* Add SHA-256.

* Tests for the various hashes.

6 files changed, 904 insertions, 5 deletions

diff --git a/src/libutil/Makefile.am b/src/libutil/Makefile.am
index 12459b1db5..6e83b85551 100644
--- a/src/libutil/Makefile.am
+++ b/src/libutil/Makefile.am
@@ -1,7 +1,7 @@
 noinst_LIBRARIES = libutil.a
 
 libutil_a_SOURCES = util.cc util.hh hash.cc hash.hh \
- archive.cc archive.hh md5.c md5.h sha1.c sha1.h \
- aterm.cc aterm.hh 
+ archive.cc archive.hh aterm.cc aterm.hh \
+ md5.c md5.h sha1.c sha1.h sha256.c sha256.h md32_common.h
 
 AM_CXXFLAGS = -Wall -I.. ${aterm_include}
diff --git a/src/libutil/hash.cc b/src/libutil/hash.cc
index 46334ad820..599d375dca 100644
--- a/src/libutil/hash.cc
+++ b/src/libutil/hash.cc
@@ -3,6 +3,7 @@
 extern "C" {
 #include "md5.h"
 #include "sha1.h"
+#include "sha256.h"
 }
 
 #include "hash.hh"
@@ -19,6 +20,7 @@ Hash::Hash(HashType type)
     this->type = type;
     if (type == htMD5) hashSize = md5HashSize;
     else if (type == htSHA1) hashSize = sha1HashSize;
+    else if (type == htSHA256) hashSize = sha256HashSize;
     else throw Error("unknown hash type");
     memset(hash, 0, hashSize);
 }
@@ -96,6 +98,7 @@ struct Ctx
 {
     md5_ctx md5;
     sha_ctx sha1;
+    SHA256_CTX sha256;
 };
 
 
@@ -103,6 +106,7 @@ static void start(HashType ht, Ctx & ctx)
 {
     if (ht == htMD5) md5_init_ctx(&ctx.md5);
     else if (ht == htSHA1) sha_init(&ctx.sha1);
+    else if (ht == htSHA256) SHA256_Init(&ctx.sha256);
 }
 
 
@@ -111,6 +115,7 @@ static void update(HashType ht, Ctx & ctx,
 {
     if (ht == htMD5) md5_process_bytes(bytes, len, &ctx.md5);
     else if (ht == htSHA1) sha_update(&ctx.sha1, bytes, len);
+    else if (ht == htSHA256) SHA256_Update(&ctx.sha256, bytes, len);
 }
 
 
@@ -121,6 +126,7 @@ static void finish(HashType ht, Ctx & ctx, unsigned char * hash)
         sha_final(&ctx.sha1);
         sha_digest(&ctx.sha1, hash);
     }
+    else if (ht == htSHA256) SHA256_Final(hash, &ctx.sha256);
 }
 
 
diff --git a/src/libutil/hash.hh b/src/libutil/hash.hh
index 1d2fce6455..d82cd87a21 100644
--- a/src/libutil/hash.hh
+++ b/src/libutil/hash.hh
@@ -8,11 +8,12 @@
 using namespace std;
 
 
-typedef enum { htMD5, htSHA1 } HashType;
+typedef enum { htMD5, htSHA1, htSHA256 } HashType;
 
 
 const int md5HashSize = 16;
 const int sha1HashSize = 20;
+const int sha256HashSize = 32;
 
 
 struct Hash
@@ -53,8 +54,7 @@ Hash hashString(const string & s, HashType ht);
 Hash hashFile(const Path & path, HashType ht);
 
 /* Compute the hash of the given path.  The hash is defined as
-   md5(dump(path)).
-*/
+   md5(dump(path)). */
 Hash hashPath(const Path & path, HashType ht);
 
