All of lore.kernel.org
 help / color / mirror / Atom feed
From: Andi Kleen <andi@firstfloor.org>
To: travis@sgi.com
Cc: Andrew Morton <akpm@linux-foundation.org>,
	mingo@elte.hu, Christoph Lameter <clameter@sgi.com>,
	Jack Steiner <steiner@sgi.com>,
	linux-mm@kvack.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 08/10] x86: Change NR_CPUS arrays in numa_64 V2
Date: Tue, 15 Jan 2008 11:54:12 +0100	[thread overview]
Message-ID: <p73wsqbl5pn.fsf@bingen.suse.de> (raw)
In-Reply-To: <20080115021737.228970000@sgi.com> (travis@sgi.com's message of "Mon\, 14 Jan 2008 18\:17\:43 -0800")

travis@sgi.com writes:
> +
>  /* Returns the number of the node containing CPU 'cpu' */
>  static inline int cpu_to_node(int cpu)
>  {
> -	return cpu_to_node_map[cpu];
> +	u16 *cpu_to_node_map = x86_cpu_to_node_map_early_ptr;
> +
> +	if (cpu_to_node_map)
> +		return cpu_to_node_map[cpu];
> +	else if(per_cpu_offset(cpu))
> +		return per_cpu(x86_cpu_to_node_map, cpu);
> +	else
> +		return NUMA_NO_NODE;

Seems a little big now to be still inlined.

Also I wonder if there are really that many early callers that it
isn't feasible to just convert them to a early_cpu_to_node(). Also
early_cpu_to_node() should really not be speed critical, so just
linearly searching some other table instead of setting up an explicit
array should be fine for that.

-Andi

WARNING: multiple messages have this Message-ID (diff)
From: Andi Kleen <andi@firstfloor.org>
To: travis@sgi.com
Cc: Andrew Morton <akpm@linux-foundation.org>,
	mingo@elte.hu, Christoph Lameter <clameter@sgi.com>,
	Jack Steiner <steiner@sgi.com>,
	linux-mm@kvack.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 08/10] x86: Change NR_CPUS arrays in numa_64 V2
Date: Tue, 15 Jan 2008 11:54:12 +0100	[thread overview]
Message-ID: <p73wsqbl5pn.fsf@bingen.suse.de> (raw)
In-Reply-To: <20080115021737.228970000@sgi.com> (travis@sgi.com's message of "Mon\, 14 Jan 2008 18\:17\:43 -0800")

travis@sgi.com writes:
> +
>  /* Returns the number of the node containing CPU 'cpu' */
>  static inline int cpu_to_node(int cpu)
>  {
> -	return cpu_to_node_map[cpu];
> +	u16 *cpu_to_node_map = x86_cpu_to_node_map_early_ptr;
> +
> +	if (cpu_to_node_map)
> +		return cpu_to_node_map[cpu];
> +	else if(per_cpu_offset(cpu))
> +		return per_cpu(x86_cpu_to_node_map, cpu);
> +	else
> +		return NUMA_NO_NODE;

Seems a little big now to be still inlined.

Also I wonder if there are really that many early callers that it
isn't feasible to just convert them to a early_cpu_to_node(). Also
early_cpu_to_node() should really not be speed critical, so just
linearly searching some other table instead of setting up an explicit
array should be fine for that.

-Andi

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

  reply	other threads:[~2008-01-15 10:54 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-01-15  2:17 [PATCH 00/10] x86: Reduce memory and intra-node effects with large count NR_CPUs V2 travis
2008-01-15  2:17 ` travis
2008-01-15  2:17 ` [PATCH 01/10] x86: Change size of APICIDs from u8 to u16 V2 travis
2008-01-15  2:17   ` travis
2008-01-15  2:17 ` [PATCH 02/10] x86: Change size of node ids " travis
2008-01-15  2:17   ` travis
2008-01-15  5:59   ` Eric Dumazet
2008-01-15  5:59     ` Eric Dumazet
2008-01-15 15:51     ` Mike Travis
2008-01-15 15:51       ` Mike Travis
2008-01-15  2:17 ` [PATCH 03/10] x86: Change NR_CPUS arrays in powernow-k8 V2 travis
2008-01-15  2:17   ` travis
2008-01-15  2:17 ` [PATCH 04/10] x86: Change NR_CPUS arrays in intel_cacheinfo V2 travis
2008-01-15  2:17   ` travis
2008-01-15  2:17 ` [PATCH 05/10] x86: Change NR_CPUS arrays in smpboot_64 V2 travis
2008-01-15  2:17   ` travis
2008-01-15  2:17 ` [PATCH 06/10] x86: Change NR_CPUS arrays in topology V2 travis
2008-01-15  2:17   ` travis
2008-01-15  2:17 ` [PATCH 07/10] x86: Cleanup x86_cpu_to_apicid references V2 travis
2008-01-15  2:17   ` travis
2008-01-15  2:17 ` [PATCH 08/10] x86: Change NR_CPUS arrays in numa_64 V2 travis
2008-01-15  2:17   ` travis
2008-01-15 10:54   ` Andi Kleen [this message]
2008-01-15 10:54     ` Andi Kleen
2008-01-15 22:23     ` Mike Travis
2008-01-15 22:23       ` Mike Travis
2008-01-15  2:17 ` [PATCH 09/10] x86: Change NR_CPUS arrays in acpi-cpufreq V2 travis
2008-01-15  2:17   ` travis
2008-01-15  2:17 ` [PATCH 10/10] x86: Change bios_cpu_apicid to percpu data variable V2 travis
2008-01-15  2:17   ` travis

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=p73wsqbl5pn.fsf@bingen.suse.de \
    --to=andi@firstfloor.org \
    --cc=akpm@linux-foundation.org \
    --cc=clameter@sgi.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mingo@elte.hu \
    --cc=steiner@sgi.com \
    --cc=travis@sgi.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.