Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit b0af9cd9 authored by Linus Torvalds's avatar Linus Torvalds
Browse files

Merge tag 'lzo-update-signature-20130226' of git://github.com/markus-oberhumer/linux

Pull LZO compression update from Markus Oberhumer:
 "Summary:
  ========

  Update the Linux kernel LZO compression and decompression code to the
  current upstream version which features significant performance
  improvements on modern machines.

  Some *synthetic* benchmarks:
  ============================

    x86_64 (Sandy Bridge), gcc-4.6 -O3, Silesia test corpus, 256 kB block-size:

                     compression speed   decompression speed

    LZO-2005    :         150 MB/sec          468 MB/sec
    LZO-2012    :         434 MB/sec         1210 MB/sec

    i386 (Sandy Bridge), gcc-4.6 -O3, Silesia test corpus, 256 kB block-size:

                     compression speed   decompression speed

    LZO-2005    :         143 MB/sec          409 MB/sec
    LZO-2012    :         372 MB/sec         1121 MB/sec

    armv7 (Cortex-A9), Linaro gcc-4.6 -O3, Silesia test corpus, 256 kB block-size:

                     compression speed   decompression speed

    LZO-2005    :          27 MB/sec           84 MB/sec
    LZO-2012    :          44 MB/sec          117 MB/sec
  **LZO-2013-UA :          47 MB/sec          167 MB/sec

  Legend:

    LZO-2005    : LZO version in current 3.8 kernel (which is based on
                     the LZO 2.02 release from 2005)
    LZO-2012    : updated LZO version available in linux-next
  **LZO-2013-UA : updated LZO version available in linux-next plus experimental
                     ARM Unaligned Access patch. This needs approval
                     from some ARM maintainer ist NOT YET INCLUDED."

Andrew Morton <akpm@linux-foundation.org> acks it and says:
 "There's a new LZ4 on the block which is even faster than the sped-up
  LZO, but various filesystems and things use LZO"

* tag 'lzo-update-signature-20130226' of git://github.com/markus-oberhumer/linux:
  crypto: testmgr - update LZO compression test vectors
  lib/lzo: Update LZO compression to current upstream version
  lib/lzo: Rename lzo1x_decompress.c to lzo1x_decompress_safe.c
parents 1e500597 0ec73820
Loading
Loading
Loading
Loading
+20 −18
Original line number Diff line number Diff line
@@ -25084,38 +25084,40 @@ static struct pcomp_testvec zlib_decomp_tv_template[] = {
static struct comp_testvec lzo_comp_tv_template[] = {
	{
		.inlen	= 70,
		.outlen	= 46,
		.outlen	= 57,
		.input	= "Join us now and share the software "
			"Join us now and share the software ",
		.output	= "\x00\x0d\x4a\x6f\x69\x6e\x20\x75"
			  "\x73\x20\x6e\x6f\x77\x20\x61\x6e"
			  "\x64\x20\x73\x68\x61\x72\x65\x20"
			  "\x74\x68\x65\x20\x73\x6f\x66\x74"
			"\x77\x70\x01\x01\x4a\x6f\x69\x6e"
			"\x3d\x88\x00\x11\x00\x00",
			  "\x77\x70\x01\x32\x88\x00\x0c\x65"
			  "\x20\x74\x68\x65\x20\x73\x6f\x66"
			  "\x74\x77\x61\x72\x65\x20\x11\x00"
			  "\x00",
	}, {
		.inlen	= 159,
		.outlen	= 133,
		.outlen	= 131,
		.input	= "This document describes a compression method based on the LZO "
			"compression algorithm.  This document defines the application of "
			"the LZO algorithm used in UBIFS.",
		.output	= "\x00\x2b\x54\x68\x69\x73\x20\x64"
		.output	= "\x00\x2c\x54\x68\x69\x73\x20\x64"
			  "\x6f\x63\x75\x6d\x65\x6e\x74\x20"
			  "\x64\x65\x73\x63\x72\x69\x62\x65"
			  "\x73\x20\x61\x20\x63\x6f\x6d\x70"
			  "\x72\x65\x73\x73\x69\x6f\x6e\x20"
			  "\x6d\x65\x74\x68\x6f\x64\x20\x62"
			  "\x61\x73\x65\x64\x20\x6f\x6e\x20"
			  "\x74\x68\x65\x20\x4c\x5a\x4f\x2b"
			  "\x8c\x00\x0d\x61\x6c\x67\x6f\x72"
			  "\x69\x74\x68\x6d\x2e\x20\x20\x54"
			  "\x68\x69\x73\x2a\x54\x01\x02\x66"
			  "\x69\x6e\x65\x73\x94\x06\x05\x61"
			  "\x70\x70\x6c\x69\x63\x61\x74\x76"
			  "\x0a\x6f\x66\x88\x02\x60\x09\x27"
			  "\xf0\x00\x0c\x20\x75\x73\x65\x64"
			  "\x20\x69\x6e\x20\x55\x42\x49\x46"
			  "\x53\x2e\x11\x00\x00",
			  "\x74\x68\x65\x20\x4c\x5a\x4f\x20"
			  "\x2a\x8c\x00\x09\x61\x6c\x67\x6f"
			  "\x72\x69\x74\x68\x6d\x2e\x20\x20"
			  "\x2e\x54\x01\x03\x66\x69\x6e\x65"
			  "\x73\x20\x74\x06\x05\x61\x70\x70"
			  "\x6c\x69\x63\x61\x74\x76\x0a\x6f"
			  "\x66\x88\x02\x60\x09\x27\xf0\x00"
			  "\x0c\x20\x75\x73\x65\x64\x20\x69"
			  "\x6e\x20\x55\x42\x49\x46\x53\x2e"
			  "\x11\x00\x00",
	},
};
+8 −7
Original line number Diff line number Diff line
@@ -4,22 +4,22 @@
 *  LZO Public Kernel Interface
 *  A mini subset of the LZO real-time data compression library
 *
 *  Copyright (C) 1996-2005 Markus F.X.J. Oberhumer <markus@oberhumer.com>
 *  Copyright (C) 1996-2012 Markus F.X.J. Oberhumer <markus@oberhumer.com>
 *
 *  The full LZO package can be found at:
 *  http://www.oberhumer.com/opensource/lzo/
 *
 *  Changed for kernel use by:
 *  Changed for Linux kernel use by:
 *  Nitin Gupta <nitingupta910@gmail.com>
 *  Richard Purdie <rpurdie@openedhand.com>
 */

