[linux.git] / lib / lzo / lzo1x_compress.c

/*
 *  LZO1X Compressor from LZO
 *
 *  Copyright (C) 1996-2012 Markus F.X.J. Oberhumer <[email protected]>
 *
 *  The full LZO package can be found at:
 *  http://www.oberhumer.com/opensource/lzo/
 *
 *  Changed for Linux kernel use by:
 *  Nitin Gupta <[email protected]>
 *  Richard Purdie <[email protected]>
 */

#include <linux/module.h>
#include <linux/kernel.h>
#include <asm/unaligned.h>
#include <linux/lzo.h>
#include "lzodefs.h"

static noinline size_t
lzo1x_1_do_compress(const unsigned char *in, size_t in_len,
		    unsigned char *out, size_t *out_len,
		    size_t ti, void *wrkmem)
{
	const unsigned char *ip;
	unsigned char *op;
	const unsigned char * const in_end = in + in_len;
	const unsigned char * const ip_end = in + in_len - 20;
	const unsigned char *ii;
	lzo_dict_t * const dict = (lzo_dict_t *) wrkmem;

	op = out;
	ip = in;
	ii = ip;
	ip += ti < 4 ? 4 - ti : 0;

	for (;;) {
		const unsigned char *m_pos;
		size_t t, m_len, m_off;
		u32 dv;
literal:
		ip += 1 + ((ip - ii) >> 5);
next:
		if (unlikely(ip >= ip_end))
			break;
		dv = get_unaligned_le32(ip);
		t = ((dv * 0x1824429d) >> (32 - D_BITS)) & D_MASK;
		m_pos = in + dict[t];
		dict[t] = (lzo_dict_t) (ip - in);
		if (unlikely(dv != get_unaligned_le32(m_pos)))
			goto literal;

		ii -= ti;
		ti = 0;
		t = ip - ii;
		if (t != 0) {
			if (t <= 3) {
				op[-2] |= t;
				COPY4(op, ii);
				op += t;
			} else if (t <= 16) {
				*op++ = (t - 3);
				COPY8(op, ii);
				COPY8(op + 8, ii + 8);
				op += t;
			} else {
				if (t <= 18) {
					*op++ = (t - 3);
				} else {
					size_t tt = t - 18;
					*op++ = 0;
					while (unlikely(tt > 255)) {
						tt -= 255;
						*op++ = 0;
					}
					*op++ = tt;
				}
				do {
					COPY8(op, ii);
					COPY8(op + 8, ii + 8);
					op += 16;
					ii += 16;
					t -= 16;
				} while (t >= 16);
				if (t > 0) do {
					*op++ = *ii++;
				} while (--t > 0);
			}
		}

		m_len = 4;
		{
#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) && defined(LZO_USE_CTZ64)
		u64 v;
		v = get_unaligned((const u64 *) (ip + m_len)) ^
		    get_unaligned((const u64 *) (m_pos + m_len));
		if (unlikely(v == 0)) {
			do {
				m_len += 8;
				v = get_unaligned((const u64 *) (ip + m_len)) ^
				    get_unaligned((const u64 *) (m_pos + m_len));
				if (unlikely(ip + m_len >= ip_end))
					goto m_len_done;
			} while (v == 0);
		}
#  if defined(__LITTLE_ENDIAN)
		m_len += (unsigned) __builtin_ctzll(v) / 8;
#  elif defined(__BIG_ENDIAN)
		m_len += (unsigned) __builtin_clzll(v) / 8;
#  else
#    error "missing endian definition"
#  endif
#elif defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) && defined(LZO_USE_CTZ32)
		u32 v;
		v = get_unaligned((const u32 *) (ip + m_len)) ^
		    get_unaligned((const u32 *) (m_pos + m_len));
		if (unlikely(v == 0)) {
			do {
				m_len += 4;
				v = get_unaligned((const u32 *) (ip + m_len)) ^
				    get_unaligned((const u32 *) (m_pos + m_len));
				if (v != 0)
					break;
				m_len += 4;
				v = get_unaligned((const u32 *) (ip + m_len)) ^
				    get_unaligned((const u32 *) (m_pos + m_len));
				if (unlikely(ip + m_len >= ip_end))
					goto m_len_done;
			} while (v == 0);
		}
#  if defined(__LITTLE_ENDIAN)
		m_len += (unsigned) __builtin_ctz(v) / 8;
#  elif defined(__BIG_ENDIAN)
		m_len += (unsigned) __builtin_clz(v) / 8;
#  else
#    error "missing endian definition"
#  endif
#else
		if (unlikely(ip[m_len] == m_pos[m_len])) {
			do {
				m_len += 1;
				if (ip[m_len] != m_pos[m_len])
					break;
				m_len += 1;
				if (ip[m_len] != m_pos[m_len])
					break;
				m_len += 1;
				if (ip[m_len] != m_pos[m_len])
					break;
				m_len += 1;
				if (ip[m_len] != m_pos[m_len])
					break;
				m_len += 1;
				if (ip[m_len] != m_pos[m_len])
					break;
				m_len += 1;
				if (ip[m_len] != m_pos[m_len])
					break;
				m_len += 1;
				if (ip[m_len] != m_pos[m_len])
					break;
				m_len += 1;
				if (unlikely(ip + m_len >= ip_end))
					goto m_len_done;
			} while (ip[m_len] == m_pos[m_len]);
		}
#endif
		}
m_len_done:

