Add support for LZ4 decompression algorithm

This patch adds support for LZ4-compressed FIT image contents. This
algorithm has a slightly worse compression ration than LZO while being
nearly twice as fast to decompress. When loading images from a fast
storage medium this usually results in a boot time win.

Sandbox-tested only since I don't have a U-Boot development system set
up right now. The code was imported unchanged from coreboot where it's
proven to work, though. I'm mostly interested in getting this recognized
by mkImage for use in a downstream project.

Signed-off-by: Julius Werner <jwerner@chromium.org>
Acked-by: Simon Glass <sjg@chromium.org>
diff --git a/lib/Kconfig b/lib/Kconfig
index 0673072..a8f8460 100644
--- a/lib/Kconfig
+++ b/lib/Kconfig
@@ -100,6 +100,24 @@
 	  is performed in hardware.
 endmenu
 
+menu "Compression Support"
+
+config LZ4
+	bool "Enable LZ4 decompression support"
+	help
+	  If this option is set, support for LZ4 compressed images
+	  is included. The LZ4 algorithm can run in-place as long as the
+	  compressed image is loaded to the end of the output buffer, and
+	  trades lower compression ratios for much faster decompression.
+	  
+	  NOTE: This implements the release version of the LZ4 frame
+	  format as generated by default by the 'lz4' command line tool.
+	  This is not the same as the outdated, less efficient legacy
+	  frame format currently (2015) implemented in the Linux kernel
+	  (generated by 'lz4 -l'). The two formats are incompatible.
+
+endmenu
+
 config ERRNO_STR
 	bool "Enable function for getting errno-related string message"
 	help
diff --git a/lib/Makefile b/lib/Makefile
index 96f832e..3eecefa 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -34,6 +34,7 @@
 obj-y += initcall.o
 obj-$(CONFIG_LMB) += lmb.o
 obj-y += ldiv.o
+obj-$(CONFIG_LZ4) += lz4_wrapper.o
 obj-$(CONFIG_MD5) += md5.o
 obj-y += net_utils.o
 obj-$(CONFIG_PHYSMEM) += physmem.o
diff --git a/lib/lz4.c b/lib/lz4.c
new file mode 100644
index 0000000..f518341
--- /dev/null
+++ b/lib/lz4.c
@@ -0,0 +1,243 @@
+/*
+   LZ4 - Fast LZ compression algorithm
+   Copyright (C) 2011-2015, Yann Collet.
+
+   SPDX-License-Identifier: BSD-2-Clause
+
+   You can contact the author at :
+   - LZ4 source repository : https://github.com/Cyan4973/lz4
+   - LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c
+*/
+
+
+/**************************************
+*  Reading and writing into memory
+**************************************/
+
+/* customized version of memcpy, which may overwrite up to 7 bytes beyond dstEnd */
+static void LZ4_wildCopy(void* dstPtr, const void* srcPtr, void* dstEnd)
+{
+    BYTE* d = (BYTE*)dstPtr;
+    const BYTE* s = (const BYTE*)srcPtr;
+    BYTE* e = (BYTE*)dstEnd;
+    do { LZ4_copy8(d,s); d+=8; s+=8; } while (d<e);
+}
+
+
+/**************************************
+*  Common Constants
+**************************************/
+#define MINMATCH 4
+
+#define COPYLENGTH 8
+#define LASTLITERALS 5
+#define MFLIMIT (COPYLENGTH+MINMATCH)
+static const int LZ4_minLength = (MFLIMIT+1);
+
+#define KB *(1 <<10)
+#define MB *(1 <<20)
+#define GB *(1U<<30)
+
+#define MAXD_LOG 16
+#define MAX_DISTANCE ((1 << MAXD_LOG) - 1)
+
+#define ML_BITS  4
+#define ML_MASK  ((1U<<ML_BITS)-1)
+#define RUN_BITS (8-ML_BITS)
+#define RUN_MASK ((1U<<RUN_BITS)-1)
+
+
+/**************************************
+*  Local Structures and types
+**************************************/
+typedef enum { noDict = 0, withPrefix64k, usingExtDict } dict_directive;
+typedef enum { endOnOutputSize = 0, endOnInputSize = 1 } endCondition_directive;
+typedef enum { full = 0, partial = 1 } earlyEnd_directive;
+
+
+
+/*******************************
+*  Decompression functions
+*******************************/
+/*
+ * This generic decompression function cover all use cases.
