v3.15.7

v3.15.6
v3.15.5
2025-09-17 23:44:27 +00:00 · 2021-03-08 22:44:44 -05:00 · 2021-02-12 15:16:53 -05:00 · 2020-12-21 13:25:33 -05:00 · 2020-12-15 13:15:02 -05:00 · 2020-12-10 18:23:49 -05:00
124 changed files with 7199 additions and 5270 deletions
--- a/145
+++ b/145
@@ -1,5 +1,9 @@
 Instructions for compiling cpuminer-opt for Windows.

+Thwaw intructions nay be out of date. Please consult the wiki for
+the latest:
+
+https://github.com/JayDDee/cpuminer-opt/wiki/Compiling-from-source

 Windows compilation using Visual Studio is not supported. Mingw64 is
 used on a Linux system (bare metal or virtual machine) to cross-compile
@@ -24,79 +28,76 @@ Refer to Linux compile instructions and install required packages.

 Additionally, install mingw-w64.

-sudo apt-get install mingw-w64
+sudo apt-get install mingw-w64 libz-mingw-w64-dev


 2. Create a local library directory for packages to be compiled in the next
   step. Suggested location is $HOME/usr/lib/

+$ mkdir $HOME/usr/lib
+
 3. Download and build other packages for mingw that don't have a mingw64
   version available in the repositories.

 Download the following source code packages from their respective and
 respected download locations, copy them to ~/usr/lib/ and uncompress them. 

-openssl
-curl
-gmp
+openssl: https://github.com/openssl/openssl/releases

-In most cases the latest vesrion is ok but it's safest to download
-the same major and minor version as included in your distribution.
+curl: https://github.com/curl/curl/releases

-Run the following commands or follow the supplied instructions.
-Do not run "make install" unless you are using ~/usr/lib, which isn't
-recommended.
+gmp: https://gmplib.org/download/gmp/

-Some instructions insist on running "make check". If make check fails
-it may still work, YMMV.
+In most cases the latest version is ok but it's safest to download the same major and minor version as included in your distribution. The following uses versions from Ubuntu 20.04. Change version numbers as required.

-You can speed up "make" by using all CPU cores available with "-j n" where
-n is the number of CPU threads you want to use.
+Run the following commands or follow the supplied instructions. Do not run "make install" unless you are using /usr/lib, which isn't recommended.
+
+Some instructions insist on running "make check". If make check fails it may still work, YMMV.
+
+You can speed up "make" by using all CPU cores available with "-j n" where n is the number of CPU threads you want to use.

 openssl:

-./Configure mingw64 shared --cross-compile-prefix=x86_64-w64-mingw32
-make
+$ ./Configure mingw64 shared --cross-compile-prefix=x86_64-w64-mingw32-
+$ make
+
+Make may fail with an ld error, just ensure libcrypto-1_1-x64.dll is created.

 curl:

-./configure --with-winssl --with-winidn --host=x86_64-w64-mingw32
-make
+$ ./configure --with-winssl --with-winidn --host=x86_64-w64-mingw32
+$ make

 gmp:

-./configure --host=x86_64-w64-mingw32 
-make
-
-
+$ ./configure --host=x86_64-w64-mingw32
+$ make

 4. Tweak the environment.

-This step is required everytime you login or the commands can be added to
-.bashrc.
+This step is required everytime you login or the commands can be added to .bashrc.

-Define some local variables to point to local library. 
+Define some local variables to point to local library.

-export LOCAL_LIB="$HOME/usr/lib"
+$ export LOCAL_LIB="$HOME/usr/lib"

-export LDFLAGS="-L$LOCAL_LIB/curl/lib/.libs -L$LOCAL_LIB/gmp/.libs -L$LOCAL_LIB/openssl"
+$ export LDFLAGS="-L$LOCAL_LIB/curl/lib/.libs -L$LOCAL_LIB/gmp/.libs -L$LOCAL_LIB/openssl"

-export CONFIGURE_ARGS="--with-curl=$LOCAL_LIB/curl --with-crypto=$LOCAL_LIB/openssl --host=x86_64-w64-mingw32"
+$ export CONFIGURE_ARGS="--with-curl=$LOCAL_LIB/curl --with-crypto=$LOCAL_LIB/openssl --host=x86_64-w64-mingw32"

-Create a release directory and copy some dll files previously built.
-This can be done outside of cpuminer-opt and only needs to be done once.
-If the release directory is in cpuminer-opt directory it needs to be
-recreated every a source package is decompressed.
+Adjust for gcc version:

-mkdir release
-cp /usr/x86_64-w64-mingw32/lib/zlib1.dll release/
-cp /usr/x86_64-w64-mingw32/lib/libwinpthread-1.dll release/
-cp /usr/lib/gcc/x86_64-w64-mingw32/7.3-win32/libstdc++-6.dll release/
-cp /usr/lib/gcc/x86_64-w64-mingw32/7.3-win32/libgcc_s_seh-1.dll release/
-cp $LOCAL_LIB/openssl/libcrypto-1_1-x64.dll release/
-cp $LOCAL_LIB/curl/lib/.libs/libcurl-4.dll release/
+$ export GCC_MINGW_LIB="/usr/lib/gcc/x86_64-w64-mingw32/9.3-win32"

+Create a release directory and copy some dll files previously built. This can be done outside of cpuminer-opt and only needs to be done once. If the release directory is in cpuminer-opt directory it needs to be recreated every time a source package is decompressed.

+$ mkdir release
+$ cp /usr/x86_64-w64-mingw32/lib/zlib1.dll release/
+$ cp /usr/x86_64-w64-mingw32/lib/libwinpthread-1.dll release/
+$ cp $GCC_MINGW_LIB/libstdc++-6.dll release/
+$ cp $GCC_MINGW_LIB/libgcc_s_seh-1.dll release/
+$ cp $LOCAL_LIB/openssl/libcrypto-1_1-x64.dll release/
+$ cp $LOCAL_LIB/curl/lib/.libs/libcurl-4.dll release/

 The following steps need to be done every time a new source package is
 opened.
@@ -110,13 +111,73 @@ https://github.com/JayDDee/cpuminer-opt/releases

 Decompress and change to the cpuminer-opt directory.

-
-
-6. Prepare to compile
+6. compile

 Create a link to the locally compiled version of gmp.h

-ln -s $LOCAL_LIB/gmp-version/gmp.h ./gmp.h
+$ ln -s $LOCAL_LIB/gmp-version/gmp.h ./gmp.h
+
+$ ./autogen.sh
+
+Configure the compiler for the CPU architecture of the host machine:
+
+CFLAGS="-O3 -march=native -Wall" ./configure $CONFIGURE_ARGS
+
+or cross compile for a specific CPU architecture:
+
+CFLAGS="-O3 -march=znver1 -Wall" ./configure $CONFIGURE_ARGS
+
+This will compile for AMD Ryzen.
+
+You can compile more generically for a set of specific CPU features if you know what features you want:
+
+CFLAGS="-O3 -maes -msse4.2 -Wall" ./configure $CONFIGURE_ARGS
+
+This will compile for an older CPU that does not have AVX.
+
+You can find several examples in README.txt
+
+If you have a CPU with more than 64 threads and Windows 7 or higher you can enable the CPU Groups feature by adding the following to CFLAGS:
+
+"-D_WIN32_WINNT=0x0601"
+
+Once you have run configure successfully run the compiler with n CPU threads:
+
+$ make -j n
+
+Copy cpuminer.exe to the release directory, compress and copy the release directory to a Windows system and run cpuminer.exe from the command line.
+
+Run cpuminer
+
+In a command windows change directories to the unzipped release folder. to get a list of all options:
+
+cpuminer.exe --help
+
+Command options are specific to where you mine. Refer to the pool's instructions on how to set them.
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+Create a link to the locally compiled version of gmp.h
+
+$ ln -s $LOCAL_LIB/gmp-version/gmp.h ./gmp.h

 Edit configure.ac to fix lipthread package name.

--- a/Makefile.am
+++ b/Makefile.am
@@ -85,6 +85,7 @@ cpuminer_SOURCES = \
  algo/groestl/aes_ni/hash-groestl.c \
  algo/groestl/aes_ni/hash-groestl256.c \
  algo/fugue/sph_fugue.c \
+  algo/fugue/fugue-aesni.c \
  algo/hamsi/sph_hamsi.c \
  algo/hamsi/hamsi-hash-4way.c \
  algo/haval/haval.c \
@@ -257,6 +258,7 @@ cpuminer_SOURCES = \
  algo/x16/hex.c \
  algo/x16/x21s-4way.c \
  algo/x16/x21s.c \
+  algo/x16/minotaur.c \
  algo/x17/x17-gate.c \
  algo/x17/x17.c \
  algo/x17/x17-4way.c \
--- a/README.md
+++ b/README.md
@@ -93,6 +93,7 @@ Supported Algorithms
                          lyra2z        
                          lyra2z330     Lyra2 330 rows, Zoin (ZOI)
                          m7m           Magi (XMG)
+                          minotaur      Ringcoin (RNG)
                          myr-gr        Myriad-Groestl
                          neoscrypt     NeoScrypt(128, 2, 1)
                          nist5         Nist5
--- a/README.txt
+++ b/README.txt
@@ -1,8 +1,12 @@
 This file is included in the Windows binary package. Compile instructions
 for Linux and Windows can be found in RELEASE_NOTES.

-cpuminer is a console program that is executed from a DOS command prompt.
-There is no GUI and no mouse support.
+This package is officially avalable only from:
+ https://github.com/JayDDee/cpuminer-opt
+No other sources should be trusted.
+
+cpuminer is a console program that is executed from a DOS or Powershell
+prompt. There is no GUI and no mouse support.

 Miner programs are often flagged as malware by antivirus programs. This is
 a false positive, they are flagged simply because they are cryptocurrency 
@@ -10,13 +14,13 @@ miners. The source code is open for anyone to inspect. If you don't trust
 the software, don't use it.

 Choose the exe that best matches you CPU's features or use trial and
-error to find the fastest one that doesn't crash. Pay attention to
+error to find the fastest one that works. Pay attention to
 the features listed at cpuminer startup to ensure you are mining at
 optimum speed using the best available features.

 Architecture names and compile options used are only provided for Intel
-Core series. Budget CPUs like Pentium and Celeron are often missing the
-latest features.
+Core series. Budget CPUs like Pentium and Celeron are often missing some
+features.

 AMD CPUs older than Piledriver, including Athlon x2 and Phenom II x4, are not
 supported by cpuminer-opt due to an incompatible implementation of SSE2 on
@@ -31,14 +35,34 @@ https://en.wikipedia.org/wiki/List_of_Intel_CPU_microarchitectures
 https://en.wikipedia.org/wiki/List_of_AMD_CPU_microarchitectures


-Exe name                Compile flags            Arch name
+Exe file name                Compile flags              Arch name

-cpuminer-sse2.exe      "-msse2"                  Core2, Nehalem   
-cpuminer-aes-sse42.exe "-march=westmere"         Westmere
-cpuminer-avx.exe       "-march=corei7-avx"       Sandybridge
-cpuminer-avx2.exe      "-march=core-avx2 -maes"  Haswell, Skylake, Coffeelake
-cpuminer-avx512.exe    "-march=skylake-avx512"   Skylake-X, Cascadelake-X
-cpuminer-zen           "-march=znver1"           AMD Ryzen, Threadripper
+cpuminer-sse2.exe            "-msse2"                   Core2, Nehalem   
+cpuminer-aes-sse42.exe       "-march=westmere"          Westmere
+cpuminer-avx.exe             "-march=corei7-avx"        Sandybridge, Ivybridge
+cpuminer-avx2.exe            "-march=core-avx2 -maes"   Haswell(1)
+cpuminer-avx512.exe          "-march=skylake-avx512"    Skylake-X, Cascadelake
+cpuminer-avx512-sha.exe      "-march=cascadelake -msha" Rocketlake(2)
+cpuminer-avx512-sha-vaes.exe "-march=icelake-client"    Icelake, Tigerlake(3)
+cpuminer-zen.exe             "-march=znver1"            AMD Zen1, Zen2
+cpuminer-zen3.exe            "-march=znver2 -mvaes"     Zen3(4)
+
+(1) Haswell includes Broadwell, Skylake, Kabylake, Coffeelake & Cometlake. 
+(2) Rocketlake build uses cascadelake+sha as a workaround until Rocketlake
+    compiler support is avalable.
+(3) Icelake & Tigerlake are only available on some laptops. Mining with a
+    laptop is not recommended.
+(4) Zen3 build uses zen2+vaes as a workaround until Zen3 compiler support is
+    available. Zen2 CPUs should use Zen1 build.
+
+Notes about included DLL files:
+
+Downloading DLL files from alternative sources presents an inherent
+security risk if their source is unknown. All DLL files included have
+been copied from the Ubuntu-20.04 installation or compiled by me from
+source code obtained from the author's official repository. The exact
+procedure is documented in the build instructions for Windows:
+https://github.com/JayDDee/cpuminer-opt/wiki/Compiling-from-source

 If you like this software feel free to donate:

--- a/113
+++ b/113
@@ -44,7 +44,7 @@ Please include the following information:
 1. CPU model, operating system, cpuminer-opt version (must be latest),
   binary file for Windows, changes to default build procedure for Linux.

-2. Exact comand line (except user and pw) and intial output showing
+2. Exact command line (except user and pw) and intial output showing
   the above requested info.

 3. Additional program output showing any error messages or other
@@ -65,6 +65,117 @@ If not what makes it happen or not happen?
 Change Log
 ----------

+v3.15.7
+
+Added accepted/stale/rejected percentage to summary log report.
+Added warning if share counters mismatch which could corrupt stats.
+Linux: CPU temperature reporting is more responsive to rising temperature.
+A few AVX2 & AVX512 tweaks.
+Removed some dead code and other cleanup.
+
+v3.15.6
+
+Implement keccak pre-hash optimization for x16* algos.
+Move conditional mining test to before get_new_work in miner thread.
+Add test for share reject reason when solo mining.
+Add support for floating point, as well as integer, "networkhasps" in
+RPC getmininginfo method.
+
+v3.15.5
+
+Fix stratum jobs lost if 2 jobs received in less than one second.
+ 
+
+v3.15.4
+
+Fixed yescryptr16 broken in v3.15.3.
+
+v3.15.3
+
+Yescrypt algos now use yespower v0.5, a little faster.
+New implementation of sha256 using SHA CPU extension.
+Replace Openssl with SPH for sha256 & sha512.
+AVX512 optimization for sha256t & sha256q.
+Faster sha256t, sha256q, x21s, x22i & x25x on CPUs with SHA without AVX512.
+AVX512+SHA build for Intel Rocketlake added to Windows binary package. 
+
+v3.15.2
+
+Zen3 AVX2+VAES optimization for x16*, x17, sonoa, xevan, x21s, x22i, x25x,
+allium.
+Zen3 (AVX2+SHA+VAES) build added to Windows binary package.
+
+v3.15.1
+
+Fix compile on AMD Zen3 CPUs with VAES.
+Force new work immediately after solving a block solo.
+
+
+v3.15.0
+
+Fugue optimized with AES, improves many sha3 algos.
+Minotaur algo optimized for all architectures.
+Fixed neoscrypt BUG log.
+ 
+v3.14.3
+
+#265: more mutex changes to reduce blocking with high thread count.
+
+#267: fixed hodl algo potential memory alignment issue,
+      add warning when thread count is not valid for mining hodl algo.
+
+v3.14.2
+
+The second line of the Share Accepted log is no longer displayed,
+new Xnonce log is added and other small log tweaks.
+
+#265: Cleanup use of mutex.
+
+v3.14.1
+
+GBT and getwork log changes:
+ fixed missing TTF in New Block log,
+ ntime no longer byte-swapped for display in New Work log,
+ fixed zero effective hash rate in Periodic Report log,
+ deleted "Current block is..." log.
+
+Renamed stratum "New Job" log to "New Work" to be consistent with the solo
+version of the log. Added more data to both versions.
+
+v3.14.0
+
+Changes to solo mining:
+  - segwit is supported by getblocktemplate,
+  - longpolling is not working and is disabled,
+  - Periodic Report log is output,
+  - New Block log includes TTF estimates,
+  - Stratum thread no longer created when using getwork or GBT.
+
+Fixed BUG log mining sha256d.
+
+v3.13.1.1
+
+Fixed Windows crash mining minotaur algo.
+
+Fixed GCC 10 compile again.
+Added -fno-common to testing to be consistent with GCC 10 default.
+
+v3.13.1
+
+Added minotaur algo for Ringcoin.
+
+v3.13.0.1
+
+Issue #262: Fixed xevan AVX2 invalid shares.
+
+v3.13.0
+
+Updated Windows binaries compiled with GCC 9. Included DLLs also updated.
+Icelake build (cpuminer-avx512-sha-vaes.exe) now included in Windows
+binaries package.
+
+No source code changes.
+
 v3.12.8.2

 Fixed x12 AVX2 rejects.
--- a/aclocal.m4
+++ b/aclocal.m4
@@ -1,6 +1,6 @@
-# generated automatically by aclocal 1.15.1 -*- Autoconf -*-
+# generated automatically by aclocal 1.16.1 -*- Autoconf -*-

-# Copyright (C) 1996-2017 Free Software Foundation, Inc.
+# Copyright (C) 1996-2018 Free Software Foundation, Inc.

 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -20,7 +20,7 @@ You have another version of autoconf.  It may work, but is not guaranteed to.
 If you have problems, you may need to regenerate the build system entirely.
 To do so, use the procedure documented by the package, typically 'autoreconf'.])])

-# Copyright (C) 2002-2017 Free Software Foundation, Inc.
+# Copyright (C) 2002-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -32,10 +32,10 @@ To do so, use the procedure documented by the package, typically 'autoreconf'.])
 # generated from the m4 files accompanying Automake X.Y.
 # (This private macro should not be called outside this file.)
 AC_DEFUN([AM_AUTOMAKE_VERSION],
-[am__api_version='1.15'
+[am__api_version='1.16'
 dnl Some users find AM_AUTOMAKE_VERSION and mistake it for a way to
 dnl require some minimum version.  Point them to the right macro.
-m4_if([$1], [1.15.1], [],
+m4_if([$1], [1.16.1], [],
      [AC_FATAL([Do not call $0, use AM_INIT_AUTOMAKE([$1]).])])dnl
 ])

@@ -51,14 +51,14 @@ m4_define([_AM_AUTOCONF_VERSION], [])
 # Call AM_AUTOMAKE_VERSION and AM_AUTOMAKE_VERSION so they can be traced.
 # This function is AC_REQUIREd by AM_INIT_AUTOMAKE.
 AC_DEFUN([AM_SET_CURRENT_AUTOMAKE_VERSION],
-[AM_AUTOMAKE_VERSION([1.15.1])dnl
+[AM_AUTOMAKE_VERSION([1.16.1])dnl
 m4_ifndef([AC_AUTOCONF_VERSION],
  [m4_copy([m4_PACKAGE_VERSION], [AC_AUTOCONF_VERSION])])dnl
 _AM_AUTOCONF_VERSION(m4_defn([AC_AUTOCONF_VERSION]))])

 # Figure out how to run the assembler.                      -*- Autoconf -*-

-# Copyright (C) 2001-2017 Free Software Foundation, Inc.
+# Copyright (C) 2001-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -78,7 +78,7 @@ _AM_IF_OPTION([no-dependencies],, [_AM_DEPENDENCIES([CCAS])])dnl

 # AM_AUX_DIR_EXPAND                                         -*- Autoconf -*-

-# Copyright (C) 2001-2017 Free Software Foundation, Inc.
+# Copyright (C) 2001-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -130,7 +130,7 @@ am_aux_dir=`cd "$ac_aux_dir" && pwd`

 # AM_CONDITIONAL                                            -*- Autoconf -*-

-# Copyright (C) 1997-2017 Free Software Foundation, Inc.
+# Copyright (C) 1997-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -161,7 +161,7 @@ AC_CONFIG_COMMANDS_PRE(
 Usually this means the macro was only invoked conditionally.]])
 fi])])

-# Copyright (C) 1999-2017 Free Software Foundation, Inc.
+# Copyright (C) 1999-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -352,13 +352,12 @@ _AM_SUBST_NOTMAKE([am__nodep])dnl

 # Generate code to set up dependency tracking.              -*- Autoconf -*-

-# Copyright (C) 1999-2017 Free Software Foundation, Inc.
+# Copyright (C) 1999-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
 # with or without modifications, as long as this notice is preserved.

-
 # _AM_OUTPUT_DEPENDENCY_COMMANDS
 # ------------------------------
 AC_DEFUN([_AM_OUTPUT_DEPENDENCY_COMMANDS],
@@ -366,49 +365,41 @@ AC_DEFUN([_AM_OUTPUT_DEPENDENCY_COMMANDS],
  # Older Autoconf quotes --file arguments for eval, but not when files
  # are listed without --file.  Let's play safe and only enable the eval
  # if we detect the quoting.
-  case $CONFIG_FILES in
-  *\'*) eval set x "$CONFIG_FILES" ;;
-  *)   set x $CONFIG_FILES ;;
-  esac
+  # TODO: see whether this extra hack can be removed once we start
+  # requiring Autoconf 2.70 or later.
+  AS_CASE([$CONFIG_FILES],
+          [*\'*], [eval set x "$CONFIG_FILES"],
+          [*], [set x $CONFIG_FILES])
  shift
-  for mf
+  # Used to flag and report bootstrapping failures.
+  am_rc=0
+  for am_mf
  do
    # Strip MF so we end up with the name of the file.
-    mf=`echo "$mf" | sed -e 's/:.*$//'`
-    # Check whether this is an Automake generated Makefile or not.
-    # We used to match only the files named 'Makefile.in', but
-    # some people rename them; so instead we look at the file content.
-    # Grep'ing the first line is not enough: some people post-process
-    # each Makefile.in and add a new line on top of each file to say so.
-    # Grep'ing the whole file is not good either: AIX grep has a line
+    am_mf=`AS_ECHO(["$am_mf"]) | sed -e 's/:.*$//'`
+    # Check whether this is an Automake generated Makefile which includes
+    # dependency-tracking related rules and includes.
+    # Grep'ing the whole file directly is not great: AIX grep has a line
    # limit of 2048, but all sed's we know have understand at least 4000.
-    if sed -n 's,^#.*generated by automake.*,X,p' "$mf" | grep X >/dev/null 2>&1; then
-      dirpart=`AS_DIRNAME("$mf")`
-    else
-      continue
-    fi
-    # Extract the definition of DEPDIR, am__include, and am__quote
-    # from the Makefile without running 'make'.
-    DEPDIR=`sed -n 's/^DEPDIR = //p' < "$mf"`
-    test -z "$DEPDIR" && continue
-    am__include=`sed -n 's/^am__include = //p' < "$mf"`
-    test -z "$am__include" && continue
-    am__quote=`sed -n 's/^am__quote = //p' < "$mf"`
-    # Find all dependency output files, they are included files with
-    # $(DEPDIR) in their names.  We invoke sed twice because it is the
-    # simplest approach to changing $(DEPDIR) to its actual value in the
-    # expansion.
-    for file in `sed -n "
-      s/^$am__include $am__quote\(.*(DEPDIR).*\)$am__quote"'$/\1/p' <"$mf" | \
-	 sed -e 's/\$(DEPDIR)/'"$DEPDIR"'/g'`; do
-      # Make sure the directory exists.
-      test -f "$dirpart/$file" && continue
-      fdir=`AS_DIRNAME(["$file"])`
-      AS_MKDIR_P([$dirpart/$fdir])
-      # echo "creating $dirpart/$file"
-      echo '# dummy' > "$dirpart/$file"
-    done
+    sed -n 's,^am--depfiles:.*,X,p' "$am_mf" | grep X >/dev/null 2>&1 \
+      || continue
+    am_dirpart=`AS_DIRNAME(["$am_mf"])`
+    am_filepart=`AS_BASENAME(["$am_mf"])`
+    AM_RUN_LOG([cd "$am_dirpart" \
+      && sed -e '/# am--include-marker/d' "$am_filepart" \
+        | $MAKE -f - am--depfiles]) || am_rc=$?
  done
+  if test $am_rc -ne 0; then
+    AC_MSG_FAILURE([Something went wrong bootstrapping makefile fragments
+    for automatic dependency tracking.  Try re-running configure with the
+    '--disable-dependency-tracking' option to at least be able to build
+    the package (albeit without support for automatic dependency tracking).])
+  fi
+  AS_UNSET([am_dirpart])
+  AS_UNSET([am_filepart])
+  AS_UNSET([am_mf])
+  AS_UNSET([am_rc])
+  rm -f conftest-deps.mk
 }
 ])# _AM_OUTPUT_DEPENDENCY_COMMANDS

@@ -417,18 +408,17 @@ AC_DEFUN([_AM_OUTPUT_DEPENDENCY_COMMANDS],
 # -----------------------------
 # This macro should only be invoked once -- use via AC_REQUIRE.
 #
-# This code is only required when automatic dependency tracking
-# is enabled.  FIXME.  This creates each '.P' file that we will
-# need in order to bootstrap the dependency handling code.
+# This code is only required when automatic dependency tracking is enabled.
+# This creates each '.Po' and '.Plo' makefile fragment that we'll need in
+# order to bootstrap the dependency handling code.
 AC_DEFUN([AM_OUTPUT_DEPENDENCY_COMMANDS],
 [AC_CONFIG_COMMANDS([depfiles],
     [test x"$AMDEP_TRUE" != x"" || _AM_OUTPUT_DEPENDENCY_COMMANDS],
-     [AMDEP_TRUE="$AMDEP_TRUE" ac_aux_dir="$ac_aux_dir"])
-])
+     [AMDEP_TRUE="$AMDEP_TRUE" MAKE="${MAKE-make}"])])

 # Do all the work for Automake.                             -*- Autoconf -*-

-# Copyright (C) 1996-2017 Free Software Foundation, Inc.
+# Copyright (C) 1996-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -515,8 +505,8 @@ AC_REQUIRE([AM_PROG_INSTALL_STRIP])dnl
 AC_REQUIRE([AC_PROG_MKDIR_P])dnl
 # For better backward compatibility.  To be removed once Automake 1.9.x
 # dies out for good.  For more background, see:
-# <http://lists.gnu.org/archive/html/automake/2012-07/msg00001.html>
-# <http://lists.gnu.org/archive/html/automake/2012-07/msg00014.html>
+# <https://lists.gnu.org/archive/html/automake/2012-07/msg00001.html>
+# <https://lists.gnu.org/archive/html/automake/2012-07/msg00014.html>
 AC_SUBST([mkdir_p], ['$(MKDIR_P)'])
 # We need awk for the "check" target (and possibly the TAP driver).  The
 # system "awk" is bad on some platforms.
@@ -583,7 +573,7 @@ END
 Aborting the configuration process, to ensure you take notice of the issue.

 You can download and install GNU coreutils to get an 'rm' implementation
-that behaves properly: <http://www.gnu.org/software/coreutils/>.
+that behaves properly: <https://www.gnu.org/software/coreutils/>.

 If you want to complete the configuration process using your problematic
 'rm' anyway, export the environment variable ACCEPT_INFERIOR_RM_PROGRAM
@@ -625,7 +615,7 @@ for _am_header in $config_headers :; do
 done
 echo "timestamp for $_am_arg" >`AS_DIRNAME(["$_am_arg"])`/stamp-h[]$_am_stamp_count])

-# Copyright (C) 2001-2017 Free Software Foundation, Inc.
+# Copyright (C) 2001-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -646,7 +636,7 @@ if test x"${install_sh+set}" != xset; then
 fi
 AC_SUBST([install_sh])])

-# Copyright (C) 2003-2017 Free Software Foundation, Inc.
+# Copyright (C) 2003-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -668,7 +658,7 @@ AC_SUBST([am__leading_dot])])
 # Add --enable-maintainer-mode option to configure.         -*- Autoconf -*-
 # From Jim Meyering

-# Copyright (C) 1996-2017 Free Software Foundation, Inc.
+# Copyright (C) 1996-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -703,7 +693,7 @@ AC_MSG_CHECKING([whether to enable maintainer-specific portions of Makefiles])

 # Check to see how 'make' treats includes.	            -*- Autoconf -*-

-# Copyright (C) 2001-2017 Free Software Foundation, Inc.
+# Copyright (C) 2001-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -711,49 +701,42 @@ AC_MSG_CHECKING([whether to enable maintainer-specific portions of Makefiles])

 # AM_MAKE_INCLUDE()
 # -----------------
-# Check to see how make treats includes.
+# Check whether make has an 'include' directive that can support all
+# the idioms we need for our automatic dependency tracking code.
 AC_DEFUN([AM_MAKE_INCLUDE],
-[am_make=${MAKE-make}
-cat > confinc << 'END'
+[AC_MSG_CHECKING([whether ${MAKE-make} supports the include directive])
+cat > confinc.mk << 'END'
 am__doit:
-	@echo this is the am__doit target
+	@echo this is the am__doit target >confinc.out
 .PHONY: am__doit
 END
-# If we don't find an include directive, just comment out the code.
-AC_MSG_CHECKING([for style of include used by $am_make])
 am__include="#"
 am__quote=
-_am_result=none
-# First try GNU make style include.
-echo "include confinc" > confmf
-# Ignore all kinds of additional output from 'make'.
-case `$am_make -s -f confmf 2> /dev/null` in #(
-*the\ am__doit\ target*)
-  am__include=include
-  am__quote=
-  _am_result=GNU
-  ;;
-esac
-# Now try BSD make style include.
-if test "$am__include" = "#"; then
-   echo '.include "confinc"' > confmf
-   case `$am_make -s -f confmf 2> /dev/null` in #(
-   *the\ am__doit\ target*)
-     am__include=.include
-     am__quote="\""
-     _am_result=BSD
-     ;;
-   esac
-fi
-AC_SUBST([am__include])
-AC_SUBST([am__quote])
-AC_MSG_RESULT([$_am_result])
-rm -f confinc confmf
-])
+# BSD make does it like this.
+echo '.include "confinc.mk" # ignored' > confmf.BSD
+# Other make implementations (GNU, Solaris 10, AIX) do it like this.
+echo 'include confinc.mk # ignored' > confmf.GNU
+_am_result=no
+for s in GNU BSD; do
+  AM_RUN_LOG([${MAKE-make} -f confmf.$s && cat confinc.out])
+  AS_CASE([$?:`cat confinc.out 2>/dev/null`],
+      ['0:this is the am__doit target'],
+      [AS_CASE([$s],
+          [BSD], [am__include='.include' am__quote='"'],
+          [am__include='include' am__quote=''])])
+  if test "$am__include" != "#"; then
+    _am_result="yes ($s style)"
+    break
+  fi
+done
+rm -f confinc.* confmf.*
+AC_MSG_RESULT([${_am_result}])
+AC_SUBST([am__include])])
+AC_SUBST([am__quote])])

 # Fake the existence of programs that GNU maintainers use.  -*- Autoconf -*-

-# Copyright (C) 1997-2017 Free Software Foundation, Inc.
+# Copyright (C) 1997-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -792,7 +775,7 @@ fi

 # Helper functions for option handling.                     -*- Autoconf -*-

-# Copyright (C) 2001-2017 Free Software Foundation, Inc.
+# Copyright (C) 2001-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -821,7 +804,7 @@ AC_DEFUN([_AM_SET_OPTIONS],
 AC_DEFUN([_AM_IF_OPTION],
 [m4_ifset(_AM_MANGLE_OPTION([$1]), [$2], [$3])])

-# Copyright (C) 1999-2017 Free Software Foundation, Inc.
+# Copyright (C) 1999-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -868,7 +851,7 @@ AC_LANG_POP([C])])
 # For backward compatibility.
 AC_DEFUN_ONCE([AM_PROG_CC_C_O], [AC_REQUIRE([AC_PROG_CC])])

-# Copyright (C) 2001-2017 Free Software Foundation, Inc.
+# Copyright (C) 2001-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -887,7 +870,7 @@ AC_DEFUN([AM_RUN_LOG],

 # Check to make sure that the build environment is sane.    -*- Autoconf -*-

-# Copyright (C) 1996-2017 Free Software Foundation, Inc.
+# Copyright (C) 1996-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -968,7 +951,7 @@ AC_CONFIG_COMMANDS_PRE(
 rm -f conftest.file
 ])

-# Copyright (C) 2009-2017 Free Software Foundation, Inc.
+# Copyright (C) 2009-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -1028,7 +1011,7 @@ AC_SUBST([AM_BACKSLASH])dnl
 _AM_SUBST_NOTMAKE([AM_BACKSLASH])dnl
 ])

-# Copyright (C) 2001-2017 Free Software Foundation, Inc.
+# Copyright (C) 2001-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -1056,7 +1039,7 @@ fi
 INSTALL_STRIP_PROGRAM="\$(install_sh) -c -s"
 AC_SUBST([INSTALL_STRIP_PROGRAM])])

-# Copyright (C) 2006-2017 Free Software Foundation, Inc.
+# Copyright (C) 2006-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
@@ -1075,7 +1058,7 @@ AC_DEFUN([AM_SUBST_NOTMAKE], [_AM_SUBST_NOTMAKE($@)])

 # Check how to create a tarball.                            -*- Autoconf -*-

-# Copyright (C) 2004-2017 Free Software Foundation, Inc.
+# Copyright (C) 2004-2018 Free Software Foundation, Inc.
 #
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
--- a/algo-gate-api.c
+++ b/algo-gate-api.c
@@ -90,13 +90,157 @@ void algo_not_implemented()
 }

 // default null functions
-
+// deprecated, use generic as default
 int null_scanhash()
 {
   applog(LOG_WARNING,"SWERR: undefined scanhash function in algo_gate");
   return 0;
 }

