linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Rafael Aquini <aquini@redhat.com>
To: Rik van Riel <riel@redhat.com>
Cc: linux-kernel@vger.kernel.org, walken@google.com,
	eric.dumazet@gmail.com, lwoodman@redhat.com, jeremy@goop.org,
	Jan Beulich <JBeulich@novell.com>,
	knoel@redhat.com, chegu_vinod@hp.com,
	raghavendra.kt@linux.vnet.ibm.com, mingo@redhat.com
Subject: Re: [PATCH 4/5] x86,smp: keep spinlock delay values per hashed spinlock address
Date: Thu, 10 Jan 2013 01:14:19 -0200	[thread overview]
Message-ID: <20130110031418.GD1636@x61.redhat.com> (raw)
In-Reply-To: <20130108173119.5e59fac2@annuminas.surriel.com>

On Tue, Jan 08, 2013 at 05:31:19PM -0500, Rik van Riel wrote:
> From: Eric Dumazet <eric.dumazet@gmail.com>
> 
> Eric Dumazet found a regression with the first version of the spinlock
> backoff code, in a workload where multiple spinlocks were contended,
> each having a different wait time.
> 
> This patch has multiple delay values per cpu, indexed on a hash
> of the lock address, to avoid that problem.
> 
> Eric Dumazet wrote:
> 
> I did some tests with your patches with following configuration :
> 
> tc qdisc add dev eth0 root htb r2q 1000 default 3
> (to force a contention on qdisc lock, even with a multi queue net
> device)
> 
> and 24 concurrent "netperf -t UDP_STREAM -H other_machine -- -m 128"
> 
> Machine : 2 Intel(R) Xeon(R) CPU X5660  @ 2.80GHz
> (24 threads), and a fast NIC (10Gbps)
> 
> Resulting in a 13 % regression (676 Mbits -> 595 Mbits)
> 
> In this workload we have at least two contended spinlocks, with
> different delays. (spinlocks are not held for the same duration)
> 
> It clearly defeats your assumption of a single per cpu delay being OK :
> Some cpus are spinning too long while the lock was released.
> 
> We might try to use a hash on lock address, and an array of 16 different
> delays so that different spinlocks have a chance of not sharing the same
> delay.
> 
> With following patch, I get 982 Mbits/s with same bench, so an increase
> of 45 % instead of a 13 % regression.
> 
> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
> Signed-off-by: Rik van Riel <riel@redhat.com>
> ---

Acked-by: Rafael Aquini <aquini@redhat.com>


>  arch/x86/kernel/smp.c |   22 +++++++++++++++++++---
>  1 files changed, 19 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/x86/kernel/smp.c b/arch/x86/kernel/smp.c
> index 05f828b..1877890 100644
> --- a/arch/x86/kernel/smp.c
> +++ b/arch/x86/kernel/smp.c
> @@ -23,6 +23,7 @@
>  #include <linux/interrupt.h>
>  #include <linux/cpu.h>
>  #include <linux/gfp.h>
> +#include <linux/hash.h>
>  
>  #include <asm/mtrr.h>
>  #include <asm/tlbflush.h>
> @@ -134,12 +135,26 @@ static bool smp_no_nmi_ipi = false;
>  #define DELAY_FIXED_1 (1<<DELAY_SHIFT)
>  #define MIN_SPINLOCK_DELAY (1 * DELAY_FIXED_1)
>  #define MAX_SPINLOCK_DELAY (16000 * DELAY_FIXED_1)
> -DEFINE_PER_CPU(unsigned, spinlock_delay) = { MIN_SPINLOCK_DELAY };
> +#define DELAY_HASH_SHIFT 6
> +struct delay_entry {
> +	u32 hash;
> +	u32 delay;
> +};
> +static DEFINE_PER_CPU(struct delay_entry [1 << DELAY_HASH_SHIFT], spinlock_delay) = {
> +	[0 ... (1 << DELAY_HASH_SHIFT) - 1] = {
> +		.hash = 0,
> +		.delay = MIN_SPINLOCK_DELAY,
> +	},
> +};
> +
>  void ticket_spin_lock_wait(arch_spinlock_t *lock, struct __raw_tickets inc)
>  {
>  	__ticket_t head = inc.head, ticket = inc.tail;
>  	__ticket_t waiters_ahead;
> -	unsigned delay = __this_cpu_read(spinlock_delay);
> +	u32 hash = hash32_ptr(lock);
> +	u32 slot = hash_32(hash, DELAY_HASH_SHIFT);
> +	struct delay_entry *ent = &__get_cpu_var(spinlock_delay[slot]);
> +	u32 delay = (ent->hash == hash) ? ent->delay : MIN_SPINLOCK_DELAY;
>  	unsigned loops = 1;
>  
>  	for (;;) {
> @@ -175,7 +190,8 @@ void ticket_spin_lock_wait(arch_spinlock_t *lock, struct __raw_tickets inc)
>  			break;
>  		}
>  	}
> -	__this_cpu_write(spinlock_delay, delay);
> +	ent->hash = hash;
> +	ent->delay = delay;
>  }
>  
>  /*
> 