+ * It shall be instantiated several times, using different sets of directives
+ * Note that it is essential this generic function is really inlined,
+ * in order to remove useless branches during compilation optimization.
+ */
+FORCE_INLINE int LZ4_decompress_generic(
+                 const char* const source,
+                 char* const dest,
+                 int inputSize,
+                 int outputSize,         /* If endOnInput==endOnInputSize, this value is the max size of Output Buffer. */
+
+                 int endOnInput,         /* endOnOutputSize, endOnInputSize */
+                 int partialDecoding,    /* full, partial */
+                 int targetOutputSize,   /* only used if partialDecoding==partial */
+                 int dict,               /* noDict, withPrefix64k, usingExtDict */
+                 const BYTE* const lowPrefix,  /* == dest if dict == noDict */
+                 const BYTE* const dictStart,  /* only if dict==usingExtDict */
+                 const size_t dictSize         /* note : = 0 if noDict */
+                 )
+{
+    /* Local Variables */
+    const BYTE* ip = (const BYTE*) source;
+    const BYTE* const iend = ip + inputSize;
+
+    BYTE* op = (BYTE*) dest;
+    BYTE* const oend = op + outputSize;
+    BYTE* cpy;
+    BYTE* oexit = op + targetOutputSize;
+    const BYTE* const lowLimit = lowPrefix - dictSize;
+
+    const BYTE* const dictEnd = (const BYTE*)dictStart + dictSize;
+    const size_t dec32table[] = {4, 1, 2, 1, 4, 4, 4, 4};
+    const size_t dec64table[] = {0, 0, 0, (size_t)-1, 0, 1, 2, 3};
+
+    const int safeDecode = (endOnInput==endOnInputSize);
+    const int checkOffset = ((safeDecode) && (dictSize < (int)(64 KB)));
+
+
+    /* Special cases */
+    if ((partialDecoding) && (oexit> oend-MFLIMIT)) oexit = oend-MFLIMIT;                         /* targetOutputSize too high => decode everything */
+    if ((endOnInput) && (unlikely(outputSize==0))) return ((inputSize==1) && (*ip==0)) ? 0 : -1;  /* Empty output buffer */
+    if ((!endOnInput) && (unlikely(outputSize==0))) return (*ip==0?1:-1);
+
+
+    /* Main Loop */
+    while (1)
+    {
+        unsigned token;
+        size_t length;
+        const BYTE* match;
+
+        /* get literal length */
+        token = *ip++;
+        if ((length=(token>>ML_BITS)) == RUN_MASK)
+        {
+            unsigned s;
+            do
+            {
+                s = *ip++;
+                length += s;
+            }
+            while (likely((endOnInput)?ip<iend-RUN_MASK:1) && (s==255));
+            if ((safeDecode) && unlikely((size_t)(op+length)<(size_t)(op))) goto _output_error;   /* overflow detection */
+            if ((safeDecode) && unlikely((size_t)(ip+length)<(size_t)(ip))) goto _output_error;   /* overflow detection */
+        }
+
+        /* copy literals */
+        cpy = op+length;
+        if (((endOnInput) && ((cpy>(partialDecoding?oexit:oend-MFLIMIT)) || (ip+length>iend-(2+1+LASTLITERALS))) )
+            || ((!endOnInput) && (cpy>oend-COPYLENGTH)))
+        {
+            if (partialDecoding)
+            {
+                if (cpy > oend) goto _output_error;                           /* Error : write attempt beyond end of output buffer */
+                if ((endOnInput) && (ip+length > iend)) goto _output_error;   /* Error : read attempt beyond end of input buffer */
+            }
+            else
+            {
+                if ((!endOnInput) && (cpy != oend)) goto _output_error;       /* Error : block decoding must stop exactly there */
+                if ((endOnInput) && ((ip+length != iend) || (cpy > oend))) goto _output_error;   /* Error : input must be consumed */
+            }
+            memcpy(op, ip, length);
