decompressors: add XZ decompressor module

In userspace, the .lzma format has become mostly a legacy file format that
got superseded by the .xz format.  Similarly, LZMA Utils was superseded by
XZ Utils.

These patches add support for XZ decompression into the kernel.  Most of
the code is as is from XZ Embedded <http://tukaani.org/xz/embedded.html>.
It was written for the Linux kernel but is usable in other projects too.

Advantages of XZ over the current LZMA code in the kernel:
  - Nice API that can be used by other kernel modules; it's
    not limited to kernel, initramfs, and initrd decompression.
  - Integrity check support (CRC32)
  - BCJ filters improve compression of executable code on
    certain architectures. These together with LZMA2 can
    produce a few percent smaller kernel or Squashfs images
    than plain LZMA without making the decompression slower.

This patch: Add the main decompression code (xz_dec), testing module
(xz_dec_test), wrapper script (xz_wrap.sh) for the xz command line tool,
and documentation.  The xz_dec module is enough to have a usable XZ
decompressor e.g.  for Squashfs.

Signed-off-by: Lasse Collin <lasse.collin@tukaani.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Alain Knaff <alain@knaff.lu>
Cc: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Cc: Phillip Lougher <phillip@lougher.demon.co.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

17 files changed, 3783 insertions, 0 deletions

diff --git a/Documentation/xz.txt b/Documentation/xz.txt
new file mode 100644
index 000000000000..2cf3e2608de3
--- /dev/null
+++ b/Documentation/xz.txt
@@ -0,0 +1,121 @@
+
+XZ data compression in Linux
+============================
+
+Introduction
+
+    XZ is a general purpose data compression format with high compression
+    ratio and relatively fast decompression. The primary compression
+    algorithm (filter) is LZMA2. Additional filters can be used to improve
+    compression ratio even further. E.g. Branch/Call/Jump (BCJ) filters
+    improve compression ratio of executable data.
+
+    The XZ decompressor in Linux is called XZ Embedded. It supports
+    the LZMA2 filter and optionally also BCJ filters. CRC32 is supported
+    for integrity checking. The home page of XZ Embedded is at
+    <http://tukaani.org/xz/embedded.html>, where you can find the
+    latest version and also information about using the code outside
+    the Linux kernel.
+
+    For userspace, XZ Utils provide a zlib-like compression library
+    and a gzip-like command line tool. XZ Utils can be downloaded from
+    <http://tukaani.org/xz/>.
+
+XZ related components in the kernel
+
+    The xz_dec module provides XZ decompressor with single-call (buffer
+    to buffer) and multi-call (stateful) APIs. The usage of the xz_dec
+    module is documented in include/linux/xz.h.
+
+    The xz_dec_test module is for testing xz_dec. xz_dec_test is not
+    useful unless you are hacking the XZ decompressor. xz_dec_test
+    allocates a char device major dynamically to which one can write
+    .xz files from userspace. The decompressed output is thrown away.
+    Keep an eye on dmesg to see diagnostics printed by xz_dec_test.
+    See the xz_dec_test source code for the details.
+
+    For decompressing the kernel image, initramfs, and initrd, there
+    is a wrapper function in lib/decompress_unxz.c. Its API is the
+    same as in other decompress_*.c files, which is defined in
+    include/linux/decompress/generic.h.
+
+    scripts/xz_wrap.sh is a wrapper for the xz command line tool found
+    from XZ Utils. The wrapper sets compression options to values suitable
+    for compressing the kernel image.
+
+    For kernel makefiles, two commands are provided for use with
+    $(call if_needed). The kernel image should be compressed with
+    $(call if_needed,xzkern) which will use a BCJ filter and a big LZMA2
+    dictionary. It will also append a four-byte trailer containing the
+    uncompressed size of the file, which is needed by the boot code.
+    Other things should be compressed with $(call if_needed,xzmisc)
+    which will use no BCJ filter and 1 MiB LZMA2 dictionary.
+
+Notes on compression options
+
+    Since the XZ Embedded supports only streams with no integrity check or
+    CRC32, make sure that you don't use some other integrity check type
+    when encoding files that are supposed to be decoded by the kernel. With
+    liblzma, you need to use either LZMA_CHECK_NONE or LZMA_CHECK_CRC32
+    when encoding. With the xz command line tool, use --check=none or
+    --check=crc32.
+
+    Using CRC32 is strongly recommended unless there is some other layer
+    which will verify the integrity of the uncompressed data anyway.
+    Double checking the integrity would probably be waste of CPU cycles.
+    Note that the headers will always have a CRC32 which will be validated
+    by the decoder; you can only change the integrity check type (or
+    disable it) for the actual uncompressed data.
