed and MC kept.
>
> Link: https://lore.kernel.org/r/20220621192034.3332546-18-sudeep.holla@arm.com
> Cc: Darren Hart <darren@os.amperecomputing.com>
> Tested-by: Ionela Voinescu <ionela.voinescu@arm.com>
> Signed-off-by: Ionela Voinescu <ionela.voinescu@arm.com>
> Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
> ---
>  drivers/base/arch_topology.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
>
> diff --git a/drivers/base/arch_topology.c b/drivers/base/arch_topology.c
> index 46fa1b70b02b..277b65cf3306 100644
> --- a/drivers/base/arch_topology.c
> +++ b/drivers/base/arch_topology.c
> @@ -686,6 +686,14 @@ const struct cpumask *cpu_coregroup_mask(int cpu)
>
>  const struct cpumask *cpu_clustergroup_mask(int cpu)
>  {
> +       /*
> +        * Forbid cpu_clustergroup_mask() to span more or the same CPUs as
> +        * cpu_coregroup_mask().
> +        */
> +       if (cpumask_subset(cpu_coregroup_mask(cpu),
> +                          &cpu_topology[cpu].cluster_sibling))
> +               return get_cpu_mask(cpu);

AFAICT, this will change the Altra scheduling topology which will now
have a MC level instead of the CLS but it's probably not a problem as
the flags are the same for now

Acked-by: Vincent Guittot <vincent.guittot@linaro.org>

> +
>         return &cpu_topology[cpu].cluster_sibling;
>  }
>
> --
> 2.36.1
>

_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

WARNING: multiple messages have this Message-ID (diff)
From: Vincent Guittot <vincent.guittot@linaro.org>
To: Sudeep Holla <sudeep.holla@arm.com>
Cc: linux-kernel@vger.kernel.org,
	 Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Atish Patra <atishp@atishpatra.org>,
	 Atish Patra <atishp@rivosinc.com>,
	Dietmar Eggemann <dietmar.eggemann@arm.com>,
	 Qing Wang <wangqing@vivo.com>, Rob Herring <robh+dt@kernel.org>,
	 "Rafael J . Wysocki" <rafael@kernel.org>,
	Ionela Voinescu <ionela.voinescu@arm.com>,
	 Pierre Gondois <pierre.gondois@arm.com>,
	linux-arm-kernel@lists.infradead.org,
	 linux-riscv@lists.infradead.org,
	Darren Hart <darren@os.amperecomputing.com>
Subject: Re: [PATCH v5 16/19] arch_topology: Limit span of cpu_clustergroup_mask()
Date: Tue, 28 Jun 2022 12:28:05 +0200	[thread overview]
Message-ID: <CAKfTPtAxNKV=L0DHuBGmBiUww5tDhyAiw-6N3cun11iQLALYHA@mail.gmail.com> (raw)
In-Reply-To: <20220627165047.336669-17-sudeep.holla@arm.com>

On Mon, 27 Jun 2022 at 18:51, Sudeep Holla <sudeep.holla@arm.com> wrote:
>
> From: Ionela Voinescu <ionela.voinescu@arm.com>
>
> Currently the cluster identifier is not set on DT based platforms.
> The reset or default value is -1 for all the CPUs. Once we assign the
> cluster identifier values correctly, the cluster_sibling mask will be
> populated and returned by cpu_clustergroup_mask() to contribute in the
> creation of the CLS scheduling domain level, if SCHED_CLUSTER is
> enabled.
>
> To avoid topologies that will result in questionable or incorrect
> scheduling domains, impose restrictions regarding the span of clusters,
> as presented to scheduling domains building code: cluster_sibling should
> not span more or the same CPUs as cpu_coregroup_mask().
>
> This is needed in order to obtain a strict separation between the MC and
> CLS levels, and maintain the same domains for existing platforms in
> the presence of CONFIG_SCHED_CLUSTER, where the new cluster information
> is redundant and irrelevant for the scheduler.
>
> While previously the scheduling domain builder code would have removed MC
> as redundant and kept CLS if SCHED_CLUSTER was enabled and the
> cpu_coregroup_mask() and cpu_clustergroup_mask() spanned the same CPUs,
> now CLS will be removed and MC kept.
>
> Link: https://lore.kernel.org/r/20220621192034.3332546-18-sudeep.holla@arm.com
> Cc: Darren Hart <darren@os.amperecomputing.com>
> Tested-by: Ionela Voinescu <ionela.voinescu@arm.com>
> Signed-off-by: Ionela Voinescu <ionela.voinescu@arm.com>
> Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
> ---
>  drivers/base/arch_topology.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
>
> diff --git a/drivers/base/arch_topology.c b/drivers/base/arch_topology.c
> index 46fa1b70b02b..277b65cf3306 100644
> --- a/drivers/base/arch_topology.c
> +++ b/drivers/base/arch_topology.c
> @@ -686,6 +686,14 @@ const struct cpumask *cpu_coregroup_mask(int cpu)
>
>  const struct cpumask *cpu_clustergroup_mask(int cpu)
>  {
> +       /*
> +        * Forbid cpu_clustergroup_mask() to span more or the same CPUs as
> +        * cpu_coregroup_mask().
> +        */
> +       if (cpumask_subset(cpu_coregroup_mask(cpu),
> +                          &cpu_topology[cpu].cluster_sibling))
> +               return get_cpu_mask(cpu);