+// Default generic scanhash can be used in many cases.
+int scanhash_generic( struct work *work, uint32_t max_nonce,
+                      uint64_t *hashes_done, struct thr_info *mythr )
+{
+   uint32_t edata[20] __attribute__((aligned(64)));
+   uint32_t hash[8] __attribute__((aligned(64)));
+   uint32_t *pdata = work->data;
+   uint32_t *ptarget = work->target;
+   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 1;
+   uint32_t n = first_nonce;
+   const int thr_id = mythr->id;
+   const bool bench = opt_benchmark;
+
+   mm128_bswap32_80( edata, pdata );
+   do
+   {
+      edata[19] = n;
+      if ( likely( algo_gate.hash( hash, edata, thr_id ) ) )
+      if ( unlikely( valid_hash( hash, ptarget ) && !bench ) )
+      {
+         pdata[19] = bswap_32( n );
+         submit_solution( work, hash, mythr );
+      }
+      n++;
+   } while ( n < last_nonce && !work_restart[thr_id].restart );
+   *hashes_done = n - first_nonce;
+   pdata[19] = n;
+   return 0;
+}
+
+#if defined(__AVX2__)
+
+//int scanhash_4way_64_64( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr )
+
+//int scanhash_4way_64_640( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr )
+
+int scanhash_4way_64in_32out( struct work *work, uint32_t max_nonce,
+                      uint64_t *hashes_done, struct thr_info *mythr )
+{
+   uint32_t hash32[8*4] __attribute__ ((aligned (64)));
+   uint32_t vdata[20*4] __attribute__ ((aligned (64)));
+   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
+   uint32_t *hash32_d7 = &(hash32[ 7*4 ]);
+   uint32_t *pdata = work->data;
+   const uint32_t *ptarget = work->target;
+   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 4;
+   __m256i  *noncev = (__m256i*)vdata + 9;
+   uint32_t n = first_nonce;
+   const int thr_id = mythr->id;
+   const uint32_t targ32_d7 = ptarget[7];
+   const bool bench = opt_benchmark;
+
+   mm256_bswap32_intrlv80_4x64( vdata, pdata );
+   *noncev = mm256_intrlv_blend_32(
+                   _mm256_set_epi32( n+3, 0, n+2, 0, n+1, 0, n, 0 ), *noncev );
+   do
+   {
+      if ( likely( algo_gate.hash( hash32, vdata, thr_id ) ) )
+      for ( int lane = 0; lane < 4; lane++ )
+      if ( unlikely( hash32_d7[ lane ] <= targ32_d7 && !bench ) )
+      {
+         extr_lane_4x32( lane_hash, hash32, lane, 256 );
+         if ( valid_hash( lane_hash, ptarget ) )
+         {
+            pdata[19] = bswap_32( n + lane );
+            submit_solution( work, lane_hash, mythr );
+         }
+      }
+      *noncev = _mm256_add_epi32( *noncev,
+                                  m256_const1_64( 0x0000000400000000 ) );
+      n += 4;
+   } while ( likely( ( n <= last_nonce ) && !work_restart[thr_id].restart ) );
+   pdata[19] = n;
+   *hashes_done = n - first_nonce;
+   return 0;
+}
+
+//int scanhash_8way_32_32( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr )
+
+#endif
+
+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+
+//int scanhash_8way_64_64( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr )
+
+//int scanhash_8way_64_640( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr )
+
+int scanhash_8way_64in_32out( struct work *work, uint32_t max_nonce,
+                      uint64_t *hashes_done, struct thr_info *mythr )
+{
+   uint32_t hash32[8*8] __attribute__ ((aligned (128)));
+   uint32_t vdata[20*8] __attribute__ ((aligned (64)));
+   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
+   uint32_t *hash32_d7 = &(hash32[7*8]);
+   uint32_t *pdata = work->data;
+   const uint32_t *ptarget = work->target;
+   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 8;
+   __m512i  *noncev = (__m512i*)vdata + 9;
+   uint32_t n = first_nonce;
+   const int thr_id = mythr->id;
+   const uint32_t targ32_d7 = ptarget[7];
+   const bool bench = opt_benchmark;
+
+   mm512_bswap32_intrlv80_8x64( vdata, pdata );
+   *noncev = mm512_intrlv_blend_32(
+              _mm512_set_epi32( n+7, 0, n+6, 0, n+5, 0, n+4, 0,
+                                n+3, 0, n+2, 0, n+1, 0, n,   0 ), *noncev );
+   do
+   {
+      if ( likely( algo_gate.hash( hash32, vdata, thr_id ) ) )
+      for ( int lane = 0; lane < 8; lane++ )
+      if ( unlikely( ( hash32_d7[ lane ] <= targ32_d7 ) && !bench ) )
+      {
+         extr_lane_8x32( lane_hash, hash32, lane, 256 );
+         if ( likely( valid_hash( lane_hash, ptarget ) ) )
+         {
+            pdata[19] = bswap_32( n + lane );
+            submit_solution( work, lane_hash, mythr );
+         }
+      }
+      *noncev = _mm512_add_epi32( *noncev,
+                                  m512_const1_64( 0x0000000800000000 ) );
+      n += 8;
+   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
+   pdata[19] = n;
+   *hashes_done = n - first_nonce;
+   return 0;
+}
+
+//int scanhash_16way_32_32( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr )
+
+#endif
+
+
+
 int null_hash()
 {
   applog(LOG_WARNING,"SWERR: null_hash unsafe null function");
@@ -106,7 +250,7 @@ int null_hash()
 void init_algo_gate( algo_gate_t* gate )
 {
   gate->miner_thread_init       = (void*)&return_true;
-   gate->scanhash                = (void*)&null_scanhash;
+   gate->scanhash                = (void*)&scanhash_generic;
   gate->hash                    = (void*)&null_hash;
   gate->get_new_work            = (void*)&std_get_new_work;
   gate->work_decode             = (void*)&std_le_work_decode;
@@ -179,6 +323,7 @@ bool register_algo_gate( int algo, algo_gate_t *gate )
    case ALGO_LYRA2Z:        register_lyra2z_algo        ( gate ); break;
    case ALGO_LYRA2Z330:     register_lyra2z330_algo     ( gate ); break;
    case ALGO_M7M:           register_m7m_algo           ( gate ); break;
+    case ALGO_MINOTAUR:      register_minotaur_algo      ( gate ); break;
    case ALGO_MYR_GR:        register_myriad_algo        ( gate ); break;
    case ALGO_NEOSCRYPT:     register_neoscrypt_algo     ( gate ); break;
    case ALGO_NIST5:         register_nist5_algo         ( gate ); break;
@@ -225,11 +370,15 @@ bool register_algo_gate( int algo, algo_gate_t *gate )
    case ALGO_X22I:          register_x22i_algo          ( gate ); break;
    case ALGO_X25X:          register_x25x_algo          ( gate ); break;
    case ALGO_XEVAN:         register_xevan_algo         ( gate ); break;
-    case ALGO_YESCRYPT:      register_yescrypt_algo      ( gate ); break;
-    case ALGO_YESCRYPTR8:    register_yescryptr8_algo    ( gate ); break;
+    case ALGO_YESCRYPT:      register_yescrypt_05_algo   ( gate ); break;
+//    case ALGO_YESCRYPT:      register_yescrypt_algo      ( gate ); break;
+    case ALGO_YESCRYPTR8:    register_yescryptr8_05_algo ( gate ); break;
+//    case ALGO_YESCRYPTR8:    register_yescryptr8_algo    ( gate ); break;
    case ALGO_YESCRYPTR8G:   register_yescryptr8g_algo   ( gate ); break;
-    case ALGO_YESCRYPTR16:   register_yescryptr16_algo   ( gate ); break;
-    case ALGO_YESCRYPTR32:   register_yescryptr32_algo   ( gate ); break;
+    case ALGO_YESCRYPTR16:   register_yescryptr16_05_algo( gate ); break;
+//    case ALGO_YESCRYPTR16:   register_yescryptr16_algo   ( gate ); break;
+    case ALGO_YESCRYPTR32:   register_yescryptr32_05_algo( gate ); break;
+//    case ALGO_YESCRYPTR32:   register_yescryptr32_algo   ( gate ); break;
    case ALGO_YESPOWER:      register_yespower_algo      ( gate ); break;
    case ALGO_YESPOWERR16:   register_yespowerr16_algo   ( gate ); break;
    case ALGO_YESPOWER_B2B:  register_yespower_b2b_algo  ( gate ); break;
@@ -251,7 +400,6 @@ bool register_algo_gate( int algo, algo_gate_t *gate )
 // restore warnings
 #pragma GCC diagnostic pop

-// run the alternate hash function for a specific algo
 void exec_hash_function( int algo, void *output, const void *pdata )
 {
  algo_gate_t gate;   
@@ -271,7 +419,6 @@ void exec_hash_function( int algo, void *output, const void *pdata )
 const char* const algo_alias_map[][2] =
 {
 //   alias                proper
-  { "argon2d-crds",      "argon2d250"     },
  { "argon2d-dyn",       "argon2d500"     },
  { "argon2d-uis",       "argon2d4096"    },
  { "bcd",               "x13bcd"         },
--- a/algo-gate-api.h
+++ b/algo-gate-api.h
@@ -90,10 +90,11 @@ typedef  uint32_t set_t;
 #define AES_OPT          2  
 #define SSE42_OPT        4
 #define AVX_OPT          8   // Sandybridge
-#define AVX2_OPT      0x10   // Haswell
-#define SHA_OPT       0x20   // sha256 (Ryzen, Ice Lake)
-#define AVX512_OPT    0x40   // AVX512- F, VL, DQ, BW (Skylake-X)
-#define VAES_OPT      0x80   // VAES (Ice Lake)
+#define AVX2_OPT      0x10   // Haswell, Zen1
+#define SHA_OPT       0x20   // Zen1, Icelake (sha256)
+#define AVX512_OPT    0x40   // Skylake-X (AVX512[F,VL,DQ,BW])
+#define VAES_OPT      0x80   // Icelake (VAES & AVX512)
+#define VAES256_OPT   0x100  // Zen3 (VAES without AVX512)


 // return set containing all elements from sets a & b
@@ -110,11 +111,13 @@ inline bool set_excl ( set_t a, set_t b ) { return (a & b) == 0; }

 typedef struct
 {
-// mandatory function, must be overwritten
+// Mandatory functions, one of these is mandatory. If a generic scanhash
+// is used a custom target hash function must be registered, with a custom
+// scanhash the target hash function can be called directly and doesn't need
+// to be registered in the gate. 
 int ( *scanhash ) ( struct work*, uint32_t, uint64_t*, struct thr_info* );

-// Deprecated, will be removed
-int ( *hash )     ( void*, const void*, uint32_t ) ;
+int ( *hash )     ( void*, const void*, int );

 //optional, safe to use default in most cases

@@ -126,7 +129,7 @@ bool ( *miner_thread_init )     ( int );
 void ( *get_new_work )          ( struct work*, struct work*, int, uint32_t* );

 // Decode getwork blockheader
-bool ( *work_decode )           ( const json_t*, struct work* );
+bool ( *work_decode )           ( struct work* );

 // Extra getwork data
 void ( *decode_extra_data )     ( struct work*, uint64_t* );
@@ -159,7 +162,7 @@ bool ( *ready_to_mine )         ( struct work*, struct stratum_ctx*, int );
 bool ( *do_this_thread )        ( int );

 // After do_this_thread
-void ( *resync_threads )        ( struct work* );
+void ( *resync_threads )        ( int, struct work* );

 // No longer needed
 json_t* (*longpoll_rpc_call)      ( CURL*, int*, char* );
@@ -201,15 +204,61 @@ void four_way_not_tested();
 #define STD_WORK_DATA_SIZE 128
 #define STD_WORK_CMP_SIZE 76

-#define JR2_NONCE_INDEX 39  // 8 bit offset
+//#define JR2_NONCE_INDEX 39  // 8 bit offset

 // These indexes are only used with JSON RPC2 and are not gated.
-#define JR2_WORK_CMP_INDEX_2 43
-#define JR2_WORK_CMP_SIZE_2 33
+//#define JR2_WORK_CMP_INDEX_2 43
+//#define JR2_WORK_CMP_SIZE_2 33

-// allways returns failure
+// deprecated, use generic instead
 int null_scanhash();

+// Default generic, may be used in many cases.
+// N-way is more complicated, requires many different implementations
+// depending on architecture, input format, and output format.
+// Naming convention is scanhash_[N]way_[input format]in_[output format]out
+// N = number of lanes
+// input/output format:
+//    32: 32 bit interleaved parallel lanes
+//    64: 64 bit interleaved parallel lanes
+//    640: input only, not interleaved, contiguous serial 640 bit lanes.
+//    256: output only, not interleaved, contiguous serial 256 bit lanes.
+
+int scanhash_generic( struct work *work, uint32_t max_nonce,
+                      uint64_t *hashes_done, struct thr_info *mythr );
+
+#if defined(__AVX2__)
+
+//int scanhash_4way_64in_64out( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr );
+
+//int scanhash_4way_64in_256out( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr );
+
+int scanhash_4way_64in_32out( struct work *work, uint32_t max_nonce,
+                      uint64_t *hashes_done, struct thr_info *mythr );
+
+//int scanhash_8way_32in_32out( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr );
+
+#endif
+
+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+
+//int scanhash_8way_64in_64out( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr );
+
+//int scanhash_8way_64in_256out( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr );
+
+int scanhash_8way_64in_32out( struct work *work, uint32_t max_nonce,
+                      uint64_t *hashes_done, struct thr_info *mythr );
+
+//int scanhash_16way_32in_32out( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr );
+
+#endif
+
 // displays warning
 int null_hash    ();

@@ -221,8 +270,8 @@ void std_get_new_work( struct work *work, struct work *g_work, int thr_id,
 void sha256d_gen_merkle_root( char *merkle_root, struct stratum_ctx *sctx );
 void SHA256_gen_merkle_root ( char *merkle_root, struct stratum_ctx *sctx );

-bool std_le_work_decode( const json_t *val, struct work *work );
-bool std_be_work_decode( const json_t *val, struct work *work );
+bool std_le_work_decode( struct work *work );
+bool std_be_work_decode( struct work *work );

 bool std_le_submit_getwork_result( CURL *curl, struct work *work );
 bool std_be_submit_getwork_result( CURL *curl, struct work *work );
@@ -257,7 +306,7 @@ int std_get_work_data_size();
 // by calling the algo's register function.
 bool register_algo_gate( int algo, algo_gate_t *gate );

-// Called by algos toverride any default gate functions that are applicable
+// Called by algos to verride any default gate functions that are applicable
 // and do any other algo-specific initialization.
 // The register functions for all the algos can be declared here to reduce
 // compiler warnings but that's just more work for devs adding new algos.
--- a/algo/blake/decred-gate.c
+++ b/algo/blake/decred-gate.c
@@ -78,7 +78,6 @@ void decred_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
   uint32_t extraheader[32] = { 0 };
   int headersize = 0;
   uint32_t* extradata = (uint32_t*) sctx->xnonce1;
-   size_t t;
   int i;

   // getwork over stratum, getwork merkle + header passed in coinb1
@@ -87,9 +86,6 @@ void decred_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
                  sizeof(extraheader) );
   memcpy( extraheader, &sctx->job.coinbase[32], headersize );

-   // Increment extranonce2 
-   for ( t = 0; t < sctx->xnonce2_size && !( ++sctx->job.xnonce2[t] ); t++ );
-
   // Assemble block header 
   memset( g_work->data, 0, sizeof(g_work->data) );
   g_work->data[0] = le32dec( sctx->job.version );
--- a/algo/echo/aes_ni/hash.c
+++ b/algo/echo/aes_ni/hash.c
@@ -55,8 +55,8 @@ MYALIGN const unsigned int	mul2ipt[]		= {0x728efc00, 0x6894e61a, 0x3fc3b14d, 0x2

 #define ECHO_SUBBYTES(state, i, j) \
 	state[i][j] = _mm_aesenc_si128(state[i][j], k1);\
-	state[i][j] = _mm_aesenc_si128(state[i][j], M128(zero));\
-	k1 = _mm_add_epi32(k1, M128(const1))
+   k1 = _mm_add_epi32(k1, M128(const1));\
+	state[i][j] = _mm_aesenc_si128(state[i][j], M128(zero))

 #define ECHO_MIXBYTES(state1, state2, j, t1, t2, s2) \
 	s2 = _mm_add_epi8(state1[0][j], state1[0][j]);\
--- a/algo/echo/echo-hash-4way.c
+++ b/algo/echo/echo-hash-4way.c
@@ -1,5 +1,4 @@
-//#if 0
-#if defined(__VAES__) && defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+#if defined(__VAES__)

 #include "simd-utils.h"
 #include "echo-hash-4way.h"
@@ -11,18 +10,20 @@ static const unsigned int mul2ipt[] __attribute__ ((aligned (64))) =
   0xfd5ba600, 0x2a8c71d7, 0x1eb845e3, 0xc96f9234
 };
 */
-// do these need to be reversed?

-#define mul2mask \
-   _mm512_set4_epi32( 0, 0, 0, 0x00001b00 ) 
-//   _mm512_set4_epi32( 0x00001b00, 0, 0, 0 )  
+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)

-#define lsbmask    m512_const1_32( 0x01010101 ) 
+
+//#define mul2mask    m512_const2_64( 0, 0x00001b00 )
+//_mm512_set4_epi32( 0, 0, 0, 0x00001b00 ) 
+//_mm512_set4_epi32( 0x00001b00, 0, 0, 0 )  
+
+//#define lsbmask    m512_const1_32( 0x01010101 ) 

 #define ECHO_SUBBYTES( state, i, j ) \
 	state[i][j] = _mm512_aesenc_epi128( state[i][j], k1 ); \
-	state[i][j] = _mm512_aesenc_epi128( state[i][j], m512_zero ); \
-	k1 = _mm512_add_epi32( k1, m512_one_128 );
+   k1 = _mm512_add_epi32( k1, one ); \
+	state[i][j] = _mm512_aesenc_epi128( state[i][j], m512_zero );

 #define ECHO_MIXBYTES( state1, state2, j, t1, t2, s2 ) do \
 { \
@@ -30,87 +31,87 @@ static const unsigned int mul2ipt[] __attribute__ ((aligned (64))) =
   const int j2 = ( (j)+2 ) & 3; \
   const int j3 = ( (j)+3 ) & 3; \
   s2 = _mm512_add_epi8( state1[ 0 ] [j ], state1[ 0 ][ j ] ); \
-	t1 = _mm512_srli_epi16( state1[ 0 ][ j ], 7 ); \
-	t1 = _mm512_and_si512( t1, lsbmask );\
-	t2 = _mm512_shuffle_epi8( mul2mask, t1 ); \
-	s2 = _mm512_xor_si512( s2, t2 ); \
-	state2[ 0 ] [j ] = s2; \
-	state2[ 1 ] [j ] = state1[ 0 ][ j ]; \
-	state2[ 2 ] [j ] = state1[ 0 ][ j ]; \
-	state2[ 3 ] [j ] = _mm512_xor_si512( s2, state1[ 0 ][ j ] );\
-	s2 = _mm512_add_epi8( state1[ 1 ][ j1 ], state1[ 1 ][ j1 ] ); \
-	t1 = _mm512_srli_epi16( state1[ 1 ][ j1 ], 7 ); \
-	t1 = _mm512_and_si512( t1, lsbmask ); \
-	t2 = _mm512_shuffle_epi8( mul2mask, t1 ); \
-	s2 = _mm512_xor_si512( s2, t2 );\
-	state2[ 0 ][ j ] = _mm512_xor_si512( state2[ 0 ][ j ], \
-                            _mm512_xor_si512( s2, state1[ 1 ][ j1 ] ) ); \
-	state2[ 1 ][ j ] = _mm512_xor_si512( state2[ 1 ][ j ], s2 ); \
-	state2[ 2 ][ j ] = _mm512_xor_si512( state2[ 2 ][ j ], state1[ 1 ][ j1 ] ); \
-	state2[ 3 ][ j ] = _mm512_xor_si512( state2[ 3 ][ j ], state1[ 1 ][ j1 ] ); \
-	s2 = _mm512_add_epi8( state1[ 2 ][ j2 ], state1[ 2 ][ j2 ] ); \
-	t1 = _mm512_srli_epi16( state1[ 2 ][ j2 ], 7 ); \
-	t1 = _mm512_and_si512( t1, lsbmask ); \
-	t2 = _mm512_shuffle_epi8( mul2mask, t1 ); \
-	s2 = _mm512_xor_si512( s2, t2 ); \
-	state2[ 0 ][ j ] = _mm512_xor_si512( state2[ 0 ][ j ], state1[ 2 ][ j2 ] ); \
-	state2[ 1 ][ j ] = _mm512_xor_si512( state2[ 1 ][ j ], \
+   t1 = _mm512_srli_epi16( state1[ 0 ][ j ], 7 ); \
+   t1 = _mm512_and_si512( t1, lsbmask );\
+   t2 = _mm512_shuffle_epi8( mul2mask, t1 ); \
+   s2 = _mm512_xor_si512( s2, t2 ); \
+   state2[ 0 ] [j ] = s2; \
+   state2[ 1 ] [j ] = state1[ 0 ][ j ]; \
+   state2[ 2 ] [j ] = state1[ 0 ][ j ]; \
+   state2[ 3 ] [j ] = _mm512_xor_si512( s2, state1[ 0 ][ j ] );\
+   s2 = _mm512_add_epi8( state1[ 1 ][ j1 ], state1[ 1 ][ j1 ] ); \
+   t1 = _mm512_srli_epi16( state1[ 1 ][ j1 ], 7 ); \
+   t1 = _mm512_and_si512( t1, lsbmask ); \
+   t2 = _mm512_shuffle_epi8( mul2mask, t1 ); \
+   s2 = _mm512_xor_si512( s2, t2 );\
+   state2[ 0 ][ j ] = _mm512_xor_si512( state2[ 0 ][ j ], \
+                              _mm512_xor_si512( s2, state1[ 1 ][ j1 ] ) ); \
+   state2[ 1 ][ j ] = _mm512_xor_si512( state2[ 1 ][ j ], s2 ); \
+   state2[ 2 ][ j ] = _mm512_xor_si512( state2[ 2 ][ j ], state1[ 1 ][ j1 ] ); \
+   state2[ 3 ][ j ] = _mm512_xor_si512( state2[ 3 ][ j ], state1[ 1 ][ j1 ] ); \
+   s2 = _mm512_add_epi8( state1[ 2 ][ j2 ], state1[ 2 ][ j2 ] ); \
+   t1 = _mm512_srli_epi16( state1[ 2 ][ j2 ], 7 ); \
+   t1 = _mm512_and_si512( t1, lsbmask ); \
+   t2 = _mm512_shuffle_epi8( mul2mask, t1 ); \
+   s2 = _mm512_xor_si512( s2, t2 ); \
+   state2[ 0 ][ j ] = _mm512_xor_si512( state2[ 0 ][ j ], state1[ 2 ][ j2 ] ); \
+   state2[ 1 ][ j ] = _mm512_xor_si512( state2[ 1 ][ j ], \
                            _mm512_xor_si512( s2, state1[ 2 ][ j2 ] ) ); \
-	state2[ 2 ][ j ] = _mm512_xor_si512( state2[ 2 ][ j ], s2 ); \
-	state2[ 3 ][ j ] = _mm512_xor_si512( state2[ 3][ j ], state1[ 2 ][ j2 ] ); \
-	s2 = _mm512_add_epi8( state1[ 3 ][ j3 ], state1[ 3 ][ j3 ] ); \
-	t1 = _mm512_srli_epi16( state1[ 3 ][ j3 ], 7 ); \
-	t1 = _mm512_and_si512( t1, lsbmask ); \
-	t2 = _mm512_shuffle_epi8( mul2mask, t1 ); \
-	s2 = _mm512_xor_si512( s2, t2 ); \
-	state2[ 0 ][ j ] = _mm512_xor_si512( state2[ 0 ][ j ], state1[ 3 ][ j3 ] ); \
-	state2[ 1 ][ j ] = _mm512_xor_si512( state2[ 1 ][ j ], state1[ 3 ][ j3 ] ); \
-	state2[ 2 ][ j ] = _mm512_xor_si512( state2[ 2 ][ j ], \
+   state2[ 2 ][ j ] = _mm512_xor_si512( state2[ 2 ][ j ], s2 ); \
+   state2[ 3 ][ j ] = _mm512_xor_si512( state2[ 3][ j ], state1[ 2 ][ j2 ] ); \
+   s2 = _mm512_add_epi8( state1[ 3 ][ j3 ], state1[ 3 ][ j3 ] ); \
+   t1 = _mm512_srli_epi16( state1[ 3 ][ j3 ], 7 ); \
+   t1 = _mm512_and_si512( t1, lsbmask ); \
+   t2 = _mm512_shuffle_epi8( mul2mask, t1 ); \
+   s2 = _mm512_xor_si512( s2, t2 ); \
+   state2[ 0 ][ j ] = _mm512_xor_si512( state2[ 0 ][ j ], state1[ 3 ][ j3 ] ); \
+   state2[ 1 ][ j ] = _mm512_xor_si512( state2[ 1 ][ j ], state1[ 3 ][ j3 ] ); \
+   state2[ 2 ][ j ] = _mm512_xor_si512( state2[ 2 ][ j ], \
                            _mm512_xor_si512( s2, state1[ 3 ][ j3] ) ); \
-	state2[ 3 ][ j ] = _mm512_xor_si512( state2[ 3 ][ j ], s2 ); \
+   state2[ 3 ][ j ] = _mm512_xor_si512( state2[ 3 ][ j ], s2 ); \
 } while(0)

 #define ECHO_ROUND_UNROLL2 \
-	ECHO_SUBBYTES(_state, 0, 0);\
+   ECHO_SUBBYTES(_state, 0, 0);\
   ECHO_SUBBYTES(_state, 1, 0);\
-	ECHO_SUBBYTES(_state, 2, 0);\
-	ECHO_SUBBYTES(_state, 3, 0);\
-	ECHO_SUBBYTES(_state, 0, 1);\
-	ECHO_SUBBYTES(_state, 1, 1);\
-	ECHO_SUBBYTES(_state, 2, 1);\
-	ECHO_SUBBYTES(_state, 3, 1);\
-	ECHO_SUBBYTES(_state, 0, 2);\
-	ECHO_SUBBYTES(_state, 1, 2);\
-	ECHO_SUBBYTES(_state, 2, 2);\
-	ECHO_SUBBYTES(_state, 3, 2);\
-	ECHO_SUBBYTES(_state, 0, 3);\
-	ECHO_SUBBYTES(_state, 1, 3);\
-	ECHO_SUBBYTES(_state, 2, 3);\
-	ECHO_SUBBYTES(_state, 3, 3);\
-	ECHO_MIXBYTES(_state, _state2, 0, t1, t2, s2);\
-	ECHO_MIXBYTES(_state, _state2, 1, t1, t2, s2);\
-	ECHO_MIXBYTES(_state, _state2, 2, t1, t2, s2);\
-	ECHO_MIXBYTES(_state, _state2, 3, t1, t2, s2);\
-	ECHO_SUBBYTES(_state2, 0, 0);\
-	ECHO_SUBBYTES(_state2, 1, 0);\
-	ECHO_SUBBYTES(_state2, 2, 0);\
-	ECHO_SUBBYTES(_state2, 3, 0);\
-	ECHO_SUBBYTES(_state2, 0, 1);\
-	ECHO_SUBBYTES(_state2, 1, 1);\
-	ECHO_SUBBYTES(_state2, 2, 1);\
-	ECHO_SUBBYTES(_state2, 3, 1);\
-	ECHO_SUBBYTES(_state2, 0, 2);\
-	ECHO_SUBBYTES(_state2, 1, 2);\
-	ECHO_SUBBYTES(_state2, 2, 2);\
-	ECHO_SUBBYTES(_state2, 3, 2);\
-	ECHO_SUBBYTES(_state2, 0, 3);\
-	ECHO_SUBBYTES(_state2, 1, 3);\
-	ECHO_SUBBYTES(_state2, 2, 3);\
-	ECHO_SUBBYTES(_state2, 3, 3);\
-	ECHO_MIXBYTES(_state2, _state, 0, t1, t2, s2);\
-	ECHO_MIXBYTES(_state2, _state, 1, t1, t2, s2);\
-	ECHO_MIXBYTES(_state2, _state, 2, t1, t2, s2);\
-	ECHO_MIXBYTES(_state2, _state, 3, t1, t2, s2)
+   ECHO_SUBBYTES(_state, 2, 0);\
+   ECHO_SUBBYTES(_state, 3, 0);\
+   ECHO_SUBBYTES(_state, 0, 1);\
+   ECHO_SUBBYTES(_state, 1, 1);\
+   ECHO_SUBBYTES(_state, 2, 1);\
+   ECHO_SUBBYTES(_state, 3, 1);\
+   ECHO_SUBBYTES(_state, 0, 2);\
+   ECHO_SUBBYTES(_state, 1, 2);\
+   ECHO_SUBBYTES(_state, 2, 2);\
+   ECHO_SUBBYTES(_state, 3, 2);\
+   ECHO_SUBBYTES(_state, 0, 3);\
+   ECHO_SUBBYTES(_state, 1, 3);\
+   ECHO_SUBBYTES(_state, 2, 3);\
+   ECHO_SUBBYTES(_state, 3, 3);\
+   ECHO_MIXBYTES(_state, _state2, 0, t1, t2, s2);\
+   ECHO_MIXBYTES(_state, _state2, 1, t1, t2, s2);\
+   ECHO_MIXBYTES(_state, _state2, 2, t1, t2, s2);\
+   ECHO_MIXBYTES(_state, _state2, 3, t1, t2, s2);\
+   ECHO_SUBBYTES(_state2, 0, 0);\
+   ECHO_SUBBYTES(_state2, 1, 0);\
+   ECHO_SUBBYTES(_state2, 2, 0);\
+   ECHO_SUBBYTES(_state2, 3, 0);\
+   ECHO_SUBBYTES(_state2, 0, 1);\
+   ECHO_SUBBYTES(_state2, 1, 1);\
+   ECHO_SUBBYTES(_state2, 2, 1);\
+   ECHO_SUBBYTES(_state2, 3, 1);\
+   ECHO_SUBBYTES(_state2, 0, 2);\
+   ECHO_SUBBYTES(_state2, 1, 2);\
+   ECHO_SUBBYTES(_state2, 2, 2);\
+   ECHO_SUBBYTES(_state2, 3, 2);\
+   ECHO_SUBBYTES(_state2, 0, 3);\
+   ECHO_SUBBYTES(_state2, 1, 3);\
+   ECHO_SUBBYTES(_state2, 2, 3);\
+   ECHO_SUBBYTES(_state2, 3, 3);\
+   ECHO_MIXBYTES(_state2, _state, 0, t1, t2, s2);\
+   ECHO_MIXBYTES(_state2, _state, 1, t1, t2, s2);\
+   ECHO_MIXBYTES(_state2, _state, 2, t1, t2, s2);\
+   ECHO_MIXBYTES(_state2, _state, 3, t1, t2, s2)

 #define SAVESTATE(dst, src)\
 	dst[0][0] = src[0][0];\
@@ -137,6 +138,9 @@ void echo_4way_compress( echo_4way_context *ctx, const __m512i *pmsg,
  unsigned int r, b, i, j;
  __m512i t1, t2, s2, k1;
  __m512i _state[4][4], _state2[4][4], _statebackup[4][4]; 
+  __m512i one = m512_one_128;
+  __m512i mul2mask = m512_const2_64( 0, 0x00001b00 );
+  __m512i lsbmask  = m512_const1_32( 0x01010101 ); 

  _state[ 0 ][ 0 ] = ctx->state[ 0 ][ 0 ];
  _state[ 0 ][ 1 ] = ctx->state[ 0 ][ 1 ];
@@ -224,43 +228,43 @@ void echo_4way_compress( echo_4way_context *ctx, const __m512i *pmsg,

 int echo_4way_init( echo_4way_context *ctx, int nHashSize )
 {
-	int i, j;
+   int i, j;

   ctx->k = m512_zero; 
-	ctx->processed_bits = 0;
-	ctx->uBufferBytes = 0;
+   ctx->processed_bits = 0;
+   ctx->uBufferBytes = 0;

-	switch( nHashSize )
-	{
-		case 256:
-			ctx->uHashSize = 256;
-			ctx->uBlockLength = 192;
-			ctx->uRounds = 8;
-			ctx->hashsize = _mm512_set4_epi32( 0, 0, 0, 0x100 );
-			ctx->const1536 = _mm512_set4_epi32( 0, 0, 0, 0x600 );
-			break;
+   switch( nHashSize )
+   {
+	case 256:
+		ctx->uHashSize = 256;
+		ctx->uBlockLength = 192;
+		ctx->uRounds = 8;
+		ctx->hashsize = m512_const2_64( 0, 0x100 );
+		ctx->const1536 = m512_const2_64( 0, 0x600 );
+		break;

-		case 512:
-			ctx->uHashSize = 512;
-			ctx->uBlockLength = 128;
-			ctx->uRounds = 10;
-			ctx->hashsize = _mm512_set4_epi32( 0, 0, 0, 0x200 );
-			ctx->const1536 = _mm512_set4_epi32( 0, 0, 0, 0x400);
-			break;
+	case 512:
+		ctx->uHashSize = 512;
+		ctx->uBlockLength = 128;
+		ctx->uRounds = 10;
+		ctx->hashsize = m512_const2_64( 0, 0x200 );
+		ctx->const1536 = m512_const2_64( 0, 0x400);
+		break;

-		default:
-			return 1;
-	}
+	default:
+        	return 1;
+   }

-	for( i = 0; i < 4; i++ )
-		for( j = 0; j < nHashSize / 256; j++ )
-			ctx->state[ i ][ j ] = ctx->hashsize;
+   for( i = 0; i < 4; i++ )
+	for( j = 0; j < nHashSize / 256; j++ )
+		ctx->state[ i ][ j ] = ctx->hashsize;

-	for( i = 0; i < 4; i++ )
-		for( j = nHashSize / 256; j < 4; j++ )
-			ctx->state[ i ][ j ] = m512_zero;
+   for( i = 0; i < 4; i++ )
+	for( j = nHashSize / 256; j < 4; j++ )
+		ctx->state[ i ][ j ] = m512_zero;

-	return 0;
+   return 0;
 }

 int echo_4way_update_close( echo_4way_context *state, void *hashval,
@@ -285,17 +289,13 @@ int echo_4way_update_close( echo_4way_context *state, void *hashval,
      vlen = databitlen / 128;  // * 4 lanes / 128 bits per lane
      memcpy_512( state->buffer, data, vlen );
      state->processed_bits += (unsigned int)( databitlen );
-      remainingbits = _mm512_set4_epi32( 0, 0, 0, databitlen );
-
+      remainingbits = m512_const2_64( 0, (uint64_t)databitlen );
   }