#define LZO1X_MEM_COMPRESS	(16384 * sizeof(unsigned char *))
#define LZO1X_1_MEM_COMPRESS	LZO1X_MEM_COMPRESS
#define LZO1X_1_MEM_COMPRESS	(8192 * sizeof(unsigned short))
#define LZO1X_MEM_COMPRESS	LZO1X_1_MEM_COMPRESS

#define lzo1x_worst_compress(x) ((x) + ((x) / 16) + 64 + 3)

/* This requires 'workmem' of size LZO1X_1_MEM_COMPRESS */
/* This requires 'wrkmem' of size LZO1X_1_MEM_COMPRESS */
int lzo1x_1_compress(const unsigned char *src, size_t src_len,
		     unsigned char *dst, size_t *dst_len, void *wrkmem);

@@ -40,5 +40,6 @@ int lzo1x_decompress_safe(const unsigned char *src, size_t src_len,
#define LZO_E_EOF_NOT_FOUND		(-7)
#define LZO_E_INPUT_NOT_CONSUMED	(-8)
#define LZO_E_NOT_YET_IMPLEMENTED	(-9)
#define LZO_E_INVALID_ARGUMENT		(-10)

#endif
+1 −1
Original line number Diff line number Diff line
@@ -31,7 +31,7 @@
 */

#ifdef STATIC
#include "lzo/lzo1x_decompress.c"
#include "lzo/lzo1x_decompress_safe.c"
#else
#include <linux/decompress/unlzo.h>
#endif
+1 −1
Original line number Diff line number Diff line
lzo_compress-objs := lzo1x_compress.o
lzo_decompress-objs := lzo1x_decompress.o
lzo_decompress-objs := lzo1x_decompress_safe.o

obj-$(CONFIG_LZO_COMPRESS) += lzo_compress.o
obj-$(CONFIG_LZO_DECOMPRESS) += lzo_decompress.o
+194 −141
Original line number Diff line number Diff line
/*
 *  LZO1X Compressor from MiniLZO
 *  LZO1X Compressor from LZO
 *
 *  Copyright (C) 1996-2005 Markus F.X.J. Oberhumer <markus@oberhumer.com>
 *  Copyright (C) 1996-2012 Markus F.X.J. Oberhumer <markus@oberhumer.com>
 *
 *  The full LZO package can be found at:
 *  http://www.oberhumer.com/opensource/lzo/
 *
 *  Changed for kernel use by:
 *  Changed for Linux kernel use by:
 *  Nitin Gupta <nitingupta910@gmail.com>
 *  Richard Purdie <rpurdie@openedhand.com>
 */

#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/lzo.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, void *wrkmem)
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 - M2_MAX_LEN - 5;
	const unsigned char ** const dict = wrkmem;
	const unsigned char *ip = in, *ii = ip;
	const unsigned char *end, *m, *m_pos;
	size_t m_off, m_len, dindex;
	unsigned char *op = out;
	const unsigned char * const ip_end = in + in_len - 20;
	const unsigned char *ii;
	lzo_dict_t * const dict = (lzo_dict_t *) wrkmem;