 
diff --git a/src/libutil/md32_common.h b/src/libutil/md32_common.h
new file mode 100644
index 0000000000..0cbcfaf8a2
--- /dev/null
+++ b/src/libutil/md32_common.h
@@ -0,0 +1,620 @@
+/* crypto/md32_common.h */
+/* ====================================================================
+ * Copyright (c) 1999-2002 The OpenSSL Project.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer. 
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * 3. All advertising materials mentioning features or use of this
+ *    software must display the following acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit. (http://www.OpenSSL.org/)"
+ *
+ * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
+ *    endorse or promote products derived from this software without
+ *    prior written permission. For written permission, please contact
+ *    licensing@OpenSSL.org.
+ *
+ * 5. Products derived from this software may not be called "OpenSSL"
+ *    nor may "OpenSSL" appear in their names without prior written
+ *    permission of the OpenSSL Project.
+ *
+ * 6. Redistributions of any form whatsoever must retain the following
+ *    acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit (http://www.OpenSSL.org/)"
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
+ * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
+ * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ * ====================================================================
+ *
+ * This product includes cryptographic software written by Eric Young
+ * (eay@cryptsoft.com).  This product includes software written by Tim
+ * Hudson (tjh@cryptsoft.com).
+ *
+ */
+
+/*
+ * This is a generic 32 bit "collector" for message digest algorithms.
+ * Whenever needed it collects input character stream into chunks of
+ * 32 bit values and invokes a block function that performs actual hash
+ * calculations.
+ *
+ * Porting guide.
+ *
+ * Obligatory macros:
+ *
+ * DATA_ORDER_IS_BIG_ENDIAN or DATA_ORDER_IS_LITTLE_ENDIAN
+ *	this macro defines byte order of input stream.
+ * HASH_CBLOCK
+ *	size of a unit chunk HASH_BLOCK operates on.
+ * HASH_LONG
+ *	has to be at lest 32 bit wide, if it's wider, then
+ *	HASH_LONG_LOG2 *has to* be defined along
+ * HASH_CTX
+ *	context structure that at least contains following
+ *	members:
+ *		typedef struct {
+ *			...
+ *			HASH_LONG	Nl,Nh;
+ *			HASH_LONG	data[HASH_LBLOCK];
+ *			unsigned int	num;
+ *			...
+ *			} HASH_CTX;
+ * HASH_UPDATE
+ *	name of "Update" function, implemented here.
+ * HASH_TRANSFORM
+ *	name of "Transform" function, implemented here.
+ * HASH_FINAL
+ *	name of "Final" function, implemented here.
+ * HASH_BLOCK_HOST_ORDER
+ *	name of "block" function treating *aligned* input message
+ *	in host byte order, implemented externally.
+ * HASH_BLOCK_DATA_ORDER
+ *	name of "block" function treating *unaligned* input message
+ *	in original (data) byte order, implemented externally (it
+ *	actually is optional if data and host are of the same
+ *	"endianess").
+ * HASH_MAKE_STRING
+ *	macro convering context variables to an ASCII hash string.
+ *
+ * Optional macros:
+ *
+ * B_ENDIAN or L_ENDIAN
+ *	defines host byte-order.
+ * HASH_LONG_LOG2
+ *	defaults to 2 if not states otherwise.
+ * HASH_LBLOCK
+ *	assumed to be HASH_CBLOCK/4 if not stated otherwise.
+ * HASH_BLOCK_DATA_ORDER_ALIGNED
+ *	alternative "block" function capable of treating
+ *	aligned input message in original (data) order,
+ *	implemented externally.
+ *
+ * MD5 example:
+ *
+ *	#define DATA_ORDER_IS_LITTLE_ENDIAN
+ *
+ *	#define HASH_LONG		MD5_LONG
+ *	#define HASH_LONG_LOG2		MD5_LONG_LOG2
+ *	#define HASH_CTX		MD5_CTX
+ *	#define HASH_CBLOCK		MD5_CBLOCK
+ *	#define HASH_LBLOCK		MD5_LBLOCK
+ *	#define HASH_UPDATE		MD5_Update
+ *	#define HASH_TRANSFORM		MD5_Transform
+ *	#define HASH_FINAL		MD5_Final
+ *	#define HASH_BLOCK_HOST_ORDER	md5_block_host_order
+ *	#define HASH_BLOCK_DATA_ORDER	md5_block_data_order
+ *
+ *					<appro@fy.chalmers.se>
+ */
+