-   state->buffer[ vlen ] = _mm512_set4_epi32( 0, 0, 0, 0x80 );
+   state->buffer[ vlen ] = m512_const2_64( 0, 0x80 );
   memset_zero_512( state->buffer + vlen + 1, vblen - vlen - 2 );
-   state->buffer[ vblen-2 ] =
-                _mm512_set4_epi32( (uint32_t)state->uHashSize << 16, 0, 0, 0 );
-   state->buffer[ vblen-1 ] =
-                   _mm512_set4_epi64( 0, state->processed_bits,
-                                      0, state->processed_bits );  
+   state->buffer[ vblen-2 ] = m512_const2_64( (uint64_t)state->uHashSize << 48, 0 );
+   state->buffer[ vblen-1 ] = m512_const2_64( 0, state->processed_bits);

   state->k = _mm512_add_epi64( state->k, remainingbits );
   state->k = _mm512_sub_epi64( state->k, state->const1536 );
@@ -328,16 +328,16 @@ int echo_4way_full( echo_4way_context *ctx, void *hashval, int nHashSize,
         ctx->uHashSize = 256;
         ctx->uBlockLength = 192;
         ctx->uRounds = 8;
-         ctx->hashsize = _mm512_set4_epi32( 0, 0, 0, 0x100 );
-         ctx->const1536 = _mm512_set4_epi32( 0, 0, 0, 0x600 );
+         ctx->hashsize = m512_const2_64( 0, 0x100 );
+         ctx->const1536 = m512_const2_64( 0, 0x600 );
         break;

      case 512:
         ctx->uHashSize = 512;
         ctx->uBlockLength = 128;
         ctx->uRounds = 10;
-         ctx->hashsize = _mm512_set4_epi32( 0, 0, 0, 0x200 );
-         ctx->const1536 = _mm512_set4_epi32( 0, 0, 0, 0x400);
+         ctx->hashsize = m512_const2_64( 0, 0x200 );
+         ctx->const1536 = m512_const2_64( 0, 0x400 );
         break;

      default:
@@ -372,17 +372,14 @@ int echo_4way_full( echo_4way_context *ctx, void *hashval, int nHashSize,
      vlen = databitlen / 128;  // * 4 lanes / 128 bits per lane
      memcpy_512( ctx->buffer, data, vlen );
      ctx->processed_bits += (unsigned int)( databitlen );
-      remainingbits = _mm512_set4_epi32( 0, 0, 0, databitlen );
-
+      remainingbits = m512_const2_64( 0, databitlen );
   }

-   ctx->buffer[ vlen ] = _mm512_set4_epi32( 0, 0, 0, 0x80 );
+   ctx->buffer[ vlen ] = m512_const2_64( 0, 0x80 );
   memset_zero_512( ctx->buffer + vlen + 1, vblen - vlen - 2 );
   ctx->buffer[ vblen-2 ] =
-                _mm512_set4_epi32( (uint32_t)ctx->uHashSize << 16, 0, 0, 0 );
-   ctx->buffer[ vblen-1 ] =
-                   _mm512_set4_epi64( 0, ctx->processed_bits,
-                                      0, ctx->processed_bits );
+                     m512_const2_64( (uint64_t)ctx->uHashSize << 48, 0 );
+   ctx->buffer[ vblen-1 ] = m512_const2_64( 0, ctx->processed_bits);

   ctx->k = _mm512_add_epi64( ctx->k, remainingbits );
   ctx->k = _mm512_sub_epi64( ctx->k, ctx->const1536 );
@@ -400,5 +397,380 @@ int echo_4way_full( echo_4way_context *ctx, void *hashval, int nHashSize,
   return 0;
 }

+#endif  // AVX512

-#endif
+// AVX2 + VAES
+
+#define mul2mask_2way   m256_const2_64( 0, 0x0000000000001b00 ) 
+
+#define lsbmask_2way    m256_const1_32( 0x01010101 ) 
+
+#define ECHO_SUBBYTES_2WAY( state, i, j ) \
+        state[i][j] = _mm256_aesenc_epi128( state[i][j], k1 ); \
+        k1 = _mm256_add_epi32( k1, m256_one_128 ); \
+        state[i][j] = _mm256_aesenc_epi128( state[i][j], m256_zero ); \
+
+#define ECHO_MIXBYTES_2WAY( state1, state2, j, t1, t2, s2 ) do \
+{ \
+   const int j1 = ( (j)+1 ) & 3; \
+   const int j2 = ( (j)+2 ) & 3; \
+   const int j3 = ( (j)+3 ) & 3; \
+   s2 = _mm256_add_epi8( state1[ 0 ] [j ], state1[ 0 ][ j ] ); \
+   t1 = _mm256_srli_epi16( state1[ 0 ][ j ], 7 ); \
+   t1 = _mm256_and_si256( t1, lsbmask_2way );\
+   t2 = _mm256_shuffle_epi8( mul2mask_2way, t1 ); \
+   s2 = _mm256_xor_si256( s2, t2 ); \
+   state2[ 0 ] [j ] = s2; \
+   state2[ 1 ] [j ] = state1[ 0 ][ j ]; \
+   state2[ 2 ] [j ] = state1[ 0 ][ j ]; \
+   state2[ 3 ] [j ] = _mm256_xor_si256( s2, state1[ 0 ][ j ] );\
+   s2 = _mm256_add_epi8( state1[ 1 ][ j1 ], state1[ 1 ][ j1 ] ); \
+   t1 = _mm256_srli_epi16( state1[ 1 ][ j1 ], 7 ); \
+   t1 = _mm256_and_si256( t1, lsbmask_2way ); \
+   t2 = _mm256_shuffle_epi8( mul2mask_2way, t1 ); \
+   s2 = _mm256_xor_si256( s2, t2 );\
+   state2[ 0 ][ j ] = _mm256_xor_si256( state2[ 0 ][ j ], \
+                              _mm256_xor_si256( s2, state1[ 1 ][ j1 ] ) ); \
+   state2[ 1 ][ j ] = _mm256_xor_si256( state2[ 1 ][ j ], s2 ); \
+   state2[ 2 ][ j ] = _mm256_xor_si256( state2[ 2 ][ j ], state1[ 1 ][ j1 ] ); \
+   state2[ 3 ][ j ] = _mm256_xor_si256( state2[ 3 ][ j ], state1[ 1 ][ j1 ] ); \
+   s2 = _mm256_add_epi8( state1[ 2 ][ j2 ], state1[ 2 ][ j2 ] ); \
+   t1 = _mm256_srli_epi16( state1[ 2 ][ j2 ], 7 ); \
+   t1 = _mm256_and_si256( t1, lsbmask_2way ); \
+   t2 = _mm256_shuffle_epi8( mul2mask_2way, t1 ); \
+   s2 = _mm256_xor_si256( s2, t2 ); \
+   state2[ 0 ][ j ] = _mm256_xor_si256( state2[ 0 ][ j ], state1[ 2 ][ j2 ] ); \
+   state2[ 1 ][ j ] = _mm256_xor_si256( state2[ 1 ][ j ], \
+                            _mm256_xor_si256( s2, state1[ 2 ][ j2 ] ) ); \
+   state2[ 2 ][ j ] = _mm256_xor_si256( state2[ 2 ][ j ], s2 ); \
+   state2[ 3 ][ j ] = _mm256_xor_si256( state2[ 3][ j ], state1[ 2 ][ j2 ] ); \
+   s2 = _mm256_add_epi8( state1[ 3 ][ j3 ], state1[ 3 ][ j3 ] ); \
+   t1 = _mm256_srli_epi16( state1[ 3 ][ j3 ], 7 ); \
+   t1 = _mm256_and_si256( t1, lsbmask_2way ); \
+   t2 = _mm256_shuffle_epi8( mul2mask_2way, t1 ); \
+   s2 = _mm256_xor_si256( s2, t2 ); \
+   state2[ 0 ][ j ] = _mm256_xor_si256( state2[ 0 ][ j ], state1[ 3 ][ j3 ] ); \
+   state2[ 1 ][ j ] = _mm256_xor_si256( state2[ 1 ][ j ], state1[ 3 ][ j3 ] ); \
+   state2[ 2 ][ j ] = _mm256_xor_si256( state2[ 2 ][ j ], \
+                            _mm256_xor_si256( s2, state1[ 3 ][ j3] ) ); \
+   state2[ 3 ][ j ] = _mm256_xor_si256( state2[ 3 ][ j ], s2 ); \
+} while(0)
+
+#define ECHO_ROUND_UNROLL2_2WAY \
+   ECHO_SUBBYTES_2WAY(_state, 0, 0);\
+   ECHO_SUBBYTES_2WAY(_state, 1, 0);\
+   ECHO_SUBBYTES_2WAY(_state, 2, 0);\
+   ECHO_SUBBYTES_2WAY(_state, 3, 0);\
+   ECHO_SUBBYTES_2WAY(_state, 0, 1);\
+   ECHO_SUBBYTES_2WAY(_state, 1, 1);\
+   ECHO_SUBBYTES_2WAY(_state, 2, 1);\
+   ECHO_SUBBYTES_2WAY(_state, 3, 1);\
+   ECHO_SUBBYTES_2WAY(_state, 0, 2);\
+   ECHO_SUBBYTES_2WAY(_state, 1, 2);\
+   ECHO_SUBBYTES_2WAY(_state, 2, 2);\
+   ECHO_SUBBYTES_2WAY(_state, 3, 2);\
+   ECHO_SUBBYTES_2WAY(_state, 0, 3);\
+   ECHO_SUBBYTES_2WAY(_state, 1, 3);\
+   ECHO_SUBBYTES_2WAY(_state, 2, 3);\
+   ECHO_SUBBYTES_2WAY(_state, 3, 3);\
+   ECHO_MIXBYTES_2WAY(_state, _state2, 0, t1, t2, s2);\
+   ECHO_MIXBYTES_2WAY(_state, _state2, 1, t1, t2, s2);\
+   ECHO_MIXBYTES_2WAY(_state, _state2, 2, t1, t2, s2);\
+   ECHO_MIXBYTES_2WAY(_state, _state2, 3, t1, t2, s2);\
+   ECHO_SUBBYTES_2WAY(_state2, 0, 0);\
+   ECHO_SUBBYTES_2WAY(_state2, 1, 0);\
+   ECHO_SUBBYTES_2WAY(_state2, 2, 0);\
+   ECHO_SUBBYTES_2WAY(_state2, 3, 0);\
+   ECHO_SUBBYTES_2WAY(_state2, 0, 1);\
+   ECHO_SUBBYTES_2WAY(_state2, 1, 1);\
+   ECHO_SUBBYTES_2WAY(_state2, 2, 1);\
+   ECHO_SUBBYTES_2WAY(_state2, 3, 1);\
+   ECHO_SUBBYTES_2WAY(_state2, 0, 2);\
+   ECHO_SUBBYTES_2WAY(_state2, 1, 2);\
+   ECHO_SUBBYTES_2WAY(_state2, 2, 2);\
+   ECHO_SUBBYTES_2WAY(_state2, 3, 2);\
+   ECHO_SUBBYTES_2WAY(_state2, 0, 3);\
+   ECHO_SUBBYTES_2WAY(_state2, 1, 3);\
+   ECHO_SUBBYTES_2WAY(_state2, 2, 3);\
+   ECHO_SUBBYTES_2WAY(_state2, 3, 3);\
+   ECHO_MIXBYTES_2WAY(_state2, _state, 0, t1, t2, s2);\
+   ECHO_MIXBYTES_2WAY(_state2, _state, 1, t1, t2, s2);\
+   ECHO_MIXBYTES_2WAY(_state2, _state, 2, t1, t2, s2);\
+   ECHO_MIXBYTES_2WAY(_state2, _state, 3, t1, t2, s2)
+
+#define SAVESTATE_2WAY(dst, src)\
+        dst[0][0] = src[0][0];\
+        dst[0][1] = src[0][1];\
+        dst[0][2] = src[0][2];\
+        dst[0][3] = src[0][3];\
+        dst[1][0] = src[1][0];\
+        dst[1][1] = src[1][1];\
+        dst[1][2] = src[1][2];\
+        dst[1][3] = src[1][3];\
+        dst[2][0] = src[2][0];\
+        dst[2][1] = src[2][1];\
+        dst[2][2] = src[2][2];\
+        dst[2][3] = src[2][3];\
+        dst[3][0] = src[3][0];\
+        dst[3][1] = src[3][1];\
+        dst[3][2] = src[3][2];\
+        dst[3][3] = src[3][3]
+
+// blockcount always 1
+void echo_2way_compress( echo_2way_context *ctx, const __m256i *pmsg,
+               unsigned int uBlockCount )
+{
+  unsigned int r, b, i, j;
+  __m256i t1, t2, s2, k1;
+  __m256i _state[4][4], _state2[4][4], _statebackup[4][4];
+
+  _state[ 0 ][ 0 ] = ctx->state[ 0 ][ 0 ];
+  _state[ 0 ][ 1 ] = ctx->state[ 0 ][ 1 ];
+  _state[ 0 ][ 2 ] = ctx->state[ 0 ][ 2 ];
+  _state[ 0 ][ 3 ] = ctx->state[ 0 ][ 3 ];
+  _state[ 1 ][ 0 ] = ctx->state[ 1 ][ 0 ];
+  _state[ 1 ][ 1 ] = ctx->state[ 1 ][ 1 ];
+  _state[ 1 ][ 2 ] = ctx->state[ 1 ][ 2 ];
+  _state[ 1 ][ 3 ] = ctx->state[ 1 ][ 3 ];
+  _state[ 2 ][ 0 ] = ctx->state[ 2 ][ 0 ];
+  _state[ 2 ][ 1 ] = ctx->state[ 2 ][ 1 ];
+  _state[ 2 ][ 2 ] = ctx->state[ 2 ][ 2 ];
+  _state[ 2 ][ 3 ] = ctx->state[ 2 ][ 3 ];
+  _state[ 3 ][ 0 ] = ctx->state[ 3 ][ 0 ];
+  _state[ 3 ][ 1 ] = ctx->state[ 3 ][ 1 ];
+  _state[ 3 ][ 2 ] = ctx->state[ 3 ][ 2 ];
+  _state[ 3 ][ 3 ] = ctx->state[ 3 ][ 3 ];
+
+  for ( b = 0; b < uBlockCount; b++ )
+  {
+    ctx->k = _mm256_add_epi64( ctx->k, ctx->const1536 );
+
+    for( j = ctx->uHashSize / 256; j < 4; j++ )
+    {
+      for ( i = 0; i < 4; i++ )
+      {
+        _state[ i ][ j ] = _mm256_load_si256(
+                     pmsg + 4 * (j - (ctx->uHashSize / 256)) + i );
+      }
+   }
+
+   // save state
+   SAVESTATE_2WAY( _statebackup, _state );
+
+   k1 = ctx->k;
+
+   for ( r = 0; r < ctx->uRounds / 2; r++ )
+   {
+       ECHO_ROUND_UNROLL2_2WAY;
+   }
+
+   if ( ctx->uHashSize == 256 )
+   {
+      for ( i = 0; i < 4; i++ )
+      {
+         _state[ i ][ 0 ] = _mm256_xor_si256( _state[ i ][ 0 ],
+                                              _state[ i ][ 1 ] );
+         _state[ i ][ 0 ] = _mm256_xor_si256( _state[ i ][ 0 ],
+                                              _state[ i ][ 2 ] );
+         _state[ i ][ 0 ] = _mm256_xor_si256( _state[ i ][ 0 ],
+                                              _state[ i ][ 3 ] );
+         _state[ i ][ 0 ] = _mm256_xor_si256( _state[ i ][ 0 ],
+                                              _statebackup[ i ][ 0 ] );
+         _state[ i ][ 0 ] = _mm256_xor_si256( _state[ i ][ 0 ],
+                                              _statebackup[ i ][ 1 ] );
+         _state[ i ][ 0 ] = _mm256_xor_si256( _state[ i ][ 0 ],
+                                              _statebackup[ i ][ 2 ] ) ;
+         _state[ i ][ 0 ] = _mm256_xor_si256( _state[ i ][ 0 ],
+                                              _statebackup[ i ][ 3 ] );
+       }
+    }
+    else
+    {
+       for ( i = 0; i < 4; i++ )
+       {
+         _state[ i ][ 0 ] = _mm256_xor_si256( _state[ i ][ 0 ],
+                                              _state[ i ][ 2 ] );
+         _state[ i ][ 1 ] = _mm256_xor_si256( _state[ i ][ 1 ],
+                                              _state[ i ][ 3 ] );
+         _state[ i ][ 0 ] = _mm256_xor_si256( _state[ i ][ 0 ],
+                                              _statebackup[ i ][ 0 ] );
+         _state[ i ][ 0 ] = _mm256_xor_si256( _state[ i ] [0 ],
+                                              _statebackup[ i ][ 2 ] );
+         _state[ i ][ 1 ] = _mm256_xor_si256( _state[ i ][ 1 ],
+                                              _statebackup[ i ][ 1 ] );
+         _state[ i ][ 1 ] = _mm256_xor_si256( _state[ i ][ 1 ],
+                                              _statebackup[ i ][ 3 ] );
+      }
+    }
+    pmsg += ctx->uBlockLength;
+  }
+  SAVESTATE_2WAY(ctx->state, _state);
+
+}
+int echo_2way_init( echo_2way_context *ctx, int nHashSize )
+{
+        int i, j;
+
+   ctx->k = m256_zero;
+   ctx->processed_bits = 0;
+   ctx->uBufferBytes = 0;
+
+   switch( nHashSize )
+   {
+                case 256:
+                        ctx->uHashSize = 256;
+                        ctx->uBlockLength = 192;
+                        ctx->uRounds = 8;
+                        ctx->hashsize = m256_const2_64( 0, 0x100 );
+                        ctx->const1536 = m256_const2_64( 0, 0x600 );
+                        break;
+
+                case 512:
+                        ctx->uHashSize = 512;
+                        ctx->uBlockLength = 128;
+                        ctx->uRounds = 10;
+                        ctx->hashsize = m256_const2_64( 0, 0x200 );
+                        ctx->const1536 = m256_const2_64( 0, 0x400 );
+                        break;
+
+                default:
+                        return 1;
+        }
+
+        for( i = 0; i < 4; i++ )
+                for( j = 0; j < nHashSize / 256; j++ )
+                        ctx->state[ i ][ j ] = ctx->hashsize;
+
+        for( i = 0; i < 4; i++ )
+                for( j = nHashSize / 256; j < 4; j++ )
+                        ctx->state[ i ][ j ] = m256_zero;
+
+        return 0;
+}
+
+int echo_2way_update_close( echo_2way_context *state, void *hashval,
+                              const void *data, int databitlen )
+{
+// bytelen is either 32 (maybe), 64 or 80 or 128!
+// all are less than full block.
+
+   int vlen = databitlen / 128;  // * 4 lanes / 128 bits per lane
+   const int vblen = state->uBlockLength / 16; //  16 bytes per lane
+   __m256i remainingbits;
+
+   if ( databitlen == 1024 )
+   {
+      echo_2way_compress( state, data, 1 );
+      state->processed_bits = 1024;
+      remainingbits = m256_const2_64( 0, -1024 );
+      vlen = 0;
+   }
+   else
+   {
+      memcpy_256( state->buffer, data, vlen );
+      state->processed_bits += (unsigned int)( databitlen );
+      remainingbits = m256_const2_64( 0, databitlen );
+   }
+
+   state->buffer[ vlen ] = m256_const2_64( 0, 0x80 );
+   memset_zero_256( state->buffer + vlen + 1, vblen - vlen - 2 );
+   state->buffer[ vblen-2 ] = m256_const2_64( (uint64_t)state->uHashSize << 48, 0 );
+   state->buffer[ vblen-1 ] = m256_const2_64( 0, state->processed_bits );
+
+   state->k = _mm256_add_epi64( state->k, remainingbits );
+   state->k = _mm256_sub_epi64( state->k, state->const1536 );
+
+   echo_2way_compress( state, state->buffer, 1 );
+
+   _mm256_store_si256( (__m256i*)hashval + 0, state->state[ 0 ][ 0] );
+   _mm256_store_si256( (__m256i*)hashval + 1, state->state[ 1 ][ 0] );
+
+   if ( state->uHashSize == 512 )
+   {
+      _mm256_store_si256( (__m256i*)hashval + 2, state->state[ 2 ][ 0 ] );
+      _mm256_store_si256( (__m256i*)hashval + 3, state->state[ 3 ][ 0 ] );
+   }
+   return 0;
+}
+
+int echo_2way_full( echo_2way_context *ctx, void *hashval, int nHashSize,
+                    const void *data, int datalen )
+{
+   int i, j;
+   int databitlen = datalen * 8;
+   ctx->k = m256_zero;
+   ctx->processed_bits = 0;
+   ctx->uBufferBytes = 0;
+
+   switch( nHashSize )
+   {
+      case 256:
+         ctx->uHashSize = 256;
+         ctx->uBlockLength = 192;
+         ctx->uRounds = 8;
+         ctx->hashsize = m256_const2_64( 0, 0x100 );
+         ctx->const1536 = m256_const2_64( 0, 0x600 );
+         break;
+
+      case 512:
+         ctx->uHashSize = 512;
+         ctx->uBlockLength = 128;
+         ctx->uRounds = 10;
+         ctx->hashsize = m256_const2_64( 0, 0x200 );
+         ctx->const1536 = m256_const2_64( 0, 0x400 );
+         break;
+
+      default:
+         return 1;
+   }
+
+   for( i = 0; i < 4; i++ )
+      for( j = 0; j < nHashSize / 256; j++ )
+         ctx->state[ i ][ j ] = ctx->hashsize;
+
+   for( i = 0; i < 4; i++ )
+      for( j = nHashSize / 256; j < 4; j++ )
+         ctx->state[ i ][ j ] = m256_zero;
+
+   int vlen = datalen / 32;
+   const int vblen = ctx->uBlockLength / 16; //  16 bytes per lane
+   __m256i remainingbits;
+
+   if ( databitlen == 1024 )
+   {
+      echo_2way_compress( ctx, data, 1 );
+      ctx->processed_bits = 1024;
+      remainingbits = m256_const2_64( 0, -1024 );
+      vlen = 0;
+   }
+   else
+   {
+      vlen = databitlen / 128;  // * 4 lanes / 128 bits per lane
+      memcpy_256( ctx->buffer, data, vlen );
+      ctx->processed_bits += (unsigned int)( databitlen );
+      remainingbits = m256_const2_64( 0, databitlen );
+   }
+
+   ctx->buffer[ vlen ] = m256_const2_64( 0, 0x80 );
+   memset_zero_256( ctx->buffer + vlen + 1, vblen - vlen - 2 );
+   ctx->buffer[ vblen-2 ] = m256_const2_64( (uint64_t)ctx->uHashSize << 48, 0 );
+   ctx->buffer[ vblen-1 ] = m256_const2_64( 0, ctx->processed_bits );
+
+   ctx->k = _mm256_add_epi64( ctx->k, remainingbits );
+   ctx->k = _mm256_sub_epi64( ctx->k, ctx->const1536 );
+
+   echo_2way_compress( ctx, ctx->buffer, 1 );
+
+   _mm256_store_si256( (__m256i*)hashval + 0, ctx->state[ 0 ][ 0] );
+   _mm256_store_si256( (__m256i*)hashval + 1, ctx->state[ 1 ][ 0] );
+
+   if ( ctx->uHashSize == 512 )
+   {
+      _mm256_store_si256( (__m256i*)hashval + 2, ctx->state[ 2 ][ 0 ] );
+      _mm256_store_si256( (__m256i*)hashval + 3, ctx->state[ 3 ][ 0 ] );
+   }
+   return 0;
+}
+
+
+#endif   // VAES
--- a/algo/echo/echo-hash-4way.h
+++ b/algo/echo/echo-hash-4way.h
@@ -1,10 +1,12 @@
 #if !defined(ECHO_HASH_4WAY_H__)
 #define ECHO_HASH_4WAY_H__ 1

-#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+#if defined(__VAES__)

 #include "simd-utils.h"

+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+
 typedef struct
 {
   __m512i    state[4][4];
@@ -20,6 +22,7 @@ typedef struct
   unsigned int   processed_bits;

 } echo_4way_context __attribute__ ((aligned (64)));
+#define echo512_4way_context echo_4way_context

 int echo_4way_init( echo_4way_context *state, int hashbitlen );
 #define echo512_4way_init( state ) echo_4way_init( state, 512 )
@@ -29,8 +32,8 @@ int echo_4way_update( echo_4way_context *state, const void *data,
    unsigned int databitlen);
 #define echo512_4way_update echo_4way_update

-int echo_close( echo_4way_context *state, void *hashval );
-#define echo512_4way_close echo_4way_close
+// int echo_4way_close( echo_4way_context *state, void *hashval );
+// #define echo512_4way_close echo_4way_close

 int echo_4way_update_close( echo_4way_context *state, void *hashval,
                              const void *data, int databitlen );
@@ -43,5 +46,45 @@ int echo_4way_full( echo_4way_context *ctx, void *hashval, int nHashSize,
 #define echo256_4way_full( state, hashval, data, datalen ) \
           echo_4way_full( state, hashval, 256, data, datalen )

