Commit ba405f86 authored by Nick Mathewson's avatar Nick Mathewson 🦀
Browse files

Merge branch 'ahf_prop278_21662_squashed'

parents db95a6e1 2210b330
Loading
Loading
Loading
Loading
+1 −1
Original line number Diff line number Diff line
@@ -16,7 +16,7 @@ noinst_PROGRAMS=
DISTCLEANFILES=
bin_SCRIPTS=
AM_CPPFLAGS=
AM_CFLAGS=@TOR_SYSTEMD_CFLAGS@ @CFLAGS_BUGTRAP@
AM_CFLAGS=@TOR_SYSTEMD_CFLAGS@ @CFLAGS_BUGTRAP@ @TOR_LZMA_CFLAGS@ @TOR_ZSTD_CFLAGS@
SHELL=@SHELL@

if COVERAGE_ENABLED
+64 −0
Original line number Diff line number Diff line
@@ -732,6 +732,70 @@ else
fi
AC_SUBST(TOR_ZLIB_LIBS)

dnl ------------------------------------------------------
dnl Where we do we find lzma?

AC_ARG_ENABLE(lzma,
      AS_HELP_STRING(--enable-lzma, [enable support for the Zstandard compression scheme.]),
      [case "${enableval}" in
        "yes") lzma=true ;;
        "no")  lzma=false ;;
        * ) AC_MSG_ERROR(bad value for --enable-lzma) ;;
      esac], [lzma=auto])

if test "x$enable_lzma" = "xno"; then
    have_lzma=no;
else
    PKG_CHECK_MODULES([LZMA],
                      [liblzma],
                      have_lzma=yes,
                      have_lzma=no)

    if test "x$have_lzma" = "xno" ; then
        AC_MSG_WARN([Unable to find liblzma.])
    fi
fi

if test "x$have_lzma" = "xyes"; then
    AC_DEFINE(HAVE_LZMA,1,[Have LZMA])
    TOR_LZMA_CFLAGS="${LZMA_CFLAGS}"
    TOR_LZMA_LIBS="${LZMA_LIBS}"
fi
AC_SUBST(TOR_LZMA_CFLAGS)
AC_SUBST(TOR_LZMA_LIBS)

dnl ------------------------------------------------------
dnl Where we do we find zstd?

AC_ARG_ENABLE(zstd,
      AS_HELP_STRING(--enable-zstd, [enable support for the Zstandard compression scheme.]),
      [case "${enableval}" in
        "yes") zstd=true ;;
        "no")  zstd=false ;;
        * ) AC_MSG_ERROR(bad value for --enable-zstd) ;;
      esac], [zstd=auto])

if test "x$enable_zstd" = "xno"; then
    have_zstd=no;
else
    PKG_CHECK_MODULES([ZSTD],
                      [libzstd],
                      have_zstd=yes,
                      have_zstd=no)

    if test "x$have_zstd" = "xno" ; then
        AC_MSG_WARN([Unable to find libzstd.])
    fi
fi

if test "x$have_zstd" = "xyes"; then
    AC_DEFINE(HAVE_ZSTD,1,[Have Zstd])
    TOR_ZSTD_CFLAGS="${ZSTD_CFLAGS}"
    TOR_ZSTD_LIBS="${ZSTD_LIBS}"
fi
AC_SUBST(TOR_ZSTD_CFLAGS)
AC_SUBST(TOR_ZSTD_LIBS)

dnl ----------------------------------------------------------------------
dnl Check if libcap is available for capabilities.