+#if !defined(DATA_ORDER_IS_BIG_ENDIAN) && !defined(DATA_ORDER_IS_LITTLE_ENDIAN)
+#error "DATA_ORDER must be defined!"
+#endif
+
+#ifndef HASH_CBLOCK
+#error "HASH_CBLOCK must be defined!"
+#endif
+#ifndef HASH_LONG
+#error "HASH_LONG must be defined!"
+#endif
+#ifndef HASH_CTX
+#error "HASH_CTX must be defined!"
+#endif
+
+#ifndef HASH_UPDATE
+#error "HASH_UPDATE must be defined!"
+#endif
+#ifndef HASH_TRANSFORM
+#error "HASH_TRANSFORM must be defined!"
+#endif
+#ifndef HASH_FINAL
+#error "HASH_FINAL must be defined!"
+#endif
+
+#ifndef HASH_BLOCK_HOST_ORDER
+#error "HASH_BLOCK_HOST_ORDER must be defined!"
+#endif
+
+#if 0
+/*
+ * Moved below as it's required only if HASH_BLOCK_DATA_ORDER_ALIGNED
+ * isn't defined.
+ */
+#ifndef HASH_BLOCK_DATA_ORDER
+#error "HASH_BLOCK_DATA_ORDER must be defined!"
+#endif
+#endif
+
+#ifndef HASH_LBLOCK
+#define HASH_LBLOCK	(HASH_CBLOCK/4)
+#endif
+
+#ifndef HASH_LONG_LOG2
+#define HASH_LONG_LOG2	2
+#endif
+
+/*
+ * Engage compiler specific rotate intrinsic function if available.
+ */
+#undef ROTATE
+#ifndef PEDANTIC
+# if defined(_MSC_VER) || defined(__ICC)
+#  define ROTATE(a,n)	_lrotl(a,n)
+# elif defined(__MWERKS__)
+#  if defined(__POWERPC__)
+#   define ROTATE(a,n)	__rlwinm(a,n,0,31)
+#  elif defined(__MC68K__)
+    /* Motorola specific tweak. <appro@fy.chalmers.se> */
+#   define ROTATE(a,n)	( n<24 ? __rol(a,n) : __ror(a,32-n) )
+#  else
+#   define ROTATE(a,n)	__rol(a,n)
+#  endif
+# elif defined(__GNUC__) && __GNUC__>=2 && !defined(OPENSSL_NO_ASM) && !defined(OPENSSL_NO_INLINE_ASM)
+  /*
+   * Some GNU C inline assembler templates. Note that these are
+   * rotates by *constant* number of bits! But that's exactly
+   * what we need here...
+   * 					<appro@fy.chalmers.se>
+   */
+#  if defined(__i386) || defined(__i386__) || defined(__x86_64) || defined(__x86_64__)
+#   define ROTATE(a,n)	({ register unsigned int ret;	\
+				asm (			\
+				"roll %1,%0"		\
+				: "=r"(ret)		\
+				: "I"(n), "0"(a)	\
+				: "cc");		\
+			   ret;				\
+			})
+#  elif defined(__powerpc) || defined(__ppc__) || defined(__powerpc64__)
+#   define ROTATE(a,n)	({ register unsigned int ret;	\
+				asm (			\
+				"rlwinm %0,%1,%2,0,31"	\
+				: "=r"(ret)		\
+				: "r"(a), "I"(n));	\
+			   ret;				\
+			})
+#  endif
+# endif
+#endif /* PEDANTIC */
+
+#if HASH_LONG_LOG2==2	/* Engage only if sizeof(HASH_LONG)== 4 */
+/* A nice byte order reversal from Wei Dai <weidai@eskimo.com> */
+#ifdef ROTATE
+/* 5 instructions with rotate instruction, else 9 */
+#define REVERSE_FETCH32(a,l)	(					\
+		l=*(const HASH_LONG *)(a),				\
+		((ROTATE(l,8)&0x00FF00FF)|(ROTATE((l&0x00FF00FF),24)))	\
+				)
+#else
+/* 6 instructions with rotate instruction, else 8 */
+#define REVERSE_FETCH32(a,l)	(				\
+		l=*(const HASH_LONG *)(a),			\
+		l=(((l>>8)&0x00FF00FF)|((l&0x00FF00FF)<<8)),	\
+		ROTATE(l,16)					\
+				)
+/*
+ * Originally the middle line started with l=(((l&0xFF00FF00)>>8)|...
+ * It's rewritten as above for two reasons:
+ *	- RISCs aren't good at long constants and have to explicitely
+ *	  compose 'em with several (well, usually 2) instructions in a
+ *	  register before performing the actual operation and (as you
+ *	  already realized:-) having same constant should inspire the
+ *	  compiler to permanently allocate the only register for it;
+ *	- most modern CPUs have two ALUs, but usually only one has
+ *	  circuitry for shifts:-( this minor tweak inspires compiler
+ *	  to schedule shift instructions in a better way...
+ *
+ *				<appro@fy.chalmers.se>
+ */
+#endif
+#endif
+
+#ifndef ROTATE
+#define ROTATE(a,n)     (((a)<<(n))|(((a)&0xffffffff)>>(32-(n))))
+#endif
+
+/*
+ * Make some obvious choices. E.g., HASH_BLOCK_DATA_ORDER_ALIGNED
+ * and HASH_BLOCK_HOST_ORDER ought to be the same if input data
+ * and host are of the same "endianess". It's possible to mask
+ * this with blank #define HASH_BLOCK_DATA_ORDER though...
+ *
+ *				<appro@fy.chalmers.se>
+ */
+#if defined(B_ENDIAN)
+#  if defined(DATA_ORDER_IS_BIG_ENDIAN)