-#endif 
-#endif
+#endif   // AVX512
+
+typedef struct
+{
+   __m256i    state[4][4];
+   __m256i    buffer[ 4 * 192 / 16 ];  // 4x128 interleaved 192 bytes
+   __m256i    k;
+   __m256i    hashsize;
+   __m256i    const1536;
+
+   unsigned int   uRounds;
+   unsigned int   uHashSize;
+   unsigned int   uBlockLength;
+   unsigned int   uBufferBytes;
+   unsigned int   processed_bits;
+
+} echo_2way_context __attribute__ ((aligned (64)));
+#define echo512_2way_context echo_2way_context
+
+int echo_2way_init( echo_2way_context *state, int hashbitlen );
+#define echo512_2way_init( state ) echo_2way_init( state, 512 )
+#define echo256_2way_init( state ) echo_2way_init( state, 256 )
+
+int echo_2way_update( echo_2way_context *state, const void *data,
+    unsigned int databitlen);
+#define echo512_2way_update echo_2way_update
+
+int echo_2way_update_close( echo_2way_context *state, void *hashval,
+                              const void *data, int databitlen );
+#define echo512_2way_update_close echo_2way_update_close
+
+int echo_2way_full( echo_2way_context *ctx, void *hashval, int nHashSize,
+                    const void *data, int datalen );
+#define echo512_2way_full( state, hashval, data, datalen ) \
+           echo_2way_full( state, hashval, 512, data, datalen )
+#define echo256_2way_full( state, hashval, data, datalen ) \
+           echo_2way_full( state, hashval, 256, data, datalen )
+
+
+#endif  // VAES
+
+#endif   // ECHO_HASH_4WAY_H__
--- a/algo/fugue/fugue-aesni.c
+++ b/algo/fugue/fugue-aesni.c
@@ -0,0 +1,565 @@
+/*
+ * file        : fugue_vperm.c
+ * version     : 1.0.208
+ * date        : 14.12.2010
+ * 
+ * - vperm and aes_ni implementations of hash function Fugue
+ * - implements NIST hash api
+ * - assumes that message lenght is multiple of 8-bits
+ * - _FUGUE_VPERM_ must be defined if compiling with ../main.c
+ * - default version is vperm, define AES_NI for aes_ni version
+ * 
+ * Cagdas Calik
+ * ccalik@metu.edu.tr
+ * Institute of Applied Mathematics, Middle East Technical University, Turkey.
+ *
+ */
+
+#if defined(__AES__)
+
+#include <x86intrin.h>
+
+#include <memory.h>
+#include "fugue-aesni.h"
+
+
+MYALIGN const unsigned long long _supermix1a[]	= {0x0202010807020100, 0x0a05000f06010c0b};
+MYALIGN const unsigned long long _supermix1b[]	= {0x0b0d080703060504, 0x0e0a090c050e0f0a};
+MYALIGN const unsigned long long _supermix1c[]	= {0x0402060c070d0003, 0x090a060580808080};
+MYALIGN const unsigned long long _supermix1d[]	= {0x808080800f0e0d0c, 0x0f0e0d0c80808080};
+MYALIGN const unsigned long long _supermix2a[]	= {0x07020d0880808080, 0x0b06010c050e0f0a};
+MYALIGN const unsigned long long _supermix4a[]	= {0x000f0a050c0b0601, 0x0302020404030e09};
+MYALIGN const unsigned long long _supermix4b[]	= {0x07020d08080e0d0d, 0x07070908050e0f0a};
+MYALIGN const unsigned long long _supermix4c[]	= {0x0706050403020000, 0x0302000007060504};
+MYALIGN const unsigned long long _supermix7a[]	= {0x010c0b060d080702, 0x0904030e03000104};
+MYALIGN const unsigned long long _supermix7b[]	= {0x8080808080808080, 0x0504070605040f06};
+MYALIGN const unsigned long long _k_n[] = {0x4E4E4E4E4E4E4E4E, 0x1B1B1B1B0E0E0E0E};
+MYALIGN const unsigned char _shift_one_mask[]   = {7, 4, 5, 6, 11, 8, 9, 10, 15, 12, 13, 14, 3, 0, 1, 2};
+MYALIGN const unsigned char _shift_four_mask[]  = {13, 14, 15, 12, 1, 2, 3, 0, 5, 6, 7, 4, 9, 10, 11, 8};
+MYALIGN const unsigned char _shift_seven_mask[] = {10, 11, 8, 9, 14, 15, 12, 13, 2, 3, 0, 1, 6, 7, 4, 5};
+MYALIGN const unsigned char _aes_shift_rows[]   = {0, 5, 10, 15, 4, 9, 14, 3, 8, 13, 2, 7, 12, 1, 6, 11};
+MYALIGN const unsigned int _inv_shift_rows[] = {0x070a0d00, 0x0b0e0104, 0x0f020508, 0x0306090c};
+MYALIGN const unsigned int _mul2mask[] = {0x1b1b0000, 0x00000000, 0x00000000, 0x00000000};
+MYALIGN const unsigned int _mul4mask[] = {0x2d361b00, 0x00000000, 0x00000000, 0x00000000};
+MYALIGN const unsigned int _lsbmask2[] = {0x03030303, 0x03030303, 0x03030303, 0x03030303};
+
+
+MYALIGN const unsigned int _IV512[] = {		
+	0x00000000, 0x00000000,	0x7ea50788, 0x00000000,
+	0x75af16e6, 0xdbe4d3c5, 0x27b09aac, 0x00000000,
+	0x17f115d9, 0x54cceeb6, 0x0b02e806, 0x00000000,
+	0xd1ef924a, 0xc9e2c6aa, 0x9813b2dd, 0x00000000,
+	0x3858e6ca, 0x3f207f43, 0xe778ea25, 0x00000000,
+	0xd6dd1f95, 0x1dd16eda, 0x67353ee1, 0x00000000};
+
+#if defined(__SSE4_1__)
+
+#define PACK_S0(s0, s1, t1)\
+   s0 = _mm_castps_si128(_mm_insert_ps(_mm_castsi128_ps(s0), _mm_castsi128_ps(s1), 0x30))
+
+#define UNPACK_S0(s0, s1, t1)\
+   s1 = _mm_castps_si128(_mm_insert_ps(_mm_castsi128_ps(s1), _mm_castsi128_ps(s0), 0xc0));\
+   s0 = mm128_mask_32( s0, 8 )
+
+#define CMIX(s1, s2, r1, r2, t1, t2)\
+   t1 = s1;\
+   t1 = _mm_castps_si128(_mm_shuffle_ps(_mm_castsi128_ps(t1), _mm_castsi128_ps(s2), _MM_SHUFFLE(3, 0, 2, 1)));\
+   r1 = _mm_xor_si128(r1, t1);\
+   r2 = _mm_xor_si128(r2, t1);
+
+#else   // SSE2
+
+#define PACK_S0(s0, s1, t1)\
+   t1 = _mm_shuffle_epi32(s1, _MM_SHUFFLE(0, 3, 3, 3));\
+   s0 = _mm_xor_si128(s0, t1);
+
+#define UNPACK_S0(s0, s1, t1)\
+   t1 = _mm_shuffle_epi32(s0, _MM_SHUFFLE(3, 3, 3, 3));\
+   s1 = _mm_castps_si128(_mm_move_ss(_mm_castsi128_ps(s1), _mm_castsi128_ps(t1)));\
+   s0 = mm128_mask_32( s0, 8 )
+
+#define CMIX(s1, s2, r1, r2, t1, t2)\
+   t1 = _mm_shuffle_epi32(s1, 0xf9);\
+   t2 = _mm_shuffle_epi32(s2, 0xcf);\
+   t1 = _mm_xor_si128(t1, t2);\
+   r1 = _mm_xor_si128(r1, t1);\
+   r2 = _mm_xor_si128(r2, t1)
+
+#endif
+
+#define TIX256(msg, s10, s8, s24, s0, t1, t2, t3)\
+	t1 = _mm_shuffle_epi32(s0, _MM_SHUFFLE(3, 3, 0, 3));\
+	s10 = _mm_xor_si128(s10, t1);\
+	t1 = _mm_castps_si128(_mm_load_ss((float*)msg));\
+	s0 = _mm_castps_si128(_mm_move_ss(_mm_castsi128_ps(s0), _mm_castsi128_ps(t1)));\
+	t1 = _mm_slli_si128(t1, 8);\
+	s8 = _mm_xor_si128(s8, t1);\
+	t1 = _mm_shuffle_epi32(s24, _MM_SHUFFLE(3, 3, 0, 3));\
+	s0 = _mm_xor_si128(s0, t1)
+
+
+#define TIX384(msg, s16, s8, s27, s30, s0, s4, t1, t2, t3)\
+	t1 = _mm_shuffle_epi32(s0, _MM_SHUFFLE(3, 3, 0, 3));\
+	s16 = _mm_xor_si128(s16, t1);\
+	t1 = _mm_castps_si128(_mm_load_ss((float*)msg));\
+	s0 = _mm_castps_si128(_mm_move_ss(_mm_castsi128_ps(s0), _mm_castsi128_ps(t1)));\
+	t1 = _mm_slli_si128(t1, 8);\
+	s8 = _mm_xor_si128(s8, t1);\
+	t1 = _mm_shuffle_epi32(s27, _MM_SHUFFLE(3, 3, 0, 3));\
+	s0 = _mm_xor_si128(s0, t1);\
+	t1 = _mm_shuffle_epi32(s30, _MM_SHUFFLE(3, 3, 0, 3));\
+	s4 = _mm_xor_si128(s4, t1)
+
+#define TIX512(msg, s22, s8, s24, s27, s30, s0, s4, s7, t1, t2, t3)\
+	t1 = _mm_shuffle_epi32(s0, _MM_SHUFFLE(3, 3, 0, 3));\
+	s22 = _mm_xor_si128(s22, t1);\
+	t1 = _mm_castps_si128(_mm_load_ss((float*)msg));\
+	s0 = _mm_castps_si128(_mm_move_ss(_mm_castsi128_ps(s0), _mm_castsi128_ps(t1)));\
+	t1 = _mm_slli_si128(t1, 8);\
+	s8 = _mm_xor_si128(s8, t1);\
+	t1 = _mm_shuffle_epi32(s24, _MM_SHUFFLE(3, 3, 0, 3));\
+	s0 = _mm_xor_si128(s0, t1);\
+	t1 = _mm_shuffle_epi32(s27, _MM_SHUFFLE(3, 3, 0, 3));\
+	s4 = _mm_xor_si128(s4, t1);\
+	t1 = _mm_shuffle_epi32(s30, _MM_SHUFFLE(3, 3, 0, 3));\
+	s7 = _mm_xor_si128(s7, t1)
+
+
+#define PRESUPERMIX(x, t1, s1, s2, t2)\
+	s1 = x;\
+	s2 = _mm_add_epi8(x, x);\
+	t2 = _mm_add_epi8(s2, s2);\
+	t1 = _mm_srli_epi16(x, 6);\
+	t1 = _mm_and_si128(t1, M128(_lsbmask2));\
+	s2 = _mm_xor_si128(s2, _mm_shuffle_epi8(M128(_mul2mask), t1));\
+	x  = _mm_xor_si128(t2, _mm_shuffle_epi8(M128(_mul4mask), t1))
+
+#define SUBSTITUTE(r0, _t1, _t2, _t3, _t0)\
+	_t2 = _mm_shuffle_epi8(r0, M128(_inv_shift_rows));\
+	_t2 = _mm_aesenclast_si128( _t2, m128_zero )
+	
+#define SUPERMIX(t0, t1, t2, t3, t4)\
+	PRESUPERMIX(t0, t1, t2, t3, t4);\
+	POSTSUPERMIX(t0, t1, t2, t3, t4)
+
+
+#define POSTSUPERMIX(t0, t1, t2, t3, t4)\
+	t1 = t2;\
+	t1 = _mm_shuffle_epi8(t1, M128(_supermix1b));\
+	t4 = t1;\
+	t1 = _mm_shuffle_epi8(t1, M128(_supermix1c));\
+	t4 = _mm_xor_si128(t4, t1);\
+	t1 = t4;\
+	t1 = _mm_shuffle_epi8(t1, M128(_supermix1d));\
+	t4 = _mm_xor_si128(t4, t1);\
+	t1 = t2;\
+	t1 = _mm_shuffle_epi8(t1, M128(_supermix1a));\
+	t4 = _mm_xor_si128(t4, t1);\
+	t2 = _mm_xor_si128(t2, t3);\
+	t2 = _mm_xor_si128(t2, t0);\
+	t2 = _mm_shuffle_epi8(t2, M128(_supermix7a));\
+	t4 = _mm_xor_si128(t4, t2);\
+	t2 = _mm_shuffle_epi8(t2, M128(_supermix7b));\
+	t4 = _mm_xor_si128(t4, t2);\
+	t3 = _mm_shuffle_epi8(t3, M128(_supermix2a));\
+	t1 = t0;\
+	t1 = _mm_shuffle_epi8(t1, M128(_supermix4a));\
+	t4 = _mm_xor_si128(t4, t1);\
+	t0 = _mm_shuffle_epi8(t0, M128(_supermix4b));\
+	t0 = _mm_xor_si128(t0, t3);\
+	t4 = _mm_xor_si128(t4, t0);\
+	t0 = _mm_shuffle_epi8(t0, M128(_supermix4c));\
+	t4 = _mm_xor_si128(t4, t0)
+
+
+#define SUBROUND512_3(r1a, r1b, r1c, r1d, r2a, r2b, r2c, r2d, r3a, r3b, r3c, r3d)\
+	CMIX(r1a, r1b, r1c, r1d, _t0, _t1);\
+	PACK_S0(r1c, r1a, _t0);\
+	SUBSTITUTE(r1c, _t1, _t2, _t3, _t0);\
+	SUPERMIX(_t2, _t3, _t0, _t1, r1c);\
+	_t0 = _mm_shuffle_epi32(r1c, 0x39);\
+	r2c = _mm_xor_si128(r2c, _t0);\
+   _t0 = mm128_mask_32( _t0, 8 ); \
+	r2d = _mm_xor_si128(r2d, _t0);\
+	UNPACK_S0(r1c, r1a, _t3);\
+	SUBSTITUTE(r2c, _t1, _t2, _t3, _t0);\
+	SUPERMIX(_t2, _t3, _t0, _t1, r2c);\
+	_t0 = _mm_shuffle_epi32(r2c, 0x39);\
+	r3c = _mm_xor_si128(r3c, _t0);\
+   _t0 = mm128_mask_32( _t0, 8 ); \
+	r3d = _mm_xor_si128(r3d, _t0);\
+	UNPACK_S0(r2c, r2a, _t3);\
+	SUBSTITUTE(r3c, _t1, _t2, _t3, _t0);\
+	SUPERMIX(_t2, _t3, _t0, _t1, r3c);\
+	UNPACK_S0(r3c, r3a, _t3)
+
+
+#define SUBROUND512_4(r1a, r1b, r1c, r1d, r2a, r2b, r2c, r2d, r3a, r3b, r3c, r3d, r4a, r4b, r4c, r4d)\
+	CMIX(r1a, r1b, r1c, r1d, _t0, _t1);\
+	PACK_S0(r1c, r1a, _t0);\
+	SUBSTITUTE(r1c, _t1, _t2, _t3, _t0);\
+	SUPERMIX(_t2, _t3, _t0, _t1, r1c);\
+	_t0 = _mm_shuffle_epi32(r1c, 0x39);\
+	r2c = _mm_xor_si128(r2c, _t0);\
+   _t0 = mm128_mask_32( _t0, 8 ); \
+	r2d = _mm_xor_si128(r2d, _t0);\
+	UNPACK_S0(r1c, r1a, _t3);\
+	SUBSTITUTE(r2c, _t1, _t2, _t3, _t0);\
+	SUPERMIX(_t2, _t3, _t0, _t1, r2c);\
+	_t0 = _mm_shuffle_epi32(r2c, 0x39);\
+	r3c = _mm_xor_si128(r3c, _t0);\
+   _t0 = mm128_mask_32( _t0, 8 ); \
+	r3d = _mm_xor_si128(r3d, _t0);\
+	UNPACK_S0(r2c, r2a, _t3);\
+	SUBSTITUTE(r3c, _t1, _t2, _t3, _t0);\
+	SUPERMIX(_t2, _t3, _t0, _t1, r3c);\
+	_t0 = _mm_shuffle_epi32(r3c, 0x39);\
+	r4c = _mm_xor_si128(r4c, _t0);\
+   _t0 = mm128_mask_32( _t0, 8 ); \
+	r4d = _mm_xor_si128(r4d, _t0);\
+	UNPACK_S0(r3c, r3a, _t3);\
+	SUBSTITUTE(r4c, _t1, _t2, _t3, _t0);\
+	SUPERMIX(_t2, _t3, _t0, _t1, r4c);\
+	UNPACK_S0(r4c, r4a, _t3)
+
+
+
+#define LOADCOLUMN(x, s, a)\
+	block[0] = col[(base + a + 0) % s];\
+	block[1] = col[(base + a + 1) % s];\
+	block[2] = col[(base + a + 2) % s];\
+	block[3] = col[(base + a + 3) % s];\
+	x = _mm_load_si128((__m128i*)block)
+
+#define STORECOLUMN(x, s)\
+	_mm_store_si128((__m128i*)block, x);\
+	col[(base + 0) % s] = block[0];\
+	col[(base + 1) % s] = block[1];\
+	col[(base + 2) % s] = block[2];\
+	col[(base + 3) % s] = block[3]
+
+void Compress512(hashState_fugue *ctx, const unsigned char *pmsg, unsigned int uBlockCount)
+{
+   __m128i _t0, _t1, _t2, _t3;
+
+   switch(ctx->base)
+   {
+      case 1:
+         TIX512( pmsg, ctx->state[3], ctx->state[10], ctx->state[4],
+                       ctx->state[5], ctx->state[ 6], ctx->state[8],
+		       ctx->state[9], ctx->state[10], _t0, _t1, _t2 );
+
+	 SUBROUND512_4( ctx->state[8], ctx->state[9], ctx->state[7],
+                        ctx->state[1], ctx->state[7], ctx->state[8],
+		       	ctx->state[6], ctx->state[0], ctx->state[6],
+		       	ctx->state[7], ctx->state[5], ctx->state[11],
+		       	ctx->state[5], ctx->state[6], ctx->state[4],
+		       	ctx->state[10] );
+         ctx->base++;
+         pmsg += 4;
+         uBlockCount--;
+      if( uBlockCount == 0 ) break;
+
+      case 2:
+         TIX512( pmsg, ctx->state[11], ctx->state[6], ctx->state[0],
+                       ctx->state[ 1], ctx->state[2], ctx->state[4],
+		       ctx->state[ 5], ctx->state[6], _t0, _t1, _t2);
+
+         SUBROUND512_4( ctx->state[4], ctx->state[5], ctx->state[3],
+                        ctx->state[9], ctx->state[3], ctx->state[4],
+		       	ctx->state[2], ctx->state[8], ctx->state[2],
+		       	ctx->state[3], ctx->state[1], ctx->state[7],
+		       	ctx->state[1], ctx->state[2], ctx->state[0],
+		       	ctx->state[6]);
+
+         ctx->base = 0;
+         pmsg += 4;
+         uBlockCount--;
+      break;
+   }
+
+
+   while( uBlockCount > 0 )
+   {
+      TIX512( pmsg, ctx->state[ 7], ctx->state[2], ctx->state[8], ctx->state[9],
+                    ctx->state[10], ctx->state[0], ctx->state[1], ctx->state[2],
+              _t0, _t1, _t2 );
+      SUBROUND512_4( ctx->state[0], ctx->state[1], ctx->state[11],
+                     ctx->state[5], ctx->state[11], ctx->state[0],
+		     ctx->state[10], ctx->state[4], ctx->state[10],
+		     ctx->state[11], ctx->state[9], ctx->state[3],
+		     ctx->state[9], ctx->state[10], ctx->state[8],
+		     ctx->state[2] );
+
+      ctx->base++;
+      pmsg += 4;
+      uBlockCount--;
+      if( uBlockCount == 0 ) break;
+
+      TIX512( pmsg, ctx->state[3], ctx->state[10], ctx->state[4], ctx->state[5],
+                    ctx->state[6], ctx->state[8], ctx->state[9], ctx->state[10],
+              _t0, _t1, _t2 );
+
+      SUBROUND512_4( ctx->state[8], ctx->state[9], ctx->state[7], ctx->state[1],                     ctx->state[7], ctx->state[8], ctx->state[6], ctx->state[0],
+		     ctx->state[6], ctx->state[7], ctx->state[5], ctx->state[11],
+		     ctx->state[5], ctx->state[6, ctx->state[4], ctx->state[10]);
+
+      ctx->base++;
+      pmsg += 4;
+      uBlockCount--;
+      if( uBlockCount == 0 ) break;
+
+      TIX512( pmsg, ctx->state[11], ctx->state[6], ctx->state[0], ctx->state[1],
+		    ctx->state[2], ctx->state[4], ctx->state[5], ctx->state[6],
+               _t0, _t1, _t2);
+      SUBROUND512_4( ctx->state[4], ctx->state[5], ctx->state[3], ctx->state[9],
+		     ctx->state[3], ctx->state[4], ctx->state[2], ctx->state[8],
+		     ctx->state[2], ctx->state[3], ctx->state[1], ctx->state[7],
+		     ctx->state[1], ctx->state[2], ctx->state[0], ctx->state[6]);
+
+      ctx->base = 0;
+      pmsg += 4;
+      uBlockCount--;
+   }
+
+}
+
+void Final512(hashState_fugue *ctx, BitSequence *hashval)
+{
+        unsigned int block[4] __attribute__ ((aligned (32)));
+        unsigned int col[36] __attribute__ ((aligned (16)));
+	unsigned int i, base;
+	__m128i r0, _t0, _t1, _t2, _t3;
+
+	for(i = 0; i < 12; i++)
+	{
+		_mm_store_si128((__m128i*)block, ctx->state[i]);
+
+		col[3 * i + 0] = block[0];
+		col[3 * i + 1] = block[1];
+		col[3 * i + 2] = block[2];
+	}
+
+	base = (36 - (12 * ctx->base)) % 36;
+
+	for(i = 0; i < 32; i++)
+	{
+		// ROR3
+		base = (base + 33) % 36;
+
+		// CMIX
+		col[(base +  0) % 36] ^= col[(base + 4) % 36];
+		col[(base +  1) % 36] ^= col[(base + 5) % 36];
+		col[(base +  2) % 36] ^= col[(base + 6) % 36];
+		col[(base +  18) % 36] ^= col[(base + 4) % 36];
+		col[(base +  19) % 36] ^= col[(base + 5) % 36];
+		col[(base +  20) % 36] ^= col[(base + 6) % 36];
+
+		// SMIX
+		LOADCOLUMN(r0, 36, 0);
+		SUBSTITUTE(r0, _t1, _t2, _t3, _t0);
+		SUPERMIX(_t2, _t3, _t0, _t1, r0);
+		STORECOLUMN(r0, 36);
+	}
+
+	for(i = 0; i < 13; i++)
+	{
+		// S4 += S0; S9 += S0; S18 += S0; S27 += S0;
+		col[(base +  4) % 36] ^= col[(base + 0) % 36];
+		col[(base +  9) % 36] ^= col[(base + 0) % 36];
+		col[(base + 18) % 36] ^= col[(base + 0) % 36];
+		col[(base + 27) % 36] ^= col[(base + 0) % 36];
+
+		// ROR9
+		base = (base + 27) % 36;
+
+		// SMIX
+		LOADCOLUMN(r0, 36, 0);
+		SUBSTITUTE(r0, _t1, _t2, _t3, _t0);
+		SUPERMIX(_t2, _t3, _t0, _t1, r0);
+		STORECOLUMN(r0, 36);
+
+		// S4 += S0; S10 += S0; S18 += S0; S27 += S0;
+		col[(base +  4) % 36] ^= col[(base + 0) % 36];
+		col[(base + 10) % 36] ^= col[(base + 0) % 36];
+		col[(base + 18) % 36] ^= col[(base + 0) % 36];
+		col[(base + 27) % 36] ^= col[(base + 0) % 36];
+
+		// ROR9
+		base = (base + 27) % 36;
+
+		// SMIX
+		LOADCOLUMN(r0, 36, 0);
+		SUBSTITUTE(r0, _t1, _t2, _t3, _t0);
+		SUPERMIX(_t2, _t3, _t0, _t1, r0);
+		STORECOLUMN(r0, 36);
+
+		// S4 += S0; S10 += S0; S19 += S0; S27 += S0;
+		col[(base +  4) % 36] ^= col[(base + 0) % 36];
+		col[(base + 10) % 36] ^= col[(base + 0) % 36];
+		col[(base + 19) % 36] ^= col[(base + 0) % 36];
+		col[(base + 27) % 36] ^= col[(base + 0) % 36];
+
+		// ROR9
+		base = (base + 27) % 36;
+
+		// SMIX
+		LOADCOLUMN(r0, 36, 0);
+		SUBSTITUTE(r0, _t1, _t2, _t3, _t0);
+		SUPERMIX(_t2, _t3, _t0, _t1, r0);
+		STORECOLUMN(r0, 36);
+
+		// S4 += S0; S10 += S0; S19 += S0; S28 += S0;
+		col[(base +  4) % 36] ^= col[(base + 0) % 36];
+		col[(base + 10) % 36] ^= col[(base + 0) % 36];
+		col[(base + 19) % 36] ^= col[(base + 0) % 36];
+		col[(base + 28) % 36] ^= col[(base + 0) % 36];
+
+		// ROR8
+		base = (base + 28) % 36;
+
+		// SMIX
+		LOADCOLUMN(r0, 36, 0);
+		SUBSTITUTE(r0, _t1, _t2, _t3, _t0);
+		SUPERMIX(_t2, _t3, _t0, _t1, r0);
+		STORECOLUMN(r0, 36);
+	}
+
+	// S4 += S0; S9 += S0; S18 += S0; S27 += S0;
+	col[(base +  4) % 36] ^= col[(base + 0) % 36];
+	col[(base +  9) % 36] ^= col[(base + 0) % 36];
+	col[(base + 18) % 36] ^= col[(base + 0) % 36];
+	col[(base + 27) % 36] ^= col[(base + 0) % 36];
+
+	// Transform to the standard basis and store output; S1 || S2 || S3 || S4
+	LOADCOLUMN(r0, 36, 1);
+	_mm_store_si128((__m128i*)hashval, r0);
+
+	// Transform to the standard basis and store output; S9 || S10 || S11 || S12
+	LOADCOLUMN(r0, 36, 9);
+	_mm_store_si128((__m128i*)hashval + 1, r0);
+
+	// Transform to the standard basis and store output; S18 || S19 || S20 || S21
+	LOADCOLUMN(r0, 36, 18);
+	_mm_store_si128((__m128i*)hashval + 2, r0);
+
+	// Transform to the standard basis and store output; S27 || S28 || S29 || S30
+	LOADCOLUMN(r0, 36, 27);
+	_mm_store_si128((__m128i*)hashval + 3, r0);
+}
+
+HashReturn fugue512_Init(hashState_fugue *ctx, int nHashSize)
+{
+	int i;
+	ctx->processed_bits = 0;
+	ctx->uBufferBytes = 0;
+	ctx->base = 0;
+
+
+	ctx->uHashSize = 512;
+	ctx->uBlockLength = 4;
+
+	for(i = 0; i < 6; i++)
+		ctx->state[i] = m128_zero;
+
+	ctx->state[6]  = _mm_load_si128((__m128i*)_IV512 + 0);
+	ctx->state[7]  = _mm_load_si128((__m128i*)_IV512 + 1);
+	ctx->state[8]  = _mm_load_si128((__m128i*)_IV512 + 2);
+	ctx->state[9]  = _mm_load_si128((__m128i*)_IV512 + 3);
+	ctx->state[10] = _mm_load_si128((__m128i*)_IV512 + 4);
+	ctx->state[11] = _mm_load_si128((__m128i*)_IV512 + 5);
+
+	return SUCCESS;
+}
+
+
+HashReturn fugue512_Update(hashState_fugue *state, const void *data, DataLength databitlen)
+{
+	unsigned int uByteLength, uBlockCount, uRemainingBytes;
+
+	uByteLength = (unsigned int)(databitlen / 8);
+
+	if(state->uBufferBytes + uByteLength >= state->uBlockLength)
+	{
+		if(state->uBufferBytes != 0)
+		{
+			// Fill the buffer
+			memcpy(state->buffer + state->uBufferBytes, (void*)data, state->uBlockLength - state->uBufferBytes);
+
+			// Process the buffer
+			Compress512(state, state->buffer, 1);
+
+			state->processed_bits += state->uBlockLength * 8;
+			data += state->uBlockLength - state->uBufferBytes;
+			uByteLength -= state->uBlockLength - state->uBufferBytes;
+		}
+
+		// buffer now does not contain any unprocessed bytes
+
+		uBlockCount = uByteLength / state->uBlockLength;
+		uRemainingBytes = uByteLength % state->uBlockLength;
+
+		if(uBlockCount > 0)
+		{
+			Compress512(state, data, uBlockCount);
+
+			state->processed_bits += uBlockCount * state->uBlockLength * 8;
+			data += uBlockCount * state->uBlockLength;
+		}
+
+		if(uRemainingBytes > 0)
+		{
+			memcpy(state->buffer, (void*)data, uRemainingBytes);
+		}
+
+		state->uBufferBytes = uRemainingBytes;
+	}
+	else
+	{
+		memcpy(state->buffer + state->uBufferBytes, (void*)data, uByteLength);
+		state->uBufferBytes += uByteLength;
+	}
+
+	return SUCCESS;
+}
+
+HashReturn fugue512_Final(hashState_fugue *state, void *hashval)
+{
+	unsigned int i;
+	BitSequence lengthbuf[8] __attribute__((aligned(64)));
+
+	// Update message bit count
+	state->processed_bits += state->uBufferBytes * 8;
+
+	// Pad the remaining buffer bytes with zero
+	if(state->uBufferBytes != 0)
+	{
+	   if ( state->uBufferBytes != state->uBlockLength)
+		memset(state->buffer + state->uBufferBytes, 0, state->uBlockLength - state->uBufferBytes);
+
+	   Compress512(state, state->buffer, 1);
+	}
+
+	// Last two blocks are message length in bits
+	for(i = 0; i < 8; i++)
+           lengthbuf[i] = ((state->processed_bits) >> (8 * (7 - i))) & 0xff;
+
+	// Process the last two blocks
+	Compress512(state, lengthbuf, 2);
+
+	// Finalization
+	Final512(state, hashval);
+
+	return SUCCESS;
+}
+
+
+HashReturn fugue512_full(hashState_fugue *hs, void *hashval, const void *data, DataLength databitlen)
+{
+	fugue512_Init(hs, 512);
+	fugue512_Update(hs, data, databitlen*8);
+	fugue512_Final(hs, hashval);
+	return SUCCESS;
+}
+
+#endif  // AES
--- a/algo/fugue/fugue-aesni.h
+++ b/algo/fugue/fugue-aesni.h
@@ -0,0 +1,50 @@
+/*
+ * file        : hash_api.h
+ * version     : 1.0.208
+ * date        : 14.12.2010
+ * 
+ * Fugue vperm implementation Hash API
+ *
+ * Cagdas Calik
+ * ccalik@metu.edu.tr
+ * Institute of Applied Mathematics, Middle East Technical University, Turkey.
+ *
+ */
+
+#ifndef FUGUE_HASH_API_H
+#define FUGUE_HASH_API_H
+
+#if defined(__AES__) 
+
+#if !defined(__SSE4_1__)
+#error "Unsupported configuration, AES needs SSE4.1. Compile without AES."
+#endif
+
+#include "algo/sha/sha3_common.h"
+#include "simd-utils.h"
+
+
+typedef struct
+{
+	__m128i			state[12];
+	unsigned int	base;
+
+	unsigned int	uHashSize;
+	unsigned int	uBlockLength;
+	unsigned int	uBufferBytes;
+	DataLength		processed_bits;
+	BitSequence		buffer[4];
+
+} hashState_fugue __attribute__ ((aligned (64)));
+
+HashReturn fugue512_Init(hashState_fugue *state, int hashbitlen);
+
+HashReturn fugue512_Update(hashState_fugue *state, const void *data, DataLength databitlen);
+
+HashReturn fugue512_Final(hashState_fugue *state, void *hashval);
+
+HashReturn fugue512_full(hashState_fugue *hs, void *hashval, const void *data, DataLength databitlen);
+
+#endif // AES
+#endif // HASH_API_H
+
--- a/algo/groestl/groestl256-hash-4way.c
+++ b/algo/groestl/groestl256-hash-4way.c
@@ -15,7 +15,9 @@
 #include "miner.h"
 #include "simd-utils.h"

-#if defined(__VAES__) && defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+#if defined(__AVX2__) && defined(__VAES__)
+
+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)


 int groestl256_4way_init( groestl256_4way_context* ctx, uint64_t hashlen )
@@ -43,13 +45,13 @@ int groestl256_4way_init( groestl256_4way_context* ctx, uint64_t hashlen )
 }

 int groestl256_4way_full( groestl256_4way_context* ctx, void* output,
-                                const void* input, uint64_t databitlen )
+                                const void* input, uint64_t datalen )
 {
-   const int len = (int)databitlen / 128;
-   const int hashlen_m128i = 32 / 16;   // bytes to __m128i
+   const int len = (int)datalen >> 4;
+   const int hashlen_m128i = 32 >> 4;   // bytes to __m128i
   const int hash_offset = SIZE256 - hashlen_m128i;
   int rem = ctx->rem_ptr;
-   int blocks = len / SIZE256;
+   uint64_t blocks = len / SIZE256;
   __m512i* in = (__m512i*)input;
   int i;

@@ -87,21 +89,21 @@ int groestl256_4way_full( groestl256_4way_context* ctx, void* output,
   if ( i == SIZE256 - 1 )
   {        
       // only 1 vector left in buffer, all padding at once
-      ctx->buffer[i] = m512_const2_64( (uint64_t)blocks << 56, 0x80 ); 
+       ctx->buffer[i] = m512_const2_64( blocks << 56, 0x80 ); 
   }   
   else
   {
       // add first padding
-       ctx->buffer[i] = m512_const4_64( 0, 0x80, 0, 0x80 );
+       ctx->buffer[i] = m512_const2_64( 0, 0x80 );
       // add zero padding
       for ( i += 1; i < SIZE256 - 1; i++ )
           ctx->buffer[i] = m512_zero;

       // add length padding, second last byte is zero unless blocks > 255
-      ctx->buffer[i] = m512_const2_64( (uint64_t)blocks << 56, 0 );
+       ctx->buffer[i] = m512_const2_64( blocks << 56, 0 );
   }

-// digest final padding block and do output transform
+   // digest final padding block and do output transform
   TF512_4way( ctx->chaining, ctx->buffer );

   OF512_4way( ctx->chaining );
@@ -120,7 +122,7 @@ int groestl256_4way_update_close( groestl256_4way_context* ctx, void* output,
   const int hashlen_m128i = ctx->hashlen / 16;   // bytes to __m128i
   const int hash_offset = SIZE256 - hashlen_m128i;
   int rem = ctx->rem_ptr;
-   int blocks = len / SIZE256;
+   uint64_t blocks = len / SIZE256;
   __m512i* in = (__m512i*)input;
   int i;

@@ -144,20 +146,18 @@ int groestl256_4way_update_close( groestl256_4way_context* ctx, void* output,
   if ( i == SIZE256 - 1 )
   {
       // only 1 vector left in buffer, all padding at once
-       ctx->buffer[i] = m512_const1_128( _mm_set_epi8(
-                      blocks, blocks>>8,0,0, 0,0,0,0, 0,0,0,0, 0,0,0,0x80 ) );
+       ctx->buffer[i] = m512_const2_64( blocks << 56, 0x80 );
   }
   else
   {
       // add first padding
-       ctx->buffer[i] = m512_const4_64( 0, 0x80, 0, 0x80 );
+       ctx->buffer[i] = m512_const2_64( 0, 0x80 );
       // add zero padding
       for ( i += 1; i < SIZE256 - 1; i++ )
           ctx->buffer[i] = m512_zero;

       // add length padding, second last byte is zero unless blocks > 255
-       ctx->buffer[i] = m512_const1_128( _mm_set_epi8(
-                   blocks, blocks>>8, 0,0, 0,0, 0,0, 0,0, 0,0, 0,0, 0,0 ) );
+       ctx->buffer[i] = m512_const2_64( blocks << 56, 0 );
   }

 // digest final padding block and do output transform
@@ -172,5 +172,159 @@ int groestl256_4way_update_close( groestl256_4way_context* ctx, void* output,
   return 0;
 }

-#endif   // VAES
+#endif   // AVX512

+// AVX2 + VAES
+
+int groestl256_2way_init( groestl256_2way_context* ctx, uint64_t hashlen )
+{
+  int i;
+
+  ctx->hashlen = hashlen;
+
+  if (ctx->chaining == NULL || ctx->buffer == NULL)
+    return 1;
+
+  for ( i = 0; i < SIZE256; i++ )
+  {
+     ctx->chaining[i] = m256_zero;
+     ctx->buffer[i]   = m256_zero;
+  }
+
+  // The only non-zero in the IV is len. It can be hard coded.
+  ctx->chaining[ 3 ] = m256_const2_64( 0, 0x0100000000000000 );
+
+  ctx->buf_ptr = 0;
+  ctx->rem_ptr = 0;
+
+  return 0;
+}
+
+int groestl256_2way_full( groestl256_2way_context* ctx, void* output,
+                                const void* input, uint64_t datalen )
+{
+   const int len = (int)datalen >> 4;
+   const int hashlen_m128i = 32 >> 4;   // bytes to __m128i
+   const int hash_offset = SIZE256 - hashlen_m128i;
+   int rem = ctx->rem_ptr;
+   uint64_t blocks = len / SIZE256;
+   __m256i* in = (__m256i*)input;
+   int i;
+
+   if (ctx->chaining == NULL || ctx->buffer == NULL)
+     return 1;
+
+   for ( i = 0; i < SIZE256; i++ )
+   {
+     ctx->chaining[i] = m256_zero;
+     ctx->buffer[i]   = m256_zero;
+   }
+
+   // The only non-zero in the IV is len. It can be hard coded.
+   ctx->chaining[ 3 ] = m256_const2_64( 0, 0x0100000000000000 );
+   ctx->buf_ptr = 0;
+   ctx->rem_ptr = 0;
+
+   // --- update ---
+
+   // digest any full blocks, process directly from input 
+   for ( i = 0; i < blocks; i++ )
+      TF512_2way( ctx->chaining, &in[ i * SIZE256 ] );
+   ctx->buf_ptr = blocks * SIZE256;
+
+   // copy any remaining data to buffer, it may already contain data
+   // from a previous update for a midstate precalc
+   for ( i = 0; i < len % SIZE256; i++ )
+       ctx->buffer[ rem + i ] = in[ ctx->buf_ptr + i ];
+   i += rem;    // use i as rem_ptr in final
+
+   //--- final ---
+
+   blocks++;      // adjust for final block
+
+   if ( i == SIZE256 - 1 )
+   {
+       // only 1 vector left in buffer, all padding at once
+      ctx->buffer[i] = m256_const2_64( blocks << 56, 0x80 );
+   }
+   else
+   {
+       // add first padding
+       ctx->buffer[i] = m256_const2_64( 0, 0x80 );
+       // add zero padding
+       for ( i += 1; i < SIZE256 - 1; i++ )
+           ctx->buffer[i] = m256_zero;
+
+       // add length padding, second last byte is zero unless blocks > 255
+       ctx->buffer[i] = m256_const2_64( blocks << 56, 0 );
+   }
+
+   // digest final padding block and do output transform
+   TF512_2way( ctx->chaining, ctx->buffer );
+
+   OF512_2way( ctx->chaining );
+
+   // store hash result in output 
+   for ( i = 0; i < hashlen_m128i; i++ )
+      casti_m256i( output, i ) = ctx->chaining[ hash_offset + i ];
+
+   return 0;
+}
+int groestl256_2way_update_close( groestl256_2way_context* ctx, void* output,
+                                const void* input, uint64_t databitlen )
+{
+   const int len = (int)databitlen / 128;
+   const int hashlen_m128i = ctx->hashlen / 16;   // bytes to __m128i
+   const int hash_offset = SIZE256 - hashlen_m128i;
+   int rem = ctx->rem_ptr;
+   uint64_t blocks = len / SIZE256;
+   __m256i* in = (__m256i*)input;
+   int i;
+
+   // --- update ---
+
+   // digest any full blocks, process directly from input 
+   for ( i = 0; i < blocks; i++ )
+      TF512_2way( ctx->chaining, &in[ i * SIZE256 ] );
+   ctx->buf_ptr = blocks * SIZE256;
+
+   // copy any remaining data to buffer, it may already contain data
+   // from a previous update for a midstate precalc
+   for ( i = 0; i < len % SIZE256; i++ )
+       ctx->buffer[ rem + i ] = in[ ctx->buf_ptr + i ];
+   i += rem;    // use i as rem_ptr in final
+
+   //--- final ---
+
+   blocks++;      // adjust for final block
+
+   if ( i == SIZE256 - 1 )
+   {
+       // only 1 vector left in buffer, all padding at once
+       ctx->buffer[i] = m256_const2_64( blocks << 56, 0x80 );
+   }
+   else
+   {
+       // add first padding
+       ctx->buffer[i] = m256_const2_64( 0, 0x80 );
+       // add zero padding
+       for ( i += 1; i < SIZE256 - 1; i++ )
+           ctx->buffer[i] = m256_zero;
+
+       // add length padding, second last byte is zero unless blocks > 255
+       ctx->buffer[i] = m256_const2_64( blocks << 56, 0 );
+   }
+
+// digest final padding block and do output transform
+   TF512_2way( ctx->chaining, ctx->buffer );
+
+   OF512_2way( ctx->chaining );
+
+   // store hash result in output 
+   for ( i = 0; i < hashlen_m128i; i++ )
+      casti_m256i( output, i ) = ctx->chaining[ hash_offset + i ];
+
+   return 0;
+}
+
+#endif  // VAES
--- a/algo/groestl/groestl256-hash-4way.h
+++ b/algo/groestl/groestl256-hash-4way.h
@@ -18,8 +18,8 @@
 #endif
 #include <stdlib.h>

-#if defined(__VAES__) && defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
-   
+#if defined(__AVX2__) && defined(__VAES__)
+
 #define LENGTH (256)

 //#include "brg_endian.h"
@@ -48,6 +48,8 @@

 #define SIZE256 (SIZE_512/16)

+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+
 typedef struct {
  __attribute__ ((aligned (128))) __m512i chaining[SIZE256];
  __attribute__ ((aligned (64))) __m512i buffer[SIZE256];
@@ -55,7 +57,7 @@ typedef struct {
  int blk_count;     // SIZE_m128i
  int buf_ptr;       // __m128i offset
  int rem_ptr;
-  int databitlen;    // bits
+//  int databitlen;    // bits
 } groestl256_4way_context;


@@ -74,5 +76,25 @@ int groestl256_4way_update_close( groestl256_4way_context*,  void*,
 int groestl256_4way_full( groestl256_4way_context*, void*,
                          const void*, uint64_t );

-#endif
-#endif 
+#endif  // AVX512
+
+typedef struct {
+  __attribute__ ((aligned (128))) __m256i chaining[SIZE256];
+  __attribute__ ((aligned (64))) __m256i buffer[SIZE256];
+  int hashlen;       // byte
+  int blk_count;     // SIZE_m128i
+  int buf_ptr;       // __m128i offset
+  int rem_ptr;
+//  int databitlen;    // bits
+} groestl256_2way_context;
+
+int groestl256_2way_init( groestl256_2way_context*, uint64_t );
+
+int groestl256_2way_update_close( groestl256_2way_context*,  void*,
+                                        const void*, uint64_t );
+
+int groestl256_2way_full( groestl256_2way_context*, void*,
+                          const void*, uint64_t );
+
+#endif  // VAES
+#endif  // GROESTL256_HASH_4WAY_H__
--- a/algo/groestl/groestl256-intr-4way.h
+++ b/algo/groestl/groestl256-intr-4way.h
@@ -7,13 +7,13 @@
 * This code is placed in the public domain
 */

-
 #if !defined(GROESTL256_INTR_4WAY_H__)
 #define GROESTL256_INTR_4WAY_H__ 1
      
 #include "groestl256-hash-4way.h"