+
+    In userspace, LZMA2 is typically used with dictionary sizes of several
+    megabytes. The decoder needs to have the dictionary in RAM, thus big
+    dictionaries cannot be used for files that are intended to be decoded
+    by the kernel. 1 MiB is probably the maximum reasonable dictionary
+    size for in-kernel use (maybe more is OK for initramfs). The presets
+    in XZ Utils may not be optimal when creating files for the kernel,
+    so don't hesitate to use custom settings. Example:
+
+        xz --check=crc32 --lzma2=dict=512KiB inputfile
+
+    An exception to above dictionary size limitation is when the decoder
+    is used in single-call mode. Decompressing the kernel itself is an
+    example of this situation. In single-call mode, the memory usage
+    doesn't depend on the dictionary size, and it is perfectly fine to
+    use a big dictionary: for maximum compression, the dictionary should
+    be at least as big as the uncompressed data itself.
+
+Future plans
+
+    Creating a limited XZ encoder may be considered if people think it is
+    useful. LZMA2 is slower to compress than e.g. Deflate or LZO even at
+    the fastest settings, so it isn't clear if LZMA2 encoder is wanted
+    into the kernel.
+
+    Support for limited random-access reading is planned for the
+    decompression code. I don't know if it could have any use in the
+    kernel, but I know that it would be useful in some embedded projects
+    outside the Linux kernel.
+
+Conformance to the .xz file format specification
+
+    There are a couple of corner cases where things have been simplified
+    at expense of detecting errors as early as possible. These should not
+    matter in practice all, since they don't cause security issues. But
+    it is good to know this if testing the code e.g. with the test files
+    from XZ Utils.
+
+Reporting bugs
+
+    Before reporting a bug, please check that it's not fixed already
+    at upstream. See <http://tukaani.org/xz/embedded.html> to get the
+    latest code.
+
+    Report bugs to <lasse.collin@tukaani.org> or visit #tukaani on
+    Freenode and talk to Larhzu. I don't actively read LKML or other
+    kernel-related mailing lists, so if there's something I should know,
+    you should email to me personally or use IRC.
+
+    Don't bother Igor Pavlov with questions about the XZ implementation
+    in the kernel or about XZ Utils. While these two implementations
+    include essential code that is directly based on Igor Pavlov's code,
+    these implementations aren't maintained nor supported by him.
diff --git a/include/linux/xz.h b/include/linux/xz.h
new file mode 100644
index 000000000000..64cffa6ddfce
--- /dev/null
+++ b/include/linux/xz.h
@@ -0,0 +1,264 @@
+/*
+ * XZ decompressor
+ *
+ * Authors: Lasse Collin <lasse.collin@tukaani.org>
+ *          Igor Pavlov <http://7-zip.org/>
+ *
+ * This file has been put into the public domain.
+ * You can do whatever you want with this file.
+ */
+
+#ifndef XZ_H
+#define XZ_H
+
+#ifdef __KERNEL__
+#	include <linux/stddef.h>
+#	include <linux/types.h>
+#else
+#	include <stddef.h>
+#	include <stdint.h>
+#endif
+
+/* In Linux, this is used to make extern functions static when needed. */
+#ifndef XZ_EXTERN
+#	define XZ_EXTERN extern
+#endif
+
+/**
+ * enum xz_mode - Operation mode
+ *
+ * @XZ_SINGLE:              Single-call mode. This uses less RAM than
+ *                          than multi-call modes, because the LZMA2
+ *                          dictionary doesn't need to be allocated as
+ *                          part of the decoder state. All required data
+ *                          structures are allocated at initialization,
+ *                          so xz_dec_run() cannot return XZ_MEM_ERROR.
+ * @XZ_PREALLOC:            Multi-call mode with preallocated LZMA2
+ *                          dictionary buffer. All data structures are
+ *                          allocated at initialization, so xz_dec_run()
+ *                          cannot return XZ_MEM_ERROR.
+ * @XZ_DYNALLOC:            Multi-call mode. The LZMA2 dictionary is
+ *                          allocated once the required size has been
+ *                          parsed from the stream headers. If the
+ *                          allocation fails, xz_dec_run() will return
+ *                          XZ_MEM_ERROR.
+ *
+ * It is possible to enable support only for a subset of the above
+ * modes at compile time by defining XZ_DEC_SINGLE, XZ_DEC_PREALLOC,
+ * or XZ_DEC_DYNALLOC. The xz_dec kernel module is always compiled
+ * with support for all operation modes, but the preboot code may
+ * be built with fewer features to minimize code size.