+#    if !defined(HASH_BLOCK_DATA_ORDER_ALIGNED) && HASH_LONG_LOG2==2
+#      define HASH_BLOCK_DATA_ORDER_ALIGNED	HASH_BLOCK_HOST_ORDER
+#    endif
+#  endif
+#elif defined(L_ENDIAN)
+#  if defined(DATA_ORDER_IS_LITTLE_ENDIAN)
+#    if !defined(HASH_BLOCK_DATA_ORDER_ALIGNED) && HASH_LONG_LOG2==2
+#      define HASH_BLOCK_DATA_ORDER_ALIGNED	HASH_BLOCK_HOST_ORDER
+#    endif
+#  endif
+#endif
+
+#if !defined(HASH_BLOCK_DATA_ORDER_ALIGNED)
+#ifndef HASH_BLOCK_DATA_ORDER
+#error "HASH_BLOCK_DATA_ORDER must be defined!"
+#endif
+#endif
+
+#if defined(DATA_ORDER_IS_BIG_ENDIAN)
+
+#ifndef PEDANTIC
+# if defined(__GNUC__) && __GNUC__>=2 && !defined(OPENSSL_NO_ASM) && !defined(OPENSSL_NO_INLINE_ASM)
+#  if defined(__i386) || defined(__i386__) || defined(__x86_64) || defined(__x86_64__)
+    /*
+     * This gives ~30-40% performance improvement in SHA-256 compiled
+     * with gcc [on P4]. Well, first macro to be frank. We can pull
+     * this trick on x86* platforms only, because these CPUs can fetch
+     * unaligned data without raising an exception.
+     */
+#   define HOST_c2l(c,l)	({ unsigned int r=*((const unsigned int *)(c));	\
+				   asm ("bswapl %0":"=r"(r):"0"(r));	\
+				   (c)+=4; (l)=r;			})
+#   define HOST_l2c(l,c)	({ unsigned int r=(l);			\
+				   asm ("bswapl %0":"=r"(r):"0"(r));	\
+				   *((unsigned int *)(c))=r; (c)+=4; r;	})
+#  endif
+# endif
+#endif
+
+#ifndef HOST_c2l
+#define HOST_c2l(c,l)	(l =(((unsigned long)(*((c)++)))<<24),		\
+			 l|=(((unsigned long)(*((c)++)))<<16),		\
+			 l|=(((unsigned long)(*((c)++)))<< 8),		\
+			 l|=(((unsigned long)(*((c)++)))    ),		\
+			 l)
+#endif
+#define HOST_p_c2l(c,l,n)	{					\
+			switch (n) {					\
+			case 0: l =((unsigned long)(*((c)++)))<<24;	\
+			case 1: l|=((unsigned long)(*((c)++)))<<16;	\
+			case 2: l|=((unsigned long)(*((c)++)))<< 8;	\
+			case 3: l|=((unsigned long)(*((c)++)));		\
+				} }
+#define HOST_p_c2l_p(c,l,sc,len) {					\
+			switch (sc) {					\
+			case 0: l =((unsigned long)(*((c)++)))<<24;	\
+				if (--len == 0) break;			\
+			case 1: l|=((unsigned long)(*((c)++)))<<16;	\
+				if (--len == 0) break;			\
+			case 2: l|=((unsigned long)(*((c)++)))<< 8;	\
+				} }
+/* NOTE the pointer is not incremented at the end of this */
+#define HOST_c2l_p(c,l,n)	{					\
+			l=0; (c)+=n;					\
+			switch (n) {					\
+			case 3: l =((unsigned long)(*(--(c))))<< 8;	\
+			case 2: l|=((unsigned long)(*(--(c))))<<16;	\
+			case 1: l|=((unsigned long)(*(--(c))))<<24;	\
+				} }
+#ifndef HOST_l2c
+#define HOST_l2c(l,c)	(*((c)++)=(unsigned char)(((l)>>24)&0xff),	\
+			 *((c)++)=(unsigned char)(((l)>>16)&0xff),	\
+			 *((c)++)=(unsigned char)(((l)>> 8)&0xff),	\
+			 *((c)++)=(unsigned char)(((l)    )&0xff),	\
+			 l)
+#endif
+
+#elif defined(DATA_ORDER_IS_LITTLE_ENDIAN)
+
+#if defined(__i386) || defined(__i386__) || defined(__x86_64) || defined(__x86_64__)
+  /* See comment in DATA_ORDER_IS_BIG_ENDIAN section. */
+# define HOST_c2l(c,l)	((l)=*((const unsigned int *)(c)), (c)+=4, l)
+# define HOST_l2c(l,c)	(*((unsigned int *)(c))=(l), (c)+=4, l)
+#endif
+
+#ifndef HOST_c2l
+#define HOST_c2l(c,l)	(l =(((unsigned long)(*((c)++)))    ),		\
+			 l|=(((unsigned long)(*((c)++)))<< 8),		\
+			 l|=(((unsigned long)(*((c)++)))<<16),		\
+			 l|=(((unsigned long)(*((c)++)))<<24),		\
+			 l)
+#endif
+#define HOST_p_c2l(c,l,n)	{					\
+			switch (n) {					\
+			case 0: l =((unsigned long)(*((c)++)));		\
+			case 1: l|=((unsigned long)(*((c)++)))<< 8;	\
+			case 2: l|=((unsigned long)(*((c)++)))<<16;	\
+			case 3: l|=((unsigned long)(*((c)++)))<<24;	\
+				} }
+#define HOST_p_c2l_p(c,l,sc,len) {					\
+			switch (sc) {					\
+			case 0: l =((unsigned long)(*((c)++)));		\
+				if (--len == 0) break;			\
+			case 1: l|=((unsigned long)(*((c)++)))<< 8;	\
+				if (--len == 0) break;			\
+			case 2: l|=((unsigned long)(*((c)++)))<<16;	\
+				} }
+/* NOTE the pointer is not incremented at the end of this */
+#define HOST_c2l_p(c,l,n)	{					\
+			l=0; (c)+=n;					\
+			switch (n) {					\
+			case 3: l =((unsigned long)(*(--(c))))<<16;	\