-#if defined(__VAES__)
+#if defined(__AVX2__) && defined(__VAES__)
+
 static const __m128i round_const_l0[] __attribute__ ((aligned (64))) =
 {
   { 0x7060504030201000, 0xffffffffffffffff },
@@ -42,6 +42,8 @@ static const __m128i round_const_l7[] __attribute__ ((aligned (64))) =
   { 0x0000000000000000, 0x8696a6b6c6d6e6f6 }
 };

+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+
 static const __m512i TRANSP_MASK = { 0x0d0509010c040800, 0x0f070b030e060a02,
                                     0x1d1519111c141810, 0x1f171b131e161a12,
                                     0x2d2529212c242820, 0x2f272b232e262a22,
@@ -499,5 +501,398 @@ void OF512_4way( __m512i* chaining )
  chaining[3] = xmm11;
 }

+#endif  // AVX512
+
+static const __m256i TRANSP_MASK_2WAY =
+             { 0x0d0509010c040800, 0x0f070b030e060a02,
+               0x1d1519111c141810, 0x1f171b131e161a12 };
+
+static const __m256i SUBSH_MASK0_2WAY =
+             { 0x0c0f0104070b0e00, 0x03060a0d08020509,
+               0x1c1f1114171b1e10, 0x13161a1d18121519 };
+
+static const __m256i SUBSH_MASK1_2WAY =
+             { 0x0e090205000d0801, 0x04070c0f0a03060b,
+               0x1e191215101d1801, 0x14171c1f1a13161b };
+
+static const __m256i SUBSH_MASK2_2WAY =
+               { 0x080b0306010f0a02, 0x05000e090c04070d,
+                 0x181b1316111f1a12, 0x15101e191c14171d };
+
+static const __m256i SUBSH_MASK3_2WAY =
+               { 0x0a0d040702090c03, 0x0601080b0e05000f,
+                 0x1a1d141712191c13, 0x1611181b1e15101f };
+
+static const __m256i SUBSH_MASK4_2WAY =
+               { 0x0b0e0500030a0d04, 0x0702090c0f060108,
+                 0x1b1e1510131a1d14, 0x1712191c1f161118 };
+
+static const __m256i SUBSH_MASK5_2WAY =
+               { 0x0d080601040c0f05, 0x00030b0e0907020a,
+                 0x1d181611141c1f15, 0x10131b1e1917121a };
+
+static const __m256i SUBSH_MASK6_2WAY =
+               { 0x0f0a0702050e0906, 0x01040d080b00030c,
+                 0x1f1a1712151e1916, 0x11141d181b10131c };
+
+static const __m256i SUBSH_MASK7_2WAY =
+               { 0x090c000306080b07, 0x02050f0a0d01040e,
+                 0x191c101316181b17, 0x12151f1a1d11141e, };
+
+#define tos(a)    #a
+#define tostr(a)  tos(a)
+
+/* xmm[i] will be multiplied by 2
+ * xmm[j] will be lost
+ * xmm[k] has to be all 0x1b */
+#define MUL2_2WAY(i, j, k){\
+  j = _mm256_xor_si256(j, j);\
+  j = _mm256_cmpgt_epi8(j, i );\
+  i = _mm256_add_epi8(i, i);\
+  j = _mm256_and_si256(j, k);\
+  i = _mm256_xor_si256(i, j);\
+}
+
+#define MixBytes_2way(a0, a1, a2, a3, a4, a5, a6, a7, b0, b1, b2, b3, b4, b5, b6, b7){\
+  /* t_i = a_i + a_{i+1} */\
+  b6 = a0;\
+  b7 = a1;\
+  a0 = _mm256_xor_si256(a0, a1);\
+  b0 = a2;\
+  a1 = _mm256_xor_si256(a1, a2);\
+  b1 = a3;\
+  a2 = _mm256_xor_si256(a2, a3);\
+  b2 = a4;\
+  a3 = _mm256_xor_si256(a3, a4);\
+  b3 = a5;\
+  a4 = _mm256_xor_si256(a4, a5);\
+  b4 = a6;\
+  a5 = _mm256_xor_si256(a5, a6);\
+  b5 = a7;\
+  a6 = _mm256_xor_si256(a6, a7);\
+  a7 = _mm256_xor_si256(a7, b6);\
+  \
+  /* build y4 y5 y6 ... in regs xmm8, xmm9, xmm10 by adding t_i*/\
+  b0 = _mm256_xor_si256(b0, a4);\
+  b6 = _mm256_xor_si256(b6, a4);\
+  b1 = _mm256_xor_si256(b1, a5);\
+  b7 = _mm256_xor_si256(b7, a5);\
+  b2 = _mm256_xor_si256(b2, a6);\
+  b0 = _mm256_xor_si256(b0, a6);\
+  /* spill values y_4, y_5 to memory */\
+  TEMP0 = b0;\
+  b3 = _mm256_xor_si256(b3, a7);\
+  b1 = _mm256_xor_si256(b1, a7);\
+  TEMP1 = b1;\
+  b4 = _mm256_xor_si256(b4, a0);\
+  b2 = _mm256_xor_si256(b2, a0);\
+  /* save values t0, t1, t2 to xmm8, xmm9 and memory */\
+  b0 = a0;\
+  b5 = _mm256_xor_si256(b5, a1);\
+  b3 = _mm256_xor_si256(b3, a1);\
+  b1 = a1;\
+  b6 = _mm256_xor_si256(b6, a2);\
+  b4 = _mm256_xor_si256(b4, a2);\
+  TEMP2 = a2;\
+  b7 = _mm256_xor_si256(b7, a3);\
+  b5 = _mm256_xor_si256(b5, a3);\
+  \
+  /* compute x_i = t_i + t_{i+3} */\
+  a0 = _mm256_xor_si256(a0, a3);\
+  a1 = _mm256_xor_si256(a1, a4);\
+  a2 = _mm256_xor_si256(a2, a5);\
+  a3 = _mm256_xor_si256(a3, a6);\
+  a4 = _mm256_xor_si256(a4, a7);\
+  a5 = _mm256_xor_si256(a5, b0);\
+  a6 = _mm256_xor_si256(a6, b1);\
+  a7 = _mm256_xor_si256(a7, TEMP2);\
+  \
+  /* compute z_i : double x_i using temp xmm8 and 1B xmm9 */\
+  /* compute w_i : add y_{i+4} */\
+  b1 = m256_const1_64( 0x1b1b1b1b1b1b1b1b );\
+  MUL2_2WAY(a0, b0, b1);\
+  a0 = _mm256_xor_si256(a0, TEMP0);\
+  MUL2_2WAY(a1, b0, b1);\
+  a1 = _mm256_xor_si256(a1, TEMP1);\
+  MUL2_2WAY(a2, b0, b1);\
+  a2 = _mm256_xor_si256(a2, b2);\
+  MUL2_2WAY(a3, b0, b1);\
+  a3 = _mm256_xor_si256(a3, b3);\
+  MUL2_2WAY(a4, b0, b1);\
+  a4 = _mm256_xor_si256(a4, b4);\
+  MUL2_2WAY(a5, b0, b1);\
+  a5 = _mm256_xor_si256(a5, b5);\
+  MUL2_2WAY(a6, b0, b1);\
+  a6 = _mm256_xor_si256(a6, b6);\
+  MUL2_2WAY(a7, b0, b1);\
+  a7 = _mm256_xor_si256(a7, b7);\
+  \
+  /* compute v_i : double w_i      */\
+  /* add to y_4 y_5 .. v3, v4, ... */\
+  MUL2_2WAY(a0, b0, b1);\
+  b5 = _mm256_xor_si256(b5, a0);\
+  MUL2_2WAY(a1, b0, b1);\
+  b6 = _mm256_xor_si256(b6, a1);\
+  MUL2_2WAY(a2, b0, b1);\
+  b7 = _mm256_xor_si256(b7, a2);\
+  MUL2_2WAY(a5, b0, b1);\
+  b2 = _mm256_xor_si256(b2, a5);\
+  MUL2_2WAY(a6, b0, b1);\
+  b3 = _mm256_xor_si256(b3, a6);\
+  MUL2_2WAY(a7, b0, b1);\
+  b4 = _mm256_xor_si256(b4, a7);\
+  MUL2_2WAY(a3, b0, b1);\
+  MUL2_2WAY(a4, b0, b1);\
+  b0 = TEMP0;\
+  b1 = TEMP1;\
+  b0 = _mm256_xor_si256(b0, a3);\
+  b1 = _mm256_xor_si256(b1, a4);\
+}/*MixBytes*/
+
+#define ROUND_2WAY(i, a0, a1, a2, a3, a4, a5, a6, a7, b0, b1, b2, b3, b4, b5, b6, b7){\
+  /* AddRoundConstant */\
+  b1 = m256_const2_64( 0xffffffffffffffff, 0 ); \
+  a0 = _mm256_xor_si256( a0, m256_const1_128( round_const_l0[i] ) );\
+  a1 = _mm256_xor_si256( a1, b1 );\
+  a2 = _mm256_xor_si256( a2, b1 );\
+  a3 = _mm256_xor_si256( a3, b1 );\
+  a4 = _mm256_xor_si256( a4, b1 );\
+  a5 = _mm256_xor_si256( a5, b1 );\
+  a6 = _mm256_xor_si256( a6, b1 );\
+  a7 = _mm256_xor_si256( a7, m256_const1_128( round_const_l7[i] ) );\
+  \
+  /* ShiftBytes + SubBytes (interleaved) */\
+  b0 = _mm256_xor_si256( b0, b0 );\
+  a0 = _mm256_shuffle_epi8( a0, SUBSH_MASK0_2WAY );\
+  a0 = _mm256_aesenclast_epi128(a0, b0 );\
+  a1 = _mm256_shuffle_epi8( a1, SUBSH_MASK1_2WAY );\
+  a1 = _mm256_aesenclast_epi128(a1, b0 );\
+  a2 = _mm256_shuffle_epi8( a2, SUBSH_MASK2_2WAY );\
+  a2 = _mm256_aesenclast_epi128(a2, b0 );\
+  a3 = _mm256_shuffle_epi8( a3, SUBSH_MASK3_2WAY );\
+  a3 = _mm256_aesenclast_epi128(a3, b0 );\
+  a4 = _mm256_shuffle_epi8( a4, SUBSH_MASK4_2WAY );\
+  a4 = _mm256_aesenclast_epi128(a4, b0 );\
+  a5 = _mm256_shuffle_epi8( a5, SUBSH_MASK5_2WAY );\
+  a5 = _mm256_aesenclast_epi128(a5, b0 );\
+  a6 = _mm256_shuffle_epi8( a6, SUBSH_MASK6_2WAY );\
+  a6 = _mm256_aesenclast_epi128(a6, b0 );\
+  a7 = _mm256_shuffle_epi8( a7, SUBSH_MASK7_2WAY );\
+  a7 = _mm256_aesenclast_epi128( a7, b0 );\
+  \
+  /* MixBytes */\
+  MixBytes_2way(a0, a1, a2, a3, a4, a5, a6, a7, b0, b1, b2, b3, b4, b5, b6, b7);\
+\
+}
+
+/* 10 rounds, P and Q in parallel */
+#define ROUNDS_P_Q_2WAY(){\
+  ROUND_2WAY(0, xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7);\
+  ROUND_2WAY(1, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7, xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15);\
+  ROUND_2WAY(2, xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7);\
+  ROUND_2WAY(3, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7, xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15);\
+  ROUND_2WAY(4, xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7);\
+  ROUND_2WAY(5, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7, xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15);\
+  ROUND_2WAY(6, xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7);\
+  ROUND_2WAY(7, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7, xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15);\
+  ROUND_2WAY(8, xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7);\
+  ROUND_2WAY(9, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7, xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15);\
+}
+
+#define Matrix_Transpose_A_2way(i0, i1, i2, i3, o1, o2, o3, t0){\
+  t0 = TRANSP_MASK_2WAY;\
+  \
+  i0 = _mm256_shuffle_epi8( i0, t0 );\
+  i1 = _mm256_shuffle_epi8( i1, t0 );\
+  i2 = _mm256_shuffle_epi8( i2, t0 );\
+  i3 = _mm256_shuffle_epi8( i3, t0 );\
+  \
+  o1 = i0;\
+  t0 = i2;\
+  \
+  i0 = _mm256_unpacklo_epi16( i0, i1 );\
+  o1 = _mm256_unpackhi_epi16( o1, i1 );\
+  i2 = _mm256_unpacklo_epi16( i2, i3 );\
+  t0 = _mm256_unpackhi_epi16( t0, i3 );\
+  \
+  i0 = _mm256_shuffle_epi32( i0, 216 );\
+  o1 = _mm256_shuffle_epi32( o1, 216 );\
+  i2 = _mm256_shuffle_epi32( i2, 216 );\
+  t0 = _mm256_shuffle_epi32( t0, 216 );\
+  \
+  o2 = i0;\
+  o3 = o1;\
+  \
+  i0 = _mm256_unpacklo_epi32( i0, i2 );\
+  o1 = _mm256_unpacklo_epi32( o1, t0 );\
+  o2 = _mm256_unpackhi_epi32( o2, i2 );\
+  o3 = _mm256_unpackhi_epi32( o3, t0 );\
+}/**/
+
+#define Matrix_Transpose_B_2way(i0, i1, i2, i3, i4, i5, i6, i7, o1, o2, o3, o4, o5, o6, o7){\
+  o1 = i0;\
+  o2 = i1;\
+  i0 = _mm256_unpacklo_epi64( i0, i4 );\
+  o1 = _mm256_unpackhi_epi64( o1, i4 );\
+  o3 = i1;\
+  o4 = i2;\
+  o2 = _mm256_unpacklo_epi64( o2, i5 );\
+  o3 = _mm256_unpackhi_epi64( o3, i5 );\
+  o5 = i2;\
+  o6 = i3;\
+  o4 = _mm256_unpacklo_epi64( o4, i6 );\
+  o5 = _mm256_unpackhi_epi64( o5, i6 );\
+  o7 = i3;\
+  o6 = _mm256_unpacklo_epi64( o6, i7 );\
+  o7 = _mm256_unpackhi_epi64( o7, i7 );\
+}/**/
+
+#define Matrix_Transpose_B_INV_2way(i0, i1, i2, i3, i4, i5, i6, i7, o0, o1, o2, o3){\
+  o0 = i0;\
+  i0 = _mm256_unpacklo_epi64( i0, i1 );\
+  o0 = _mm256_unpackhi_epi64( o0, i1 );\
+  o1 = i2;\
+  i2 = _mm256_unpacklo_epi64( i2, i3 );\
+  o1 = _mm256_unpackhi_epi64( o1, i3 );\
+  o2 = i4;\
+  i4 = _mm256_unpacklo_epi64( i4, i5 );\
+  o2 = _mm256_unpackhi_epi64( o2, i5 );\
+  o3 = i6;\
+  i6 = _mm256_unpacklo_epi64( i6, i7 );\
+  o3 = _mm256_unpackhi_epi64( o3, i7 );\
+}/**/
+
+#define Matrix_Transpose_O_B_2way(i0, i1, i2, i3, i4, i5, i6, i7, t0){\
+  t0 = _mm256_xor_si256( t0, t0 );\
+  i1 = i0;\
+  i3 = i2;\
+  i5 = i4;\
+  i7 = i6;\
+  i0 = _mm256_unpacklo_epi64( i0, t0 );\
+  i1 = _mm256_unpackhi_epi64( i1, t0 );\
+  i2 = _mm256_unpacklo_epi64( i2, t0 );\
+  i3 = _mm256_unpackhi_epi64( i3, t0 );\
+  i4 = _mm256_unpacklo_epi64( i4, t0 );\
+  i5 = _mm256_unpackhi_epi64( i5, t0 );\
+  i6 = _mm256_unpacklo_epi64( i6, t0 );\
+  i7 = _mm256_unpackhi_epi64( i7, t0 );\
+}/**/
+
+#define Matrix_Transpose_O_B_INV_2way(i0, i1, i2, i3, i4, i5, i6, i7){\
+  i0 = _mm256_unpacklo_epi64( i0, i1 );\
+  i2 = _mm256_unpacklo_epi64( i2, i3 );\
+  i4 = _mm256_unpacklo_epi64( i4, i5 );\
+  i6 = _mm256_unpacklo_epi64( i6, i7 );\
+}/**/
+
+void TF512_2way( __m256i* chaining, __m256i* message )
+{
+  static __m256i xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7;
+  static __m256i xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15;
+  static __m256i TEMP0;
+  static __m256i TEMP1;
+  static __m256i TEMP2;
+
+  /* load message into registers xmm12 - xmm15 */
+  xmm12 = message[0];
+  xmm13 = message[1];
+  xmm14 = message[2];
+  xmm15 = message[3];
+
+  /* transform message M from column ordering into row ordering */
+  /* we first put two rows (64 bit) of the message into one 128-bit xmm register */
+  Matrix_Transpose_A_2way(xmm12, xmm13, xmm14, xmm15, xmm2, xmm6, xmm7, xmm0);
+
+  /* load previous chaining value */
+  /* we first put two rows (64 bit) of the CV into one 128-bit xmm register */
+  xmm8 = chaining[0];
+  xmm0 = chaining[1];
+  xmm4 = chaining[2];
+  xmm5 = chaining[3];
+
+  /* xor message to CV get input of P */
+  /* result: CV+M in xmm8, xmm0, xmm4, xmm5 */
+  xmm8 = _mm256_xor_si256( xmm8, xmm12 );
+  xmm0 = _mm256_xor_si256( xmm0, xmm2 );
+  xmm4 = _mm256_xor_si256( xmm4, xmm6 );
+  xmm5 = _mm256_xor_si256( xmm5, xmm7 );
+
+  /* there are now 2 rows of the Groestl state (P and Q) in each xmm register */
+  /* unpack to get 1 row of P (64 bit) and Q (64 bit) into one xmm register */
+  /* result: the 8 rows of P and Q in xmm8 - xmm12 */
+  Matrix_Transpose_B_2way(xmm8, xmm0, xmm4, xmm5, xmm12, xmm2, xmm6, xmm7, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15);
+
+  /* compute the two permutations P and Q in parallel */
+  ROUNDS_P_Q_2WAY();
+
+  /* unpack again to get two rows of P or two rows of Q in one xmm register */
+  Matrix_Transpose_B_INV_2way(xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15, xmm0, xmm1, xmm2, xmm3);
+
+  /* xor output of P and Q */
+  /* result: P(CV+M)+Q(M) in xmm0...xmm3 */
+  xmm0 = _mm256_xor_si256( xmm0, xmm8 );
+  xmm1 = _mm256_xor_si256( xmm1, xmm10 );
+  xmm2 = _mm256_xor_si256( xmm2, xmm12 );
+  xmm3 = _mm256_xor_si256( xmm3, xmm14 );
+
+  /* xor CV (feed-forward) */
+  /* result: P(CV+M)+Q(M)+CV in xmm0...xmm3 */
+  xmm0 = _mm256_xor_si256( xmm0, (chaining[0]) );
+  xmm1 = _mm256_xor_si256( xmm1, (chaining[1]) );
+  xmm2 = _mm256_xor_si256( xmm2, (chaining[2]) );
+  xmm3 = _mm256_xor_si256( xmm3, (chaining[3]) );
+
+  /* store CV */
+  chaining[0] = xmm0;
+  chaining[1] = xmm1;
+  chaining[2] = xmm2;
+  chaining[3] = xmm3;
+
+  return;
+}
+  
+void OF512_2way( __m256i* chaining )
+{
+  static __m256i xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7;
+  static __m256i xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15;
+  static __m256i TEMP0;
+  static __m256i TEMP1;
+  static __m256i TEMP2;
+
+  /* load CV into registers xmm8, xmm10, xmm12, xmm14 */
+  xmm8 = chaining[0];
+  xmm10 = chaining[1];
+  xmm12 = chaining[2];
+  xmm14 = chaining[3];
+
+  /* there are now 2 rows of the CV in one xmm register */
+  /* unpack to get 1 row of P (64 bit) into one half of an xmm register */
+  /* result: the 8 input rows of P in xmm8 - xmm15 */
+  Matrix_Transpose_O_B_2way(xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15, xmm0);
+
+  /* compute the permutation P */
+  /* result: the output of P(CV) in xmm8 - xmm15 */
+  ROUNDS_P_Q_2WAY();
+
+  /* unpack again to get two rows of P in one xmm register */
+  /* result: P(CV) in xmm8, xmm10, xmm12, xmm14 */
+  Matrix_Transpose_O_B_INV_2way(xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15);
+
+  /* xor CV to P output (feed-forward) */
+  /* result: P(CV)+CV in xmm8, xmm10, xmm12, xmm14 */
+  xmm8  = _mm256_xor_si256( xmm8,  (chaining[0]) );
+  xmm10 = _mm256_xor_si256( xmm10, (chaining[1]) );
+  xmm12 = _mm256_xor_si256( xmm12, (chaining[2]) );
+  xmm14 = _mm256_xor_si256( xmm14, (chaining[3]) );
+
+  /* transform state back from row ordering into column ordering */
+  /* result: final hash value in xmm9, xmm11 */
+  Matrix_Transpose_A_2way(xmm8, xmm10, xmm12, xmm14, xmm4, xmm9, xmm11, xmm0);
+
+  /* we only need to return the truncated half of the state */
+  chaining[2] = xmm9;
+  chaining[3] = xmm11;
+}
+
 #endif  // VAES
-#endif  // GROESTL512_INTR_4WAY_H__
+#endif  // GROESTL256_INTR_4WAY_H__
--- a/algo/groestl/groestl512-hash-4way.c
+++ b/algo/groestl/groestl512-hash-4way.c
@@ -15,7 +15,9 @@
 #include "miner.h"
 #include "simd-utils.h"

-#if defined(__VAES__) && defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+#if defined(__AVX2__) && defined(__VAES__)
+
+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)

 int groestl512_4way_init( groestl512_4way_context* ctx, uint64_t hashlen )
 {
@@ -41,7 +43,7 @@ int groestl512_4way_update_close( groestl512_4way_context* ctx, void* output,
   const int hashlen_m128i = 64 / 16;   // bytes to __m128i
   const int hash_offset = SIZE512 - hashlen_m128i;
   int rem = ctx->rem_ptr;
-   int blocks = len / SIZE512;
+   uint64_t blocks = len / SIZE512;
   __m512i* in = (__m512i*)input;
   int i;

@@ -62,16 +64,14 @@ int groestl512_4way_update_close( groestl512_4way_context* ctx, void* output,
   if ( i == SIZE512 - 1 )
   {        
       // only 1 vector left in buffer, all padding at once
-       ctx->buffer[i] = m512_const1_128( _mm_set_epi8(
-                      blocks, blocks>>8,0,0, 0,0,0,0, 0,0,0,0, 0,0,0,0x80 ) );
+       ctx->buffer[i] = m512_const2_64( blocks << 56, 0x80 );
   }   
   else
   {
-       ctx->buffer[i] = m512_const4_64( 0, 0x80, 0, 0x80 );
+       ctx->buffer[i] = m512_const2_64( 0, 0x80 );
       for ( i += 1; i < SIZE512 - 1; i++ )
           ctx->buffer[i] = m512_zero;
-       ctx->buffer[i] = m512_const1_128( _mm_set_epi8(
-                   blocks, blocks>>8, 0,0, 0,0, 0,0, 0,0, 0,0, 0,0, 0,0 ) );
+       ctx->buffer[i] = m512_const2_64( blocks << 56, 0 );
   }

   TF1024_4way( ctx->chaining, ctx->buffer );
@@ -122,7 +122,7 @@ int groestl512_4way_full( groestl512_4way_context* ctx, void* output,
   }
   else
   {
-       ctx->buffer[i] = m512_const4_64( 0, 0x80, 0, 0x80 );
+       ctx->buffer[i] = m512_const2_64( 0, 0x80 );
       for ( i += 1; i < SIZE512 - 1; i++ )
           ctx->buffer[i] = m512_zero;
       ctx->buffer[i] = m512_const2_64( blocks << 56, 0 );
@@ -137,5 +137,128 @@ int groestl512_4way_full( groestl512_4way_context* ctx, void* output,
   return 0;
 }

+#endif   // AVX512
+
+
+// AVX2 + VAES
+
+int groestl512_2way_init( groestl512_2way_context* ctx, uint64_t hashlen )
+{
+  if (ctx->chaining == NULL || ctx->buffer == NULL)
+    return 1;
+
+  memset_zero_256( ctx->chaining, SIZE512 );
+  memset_zero_256( ctx->buffer, SIZE512 );
+
+  // The only non-zero in the IV is len. It can be hard coded.
+  ctx->chaining[ 6 ] = m256_const2_64( 0x0200000000000000, 0 );
+
+  ctx->buf_ptr = 0;
+  ctx->rem_ptr = 0;
+
+  return 0;
+}
+
+int groestl512_2way_update_close( groestl512_2way_context* ctx, void* output,
+                                const void* input, uint64_t databitlen )
+{
+   const int len = (int)databitlen / 128;
+   const int hashlen_m128i = 64 / 16;   // bytes to __m128i
+   const int hash_offset = SIZE512 - hashlen_m128i;
+   int rem = ctx->rem_ptr;
+   uint64_t blocks = len / SIZE512;
+   __m256i* in = (__m256i*)input;
+   int i;
+
+   // --- update ---
+
+   for ( i = 0; i < blocks; i++ )
+      TF1024_2way( ctx->chaining, &in[ i * SIZE512 ] );
+   ctx->buf_ptr = blocks * SIZE512;
+
+   for ( i = 0; i < len % SIZE512; i++ )
+       ctx->buffer[ rem + i ] = in[ ctx->buf_ptr + i ];
+   i += rem;
+
+   //--- final ---
+
+   blocks++;      // adjust for final block
+
+   if ( i == SIZE512 - 1 )
+   {
+       // only 1 vector left in buffer, all padding at once
+       ctx->buffer[i] = m256_const2_64( blocks << 56, 0x80 );
+   }
+   else
+   {
+       ctx->buffer[i] = m256_const2_64( 0, 0x80 );
+       for ( i += 1; i < SIZE512 - 1; i++ )
+           ctx->buffer[i] = m256_zero;
+       ctx->buffer[i] = m256_const2_64( blocks << 56, 0 );
+   }
+
+   TF1024_2way( ctx->chaining, ctx->buffer );
+   OF1024_2way( ctx->chaining );
+
+   for ( i = 0; i < hashlen_m128i; i++ )
+      casti_m256i( output, i ) = ctx->chaining[ hash_offset + i ];
+
+   return 0;
+}
+
+int groestl512_2way_full( groestl512_2way_context* ctx, void* output,
+                          const void* input, uint64_t datalen )
+{
+   const int len = (int)datalen >> 4;
+   const int hashlen_m128i = 64 >> 4;   // bytes to __m128i
+   const int hash_offset = SIZE512 - hashlen_m128i;
+   uint64_t blocks = len / SIZE512;
+   __m256i* in = (__m256i*)input;
+   int i;
+
+   // --- init ---
+
+   memset_zero_256( ctx->chaining, SIZE512 );
+   memset_zero_256( ctx->buffer, SIZE512 );
+   ctx->chaining[ 6 ] = m256_const2_64( 0x0200000000000000, 0 );
+   ctx->buf_ptr = 0;
+   ctx->rem_ptr = 0;
+
+   // --- update ---
+
+   for ( i = 0; i < blocks; i++ )
+      TF1024_2way( ctx->chaining, &in[ i * SIZE512 ] );
+   ctx->buf_ptr = blocks * SIZE512;
+
+   for ( i = 0; i < len % SIZE512; i++ )
+       ctx->buffer[ ctx->rem_ptr + i ] = in[ ctx->buf_ptr + i ];
+   i += ctx->rem_ptr;
+
+   // --- close ---
+
+   blocks++;
+
+   if ( i == SIZE512 - 1 )
+   {
+       // only 1 vector left in buffer, all padding at once
+       ctx->buffer[i] = m256_const2_64( blocks << 56, 0x80 );
+   }
+   else
+   {
+       ctx->buffer[i] = m256_const2_64( 0, 0x80 );
+       for ( i += 1; i < SIZE512 - 1; i++ )
+           ctx->buffer[i] = m256_zero;
+       ctx->buffer[i] = m256_const2_64( blocks << 56, 0 );
+   }
+
+   TF1024_2way( ctx->chaining, ctx->buffer );
+   OF1024_2way( ctx->chaining );
+
+   for ( i = 0; i < hashlen_m128i; i++ )
+      casti_m256i( output, i ) = ctx->chaining[ hash_offset + i ];
+
+   return 0;
+}
+   
 #endif   // VAES

--- a/algo/groestl/groestl512-hash-4way.h
+++ b/algo/groestl/groestl512-hash-4way.h
@@ -10,7 +10,7 @@
 #endif
 #include <stdlib.h>

-#if defined(__VAES__) && defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+#if defined(__AVX2__) && defined(__VAES__)

 #define LENGTH (512)

@@ -36,20 +36,19 @@

 #define SIZE512 (SIZE_1024/16)

+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+
 typedef struct {
  __attribute__ ((aligned (128))) __m512i chaining[SIZE512];
  __attribute__ ((aligned (64))) __m512i buffer[SIZE512];
  int blk_count;     // SIZE_m128i
  int buf_ptr;       // __m128i offset
  int rem_ptr;
-  int databitlen;    // bits
 } groestl512_4way_context;


 int groestl512_4way_init( groestl512_4way_context*, uint64_t );

-//int reinit_groestl( hashState_groestl* );
-
 int groestl512_4way_update( groestl512_4way_context*, const void*,
                              uint64_t );
 int groestl512_4way_close( groestl512_4way_context*, void* );
@@ -58,5 +57,29 @@ int groestl512_4way_update_close( groestl512_4way_context*,  void*,
 int groestl512_4way_full( groestl512_4way_context*,  void*,
                          const void*, uint64_t );

+#endif   // AVX512
+
+// AVX2 + VAES
+
+typedef struct {
+  __attribute__ ((aligned (128))) __m256i chaining[SIZE512];
+  __attribute__ ((aligned (64))) __m256i buffer[SIZE512];
+  int blk_count;     // SIZE_m128i
+  int buf_ptr;       // __m128i offset
+  int rem_ptr;
+} groestl512_2way_context;
+
+
+int groestl512_2way_init( groestl512_2way_context*, uint64_t );
+
+int groestl512_2way_update( groestl512_2way_context*, const void*,
+                              uint64_t );
+int groestl512_2way_close( groestl512_2way_context*, void* );
+int groestl512_2way_update_close( groestl512_2way_context*,  void*,
+                                        const void*, uint64_t );
+int groestl512_2way_full( groestl512_2way_context*,  void*,
+                          const void*, uint64_t );
+
+
 #endif   // VAES
 #endif   // GROESTL512_HASH_4WAY_H__
--- a/algo/groestl/groestl512-intr-4way.h
+++ b/algo/groestl/groestl512-intr-4way.h
@@ -7,13 +7,12 @@
 * This code is placed in the public domain
 */

-
 #if !defined(GROESTL512_INTR_4WAY_H__)
 #define GROESTL512_INTR_4WAY_H__ 1
      
 #include "groestl512-hash-4way.h"

-#if defined(__VAES__)
+#if defined(__AVX2__) && defined(__VAES__)

 static const __m128i round_const_p[] __attribute__ ((aligned (64))) =
 {
@@ -51,6 +50,8 @@ static const __m128i round_const_q[] __attribute__ ((aligned (64))) =
   { 0x8292a2b2c2d2e2f2, 0x0212223242526272 }
 };

+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+
 static const __m512i TRANSP_MASK = { 0x0d0509010c040800, 0x0f070b030e060a02,
                                     0x1d1519111c141810, 0x1f171b131e161a12,
                                     0x2d2529212c242820, 0x2f272b232e262a22,
@@ -661,5 +662,578 @@ void OF1024_4way( __m512i* chaining )
  return;
 }

