linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: dietmar.eggemann@arm.com (Dietmar Eggemann)
To: linux-arm-kernel@lists.infradead.org
Subject: [RFC PATCH 3/8] arm: parse cpu capacity from DT
Date: Thu, 10 Dec 2015 14:14:53 +0000	[thread overview]
Message-ID: <566988DD.9080005@arm.com> (raw)
In-Reply-To: <1448288921-30307-4-git-send-email-juri.lelli@arm.com>

On 23/11/15 14:28, Juri Lelli wrote:
> With the introduction of cpu capacity bindings, CPU capacities can now be
> extracted from DT. Add parsing of such information at boot time. We keep
> code that can produce same information, based on different DT properties
> and hard-coded values, as fall-back for backward compatibility.

This patch-set should define _one_ way to be able to specify
heterogeneous cpu capacity values for all arm and arm64 systems. So I
would really like that we can agree on this solution ('capacity'
property) and delete the old (cortex-a[15,7] only) solution based on
struct cpu_efficiency table_efficiency[] and 'clock-frequency' property
in arch/arm/kernel/topology.c. The appropriate code in
arch/arm[,64]/kernel/topology.c should be the same. Everything else is
highly confusing.

> 
> Cc: Russell King <linux@arm.linux.org.uk>
> Signed-off-by: Juri Lelli <juri.lelli@arm.com>
> ---
>  arch/arm/kernel/topology.c | 50 +++++++++++++++++++++++++++++++++++++++++++++-
>  1 file changed, 49 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/arm/kernel/topology.c b/arch/arm/kernel/topology.c
> index ec279d1..ecbff03 100644
> --- a/arch/arm/kernel/topology.c
> +++ b/arch/arm/kernel/topology.c
> @@ -78,6 +78,35 @@ static unsigned long *__cpu_capacity;
>  #define cpu_capacity(cpu)	__cpu_capacity[cpu]
>  
>  static unsigned long middle_capacity = 1;
> +static bool capacity_from_dt = true;
> +static u32 capacity_scale = SCHED_CAPACITY_SCALE;
> +
> +static int __init parse_cpu_capacity(struct device_node *cpu_node, int cpu)
> +{
> +	int ret = 1;
> +	u32 cpu_capacity;
> +
> +	ret = of_property_read_u32(cpu_node,
> +				   "capacity",
> +				   &cpu_capacity);
> +	if (!ret) {
> +		u64 capacity;
> +
> +		/*
> +		 * Enforce capacity <= capacity-scale.
> +		 */
> +		cpu_capacity = cpu_capacity <= capacity_scale ? cpu_capacity :
> +			capacity_scale;
> +		capacity = (cpu_capacity << SCHED_CAPACITY_SHIFT) /
> +			capacity_scale;
> +
> +		set_capacity_scale(cpu, capacity);
> +		pr_info("CPU%d: DT cpu capacity %lu\n",
> +			cpu, arch_scale_cpu_capacity(NULL, cpu));
> +	}
> +
> +	return !ret;
> +}
>  
>  /*
>   * Iterate all CPUs' descriptor in DT and compute the efficiency
> @@ -99,6 +128,18 @@ static void __init parse_dt_topology(void)
>  	__cpu_capacity = kcalloc(nr_cpu_ids, sizeof(*__cpu_capacity),
>  				 GFP_NOWAIT);
>  
> +	cn = of_find_node_by_path("/cpus");
> +	if (!cn) {
> +		pr_err("No CPU information found in DT\n");
> +		return;
> +	}
> +
> +	if (!of_property_read_u32(cn, "capacity-scale", &capacity_scale))
> +		pr_info("DT cpus capacity-scale %u\n", capacity_scale);
> +	else
> +		pr_debug("DT cpus capacity-scale not found: assuming %u\n",
> +			capacity_scale);
> +
>  	for_each_possible_cpu(cpu) {
>  		const u32 *rate;
>  		int len;
> @@ -110,6 +151,13 @@ static void __init parse_dt_topology(void)
>  			continue;
>  		}
>  
> +		if (parse_cpu_capacity(cn, cpu)) {
> +			of_node_put(cn);
> +			continue;
> +		}
> +
> +		capacity_from_dt = false;
> +
>  		for (cpu_eff = table_efficiency; cpu_eff->compatible; cpu_eff++)
>  			if (of_device_is_compatible(cn, cpu_eff->compatible))
>  				break;
> @@ -160,7 +208,7 @@ static void __init parse_dt_topology(void)
>   */
>  static void update_cpu_capacity(unsigned int cpu)
>  {
> -	if (!cpu_capacity(cpu))
> +	if (!cpu_capacity(cpu) || capacity_from_dt)
>  		return;
>  
>  	set_capacity_scale(cpu, cpu_capacity(cpu) / middle_capacity);
> 

  reply	other threads:[~2015-12-10 14:14 UTC|newest]

