lkml.org 
[lkml]   [2010]   [Nov]   [3]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
Subject[PATCH 14/20] x86/ticketlock: loosen ordering restraints on unlock
Date
From: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>

It's only necessary to prevent the compiler from reordering code out of
the locked region past the unlock. Putting too many barriers in prevents
the compiler from generating the best code when PV spinlocks are enabled.

Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
arch/x86/include/asm/spinlock.h | 8 +++++---
1 files changed, 5 insertions(+), 3 deletions(-)

diff --git a/arch/x86/include/asm/spinlock.h b/arch/x86/include/asm/spinlock.h
index b9a1aae..d6de5e7 100644
--- a/arch/x86/include/asm/spinlock.h
+++ b/arch/x86/include/asm/spinlock.h
@@ -46,9 +46,7 @@ static __always_inline void __ticket_unlock_release(struct arch_spinlock *lock)
#else
static __always_inline void __ticket_unlock_release(struct arch_spinlock *lock)
{
- barrier();
lock->tickets.head++;
- barrier();
}
#endif

@@ -184,7 +182,11 @@ static __always_inline int arch_spin_trylock(arch_spinlock_t *lock)

static __always_inline void arch_spin_unlock(arch_spinlock_t *lock)
{
- __ticket_t next = lock->tickets.head + 1;
+ __ticket_t next;
+
+ barrier(); /* prevent reordering out of locked region */
+
+ next = lock->tickets.head + 1;
__ticket_unlock_release(lock);
__ticket_unlock_kick(lock, next);
}
--
1.7.2.3


\
 
 \ /
  Last update: 2010-11-03 16:05    [W:0.387 / U:0.144 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site