+#endif  // AVX512
+
+// AVX2 + VAES
+
+static const __m256i TRANSP_MASK_2WAY =
+    { 0x0d0509010c040800, 0x0f070b030e060a02,
+      0x1d1519111c141810, 0x1f171b131e161a12 };
+
+static const __m256i SUBSH_MASK0_2WAY =
+    { 0x0b0e0104070a0d00, 0x0306090c0f020508,
+      0x1b1e1114171a1d10, 0x1316191c1f121518 };
+
+static const __m256i SUBSH_MASK1_2WAY =
+    { 0x0c0f0205080b0e01, 0x04070a0d00030609,
+      0x1c1f1215181b1e11, 0x14171a1d10131619 };
+
+static const __m256i SUBSH_MASK2_2WAY =
+    { 0x0d000306090c0f02, 0x05080b0e0104070a,
+      0x1d101316191c1f12, 0x15181b1e1114171a };
+
+static const __m256i SUBSH_MASK3_2WAY =
+    { 0x0e0104070a0d0003, 0x06090c0f0205080b,
+      0x1e1114171a1d1013, 0x16191c1f1215181b };
+
+static const __m256i SUBSH_MASK4_2WAY = 
+    { 0x0f0205080b0e0104, 0x070a0d000306090c,
+      0x1f1215181b1e1114, 0x171a1d101316191c };
+
+static const __m256i SUBSH_MASK5_2WAY =
+    { 0x000306090c0f0205, 0x080b0e0104070a0d,
+      0x101316191c1f1215, 0x181b1e1114171a1d };
+
+static const __m256i SUBSH_MASK6_2WAY =
+    { 0x0104070a0d000306, 0x090c0f0205080b0e,
+      0x1114171a1d101316, 0x191c1f1215181b1e };
+
+static const __m256i SUBSH_MASK7_2WAY =
+    { 0x06090c0f0205080b, 0x0e0104070a0d0003,
+      0x16191c1f1215181b, 0x1e1114171a1d1013 };
+
+#define tos(a)    #a
+#define tostr(a)  tos(a)
+
+/* xmm[i] will be multiplied by 2
+ * xmm[j] will be lost
+ * xmm[k] has to be all 0x1b */
+#define MUL2_2WAY(i, j, k){\
+  j = _mm256_xor_si256(j, j);\
+  j = _mm256_cmpgt_epi8(j, i );\
+  i = _mm256_add_epi8(i, i);\
+  j = _mm256_and_si256(j, k);\
+  i = _mm256_xor_si256(i, j);\
+}
+
+#define MixBytes_2way(a0, a1, a2, a3, a4, a5, a6, a7, b0, b1, b2, b3, b4, b5, b6, b7){\
+  /* t_i = a_i + a_{i+1} */\
+  b6 = a0;\
+  b7 = a1;\
+  a0 = _mm256_xor_si256(a0, a1);\
+  b0 = a2;\
+  a1 = _mm256_xor_si256(a1, a2);\
+  b1 = a3;\
+  a2 = _mm256_xor_si256(a2, a3);\
+  b2 = a4;\
+  a3 = _mm256_xor_si256(a3, a4);\
+  b3 = a5;\
+  a4 = _mm256_xor_si256(a4, a5);\
+  b4 = a6;\
+  a5 = _mm256_xor_si256(a5, a6);\
+  b5 = a7;\
+  a6 = _mm256_xor_si256(a6, a7);\
+  a7 = _mm256_xor_si256(a7, b6);\
+  \
+  /* build y4 y5 y6 ... in regs xmm8, xmm9, xmm10 by adding t_i*/\
+  b0 = _mm256_xor_si256(b0, a4);\
+  b6 = _mm256_xor_si256(b6, a4);\
+  b1 = _mm256_xor_si256(b1, a5);\
+  b7 = _mm256_xor_si256(b7, a5);\
+  b2 = _mm256_xor_si256(b2, a6);\
+  b0 = _mm256_xor_si256(b0, a6);\
+  /* spill values y_4, y_5 to memory */\
+  TEMP0 = b0;\
+  b3 = _mm256_xor_si256(b3, a7);\
+  b1 = _mm256_xor_si256(b1, a7);\
+  TEMP1 = b1;\
+  b4 = _mm256_xor_si256(b4, a0);\
+  b2 = _mm256_xor_si256(b2, a0);\
+  /* save values t0, t1, t2 to xmm8, xmm9 and memory */\
+  b0 = a0;\
+  b5 = _mm256_xor_si256(b5, a1);\
+  b3 = _mm256_xor_si256(b3, a1);\
+  b1 = a1;\
+  b6 = _mm256_xor_si256(b6, a2);\
+  b4 = _mm256_xor_si256(b4, a2);\
+  TEMP2 = a2;\
+  b7 = _mm256_xor_si256(b7, a3);\
+  b5 = _mm256_xor_si256(b5, a3);\
+  \
+  /* compute x_i = t_i + t_{i+3} */\
+  a0 = _mm256_xor_si256(a0, a3);\
+  a1 = _mm256_xor_si256(a1, a4);\
+  a2 = _mm256_xor_si256(a2, a5);\
+  a3 = _mm256_xor_si256(a3, a6);\
+  a4 = _mm256_xor_si256(a4, a7);\
+  a5 = _mm256_xor_si256(a5, b0);\
+  a6 = _mm256_xor_si256(a6, b1);\
+  a7 = _mm256_xor_si256(a7, TEMP2);\
+  \
+  /* compute z_i : double x_i using temp xmm8 and 1B xmm9 */\
+  /* compute w_i : add y_{i+4} */\
+  b1 = m256_const1_64( 0x1b1b1b1b1b1b1b1b );\
+  MUL2_2WAY(a0, b0, b1);\
+  a0 = _mm256_xor_si256(a0, TEMP0);\
+  MUL2_2WAY(a1, b0, b1);\
+  a1 = _mm256_xor_si256(a1, TEMP1);\
+  MUL2_2WAY(a2, b0, b1);\
+  a2 = _mm256_xor_si256(a2, b2);\
+  MUL2_2WAY(a3, b0, b1);\
+  a3 = _mm256_xor_si256(a3, b3);\
+  MUL2_2WAY(a4, b0, b1);\
+  a4 = _mm256_xor_si256(a4, b4);\
+  MUL2_2WAY(a5, b0, b1);\
+  a5 = _mm256_xor_si256(a5, b5);\
+  MUL2_2WAY(a6, b0, b1);\
+  a6 = _mm256_xor_si256(a6, b6);\
+  MUL2_2WAY(a7, b0, b1);\
+  a7 = _mm256_xor_si256(a7, b7);\
+  \
+  /* compute v_i : double w_i      */\
+  /* add to y_4 y_5 .. v3, v4, ... */\
+  MUL2_2WAY(a0, b0, b1);\
+  b5 = _mm256_xor_si256(b5, a0);\
+  MUL2_2WAY(a1, b0, b1);\
+  b6 = _mm256_xor_si256(b6, a1);\
+  MUL2_2WAY(a2, b0, b1);\
+  b7 = _mm256_xor_si256(b7, a2);\
+  MUL2_2WAY(a5, b0, b1);\
+  b2 = _mm256_xor_si256(b2, a5);\
+  MUL2_2WAY(a6, b0, b1);\
+  b3 = _mm256_xor_si256(b3, a6);\
+  MUL2_2WAY(a7, b0, b1);\
+  b4 = _mm256_xor_si256(b4, a7);\
+  MUL2_2WAY(a3, b0, b1);\
+  MUL2_2WAY(a4, b0, b1);\
+  b0 = TEMP0;\
+  b1 = TEMP1;\
+  b0 = _mm256_xor_si256(b0, a3);\
+  b1 = _mm256_xor_si256(b1, a4);\
+}/*MixBytes*/
+
+/* one round
+ * a0-a7 = input rows
+ * b0-b7 = output rows
+ */
+#define SUBMIX_2WAY(a0, a1, a2, a3, a4, a5, a6, a7, b0, b1, b2, b3, b4, b5, b6, b7){\
+  /* SubBytes */\
+  b0 = _mm256_xor_si256( b0, b0 );\
+  a0 = _mm256_aesenclast_epi128( a0, b0 );\
+  a1 = _mm256_aesenclast_epi128( a1, b0 );\
+  a2 = _mm256_aesenclast_epi128( a2, b0 );\
+  a3 = _mm256_aesenclast_epi128( a3, b0 );\
+  a4 = _mm256_aesenclast_epi128( a4, b0 );\
+  a5 = _mm256_aesenclast_epi128( a5, b0 );\
+  a6 = _mm256_aesenclast_epi128( a6, b0 );\
+  a7 = _mm256_aesenclast_epi128( a7, b0 );\
+  /* MixBytes */\
+  MixBytes_2way(a0, a1, a2, a3, a4, a5, a6, a7, b0, b1, b2, b3, b4, b5, b6, b7);\
+}
+
+#define ROUNDS_P_2WAY(){\
+  uint8_t round_counter = 0;\
+  for ( round_counter = 0; round_counter < 14; round_counter += 2 ) \
+  { \
+    /* AddRoundConstant P1024 */\
+    xmm8 = _mm256_xor_si256( xmm8, m256_const1_128( \
+             casti_m128i( round_const_p, round_counter ) ) ); \
+    /* ShiftBytes P1024 + pre-AESENCLAST */\
+    xmm8  = _mm256_shuffle_epi8( xmm8,  SUBSH_MASK0_2WAY ); \
+    xmm9  = _mm256_shuffle_epi8( xmm9,  SUBSH_MASK1_2WAY );\
+    xmm10 = _mm256_shuffle_epi8( xmm10, SUBSH_MASK2_2WAY );\
+    xmm11 = _mm256_shuffle_epi8( xmm11, SUBSH_MASK3_2WAY );\
+    xmm12 = _mm256_shuffle_epi8( xmm12, SUBSH_MASK4_2WAY );\
+    xmm13 = _mm256_shuffle_epi8( xmm13, SUBSH_MASK5_2WAY );\
+    xmm14 = _mm256_shuffle_epi8( xmm14, SUBSH_MASK6_2WAY );\
+    xmm15 = _mm256_shuffle_epi8( xmm15, SUBSH_MASK7_2WAY );\
+    /* SubBytes + MixBytes */\
+    SUBMIX_2WAY(xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7);\
+    \
+     /* AddRoundConstant P1024 */\
+    xmm0 = _mm256_xor_si256( xmm0, m256_const1_128( \
+             casti_m128i( round_const_p, round_counter+1 ) ) ); \
+    /* ShiftBytes P1024 + pre-AESENCLAST */\
+    xmm0 = _mm256_shuffle_epi8( xmm0, SUBSH_MASK0_2WAY );\
+    xmm1 = _mm256_shuffle_epi8( xmm1, SUBSH_MASK1_2WAY );\
+    xmm2 = _mm256_shuffle_epi8( xmm2, SUBSH_MASK2_2WAY );\
+    xmm3 = _mm256_shuffle_epi8( xmm3, SUBSH_MASK3_2WAY );\
+    xmm4 = _mm256_shuffle_epi8( xmm4, SUBSH_MASK4_2WAY );\
+    xmm5 = _mm256_shuffle_epi8( xmm5, SUBSH_MASK5_2WAY );\
+    xmm6 = _mm256_shuffle_epi8( xmm6, SUBSH_MASK6_2WAY );\
+    xmm7 = _mm256_shuffle_epi8( xmm7, SUBSH_MASK7_2WAY );\
+    /* SubBytes + MixBytes */\
+     SUBMIX_2WAY(xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7, xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15);\
+  }\
+}
+
+#define ROUNDS_Q_2WAY(){\
+  uint8_t round_counter = 0;\
+  for ( round_counter = 0; round_counter < 14; round_counter += 2) \
+  { \
+    /* AddRoundConstant Q1024 */\
+    xmm1 = m256_neg1;\
+    xmm8  = _mm256_xor_si256( xmm8,  xmm1 );\
+    xmm9  = _mm256_xor_si256( xmm9,  xmm1 );\
+    xmm10 = _mm256_xor_si256( xmm10, xmm1 );\
+    xmm11 = _mm256_xor_si256( xmm11, xmm1 );\
+    xmm12 = _mm256_xor_si256( xmm12, xmm1 );\
+    xmm13 = _mm256_xor_si256( xmm13, xmm1 );\
+    xmm14 = _mm256_xor_si256( xmm14, xmm1 );\
+    xmm15 = _mm256_xor_si256( xmm15, m256_const1_128( \
+                 casti_m128i( round_const_q, round_counter ) ) ); \
+    /* ShiftBytes Q1024 + pre-AESENCLAST */\
+    xmm8  = _mm256_shuffle_epi8( xmm8,  SUBSH_MASK1_2WAY );\
+    xmm9  = _mm256_shuffle_epi8( xmm9,  SUBSH_MASK3_2WAY );\
+    xmm10 = _mm256_shuffle_epi8( xmm10, SUBSH_MASK5_2WAY );\
+    xmm11 = _mm256_shuffle_epi8( xmm11, SUBSH_MASK7_2WAY );\
+    xmm12 = _mm256_shuffle_epi8( xmm12, SUBSH_MASK0_2WAY );\
+    xmm13 = _mm256_shuffle_epi8( xmm13, SUBSH_MASK2_2WAY );\
+    xmm14 = _mm256_shuffle_epi8( xmm14, SUBSH_MASK4_2WAY );\
+    xmm15 = _mm256_shuffle_epi8( xmm15, SUBSH_MASK6_2WAY );\
+    /* SubBytes + MixBytes */\
+    SUBMIX_2WAY(xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7);\
+    \
+    /* AddRoundConstant Q1024 */\
+    xmm9 = m256_neg1;\
+    xmm0 = _mm256_xor_si256( xmm0, xmm9 );\
+    xmm1 = _mm256_xor_si256( xmm1, xmm9 );\
+    xmm2 = _mm256_xor_si256( xmm2, xmm9 );\
+    xmm3 = _mm256_xor_si256( xmm3, xmm9 );\
+    xmm4 = _mm256_xor_si256( xmm4, xmm9 );\
+    xmm5 = _mm256_xor_si256( xmm5, xmm9 );\
+    xmm6 = _mm256_xor_si256( xmm6, xmm9 );\
+    xmm7 = _mm256_xor_si256( xmm7, m256_const1_128( \
+             casti_m128i( round_const_q, round_counter+1 ) ) ); \
+    /* ShiftBytes Q1024 + pre-AESENCLAST */\
+    xmm0 = _mm256_shuffle_epi8( xmm0, SUBSH_MASK1_2WAY );\
+    xmm1 = _mm256_shuffle_epi8( xmm1, SUBSH_MASK3_2WAY );\
+    xmm2 = _mm256_shuffle_epi8( xmm2, SUBSH_MASK5_2WAY );\
+    xmm3 = _mm256_shuffle_epi8( xmm3, SUBSH_MASK7_2WAY );\
+    xmm4 = _mm256_shuffle_epi8( xmm4, SUBSH_MASK0_2WAY );\
+    xmm5 = _mm256_shuffle_epi8( xmm5, SUBSH_MASK2_2WAY );\
+    xmm6 = _mm256_shuffle_epi8( xmm6, SUBSH_MASK4_2WAY );\
+    xmm7 = _mm256_shuffle_epi8( xmm7, SUBSH_MASK6_2WAY );\
+    /* SubBytes + MixBytes */\
+    SUBMIX_2WAY(xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7, xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15);\
+  }\
+}
+
+#define Matrix_Transpose_2way(i0, i1, i2, i3, i4, i5, i6, i7, t0, t1, t2, t3, t4, t5, t6, t7){\
+  t0 = TRANSP_MASK_2WAY;\
+\
+  i6 = _mm256_shuffle_epi8(i6, t0);\
+  i0 = _mm256_shuffle_epi8(i0, t0);\
+  i1 = _mm256_shuffle_epi8(i1, t0);\
+  i2 = _mm256_shuffle_epi8(i2, t0);\
+  i3 = _mm256_shuffle_epi8(i3, t0);\
+  t1 = i2;\
+  i4 = _mm256_shuffle_epi8(i4, t0);\
+  i5 = _mm256_shuffle_epi8(i5, t0);\
+  t2 = i4;\
+  t3 = i6;\
+  i7 = _mm256_shuffle_epi8(i7, t0);\
+\
+  /* continue with unpack using 4 temp registers */\
+  t0 = i0;\
+  t2 = _mm256_unpackhi_epi16(t2, i5);\
+  i4 = _mm256_unpacklo_epi16(i4, i5);\
+  t3 = _mm256_unpackhi_epi16(t3, i7);\
+  i6 = _mm256_unpacklo_epi16(i6, i7);\
+  t0 = _mm256_unpackhi_epi16(t0, i1);\
+  t1 = _mm256_unpackhi_epi16(t1, i3);\
+  i2 = _mm256_unpacklo_epi16(i2, i3);\
+  i0 = _mm256_unpacklo_epi16(i0, i1);\
+\
+  /* shuffle with immediate */\
+  t0 = _mm256_shuffle_epi32(t0, 216);\
+  t1 = _mm256_shuffle_epi32(t1, 216);\
+  t2 = _mm256_shuffle_epi32(t2, 216);\
+  t3 = _mm256_shuffle_epi32(t3, 216);\
+  i0 = _mm256_shuffle_epi32(i0, 216);\
+  i2 = _mm256_shuffle_epi32(i2, 216);\
+  i4 = _mm256_shuffle_epi32(i4, 216);\
+  i6 = _mm256_shuffle_epi32(i6, 216);\
+\
+  /* continue with unpack */\
+  t4 = i0;\
+  i0 = _mm256_unpacklo_epi32(i0, i2);\
+  t4 = _mm256_unpackhi_epi32(t4, i2);\
+  t5 = t0;\
+  t0 = _mm256_unpacklo_epi32(t0, t1);\
+  t5 = _mm256_unpackhi_epi32(t5, t1);\
+  t6 = i4;\
+  i4 = _mm256_unpacklo_epi32(i4, i6);\
+  t7 = t2;\
+  t6 = _mm256_unpackhi_epi32(t6, i6);\
+  i2 = t0;\
+  t2 = _mm256_unpacklo_epi32(t2, t3);\
+  i3 = t0;\
+  t7 = _mm256_unpackhi_epi32(t7, t3);\
+\
+  /* there are now 2 rows in each xmm */\
+  /* unpack to get 1 row of CV in each xmm */\
+  i1 = i0;\
+  i1 = _mm256_unpackhi_epi64(i1, i4);\
+  i0 = _mm256_unpacklo_epi64(i0, i4);\
+  i4 = t4;\
+  i3 = _mm256_unpackhi_epi64(i3, t2);\
+  i5 = t4;\
+  i2 = _mm256_unpacklo_epi64(i2, t2);\
+  i6 = t5;\
+  i5 = _mm256_unpackhi_epi64(i5, t6);\
+  i7 = t5;\
+  i4 = _mm256_unpacklo_epi64(i4, t6);\
+  i7 = _mm256_unpackhi_epi64(i7, t7);\
+  i6 = _mm256_unpacklo_epi64(i6, t7);\
+  /* transpose done */\
+}/**/
+
+#define Matrix_Transpose_INV_2way(i0, i1, i2, i3, i4, i5, i6, i7, o0, o1, o2, t0, t1, t2, t3, t4){\
+  /*  transpose matrix to get output format */\
+  o1 = i0;\
+  i0 = _mm256_unpacklo_epi64(i0, i1);\
+  o1 = _mm256_unpackhi_epi64(o1, i1);\
+  t0 = i2;\
+  i2 = _mm256_unpacklo_epi64(i2, i3);\
+  t0 = _mm256_unpackhi_epi64(t0, i3);\
+  t1 = i4;\
+  i4 = _mm256_unpacklo_epi64(i4, i5);\
+  t1 = _mm256_unpackhi_epi64(t1, i5);\
+  t2 = i6;\
+  o0 = TRANSP_MASK_2WAY;\
+  i6 = _mm256_unpacklo_epi64(i6, i7);\
+  t2 = _mm256_unpackhi_epi64(t2, i7);\
+  /* load transpose mask into a register, because it will be used 8 times */\
+  i0 = _mm256_shuffle_epi8(i0, o0);\
+  i2 = _mm256_shuffle_epi8(i2, o0);\
+  i4 = _mm256_shuffle_epi8(i4, o0);\
+  i6 = _mm256_shuffle_epi8(i6, o0);\
+  o1 = _mm256_shuffle_epi8(o1, o0);\
+  t0 = _mm256_shuffle_epi8(t0, o0);\
+  t1 = _mm256_shuffle_epi8(t1, o0);\
+  t2 = _mm256_shuffle_epi8(t2, o0);\
+  /* continue with unpack using 4 temp registers */\
+  t3 = i4;\
+  o2 = o1;\
+  o0 = i0;\
+  t4 = t1;\
+  \
+  t3 = _mm256_unpackhi_epi16(t3, i6);\
+  i4 = _mm256_unpacklo_epi16(i4, i6);\
+  o0 = _mm256_unpackhi_epi16(o0, i2);\
+  i0 = _mm256_unpacklo_epi16(i0, i2);\
+  o2 = _mm256_unpackhi_epi16(o2, t0);\
+  o1 = _mm256_unpacklo_epi16(o1, t0);\
+  t4 = _mm256_unpackhi_epi16(t4, t2);\
+  t1 = _mm256_unpacklo_epi16(t1, t2);\
+  /* shuffle with immediate */\
+  i4 = _mm256_shuffle_epi32(i4, 216);\
+  t3 = _mm256_shuffle_epi32(t3, 216);\
+  o1 = _mm256_shuffle_epi32(o1, 216);\
+  o2 = _mm256_shuffle_epi32(o2, 216);\
+  i0 = _mm256_shuffle_epi32(i0, 216);\
+  o0 = _mm256_shuffle_epi32(o0, 216);\
+  t1 = _mm256_shuffle_epi32(t1, 216);\
+  t4 = _mm256_shuffle_epi32(t4, 216);\
+  /* continue with unpack */\
+  i1 = i0;\
+  i3 = o0;\
+  i5 = o1;\
+  i7 = o2;\
+  i0 = _mm256_unpacklo_epi32(i0, i4);\
+  i1 = _mm256_unpackhi_epi32(i1, i4);\
+  o0 = _mm256_unpacklo_epi32(o0, t3);\
+  i3 = _mm256_unpackhi_epi32(i3, t3);\
+  o1 = _mm256_unpacklo_epi32(o1, t1);\
+  i5 = _mm256_unpackhi_epi32(i5, t1);\
+  o2 = _mm256_unpacklo_epi32(o2, t4);\
+  i7 = _mm256_unpackhi_epi32(i7, t4);\
+  /* transpose done */\
+}/**/
+
+void INIT_2way( __m256i *chaining )
+{
+  static __m256i xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7;
+  static __m256i xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15;
+
+  /* load IV into registers xmm8 - xmm15 */
+  xmm8 = chaining[0];
+  xmm9 = chaining[1];
+  xmm10 = chaining[2];
+  xmm11 = chaining[3];
+  xmm12 = chaining[4];
+  xmm13 = chaining[5];
+  xmm14 = chaining[6];
+  xmm15 = chaining[7];
+
+  /* transform chaining value from column ordering into row ordering */
+  Matrix_Transpose_2way(xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7);
+
+  /* store transposed IV */
+  chaining[0] = xmm8;
+  chaining[1] = xmm9;
+  chaining[2] = xmm10;
+  chaining[3] = xmm11;
+  chaining[4] = xmm12;
+  chaining[5] = xmm13;
+  chaining[6] = xmm14;
+  chaining[7] = xmm15;
+}
+
+void TF1024_2way( __m256i *chaining, const __m256i *message )
+{
+  static __m256i xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7;
+  static __m256i xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15;
+  static __m256i QTEMP[8];
+  static __m256i TEMP0;
+  static __m256i TEMP1;
+  static __m256i TEMP2;
+
+  /* load message into registers xmm8 - xmm15 (Q = message) */
+  xmm8 = message[0];
+  xmm9 = message[1];
+  xmm10 = message[2];
+  xmm11 = message[3];
+  xmm12 = message[4];
+  xmm13 = message[5];
+  xmm14 = message[6];
+  xmm15 = message[7];
+
+  /* transform message M from column ordering into row ordering */
+  Matrix_Transpose_2way(xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15, xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7);
+
+  /* store message M (Q input) for later */
+  QTEMP[0] = xmm8;
+  QTEMP[1] = xmm9;
+  QTEMP[2] = xmm10;
+  QTEMP[3] = xmm11;
+  QTEMP[4] = xmm12;
+  QTEMP[5] = xmm13;
+  QTEMP[6] = xmm14;
+  QTEMP[7] = xmm15;
+
+  /* xor CV to message to get P input */
+  /* result: CV+M in xmm8...xmm15 */
+  xmm8 = _mm256_xor_si256( xmm8,  (chaining[0]) );
+  xmm9 = _mm256_xor_si256( xmm9,  (chaining[1]) );
+  xmm10 = _mm256_xor_si256( xmm10, (chaining[2]) );
+  xmm11 = _mm256_xor_si256( xmm11, (chaining[3]) );
+  xmm12 = _mm256_xor_si256( xmm12, (chaining[4]) );
+  xmm13 = _mm256_xor_si256( xmm13, (chaining[5]) );
+  xmm14 = _mm256_xor_si256( xmm14, (chaining[6]) );
+  xmm15 = _mm256_xor_si256( xmm15, (chaining[7]) );
+
+  /* compute permutation P */
+  /* result: P(CV+M) in xmm8...xmm15 */
+  ROUNDS_P_2WAY();
+
+  /* xor CV to P output (feed-forward) */
+  /* result: P(CV+M)+CV in xmm8...xmm15 */
+  xmm8 = _mm256_xor_si256( xmm8,  (chaining[0]) );
+  xmm9 = _mm256_xor_si256( xmm9,  (chaining[1]) );
+  xmm10 = _mm256_xor_si256( xmm10, (chaining[2]) );
+  xmm11 = _mm256_xor_si256( xmm11, (chaining[3]) );
+  xmm12 = _mm256_xor_si256( xmm12, (chaining[4]) );
+  xmm13 = _mm256_xor_si256( xmm13, (chaining[5]) );
+  xmm14 = _mm256_xor_si256( xmm14, (chaining[6]) );
+  xmm15 = _mm256_xor_si256( xmm15, (chaining[7]) );
+
+  /* store P(CV+M)+CV */
+  chaining[0] = xmm8;
+  chaining[1] = xmm9;
+  chaining[2] = xmm10;
+  chaining[3] = xmm11;
+  chaining[4] = xmm12;
+  chaining[5] = xmm13;
+  chaining[6] = xmm14;
+  chaining[7] = xmm15;
+
+  /* load message M (Q input) into xmm8-15 */
+  xmm8 = QTEMP[0];
+  xmm9 = QTEMP[1];
+  xmm10 = QTEMP[2];
+  xmm11 = QTEMP[3];
+  xmm12 = QTEMP[4];
+  xmm13 = QTEMP[5];
+  xmm14 = QTEMP[6];
+  xmm15 = QTEMP[7];
+
+  /* compute permutation Q */
+  /* result: Q(M) in xmm8...xmm15 */
+  ROUNDS_Q_2WAY();
+
+  /* xor Q output */
+  /* result: P(CV+M)+CV+Q(M) in xmm8...xmm15 */
+  xmm8 = _mm256_xor_si256( xmm8,  (chaining[0]) );
+  xmm9 = _mm256_xor_si256( xmm9,  (chaining[1]) );
+  xmm10 = _mm256_xor_si256( xmm10, (chaining[2]) );
+  xmm11 = _mm256_xor_si256( xmm11, (chaining[3]) );
+  xmm12 = _mm256_xor_si256( xmm12, (chaining[4]) );
+  xmm13 = _mm256_xor_si256( xmm13, (chaining[5]) );
+  xmm14 = _mm256_xor_si256( xmm14, (chaining[6]) );
+  xmm15 = _mm256_xor_si256( xmm15, (chaining[7]) );
+
+  /* store CV */
+  chaining[0] = xmm8;
+  chaining[1] = xmm9;
+  chaining[2] = xmm10;
+  chaining[3] = xmm11;
+  chaining[4] = xmm12;
+  chaining[5] = xmm13;
+  chaining[6] = xmm14;
+  chaining[7] = xmm15;
+
+  return;
+}
+
+void OF1024_2way( __m256i* chaining )
+{
+  static __m256i xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7;
+  static __m256i xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15;
+  static __m256i TEMP0;
+  static __m256i TEMP1;
+  static __m256i TEMP2;
+
+  /* load CV into registers xmm8 - xmm15 */
+  xmm8 = chaining[0];
+  xmm9 = chaining[1];
+  xmm10 = chaining[2];
+  xmm11 = chaining[3];
+  xmm12 = chaining[4];
+  xmm13 = chaining[5];
+  xmm14 = chaining[6];
+  xmm15 = chaining[7];
+
+  /* compute permutation P */
+  /* result: P(CV) in xmm8...xmm15 */
+  ROUNDS_P_2WAY();
+
+  /* xor CV to P output (feed-forward) */
+  /* result: P(CV)+CV in xmm8...xmm15 */
+  xmm8 = _mm256_xor_si256( xmm8,  (chaining[0]) );
+  xmm9 = _mm256_xor_si256( xmm9,  (chaining[1]) );
+  xmm10 = _mm256_xor_si256( xmm10, (chaining[2]) );
+  xmm11 = _mm256_xor_si256( xmm11, (chaining[3]) );
+  xmm12 = _mm256_xor_si256( xmm12, (chaining[4]) );
+  xmm13 = _mm256_xor_si256( xmm13, (chaining[5]) );
+  xmm14 = _mm256_xor_si256( xmm14, (chaining[6]) );
+  xmm15 = _mm256_xor_si256( xmm15, (chaining[7]) );
+
+  /* transpose CV back from row ordering to column ordering */
+  /* result: final hash value in xmm0, xmm6, xmm13, xmm15 */
+  Matrix_Transpose_INV_2way(xmm8, xmm9, xmm10, xmm11, xmm12, xmm13, xmm14, xmm15, xmm4, xmm0, xmm6, xmm1, xmm2, xmm3, xmm5, xmm7);
+
+  /* we only need to return the truncated half of the state */
+  chaining[4] = xmm0;
+  chaining[5] = xmm6;
+  chaining[6] = xmm13;
+  chaining[7] = xmm15;
+
+  return;
+}
+
+
+
 #endif  // VAES
 #endif  // GROESTL512_INTR_4WAY_H__
--- a/algo/groestl/myr-groestl.c
+++ b/algo/groestl/myr-groestl.c
@@ -11,7 +11,7 @@
 #else
  #include "sph_groestl.h"
 #endif
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"

 typedef struct {
 #ifdef __AES__
@@ -19,7 +19,7 @@ typedef struct {
 #else
    sph_groestl512_context  groestl;
 #endif
-    SHA256_CTX              sha;
+    sph_sha256_context      sha;
 } myrgr_ctx_holder;

 myrgr_ctx_holder myrgr_ctx;
@@ -31,7 +31,7 @@ void init_myrgr_ctx()
 #else
     sph_groestl512_init( &myrgr_ctx.groestl );
 #endif
-     SHA256_Init( &myrgr_ctx.sha );
+     sph_sha256_init( &myrgr_ctx.sha );
 }

 void myriad_hash(void *output, const void *input)
@@ -39,54 +39,55 @@ void myriad_hash(void *output, const void *input)
   myrgr_ctx_holder ctx;
   memcpy( &ctx, &myrgr_ctx, sizeof(myrgr_ctx) );

- 	uint32_t _ALIGN(32) hash[16];
+   uint32_t _ALIGN(32) hash[16];

 #ifdef __AES__
   update_groestl( &ctx.groestl, (char*)input, 640 );
   final_groestl( &ctx.groestl, (char*)hash);
 #else
-	sph_groestl512(&ctx.groestl, input, 80);
-	sph_groestl512_close(&ctx.groestl, hash);
+   sph_groestl512(&ctx.groestl, input, 80);
+   sph_groestl512_close(&ctx.groestl, hash);
 #endif

-   SHA256_Update( &ctx.sha, (unsigned char*)hash, 64 );
-   SHA256_Final( (unsigned char*)hash, &ctx.sha );
+   sph_sha256( &ctx.sha, hash, 64 );
+   sph_sha256_close( &ctx.sha, hash );

-	memcpy(output, hash, 32);
+   memcpy(output, hash, 32);
 }

 int scanhash_myriad( struct work *work, uint32_t max_nonce,
                     uint64_t *hashes_done, struct thr_info *mythr )
 {
-	uint32_t _ALIGN(64) endiandata[20];
+   uint32_t _ALIGN(64) endiandata[20];
   uint32_t *pdata = work->data;
   uint32_t *ptarget = work->target;
-	const uint32_t first_nonce = pdata[19];
-	uint32_t nonce = first_nonce;
-   int thr_id = mythr->id;  // thr_id arg is deprecated
+   const uint32_t first_nonce = pdata[19];
+   uint32_t nonce = first_nonce;
+   int thr_id = mythr->id;

-	if (opt_benchmark)
-		((uint32_t*)ptarget)[7] = 0x0000ff;
+   if (opt_benchmark)
+      ((uint32_t*)ptarget)[7] = 0x0000ff;

-        swab32_array( endiandata, pdata, 20 );
+   swab32_array( endiandata, pdata, 20 );

-	do {
-		const uint32_t Htarg = ptarget[7];
-		uint32_t hash[8];
-		be32enc(&endiandata[19], nonce);
-		myriad_hash(hash, endiandata);
+   do {
+      const uint32_t Htarg = ptarget[7];
+      uint32_t hash[8];
+      be32enc(&endiandata[19], nonce);
+      myriad_hash(hash, endiandata);

-		if (hash[7] <= Htarg && fulltest(hash, ptarget)) {
-			pdata[19] = nonce;
-			*hashes_done = pdata[19] - first_nonce;
-			return 1;
-		}
-		nonce++;
+      if (hash[7] <= Htarg && fulltest(hash, ptarget))
+      {
+         pdata[19] = nonce;
+         *hashes_done = pdata[19] - first_nonce;
+         return 1;
+      }
+      nonce++;

-	} while (nonce < max_nonce && !work_restart[thr_id].restart);
+   } while (nonce < max_nonce && !work_restart[thr_id].restart);

-	pdata[19] = nonce;
-	*hashes_done = pdata[19] - first_nonce + 1;
-	return 0;
+   pdata[19] = nonce;
+   *hashes_done = pdata[19] - first_nonce + 1;
+   return 0;
 }
 #endif
--- a/algo/hamsi/hamsi-hash-4way.c
+++ b/algo/hamsi/hamsi-hash-4way.c
@@ -548,7 +548,7 @@ static const sph_u32 T512[64][16] = {

 #if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)

-// Hamsi 8 way 
+// Hamsi 8 way AVX512 

 #define INPUT_BIG8 \
 do { \
@@ -849,13 +849,11 @@ void hamsi512_8way_update( hamsi_8way_big_context *sc, const void *data,
 void hamsi512_8way_close( hamsi_8way_big_context *sc, void *dst )
 {
   __m512i pad[1];
-   int ch, cl;
+   uint32_t ch, cl;

   sph_enc32be( &ch, sc->count_high );
   sph_enc32be( &cl, sc->count_low + ( sc->partial_len << 3 ) );
-   pad[0] =  _mm512_set_epi32( cl, ch, cl, ch, cl, ch, cl, ch,
-                               cl, ch, cl, ch, cl, ch, cl, ch );
-//   pad[0] =  m512_const2_32( cl, ch );
+   pad[0] = _mm512_set1_epi64( ((uint64_t)cl << 32 ) | (uint64_t)ch );
   sc->buf[0] = m512_const1_64( 0x80 );
   hamsi_8way_big( sc, sc->buf, 1 );
   hamsi_8way_big_final( sc, pad );
@@ -863,11 +861,9 @@ void hamsi512_8way_close( hamsi_8way_big_context *sc, void *dst )
   mm512_block_bswap_32( (__m512i*)dst, sc->h );
 }

-
 #endif // AVX512

-
-// Hamsi 4 way
+// Hamsi 4 way AVX2

 #define INPUT_BIG \
 do { \
@@ -1186,14 +1182,12 @@ void hamsi512_4way_update( hamsi_4way_big_context *sc, const void *data,
 void hamsi512_4way_close( hamsi_4way_big_context *sc, void *dst )
 {
   __m256i pad[1];
-   int ch, cl;
+   uint32_t ch, cl;

   sph_enc32be( &ch, sc->count_high );
   sph_enc32be( &cl, sc->count_low + ( sc->partial_len << 3 ) );
-   pad[0] =  _mm256_set_epi32( cl, ch, cl, ch, cl, ch, cl, ch );
+   pad[0] = _mm256_set1_epi64x( ((uint64_t)cl << 32 ) | (uint64_t)ch );
   sc->buf[0] = m256_const1_64( 0x80 );
-//      sc->buf[0] = _mm256_set_epi32( 0UL, 0x80UL, 0UL, 0x80UL,
-//                                  0UL, 0x80UL, 0UL, 0x80UL );
   hamsi_big( sc, sc->buf, 1 );
   hamsi_big_final( sc, pad );

