[mirror_ubuntu-hirsute-kernel.git] / arch / powerpc / include / asm / bitops.h

/*
 * PowerPC atomic bit operations.
 *
 * Merged version by David Gibson <david@gibson.dropbear.id.au>.
 * Based on ppc64 versions by: Dave Engebretsen, Todd Inglett, Don
 * Reed, Pat McCarthy, Peter Bergner, Anton Blanchard.  They
 * originally took it from the ppc32 code.
 *
 * Within a word, bits are numbered LSB first.  Lot's of places make
 * this assumption by directly testing bits with (val & (1<<nr)).
 * This can cause confusion for large (> 1 word) bitmaps on a
 * big-endian system because, unlike little endian, the number of each
 * bit depends on the word size.
 *
 * The bitop functions are defined to work on unsigned longs, so for a
 * ppc64 system the bits end up numbered:
 *   |63..............0|127............64|191...........128|255...........196|
 * and on ppc32:
 *   |31.....0|63....31|95....64|127...96|159..128|191..160|223..192|255..224|
 *
 * There are a few little-endian macros used mostly for filesystem
 * bitmaps, these work on similar bit arrays layouts, but
 * byte-oriented:
 *   |7...0|15...8|23...16|31...24|39...32|47...40|55...48|63...56|
 *
 * The main difference is that bit 3-5 (64b) or 3-4 (32b) in the bit
 * number field needs to be reversed compared to the big-endian bit
 * fields. This can be achieved by XOR with 0x38 (64b) or 0x18 (32b).
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version
 * 2 of the License, or (at your option) any later version.
 */

#ifndef _ASM_POWERPC_BITOPS_H
#define _ASM_POWERPC_BITOPS_H

#ifdef __KERNEL__

#ifndef _LINUX_BITOPS_H
#error only <linux/bitops.h> can be included directly
#endif

#include <linux/compiler.h>
#include <asm/asm-compat.h>
#include <asm/synch.h>

/* PPC bit number conversion */
#define PPC_BITLSHIFT(be)	(BITS_PER_LONG - 1 - (be))
#define PPC_BIT(bit)		(1UL << PPC_BITLSHIFT(bit))
#define PPC_BITMASK(bs, be)	((PPC_BIT(bs) - PPC_BIT(be)) | PPC_BIT(bs))

/*
 * clear_bit doesn't imply a memory barrier
 */
#define smp_mb__before_clear_bit()	smp_mb()
#define smp_mb__after_clear_bit()	smp_mb()

/* Macro for generating the ***_bits() functions */
#define DEFINE_BITOP(fn, op, prefix)		\
static __inline__ void fn(unsigned long mask,	\
		volatile unsigned long *_p)	\
{						\
	unsigned long old;			\
	unsigned long *p = (unsigned long *)_p;	\
	__asm__ __volatile__ (			\
	prefix					\
"1:"	PPC_LLARX(%0,0,%3,0) "\n"		\
	stringify_in_c(op) "%0,%0,%2\n"		\
	PPC405_ERR77(0,%3)			\
	PPC_STLCX "%0,0,%3\n"			\
	"bne- 1b\n"				\
	: "=&r" (old), "+m" (*p)		\
	: "r" (mask), "r" (p)			\
	: "cc", "memory");			\
}

DEFINE_BITOP(set_bits, or, "")
DEFINE_BITOP(clear_bits, andc, "")
DEFINE_BITOP(clear_bits_unlock, andc, PPC_RELEASE_BARRIER)
DEFINE_BITOP(change_bits, xor, "")

static __inline__ void set_bit(int nr, volatile unsigned long *addr)
{
	set_bits(BIT_MASK(nr), addr + BIT_WORD(nr));
}

static __inline__ void clear_bit(int nr, volatile unsigned long *addr)
{
	clear_bits(BIT_MASK(nr), addr + BIT_WORD(nr));
}

static __inline__ void clear_bit_unlock(int nr, volatile unsigned long *addr)
{
	clear_bits_unlock(BIT_MASK(nr), addr + BIT_WORD(nr));
}

static __inline__ void change_bit(int nr, volatile unsigned long *addr)
{
	change_bits(BIT_MASK(nr), addr + BIT_WORD(nr));
}