+ */
+enum xz_mode {
+	XZ_SINGLE,
+	XZ_PREALLOC,
+	XZ_DYNALLOC
+};
+
+/**
+ * enum xz_ret - Return codes
+ * @XZ_OK:                  Everything is OK so far. More input or more
+ *                          output space is required to continue. This
+ *                          return code is possible only in multi-call mode
+ *                          (XZ_PREALLOC or XZ_DYNALLOC).
+ * @XZ_STREAM_END:          Operation finished successfully.
+ * @XZ_UNSUPPORTED_CHECK:   Integrity check type is not supported. Decoding
+ *                          is still possible in multi-call mode by simply
+ *                          calling xz_dec_run() again.
+ *                          Note that this return value is used only if
+ *                          XZ_DEC_ANY_CHECK was defined at build time,
+ *                          which is not used in the kernel. Unsupported
+ *                          check types return XZ_OPTIONS_ERROR if
+ *                          XZ_DEC_ANY_CHECK was not defined at build time.
+ * @XZ_MEM_ERROR:           Allocating memory failed. This return code is
+ *                          possible only if the decoder was initialized
+ *                          with XZ_DYNALLOC. The amount of memory that was
+ *                          tried to be allocated was no more than the
+ *                          dict_max argument given to xz_dec_init().
+ * @XZ_MEMLIMIT_ERROR:      A bigger LZMA2 dictionary would be needed than
+ *                          allowed by the dict_max argument given to
+ *                          xz_dec_init(). This return value is possible
+ *                          only in multi-call mode (XZ_PREALLOC or
+ *                          XZ_DYNALLOC); the single-call mode (XZ_SINGLE)
+ *                          ignores the dict_max argument.
+ * @XZ_FORMAT_ERROR:        File format was not recognized (wrong magic
+ *                          bytes).
+ * @XZ_OPTIONS_ERROR:       This implementation doesn't support the requested
+ *                          compression options. In the decoder this means
+ *                          that the header CRC32 matches, but the header
+ *                          itself specifies something that we don't support.
+ * @XZ_DATA_ERROR:          Compressed data is corrupt.
+ * @XZ_BUF_ERROR:           Cannot make any progress. Details are slightly
+ *                          different between multi-call and single-call
+ *                          mode; more information below.
+ *
+ * In multi-call mode, XZ_BUF_ERROR is returned when two consecutive calls
+ * to XZ code cannot consume any input and cannot produce any new output.
+ * This happens when there is no new input available, or the output buffer
+ * is full while at least one output byte is still pending. Assuming your
+ * code is not buggy, you can get this error only when decoding a compressed
+ * stream that is truncated or otherwise corrupt.
+ *
+ * In single-call mode, XZ_BUF_ERROR is returned only when the output buffer
+ * is too small or the compressed input is corrupt in a way that makes the
+ * decoder produce more output than the caller expected. When it is
+ * (relatively) clear that the compressed input is truncated, XZ_DATA_ERROR
+ * is used instead of XZ_BUF_ERROR.
+ */
+enum xz_ret {
+	XZ_OK,
+	XZ_STREAM_END,
+	XZ_UNSUPPORTED_CHECK,
+	XZ_MEM_ERROR,
+	XZ_MEMLIMIT_ERROR,
+	XZ_FORMAT_ERROR,
+	XZ_OPTIONS_ERROR,
+	XZ_DATA_ERROR,
+	XZ_BUF_ERROR
+};
+
+/**
+ * struct xz_buf - Passing input and output buffers to XZ code
+ * @in:         Beginning of the input buffer. This may be NULL if and only
+ *              if in_pos is equal to in_size.
+ * @in_pos:     Current position in the input buffer. This must not exceed
+ *              in_size.
+ * @in_size:    Size of the input buffer
+ * @out:        Beginning of the output buffer. This may be NULL if and only
+ *              if out_pos is equal to out_size.
+ * @out_pos:    Current position in the output buffer. This must not exceed
+ *              out_size.
+ * @out_size:   Size of the output buffer
+ *
+ * Only the contents of the output buffer from out[out_pos] onward, and
+ * the variables in_pos and out_pos are modified by the XZ code.
+ */
+struct xz_buf {
+	const uint8_t *in;
+	size_t in_pos;
+	size_t in_size;
+
+	uint8_t *out;
+	size_t out_pos;
+	size_t out_size;
+};
+
+/**
+ * struct xz_dec - Opaque type to hold the XZ decoder state
+ */
+struct xz_dec;
+
+/**
+ * xz_dec_init() - Allocate and initialize a XZ decoder state
+ * @mode:       Operation mode
+ * @dict_max:   Maximum size of the LZMA2 dictionary (history buffer) for
+ *              multi-call decoding. This is ignored in single-call mode
+ *              (mode == XZ_SINGLE). LZMA2 dictionary is always 2^n bytes
+ *              or 2^n + 2^(n-1) bytes (the latter sizes are less common
+ *              in practice), so other values for dict_max don't make sense.