+			case 2: l|=((unsigned long)(*(--(c))))<< 8;	\
+			case 1: l|=((unsigned long)(*(--(c))));		\
+				} }
+#ifndef HOST_l2c
+#define HOST_l2c(l,c)	(*((c)++)=(unsigned char)(((l)    )&0xff),	\
+			 *((c)++)=(unsigned char)(((l)>> 8)&0xff),	\
+			 *((c)++)=(unsigned char)(((l)>>16)&0xff),	\
+			 *((c)++)=(unsigned char)(((l)>>24)&0xff),	\
+			 l)
+#endif
+
+#endif
+
+/*
+ * Time for some action:-)
+ */
+
+int HASH_UPDATE (HASH_CTX *c, const void *data_, size_t len)
+	{
+	const unsigned char *data=data_;
+	register HASH_LONG * p;
+	register HASH_LONG l;
+	size_t sw,sc,ew,ec;
+
+	if (len==0) return 1;
+
+	l=(c->Nl+(((HASH_LONG)len)<<3))&0xffffffffUL;
+	/* 95-05-24 eay Fixed a bug with the overflow handling, thanks to
+	 * Wei Dai <weidai@eskimo.com> for pointing it out. */
+	if (l < c->Nl) /* overflow */
+		c->Nh++;
+	c->Nh+=(len>>29);	/* might cause compiler warning on 16-bit */
+	c->Nl=l;
+
+	if (c->num != 0)
+		{
+		p=c->data;
+		sw=c->num>>2;
+		sc=c->num&0x03;
+
+		if ((c->num+len) >= HASH_CBLOCK)
+			{
+			l=p[sw]; HOST_p_c2l(data,l,sc); p[sw++]=l;
+			for (; sw<HASH_LBLOCK; sw++)
+				{
+				HOST_c2l(data,l); p[sw]=l;
+				}
+			HASH_BLOCK_HOST_ORDER (c,p,1);
+			len-=(HASH_CBLOCK-c->num);
+			c->num=0;
+			/* drop through and do the rest */
+			}
+		else
+			{
+			c->num+=(unsigned int)len;
+			if ((sc+len) < 4) /* ugly, add char's to a word */
+				{
+				l=p[sw]; HOST_p_c2l_p(data,l,sc,len); p[sw]=l;
+				}
+			else
+				{
+				ew=(c->num>>2);
+				ec=(c->num&0x03);
+				if (sc)
+					l=p[sw];
+				HOST_p_c2l(data,l,sc);
+				p[sw++]=l;
+				for (; sw < ew; sw++)
+					{
+					HOST_c2l(data,l); p[sw]=l;
+					}
+				if (ec)
+					{
+					HOST_c2l_p(data,l,ec); p[sw]=l;
+					}
+				}
+			return 1;
+			}
+		}
+
+	sw=len/HASH_CBLOCK;
+	if (sw > 0)
+		{
+#if defined(HASH_BLOCK_DATA_ORDER_ALIGNED)
+		/*
+		 * Note that HASH_BLOCK_DATA_ORDER_ALIGNED gets defined
+		 * only if sizeof(HASH_LONG)==4.
+		 */
+		if ((((size_t)data)%4) == 0)
+			{
+			/* data is properly aligned so that we can cast it: */
+			HASH_BLOCK_DATA_ORDER_ALIGNED (c,(const HASH_LONG *)data,sw);
+			sw*=HASH_CBLOCK;
+			data+=sw;
+			len-=sw;
+			}
+		else
+#if !defined(HASH_BLOCK_DATA_ORDER)
+			while (sw--)
+				{
+				memcpy (p=c->data,data,HASH_CBLOCK);
+				HASH_BLOCK_DATA_ORDER_ALIGNED(c,p,1);
+				data+=HASH_CBLOCK;
+				len-=HASH_CBLOCK;
+				}
+#endif
+#endif
+#if defined(HASH_BLOCK_DATA_ORDER)
+			{
+			HASH_BLOCK_DATA_ORDER(c,data,sw);
+			sw*=HASH_CBLOCK;
+			data+=sw;
+			len-=sw;
+			}
+#endif
+		}
+
+	if (len!=0)
+		{
+		p = c->data;
+		c->num = len;
+		ew=len>>2;	/* words to copy */
+		ec=len&0x03;
+		for (; ew; ew--,p++)
+			{
+			HOST_c2l(data,l); *p=l;
+			}
+		HOST_c2l_p(data,l,ec);
+		*p=l;
+		}
+	return 1;
+	}
+
+
+void HASH_TRANSFORM (HASH_CTX *c, const unsigned char *data)
+	{
+#if defined(HASH_BLOCK_DATA_ORDER_ALIGNED)
+	if ((((size_t)data)%4) == 0)
+		/* data is properly aligned so that we can cast it: */
+		HASH_BLOCK_DATA_ORDER_ALIGNED (c,(const HASH_LONG *)data,1);
+	else
+#if !defined(HASH_BLOCK_DATA_ORDER)
+		{
+		memcpy (c->data,data,HASH_CBLOCK);
+		HASH_BLOCK_DATA_ORDER_ALIGNED (c,c->data,1);
+		}
+#endif
+#endif
+#if defined(HASH_BLOCK_DATA_ORDER)
+	HASH_BLOCK_DATA_ORDER (c,data,1);
+#endif
+	}
+
+
+int HASH_FINAL (unsigned char *md, HASH_CTX *c)
+	{
+	register HASH_LONG *p;
+	register unsigned long l;
+	register int i,j;
+	static const unsigned char end[4]={0x80,0x00,0x00,0x00};
+	const unsigned char *cp=end;
+
+	/* c->num should definitly have room for at least one more byte. */
+	p=c->data;
+	i=c->num>>2;
+	j=c->num&0x03;
+
+#if 0
+	/* purify often complains about the following line as an
+	 * Uninitialized Memory Read.  While this can be true, the
+	 * following p_c2l macro will reset l when that case is true.
+	 * This is because j&0x03 contains the number of 'valid' bytes
+	 * already in p[i].  If and only if j&0x03 == 0, the UMR will
+	 * occur but this is also the only time p_c2l will do
+	 * l= *(cp++) instead of l|= *(cp++)