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

	for (;;) {
		dindex = ((size_t)(0x21 * DX3(ip, 5, 5, 6)) >> 5) & D_MASK;
		m_pos = dict[dindex];

		if (m_pos < in)
			goto literal;

		if (ip == m_pos || ((size_t)(ip - m_pos) > M4_MAX_OFFSET))
			goto literal;

		m_off = ip - m_pos;
		if (m_off <= M2_MAX_OFFSET || m_pos[3] == ip[3])
			goto try_match;

		dindex = (dindex & (D_MASK & 0x7ff)) ^ (D_HIGH | 0x1f);
		m_pos = dict[dindex];

		if (m_pos < in)
			goto literal;

		if (ip == m_pos || ((size_t)(ip - m_pos) > M4_MAX_OFFSET))
			goto literal;

		m_off = ip - m_pos;
		if (m_off <= M2_MAX_OFFSET || m_pos[3] == ip[3])
			goto try_match;

		goto literal;

try_match:
		if (get_unaligned((const unsigned short *)m_pos)
				== get_unaligned((const unsigned short *)ip)) {
			if (likely(m_pos[2] == ip[2]))
					goto match;
		}

		const unsigned char *m_pos;
		size_t t, m_len, m_off;
		u32 dv;
literal:
		dict[dindex] = ip;
		++ip;
		ip += 1 + ((ip - ii) >> 5);
next:
		if (unlikely(ip >= ip_end))
			break;
		continue;

match:
		dict[dindex] = ip;
		if (ip != ii) {
			size_t t = ip - ii;
		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;
			} else if (t <= 18) {
				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 (tt > 255) {
					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);
			}
		}

		ip += 3;
		if (m_pos[3] != *ip++ || m_pos[4] != *ip++
				|| m_pos[5] != *ip++ || m_pos[6] != *ip++
				|| m_pos[7] != *ip++ || m_pos[8] != *ip++) {
			--ip;
			m_len = ip - ii;
		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:

			if (m_off <= M2_MAX_OFFSET) {
		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_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));
				goto m3_m4_offset;
			} else {
				m_off -= 0x4000;

				*op++ = (M4_MARKER | ((m_off & 0x4000) >> 11)
						| (m_len - 2));
				goto m3_m4_offset;
			}
		} else {
			end = in_end;
			m = m_pos + M2_MAX_LEN + 1;

			while (ip < end && *m == *ip) {
				m++;
				ip++;
			}
			m_len = ip - ii;

			if (m_off <= M3_MAX_OFFSET) {
				m_off -= 1;
				if (m_len <= 33) {
					*op++ = (M3_MARKER | (m_len - 2));
				} else {
					m_len -= 33;
			else {
				m_len -= M3_MAX_LEN;
				*op++ = M3_MARKER | 0;
					goto m3_m4_len;
				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 & 0x4000) >> 11)
			if (m_len <= M4_MAX_LEN)
				*op++ = (M4_MARKER | ((m_off >> 11) & 8)
						| (m_len - 2));
				} else {
			else {
				m_len -= M4_MAX_LEN;
					*op++ = (M4_MARKER
						| ((m_off & 0x4000) >> 11));
m3_m4_len:
					while (m_len > 255) {
				*op++ = (M4_MARKER | ((m_off >> 11) & 8));
				while (unlikely(m_len > 255)) {
					m_len -= 255;
					*op++ = 0;
				}

				*op++ = (m_len);
			}
			}
m3_m4_offset:
			*op++ = ((m_off & 63) << 2);
			*op++ = (m_off << 2);
			*op++ = (m_off >> 6);
		}

		ii = ip;
		if (unlikely(ip >= ip_end))
			break;
		goto next;
	}

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

	if (unlikely(in_len <= M2_MAX_LEN + 5)) {
		t = in_len;
	} else {
		t = _lzo1x_1_do_compress(in, in_len, op, out_len, wrkmem);
	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) {
		ii = in + in_len - t;
		const unsigned char *ii = in + in_len - t;

		if (op == out && t <= 238) {
			*op++ = (17 + t);
@@ -198,16 +247,21 @@ int lzo1x_1_compress(const unsigned char *in, size_t in_len, unsigned char *out,
			*op++ = (t - 3);
		} else {
			size_t tt = t - 18;

			*op++ = 0;
			while (tt > 255) {
				tt -= 255;
				*op++ = 0;
			}

			*op++ = tt;
		}
		do {
		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);
	}
@@ -223,4 +277,3 @@ EXPORT_SYMBOL_GPL(lzo1x_1_compress);

MODULE_LICENSE("GPL");
MODULE_DESCRIPTION("LZO1X-1 Compressor");
Loading