+ *              In the kernel, dictionary sizes of 64 KiB, 128 KiB, 256 KiB,
+ *              512 KiB, and 1 MiB are probably the only reasonable values,
+ *              except for kernel and initramfs images where a bigger
+ *              dictionary can be fine and useful.
+ *
+ * Single-call mode (XZ_SINGLE): xz_dec_run() decodes the whole stream at
+ * once. The caller must provide enough output space or the decoding will
+ * fail. The output space is used as the dictionary buffer, which is why
+ * there is no need to allocate the dictionary as part of the decoder's
+ * internal state.
+ *
+ * Because the output buffer is used as the workspace, streams encoded using
+ * a big dictionary are not a problem in single-call mode. It is enough that
+ * the output buffer is big enough to hold the actual uncompressed data; it
+ * can be smaller than the dictionary size stored in the stream headers.
+ *
+ * Multi-call mode with preallocated dictionary (XZ_PREALLOC): dict_max bytes
+ * of memory is preallocated for the LZMA2 dictionary. This way there is no
+ * risk that xz_dec_run() could run out of memory, since xz_dec_run() will
+ * never allocate any memory. Instead, if the preallocated dictionary is too
+ * small for decoding the given input stream, xz_dec_run() will return
+ * XZ_MEMLIMIT_ERROR. Thus, it is important to know what kind of data will be
+ * decoded to avoid allocating excessive amount of memory for the dictionary.
+ *
+ * Multi-call mode with dynamically allocated dictionary (XZ_DYNALLOC):
+ * dict_max specifies the maximum allowed dictionary size that xz_dec_run()
+ * may allocate once it has parsed the dictionary size from the stream
+ * headers. This way excessive allocations can be avoided while still
+ * limiting the maximum memory usage to a sane value to prevent running the
+ * system out of memory when decompressing streams from untrusted sources.
+ *
+ * On success, xz_dec_init() returns a pointer to struct xz_dec, which is
+ * ready to be used with xz_dec_run(). If memory allocation fails,
+ * xz_dec_init() returns NULL.
+ */
+XZ_EXTERN struct xz_dec *xz_dec_init(enum xz_mode mode, uint32_t dict_max);
+
+/**
+ * xz_dec_run() - Run the XZ decoder
+ * @s:          Decoder state allocated using xz_dec_init()
+ * @b:          Input and output buffers
+ *
+ * The possible return values depend on build options and operation mode.
+ * See enum xz_ret for details.
+ *
+ * Note that if an error occurs in single-call mode (return value is not
+ * XZ_STREAM_END), b->in_pos and b->out_pos are not modified and the
+ * contents of the output buffer from b->out[b->out_pos] onward are
+ * undefined. This is true even after XZ_BUF_ERROR, because with some filter
+ * chains, there may be a second pass over the output buffer, and this pass
+ * cannot be properly done if the output buffer is truncated. Thus, you
+ * cannot give the single-call decoder a too small buffer and then expect to
+ * get that amount valid data from the beginning of the stream. You must use
+ * the multi-call decoder if you don't want to uncompress the whole stream.
+ */
+XZ_EXTERN enum xz_ret xz_dec_run(struct xz_dec *s, struct xz_buf *b);
+
+/**
+ * xz_dec_reset() - Reset an already allocated decoder state
+ * @s:          Decoder state allocated using xz_dec_init()
+ *
+ * This function can be used to reset the multi-call decoder state without
+ * freeing and reallocating memory with xz_dec_end() and xz_dec_init().
+ *
+ * In single-call mode, xz_dec_reset() is always called in the beginning of
+ * xz_dec_run(). Thus, explicit call to xz_dec_reset() is useful only in
+ * multi-call mode.
+ */
+XZ_EXTERN void xz_dec_reset(struct xz_dec *s);
+
+/**
+ * xz_dec_end() - Free the memory allocated for the decoder state
+ * @s:          Decoder state allocated using xz_dec_init(). If s is NULL,
+ *              this function does nothing.
+ */
+XZ_EXTERN void xz_dec_end(struct xz_dec *s);
+
+/*
+ * Standalone build (userspace build or in-kernel build for boot time use)
+ * needs a CRC32 implementation. For normal in-kernel use, kernel's own
+ * CRC32 module is used instead, and users of this module don't need to
+ * care about the functions below.