+	 * Many thanks to Alex Tang <altitude@cic.net> for pickup this
+	 * 'potential bug' */
+#ifdef PURIFY
+	if (j==0) p[i]=0; /* Yeah, but that's not the way to fix it:-) */
+#endif
+	l=p[i];
+#else
+	l = (j==0) ? 0 : p[i];
+#endif
+	HOST_p_c2l(cp,l,j); p[i++]=l; /* i is the next 'undefined word' */
+
+	if (i>(HASH_LBLOCK-2)) /* save room for Nl and Nh */
+		{
+		if (i<HASH_LBLOCK) p[i]=0;
+		HASH_BLOCK_HOST_ORDER (c,p,1);
+		i=0;
+		}
+	for (; i<(HASH_LBLOCK-2); i++)
+		p[i]=0;
+
+#if   defined(DATA_ORDER_IS_BIG_ENDIAN)
+	p[HASH_LBLOCK-2]=c->Nh;
+	p[HASH_LBLOCK-1]=c->Nl;
+#elif defined(DATA_ORDER_IS_LITTLE_ENDIAN)
+	p[HASH_LBLOCK-2]=c->Nl;
+	p[HASH_LBLOCK-1]=c->Nh;
+#endif
+	HASH_BLOCK_HOST_ORDER (c,p,1);
+
+#ifndef HASH_MAKE_STRING
+#error "HASH_MAKE_STRING must be defined!"
+#else
+	HASH_MAKE_STRING(c,md);
+#endif
+
+	c->num=0;
+	/* clear stuff, HASH_BLOCK may be leaving some stuff on the stack
+	 * but I'm not worried :-)
+	OPENSSL_cleanse((void *)c,sizeof(HASH_CTX));
+	 */
+	return 1;
+	}
+
+#ifndef MD32_REG_T
+#define MD32_REG_T long
+/*
+ * This comment was originaly written for MD5, which is why it
+ * discusses A-D. But it basically applies to all 32-bit digests,
+ * which is why it was moved to common header file.
+ *
+ * In case you wonder why A-D are declared as long and not
+ * as MD5_LONG. Doing so results in slight performance
+ * boost on LP64 architectures. The catch is we don't
+ * really care if 32 MSBs of a 64-bit register get polluted
+ * with eventual overflows as we *save* only 32 LSBs in
+ * *either* case. Now declaring 'em long excuses the compiler
+ * from keeping 32 MSBs zeroed resulting in 13% performance
+ * improvement under SPARC Solaris7/64 and 5% under AlphaLinux.
+ * Well, to be honest it should say that this *prevents* 
+ * performance degradation.
+ *				<appro@fy.chalmers.se>
+ * Apparently there're LP64 compilers that generate better
+ * code if A-D are declared int. Most notably GCC-x86_64
+ * generates better code.
+ *				<appro@fy.chalmers.se>
+ */
+#endif
diff --git a/src/libutil/sha256.c b/src/libutil/sha256.c
new file mode 100644
index 0000000000..63ed0ba430
--- /dev/null
+++ b/src/libutil/sha256.c
@@ -0,0 +1,238 @@
+/* crypto/sha/sha256.c */
+/* ====================================================================
+ * Copyright (c) 2004 The OpenSSL Project.  All rights reserved
+ * according to the OpenSSL license [found in ./md32_common.h].
+ * ====================================================================
+ */
+
+#include <stdlib.h>
+#include <string.h>
+
+#include "sha256.h"
+
+int SHA224_Init (SHA256_CTX *c)
+	{
+	c->h[0]=0xc1059ed8UL;	c->h[1]=0x367cd507UL;
+	c->h[2]=0x3070dd17UL;	c->h[3]=0xf70e5939UL;
+	c->h[4]=0xffc00b31UL;	c->h[5]=0x68581511UL;
+	c->h[6]=0x64f98fa7UL;	c->h[7]=0xbefa4fa4UL;
+	c->Nl=0;	c->Nh=0;
+	c->num=0;	c->md_len=SHA224_DIGEST_LENGTH;
+	return 1;
+	}
+
+int SHA256_Init (SHA256_CTX *c)
+	{
+	c->h[0]=0x6a09e667UL;	c->h[1]=0xbb67ae85UL;
+	c->h[2]=0x3c6ef372UL;	c->h[3]=0xa54ff53aUL;
+	c->h[4]=0x510e527fUL;	c->h[5]=0x9b05688cUL;
+	c->h[6]=0x1f83d9abUL;	c->h[7]=0x5be0cd19UL;
+	c->Nl=0;	c->Nh=0;
+	c->num=0;	c->md_len=SHA256_DIGEST_LENGTH;
+	return 1;
+	}
+
+unsigned char *SHA224(const unsigned char *d, size_t n, unsigned char *md)
+	{
+	SHA256_CTX c;
+	static unsigned char m[SHA224_DIGEST_LENGTH];
+
+	if (md == NULL) md=m;
+	SHA224_Init(&c);
+	SHA256_Update(&c,d,n);
+	SHA256_Final(md,&c);
+	return(md);
+	}
+
+unsigned char *SHA256(const unsigned char *d, size_t n, unsigned char *md)
+	{
+	SHA256_CTX c;
+	static unsigned char m[SHA256_DIGEST_LENGTH];
+
+	if (md == NULL) md=m;
+	SHA256_Init(&c);
+	SHA256_Update(&c,d,n);
+	SHA256_Final(md,&c);
+	return(md);
+	}
+
+int SHA224_Update(SHA256_CTX *c, const void *data, size_t len)
+{   return SHA256_Update (c,data,len);   }
+int SHA224_Final (unsigned char *md, SHA256_CTX *c)
+{   return SHA256_Final (md,c);   }
+
+#define	DATA_ORDER_IS_BIG_ENDIAN
+
+#define	HASH_LONG		uint32_t
+#define	HASH_LONG_LOG2		2
+#define	HASH_CTX		SHA256_CTX