AFAICT, this will change the Altra scheduling topology which will now
have a MC level instead of the CLS but it's probably not a problem as
the flags are the same for now

Acked-by: Vincent Guittot <vincent.guittot@linaro.org>

> +
>         return &cpu_topology[cpu].cluster_sibling;
>  }
>
> --
> 2.36.1
>

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

  reply	other threads:[~2022-06-28 10:28 UTC|newest]

Thread overview: 144+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-27 16:50 [PATCH v5 00/19] arch_topology: Updates to add socket support and fix cluster ids Sudeep Holla
2022-06-27 16:50 ` Sudeep Holla
2022-06-27 16:50 ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 01/19] ACPI: PPTT: Use table offset as fw_token instead of virtual address Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 02/19] cacheinfo: Use of_cpu_device_node_get instead cpu_dev->of_node Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 03/19] cacheinfo: Add helper to access any cache index for a given CPU Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 04/19] cacheinfo: Move cache_leaves_are_shared out of CONFIG_OF Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 05/19] cacheinfo: Add support to check if last level cache(LLC) is valid or shared Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 06/19] cacheinfo: Allow early detection and population of cache attributes Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 07/19] cacheinfo: Use cache identifiers to check if the caches are shared if available Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 08/19] arch_topology: Add support to parse and detect cache attributes Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 09/19] arch_topology: Use the last level cache information from the cacheinfo Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-29 17:49   ` Conor.Dooley
2022-06-29 17:49     ` Conor.Dooley
2022-06-29 17:49     ` Conor.Dooley
2022-06-29 18:18     ` Conor.Dooley
2022-06-29 18:18       ` Conor.Dooley
2022-06-29 18:18       ` Conor.Dooley
2022-06-29 18:33       ` Sudeep Holla
2022-06-29 18:33         ` Sudeep Holla
2022-06-29 18:33         ` Sudeep Holla
2022-06-29 18:42       ` Sudeep Holla
2022-06-29 18:42         ` Sudeep Holla
2022-06-29 18:42         ` Sudeep Holla
2022-06-29 19:39         ` Conor.Dooley
2022-06-29 19:39           ` Conor.Dooley
2022-06-29 19:39           ` Conor.Dooley
2022-06-29 19:54           ` Sudeep Holla
2022-06-29 19:54             ` Sudeep Holla
2022-06-29 19:54             ` Sudeep Holla
2022-06-29 20:32             ` Conor.Dooley
2022-06-29 20:32               ` Conor.Dooley
2022-06-29 20:32               ` Conor.Dooley
2022-06-29 23:25               ` Conor.Dooley
2022-06-29 23:25                 ` Conor.Dooley
2022-06-29 23:25                 ` Conor.Dooley
2022-06-30 10:39                 ` Sudeep Holla
2022-06-30 10:39                   ` Sudeep Holla
2022-06-30 10:39                   ` Sudeep Holla
2022-06-30 16:37                   ` Conor.Dooley
2022-06-30 16:37                     ` Conor.Dooley
2022-06-30 16:37                     ` Conor.Dooley
2022-06-30 17:35                     ` Sudeep Holla
2022-06-30 17:35                       ` Sudeep Holla
2022-06-30 17:35                       ` Sudeep Holla
2022-06-30 19:20                       ` Conor.Dooley
2022-06-30 19:20                         ` Conor.Dooley
2022-06-30 19:20                         ` Conor.Dooley
2022-06-30 20:07                         ` Sudeep Holla
2022-06-30 20:07                           ` Sudeep Holla
2022-06-30 20:07                           ` Sudeep Holla
2022-06-30 20:13                           ` Conor.Dooley
2022-06-30 20:13                             ` Conor.Dooley
2022-06-30 20:13                             ` Conor.Dooley
2022-06-30 20:21                             ` Sudeep Holla
2022-06-30 20:21                               ` Sudeep Holla
2022-06-30 20:21                               ` Sudeep Holla