		m_off = ip - m_pos;
		ip += m_len;
		ii = ip;
		if (m_len <= M2_MAX_LEN && m_off <= M2_MAX_OFFSET) {
			m_off -= 1;
			*op++ = (((m_len - 1) << 5) | ((m_off & 7) << 2));
			*op++ = (m_off >> 3);
		} else if (m_off <= M3_MAX_OFFSET) {
			m_off -= 1;
			if (m_len <= M3_MAX_LEN)
				*op++ = (M3_MARKER | (m_len - 2));
			else {
				m_len -= M3_MAX_LEN;
				*op++ = M3_MARKER | 0;
				while (unlikely(m_len > 255)) {
					m_len -= 255;
					*op++ = 0;
				}
				*op++ = (m_len);
			}
			*op++ = (m_off << 2);
			*op++ = (m_off >> 6);
		} else {
			m_off -= 0x4000;
			if (m_len <= M4_MAX_LEN)
				*op++ = (M4_MARKER | ((m_off >> 11) & 8)
						| (m_len - 2));
			else {
				m_len -= M4_MAX_LEN;
				*op++ = (M4_MARKER | ((m_off >> 11) & 8));
				while (unlikely(m_len > 255)) {
					m_len -= 255;
					*op++ = 0;
				}
				*op++ = (m_len);
			}
			*op++ = (m_off << 2);
			*op++ = (m_off >> 6);
		}
		goto next;
	}
	*out_len = op - out;
	return in_end - (ii - ti);
}

int lzo1x_1_compress(const unsigned char *in, size_t in_len,
		     unsigned char *out, size_t *out_len,
		     void *wrkmem)
{
	const unsigned char *ip = in;
	unsigned char *op = out;
	size_t l = in_len;
	size_t t = 0;

	while (l > 20) {
		size_t ll = l <= (M4_MAX_OFFSET + 1) ? l : (M4_MAX_OFFSET + 1);
		uintptr_t ll_end = (uintptr_t) ip + ll;
		if ((ll_end + ((t + ll) >> 5)) <= ll_end)
			break;
		BUILD_BUG_ON(D_SIZE * sizeof(lzo_dict_t) > LZO1X_1_MEM_COMPRESS);
		memset(wrkmem, 0, D_SIZE * sizeof(lzo_dict_t));
		t = lzo1x_1_do_compress(ip, ll, op, out_len, t, wrkmem);
		ip += ll;
		op += *out_len;
		l  -= ll;
	}
	t += l;

	if (t > 0) {
		const unsigned char *ii = in + in_len - t;

		if (op == out && t <= 238) {
			*op++ = (17 + t);
		} else if (t <= 3) {
			op[-2] |= t;
		} else if (t <= 18) {
			*op++ = (t - 3);
		} else {
			size_t tt = t - 18;
			*op++ = 0;
			while (tt > 255) {
				tt -= 255;
				*op++ = 0;
			}
			*op++ = tt;
		}
		if (t >= 16) do {
			COPY8(op, ii);
			COPY8(op + 8, ii + 8);
			op += 16;
			ii += 16;
			t -= 16;
		} while (t >= 16);
		if (t > 0) do {
			*op++ = *ii++;
		} while (--t > 0);
	}