+ */
+#ifndef XZ_INTERNAL_CRC32
+#	ifdef __KERNEL__
+#		define XZ_INTERNAL_CRC32 0
+#	else
+#		define XZ_INTERNAL_CRC32 1
+#	endif
+#endif
+
+#if XZ_INTERNAL_CRC32
+/*
+ * This must be called before any other xz_* function to initialize
+ * the CRC32 lookup table.
+ */
+XZ_EXTERN void xz_crc32_init(void);
+
+/*
+ * Update CRC32 value using the polynomial from IEEE-802.3. To start a new
+ * calculation, the third argument must be zero. To continue the calculation,
+ * the previously returned value is passed as the third argument.
+ */
+XZ_EXTERN uint32_t xz_crc32(const uint8_t *buf, size_t size, uint32_t crc);
+#endif
+#endif
diff --git a/lib/Kconfig b/lib/Kconfig
index 3116aa631af6..2b8f8540d670 100644
--- a/lib/Kconfig
+++ b/lib/Kconfig
@@ -106,6 +106,8 @@ config LZO_COMPRESS
 config LZO_DECOMPRESS
 	tristate
 
+source "lib/xz/Kconfig"
+
 #
 # These all provide a common interface (hence the apparent duplication with
 # ZLIB_INFLATE; DECOMPRESS_GZIP is just a wrapper.)
diff --git a/lib/Makefile b/lib/Makefile
index 2f59e0a1dd8d..4df2d0297721 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -69,6 +69,7 @@ obj-$(CONFIG_ZLIB_DEFLATE) += zlib_deflate/
 obj-$(CONFIG_REED_SOLOMON) += reed_solomon/
 obj-$(CONFIG_LZO_COMPRESS) += lzo/
 obj-$(CONFIG_LZO_DECOMPRESS) += lzo/
+obj-$(CONFIG_XZ_DEC) += xz/
 obj-$(CONFIG_RAID6_PQ) += raid6/
 
 lib-$(CONFIG_DECOMPRESS_GZIP) += decompress_inflate.o
diff --git a/lib/xz/Kconfig b/lib/xz/Kconfig
new file mode 100644
index 000000000000..e3b6e18fdac5
--- /dev/null
+++ b/lib/xz/Kconfig
@@ -0,0 +1,59 @@
+config XZ_DEC
+	tristate "XZ decompression support"
+	select CRC32
+	help
+	  LZMA2 compression algorithm and BCJ filters are supported using
+	  the .xz file format as the container. For integrity checking,
+	  CRC32 is supported. See Documentation/xz.txt for more information.
+
+config XZ_DEC_X86
+	bool "x86 BCJ filter decoder" if EMBEDDED
+	default y
+	depends on XZ_DEC
+	select XZ_DEC_BCJ
+
+config XZ_DEC_POWERPC
+	bool "PowerPC BCJ filter decoder" if EMBEDDED
+	default y
+	depends on XZ_DEC
+	select XZ_DEC_BCJ
+
+config XZ_DEC_IA64
+	bool "IA-64 BCJ filter decoder" if EMBEDDED
+	default y
+	depends on XZ_DEC
+	select XZ_DEC_BCJ
+
+config XZ_DEC_ARM
+	bool "ARM BCJ filter decoder" if EMBEDDED
+	default y
+	depends on XZ_DEC
+	select XZ_DEC_BCJ
+
+config XZ_DEC_ARMTHUMB
+	bool "ARM-Thumb BCJ filter decoder" if EMBEDDED
+	default y
+	depends on XZ_DEC
+	select XZ_DEC_BCJ
+
+config XZ_DEC_SPARC
+	bool "SPARC BCJ filter decoder" if EMBEDDED
+	default y
+	depends on XZ_DEC
+	select XZ_DEC_BCJ
+
+config XZ_DEC_BCJ
+	bool
+	default n
+
+config XZ_DEC_TEST
+	tristate "XZ decompressor tester"
+	default n
+	depends on XZ_DEC
+	help
+	  This allows passing .xz files to the in-kernel XZ decoder via
+	  a character special file. It calculates CRC32 of the decompressed
+	  data and writes diagnostics to the system log.
+
+	  Unless you are developing the XZ decoder, you don't need this
+	  and should say N.
diff --git a/lib/xz/Makefile b/lib/xz/Makefile
new file mode 100644
index 000000000000..a7fa7693f0f3
--- /dev/null
+++ b/lib/xz/Makefile
@@ -0,0 +1,5 @@
+obj-$(CONFIG_XZ_DEC) += xz_dec.o
+xz_dec-y := xz_dec_syms.o xz_dec_stream.o xz_dec_lzma2.o
+xz_dec-$(CONFIG_XZ_DEC_BCJ) += xz_dec_bcj.o
+
+obj-$(CONFIG_XZ_DEC_TEST) += xz_dec_test.o
diff --git a/lib/xz/xz_crc32.c b/lib/xz/xz_crc32.c
new file mode 100644
index 000000000000..34532d14fd4c
--- /dev/null
+++ b/lib/xz/xz_crc32.c
@@ -0,0 +1,59 @@
+/*
+ * CRC32 using the polynomial from IEEE-802.3
+ *
+ * Authors: Lasse Collin <lasse.collin@tukaani.org>
+ *          Igor Pavlov <http://7-zip.org/>
+ *
+ * This file has been put into the public domain.