+2 −2
Original line number Diff line number Diff line
@@ -7,8 +7,8 @@ LIBOR_OBJECTS = address.obj backtrace.obj compat.obj container.obj di_ops.obj \
	log.obj memarea.obj mempool.obj procmon.obj sandbox.obj util.obj \
	util_codedigest.obj

LIBOR_CRYPTO_OBJECTS = aes.obj crypto.obj crypto_format.obj torgzip.obj tortls.obj \
	crypto_curve25519.obj curve25519-donna.obj
LIBOR_CRYPTO_OBJECTS = aes.obj crypto.obj crypto_format.obj compress.obj compress_zlib.obj \
	tortls.obj crypto_curve25519.obj curve25519-donna.obj

LIBOR_EVENT_OBJECTS = compat_libevent.obj

src/common/compress.c

0 → 100644
+377 −0
Original line number Diff line number Diff line
/* Copyright (c) 2004, Roger Dingledine.
 * Copyright (c) 2004-2006, Roger Dingledine, Nick Mathewson.
 * Copyright (c) 2007-2017, The Tor Project, Inc. */
/* See LICENSE for licensing information */

/**
 * \file compress.c
 * \brief Common compression API.
 **/

#include "orconfig.h"

#include <stdlib.h>
#include <stdio.h>
#include <assert.h>
#include <string.h>
#include "torint.h"

#ifdef HAVE_NETINET_IN_H
#include <netinet/in.h>
#endif

#include "util.h"
#include "torlog.h"
#include "compress.h"
#include "compress_lzma.h"
#include "compress_zlib.h"
#include "compress_zstd.h"

/** @{ */
/* These macros define the maximum allowable compression factor.  Anything of
 * size greater than CHECK_FOR_COMPRESSION_BOMB_AFTER is not allowed to
 * have an uncompression factor (uncompressed size:compressed size ratio) of
 * any greater than MAX_UNCOMPRESSION_FACTOR.
 *
 * Picking a value for MAX_UNCOMPRESSION_FACTOR is a trade-off: we want it to
 * be small to limit the attack multiplier, but we also want it to be large
 * enough so that no legitimate document --even ones we might invent in the
 * future -- ever compresses by a factor of greater than
 * MAX_UNCOMPRESSION_FACTOR. Within those parameters, there's a reasonably
 * large range of possible values. IMO, anything over 8 is probably safe; IMO
 * anything under 50 is probably sufficient.
 */
#define MAX_UNCOMPRESSION_FACTOR 25
#define CHECK_FOR_COMPRESSION_BOMB_AFTER (1024*64)
/** @} */

/** Return true if uncompressing an input of size <b>in_size</b> to an input of
 * size at least <b>size_out</b> looks like a compression bomb. */
int
tor_compress_is_compression_bomb(size_t size_in, size_t size_out)
{
  if (size_in == 0 || size_out < CHECK_FOR_COMPRESSION_BOMB_AFTER)
    return 0;

  return (size_out / size_in > MAX_UNCOMPRESSION_FACTOR);
}

/** Given <b>in_len</b> bytes at <b>in</b>, compress them into a newly
 * allocated buffer, using the method described in <b>method</b>.  Store the
 * compressed string in *<b>out</b>, and its length in *<b>out_len</b>.
 * Return 0 on success, -1 on failure.
 */
int
tor_compress(char **out, size_t *out_len,
             const char *in, size_t in_len,
             compress_method_t method)
{
  switch (method) {
    case GZIP_METHOD:
    case ZLIB_METHOD:
      return tor_zlib_compress(out, out_len, in, in_len, method);
    case LZMA_METHOD:
      return tor_lzma_compress(out, out_len, in, in_len, method);
    case ZSTD_METHOD:
      return tor_zstd_compress(out, out_len, in, in_len, method);
    case NO_METHOD:
    case UNKNOWN_METHOD:
    default:
      return -1;
  }
}