	*op++ = M4_MARKER | 1;
	*op++ = 0;
	*op++ = 0;

	*out_len = op - out;
	return LZO_E_OK;
}
EXPORT_SYMBOL_GPL(lzo1x_1_compress);

MODULE_LICENSE("GPL");
MODULE_DESCRIPTION("LZO1X-1 Compressor");
Commit	Line	Data
64c70b1c	1	/*
8b975bd3	2	* LZO1X Compressor from LZO
64c70b1c	3	*
8b975bd3	4	* Copyright (C) 1996-2012 Markus F.X.J. Oberhumer <[email protected]>
64c70b1c RP	5	*
	6	* The full LZO package can be found at:
	7	* http://www.oberhumer.com/opensource/lzo/
	8	*
8b975bd3	9	* Changed for Linux kernel use by:
64c70b1c RP	10	* Nitin Gupta <[email protected]>
	11	* Richard Purdie <[email protected]>
	12	*/
	13
	14	#include <linux/module.h>
	15	#include <linux/kernel.h>
64c70b1c	16	#include <asm/unaligned.h>
8b975bd3	17	#include <linux/lzo.h>
64c70b1c RP	18	#include "lzodefs.h"
	19
	20	static noinline size_t
8b975bd3 MO	21	lzo1x_1_do_compress(const unsigned char *in, size_t in_len,
	22	unsigned char out, size_t out_len,
	23	size_t ti, void *wrkmem)
64c70b1c	24	{
8b975bd3 MO	25	const unsigned char *ip;
8b975bd3 MO	26	unsigned char *op;
64c70b1c	27	const unsigned char * const in_end = in + in_len;
8b975bd3 MO	28	const unsigned char * const ip_end = in + in_len - 20;
	29	const unsigned char *ii;
	30	lzo_dict_t * const dict = (lzo_dict_t *) wrkmem;
64c70b1c	31
8b975bd3 MO	32	op = out;
	33	ip = in;
	34	ii = ip;
	35	ip += ti < 4 ? 4 - ti : 0;
64c70b1c RP	36
64c70b1c RP	37	for (;;) {
8b975bd3 MO	38	const unsigned char *m_pos;
	39	size_t t, m_len, m_off;
	40	u32 dv;
64c70b1c	41	literal:
8b975bd3 MO	42	ip += 1 + ((ip - ii) >> 5);
8b975bd3 MO	43	next:
64c70b1c RP	44	if (unlikely(ip >= ip_end))
64c70b1c RP	45	break;
8b975bd3 MO	46	dv = get_unaligned_le32(ip);
	47	t = ((dv * 0x1824429d) >> (32 - D_BITS)) & D_MASK;
	48	m_pos = in + dict[t];
	49	dict[t] = (lzo_dict_t) (ip - in);
	50	if (unlikely(dv != get_unaligned_le32(m_pos)))
	51	goto literal;
64c70b1c	52
8b975bd3 MO	53	ii -= ti;
	54	ti = 0;
	55	t = ip - ii;
	56	if (t != 0) {
64c70b1c RP	57	if (t <= 3) {
64c70b1c RP	58	op[-2] \|= t;
8b975bd3 MO	59	COPY4(op, ii);
	60	op += t;
	61	} else if (t <= 16) {
64c70b1c	62	*op++ = (t - 3);
8b975bd3 MO	63	COPY8(op, ii);
	64	COPY8(op + 8, ii + 8);
	65	op += t;
64c70b1c	66	} else {
8b975bd3 MO	67	if (t <= 18) {
	68	*op++ = (t - 3);
	69	} else {
	70	size_t tt = t - 18;
64c70b1c	71	*op++ = 0;
8b975bd3 MO	72	while (unlikely(tt > 255)) {
	73	tt -= 255;
	74	*op++ = 0;
	75	}
	76	*op++ = tt;
64c70b1c	77	}
8b975bd3 MO	78	do {
	79	COPY8(op, ii);
	80	COPY8(op + 8, ii + 8);
	81	op += 16;
	82	ii += 16;
	83	t -= 16;
	84	} while (t >= 16);
	85	if (t > 0) do {
	86	op++ = ii++;
	87	} while (--t > 0);
64c70b1c	88	}
64c70b1c RP	89	}