+#define	HASH_CBLOCK		SHA_CBLOCK
+#define	HASH_LBLOCK		SHA_LBLOCK
+/*
+ * Note that FIPS180-2 discusses "Truncation of the Hash Function Output."
+ * default: case below covers for it. It's not clear however if it's
+ * permitted to truncate to amount of bytes not divisible by 4. I bet not,
+ * but if it is, then default: case shall be extended. For reference.
+ * Idea behind separate cases for pre-defined lenghts is to let the
+ * compiler decide if it's appropriate to unroll small loops.
+ */
+#define	HASH_MAKE_STRING(c,s)	do {	\
+	unsigned long ll;		\
+	unsigned int  n;		\
+	switch ((c)->md_len)		\
+	{   case SHA224_DIGEST_LENGTH:	\
+		for (n=0;n<SHA224_DIGEST_LENGTH/4;n++)	\
+		{   ll=(c)->h[n]; HOST_l2c(ll,(s));   }	\
+		break;			\
+	    case SHA256_DIGEST_LENGTH:	\
+		for (n=0;n<SHA256_DIGEST_LENGTH/4;n++)	\
+		{   ll=(c)->h[n]; HOST_l2c(ll,(s));   }	\
+		break;			\
+	    default:			\
+		if ((c)->md_len > SHA256_DIGEST_LENGTH)	\
+		    return 0;				\
+		for (n=0;n<(c)->md_len/4;n++)		\
+		{   ll=(c)->h[n]; HOST_l2c(ll,(s));   }	\
+		break;			\
+	}				\
+	} while (0)
+
+#define	HASH_UPDATE		SHA256_Update
+#define	HASH_TRANSFORM		SHA256_Transform
+#define	HASH_FINAL		SHA256_Final
+#define	HASH_BLOCK_HOST_ORDER	sha256_block_host_order
+#define	HASH_BLOCK_DATA_ORDER	sha256_block_data_order
+void sha256_block_host_order (SHA256_CTX *ctx, const void *in, size_t num);
+void sha256_block_data_order (SHA256_CTX *ctx, const void *in, size_t num);
+
+#include "md32_common.h"
+
+static const uint32_t K256[64] = {
+	0x428a2f98UL,0x71374491UL,0xb5c0fbcfUL,0xe9b5dba5UL,
+	0x3956c25bUL,0x59f111f1UL,0x923f82a4UL,0xab1c5ed5UL,
+	0xd807aa98UL,0x12835b01UL,0x243185beUL,0x550c7dc3UL,
+	0x72be5d74UL,0x80deb1feUL,0x9bdc06a7UL,0xc19bf174UL,
+	0xe49b69c1UL,0xefbe4786UL,0x0fc19dc6UL,0x240ca1ccUL,
+	0x2de92c6fUL,0x4a7484aaUL,0x5cb0a9dcUL,0x76f988daUL,
+	0x983e5152UL,0xa831c66dUL,0xb00327c8UL,0xbf597fc7UL,
+	0xc6e00bf3UL,0xd5a79147UL,0x06ca6351UL,0x14292967UL,
+	0x27b70a85UL,0x2e1b2138UL,0x4d2c6dfcUL,0x53380d13UL,
+	0x650a7354UL,0x766a0abbUL,0x81c2c92eUL,0x92722c85UL,
+	0xa2bfe8a1UL,0xa81a664bUL,0xc24b8b70UL,0xc76c51a3UL,
+	0xd192e819UL,0xd6990624UL,0xf40e3585UL,0x106aa070UL,
+	0x19a4c116UL,0x1e376c08UL,0x2748774cUL,0x34b0bcb5UL,
+	0x391c0cb3UL,0x4ed8aa4aUL,0x5b9cca4fUL,0x682e6ff3UL,
+	0x748f82eeUL,0x78a5636fUL,0x84c87814UL,0x8cc70208UL,
+	0x90befffaUL,0xa4506cebUL,0xbef9a3f7UL,0xc67178f2UL };
+
+/*
+ * FIPS specification refers to right rotations, while our ROTATE macro
+ * is left one. This is why you might notice that rotation coefficients
+ * differ from those observed in FIPS document by 32-N...
+ */
+#define Sigma0(x)	(ROTATE((x),30) ^ ROTATE((x),19) ^ ROTATE((x),10))
+#define Sigma1(x)	(ROTATE((x),26) ^ ROTATE((x),21) ^ ROTATE((x),7))
+#define sigma0(x)	(ROTATE((x),25) ^ ROTATE((x),14) ^ ((x)>>3))
+#define sigma1(x)	(ROTATE((x),15) ^ ROTATE((x),13) ^ ((x)>>10))
+
+#define Ch(x,y,z)	(((x) & (y)) ^ ((~(x)) & (z)))
+#define Maj(x,y,z)	(((x) & (y)) ^ ((x) & (z)) ^ ((y) & (z)))
+
+#define	ROUND_00_15(i,a,b,c,d,e,f,g,h)		do {	\
+	T1 += h + Sigma1(e) + Ch(e,f,g) + K256[i];	\
+	h = Sigma0(a) + Maj(a,b,c);			\
+	d += T1;	h += T1;		} while (0)
+
+#define	ROUND_16_63(i,a,b,c,d,e,f,g,h,X)	do {	\
+	s0 = X[(i+1)&0x0f];	s0 = sigma0(s0);	\
+	s1 = X[(i+14)&0x0f];	s1 = sigma1(s1);	\
+	T1 = X[(i)&0x0f] += s0 + s1 + X[(i+9)&0x0f];	\
+	ROUND_00_15(i,a,b,c,d,e,f,g,h);		} while (0)
+
+static void sha256_block (SHA256_CTX *ctx, const void *in, size_t num, int host)
+	{
+	uint32_t a,b,c,d,e,f,g,h,s0,s1,T1;
+	uint32_t	X[16];
+	int i;
+	const unsigned char *data=in;
+
+			while (num--) {
+
+	a = ctx->h[0];	b = ctx->h[1];	c = ctx->h[2];	d = ctx->h[3];
+	e = ctx->h[4];	f = ctx->h[5];	g = ctx->h[6];	h = ctx->h[7];
+
+	if (host)
+		{
+		const uint32_t *W=(const uint32_t *)data;
+
+		T1 = X[0] = W[0];	ROUND_00_15(0,a,b,c,d,e,f,g,h);
+		T1 = X[1] = W[1];	ROUND_00_15(1,h,a,b,c,d,e,f,g);
+		T1 = X[2] = W[2];	ROUND_00_15(2,g,h,a,b,c,d,e,f);
+		T1 = X[3] = W[3];	ROUND_00_15(3,f,g,h,a,b,c,d,e);
