[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20070518101449.GH10873@zakalwe.fi>
Date: Fri, 18 May 2007 13:14:49 +0300
From: Heikki Orsila <shdl@...alwe.fi>
To: Nitin Gupta <nitingupta910@...il.com>
Cc: linux-kernel@...r.kernel.org,
Richard Purdie <richard@...nedhand.com>
Subject: Re: [RFC] LZO1X de/compression support
Good work..
On Fri, May 18, 2007 at 03:28:31PM +0530, Nitin Gupta wrote:
> Facts for LZO (at least for original code. Should hold true for this
> port also - hence the RFC!):
> - The compressor can never overrun buffer.
> - The "non-safe" version of decompressor can never overrun buffer if
> compressed data is unmodified. I am not sure about this if compressed
> data is malicious (to be confirmed from the author).
> - The "safe" version can never crash (buffer overrun etc.) - confirmed
> from the author.
What's the proof?
> +/* LZO1X_1 compression */
> +int
> +lzo1x_compress(const unsigned char *src, size_t src_len,
> + unsigned char *dst, size_t *dst_len,
> + void *workmem);
int lzo1x_compress(const unsigned char *src, size_t src_len,
unsigned char *dst, size_t *dst_len,
void *workmem);
is the preferred style.
> + Copyright (C) 1997 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 1996 Markus Franz Xaver Johannes Oberhumer
> + All Rights Reserved.
> +
> + The LZO library is free software; you can redistribute it and/or
> + modify it under the terms of the GNU General Public License,
> + version 2, as published by the Free Software Foundation.
> +
> + The LZO library is distributed in the hope that it will be useful,
> + but WITHOUT ANY WARRANTY; without even the implied warranty of
> + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + GNU General Public License for more details.
> +
> + You should have received a copy of the GNU General Public License
> + along with the LZO library; see the file COPYING.
> + If not, write to the Free Software Foundation, Inc.,
> + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
> +
> + Markus F.X.J. Oberhumer
> + <markus@...rhumer.com>
> + http://www.oberhumer.com/opensource/lzo/
> +
> +
> + This file is derived from lzo1x_1.c and lzo1x_c.ch found in original
> + LZO 2.02 code. Some additional changes have also been made to make
> + it work in kernel space.
> +
> + Nitin Gupta
> + <nitingupta910@...il.com>
> + */
> +
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/compiler.h>
> +
> +#include "lzo1x_int.h"
> +
> +MODULE_LICENSE("GPL");
> +MODULE_DESCRIPTION("LZO1X Compression");
> +
> +/* compress a block of data. */
> +static unsigned int
> +lzo1x_compress_worker(const unsigned char *in, size_t in_len,
> + unsigned char *out, size_t *out_len,
> + void *workmem)
> +{
> + register const unsigned char *ip;
Is the register directive really useful? Or any subsequent usage of that
directive?
> + 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 *ii;
> + const unsigned char ** const dict = (const unsigned char **)workmem;
> +
> + op = out;
> + ip = in;
> + ii = ip;
> +
> + ip += 4;
> + for (;;) {
> + register const unsigned char *m_pos;
> + size_t m_off;
> + size_t m_len;
> + size_t dindex;
> +
> + DINDEX1(dindex,ip);
Put a space after the delimiter: DINDEX1(dindex, ip); This happens in
many places in the source, fix them all.