+ * You can do whatever you want with this file.
+ */
+
+/*
+ * This is not the fastest implementation, but it is pretty compact.
+ * The fastest versions of xz_crc32() on modern CPUs without hardware
+ * accelerated CRC instruction are 3-5 times as fast as this version,
+ * but they are bigger and use more memory for the lookup table.
+ */
+
+#include "xz_private.h"
+
+/*
+ * STATIC_RW_DATA is used in the pre-boot environment on some architectures.
+ * See <linux/decompress/mm.h> for details.
+ */
+#ifndef STATIC_RW_DATA
+#	define STATIC_RW_DATA static
+#endif
+
+STATIC_RW_DATA uint32_t xz_crc32_table[256];
+
+XZ_EXTERN void xz_crc32_init(void)
+{
+	const uint32_t poly = 0xEDB88320;
+
+	uint32_t i;
+	uint32_t j;
+	uint32_t r;
+
+	for (i = 0; i < 256; ++i) {
+		r = i;
+		for (j = 0; j < 8; ++j)
+			r = (r >> 1) ^ (poly & ~((r & 1) - 1));
+
+		xz_crc32_table[i] = r;
+	}
+
+	return;
+}
+
+XZ_EXTERN uint32_t xz_crc32(const uint8_t *buf, size_t size, uint32_t crc)
+{
+	crc = ~crc;
+
+	while (size != 0) {
+		crc = xz_crc32_table[*buf++ ^ (crc & 0xFF)] ^ (crc >> 8);
+		--size;
+	}
+
+	return ~crc;
+}
diff --git a/lib/xz/xz_dec_bcj.c b/lib/xz/xz_dec_bcj.c
new file mode 100644
index 000000000000..e51e2558ca9d
--- /dev/null
+++ b/lib/xz/xz_dec_bcj.c
@@ -0,0 +1,561 @@
+/*
+ * Branch/Call/Jump (BCJ) filter decoders
+ *
+ * Authors: Lasse Collin <lasse.collin@tukaani.org>
+ *          Igor Pavlov <http://7-zip.org/>
+ *
+ * This file has been put into the public domain.
+ * You can do whatever you want with this file.
+ */
+
+#include "xz_private.h"
+
+/*
+ * The rest of the file is inside this ifdef. It makes things a little more
+ * convenient when building without support for any BCJ filters.
+ */
+#ifdef XZ_DEC_BCJ
+
+struct xz_dec_bcj {
+	/* Type of the BCJ filter being used */
+	enum {
+		BCJ_X86 = 4,        /* x86 or x86-64 */
+		BCJ_POWERPC = 5,    /* Big endian only */
+		BCJ_IA64 = 6,       /* Big or little endian */
+		BCJ_ARM = 7,        /* Little endian only */
+		BCJ_ARMTHUMB = 8,   /* Little endian only */
+		BCJ_SPARC = 9       /* Big or little endian */
+	} type;
+
+	/*
+	 * Return value of the next filter in the chain. We need to preserve
+	 * this information across calls, because we must not call the next
+	 * filter anymore once it has returned XZ_STREAM_END.
+	 */
+	enum xz_ret ret;
+
+	/* True if we are operating in single-call mode. */
+	bool single_call;
+
+	/*
+	 * Absolute position relative to the beginning of the uncompressed
+	 * data (in a single .xz Block). We care only about the lowest 32
+	 * bits so this doesn't need to be uint64_t even with big files.
+	 */
+	uint32_t pos;
+
+	/* x86 filter state */
+	uint32_t x86_prev_mask;
+
+	/* Temporary space to hold the variables from struct xz_buf */
+	uint8_t *out;
+	size_t out_pos;
+	size_t out_size;
+
+	struct {
+		/* Amount of already filtered data in the beginning of buf */
+		size_t filtered;
+
+		/* Total amount of data currently stored in buf  */
+		size_t size;
+
+		/*
+		 * Buffer to hold a mix of filtered and unfiltered data. This
+		 * needs to be big enough to hold Alignment + 2 * Look-ahead:
+		 *
+		 * Type         Alignment   Look-ahead
+		 * x86              1           4
+		 * PowerPC          4           0
+		 * IA-64           16           0
+		 * ARM              4           0
+		 * ARM-Thumb        2           2
+		 * SPARC            4           0
+		 */
+		uint8_t buf[16];
+	} temp;
+};
+
+#ifdef XZ_DEC_X86
+/*
+ * This is used to test the most significant byte of a memory address
+ * in an x86 instruction.