+		T1 = X[4] = W[4];	ROUND_00_15(4,e,f,g,h,a,b,c,d);
+		T1 = X[5] = W[5];	ROUND_00_15(5,d,e,f,g,h,a,b,c);
+		T1 = X[6] = W[6];	ROUND_00_15(6,c,d,e,f,g,h,a,b);
+		T1 = X[7] = W[7];	ROUND_00_15(7,b,c,d,e,f,g,h,a);
+		T1 = X[8] = W[8];	ROUND_00_15(8,a,b,c,d,e,f,g,h);
+		T1 = X[9] = W[9];	ROUND_00_15(9,h,a,b,c,d,e,f,g);
+		T1 = X[10] = W[10];	ROUND_00_15(10,g,h,a,b,c,d,e,f);
+		T1 = X[11] = W[11];	ROUND_00_15(11,f,g,h,a,b,c,d,e);
+		T1 = X[12] = W[12];	ROUND_00_15(12,e,f,g,h,a,b,c,d);
+		T1 = X[13] = W[13];	ROUND_00_15(13,d,e,f,g,h,a,b,c);
+		T1 = X[14] = W[14];	ROUND_00_15(14,c,d,e,f,g,h,a,b);
+		T1 = X[15] = W[15];	ROUND_00_15(15,b,c,d,e,f,g,h,a);
+
+		data += SHA256_CBLOCK;
+		}
+	else
+		{
+		uint32_t l;
+
+		HOST_c2l(data,l); T1 = X[0] = l;  ROUND_00_15(0,a,b,c,d,e,f,g,h);
+		HOST_c2l(data,l); T1 = X[1] = l;  ROUND_00_15(1,h,a,b,c,d,e,f,g);
+		HOST_c2l(data,l); T1 = X[2] = l;  ROUND_00_15(2,g,h,a,b,c,d,e,f);
+		HOST_c2l(data,l); T1 = X[3] = l;  ROUND_00_15(3,f,g,h,a,b,c,d,e);
+		HOST_c2l(data,l); T1 = X[4] = l;  ROUND_00_15(4,e,f,g,h,a,b,c,d);
+		HOST_c2l(data,l); T1 = X[5] = l;  ROUND_00_15(5,d,e,f,g,h,a,b,c);
+		HOST_c2l(data,l); T1 = X[6] = l;  ROUND_00_15(6,c,d,e,f,g,h,a,b);
+		HOST_c2l(data,l); T1 = X[7] = l;  ROUND_00_15(7,b,c,d,e,f,g,h,a);
+		HOST_c2l(data,l); T1 = X[8] = l;  ROUND_00_15(8,a,b,c,d,e,f,g,h);
+		HOST_c2l(data,l); T1 = X[9] = l;  ROUND_00_15(9,h,a,b,c,d,e,f,g);
+		HOST_c2l(data,l); T1 = X[10] = l; ROUND_00_15(10,g,h,a,b,c,d,e,f);
+		HOST_c2l(data,l); T1 = X[11] = l; ROUND_00_15(11,f,g,h,a,b,c,d,e);
+		HOST_c2l(data,l); T1 = X[12] = l; ROUND_00_15(12,e,f,g,h,a,b,c,d);
+		HOST_c2l(data,l); T1 = X[13] = l; ROUND_00_15(13,d,e,f,g,h,a,b,c);
+		HOST_c2l(data,l); T1 = X[14] = l; ROUND_00_15(14,c,d,e,f,g,h,a,b);
+		HOST_c2l(data,l); T1 = X[15] = l; ROUND_00_15(15,b,c,d,e,f,g,h,a);
+		}
+
+	for (i=16;i<64;i+=8)
+		{
+		ROUND_16_63(i+0,a,b,c,d,e,f,g,h,X);
+		ROUND_16_63(i+1,h,a,b,c,d,e,f,g,X);
+		ROUND_16_63(i+2,g,h,a,b,c,d,e,f,X);
+		ROUND_16_63(i+3,f,g,h,a,b,c,d,e,X);
+		ROUND_16_63(i+4,e,f,g,h,a,b,c,d,X);
+		ROUND_16_63(i+5,d,e,f,g,h,a,b,c,X);
+		ROUND_16_63(i+6,c,d,e,f,g,h,a,b,X);
+		ROUND_16_63(i+7,b,c,d,e,f,g,h,a,X);
+		}
+
+	ctx->h[0] += a;	ctx->h[1] += b;	ctx->h[2] += c;	ctx->h[3] += d;
+	ctx->h[4] += e;	ctx->h[5] += f;	ctx->h[6] += g;	ctx->h[7] += h;
+
+			}
+	}
+
+/*
+ * Idea is to trade couple of cycles for some space. On IA-32 we save
+ * about 4K in "big footprint" case. In "small footprint" case any gain
+ * is appreciated:-)
+ */
+void HASH_BLOCK_HOST_ORDER (SHA256_CTX *ctx, const void *in, size_t num)
+{   sha256_block (ctx,in,num,1);   }
+
+void HASH_BLOCK_DATA_ORDER (SHA256_CTX *ctx, const void *in, size_t num)
+{   sha256_block (ctx,in,num,0);   }
+
+
diff --git a/src/libutil/sha256.h b/src/libutil/sha256.h
new file mode 100644
index 0000000000..39b301342f
--- /dev/null
+++ b/src/libutil/sha256.h
@@ -0,0 +1,35 @@
+#ifndef _SHA256_H
+#define _SHA256_H 1
+
+#include <stdint.h>
+
+#define SHA_LBLOCK	16
+#define SHA_CBLOCK	(SHA_LBLOCK*4)	/* SHA treats input data as a
+					 * contiguous array of 32 bit
+					 * wide big-endian values. */
+
+#define SHA256_CBLOCK	(SHA_LBLOCK*4)	/* SHA-256 treats input data as a
+					 * contiguous array of 32 bit
+					 * wide big-endian values. */
+#define SHA224_DIGEST_LENGTH	28
+#define SHA256_DIGEST_LENGTH	32
+
+typedef struct SHA256state_st
+	{
+	uint32_t h[8];
+	uint32_t Nl,Nh;
+	uint32_t data[SHA_LBLOCK];
+	unsigned int num,md_len;
+	} SHA256_CTX;
+
+int SHA224_Init(SHA256_CTX *c);
+int SHA224_Update(SHA256_CTX *c, const void *data, size_t len);
+int SHA224_Final(unsigned char *md, SHA256_CTX *c);
+unsigned char *SHA224(const unsigned char *d, size_t n,unsigned char *md);
+int SHA256_Init(SHA256_CTX *c);
+int SHA256_Update(SHA256_CTX *c, const void *data, size_t len);
+int SHA256_Final(unsigned char *md, SHA256_CTX *c);
+unsigned char *SHA256(const unsigned char *d, size_t n,unsigned char *md);
+void SHA256_Transform(SHA256_CTX *c, const unsigned char *data);
+
+#endif