> + m_pos = dict[dindex];
> +
> + if (LZO_CHECK_MPOS_NON_DET(m_pos,m_off,in,ip,M4_MAX_OFFSET))
> + goto literal;
> + if (m_off <= M2_MAX_OFFSET || m_pos[3] == ip[3])
> + goto try_match;
> +
> + DINDEX2(dindex,ip);
> + m_pos = dict[dindex];
> +
> + if (LZO_CHECK_MPOS_NON_DET(m_pos,m_off,in,ip,M4_MAX_OFFSET))
> + goto literal;
> + if (m_off <= M2_MAX_OFFSET || m_pos[3] == ip[3])
> + goto try_match;
> +
> + goto literal;
> +
> +try_match:
> + if (*(const unsigned short *)m_pos != *(const unsigned short *)ip)
> {
> + } else {
> + if (likely(m_pos[2] == ip[2]))
> + goto match;
> + }
> +
> + /* a literal */
> +literal:
> + dict[dindex] = ip;
> + ++ip;
> + if (unlikely(ip >= ip_end))
> + break;
> + continue;
> +
> +
> + /* a match */
> +match:
> + dict[dindex] = ip;
> + /* store current literal run */
> + if (pd(ip,ii) > 0) {
> + register size_t t = pd(ip,ii);
> + if (t <= 3)
> + op[-2] |= (unsigned char)(t);
> + else if (t <= 18)
> + *op++ = (unsigned char)(t - 3);
> + else {
> + register size_t tt = t - 18;
> + *op++ = 0;
> + while (tt > 255) {
> + tt -= 255;
> + *op++ = 0;
> + }
> + *op++ = (unsigned char)(tt);
Useless brackets: (unsigned char) tt
> + }
> + do *op++ = *ii++; while (--t > 0);
memcpy(op, ii, t); ? Happens in other places as well.
> + }
> +
> + /* code the match */
> + 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 = pd(ip, ii);
> +
> + if (m_off <= M2_MAX_OFFSET) {
> + m_off -= 1;
> + *op++ = (unsigned char)(((m_len - 1) << 5) | ((m_off
> & 7) << 2));
> + *op++ = (unsigned char)(m_off >> 3);
> + }
> + else if (m_off <= M3_MAX_OFFSET) {
> + m_off -= 1;
> + *op++ = (unsigned char)(M3_MARKER | (m_len - 2));
> + goto m3_m4_offset;
> + } else {
> + m_off -= 0x4000;
> + *op++ = (unsigned char)(M4_MARKER |
> + ((m_off & 0x4000) >> 11) | (m_len - 2));
> + goto m3_m4_offset;
> + }
> + } else {
> + const unsigned char *end = in_end;
> + const unsigned char *m = m_pos + M2_MAX_LEN + 1;
> + while (ip < end && *m == *ip)
> + m++, ip++;
> + m_len = pd(ip, ii);
> +
> + if (m_off <= M3_MAX_OFFSET) {
> + m_off -= 1;
> + if (m_len <= 33)
> + *op++ = (unsigned char)(M3_MARKER | (m_len -
> 2));
> + else {
> + m_len -= 33;
> + *op++ = M3_MARKER | 0;
> + goto m3_m4_len;
> + }
> + } else {
> + m_off -= 0x4000;
> + if (m_len <= M4_MAX_LEN)
> + *op++ = (unsigned char)(M4_MARKER |
> + ((m_off & 0x4000) >> 11) | (m_len -
> 2));
> + else {
> + m_len -= M4_MAX_LEN;
> + *op++ = (unsigned char)(M4_MARKER | ((m_off
> & 0x4000) >> 11));
> +m3_m4_len:
> + while (m_len > 255) {
> + m_len -= 255;
> + *op++ = 0;
> + }
> + *op++ = (unsigned char)(m_len);
> + }
> + }
> +
> +m3_m4_offset:
> + *op++ = (unsigned char)((m_off & 63) << 2);
> + *op++ = (unsigned char)(m_off >> 6);
> + }
> +
> + ii = ip;
> + if (unlikely(ip >= ip_end))
> + break;
> + }
> +
> + *out_len = pd(op, out);
> + return pd(in_end,ii);
> +}
> +
> +
> +/*
> + * This requires buffer (workmem) of size LZO1X_WORKMEM_SIZE
> + * (exported by lzo1x.h).