2022-06-30 22:07                               ` Conor.Dooley
2022-06-30 22:07                                 ` Conor.Dooley
2022-06-30 22:07                                 ` Conor.Dooley
2022-07-01 11:11                                 ` Sudeep Holla
2022-07-01 11:11                                   ` Sudeep Holla
2022-07-01 11:11                                   ` Sudeep Holla
2022-07-01 14:47                                   ` Conor.Dooley
2022-07-01 14:47                                     ` Conor.Dooley
2022-07-01 14:47                                     ` Conor.Dooley
2022-06-29 18:47       ` Sudeep Holla
2022-06-29 18:47         ` Sudeep Holla
2022-06-29 18:47         ` Sudeep Holla
2022-06-29 18:56         ` Conor.Dooley
2022-06-29 18:56           ` Conor.Dooley
2022-06-29 18:56           ` Conor.Dooley
2022-06-29 19:12           ` Sudeep Holla
2022-06-29 19:12             ` Sudeep Holla
2022-06-29 19:12             ` Sudeep Holla
2022-06-29 19:25             ` Conor.Dooley
2022-06-29 19:25               ` Conor.Dooley
2022-06-29 19:25               ` Conor.Dooley
2022-06-29 19:43               ` Sudeep Holla
2022-06-29 19:43                 ` Sudeep Holla
2022-06-29 19:43                 ` Sudeep Holla
2022-06-29 19:52                 ` Conor.Dooley
2022-06-29 19:52                   ` Conor.Dooley
2022-06-29 19:52                   ` Conor.Dooley
2022-06-29 18:29     ` Sudeep Holla
2022-06-29 18:29       ` Sudeep Holla
2022-06-29 18:29       ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 10/19] arm64: topology: Remove redundant setting of llc_id in CPU topology Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 11/19] arch_topology: Drop LLC identifier stash from the " Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 12/19] arch_topology: Set thread sibling cpumask only within the cluster Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 13/19] arch_topology: Check for non-negative value rather than -1 for IDs validity Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 14/19] arch_topology: Avoid parsing through all the CPUs once a outlier CPU is found Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 15/19] arch_topology: Don't set cluster identifier as physical package identifier Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 16/19] arch_topology: Limit span of cpu_clustergroup_mask() Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-28 10:28   ` Vincent Guittot [this message]
2022-06-28 10:28     ` Vincent Guittot
2022-06-28 10:28     ` Vincent Guittot
2022-06-27 16:50 ` [PATCH v5 17/19] arch_topology: Set cluster identifier in each core/thread from /cpu-map Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 18/19] arch_topology: Add support for parsing sockets in /cpu-map Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50 ` [PATCH v5 19/19] arch_topology: Warn that topology for nested clusters is not supported Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-27 16:50   ` Sudeep Holla
2022-06-29 13:06 ` [PATCH] ACPI: Remove the unused find_acpi_cpu_cache_topology() Sudeep Holla
2022-06-29 13:06   ` Sudeep Holla
2022-06-29 13:06   ` Sudeep Holla
2022-06-29 13:50   ` Rafael J. Wysocki
2022-06-29 13:50     ` Rafael J. Wysocki
2022-06-29 13:50     ` Rafael J. Wysocki

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='CAKfTPtAxNKV=L0DHuBGmBiUww5tDhyAiw-6N3cun11iQLALYHA@mail.gmail.com' \
    --to=vincent.guittot@linaro.org \
    --cc=atishp@atishpatra.org \
    --cc=atishp@rivosinc.com \
    --cc=darren@os.amperecomputing.com \
    --cc=dietmar.eggemann@arm.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=ionela.voinescu@arm.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-riscv@lists.infradead.org \
    --cc=pierre.gondois@arm.com \
    --cc=rafael@kernel.org \
    --cc=robh+dt@kernel.org \
    --cc=sudeep.holla@arm.com \
    --cc=wangqing@vivo.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.