+ */
+static inline int bcj_x86_test_msbyte(uint8_t b)
+{
+	return b == 0x00 || b == 0xFF;
+}
+
+static size_t bcj_x86(struct xz_dec_bcj *s, uint8_t *buf, size_t size)
+{
+	static const bool mask_to_allowed_status[8]
+		= { true, true, true, false, true, false, false, false };
+
+	static const uint8_t mask_to_bit_num[8] = { 0, 1, 2, 2, 3, 3, 3, 3 };
+
+	size_t i;
+	size_t prev_pos = (size_t)-1;
+	uint32_t prev_mask = s->x86_prev_mask;
+	uint32_t src;
+	uint32_t dest;
+	uint32_t j;
+	uint8_t b;
+
+	if (size <= 4)
+		return 0;
+
+	size -= 4;
+	for (i = 0; i < size; ++i) {
+		if ((buf[i] & 0xFE) != 0xE8)
+			continue;
+
+		prev_pos = i - prev_pos;
+		if (prev_pos > 3) {
+			prev_mask = 0;
+		} else {
+			prev_mask = (prev_mask << (prev_pos - 1)) & 7;
+			if (prev_mask != 0) {
+				b = buf[i + 4 - mask_to_bit_num[prev_mask]];
+				if (!mask_to_allowed_status[prev_mask]
+						|| bcj_x86_test_msbyte(b)) {
+					prev_pos = i;
+					prev_mask = (prev_mask << 1) | 1;
+					continue;
+				}
+			}
+		}
+
+		prev_pos = i;
+
+		if (bcj_x86_test_msbyte(buf[i + 4])) {
+			src = get_unaligned_le32(buf + i + 1);
+			while (true) {
+				dest = src - (s->pos + (uint32_t)i + 5);
+				if (prev_mask == 0)
+					break;
+
+				j = mask_to_bit_num[prev_mask] * 8;
+				b = (uint8_t)(dest >> (24 - j));
+				if (!bcj_x86_test_msbyte(b))
+					break;
+
+				src = dest ^ (((uint32_t)1 << (32 - j)) - 1);
+			}
+
+			dest &= 0x01FFFFFF;
+			dest |= (uint32_t)0 - (dest & 0x01000000);
+			put_unaligned_le32(dest, buf + i + 1);
+			i += 4;
+		} else {
+			prev_mask = (prev_mask << 1) | 1;
+		}
+	}
+
+	prev_pos = i - prev_pos;
+	s->x86_prev_mask = prev_pos > 3 ? 0 : prev_mask << (prev_pos - 1);
+	return i;
+}
+#endif
+
+#ifdef XZ_DEC_POWERPC
+static size_t bcj_powerpc(struct xz_dec_bcj *s, uint8_t *buf, size_t size)
+{
+	size_t i;
+	uint32_t instr;
+
+	for (i = 0; i + 4 <= size; i += 4) {
+		instr = get_unaligned_be32(buf + i);
+		if ((instr & 0xFC000003) == 0x48000001) {
+			instr &= 0x03FFFFFC;
+			instr -= s->pos + (uint32_t)i;
+			instr &= 0x03FFFFFC;
+			instr |= 0x48000001;
+			put_unaligned_be32(instr, buf + i);
+		}
+	}
+
+	return i;
+}
+#endif
+
+#ifdef XZ_DEC_IA64
+static size_t bcj_ia64(struct xz_dec_bcj *s, uint8_t *buf, size_t size)
+{
+	static const uint8_t branch_table[32] = {
+		0, 0, 0, 0, 0, 0, 0, 0,
+		0, 0, 0, 0, 0, 0, 0, 0,
+		4, 4, 6, 6, 0, 0, 7, 7,
+		4, 4, 0, 0, 4, 4, 0, 0
+	};
+
+	/*
+	 * The local variables take a little bit stack space, but it's less
+	 * than what LZMA2 decoder takes, so it doesn't make sense to reduce
+	 * stack usage here without doing that for the LZMA2 decoder too.