/* Like DEFINE_BITOP(), with changes to the arguments to 'op' and the output
 * operands. */
#define DEFINE_TESTOP(fn, op, prefix, postfix, eh)	\
static __inline__ unsigned long fn(			\
		unsigned long mask,			\
		volatile unsigned long *_p)		\
{							\
	unsigned long old, t;				\
	unsigned long *p = (unsigned long *)_p;		\
	__asm__ __volatile__ (				\
	prefix						\
"1:"	PPC_LLARX(%0,0,%3,eh) "\n"			\
	stringify_in_c(op) "%1,%0,%2\n"			\
	PPC405_ERR77(0,%3)				\
	PPC_STLCX "%1,0,%3\n"				\
	"bne- 1b\n"					\
	postfix						\
	: "=&r" (old), "=&r" (t)			\
	: "r" (mask), "r" (p)				\
	: "cc", "memory");				\
	return (old & mask);				\
}

DEFINE_TESTOP(test_and_set_bits, or, PPC_ATOMIC_ENTRY_BARRIER,
	      PPC_ATOMIC_EXIT_BARRIER, 0)
DEFINE_TESTOP(test_and_set_bits_lock, or, "",
	      PPC_ACQUIRE_BARRIER, 1)
DEFINE_TESTOP(test_and_clear_bits, andc, PPC_ATOMIC_ENTRY_BARRIER,
	      PPC_ATOMIC_EXIT_BARRIER, 0)
DEFINE_TESTOP(test_and_change_bits, xor, PPC_ATOMIC_ENTRY_BARRIER,
	      PPC_ATOMIC_EXIT_BARRIER, 0)

static __inline__ int test_and_set_bit(unsigned long nr,
				       volatile unsigned long *addr)
{
	return test_and_set_bits(BIT_MASK(nr), addr + BIT_WORD(nr)) != 0;
}

static __inline__ int test_and_set_bit_lock(unsigned long nr,
				       volatile unsigned long *addr)
{
	return test_and_set_bits_lock(BIT_MASK(nr),
				addr + BIT_WORD(nr)) != 0;
}

static __inline__ int test_and_clear_bit(unsigned long nr,
					 volatile unsigned long *addr)
{
	return test_and_clear_bits(BIT_MASK(nr), addr + BIT_WORD(nr)) != 0;
}

static __inline__ int test_and_change_bit(unsigned long nr,
					  volatile unsigned long *addr)
{
	return test_and_change_bits(BIT_MASK(nr), addr + BIT_WORD(nr)) != 0;
}

#include <asm-generic/bitops/non-atomic.h>

static __inline__ void __clear_bit_unlock(int nr, volatile unsigned long *addr)
{
	__asm__ __volatile__(PPC_RELEASE_BARRIER "" ::: "memory");
	__clear_bit(nr, addr);
}

/*
 * Return the zero-based bit position (LE, not IBM bit numbering) of
 * the most significant 1-bit in a double word.
 */
static __inline__ __attribute__((const))
int __ilog2(unsigned long x)
{
	int lz;

	asm (PPC_CNTLZL "%0,%1" : "=r" (lz) : "r" (x));
	return BITS_PER_LONG - 1 - lz;
}

static inline __attribute__((const))
int __ilog2_u32(u32 n)
{
	int bit;
	asm ("cntlzw %0,%1" : "=r" (bit) : "r" (n));
	return 31 - bit;
}

#ifdef __powerpc64__
static inline __attribute__((const))
int __ilog2_u64(u64 n)
{
	int bit;
	asm ("cntlzd %0,%1" : "=r" (bit) : "r" (n));
	return 63 - bit;
}
#endif

/*
 * Determines the bit position of the least significant 0 bit in the
 * specified double word. The returned bit position will be
 * zero-based, starting from the right side (63/31 - 0).
 */
static __inline__ unsigned long ffz(unsigned long x)
{
	/* no zero exists anywhere in the 8 byte area. */
	if ((x = ~x) == 0)
		return BITS_PER_LONG;