/** Given zero or more zlib-compressed or gzip-compressed strings of
 * total length
 * <b>in_len</b> bytes at <b>in</b>, uncompress them into a newly allocated
 * buffer, using the method described in <b>method</b>.  Store the uncompressed
 * string in *<b>out</b>, and its length in *<b>out_len</b>.  Return 0 on
 * success, -1 on failure.
 *
 * If <b>complete_only</b> is true, we consider a truncated input as a
 * failure; otherwise we decompress as much as we can.  Warn about truncated
 * or corrupt inputs at <b>protocol_warn_level</b>.
 */
int
tor_uncompress(char **out, size_t *out_len,
               const char *in, size_t in_len,
               compress_method_t method,
               int complete_only,
               int protocol_warn_level)
{
  switch (method) {
    case GZIP_METHOD:
    case ZLIB_METHOD:
      return tor_zlib_uncompress(out, out_len, in, in_len,
                                 method,
                                 complete_only,
                                 protocol_warn_level);
    case LZMA_METHOD:
      return tor_lzma_uncompress(out, out_len, in, in_len,
                                 method,
                                 complete_only,
                                 protocol_warn_level);
    case ZSTD_METHOD:
      return tor_zstd_uncompress(out, out_len, in, in_len,
                                 method,
                                 complete_only,
                                 protocol_warn_level);
    case NO_METHOD:
    case UNKNOWN_METHOD:
    default:
      return -1;
  }
}

/** Try to tell whether the <b>in_len</b>-byte string in <b>in</b> is likely
 * to be compressed or not.  If it is, return the likeliest compression method.
 * Otherwise, return UNKNOWN_METHOD.
 */
compress_method_t
detect_compression_method(const char *in, size_t in_len)
{
  if (in_len > 2 && fast_memeq(in, "\x1f\x8b", 2)) {
    return GZIP_METHOD;
  } else if (in_len > 2 && (in[0] & 0x0f) == 8 &&
             (ntohs(get_uint16(in)) % 31) == 0) {
    return ZLIB_METHOD;
  } else if (in_len > 3 &&
             fast_memeq(in, "\x5d\x00\x00\x00", 4)) {
    return LZMA_METHOD;
  } else if (in_len > 3 &&
             fast_memeq(in, "\x28\xb5\x2f\xfd", 4)) {
    return ZSTD_METHOD;
  } else {
    return UNKNOWN_METHOD;
  }
}

/** Return 1 if a given <b>method</b> is supported; otherwise 0. */
int
tor_compress_supports_method(compress_method_t method)
{
  switch (method) {
    case GZIP_METHOD:
    case ZLIB_METHOD:
      return tor_zlib_method_supported();
    case LZMA_METHOD:
      return tor_lzma_method_supported();
    case ZSTD_METHOD:
      return tor_zstd_method_supported();
    case NO_METHOD:
    case UNKNOWN_METHOD:
    default:
      return 0;
  }
}

/** Return a string representation of the version of the library providing the
 * compression method given in <b>method</b>. Returns NULL if <b>method</b> is
 * unknown or unsupported. */
const char *
tor_compress_version_str(compress_method_t method)
{
  switch (method) {
    case GZIP_METHOD:
    case ZLIB_METHOD:
      return tor_zlib_get_version_str();
    case LZMA_METHOD:
      return tor_lzma_get_version_str();
    case ZSTD_METHOD:
      return tor_zstd_get_version_str();
    case NO_METHOD:
    case UNKNOWN_METHOD:
    default:
      return NULL;
  }
}

/** Return a string representation of the version of the library, found at
 * compile time, providing the compression method given in <b>method</b>.
 * Returns NULL if <b>method</b> is unknown or unsupported. */
const char *
tor_compress_header_version_str(compress_method_t method)
{
  switch (method) {
    case GZIP_METHOD:
    case ZLIB_METHOD:
      return tor_zlib_get_header_version_str();
    case LZMA_METHOD:
      return tor_lzma_get_header_version_str();
    case ZSTD_METHOD:
      return tor_zstd_get_header_version_str();
    case NO_METHOD:
    case UNKNOWN_METHOD:
    default:
      return NULL;
  }
}