> + */
> +int
> +lzo1x_compress(const unsigned char *in, size_t in_len,
> + unsigned char *out, size_t *out_len,
> + void *workmem)
> +{
> + unsigned char *op = out;
> + size_t t;
> +
> + if (!workmem)
> + return -EINVAL;
> +
> + if (unlikely(in_len <= M2_MAX_LEN + 5))
> + t = in_len;
> + else {
> + t = lzo1x_compress_worker(in,in_len,op,out_len,workmem);
> + op += *out_len;
> + }
> +
> + if (t > 0) {
> + const unsigned char *ii = in + in_len - t;
> +
> + if (op == out && t <= 238)
> + *op++ = (unsigned char)(17 + t);
> + else if (t <= 3)
> + op[-2] |= (unsigned char)(t);
> + else if (t <= 18)
> + *op++ = (unsigned char)(t - 3);
> + else {
> + size_t tt = t - 18;
> + *op++ = 0;
> + while (tt > 255) {
> + tt -= 255;
> + *op++ = 0;
> + }
> + *op++ = (unsigned char)(tt);
> + }
> + do {
> + *op++ = *ii++;
> + } while (--t > 0);
> + }
> + *op++ = M4_MARKER | 1;
> + *op++ = 0;
> + *op++ = 0;
> +
> + *out_len = pd(op, out);
> + return 0;
> +}
> +
> +EXPORT_SYMBOL(lzo1x_compress);
> diff --git a/lib/lzo1x/lzo1x_decompress.c b/lib/lzo1x/lzo1x_decompress.c
> new file mode 100755
> index 0000000..fbd4d69
> --- /dev/null
> +++ b/lib/lzo1x/lzo1x_decompress.c
> @@ -0,0 +1,216 @@
> +/* lzo1x_decompress.c -- LZO1X decompression
> +
> + This file is part of the LZO real-time data compression library.
> +
> + Copyright (C) 2005 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 2004 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 2003 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 2002 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 2001 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 2000 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 1999 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 1998 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 1997 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 1996 Markus Franz Xaver Johannes Oberhumer
> + All Rights Reserved.
> +
> + The LZO library is free software; you can redistribute it and/or
> + modify it under the terms of the GNU General Public License,
> + version 2, as published by the Free Software Foundation.
> +
> + The LZO library is distributed in the hope that it will be useful,
> + but WITHOUT ANY WARRANTY; without even the implied warranty of
> + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + GNU General Public License for more details.
> +
> + You should have received a copy of the GNU General Public License
> + along with the LZO library; see the file COPYING.
> + If not, write to the Free Software Foundation, Inc.,
> + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
> +
> + Markus F.X.J. Oberhumer
> + <markus@...rhumer.com>
> + http://www.oberhumer.com/opensource/lzo/
> +
> +
> + This file is derived from lzo1x_d1.c and lzo1x_d.ch found in original
> + LZO 2.02 code. Some additional changes have also been made to make
> + it work in kernel space.