+	 */
+
+	/* Loop counters */
+	size_t i;
+	size_t j;
+
+	/* Instruction slot (0, 1, or 2) in the 128-bit instruction word */
+	uint32_t slot;
+
+	/* Bitwise offset of the instruction indicated by slot */
+	uint32_t bit_pos;
+
+	/* bit_pos split into byte and bit parts */
+	uint32_t byte_pos;
+	uint32_t bit_res;
+
+	/* Address part of an instruction */
+	uint32_t addr;
+
+	/* Mask used to detect which instructions to convert */
+	uint32_t mask;
+
+	/* 41-bit instruction stored somewhere in the lowest 48 bits */
+	uint64_t instr;
+
+	/* Instruction normalized with bit_res for easier manipulation */
+	uint64_t norm;
+
+	for (i = 0; i + 16 <= size; i += 16) {
+		mask = branch_table[buf[i] & 0x1F];
+		for (slot = 0, bit_pos = 5; slot < 3; ++slot, bit_pos += 41) {
+			if (((mask >> slot) & 1) == 0)
+				continue;
+
+			byte_pos = bit_pos >> 3;
+			bit_res = bit_pos & 7;
+			instr = 0;
+			for (j = 0; j < 6; ++j)
+				instr |= (uint64_t)(buf[i + j + byte_pos])
+						<< (8 * j);
+
+			norm = instr >> bit_res;
+
+			if (((norm >> 37) & 0x0F) == 0x05
+					&& ((norm >> 9) & 0x07) == 0) {
+				addr = (norm >> 13) & 0x0FFFFF;
+				addr |= ((uint32_t)(norm >> 36) & 1) << 20;
+				addr <<= 4;
+				addr -= s->pos + (uint32_t)i;
+				addr >>= 4;
+
+				norm &= ~((uint64_t)0x8FFFFF << 13);
+				norm |= (uint64_t)(addr & 0x0FFFFF) << 13;
+				norm |= (uint64_t)(addr & 0x100000)
+						<< (36 - 20);
+
+				instr &= (1 << bit_res) - 1;
+				instr |= norm << bit_res;
+
+				for (j = 0; j < 6; j++)
+					buf[i + j + byte_pos]
+						= (uint8_t)(instr >> (8 * j));
+			}
+		}
+	}
+
+	return i;
+}
+#endif
+
+#ifdef XZ_DEC_ARM
+static size_t bcj_arm(struct xz_dec_bcj *s, uint8_t *buf, size_t size)
+{
+	size_t i;
+	uint32_t addr;
+
+	for (i = 0; i + 4 <= size; i += 4) {
+		if (buf[i + 3] == 0xEB) {
+			addr = (uint32_t)buf[i] | ((uint32_t)buf[i + 1] << 8)
+					| ((uint32_t)buf[i + 2] << 16);
+			addr <<= 2;
+			addr -= s->pos + (uint32_t)i + 8;
+			addr >>= 2;
+			buf[i] = (uint8_t)addr;
+			buf[i + 1] = (uint8_t)(addr >> 8);
+			buf[i + 2] = (uint8_t)(addr >> 16);
+		}
+	}
+
+	return i;
+}
+#endif
+
+#ifdef XZ_DEC_ARMTHUMB
+static size_t bcj_armthumb(struct xz_dec_bcj *s, uint8_t *buf, size_t size)
+{
+	size_t i;
+	uint32_t addr;
+
+	for (i = 0; i + 4 <= size; i += 2) {
+		if ((buf[i + 1] & 0xF8) == 0xF0
+				&& (buf[i + 3] & 0xF8) == 0xF8) {
+			addr = (((uint32_t)buf[i + 1] & 0x07) << 19)
+					| ((uint32_t)buf[i] << 11)
+					| (((uint32_t)buf[i + 3] & 0x07) << 8)
+					| (uint32_t)buf[i + 2];
+			addr <<= 1;
+			addr -= s->pos + (uint32_t)i + 4;
+			addr >>= 1;
+			buf[i + 1] = (uint8_t)(0xF0 | ((addr >> 19) & 0x07));
+			buf[i] = (uint8_t)(addr >> 11);
+			buf[i + 3] = (uint8_t)(0xF8 | ((addr >> 8) & 0x07));
+			buf[i + 2] = (uint8_t)addr;
+			i += 2;
+		}
+	}
+
+	return i;
+}
+#endif
+
+#ifdef XZ_DEC_SPARC
+static size_t bcj_sparc(struct xz_dec_bcj *s, uint8_t *buf, size_t size)
+{
+	size_t i;
+	uint32_t instr;
+
+	for (i = 0; i + 4 <= size; i += 4) {
+		instr = get_unaligned_be32(buf + i);
+		if ((instr >> 22) == 0x100 || (instr >> 22) == 0x1FF) {
+			instr <<= 2;
+			instr -= s->pos + (uint32_t)i;
+			instr >>= 2;
+			instr = ((uint32_t)0x40000000 - (instr & 0x400000))
+					| 0x40000