64c70b1c RP	90
8b975bd3 MO	91	m_len = 4;
	92	{
	93	#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) && defined(LZO_USE_CTZ64)
	94	u64 v;
	95	v = get_unaligned((const u64 *) (ip + m_len)) ^
	96	get_unaligned((const u64 *) (m_pos + m_len));
	97	if (unlikely(v == 0)) {
	98	do {
	99	m_len += 8;
	100	v = get_unaligned((const u64 *) (ip + m_len)) ^
	101	get_unaligned((const u64 *) (m_pos + m_len));
	102	if (unlikely(ip + m_len >= ip_end))
	103	goto m_len_done;
	104	} while (v == 0);
	105	}
	106	# if defined(__LITTLE_ENDIAN)
	107	m_len += (unsigned) __builtin_ctzll(v) / 8;
	108	# elif defined(__BIG_ENDIAN)
	109	m_len += (unsigned) __builtin_clzll(v) / 8;
	110	# else
	111	# error "missing endian definition"
	112	# endif
	113	#elif defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) && defined(LZO_USE_CTZ32)
	114	u32 v;
	115	v = get_unaligned((const u32 *) (ip + m_len)) ^
	116	get_unaligned((const u32 *) (m_pos + m_len));
	117	if (unlikely(v == 0)) {
	118	do {
	119	m_len += 4;
	120	v = get_unaligned((const u32 *) (ip + m_len)) ^
	121	get_unaligned((const u32 *) (m_pos + m_len));
	122	if (v != 0)
	123	break;
	124	m_len += 4;
	125	v = get_unaligned((const u32 *) (ip + m_len)) ^
	126	get_unaligned((const u32 *) (m_pos + m_len));
	127	if (unlikely(ip + m_len >= ip_end))
	128	goto m_len_done;
	129	} while (v == 0);
	130	}
	131	# if defined(__LITTLE_ENDIAN)
	132	m_len += (unsigned) __builtin_ctz(v) / 8;
	133	# elif defined(__BIG_ENDIAN)
	134	m_len += (unsigned) __builtin_clz(v) / 8;
	135	# else
	136	# error "missing endian definition"
	137	# endif
	138	#else
	139	if (unlikely(ip[m_len] == m_pos[m_len])) {
	140	do {
	141	m_len += 1;
	142	if (ip[m_len] != m_pos[m_len])
	143	break;
	144	m_len += 1;
	145	if (ip[m_len] != m_pos[m_len])
	146	break;
	147	m_len += 1;
	148	if (ip[m_len] != m_pos[m_len])
	149	break;
	150	m_len += 1;
	151	if (ip[m_len] != m_pos[m_len])
	152	break;
	153	m_len += 1;
	154	if (ip[m_len] != m_pos[m_len])
155	break;
156	m_len += 1;
157	if (ip[m_len] != m_pos[m_len])
158	break;
159	m_len += 1;
160	if (ip[m_len] != m_pos[m_len])
161	break;
162	m_len += 1;
163	if (unlikely(ip + m_len >= ip_end))
164	goto m_len_done;
165	} while (ip[m_len] == m_pos[m_len]);
166	}
167	#endif
168	}
169	m_len_done:
64c70b1c	170
8b975bd3 MO	171	m_off = ip - m_pos;
	172	ip += m_len;
	173	ii = ip;
	174	if (m_len <= M2_MAX_LEN && m_off <= M2_MAX_OFFSET) {
	175	m_off -= 1;
	176	*op++ = (((m_len - 1) << 5) \| ((m_off & 7) << 2));
	177	*op++ = (m_off >> 3);
	178	} else if (m_off <= M3_MAX_OFFSET) {
	179	m_off -= 1;
	180	if (m_len <= M3_MAX_LEN)
64c70b1c	181	*op++ = (M3_MARKER \| (m_len - 2));
