Displaying 4 results from an estimated 4 matches for "pv_ticket_clear_slowpath".
Did you mean:
__ticket_clear_slowpath
2015 Apr 30
0
[PATCH 3/6] x86: introduce new pvops function clear_slowpath
...p(void)
diff --git a/arch/x86/kernel/paravirt-spinlocks.c b/arch/x86/kernel/paravirt-spinlocks.c
index bbb6c73..5ece813 100644
--- a/arch/x86/kernel/paravirt-spinlocks.c
+++ b/arch/x86/kernel/paravirt-spinlocks.c
@@ -8,10 +8,32 @@
#include <asm/paravirt.h>
+#ifdef CONFIG_SMP
+static void pv_ticket_clear_slowpath(arch_spinlock_t *lock, __ticket_t head)
+{
+ arch_spinlock_t old, new;
+
+ old.tickets.head = head;
+ new.tickets.head = head & ~TICKET_SLOWPATH_FLAG;
+ old.tickets.tail = new.tickets.head + TICKET_LOCK_INC;
+ new.tickets.tail = old.tickets.tail;
+
+ /* try to clear slowpath flag when there are...
2015 Apr 30
0
[PATCH 4/6] x86: introduce new pvops function spin_unlock
...static inline int arch_spin_is_locked(arch_spinlock_t *lock)
diff --git a/arch/x86/kernel/paravirt-spinlocks.c b/arch/x86/kernel/paravirt-spinlocks.c
index 5ece813..91273fb 100644
--- a/arch/x86/kernel/paravirt-spinlocks.c
+++ b/arch/x86/kernel/paravirt-spinlocks.c
@@ -22,9 +22,24 @@ static void pv_ticket_clear_slowpath(arch_spinlock_t *lock, __ticket_t head)
cmpxchg(&lock->head_tail, old.head_tail, new.head_tail);
}
+static void pv_ticket_unlock(arch_spinlock_t *lock)
+{
+ __ticket_t head;
+
+ BUILD_BUG_ON(((__ticket_t)NR_CPUS) != NR_CPUS);
+
+ head = xadd(&lock->tickets.head, TICKET_LOCK_INC);...
2015 Apr 30
12
[PATCH 0/6] x86: reduce paravirtualized spinlock overhead
Paravirtualized spinlocks produce some overhead even if the kernel is
running on bare metal. The main reason are the more complex locking
and unlocking functions. Especially unlocking is no longer just one
instruction but so complex that it is no longer inlined.
This patch series addresses this issue by adding two more pvops
functions to reduce the size of the inlined spinlock functions. When
2015 Apr 30
12
[PATCH 0/6] x86: reduce paravirtualized spinlock overhead
Paravirtualized spinlocks produce some overhead even if the kernel is
running on bare metal. The main reason are the more complex locking
and unlocking functions. Especially unlocking is no longer just one
instruction but so complex that it is no longer inlined.
This patch series addresses this issue by adding two more pvops
functions to reduce the size of the inlined spinlock functions. When