	/*
	 * Calculate the bit position of the least significant '1' bit in x
	 * (since x has been changed this will actually be the least significant
	 * '0' bit in * the original x).  Note: (x & -x) gives us a mask that
	 * is the least significant * (RIGHT-most) 1-bit of the value in x.
	 */
	return __ilog2(x & -x);
}

static __inline__ int __ffs(unsigned long x)
{
	return __ilog2(x & -x);
}

/*
 * ffs: find first bit set. This is defined the same way as
 * the libc and compiler builtin ffs routines, therefore
 * differs in spirit from the above ffz (man ffs).
 */
static __inline__ int ffs(int x)
{
	unsigned long i = (unsigned long)x;
	return __ilog2(i & -i) + 1;
}

/*
 * fls: find last (most-significant) bit set.
 * Note fls(0) = 0, fls(1) = 1, fls(0x80000000) = 32.
 */
static __inline__ int fls(unsigned int x)
{
	int lz;

	asm ("cntlzw %0,%1" : "=r" (lz) : "r" (x));
	return 32 - lz;
}

static __inline__ unsigned long __fls(unsigned long x)
{
	return __ilog2(x);
}

/*
 * 64-bit can do this using one cntlzd (count leading zeroes doubleword)
 * instruction; for 32-bit we use the generic version, which does two
 * 32-bit fls calls.
 */
#ifdef __powerpc64__
static __inline__ int fls64(__u64 x)
{
	int lz;

	asm ("cntlzd %0,%1" : "=r" (lz) : "r" (x));
	return 64 - lz;
}
#else
#include <asm-generic/bitops/fls64.h>
#endif /* __powerpc64__ */

#ifdef CONFIG_PPC64
unsigned int __arch_hweight8(unsigned int w);
unsigned int __arch_hweight16(unsigned int w);
unsigned int __arch_hweight32(unsigned int w);
unsigned long __arch_hweight64(__u64 w);
#include <asm-generic/bitops/const_hweight.h>
#else
#include <asm-generic/bitops/hweight.h>
#endif