--- a/algo/hodl/hodl-gate.c
+++ b/algo/hodl/hodl-gate.c
@@ -99,9 +99,13 @@ void hodl_build_block_header( struct work* g_work, uint32_t version,
 // called only by thread 0, saves a backup of g_work
 void hodl_get_new_work( struct work* work, struct work* g_work)
 {
-     work_free( &hodl_work );
-     work_copy( &hodl_work, g_work );
-     hodl_work.data[ algo_gate.nonce_index ] = ( clock() + rand() ) % 9999;
+//   pthread_rwlock_rdlock( &g_work_lock );
+
+   work_free( &hodl_work );
+   work_copy( &hodl_work, g_work );
+   hodl_work.data[ algo_gate.nonce_index ] = ( clock() + rand() ) % 9999;
+
+//   pthread_rwlock_unlock( &g_work_lock );
 }

 json_t *hodl_longpoll_rpc_call( CURL *curl, int *err, char* lp_url )
@@ -121,7 +125,7 @@ json_t *hodl_longpoll_rpc_call( CURL *curl, int *err, char* lp_url )
 }

 // called by every thread, copies the backup to each thread's work.
-void hodl_resync_threads( struct work* work )
+void hodl_resync_threads( int thr_id, struct work* work )
 {
   int nonce_index = algo_gate.nonce_index;
   pthread_barrier_wait( &hodl_barrier );
@@ -131,6 +135,7 @@ void hodl_resync_threads( struct work* work )
      work_copy( work, &hodl_work );
   }
   work->data[ nonce_index ] = swab32( hodl_work.data[ nonce_index ] );
+   work_restart[thr_id].restart = 0;
 }

 bool hodl_do_this_thread( int thr_id )
@@ -155,11 +160,10 @@ bool register_hodl_algo( algo_gate_t* gate )
  applog( LOG_ERR, "Only CPUs with AES are supported, use legacy version.");
  return false;
 #endif
-//  if ( TOTAL_CHUNKS % opt_n_threads )
-//  {
-//     applog(LOG_ERR,"Thread count must be power of 2.");
-//     return false;
-//  }
+
+  if ( GARBAGE_SIZE % opt_n_threads )
+     applog( LOG_WARNING,"WARNING: Thread count must be power of 2. Miner may crash or produce invalid hash!" );
+
  pthread_barrier_init( &hodl_barrier, NULL, opt_n_threads );
  gate->optimizations         = SSE42_OPT | AES_OPT | AVX2_OPT;
  gate->scanhash              = (void*)&hodl_scanhash;
@@ -171,7 +175,7 @@ bool register_hodl_algo( algo_gate_t* gate )
  gate->resync_threads        = (void*)&hodl_resync_threads;
  gate->do_this_thread        = (void*)&hodl_do_this_thread;
  gate->work_cmp_size         = 76;
-  hodl_scratchbuf = (unsigned char*)malloc( 1 << 30 );
+  hodl_scratchbuf = (unsigned char*)_mm_malloc( 1 << 30, 64 );
  allow_getwork = false;
  opt_target_factor = 8388608.0;
  return ( hodl_scratchbuf != NULL );
--- a/algo/hodl/hodl-wolf.c
+++ b/algo/hodl/hodl-wolf.c
@@ -70,7 +70,7 @@ int scanhash_hodl_wolf( struct work* work, uint32_t max_nonce,
    uint32_t *ptarget = work->target;
    int threadNumber = mythr->id;
    CacheEntry *Garbage = (CacheEntry*)hodl_scratchbuf;
-    CacheEntry Cache[AES_PARALLEL_N];
+    CacheEntry Cache[AES_PARALLEL_N] __attribute__ ((aligned (64)));
    __m128i* data[AES_PARALLEL_N];
    const __m128i* next[AES_PARALLEL_N];
    uint32_t CollisionCount = 0;
--- a/algo/keccak/keccak-macros.c
+++ b/algo/keccak/keccak-macros.c
@@ -134,65 +134,47 @@
        do { \
                DECL64(c0); \
                DECL64(c1); \
-                DECL64(c2); \
-                DECL64(c3); \
-                DECL64(c4); \
                DECL64(bnn); \
                NOT64(bnn, b20); \
                KHI_XO(c0, b00, b10, b20); \
                KHI_XO(c1, b10, bnn, b30); \
-                KHI_XA(c2, b20, b30, b40); \
-                KHI_XO(c3, b30, b40, b00); \
-                KHI_XA(c4, b40, b00, b10); \
+                KHI_XA(b20, b20, b30, b40); \
+                KHI_XO(b30, b30, b40, b00); \
+                KHI_XA(b40, b40, b00, b10); \
                MOV64(b00, c0); \
                MOV64(b10, c1); \
-                MOV64(b20, c2); \
-                MOV64(b30, c3); \
-                MOV64(b40, c4); \
                NOT64(bnn, b41); \
                KHI_XO(c0, b01, b11, b21); \
                KHI_XA(c1, b11, b21, b31); \
-                KHI_XO(c2, b21, b31, bnn); \
-                KHI_XO(c3, b31, b41, b01); \
-                KHI_XA(c4, b41, b01, b11); \
+                KHI_XO(b21, b21, b31, bnn); \
+                KHI_XO(b31, b31, b41, b01); \
+                KHI_XA(b41, b41, b01, b11); \
                MOV64(b01, c0); \
                MOV64(b11, c1); \
-                MOV64(b21, c2); \
-                MOV64(b31, c3); \
-                MOV64(b41, c4); \
                NOT64(bnn, b32); \
                KHI_XO(c0, b02, b12, b22); \
                KHI_XA(c1, b12, b22, b32); \
-                KHI_XA(c2, b22, bnn, b42); \
-                KHI_XO(c3, bnn, b42, b02); \
-                KHI_XA(c4, b42, b02, b12); \
+                KHI_XA(b22, b22, bnn, b42); \
+                KHI_XO(b32, bnn, b42, b02); \
+                KHI_XA(b42, b42, b02, b12); \
                MOV64(b02, c0); \
                MOV64(b12, c1); \
-                MOV64(b22, c2); \
-                MOV64(b32, c3); \
-                MOV64(b42, c4); \
                NOT64(bnn, b33); \
                KHI_XA(c0, b03, b13, b23); \
                KHI_XO(c1, b13, b23, b33); \
-                KHI_XO(c2, b23, bnn, b43); \
-                KHI_XA(c3, bnn, b43, b03); \
-                KHI_XO(c4, b43, b03, b13); \
+                KHI_XO(b23, b23, bnn, b43); \
+                KHI_XA(b33, bnn, b43, b03); \
+                KHI_XO(b43, b43, b03, b13); \
                MOV64(b03, c0); \
                MOV64(b13, c1); \
-                MOV64(b23, c2); \
-                MOV64(b33, c3); \
-                MOV64(b43, c4); \
                NOT64(bnn, b14); \
                KHI_XA(c0, b04, bnn, b24); \
                KHI_XO(c1, bnn, b24, b34); \
-                KHI_XA(c2, b24, b34, b44); \
-                KHI_XO(c3, b34, b44, b04); \
-                KHI_XA(c4, b44, b04, b14); \
+                KHI_XA(b24, b24, b34, b44); \
+                KHI_XO(b34, b34, b44, b04); \
+                KHI_XA(b44, b44, b04, b14); \
                MOV64(b04, c0); \
                MOV64(b14, c1); \
-                MOV64(b24, c2); \
-                MOV64(b34, c3); \
-                MOV64(b44, c4); \
        } while (0)

 #ifdef IOTA
@@ -201,6 +183,7 @@
 #define IOTA(r)   XOR64_IOTA(a00, a00, r)

 #ifdef P0
+#undef P0
 #undef P1
 #undef P2
 #undef P3
--- a/algo/luffa/luffa-hash-2way.c
+++ b/algo/luffa/luffa-hash-2way.c
@@ -66,6 +66,17 @@ static const uint32 CNS_INIT[128] __attribute((aligned(64))) = {
    a = _mm512_xor_si512(a,c0);\
    b = _mm512_xor_si512(b,c1);

+#define MULT24W( a0, a1 ) \
+do { \
+  __m512i b = _mm512_xor_si512( a0, \
+                     _mm512_maskz_shuffle_epi32( 0xbbbb, a1, 16 ) ); \
+  a0 = _mm512_or_si512( _mm512_bsrli_epi128(  b, 4 ), \
+                        _mm512_bslli_epi128( a1,12 ) ); \
+  a1 = _mm512_or_si512( _mm512_bsrli_epi128( a1, 4 ), \
+                        _mm512_bslli_epi128(  b,12 ) ); \
+} while(0)
+
+/*
 #define MULT24W( a0, a1, mask ) \
 do { \
  __m512i b = _mm512_xor_si512( a0, \
@@ -73,6 +84,7 @@ do { \
  a0 = _mm512_or_si512( _mm512_bsrli_epi128(b,4), _mm512_bslli_epi128(a1,12) );\
  a1 = _mm512_or_si512( _mm512_bsrli_epi128(a1,4), _mm512_bslli_epi128(b,12) );\
 } while(0)
+*/

 // confirm pointer arithmetic
 // ok but use array indexes
@@ -235,7 +247,6 @@ void rnd512_4way( luffa_4way_context *state, __m512i *msg )
    __m512i msg0, msg1;
    __m512i tmp[2];
    __m512i x[8];
-    const __m512i MASK = m512_const2_64( 0, 0x00000000ffffffff );

    t0 = chainv[0];
    t1 = chainv[1];
@@ -249,7 +260,7 @@ void rnd512_4way( luffa_4way_context *state, __m512i *msg )
    t0 = _mm512_xor_si512( t0, chainv[8] );
    t1 = _mm512_xor_si512( t1, chainv[9] );

-    MULT24W( t0, t1, MASK );
+    MULT24W( t0, t1 );

    msg0 = _mm512_shuffle_epi32( msg[0], 27 );
    msg1 = _mm512_shuffle_epi32( msg[1], 27 );
@@ -268,68 +279,67 @@ void rnd512_4way( luffa_4way_context *state, __m512i *msg )
    t0 = chainv[0];
    t1 = chainv[1];

-    MULT24W( chainv[0], chainv[1], MASK );
+    MULT24W( chainv[0], chainv[1] );
    chainv[0] = _mm512_xor_si512( chainv[0], chainv[2] );
    chainv[1] = _mm512_xor_si512( chainv[1], chainv[3] );

-    MULT24W( chainv[2], chainv[3], MASK );
+    MULT24W( chainv[2], chainv[3] );
    chainv[2] = _mm512_xor_si512(chainv[2], chainv[4]);
    chainv[3] = _mm512_xor_si512(chainv[3], chainv[5]);

-    MULT24W( chainv[4], chainv[5], MASK );
+    MULT24W( chainv[4], chainv[5] );
    chainv[4] = _mm512_xor_si512(chainv[4], chainv[6]);
    chainv[5] = _mm512_xor_si512(chainv[5], chainv[7]);

-    MULT24W( chainv[6], chainv[7], MASK );
+    MULT24W( chainv[6], chainv[7] );
    chainv[6] = _mm512_xor_si512(chainv[6], chainv[8]);
    chainv[7] = _mm512_xor_si512(chainv[7], chainv[9]);

-    MULT24W( chainv[8], chainv[9], MASK );
+    MULT24W( chainv[8], chainv[9] );
    chainv[8] = _mm512_xor_si512( chainv[8], t0 );
    chainv[9] = _mm512_xor_si512( chainv[9], t1 );

    t0 = chainv[8];
    t1 = chainv[9];

-    MULT24W( chainv[8], chainv[9], MASK );
+    MULT24W( chainv[8], chainv[9] );
    chainv[8] = _mm512_xor_si512( chainv[8], chainv[6] );
    chainv[9] = _mm512_xor_si512( chainv[9], chainv[7] );

-    MULT24W( chainv[6], chainv[7], MASK );
+    MULT24W( chainv[6], chainv[7] );
    chainv[6] = _mm512_xor_si512( chainv[6], chainv[4] );
    chainv[7] = _mm512_xor_si512( chainv[7], chainv[5] );

-    MULT24W( chainv[4], chainv[5], MASK );
+    MULT24W( chainv[4], chainv[5] );
    chainv[4] = _mm512_xor_si512( chainv[4], chainv[2] );
    chainv[5] = _mm512_xor_si512( chainv[5], chainv[3] );

-    MULT24W( chainv[2], chainv[3], MASK );
+    MULT24W( chainv[2], chainv[3] );
    chainv[2] = _mm512_xor_si512( chainv[2], chainv[0] );
    chainv[3] = _mm512_xor_si512( chainv[3], chainv[1] );

-    MULT24W( chainv[0], chainv[1], MASK );
+    MULT24W( chainv[0], chainv[1] );
    chainv[0] = _mm512_xor_si512( _mm512_xor_si512( chainv[0], t0 ), msg0 );
    chainv[1] = _mm512_xor_si512( _mm512_xor_si512( chainv[1], t1 ), msg1 );

-    MULT24W( msg0, msg1, MASK );
+    MULT24W( msg0, msg1 );
    chainv[2] = _mm512_xor_si512( chainv[2], msg0 );
    chainv[3] = _mm512_xor_si512( chainv[3], msg1 );

-    MULT24W( msg0, msg1, MASK );
+    MULT24W( msg0, msg1 );
    chainv[4] = _mm512_xor_si512( chainv[4], msg0 );
    chainv[5] = _mm512_xor_si512( chainv[5], msg1 );

-    MULT24W( msg0, msg1, MASK );
+    MULT24W( msg0, msg1 );
    chainv[6] = _mm512_xor_si512( chainv[6], msg0 );
    chainv[7] = _mm512_xor_si512( chainv[7], msg1 );

-    MULT24W( msg0, msg1, MASK );
+    MULT24W( msg0, msg1);
    chainv[8] = _mm512_xor_si512( chainv[8], msg0 );
    chainv[9] = _mm512_xor_si512( chainv[9], msg1 );

-    MULT24W( msg0, msg1, MASK );
+    MULT24W( msg0, msg1 );

-    // replace with ror
    chainv[3] = _mm512_rol_epi32( chainv[3], 1 );
    chainv[5] = _mm512_rol_epi32( chainv[5], 2 );
    chainv[7] = _mm512_rol_epi32( chainv[7], 3 );
@@ -496,7 +506,7 @@ int luffa_4way_update( luffa_4way_context *state, const void *data,
    {
      // remaining data bytes
      buffer[0] = _mm512_shuffle_epi8( vdata[0], shuff_bswap32 );
-      buffer[1] = m512_const2_64( 0, 0x0000000080000000 );
+      buffer[1] = m512_const1_i128(  0x0000000080000000 );
    }
    return 0;
 }
@@ -520,7 +530,7 @@ int luffa_4way_close( luffa_4way_context *state, void *hashval )
      rnd512_4way( state, buffer );
    else
    {     // empty pad block, constant data
-      msg[0] = m512_const2_64( 0, 0x0000000080000000 );
+      msg[0] = m512_const1_i128(  0x0000000080000000 );
      msg[1] = m512_zero;
      rnd512_4way( state, msg );
    }
@@ -583,13 +593,13 @@ int luffa512_4way_full( luffa_4way_context *state, void *output,
    {
       // padding of partial block
       msg[0] = _mm512_shuffle_epi8( vdata[ 0 ], shuff_bswap32 );
-       msg[1] = m512_const2_64( 0, 0x0000000080000000 );
+       msg[1] = m512_const1_i128(  0x0000000080000000 );
       rnd512_4way( state, msg );
    }
    else
    {
       // empty pad block
-       msg[0] = m512_const2_64( 0, 0x0000000080000000 );
+       msg[0] = m512_const1_i128( 0x0000000080000000 );
       msg[1] = m512_zero;
       rnd512_4way( state, msg );
    }
@@ -631,13 +641,13 @@ int luffa_4way_update_close( luffa_4way_context *state,
    {
       // padding of partial block
       msg[0] = _mm512_shuffle_epi8( vdata[ 0 ], shuff_bswap32 );
-       msg[1] = m512_const2_64( 0, 0x0000000080000000 );
+       msg[1] = m512_const1_i128( 0x0000000080000000 );
       rnd512_4way( state, msg );
    }
    else
    {
       // empty pad block
-       msg[0] = m512_const2_64( 0, 0x0000000080000000 );
+       msg[0] = m512_const1_i128( 0x0000000080000000 );
       msg[1] = m512_zero;
       rnd512_4way( state, msg );
    }
@@ -832,7 +842,7 @@ void rnd512_2way( luffa_2way_context *state, __m256i *msg )
    __m256i msg0, msg1;
    __m256i tmp[2];
    __m256i x[8];
-    const __m256i MASK = m256_const2_64( 0, 0x00000000ffffffff );
+    const __m256i MASK = m256_const1_i128( 0x00000000ffffffff );

    t0 = chainv[0];
    t1 = chainv[1];
@@ -1088,7 +1098,7 @@ int luffa_2way_update( luffa_2way_context *state, const void *data,
    {
      // remaining data bytes
      buffer[0] = _mm256_shuffle_epi8( vdata[0], shuff_bswap32 );
-      buffer[1] = m256_const2_64( 0, 0x0000000080000000 );
+      buffer[1] = m256_const1_i128( 0x0000000080000000 );
    }
    return 0;
 }
@@ -1104,7 +1114,7 @@ int luffa_2way_close( luffa_2way_context *state, void *hashval )
      rnd512_2way( state, buffer );
    else
    {     // empty pad block, constant data
-      msg[0] = m256_const2_64( 0, 0x0000000080000000 );
+      msg[0] = m256_const1_i128( 0x0000000080000000 );
      msg[1] = m256_zero;
      rnd512_2way( state, msg );
    }
@@ -1159,13 +1169,13 @@ int luffa512_2way_full( luffa_2way_context *state, void *output,
    {
       // padding of partial block
       msg[0] = _mm256_shuffle_epi8( vdata[ 0 ], shuff_bswap32 );
-       msg[1] = m256_const2_64( 0, 0x0000000080000000 );
+       msg[1] = m256_const1_i128( 0x0000000080000000 );
       rnd512_2way( state, msg );
    }
    else
    {
       // empty pad block
-       msg[0] = m256_const2_64( 0, 0x0000000080000000 );
+       msg[0] = m256_const1_i128( 0x0000000080000000 );
       msg[1] = m256_zero;
       rnd512_2way( state, msg );
    }
@@ -1206,13 +1216,13 @@ int luffa_2way_update_close( luffa_2way_context *state,
    {
       // padding of partial block
       msg[0] = _mm256_shuffle_epi8( vdata[ 0 ], shuff_bswap32 );
-       msg[1] = m256_const2_64( 0, 0x0000000080000000 );
+       msg[1] = m256_const1_i128( 0x0000000080000000 );
       rnd512_2way( state, msg );
    }
    else
    {
       // empty pad block
-       msg[0] = m256_const2_64( 0, 0x0000000080000000 );
+       msg[0] = m256_const1_i128( 0x0000000080000000 );
       msg[1] = m256_zero;
       rnd512_2way( state, msg );
    }
--- a/algo/luffa/luffa_for_sse2.c
+++ b/algo/luffa/luffa_for_sse2.c
@@ -23,7 +23,7 @@
 #include "simd-utils.h"
 #include "luffa_for_sse2.h"

-#define MULT2(a0,a1) do \
+#define MULT2( a0, a1 ) do \
 { \
  __m128i b =  _mm_xor_si128( a0, _mm_shuffle_epi32( _mm_and_si128(a1,MASK), 16 ) ); \
  a0 = _mm_or_si128( _mm_srli_si128(b,4), _mm_slli_si128(a1,12) ); \
@@ -345,11 +345,11 @@ HashReturn update_and_final_luffa( hashState_luffa *state, BitSequence* output,
    // 16 byte partial block exists for 80 byte len
    if ( state->rembytes  )
       // padding of partial block
-       rnd512( state, m128_const_64( 0, 0x80000000 ),
+       rnd512( state, m128_const_i128(  0x80000000 ),
                      mm128_bswap_32( cast_m128i( data ) ) );
    else
       // empty pad block
-       rnd512( state, m128_zero, m128_const_64( 0, 0x80000000 ) );
+       rnd512( state, m128_zero, m128_const_i128( 0x80000000 ) );

    finalization512( state, (uint32*) output );
    if ( state->hashbitlen > 512 )
@@ -394,11 +394,11 @@ int luffa_full( hashState_luffa *state, BitSequence* output, int hashbitlen,
    // 16 byte partial block exists for 80 byte len
    if ( state->rembytes  )
       // padding of partial block
-       rnd512( state, m128_const_64( 0, 0x80000000 ),
+       rnd512( state, m128_const_i128( 0x80000000 ),
                      mm128_bswap_32( cast_m128i( data ) ) );
    else
       // empty pad block
-       rnd512( state, m128_zero, m128_const_64( 0, 0x80000000 ) );
+       rnd512( state, m128_zero, m128_const_i128( 0x80000000 ) );

    finalization512( state, (uint32*) output );
    if ( state->hashbitlen > 512 )
@@ -606,7 +606,6 @@ static void finalization512( hashState_luffa *state, uint32 *b )

    casti_m256i( b, 0 ) = _mm256_shuffle_epi8(
                                 casti_m256i( hash, 0 ), shuff_bswap32 );
-//    casti_m256i( b, 0 ) = mm256_bswap_32( casti_m256i( hash, 0 ) );

    rnd512( state, zero, zero );

@@ -621,7 +620,6 @@ static void finalization512( hashState_luffa *state, uint32 *b )

    casti_m256i( b, 1 ) = _mm256_shuffle_epi8( 
                                 casti_m256i( hash, 0 ), shuff_bswap32 );
-//    casti_m256i( b, 1 ) = mm256_bswap_32( casti_m256i( hash, 0 ) );
 }

 #else
--- a/algo/lyra2/allium-4way.c
+++ b/algo/lyra2/allium-4way.c
@@ -174,24 +174,19 @@ void allium_16way_hash( void *state, const void *input )
 #if defined(__VAES__)

   intrlv_4x128( vhash, hash0, hash1, hash2, hash3, 256 );
-
-   groestl256_4way_full( &ctx.groestl, vhash, vhash, 256 );
-
+   groestl256_4way_full( &ctx.groestl, vhash, vhash, 32 );
   dintrlv_4x128( state, state+32, state+64, state+96, vhash, 256 );
+
   intrlv_4x128( vhash, hash4, hash5, hash6, hash7, 256 );
-
-   groestl256_4way_full( &ctx.groestl, vhash, vhash, 256 );
-   
+   groestl256_4way_full( &ctx.groestl, vhash, vhash, 32 );
   dintrlv_4x128( state+128, state+160, state+192, state+224, vhash, 256 );
+
   intrlv_4x128( vhash, hash8, hash9, hash10, hash11, 256 );
-
-   groestl256_4way_full( &ctx.groestl, vhash, vhash, 256 );
-
+   groestl256_4way_full( &ctx.groestl, vhash, vhash, 32 );
   dintrlv_4x128( state+256, state+288, state+320, state+352, vhash, 256 );
-   intrlv_4x128( vhash, hash12, hash13, hash14, hash15, 256 );

-   groestl256_4way_full( &ctx.groestl, vhash, vhash, 256 );
- 
+   intrlv_4x128( vhash, hash12, hash13, hash14, hash15, 256 );
+   groestl256_4way_full( &ctx.groestl, vhash, vhash, 32 );
   dintrlv_4x128( state+384, state+416, state+448, state+480, vhash, 256 );
   
 #else
@@ -262,8 +257,11 @@ typedef struct {
   keccak256_4way_context    keccak;
   cubehashParam             cube;
   skein256_4way_context     skein;
+#if defined(__VAES__)
+   groestl256_2way_context   groestl;
+#else
   hashState_groestl256      groestl;
-
+#endif
 } allium_8way_ctx_holder;

 static __thread allium_8way_ctx_holder allium_8way_ctx;
@@ -273,7 +271,11 @@ bool init_allium_8way_ctx()
   keccak256_4way_init( &allium_8way_ctx.keccak );
   cubehashInit( &allium_8way_ctx.cube, 256, 16, 32 );
   skein256_4way_init( &allium_8way_ctx.skein );
+#if defined(__VAES__)
+   groestl256_2way_init( &allium_8way_ctx.groestl, 32 );
+#else
   init_groestl256( &allium_8way_ctx.groestl, 32 );
+#endif
   return true;
 }

@@ -352,9 +354,28 @@ void allium_8way_hash( void *hash, const void *input )
   skein256_4way_update( &ctx.skein, vhashB, 32 );
   skein256_4way_close( &ctx.skein, vhashB );

+#if defined(__VAES__)
+
+   uint64_t vhashC[4*2] __attribute__ ((aligned (64)));
+   uint64_t vhashD[4*2] __attribute__ ((aligned (64)));
+   
+   rintrlv_4x64_2x128( vhashC, vhashD, vhashA, 256 );
+   groestl256_2way_full( &ctx.groestl, vhashC, vhashC, 32 );
+   groestl256_2way_full( &ctx.groestl, vhashD, vhashD, 32 );
+   dintrlv_2x128( hash0, hash1, vhashC, 256 );
+   dintrlv_2x128( hash2, hash3, vhashD, 256 );
+
+   rintrlv_4x64_2x128( vhashC, vhashD, vhashB, 256 );
+   groestl256_2way_full( &ctx.groestl, vhashC, vhashC, 32 );
+   groestl256_2way_full( &ctx.groestl, vhashD, vhashD, 32 );
+   dintrlv_2x128( hash4, hash5, vhashC, 256 );
+   dintrlv_2x128( hash6, hash7, vhashD, 256 );
+
+#else
+
   dintrlv_4x64( hash0, hash1, hash2, hash3, vhashA, 256 );
   dintrlv_4x64( hash4, hash5, hash6, hash7, vhashB, 256 );
-
+   
   groestl256_full( &ctx.groestl, hash0, hash0, 256 );
   groestl256_full( &ctx.groestl, hash1, hash1, 256 );
   groestl256_full( &ctx.groestl, hash2, hash2, 256 );
@@ -363,6 +384,8 @@ void allium_8way_hash( void *hash, const void *input )
   groestl256_full( &ctx.groestl, hash5, hash5, 256 );
   groestl256_full( &ctx.groestl, hash6, hash6, 256 );
   groestl256_full( &ctx.groestl, hash7, hash7, 256 );
+
+#endif
 }

 int scanhash_allium_8way( struct work *work, uint32_t max_nonce,
--- a/algo/lyra2/lyra2-gate.c
+++ b/algo/lyra2/lyra2-gate.c
@@ -187,7 +187,8 @@ bool register_allium_algo( algo_gate_t* gate )
  gate->scanhash  = (void*)&scanhash_allium;
  gate->hash      = (void*)&allium_hash;
 #endif
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
+  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT
+	              | VAES_OPT | VAES256_OPT;
  opt_target_factor = 256.0;
  return true;
 };
@@ -215,9 +216,6 @@ void phi2_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
   size_t t;

   algo_gate.gen_merkle_root( merkle_tree, sctx );
-   // Increment extranonce2
-   for ( t = 0; t < sctx->xnonce2_size && !( ++sctx->job.xnonce2[t] ); t++ );
-   // Assemble block header
   algo_gate.build_block_header( g_work, le32dec( sctx->job.version ),
                  (uint32_t*) sctx->job.prevhash, (uint32_t*) merkle_tree,
                  le32dec( sctx->job.ntime ), le32dec(sctx->job.nbits), NULL );
@@ -225,7 +223,6 @@ void phi2_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
      g_work->data[ 20+t ] = ((uint32_t*)sctx->job.extra)[t];
 }

-
 bool register_phi2_algo( algo_gate_t* gate )
 {
   gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
--- a/algo/lyra2/phi2-4way.c
+++ b/algo/lyra2/phi2-4way.c
@@ -4,7 +4,7 @@
 #include "algo/gost/sph_gost.h"
 #include "algo/cubehash/cubehash_sse2.h"
 #include "lyra2.h"
-#if defined(__VAES__)
+#if defined(__VAES__) && defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
  #include "algo/echo/echo-hash-4way.h"
 #elif defined(__AES__)
  #include "algo/echo/aes_ni/hash_api.h"
--- a/algo/m7m/m7m.c
+++ b/algo/m7m/m7m.c
@@ -12,8 +12,7 @@
 #include "algo/tiger/sph_tiger.h"
 #include "algo/whirlpool/sph_whirlpool.h"
 #include "algo/ripemd/sph_ripemd.h"
-#include <openssl/sha.h>
-
+#include "algo/sha/sph_sha2.h"

 #define EPSa DBL_EPSILON
 #define EPS1 DBL_EPSILON
@@ -105,8 +104,8 @@ uint32_t sw2_( int nnounce )
 }

 typedef struct {
-    SHA256_CTX               sha256;
-    SHA512_CTX               sha512;
+    sph_sha256_context      sha256;
+    sph_sha512_context      sha512;
    sph_keccak512_context    keccak;
    sph_whirlpool_context    whirlpool;
    sph_haval256_5_context   haval;
@@ -118,8 +117,8 @@ m7m_ctx_holder m7m_ctx;

 void init_m7m_ctx()
 {
-    SHA256_Init( &m7m_ctx.sha256 );
-    SHA512_Init( &m7m_ctx.sha512 );
+    sph_sha256_init( &m7m_ctx );
+    sph_sha512_init( &m7m_ctx.sha512 );
    sph_keccak512_init( &m7m_ctx.keccak );
    sph_whirlpool_init( &m7m_ctx.whirlpool );
    sph_haval256_5_init( &m7m_ctx.haval );
@@ -143,11 +142,10 @@ int scanhash_m7m_hash( struct work* work, uint64_t max_nonce,
    uint32_t hash[8] __attribute__((aligned(64)));
    uint8_t bhash[7][64] __attribute__((aligned(64)));
    uint32_t n = pdata[19] - 1;
-    int thr_id = mythr->id;  // thr_id arg is deprecated
+    int thr_id = mythr->id;
    uint32_t usw_, mpzscale;
    const uint32_t first_nonce = pdata[19];
    char data_str[161], hash_str[65], target_str[65];
-    //uint8_t *bdata = 0;
    uint8_t bdata[8192] __attribute__ ((aligned (64)));
    int i, digits;
    int bytes;
@@ -155,12 +153,12 @@ int scanhash_m7m_hash( struct work* work, uint64_t max_nonce,

    m7m_ctx_holder ctx1, ctx2 __attribute__ ((aligned (64)));
    memcpy( &ctx1, &m7m_ctx, sizeof(m7m_ctx) );
-    SHA256_CTX         ctxf_sha256;
+    sph_sha256_context ctxf_sha256;

    memcpy(data, pdata, 80);

-    SHA256_Update(  &ctx1.sha256,    data, M7_MIDSTATE_LEN );
-    SHA512_Update(  &ctx1.sha512,    data, M7_MIDSTATE_LEN );
+    sph_sha256(     &ctx1.sha256,    data, M7_MIDSTATE_LEN );
+    sph_sha512(     &ctx1.sha512,    data, M7_MIDSTATE_LEN );
    sph_keccak512(  &ctx1.keccak,    data, M7_MIDSTATE_LEN );
    sph_whirlpool(  &ctx1.whirlpool, data, M7_MIDSTATE_LEN );
    sph_haval256_5( &ctx1.haval,     data, M7_MIDSTATE_LEN );
@@ -191,11 +189,11 @@ int scanhash_m7m_hash( struct work* work, uint64_t max_nonce,

        memcpy( &ctx2, &ctx1, sizeof(m7m_ctx) );

-        SHA256_Update(  &ctx2.sha256, data_p64, 80 - M7_MIDSTATE_LEN );
-        SHA256_Final( (unsigned char*) (bhash[0]), &ctx2.sha256 );
+        sph_sha256( &ctx2.sha256, data_p64, 80 - M7_MIDSTATE_LEN );
+        sph_sha256_close( &ctx2.sha256, bhash[0] );

-        SHA512_Update(  &ctx2.sha512, data_p64, 80 - M7_MIDSTATE_LEN );
-        SHA512_Final( (unsigned char*) (bhash[1]), &ctx2.sha512 );
+        sph_sha512(  &ctx2.sha512, data_p64, 80 - M7_MIDSTATE_LEN );
+        sph_sha512_close( &ctx2.sha512, bhash[1] );

        sph_keccak512( &ctx2.keccak, data_p64, 80 - M7_MIDSTATE_LEN );
        sph_keccak512_close( &ctx2.keccak, (void*)(bhash[2]) );
@@ -227,9 +225,9 @@ int scanhash_m7m_hash( struct work* work, uint64_t max_nonce,
        bytes = mpz_sizeinbase(product, 256);
        mpz_export((void *)bdata, NULL, -1, 1, 0, 0, product);

-        SHA256_Init( &ctxf_sha256 );
-        SHA256_Update(  &ctxf_sha256, bdata, bytes );
-        SHA256_Final( (unsigned char*) hash, &ctxf_sha256 );
+        sph_sha256_init( &ctxf_sha256 );
+        sph_sha256( &ctxf_sha256, bdata, bytes );
+        sph_sha256_close( &ctxf_sha256, hash );

        digits=(int)((sqrt((double)(n/2))*(1.+EPS))/9000+75);
        mp_bitcnt_t prec = (long int)(digits*BITS_PER_DIGIT+16);
@@ -262,18 +260,13 @@ int scanhash_m7m_hash( struct work* work, uint64_t max_nonce,
            mpzscale=bytes;
            mpz_export(bdata, NULL, -1, 1, 0, 0, product);

-            SHA256_Init( &ctxf_sha256 );
-            SHA256_Update(  &ctxf_sha256, bdata, bytes );
-            SHA256_Final( (unsigned char*) hash, &ctxf_sha256 );
-        }
-
+            sph_sha256_init( &ctxf_sha256 );
+            sph_sha256( &ctxf_sha256, bdata, bytes );
+            sph_sha256_close( &ctxf_sha256, hash );
+	}

        if ( unlikely( valid_hash( (uint64_t*)hash, (uint64_t*)ptarget ) 
             && !opt_benchmark ) )
-
-
-//        if ( unlikely( hash[7] <= ptarget[7] ) )
-//        if ( likely( fulltest( hash, ptarget ) && !opt_benchmark ) )        
        {
           if ( opt_debug )
           {
--- a/algo/nist5/zr5.c
+++ b/algo/nist5/zr5.c
@@ -156,6 +156,8 @@ int scanhash_zr5( struct work *work, uint32_t max_nonce,
 void zr5_get_new_work( struct work* work, struct work* g_work, int thr_id,
                       uint32_t* end_nonce_ptr )
 {
+//   pthread_rwlock_rdlock( &g_work_lock );
+
   // ignore POK in first word
   const int wkcmp_sz = 72;  // (19-1) * sizeof(uint32_t)
   uint32_t *nonceptr = work->data + algo_gate.nonce_index;
@@ -171,6 +173,8 @@ void zr5_get_new_work( struct work* work, struct work* g_work, int thr_id,
   }
   else
       ++(*nonceptr);
+
+//   pthread_rwlock_unlock( &g_work_lock );
 }

 void zr5_display_pok( struct work* work )
--- a/algo/quark/hmq1725-4way.c
+++ b/algo/quark/hmq1725-4way.c
@@ -16,7 +16,7 @@
 #include "algo/simd/simd-hash-2way.h"
 #include "algo/echo/aes_ni/hash_api.h"
 #include "algo/hamsi/hamsi-hash-4way.h"
-#include "algo/fugue/sph_fugue.h"
+#include "algo/fugue/fugue-aesni.h"
 #include "algo/shabal/shabal-hash-4way.h"
 #include "algo/whirlpool/sph_whirlpool.h"
 #include "algo/haval/haval-hash-4way.h"
@@ -40,7 +40,7 @@ union _hmq1725_8way_context_overlay
    cube_4way_context       cube;
    simd_4way_context       simd;
    hamsi512_8way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_8way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_8way_context     sha512;
@@ -363,14 +363,14 @@ extern void hmq1725_8way_hash(void *state, const void *input)
   dintrlv_8x64_512( hash0, hash1, hash2, hash3,
                     hash4, hash5, hash6, hash7, vhash );

-   sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-   sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-   sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-   sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
-   sph_fugue512_full( &ctx.fugue, hash4, hash4, 64 );
-   sph_fugue512_full( &ctx.fugue, hash5, hash5, 64 );
-   sph_fugue512_full( &ctx.fugue, hash6, hash6, 64 );
-   sph_fugue512_full( &ctx.fugue, hash7, hash7, 64 );
+   fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+   fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+   fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+   fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+   fugue512_full( &ctx.fugue, hash4, hash4, 64 );
+   fugue512_full( &ctx.fugue, hash5, hash5, 64 );
+   fugue512_full( &ctx.fugue, hash6, hash6, 64 );
+   fugue512_full( &ctx.fugue, hash7, hash7, 64 );

   intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3,
                           hash4, hash5, hash6, hash7 );
@@ -459,21 +459,21 @@ extern void hmq1725_8way_hash(void *state, const void *input)
                                       m512_zero );

   if ( hash0[0] & mask )
-      sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+      fugue512_full( &ctx.fugue, hash0, hash0, 64 );
   if ( hash1[0] & mask )
-      sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+      fugue512_full( &ctx.fugue, hash1, hash1, 64 );
   if ( hash2[0] & mask )
-      sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+      fugue512_full( &ctx.fugue, hash2, hash2, 64 );
   if ( hash3[0] & mask )
-      sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+      fugue512_full( &ctx.fugue, hash3, hash3, 64 );
   if ( hash4[0] & mask )
-      sph_fugue512_full( &ctx.fugue, hash4, hash4, 64 );
+      fugue512_full( &ctx.fugue, hash4, hash4, 64 );
   if ( hash5[0] & mask )
-      sph_fugue512_full( &ctx.fugue, hash5, hash5, 64 );
+      fugue512_full( &ctx.fugue, hash5, hash5, 64 );
   if ( hash6[0] & mask )
-      sph_fugue512_full( &ctx.fugue, hash6, hash6, 64 );
+      fugue512_full( &ctx.fugue, hash6, hash6, 64 );
   if ( hash7[0] & mask )
-      sph_fugue512_full( &ctx.fugue, hash7, hash7, 64 );
+      fugue512_full( &ctx.fugue, hash7, hash7, 64 );

   intrlv_8x64_512( vhashA, hash0, hash1, hash2, hash3,
                            hash4, hash5, hash6, hash7 );
@@ -628,7 +628,7 @@ union _hmq1725_4way_context_overlay
    simd_2way_context       simd;
    hashState_echo          echo;
    hamsi512_4way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_4way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_4way_context     sha512;
@@ -846,10 +846,10 @@ extern void hmq1725_4way_hash(void *state, const void *input)

    dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );

-    sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-    sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-    sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-    sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+    fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+    fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+    fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+    fugue512_full( &ctx.fugue, hash3, hash3, 64 );

    // In this situation serial simd seems to be faster.