/** Return the approximate number of bytes allocated for all
 * supported compression schemas. */
size_t
tor_compress_get_total_allocation(void)
{
  return tor_zlib_get_total_allocation() +
         tor_lzma_get_total_allocation() +
         tor_zstd_get_total_allocation();
}

/** Internal state for an incremental compression/decompression.  The body of
 * this struct is not exposed. */
struct tor_compress_state_t {
  compress_method_t method; /**< The compression method. */

  union {
    tor_zlib_compress_state_t *zlib_state;
    tor_lzma_compress_state_t *lzma_state;
    tor_zstd_compress_state_t *zstd_state;
  } u; /**< Compression backend state. */
};

/** Construct and return a tor_compress_state_t object using <b>method</b>.  If
 * <b>compress</b>, it's for compression; otherwise it's for decompression. */
tor_compress_state_t *
tor_compress_new(int compress, compress_method_t method,
                 compression_level_t compression_level)
{
  tor_compress_state_t *state;

  state = tor_malloc_zero(sizeof(tor_compress_state_t));
  state->method = method;

  switch (method) {
    case GZIP_METHOD:
    case ZLIB_METHOD: {
      tor_zlib_compress_state_t *zlib_state =
        tor_zlib_compress_new(compress, method, compression_level);

      if (zlib_state == NULL)
        goto err;

      state->u.zlib_state = zlib_state;
      break;
    }
    case LZMA_METHOD: {
      tor_lzma_compress_state_t *lzma_state =
        tor_lzma_compress_new(compress, method, compression_level);

      if (lzma_state == NULL)
        goto err;

      state->u.lzma_state = lzma_state;
      break;
    }
    case ZSTD_METHOD: {
      tor_zstd_compress_state_t *zstd_state =
        tor_zstd_compress_new(compress, method, compression_level);

      if (zstd_state == NULL)
        goto err;

      state->u.zstd_state = zstd_state;
      break;
    }
    case NO_METHOD:
    case UNKNOWN_METHOD:
      goto err;
  }

  return state;

 err:
  tor_free(state);
  return NULL;
}

/** Compress/decompress some bytes using <b>state</b>.  Read up to
 * *<b>in_len</b> bytes from *<b>in</b>, and write up to *<b>out_len</b> bytes
 * to *<b>out</b>, adjusting the values as we go.  If <b>finish</b> is true,
 * we've reached the end of the input.
 *
 * Return TOR_COMPRESS_DONE if we've finished the entire
 * compression/decompression.
 * Return TOR_COMPRESS_OK if we're processed everything from the input.
 * Return TOR_COMPRESS_BUFFER_FULL if we're out of space on <b>out</b>.
 * Return TOR_COMPRESS_ERROR if the stream is corrupt.
 */
tor_compress_output_t
tor_compress_process(tor_compress_state_t *state,
                     char **out, size_t *out_len,
                     const char **in, size_t *in_len,
                     int finish)
{
  tor_assert(state != NULL);

  switch (state->method) {
    case GZIP_METHOD:
    case ZLIB_METHOD:
      return tor_zlib_compress_process(state->u.zlib_state,
                                       out, out_len, in, in_len,
                                       finish);
    case LZMA_METHOD:
      return tor_lzma_compress_process(state->u.lzma_state,
                                       out, out_len, in, in_len,
                                       finish);
    case ZSTD_METHOD:
      return tor_zstd_compress_process(state->u.zstd_state,
                                       out, out_len, in, in_len,
                                       finish);
    case NO_METHOD:
    case UNKNOWN_METHOD:
      goto err;
  }

 err:
  return TOR_COMPRESS_ERROR;
}

/** Deallocate <b>state</b>. */
void
tor_compress_free(tor_compress_state_t *state)
{
  if (state == NULL)
    return;