+            ip += length;
+            op += length;
+            break;     /* Necessarily EOF, due to parsing restrictions */
+        }
+        LZ4_wildCopy(op, ip, cpy);
+        ip += length; op = cpy;
+
+        /* get offset */
+        match = cpy - LZ4_readLE16(ip); ip+=2;
+        if ((checkOffset) && (unlikely(match < lowLimit))) goto _output_error;   /* Error : offset outside destination buffer */
+
+        /* get matchlength */
+        length = token & ML_MASK;
+        if (length == ML_MASK)
+        {
+            unsigned s;
+            do
+            {
+                if ((endOnInput) && (ip > iend-LASTLITERALS)) goto _output_error;
+                s = *ip++;
+                length += s;
+            } while (s==255);
+            if ((safeDecode) && unlikely((size_t)(op+length)<(size_t)op)) goto _output_error;   /* overflow detection */
+        }
+        length += MINMATCH;
+
+        /* check external dictionary */
+        if ((dict==usingExtDict) && (match < lowPrefix))
+        {
+            if (unlikely(op+length > oend-LASTLITERALS)) goto _output_error;   /* doesn't respect parsing restriction */
+
+            if (length <= (size_t)(lowPrefix-match))
+            {
+                /* match can be copied as a single segment from external dictionary */
+                match = dictEnd - (lowPrefix-match);
+                memmove(op, match, length); op += length;
+            }
+            else
+            {
+                /* match encompass external dictionary and current segment */
+                size_t copySize = (size_t)(lowPrefix-match);
+                memcpy(op, dictEnd - copySize, copySize);
+                op += copySize;
+                copySize = length - copySize;
+                if (copySize > (size_t)(op-lowPrefix))   /* overlap within current segment */
+                {
+                    BYTE* const endOfMatch = op + copySize;
+                    const BYTE* copyFrom = lowPrefix;
+                    while (op < endOfMatch) *op++ = *copyFrom++;
+                }
+                else
+                {
+                    memcpy(op, lowPrefix, copySize);
+                    op += copySize;
+                }
+            }
+            continue;
+        }
+
+        /* copy repeated sequence */
+        cpy = op + length;
+        if (unlikely((op-match)<8))
+        {
+            const size_t dec64 = dec64table[op-match];
+            op[0] = match[0];
+            op[1] = match[1];
+            op[2] = match[2];
+            op[3] = match[3];
+            match += dec32table[op-match];
+            LZ4_copy4(op+4, match);
+            op += 8; match -= dec64;
+        } else { LZ4_copy8(op, match); op+=8; match+=8; }
+
+        if (unlikely(cpy>oend-12))
+        {
+            if (cpy > oend-LASTLITERALS) goto _output_error;    /* Error : last LASTLITERALS bytes must be literals */
+            if (op < oend-8)
+            {
+                LZ4_wildCopy(op, match, oend-8);
+                match += (oend-8) - op;
+                op = oend-8;
+            }
+            while (op<cpy) *op++ = *match++;
+        }
+        else
+            LZ4_wildCopy(op, match, cpy);
+        op=cpy;   /* correction */
+    }
+
+    /* end of decoding */
+    if (endOnInput)
+       return (int) (((char*)op)-dest);     /* Nb of output bytes decoded */
+    else
+       return (int) (((const char*)ip)-source);   /* Nb of input bytes read */
+
+    /* Overflow error detected */
+_output_error:
+    return (int) (-(((const char*)ip)-source))-1;
+}
diff --git a/lib/lz4_wrapper.c b/lib/lz4_wrapper.c
new file mode 100644
index 0000000..0739663
--- /dev/null
+++ b/lib/lz4_wrapper.c
@@ -0,0 +1,137 @@
+/*
+ * Copyright 2015 Google Inc.