> +
> + Nitin Gupta
> + <nitingupta910@...il.com>
> + */
> +
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <asm/byteorder.h>
> +
> +#include "lzo1x_int.h"
> +
> +MODULE_LICENSE("GPL");
> +MODULE_DESCRIPTION("LZO1X Decompression");
> +
> +int
> +lzo1x_decompress(const unsigned char *in, size_t in_len,
> + unsigned char *out, size_t *out_len)
> +{
> + register size_t t;
> + register unsigned char *op;
> + register const unsigned char *ip, *m_pos;
> + const unsigned char * const ip_end = in + in_len;
> +
> + *out_len = 0;
> +
> + op = out;
> + ip = in;
> +
> + if (*ip > 17) {
> + t = *ip++ - 17;
> + if (t < 4)
> + goto match_next;
> + do
> + *op++ = *ip++;
> + while (--t > 0);
> + goto first_literal_run;
> + }
> +
> + while (1) {
> + t = *ip++;
> + if (t >= 16)
> + goto match;
> + /* a literal run */
> + if (t == 0) {
> + while (*ip == 0) {
> + t += 255;
> + ip++;
> + }
> + t += 15 + *ip++;
> + }
> + /* copy literals */
> + COPY4(op,ip);
> + op += 4;
> + ip += 4;
> + if (--t > 0) {
> + if (t >= 4) {
> + do {
> + COPY4(op,ip);
> + op += 4; ip += 4; t -= 4;
> + } while (t >= 4);
> + if (t > 0)
> + do
> + *op++ = *ip++;
> + while (--t > 0);
> + } else
> + do
> + *op++ = *ip++;
> + while (--t > 0);
> + }
> +
> +first_literal_run:
> +
> + t = *ip++;
> + if (t >= 16)
> + goto match;
> + m_pos = op - (1 + M2_MAX_OFFSET);
> + m_pos -= t >> 2;
> + m_pos -= *ip++ << 2;
> + *op++ = *m_pos++; *op++ = *m_pos++; *op++ = *m_pos;
> + goto match_done;
> +
> + /* handle matches */
> + do {
> + match:
> + if (t >= 64) { /* a M2 match */
> + m_pos = op - 1;
> + m_pos -= (t >> 2) & 7;
> + m_pos -= *ip++ << 3;
> + t = (t >> 5) - 1;
> + goto copy_match;
> + } else if (t >= 32) { /* a M3 match */
> + t &= 31;
> + if (t == 0) {
> + while (*ip == 0) {
> + t += 255;
> + ip++;
> + }
> + t += 31 + *ip++;
> + }
> +#if defined(__LITTLE_ENDIAN)
> + m_pos = op - 1;
> + m_pos -= (*(const unsigned short *)ip) >> 2;
> +#else
> + m_pos = op - 1;
> + m_pos -= (ip[0] >> 2) + (ip[1] << 6);
> +#endif
> + ip += 2;
> + } else if (t >= 16) { /* a M4 match */
> + m_pos = op;
> + m_pos -= (t & 8) << 11;
> + t &= 7;
> + if (t == 0) {
> + while (*ip == 0) {
> + t += 255;
> + ip++;
> + }
> + t += 7 + *ip++;
> + }
> +#if defined(__LITTLE_ENDIAN)
> + m_pos -= (*(const unsigned short *)ip) >> 2;
> +#else
> + m_pos -= (ip[0] >> 2) + (ip[1] << 6);
> +#endif
> + ip += 2;
> + if (m_pos == op)
> + goto eof_found;
> + m_pos -= 0x4000;
> + } else { /* a M1 match */
> + m_pos = op - 1;
> + m_pos -= t >> 2;
> + m_pos -= *ip++ << 2;
> + *op++ = *m_pos++; *op++ = *m_pos;
> + goto match_done;
> + }
> +
> + /* copy match */
> + if (t >= 2 * 4 - (3 - 1) && (op - m_pos) >= 4) {
> + COPY4(op,m_pos);
> + op += 4; m_pos += 4; t -= 4 - (3 - 1);
> + do {
> + COPY4(op,m_pos);
> + op += 4; m_pos += 4; t -= 4;
> + } while (t >= 4);
> + if (t > 0)
> + do
> + *op++ = *m_pos++;
> + while (--t > 0);
> + } else {
> +copy_match:
> + *op++ = *m_pos++; *op++ = *m_pos++;
> + do
> + *op++ = *m_pos++;
> + while (--t > 0);
> + }
> +
> +match_done:
> + t = ip[-2] & 3;
> + if (t == 0)
> + break;
> +
> + /* copy literals */
> +match_next:
> + *op++ = *ip++;
> + if (t > 1) {
> + *op++ = *ip++;
> + if (t > 2)
> + *op++ = *ip++;
> + }
> + t = *ip++;
> + } while (1);
> + }
> +
> +eof_found:
> + *out_len = pd(op, out);
> + return (ip == ip_end ? 0 : -1);
> +}
> +
> +EXPORT_SYMBOL(lzo1x_decompress);
> diff --git a/lib/lzo1x/lzo1x_int.h b/lib/lzo1x/lzo1x_int.h
> new file mode 100755
> index 0000000..4dd993d
> --- /dev/null
> +++ b/lib/lzo1x/lzo1x_int.h
> @@ -0,0 +1,105 @@
> +/* lzo1x_int.h -- to be used internally by LZO de/compression algorithms
> +
> + This file is part of the LZO real-time data compression library.