#include <asm-generic/bitops/find.h>

/* Little-endian versions */
#include <asm-generic/bitops/le.h>

/* Bitmap functions for the ext2 filesystem */

#include <asm-generic/bitops/ext2-atomic-setbit.h>

#include <asm-generic/bitops/sched.h>

#endif /* __KERNEL__ */

#endif /* _ASM_POWERPC_BITOPS_H */
Commit	Line	Data
a0e60b20 DG	1	/*
	2	* PowerPC atomic bit operations.
	3	*
	4	* Merged version by David Gibson <david@gibson.dropbear.id.au>.
	5	* Based on ppc64 versions by: Dave Engebretsen, Todd Inglett, Don
	6	* Reed, Pat McCarthy, Peter Bergner, Anton Blanchard. They
	7	* originally took it from the ppc32 code.
	8	*
	9	* Within a word, bits are numbered LSB first. Lot's of places make
	10	* this assumption by directly testing bits with (val & (1<<nr)).
	11	* This can cause confusion for large (> 1 word) bitmaps on a
	12	* big-endian system because, unlike little endian, the number of each
	13	* bit depends on the word size.
	14	*
	15	* The bitop functions are defined to work on unsigned longs, so for a
	16	* ppc64 system the bits end up numbered:
	17	* \|63..............0\|127............64\|191...........128\|255...........196\|
	18	* and on ppc32:
	19	* \|31.....0\|63....31\|95....64\|127...96\|159..128\|191..160\|223..192\|255..224\|
	20	*
	21	* There are a few little-endian macros used mostly for filesystem
	22	* bitmaps, these work on similar bit arrays layouts, but
	23	* byte-oriented:
	24	* \|7...0\|15...8\|23...16\|31...24\|39...32\|47...40\|55...48\|63...56\|
	25	*
	26	* The main difference is that bit 3-5 (64b) or 3-4 (32b) in the bit
	27	* number field needs to be reversed compared to the big-endian bit
	28	* fields. This can be achieved by XOR with 0x38 (64b) or 0x18 (32b).
	29	*
	30	* This program is free software; you can redistribute it and/or
	31	* modify it under the terms of the GNU General Public License
	32	* as published by the Free Software Foundation; either version
	33	* 2 of the License, or (at your option) any later version.
	34	*/
	35
	36	#ifndef _ASM_POWERPC_BITOPS_H
	37	#define _ASM_POWERPC_BITOPS_H
	38
	39	#ifdef __KERNEL__
	40
0624517d JS	41	#ifndef _LINUX_BITOPS_H
	42	#error only <linux/bitops.h> can be included directly
	43	#endif
	44
a0e60b20	45	#include <linux/compiler.h>
3ddfbcf1	46	#include <asm/asm-compat.h>
a0e60b20 DG	47	#include <asm/synch.h>
a0e60b20 DG	48
e22a2274 MS	49	/* PPC bit number conversion */
	50	#define PPC_BITLSHIFT(be) (BITS_PER_LONG - 1 - (be))
	51	#define PPC_BIT(bit) (1UL << PPC_BITLSHIFT(bit))
	52	#define PPC_BITMASK(bs, be) ((PPC_BIT(bs) - PPC_BIT(be)) \| PPC_BIT(bs))
	53
a0e60b20 DG	54	/*
	55	* clear_bit doesn't imply a memory barrier
	56	*/
	57	#define smp_mb__before_clear_bit() smp_mb()
	58	#define smp_mb__after_clear_bit() smp_mb()
	59
0d2d3e38	60	/* Macro for generating the **_bits() functions /
576be130	61	#define DEFINE_BITOP(fn, op, prefix) \
0d2d3e38 GT	62	static __inline__ void fn(unsigned long mask, \
	63	volatile unsigned long *_p) \
	64	{ \
	65	unsigned long old; \
	66	unsigned long p = (unsigned long )_p; \
	67	__asm__ __volatile__ ( \
	68	prefix \
864b9e6f	69	"1:" PPC_LLARX(%0,0,%3,0) "\n" \
0d2d3e38 GT	70	stringify_in_c(op) "%0,%0,%2\n" \
	71	PPC405_ERR77(0,%3) \
	72	PPC_STLCX "%0,0,%3\n" \
	73	"bne- 1b\n" \
0d2d3e38 GT	74	: "=&r" (old), "+m" (*p) \
	75	: "r" (mask), "r" (p) \
	76	: "cc", "memory"); \
	77	}
	78
576be130 ME	79	DEFINE_BITOP(set_bits, or, "")
	80	DEFINE_BITOP(clear_bits, andc, "")
	81	DEFINE_BITOP(clear_bits_unlock, andc, PPC_RELEASE_BARRIER)
	82	DEFINE_BITOP(change_bits, xor, "")
0d2d3e38	83
a0e60b20 DG	84	static __inline__ void set_bit(int nr, volatile unsigned long *addr)
a0e60b20 DG	85	{
2237f4f4	86	set_bits(BIT_MASK(nr), addr + BIT_WORD(nr));
a0e60b20 DG	87	}
	88
	89	static __inline__ void clear_bit(int nr, volatile unsigned long *addr)
	90	{
2237f4f4	91	clear_bits(BIT_MASK(nr), addr + BIT_WORD(nr));
a0e60b20 DG	92	}
a0e60b20 DG	93
66ffb04c NP	94	static __inline__ void clear_bit_unlock(int nr, volatile unsigned long *addr)
66ffb04c NP	95	{
2237f4f4	96	clear_bits_unlock(BIT_MASK(nr), addr + BIT_WORD(nr));
66ffb04c NP	97	}
66ffb04c NP	98
a0e60b20 DG	99	static __inline__ void change_bit(int nr, volatile unsigned long *addr)
a0e60b20 DG	100	{
2237f4f4	101	change_bits(BIT_MASK(nr), addr + BIT_WORD(nr));
0d2d3e38	102	}
a0e60b20	103
0d2d3e38 GT	104	/* Like DEFINE_BITOP(), with changes to the arguments to 'op' and the output
0d2d3e38 GT	105	* operands. */
864b9e6f AB	106	#define DEFINE_TESTOP(fn, op, prefix, postfix, eh) \
	107	static __inline__ unsigned long fn( \
	108	unsigned long mask, \
	109	volatile unsigned long *_p) \
	110	{ \
	111	unsigned long old, t; \
	112	unsigned long p = (unsigned long )_p; \
	113	__asm__ __volatile__ ( \
	114	prefix \
	115	"1:" PPC_LLARX(%0,0,%3,eh) "\n" \
	116	stringify_in_c(op) "%1,%0,%2\n" \
	117	PPC405_ERR77(0,%3) \
	118	PPC_STLCX "%1,0,%3\n" \
	119	"bne- 1b\n" \
	120	postfix \
	121	: "=&r" (old), "=&r" (t) \
	122	: "r" (mask), "r" (p) \
	123	: "cc", "memory"); \
	124	return (old & mask); \
a0e60b20 DG	125	}
a0e60b20 DG	126
b97021f8 BH	127	DEFINE_TESTOP(test_and_set_bits, or, PPC_ATOMIC_ENTRY_BARRIER,
b97021f8 BH	128	PPC_ATOMIC_EXIT_BARRIER, 0)
f10e2e5b AB	129	DEFINE_TESTOP(test_and_set_bits_lock, or, "",
f10e2e5b AB	130	PPC_ACQUIRE_BARRIER, 1)
b97021f8 BH	131	DEFINE_TESTOP(test_and_clear_bits, andc, PPC_ATOMIC_ENTRY_BARRIER,
	132	PPC_ATOMIC_EXIT_BARRIER, 0)
	133	DEFINE_TESTOP(test_and_change_bits, xor, PPC_ATOMIC_ENTRY_BARRIER,
	134	PPC_ATOMIC_EXIT_BARRIER, 0)
0d2d3e38	135
a0e60b20 DG	136	static __inline__ int test_and_set_bit(unsigned long nr,
	137	volatile unsigned long *addr)
	138	{
2237f4f4	139	return test_and_set_bits(BIT_MASK(nr), addr + BIT_WORD(nr)) != 0;
a0e60b20 DG	140	}
a0e60b20 DG	141
66ffb04c NP	142	static __inline__ int test_and_set_bit_lock(unsigned long nr,
	143	volatile unsigned long *addr)
	144	{
2237f4f4 AM	145	return test_and_set_bits_lock(BIT_MASK(nr),
2237f4f4 AM	146	addr + BIT_WORD(nr)) != 0;
66ffb04c NP	147	}
66ffb04c NP	148
a0e60b20 DG	149	static __inline__ int test_and_clear_bit(unsigned long nr,
	150	volatile unsigned long *addr)
	151	{
2237f4f4	152	return test_and_clear_bits(BIT_MASK(nr), addr + BIT_WORD(nr)) != 0;
a0e60b20 DG	153	}
	154
	155	static __inline__ int test_and_change_bit(unsigned long nr,
	156	volatile unsigned long *addr)
	157	{
2237f4f4	158	return test_and_change_bits(BIT_MASK(nr), addr + BIT_WORD(nr)) != 0;
a0e60b20 DG	159	}
a0e60b20 DG	160
e779b2f9	161	#include <asm-generic/bitops/non-atomic.h>
a0e60b20	162
66ffb04c NP	163	static __inline__ void __clear_bit_unlock(int nr, volatile unsigned long *addr)
66ffb04c NP	164	{
f10e2e5b	165	__asm__ __volatile__(PPC_RELEASE_BARRIER "" ::: "memory");
66ffb04c NP	166	__clear_bit(nr, addr);
	167	}
	168
a0e60b20 DG	169	/*
	170	* Return the zero-based bit position (LE, not IBM bit numbering) of
	171	* the most significant 1-bit in a double word.
	172	*/
ef55d53c DH	173	static __inline__ __attribute__((const))
ef55d53c DH	174	int __ilog2(unsigned long x)
a0e60b20 DG	175	{
	176	int lz;
	177
3ddfbcf1	178	asm (PPC_CNTLZL "%0,%1" : "=r" (lz) : "r" (x));
a0e60b20 DG	179	return BITS_PER_LONG - 1 - lz;
	180	}
	181
ef55d53c DH	182	static inline __attribute__((const))
	183	int __ilog2_u32(u32 n)
	184	{
	185	int bit;
	186	asm ("cntlzw %0,%1" : "=r" (bit) : "r" (n));
	187	return 31 - bit;
	188	}
	189
	190	#ifdef __powerpc64__
	191	static inline __attribute__((const))
02241696	192	int __ilog2_u64(u64 n)
ef55d53c DH	193	{
	194	int bit;
	195	asm ("cntlzd %0,%1" : "=r" (bit) : "r" (n));
	196	return 63 - bit;
	197	}
	198	#endif
	199
a0e60b20 DG	200	/*
	201	* Determines the bit position of the least significant 0 bit in the
	202	* specified double word. The returned bit position will be
	203	* zero-based, starting from the right side (63/31 - 0).
	204	*/
	205	static __inline__ unsigned long ffz(unsigned long x)
	206	{
	207	/* no zero exists anywhere in the 8 byte area. */
	208	if ((x = ~x) == 0)
	209	return BITS_PER_LONG;
	210
	211	/*
25985edc LDM	212	* Calculate the bit position of the least significant '1' bit in x
25985edc LDM	213	* (since x has been changed this will actually be the least significant
a0e60b20 DG	214	* '0' bit in * the original x). Note: (x & -x) gives us a mask that
	215	* is the least significant * (RIGHT-most) 1-bit of the value in x.
	216	*/
	217	return __ilog2(x & -x);
	218	}
	219
	220	static __inline__ int __ffs(unsigned long x)
	221	{
	222	return __ilog2(x & -x);
	223	}
	224
	225	/*
	226	* ffs: find first bit set. This is defined the same way as
	227	* the libc and compiler builtin ffs routines, therefore
	228	* differs in spirit from the above ffz (man ffs).
	229	*/
	230	static __inline__ int ffs(int x)
	231	{
	232	unsigned long i = (unsigned long)x;
	233	return __ilog2(i & -i) + 1;
	234	}
	235
	236	/*
	237	* fls: find last (most-significant) bit set.
	238	* Note fls(0) = 0, fls(1) = 1, fls(0x80000000) = 32.
	239	*/
	240	static __inline__ int fls(unsigned int x)
	241	{
	242	int lz;
	243
	244	asm ("cntlzw %0,%1" : "=r" (lz) : "r" (x));
	245	return 32 - lz;
	246	}
9f264be6	247
56a6b1eb AH	248	static __inline__ unsigned long __fls(unsigned long x)
	249	{
	250	return __ilog2(x);
	251	}
	252
9f264be6 PM	253	/*
	254	* 64-bit can do this using one cntlzd (count leading zeroes doubleword)
	255	* instruction; for 32-bit we use the generic version, which does two
	256	* 32-bit fls calls.
	257	*/
	258	#ifdef __powerpc64__
	259	static __inline__ int fls64(__u64 x)
	260	{
	261	int lz;
	262
	263	asm ("cntlzd %0,%1" : "=r" (lz) : "r" (x));
	264	return 64 - lz;
	265	}
	266	#else
e779b2f9	267	#include <asm-generic/bitops/fls64.h>
9f264be6 PM	268	#endif /* __powerpc64__ */
9f264be6 PM	269
64ff3128 AB	270	#ifdef CONFIG_PPC64
	271	unsigned int __arch_hweight8(unsigned int w);
	272	unsigned int __arch_hweight16(unsigned int w);
	273	unsigned int __arch_hweight32(unsigned int w);
	274	unsigned long __arch_hweight64(__u64 w);
	275	#include <asm-generic/bitops/const_hweight.h>
	276	#else
e779b2f9	277	#include <asm-generic/bitops/hweight.h>
64ff3128 AB	278	#endif
64ff3128 AB	279
47b9d9bd	280	#include <asm-generic/bitops/find.h>
a0e60b20 DG	281
a0e60b20 DG	282	/* Little-endian versions */
79597be9	283	#include <asm-generic/bitops/le.h>
a0e60b20	284
a0e60b20 DG	285	/* Bitmap functions for the ext2 filesystem */
a0e60b20 DG	286
148817ba	287	#include <asm-generic/bitops/ext2-atomic-setbit.h>
a0e60b20	288
e779b2f9	289	#include <asm-generic/bitops/sched.h>
a0e60b20 DG	290
	291	#endif /* __KERNEL__ */
	292
	293	#endif /* _ASM_POWERPC_BITOPS_H */