Thread overview: 57+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-11-23 14:28 [RFC PATCH 0/8] CPUs capacity information for heterogeneous systems Juri Lelli
2015-11-23 14:28 ` [RFC PATCH 1/8] ARM: initialize cpu_scale to its default Juri Lelli
2015-11-30 11:13   ` Vincent Guittot
2015-11-23 14:28 ` [RFC PATCH 2/8] Documentation: arm: define DT cpu capacity bindings Juri Lelli
2015-11-24  2:06   ` Rob Herring
2015-11-24 10:54     ` Juri Lelli
2015-11-30  9:59       ` Vincent Guittot
2015-12-01 11:20         ` Juri Lelli
2015-12-10 14:14           ` Dietmar Eggemann
2015-12-11 10:09             ` Juri Lelli
2015-12-10 15:30     ` Mark Brown
2015-12-10 17:58       ` Juri Lelli
2015-12-11 17:49         ` Mark Brown
2015-12-14 12:36           ` Juri Lelli
2015-12-14 16:59             ` Mark Brown
2015-12-15 12:22               ` Juri Lelli
2015-12-15 13:39                 ` Mark Brown
2015-12-15 14:01                   ` Mark Rutland
2015-12-15 14:24                     ` Juri Lelli
2015-12-15 14:50                       ` Mark Rutland
2015-12-15 15:36                         ` Juri Lelli
2015-12-15 15:08                     ` Mark Brown
2015-12-15 15:32                       ` Mark Rutland
2015-12-15 15:46                         ` Juri Lelli
2015-12-15 15:57                           ` Mark Rutland
2015-12-15 16:23                             ` Catalin Marinas
2015-12-15 16:41                               ` Mark Rutland
2015-12-15 16:59                                 ` Vincent Guittot
2015-12-15 17:15                                   ` Mark Rutland
2015-12-15 17:47                                     ` Vincent Guittot
2015-12-15 18:39                                       ` Mark Rutland
2015-12-15 17:17                         ` Mark Brown
2015-12-15 17:28                           ` Mark Rutland
2015-12-15 17:45                             ` Mark Brown
2015-12-15 18:10                               ` Mark Rutland
2015-12-15 18:45                                 ` Mark Brown
2015-12-17  9:07                               ` Juri Lelli
2015-12-15 13:55                 ` Vincent Guittot
2015-11-23 14:28 ` [RFC PATCH 3/8] arm: parse cpu capacity from DT Juri Lelli
2015-12-10 14:14   ` Dietmar Eggemann [this message]
2015-12-11 10:12     ` Juri Lelli
2015-11-23 14:28 ` [RFC PATCH 4/8] arm, dts: add TC2 cpu capacity information Juri Lelli
2015-11-23 14:28 ` [RFC PATCH 5/8] arm64: parse cpu capacity from DT Juri Lelli
2015-12-10 14:15   ` Dietmar Eggemann
2015-12-11 10:07     ` Juri Lelli
2015-11-23 14:28 ` [RFC PATCH 6/8] arm64, dts: add Juno cpu capacity information Juri Lelli
2015-11-23 14:28 ` [RFC PATCH 7/8] arm: add sysfs cpu_capacity attribute Juri Lelli
2015-11-23 14:28 ` [RFC PATCH 8/8] arm64: " Juri Lelli
2015-12-10 14:15   ` Dietmar Eggemann
2015-12-10 15:59     ` Mark Brown
2015-12-10 18:01       ` Juri Lelli
2015-12-11 17:54         ` Mark Brown
2015-12-07 12:02 ` [RFC PATCH 0/8] CPUs capacity information for heterogeneous systems Juri Lelli
2015-12-07 12:11   ` Russell King - ARM Linux
2015-12-07 12:36     ` Juri Lelli
2015-12-07 13:18       ` Russell King - ARM Linux
2015-12-07 15:41         ` Juri Lelli

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=566988DD.9080005@arm.com \
    --to=dietmar.eggemann@arm.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    /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;
as well as URLs for NNTP newsgroup(s).