+ *
+ * SPDX-License-Identifier: GPL 2.0+ BSD-3-Clause
+ */
+
+#include <common.h>
+#include <compiler.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+
+static u16 LZ4_readLE16(const void *src) { return le16_to_cpu(*(u16 *)src); }
+static void LZ4_copy4(void *dst, const void *src) { *(u32 *)dst = *(u32 *)src; }
+static void LZ4_copy8(void *dst, const void *src) { *(u64 *)dst = *(u64 *)src; }
+
+typedef  uint8_t BYTE;
+typedef uint16_t U16;
+typedef uint32_t U32;
+typedef  int32_t S32;
+typedef uint64_t U64;
+
+#define FORCE_INLINE static inline __attribute__((always_inline))
+
+/* Unaltered (except removing unrelated code) from github.com/Cyan4973/lz4. */
+#include "lz4.c"	/* #include for inlining, do not link! */
+
+#define LZ4F_MAGIC 0x184D2204
+
+struct lz4_frame_header {
+	u32 magic;
+	union {
+		u8 flags;
+		struct {
+			u8 reserved0:2;
+			u8 has_content_checksum:1;
+			u8 has_content_size:1;
+			u8 has_block_checksum:1;
+			u8 independent_blocks:1;
+			u8 version:2;
+		};
+	};
+	union {
+		u8 block_descriptor;
+		struct {
+			u8 reserved1:4;
+			u8 max_block_size:3;
+			u8 reserved2:1;
+		};
+	};
+	/* + u64 content_size iff has_content_size is set */
+	/* + u8 header_checksum */
+} __packed;
+
+struct lz4_block_header {
+	union {
+		u32 raw;
+		struct {
+			u32 size:31;
+			u32 not_compressed:1;
+		};
+	};
+	/* + size bytes of data */
+	/* + u32 block_checksum iff has_block_checksum is set */
+} __packed;
+
+int ulz4fn(const void *src, size_t srcn, void *dst, size_t *dstn)
+{
+	const void *end = dst + *dstn;
+	const void *in = src;
+	void *out = dst;
+	int has_block_checksum;
+	int ret;
+	*dstn = 0;
+
+	{ /* With in-place decompression the header may become invalid later. */
+		const struct lz4_frame_header *h = in;
+
+		if (srcn < sizeof(*h) + sizeof(u64) + sizeof(u8))
+			return -EINVAL;	/* input overrun */
+
+		/* We assume there's always only a single, standard frame. */
+		if (le32_to_cpu(h->magic) != LZ4F_MAGIC || h->version != 1)
+			return -EPROTONOSUPPORT;	/* unknown format */
+		if (h->reserved0 || h->reserved1 || h->reserved2)
+			return -EINVAL;	/* reserved must be zero */
+		if (!h->independent_blocks)
+			return -EPROTONOSUPPORT; /* we can't support this yet */
+		has_block_checksum = h->has_block_checksum;
+
+		in += sizeof(*h);
+		if (h->has_content_size)
+			in += sizeof(u64);
+		in += sizeof(u8);
+	}
+
+	while (1) {
+		struct lz4_block_header b = { .raw = le32_to_cpu(*(u32 *)in) };
+		in += sizeof(struct lz4_block_header);
+
+		if (in - src + b.size > srcn) {
+			ret = -EINVAL;		/* input overrun */
+			break;
+		}
+
+		if (!b.size) {
+			ret = 0;	/* decompression successful */
+			break;
+		}
+
+		if (b.not_compressed) {
+			size_t size = min((ptrdiff_t)b.size, end - out);
+			memcpy(out, in, size);
+			out += size;
+			if (size < b.size) {
+				ret = -ENOBUFS;	/* output overrun */
+				break;
+			}
+		} else {
+			/* constant folding essential, do not touch params! */
+			ret = LZ4_decompress_generic(in, out, b.size,
+					end - out, endOnInputSize,
+					full, 0, noDict, out, NULL, 0);
+			if (ret < 0) {
+				ret = -EPROTO;	/* decompression error */
+				break;
+			}
+			out += ret;
+		}
+
+		in += b.size;
+		if (has_block_checksum)
+			in += sizeof(u32);
+	}
+
+	*dstn = out - dst;
+	return ret;
+}