  reply	other threads:[~2013-01-10  3:14 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-01-08 22:26 [PATCH 0/5] x86,smp: make ticket spinlock proportional backoff w/ auto tuning Rik van Riel
2013-01-08 22:30 ` [PATCH 3/5] x86,smp: auto tune spinlock backoff delay factor Rik van Riel
2013-01-10  3:13   ` Rafael Aquini
2013-01-10 12:49   ` Michel Lespinasse
2013-01-08 22:31 ` [PATCH 4/5] x86,smp: keep spinlock delay values per hashed spinlock address Rik van Riel
2013-01-10  3:14   ` Rafael Aquini [this message]
2013-01-10 13:01   ` Michel Lespinasse
2013-01-10 13:05     ` Rik van Riel
2013-01-10 13:15       ` Michel Lespinasse
2013-01-08 22:32 ` [DEBUG PATCH 5/5] x86,smp: add debugging code to track spinlock delay value Rik van Riel
2013-01-08 22:32 ` [PATCH 2/5] x86,smp: proportional backoff for ticket spinlocks Rik van Riel
2013-01-08 22:50   ` Eric Dumazet
2013-01-08 22:54     ` Rik van Riel
2013-01-10  2:30   ` Rafael Aquini
2013-01-08 22:32 ` [PATCH 1/5] x86,smp: move waiting on contended ticket lock out of line Rik van Riel
2013-01-08 22:43   ` Eric Dumazet
2013-01-10 17:38   ` Raghavendra K T
2013-01-09 12:50 ` [PATCH 0/5] x86,smp: make ticket spinlock proportional backoff w/ auto tuning Raghavendra K T
2013-01-10  2:27   ` Rafael Aquini
2013-01-10 17:36     ` Raghavendra K T
2013-01-11 20:11       ` Rik van Riel
2013-01-13 18:07         ` Raghavendra K T
2013-01-10 15:19 ` Mike Galbraith
2013-01-10 15:31   ` Rik van Riel
2013-01-10 19:30     ` Mike Galbraith
2013-01-24 13:28       ` Ingo Molnar
2013-01-10 22:24 ` Chegu Vinod

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=20130110031418.GD1636@x61.redhat.com \
    --to=aquini@redhat.com \
    --cc=JBeulich@novell.com \
    --cc=chegu_vinod@hp.com \
    --cc=eric.dumazet@gmail.com \
    --cc=jeremy@goop.org \
    --cc=knoel@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lwoodman@redhat.com \
    --cc=mingo@redhat.com \
    --cc=raghavendra.kt@linux.vnet.ibm.com \
    --cc=riel@redhat.com \
    --cc=walken@google.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;
as well as URLs for NNTP newsgroup(s).