/*
 * Copyright (C) 2001-2004 Jakub Jermar
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 *
 * - Redistributions of source code must retain the above copyright
 *   notice, this list of conditions and the following disclaimer.
 * - Redistributions in binary form must reproduce the above copyright
 *   notice, this list of conditions and the following disclaimer in the
 *   documentation and/or other materials provided with the distribution.
 * - The name of the author may not be used to endorse or promote products
 *   derived from this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

#ifndef __amd64_ATOMIC_H__
#define __amd64_ATOMIC_H__

#include <arch/types.h>
#include <arch/barrier.h>
#include <preemption.h>
#include <typedefs.h>

static inline void atomic_inc(atomic_t *val) {
#ifdef CONFIG_SMP
	__asm__ volatile ("lock incq %0\n" : "=m" (val->count));
#else
	__asm__ volatile ("incq %0\n" : "=m" (val->count));
#endif /* CONFIG_SMP */
}

static inline void atomic_dec(atomic_t *val) {
#ifdef CONFIG_SMP
	__asm__ volatile ("lock decq %0\n" : "=m" (val->count));
#else
	__asm__ volatile ("decq %0\n" : "=m" (val->count));
#endif /* CONFIG_SMP */
}

static inline long atomic_postinc(atomic_t *val) 
{
	long r = 1;

	__asm__ volatile (
		"lock xaddq %1, %0\n"
		: "=m" (val->count) : "r" (r)
	);

	return r;
}

static inline long atomic_postdec(atomic_t *val) 
{
	long r = -1;
	
	__asm__ volatile (
		"lock xaddq %1, %0\n"
		: "=m" (val->count) : "r" (r)
	);
	
	return r;
}

#define atomic_preinc(val) (atomic_postinc(val)+1)
#define atomic_predec(val) (atomic_postdec(val)-1)

static inline __u64 test_and_set(atomic_t *val) {
	__u64 v;
	
	__asm__ volatile (
		"movq $1, %0\n"
		"xchgq %0, %1\n"
		: "=r" (v),"=m" (val->count)
	);
	
	return v;
}


/** amd64 specific fast spinlock */
static inline void atomic_lock_arch(atomic_t *val)
{
	__u64 tmp;

	preemption_disable();
	__asm__ volatile (
		"0:;"
#ifdef CONFIG_HT
		"pause;"
#endif
		"mov %0, %1;"
		"testq %1, %1;"
		"jnz 0b;"       /* Lightweight looping on locked spinlock */
		
		"incq %1;"      /* now use the atomic operation */
		"xchgq %0, %1;"
		"testq %1, %1;"
		"jnz 0b;"
                : "=m"(val->count),"=r"(tmp)
		);
	/*
	 * Prevent critical section code from bleeding out this way up.
	 */
	CS_ENTER_BARRIER();
}

#endif