  switch (state->method) {
    case GZIP_METHOD:
    case ZLIB_METHOD:
      tor_zlib_compress_free(state->u.zlib_state);
      break;
    case LZMA_METHOD:
      tor_lzma_compress_free(state->u.lzma_state);
      break;
    case ZSTD_METHOD:
      tor_zstd_compress_free(state->u.zstd_state);
      break;
    case NO_METHOD:
    case UNKNOWN_METHOD:
      break;
  }

  tor_free(state);
}

/** Return the approximate number of bytes allocated for <b>state</b>. */
size_t
tor_compress_state_size(const tor_compress_state_t *state)
{
  tor_assert(state != NULL);

  switch (state->method) {
    case GZIP_METHOD:
    case ZLIB_METHOD:
      return tor_zlib_compress_state_size(state->u.zlib_state);
    case LZMA_METHOD:
      return tor_lzma_compress_state_size(state->u.lzma_state);
    case ZSTD_METHOD:
      return tor_zstd_compress_state_size(state->u.zstd_state);
    case NO_METHOD:
    case UNKNOWN_METHOD:
      goto err;
  }

 err:
  return 0;
}

src/common/compress.h

0 → 100644
+83 −0
Original line number Diff line number Diff line
/* Copyright (c) 2003, Roger Dingledine
 * Copyright (c) 2004-2006, Roger Dingledine, Nick Mathewson.
 * Copyright (c) 2007-2017, The Tor Project, Inc. */
/* See LICENSE for licensing information */

/**
 * \file compress.h
 * \brief Headers for compress.c
 **/

#ifndef TOR_COMPRESS_H
#define TOR_COMPRESS_H

/** Enumeration of what kind of compression to use.  Only ZLIB_METHOD and
 * GZIP_METHOD is guaranteed to be supported by the compress/uncompress
 * functions here. Call tor_compress_supports_method() to check if a given
 * compression schema is supported by Tor. */
typedef enum {
  NO_METHOD=0,
  GZIP_METHOD=1,
  ZLIB_METHOD=2,
  LZMA_METHOD=3,
  ZSTD_METHOD=4,
  UNKNOWN_METHOD=5
} compress_method_t;

/**
 * Enumeration to define tradeoffs between memory usage and compression level.
 * HIGH_COMPRESSION saves the most bandwidth; LOW_COMPRESSION saves the most
 * memory.
 **/
typedef enum {
  HIGH_COMPRESSION, MEDIUM_COMPRESSION, LOW_COMPRESSION
} compression_level_t;

int tor_compress(char **out, size_t *out_len,
                 const char *in, size_t in_len,
                 compress_method_t method);

int tor_uncompress(char **out, size_t *out_len,
                   const char *in, size_t in_len,
                   compress_method_t method,
                   int complete_only,
                   int protocol_warn_level);

compress_method_t detect_compression_method(const char *in, size_t in_len);

int tor_compress_is_compression_bomb(size_t size_in, size_t size_out);

int tor_compress_supports_method(compress_method_t method);

const char *tor_compress_version_str(compress_method_t method);

const char *tor_compress_header_version_str(compress_method_t method);

size_t tor_compress_get_total_allocation(void);

/** Return values from tor_compress_process; see that function's documentation
 * for details. */
typedef enum {
  TOR_COMPRESS_OK,
  TOR_COMPRESS_DONE,
  TOR_COMPRESS_BUFFER_FULL,
  TOR_COMPRESS_ERROR
} tor_compress_output_t;

/** Internal state for an incremental compression/decompression. */
typedef struct tor_compress_state_t tor_compress_state_t;

tor_compress_state_t *tor_compress_new(int compress,
                                       compress_method_t method,
                                       compression_level_t level);

tor_compress_output_t tor_compress_process(tor_compress_state_t *state,
                                           char **out, size_t *out_len,
                                           const char **in, size_t *in_len,
                                           int finish);
void tor_compress_free(tor_compress_state_t *state);

size_t tor_compress_state_size(const tor_compress_state_t *state);

#endif // TOR_COMPRESS_H.
Loading