8b975bd3 MO	182	else {
	183	m_len -= M3_MAX_LEN;
	184	*op++ = M3_MARKER \| 0;
	185	while (unlikely(m_len > 255)) {
	186	m_len -= 255;
	187	*op++ = 0;
	188	}
	189	*op++ = (m_len);
64c70b1c	190	}
8b975bd3 MO	191	*op++ = (m_off << 2);
8b975bd3 MO	192	*op++ = (m_off >> 6);
64c70b1c	193	} else {
8b975bd3 MO	194	m_off -= 0x4000;
	195	if (m_len <= M4_MAX_LEN)
	196	*op++ = (M4_MARKER \| ((m_off >> 11) & 8)
64c70b1c	197	\| (m_len - 2));
8b975bd3 MO	198	else {
	199	m_len -= M4_MAX_LEN;
	200	*op++ = (M4_MARKER \| ((m_off >> 11) & 8));
	201	while (unlikely(m_len > 255)) {
	202	m_len -= 255;
	203	*op++ = 0;
64c70b1c	204	}
8b975bd3	205	*op++ = (m_len);
64c70b1c	206	}
8b975bd3	207	*op++ = (m_off << 2);
64c70b1c RP	208	*op++ = (m_off >> 6);
64c70b1c RP	209	}
8b975bd3	210	goto next;
64c70b1c	211	}
64c70b1c	212	*out_len = op - out;
8b975bd3	213	return in_end - (ii - ti);
64c70b1c RP	214	}
64c70b1c RP	215
8b975bd3 MO	216	int lzo1x_1_compress(const unsigned char *in, size_t in_len,
	217	unsigned char out, size_t out_len,
	218	void *wrkmem)
64c70b1c	219	{
8b975bd3	220	const unsigned char *ip = in;
64c70b1c	221	unsigned char *op = out;
8b975bd3 MO	222	size_t l = in_len;
8b975bd3 MO	223	size_t t = 0;
64c70b1c	224
8b975bd3 MO	225	while (l > 20) {
	226	size_t ll = l <= (M4_MAX_OFFSET + 1) ? l : (M4_MAX_OFFSET + 1);
	227	uintptr_t ll_end = (uintptr_t) ip + ll;
	228	if ((ll_end + ((t + ll) >> 5)) <= ll_end)
	229	break;
	230	BUILD_BUG_ON(D_SIZE * sizeof(lzo_dict_t) > LZO1X_1_MEM_COMPRESS);
	231	memset(wrkmem, 0, D_SIZE * sizeof(lzo_dict_t));
	232	t = lzo1x_1_do_compress(ip, ll, op, out_len, t, wrkmem);
	233	ip += ll;
64c70b1c	234	op += *out_len;
8b975bd3	235	l -= ll;
64c70b1c	236	}
8b975bd3	237	t += l;
64c70b1c RP	238
64c70b1c RP	239	if (t > 0) {
8b975bd3	240	const unsigned char *ii = in + in_len - t;
64c70b1c RP	241
	242	if (op == out && t <= 238) {
	243	*op++ = (17 + t);
	244	} else if (t <= 3) {
	245	op[-2] \|= t;
	246	} else if (t <= 18) {
	247	*op++ = (t - 3);
	248	} else {
	249	size_t tt = t - 18;
64c70b1c RP	250	*op++ = 0;
	251	while (tt > 255) {
	252	tt -= 255;
	253	*op++ = 0;
	254	}
64c70b1c RP	255	*op++ = tt;
64c70b1c RP	256	}
8b975bd3 MO	257	if (t >= 16) do {
	258	COPY8(op, ii);
	259	COPY8(op + 8, ii + 8);
	260	op += 16;
	261	ii += 16;
	262	t -= 16;
	263	} while (t >= 16);
	264	if (t > 0) do {
64c70b1c RP	265	op++ = ii++;
	266	} while (--t > 0);
	267	}
	268
	269	*op++ = M4_MARKER \| 1;
	270	*op++ = 0;
	271	*op++ = 0;
	272
	273	*out_len = op - out;
	274	return LZO_E_OK;
	275	}
	276	EXPORT_SYMBOL_GPL(lzo1x_1_compress);
	277
	278	MODULE_LICENSE("GPL");
	279	MODULE_DESCRIPTION("LZO1X-1 Compressor");