> +
> + Copyright (C) 2005 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 2004 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 2003 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 2002 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 2001 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 2000 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 1999 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 1998 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 1997 Markus Franz Xaver Johannes Oberhumer
> + Copyright (C) 1996 Markus Franz Xaver Johannes Oberhumer
> + All Rights Reserved.
> +
> + The LZO library is free software; you can redistribute it and/or
> + modify it under the terms of the GNU General Public License,
> + version 2, as published by the Free Software Foundation.
> +
> + The LZO library is distributed in the hope that it will be useful,
> + but WITHOUT ANY WARRANTY; without even the implied warranty of
> + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + GNU General Public License for more details.
> +
> + You should have received a copy of the GNU General Public License
> + along with the LZO library; see the file COPYING.
> + If not, write to the Free Software Foundation, Inc.,
> + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
> +
> + Markus F.X.J. Oberhumer
> + <markus@...rhumer.com>
> + http://www.oberhumer.com/opensource/lzo/
> +
> +
> + This file was derived from several header files found in original
> + LZO 2.02 code. Some additional changes have also been made to make
> + it work in kernel space.
> +
> + Nitin Gupta
> + <nitingupta910@...il.com>
> + */
> +
> +#ifndef __LZO1X_INT_H
> +#define __LZO1X_INT_H
> +
> +#include <linux/types.h>
> +
> +#define D_SIZE (1u << D_BITS)
> +#define D_MASK (D_SIZE - 1)
> +#define D_HIGH ((D_MASK >> 1) + 1)
> +
> +#define DX2(p,s1,s2) \
> + (((((size_t)((p)[2]) << (s2)) ^ (p)[1]) << (s1)) ^ (p)[0])
> +#define DX3(p,s1,s2,s3) ((DX2((p)+1,s2,s3) << (s1)) ^ (p)[0])
> +#define DMUL(a,b) ((size_t) ((a) * (b)))
> +#define DMS(v,s) ((size_t) (((v) & (D_MASK >> (s))) << (s)))
> +#define DM(v) DMS(v,0)
> +
> +#define D_BITS 14
> +#define DINDEX1(d,p) d = DM(DMUL(0x21,DX3(p,5,5,6)) >> 5)
> +#define DINDEX2(d,p) d = (d & (D_MASK & 0x7ff)) ^ (D_HIGH | 0x1f)
> +#define DENTRY(p,in) (p)
> +
> +#define PTR(a) ((unsigned long) (a))
> +#define PTR_LT(a,b) (PTR(a) < PTR(b))
> +#define PTR_GE(a,b) (PTR(a) >= PTR(b))
> +#define PTR_DIFF(a,b) (PTR(a) - PTR(b))
> +#define pd(a,b) ((size_t) ((a)-(b)))
> +
> +#define LZO_CHECK_MPOS_NON_DET(m_pos,m_off,in,ip,max_offset) \
> + ( m_pos = ip - (size_t) PTR_DIFF(ip,m_pos), \
> + PTR_LT(m_pos,in) || \
> + (m_off = (size_t) PTR_DIFF(ip,m_pos)) <= 0 || \
> + m_off > max_offset )
> +
> +#define COPY4(dst,src) *(uint32_t *)(dst) = *(uint32_t *)(src)
Use u32.
--
Heikki Orsila Barbie's law:
heikki.orsila@....fi "Math is hard, let's go shopping!"
http://www.iki.fi/shd
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists