lkml.org 
[lkml]   [2013]   [Feb]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
Subject[tip:core/locking] x86/smp: Keep spinlock delay values per hashed spinlock address
Commit-ID:  e3c94ba97ac742a4c03713629da131fef53b2237
Gitweb: http://git.kernel.org/tip/e3c94ba97ac742a4c03713629da131fef53b2237
Author: Eric Dumazet <eric.dumazet@gmail.com>
AuthorDate: Wed, 6 Feb 2013 15:06:47 -0500
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Wed, 13 Feb 2013 09:06:30 +0100

x86/smp: Keep spinlock delay values per hashed spinlock address

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>
Reviewed-by: Michel Lespinasse <walken@google.com>
Cc: lwoodman@redhat.com
Cc: knoel@redhat.com
Cc: chegu_vinod@hp.com
Cc: raghavendra.kt@linux.vnet.ibm.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20130206150647.633f7b1c@cuia.bos.redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
arch/x86/kernel/smp.c | 24 ++++++++++++++++++++----
1 file changed, 20 insertions(+), 4 deletions(-)

diff --git a/arch/x86/kernel/smp.c b/arch/x86/kernel/smp.c
index c1fce41..8e94469 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>
@@ -116,6 +117,18 @@ 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 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,
+ },
+};
+
/*
* Wait on a congested ticket spinlock. Many spinlocks are embedded in
* data structures; having many CPUs pounce on the cache line with the
@@ -134,12 +147,14 @@ static bool smp_no_nmi_ipi = false;
* the queue, to slowly increase the delay if we sleep for too short a
* time, and to decrease the delay if we slept for too long.
*/
-DEFINE_PER_CPU(unsigned, spinlock_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;

for (;;) {
@@ -171,11 +186,12 @@ void ticket_spin_lock_wait(arch_spinlock_t *lock, struct __raw_tickets inc)
* to get it back to a good value quickly.
*/
if (delay >= 2 * DELAY_FIXED_1)
- delay -= max(delay/32, DELAY_FIXED_1);
+ delay -= max_t(u32, delay/32, DELAY_FIXED_1);
break;
}
}
- __this_cpu_write(spinlock_delay, delay);
+ ent->hash = hash;
+ ent->delay = delay;
}

/*

\
 
 \ /
  Last update: 2013-02-13 14:01    [W:0.163 / U:0.692 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site