@@ -912,7 +912,7 @@ extern void hmq1725_4way_hash(void *state, const void *input)
   sph_whirlpool512_full( &ctx.whirlpool, hash2, hash2, 64 );
   sph_whirlpool512_full( &ctx.whirlpool, hash3, hash3, 64 );

-// A = fugue serial, B = sha512 prarallel
+// A = fugue serial, B = sha512 parallel
   
   intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );

@@ -920,13 +920,13 @@ extern void hmq1725_4way_hash(void *state, const void *input)
   h_mask = _mm256_movemask_epi8( vh_mask );

   if ( hash0[0] & mask ) 
-      sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+      fugue512_full( &ctx.fugue, hash0, hash0, 64 );
   if ( hash1[0] & mask ) 
-      sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+      fugue512_full( &ctx.fugue, hash1, hash1, 64 );
   if ( hash2[0] & mask ) 
-      sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+      fugue512_full( &ctx.fugue, hash2, hash2, 64 );
   if ( hash3[0] & mask ) 
-      sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+      fugue512_full( &ctx.fugue, hash3, hash3, 64 );

   intrlv_4x64( vhashA, hash0, hash1, hash2, hash3, 512 );

--- a/algo/quark/hmq1725.c
+++ b/algo/quark/hmq1725.c
@@ -17,13 +17,15 @@
 #include "algo/shabal/sph_shabal.h"
 #include "algo/whirlpool/sph_whirlpool.h"
 #include "algo/haval/sph-haval.h"
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"
 #if defined(__AES__)
  #include "algo/groestl/aes_ni/hash-groestl.h"
  #include "algo/echo/aes_ni/hash_api.h"
+  #include "algo/fugue/fugue-aesni.h"
 #else
  #include "algo/groestl/sph_groestl.h"
  #include "algo/echo/sph_echo.h"
+  #include "algo/fugue/sph_fugue.h"
 #endif
 #include "algo/luffa/luffa_for_sse2.h"
 #include "algo/cubehash/cubehash_sse2.h"
@@ -40,17 +42,18 @@ typedef struct {
  sph_shavite512_context  shavite1, shavite2;
  hashState_sd            simd1, simd2;
  sph_hamsi512_context    hamsi1;
-  sph_fugue512_context    fugue1, fugue2;
  sph_shabal512_context   shabal1;
  sph_whirlpool_context   whirlpool1, whirlpool2, whirlpool3, whirlpool4;
-  SHA512_CTX              sha1, sha2;
+  sph_sha512_context      sha1, sha2;
  sph_haval256_5_context  haval1, haval2;
 #if defined(__AES__)
  hashState_echo          echo1, echo2;
  hashState_groestl       groestl1, groestl2;
+  hashState_fugue         fugue1, fugue2;
 #else
  sph_groestl512_context  groestl1, groestl2;
  sph_echo512_context     echo1, echo2;
+  sph_fugue512_context    fugue1, fugue2;
 #endif
 } hmq1725_ctx_holder;

@@ -88,8 +91,13 @@ void init_hmq1725_ctx()

    sph_hamsi512_init(&hmq1725_ctx.hamsi1);

+#if defined(__AES__)
+    fugue512_Init( &hmq1725_ctx.fugue1, 512 );
+    fugue512_Init( &hmq1725_ctx.fugue2, 512 );
+#else
    sph_fugue512_init(&hmq1725_ctx.fugue1);
    sph_fugue512_init(&hmq1725_ctx.fugue2);
+#endif

    sph_shabal512_init(&hmq1725_ctx.shabal1);

@@ -98,8 +106,8 @@ void init_hmq1725_ctx()
    sph_whirlpool_init(&hmq1725_ctx.whirlpool3);
    sph_whirlpool_init(&hmq1725_ctx.whirlpool4);

-    SHA512_Init( &hmq1725_ctx.sha1 );
-    SHA512_Init( &hmq1725_ctx.sha2 );
+    sph_sha512_init( &hmq1725_ctx.sha1 );
+    sph_sha512_init( &hmq1725_ctx.sha2 );

    sph_haval256_5_init(&hmq1725_ctx.haval1);
    sph_haval256_5_init(&hmq1725_ctx.haval2);
@@ -235,8 +243,13 @@ extern void hmq1725hash(void *state, const void *input)
    sph_hamsi512 (&h_ctx.hamsi1, hashA, 64); //3
    sph_hamsi512_close(&h_ctx.hamsi1, hashB); //4

+#if defined(__AES__)
+    fugue512_Update( &h_ctx.fugue1, hashB, 512 ); //2   ////
+    fugue512_Final( &h_ctx.fugue1, hashA ); //3 
+#else
    sph_fugue512 (&h_ctx.fugue1, hashB, 64); //2   ////
    sph_fugue512_close(&h_ctx.fugue1, hashA); //3 
+#endif

    if ( hashA[0] & mask ) //4
    {
@@ -262,13 +275,18 @@ extern void hmq1725hash(void *state, const void *input)

    if ( hashB[0] & mask ) //7
    {
+#if defined(__AES__)
+        fugue512_Update( &h_ctx.fugue2, hashB, 512 ); //
+        fugue512_Final( &h_ctx.fugue2, hashA ); //8
+#else
        sph_fugue512 (&h_ctx.fugue2, hashB, 64); //
        sph_fugue512_close(&h_ctx.fugue2, hashA); //8
+#endif
    }
    else
    {
-        SHA512_Update( &h_ctx.sha1, hashB, 64 );
-        SHA512_Final( (unsigned char*) hashA, &h_ctx.sha1 );
+        sph_sha512( &h_ctx.sha1, hashB, 64 );
+        sph_sha512_close( &h_ctx.sha1, hashA );
    }

 #if defined(__AES__)
@@ -279,8 +297,8 @@ extern void hmq1725hash(void *state, const void *input)
    sph_groestl512_close(&h_ctx.groestl2, hashB); //4
 #endif

-    SHA512_Update( &h_ctx.sha2, hashB, 64 );
-    SHA512_Final( (unsigned char*) hashA, &h_ctx.sha2 );
+    sph_sha512( &h_ctx.sha2, hashB, 64 );
+    sph_sha512_close( &h_ctx.sha2, hashA );

    if ( hashA[0] & mask ) //4
    {
--- a/algo/ripemd/lbry-gate.c
+++ b/algo/ripemd/lbry-gate.c
@@ -69,13 +69,9 @@ void lbry_build_block_header( struct work* g_work, uint32_t version,
 void lbry_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
 {
   unsigned char merkle_root[64] = { 0 };
-   size_t t;
   int i;

   algo_gate.gen_merkle_root( merkle_root, sctx );
-   // Increment extranonce2 
-   for ( t = 0; t < sctx->xnonce2_size && !( ++sctx->job.xnonce2[t] ); t++ );
-   // Assemble block header 

   memset( g_work->data, 0, sizeof(g_work->data) );
   g_work->data[0] = le32dec( sctx->job.version );
--- a/algo/ripemd/lbry.c
+++ b/algo/ripemd/lbry.c
@@ -7,28 +7,28 @@
 #include <string.h>
 #include <stdio.h>
 #include "sph_ripemd.h"
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"

 void lbry_hash(void* output, const void* input)
 {
-   SHA256_CTX              ctx_sha256 __attribute__ ((aligned (64)));
-   SHA512_CTX              ctx_sha512 __attribute__ ((aligned (64)));
-   sph_ripemd160_context   ctx_ripemd __attribute__ ((aligned (64)));
+   sph_sha256_context    ctx_sha256 __attribute__ ((aligned (64)));
+   sph_sha512_context    ctx_sha512 __attribute__ ((aligned (64)));
+   sph_ripemd160_context ctx_ripemd __attribute__ ((aligned (64)));
   uint32_t _ALIGN(64) hashA[16];
   uint32_t _ALIGN(64) hashB[16];
   uint32_t _ALIGN(64) hashC[16];

-   SHA256_Init( &ctx_sha256 );
-   SHA256_Update( &ctx_sha256, input, 112 );
-   SHA256_Final( (unsigned char*) hashA, &ctx_sha256 );
+   sph_sha256_init( &ctx_sha256 );
+   sph_sha256( &ctx_sha256, input, 112 );
+   sph_sha256_close( &ctx_sha256, hashA );

-   SHA256_Init( &ctx_sha256 );
-   SHA256_Update( &ctx_sha256, hashA, 32 );
-   SHA256_Final( (unsigned char*) hashA, &ctx_sha256 );
+   sph_sha256_init( &ctx_sha256 );
+   sph_sha256( &ctx_sha256, hashA, 32 );
+   sph_sha256_close( &ctx_sha256, hashA );

-   SHA512_Init( &ctx_sha512 );
-   SHA512_Update( &ctx_sha512, hashA, 32 );
-   SHA512_Final( (unsigned char*) hashA, &ctx_sha512 );
+   sph_sha512_init( &ctx_sha512 );
+   sph_sha512( &ctx_sha512, hashA, 32 );
+   sph_sha512_close( &ctx_sha512, hashA );

   sph_ripemd160_init( &ctx_ripemd );
   sph_ripemd160 ( &ctx_ripemd, hashA, 32 );
@@ -38,14 +38,14 @@ void lbry_hash(void* output, const void* input)
   sph_ripemd160 ( &ctx_ripemd, hashA+8, 32 );
   sph_ripemd160_close( &ctx_ripemd, hashC );

-   SHA256_Init( &ctx_sha256 );
-   SHA256_Update( &ctx_sha256, hashB, 20 );
-   SHA256_Update( &ctx_sha256, hashC, 20 );
-   SHA256_Final( (unsigned char*) hashA, &ctx_sha256 );
+   sph_sha256_init( &ctx_sha256 );
+   sph_sha256( &ctx_sha256, hashB, 20 );
+   sph_sha256( &ctx_sha256, hashC, 20 );
+   sph_sha256_close( &ctx_sha256, hashA );

-   SHA256_Init( &ctx_sha256 );
-   SHA256_Update( &ctx_sha256, hashA, 32 );
-   SHA256_Final( (unsigned char*) hashA, &ctx_sha256 );
+   sph_sha256_init( &ctx_sha256 );
+   sph_sha256( &ctx_sha256, hashA, 32 );
+   sph_sha256_close( &ctx_sha256, hashA );

   memcpy( output, hashA, 32 );
 }
--- a/algo/scrypt/neoscrypt.c
+++ b/algo/scrypt/neoscrypt.c
@@ -1051,16 +1051,16 @@ int scanhash_neoscrypt( struct work *work,
    uint32_t _ALIGN(64) hash[8];
    const uint32_t Htarg = ptarget[7];
    const uint32_t first_nonce = pdata[19];
-    int thr_id = mythr->id;  // thr_id arg is deprecated
+    int thr_id = mythr->id; 

    while (pdata[19] < max_nonce && !work_restart[thr_id].restart)
    {
        neoscrypt((uint8_t *) hash, (uint8_t *) pdata );

        /* Quick hash check */
-        if (hash[7] <= Htarg && fulltest_le(hash, ptarget)) {
-            *hashes_done = pdata[19] - first_nonce + 1;
-            return 1;
+        if (hash[7] <= Htarg && fulltest_le(hash, ptarget))
+        {
+          submit_solution( work, hash, mythr );
        }

        pdata[19]++;
--- a/algo/sha/hmac-sha256-hash.c
+++ b/algo/sha/hmac-sha256-hash.c
@@ -39,10 +39,17 @@
 void
 SHA256_Buf( const void * in, size_t len, uint8_t digest[32] )
 {
-	SHA256_CTX ctx;
+#if defined(HMAC_SPH_SHA)
+   sph_sha256_context ctx;
+   sph_sha256_init( &ctx );
+   sph_sha256( &ctx, in, len );
+   sph_sha256_close( &ctx, digest );
+#else
+   SHA256_CTX ctx;
   SHA256_Init( &ctx );
   SHA256_Update( &ctx, in, len );
   SHA256_Final( digest, &ctx );
+#endif
 }

 /**
@@ -64,35 +71,59 @@ HMAC_SHA256_Buf( const void *K, size_t Klen, const void *in, size_t len,
 void
 HMAC_SHA256_Init( HMAC_SHA256_CTX *ctx, const void *_K, size_t Klen )
 {
-	unsigned char pad[64];
-	unsigned char khash[32];
-	const unsigned char * K = _K;
-	size_t i;
+   unsigned char pad[64];
+   unsigned char khash[32];
+   const unsigned char * K = _K;
+   size_t i;

-	/* If Klen > 64, the key is really SHA256(K). */
-	if ( Klen > 64 )
+   /* If Klen > 64, the key is really SHA256(K). */
+   if ( Klen > 64 )
   {
-		SHA256_Init( &ctx->ictx );
-		SHA256_Update( &ctx->ictx, K, Klen );
-		SHA256_Final( khash, &ctx->ictx );
-		K = khash;
-		Klen = 32;
-	}
+	   
+#if defined(HMAC_SPH_SHA)
+      sph_sha256_init( &ctx->ictx );
+      sph_sha256( &ctx->ictx, K, Klen );
+      sph_sha256_close( &ctx->ictx, khash );
+#else
+      SHA256_Init( &ctx->ictx );
+      SHA256_Update( &ctx->ictx, K, Klen );
+      SHA256_Final( khash, &ctx->ictx );
+#endif
+       K = khash;
+       Klen = 32;
+   }

-	/* Inner SHA256 operation is SHA256(K xor [block of 0x36] || data). */
+   /* Inner SHA256 operation is SHA256(K xor [block of 0x36] || data). */
+#if defined(HMAC_SPH_SHA)
+   sph_sha256_init( &ctx->ictx );
+#else
   SHA256_Init( &ctx->ictx );
-
+#endif

   for ( i = 0; i < Klen; i++ )  pad[i] = K[i] ^ 0x36;
-   memset( pad + Klen, 0x36, 64 - Klen );
-	SHA256_Update( &ctx->ictx, pad, 64 );

-	/* Outer SHA256 operation is SHA256(K xor [block of 0x5c] || hash). */
-	SHA256_Init( &ctx->octx );
+   memset( pad + Klen, 0x36, 64 - Klen );
+#if defined(HMAC_SPH_SHA)
+   sph_sha256( &ctx->ictx, pad, 64 );
+#else
+   SHA256_Update( &ctx->ictx, pad, 64 );
+#endif
+
+   /* Outer SHA256 operation is SHA256(K xor [block of 0x5c] || hash). */
+#if defined(HMAC_SPH_SHA)
+   sph_sha256_init( &ctx->octx );
+#else   
+   SHA256_Init( &ctx->octx );
+#endif

   for ( i = 0; i < Klen; i++ )  pad[i] = K[i] ^ 0x5c;
+
   memset( pad + Klen, 0x5c, 64 - Klen );
-	SHA256_Update( &ctx->octx, pad, 64 );
+#if defined(HMAC_SPH_SHA)
+   sph_sha256( &ctx->octx, pad, 64 );
+#else
+   SHA256_Update( &ctx->octx, pad, 64 );
+#endif
 }

 /* Add bytes to the HMAC-SHA256 operation. */
@@ -100,23 +131,33 @@ void
 HMAC_SHA256_Update( HMAC_SHA256_CTX *ctx, const void *in, size_t len )
 {
 	/* Feed data to the inner SHA256 operation. */
-	SHA256_Update( &ctx->ictx, in, len );
+#if defined(HMAC_SPH_SHA)
+   sph_sha256( &ctx->ictx, in, len );
+#else
+   SHA256_Update( &ctx->ictx, in, len );
+#endif
 }

 /* Finish an HMAC-SHA256 operation. */
 void
 HMAC_SHA256_Final( unsigned char digest[32], HMAC_SHA256_CTX *ctx )
 {
-	unsigned char ihash[32];
+   unsigned char ihash[32];

-	/* Finish the inner SHA256 operation. */
-	SHA256_Final( ihash, &ctx->ictx );
+#if defined(HMAC_SPH_SHA)
+   sph_sha256_close( &ctx->ictx, ihash );
+   sph_sha256( &ctx->octx, ihash, 32 );
+   sph_sha256_close( &ctx->octx, digest );
+#else
+   /* Finish the inner SHA256 operation. */
+   SHA256_Final( ihash, &ctx->ictx );

-	/* Feed the inner hash to the outer SHA256 operation. */
-	SHA256_Update( &ctx->octx, ihash, 32 );
+   /* Feed the inner hash to the outer SHA256 operation. */
+   SHA256_Update( &ctx->octx, ihash, 32 );

-	/* Finish the outer SHA256 operation. */
-	SHA256_Final( digest, &ctx->octx );
+   /* Finish the outer SHA256 operation. */
+   SHA256_Final( digest, &ctx->octx );
+#endif
 }

 /**
--- a/algo/sha/hmac-sha256-hash.h
+++ b/algo/sha/hmac-sha256-hash.h
@@ -29,14 +29,24 @@
 #ifndef HMAC_SHA256_H__
 #define HMAC_SHA256_H__

+//#define HMAC_SSL_SHA 1
+#define HMAC_SPH_SHA 1
+
 #include <sys/types.h>
 #include <stdint.h>
+#include "sph_sha2.h"
 #include <openssl/sha.h>

+
 typedef struct HMAC_SHA256Context
 {
-     SHA256_CTX ictx;
-     SHA256_CTX octx;
+#if defined(HMAC_SPH_SHA)
+   sph_sha256_context ictx;
+   sph_sha256_context octx;
+#else
+   SHA256_CTX ictx;
+   SHA256_CTX octx;
+#endif
 } HMAC_SHA256_CTX;

 void SHA256_Buf( const void *, size_t len, uint8_t digest[32] );
--- a/algo/sha/sha2.c
+++ b/algo/sha/sha2.c
@@ -12,7 +12,6 @@

 #include <string.h>
 #include <inttypes.h>
-#include <openssl/sha.h>

 #if defined(USE_ASM) && defined(__arm__) && defined(__APCS_32__)
 #define EXTERN_SHA256
@@ -198,16 +197,6 @@ static void sha256d_80_swap(uint32_t *hash, const uint32_t *data)

 extern void sha256d(unsigned char *hash, const unsigned char *data, int len)
 {
-#if defined(__SHA__)
-   SHA256_CTX ctx;
-   SHA256_Init( &ctx );
-   SHA256_Update( &ctx, data, len );
-   SHA256_Final( (unsigned char*)hash, &ctx );
-   SHA256_Init( &ctx );
-   SHA256_Update( &ctx, hash, 32 );
-   SHA256_Final( (unsigned char*)hash, &ctx );
-#else
-
   uint32_t S[16], T[16];
 	int i, r;

@@ -229,7 +218,6 @@ extern void sha256d(unsigned char *hash, const unsigned char *data, int len)
 	sha256_transform(T, S, 0);
 	for (i = 0; i < 8; i++)
 		be32enc((uint32_t *)hash + i, T[i]);
-#endif
 }

 static inline void sha256d_preextend(uint32_t *W)
@@ -479,8 +467,8 @@ static inline void sha256d_ms(uint32_t *hash, uint32_t *W,
 void sha256d_ms_4way(uint32_t *hash,  uint32_t *data,
 	const uint32_t *midstate, const uint32_t *prehash);

-static inline int scanhash_sha256d_4way(int thr_id, struct work *work,
-             uint32_t max_nonce, uint64_t *hashes_done)
+static inline int scanhash_sha256d_4way( struct work *work,
+             uint32_t max_nonce, uint64_t *hashes_done, struct thr_info *mythr )
 {
        uint32_t *pdata = work->data;
        uint32_t *ptarget = work->target;
@@ -492,6 +480,7 @@ static inline int scanhash_sha256d_4way(int thr_id, struct work *work,
 	uint32_t n = pdata[19] - 1;
 	const uint32_t first_nonce = pdata[19];
 	const uint32_t Htarg = ptarget[7];
+   int thr_id = mythr->id;
 	int i, j;
 	
 	memcpy(data, pdata + 16, 64);
@@ -521,10 +510,8 @@ static inline int scanhash_sha256d_4way(int thr_id, struct work *work,
 			if (swab32(hash[4 * 7 + i]) <= Htarg) {
 				pdata[19] = data[4 * 3 + i];
 				sha256d_80_swap(hash, pdata);
-				if (fulltest(hash, ptarget)) {
-					*hashes_done = n - first_nonce + 1;
-					return 1;
-				}
+            if ( fulltest( hash, ptarget ) && !opt_benchmark )
+               submit_solution( work, hash, mythr );
 			}
 		}
 	} while (n < max_nonce && !work_restart[thr_id].restart);
@@ -541,8 +528,8 @@ static inline int scanhash_sha256d_4way(int thr_id, struct work *work,
 void sha256d_ms_8way(uint32_t *hash,  uint32_t *data,
 	const uint32_t *midstate, const uint32_t *prehash);

-static inline int scanhash_sha256d_8way(int thr_id, struct work *work,
-                              uint32_t max_nonce, uint64_t *hashes_done)
+static inline int scanhash_sha256d_8way( struct work *work,
+            uint32_t max_nonce, uint64_t *hashes_done, struct thr_info *mythr )
 {
        uint32_t *pdata = work->data;
        uint32_t *ptarget = work->target;
@@ -554,6 +541,7 @@ static inline int scanhash_sha256d_8way(int thr_id, struct work *work,
 	uint32_t n = pdata[19] - 1;
 	const uint32_t first_nonce = pdata[19];
 	const uint32_t Htarg = ptarget[7];
+   int thr_id = mythr->id;
 	int i, j;
 	
 	memcpy(data, pdata + 16, 64);
@@ -583,10 +571,8 @@ static inline int scanhash_sha256d_8way(int thr_id, struct work *work,
 			if (swab32(hash[8 * 7 + i]) <= Htarg) {
 				pdata[19] = data[8 * 3 + i];
 				sha256d_80_swap(hash, pdata);
-				if (fulltest(hash, ptarget)) {
-					*hashes_done = n - first_nonce + 1;
-					return 1;
-				}
+            if ( fulltest( hash, ptarget ) && !opt_benchmark )
+               submit_solution( work, hash, mythr );
 			}
 		}
 	} while (n < max_nonce && !work_restart[thr_id].restart);
@@ -614,13 +600,11 @@ int scanhash_sha256d( struct work *work,

 #ifdef HAVE_SHA256_8WAY
 	if (sha256_use_8way())
-		return scanhash_sha256d_8way(thr_id, work,
-			max_nonce, hashes_done);
+		return scanhash_sha256d_8way( work,	max_nonce, hashes_done, mythr );
 #endif
 #ifdef HAVE_SHA256_4WAY
 	if (sha256_use_4way())
-		return scanhash_sha256d_4way(thr_id, work,
-			max_nonce, hashes_done);
+		return scanhash_sha256d_4way( work,	max_nonce, hashes_done, mythr );
 #endif
 	
 	memcpy(data, pdata + 16, 64);
@@ -657,7 +641,7 @@ int scanhash_SHA256d( struct work *work, const uint32_t max_nonce,
   uint32_t n = pdata[19] - 1;
   const uint32_t first_nonce = pdata[19];
   const uint32_t Htarg = ptarget[7];
-   int thr_id = mythr->id;  // thr_id arg is deprecated
+   int thr_id = mythr->id;

   memcpy( data, pdata, 80 );

@@ -680,14 +664,9 @@ int scanhash_SHA256d( struct work *work, const uint32_t max_nonce,

 bool register_sha256d_algo( algo_gate_t* gate )
 {
-#if defined(__SHA__)
-   gate->optimizations = SHA_OPT;
-   gate->scanhash = (void*)&scanhash_SHA256d;
-#else
   gate->optimizations = SSE2_OPT | AVX2_OPT;
   gate->scanhash = (void*)&scanhash_sha256d;
-#endif
-    gate->hash     = (void*)&sha256d;
-    return true;
+   gate->hash     = (void*)&sha256d;
+   return true;
 };

--- a/algo/sha/sha256-hash-opt.c
+++ b/algo/sha/sha256-hash-opt.c
@@ -0,0 +1,200 @@
+/*   Intel SHA extensions using C intrinsics               */
+/*   Written and place in public domain by Jeffrey Walton  */
+/*   Based on code from Intel, and by Sean Gulley for      */
+/*   the miTLS project.                                    */
+
+// A drop in replacement for the function of the same name in sph_sha2.c.
+
+#if defined(__SHA__)
+
+#include "simd-utils.h"
+
+static void sha2_round( const uint8_t input[], uint32_t state[8] )
+{
+    __m128i STATE0, STATE1;
+    __m128i MSG, TMP, MASK;
+    __m128i TMSG0, TMSG1, TMSG2, TMSG3;
+    __m128i ABEF_SAVE, CDGH_SAVE;
+
+    // Load initial values
+    TMP = _mm_load_si128((__m128i*) &state[0]);
+    STATE1 = _mm_load_si128((__m128i*) &state[4]);
+    MASK = _mm_set_epi64x(0x0c0d0e0f08090a0bULL, 0x0405060700010203ULL);
+
+    TMP = _mm_shuffle_epi32(TMP, 0xB1); // CDAB
+    STATE1 = _mm_shuffle_epi32(STATE1, 0x1B); // EFGH
+    STATE0 = _mm_alignr_epi8(TMP, STATE1, 8); // ABEF
+    STATE1 = _mm_blend_epi16(STATE1, TMP, 0xF0); // CDGH
+
+    // Save current hash
+    ABEF_SAVE = STATE0;
+    CDGH_SAVE = STATE1;
+
+    // Rounds 0-3
+    MSG = _mm_load_si128((const __m128i*) (input+0));
+    TMSG0 = _mm_shuffle_epi8(MSG, MASK);
+    MSG = _mm_add_epi32(TMSG0, _mm_set_epi64x(0xE9B5DBA5B5C0FBCFULL, 0x71374491428A2F98ULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+
+    // Rounds 4-7
+    TMSG1 = _mm_load_si128((const __m128i*) (input+16));
+    TMSG1 = _mm_shuffle_epi8(TMSG1, MASK);
+    MSG = _mm_add_epi32(TMSG1, _mm_set_epi64x(0xAB1C5ED5923F82A4ULL, 0x59F111F13956C25BULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+    TMSG0 = _mm_sha256msg1_epu32(TMSG0, TMSG1);
+
+    // Rounds 8-11
+    TMSG2 = _mm_load_si128((const __m128i*) (input+32));
+    TMSG2 = _mm_shuffle_epi8(TMSG2, MASK);
+    MSG = _mm_add_epi32(TMSG2, _mm_set_epi64x(0x550C7DC3243185BEULL, 0x12835B01D807AA98ULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+    TMSG1 = _mm_sha256msg1_epu32(TMSG1, TMSG2);
+
+    // Rounds 12-15
+    TMSG3 = _mm_load_si128((const __m128i*) (input+48));
+    TMSG3 = _mm_shuffle_epi8(TMSG3, MASK);
+    MSG = _mm_add_epi32(TMSG3, _mm_set_epi64x(0xC19BF1749BDC06A7ULL, 0x80DEB1FE72BE5D74ULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    TMP = _mm_alignr_epi8(TMSG3, TMSG2, 4);
+    TMSG0 = _mm_add_epi32(TMSG0, TMP);
+    TMSG0 = _mm_sha256msg2_epu32(TMSG0, TMSG3);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+    TMSG2 = _mm_sha256msg1_epu32(TMSG2, TMSG3);
+
+    // Rounds 16-19
+    MSG = _mm_add_epi32(TMSG0, _mm_set_epi64x(0x240CA1CC0FC19DC6ULL, 0xEFBE4786E49B69C1ULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    TMP = _mm_alignr_epi8(TMSG0, TMSG3, 4);
+    TMSG1 = _mm_add_epi32(TMSG1, TMP);
+    TMSG1 = _mm_sha256msg2_epu32(TMSG1, TMSG0);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+    TMSG3 = _mm_sha256msg1_epu32(TMSG3, TMSG0);
+
+    // Rounds 20-23
+    MSG = _mm_add_epi32(TMSG1, _mm_set_epi64x(0x76F988DA5CB0A9DCULL, 0x4A7484AA2DE92C6FULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    TMP = _mm_alignr_epi8(TMSG1, TMSG0, 4);
+    TMSG2 = _mm_add_epi32(TMSG2, TMP);
+    TMSG2 = _mm_sha256msg2_epu32(TMSG2, TMSG1);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+    TMSG0 = _mm_sha256msg1_epu32(TMSG0, TMSG1);
+
+    // Rounds 24-27
+    MSG = _mm_add_epi32(TMSG2, _mm_set_epi64x(0xBF597FC7B00327C8ULL, 0xA831C66D983E5152ULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    TMP = _mm_alignr_epi8(TMSG2, TMSG1, 4);
+    TMSG3 = _mm_add_epi32(TMSG3, TMP);
+    TMSG3 = _mm_sha256msg2_epu32(TMSG3, TMSG2);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+    TMSG1 = _mm_sha256msg1_epu32(TMSG1, TMSG2);
+
+    // Rounds 28-31
+    MSG = _mm_add_epi32(TMSG3, _mm_set_epi64x(0x1429296706CA6351ULL,  0xD5A79147C6E00BF3ULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    TMP = _mm_alignr_epi8(TMSG3, TMSG2, 4);
+    TMSG0 = _mm_add_epi32(TMSG0, TMP);
+    TMSG0 = _mm_sha256msg2_epu32(TMSG0, TMSG3);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+    TMSG2 = _mm_sha256msg1_epu32(TMSG2, TMSG3);
+
+    // Rounds 32-35
+    MSG = _mm_add_epi32(TMSG0, _mm_set_epi64x(0x53380D134D2C6DFCULL, 0x2E1B213827B70A85ULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    TMP = _mm_alignr_epi8(TMSG0, TMSG3, 4);
+    TMSG1 = _mm_add_epi32(TMSG1, TMP);
+    TMSG1 = _mm_sha256msg2_epu32(TMSG1, TMSG0);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+    TMSG3 = _mm_sha256msg1_epu32(TMSG3, TMSG0);
+
+    // Rounds 36-39
+    MSG = _mm_add_epi32(TMSG1, _mm_set_epi64x(0x92722C8581C2C92EULL, 0x766A0ABB650A7354ULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    TMP = _mm_alignr_epi8(TMSG1, TMSG0, 4);
+    TMSG2 = _mm_add_epi32(TMSG2, TMP);
+    TMSG2 = _mm_sha256msg2_epu32(TMSG2, TMSG1);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+    TMSG0 = _mm_sha256msg1_epu32(TMSG0, TMSG1);
+
+    // Rounds 40-43
+    MSG = _mm_add_epi32(TMSG2, _mm_set_epi64x(0xC76C51A3C24B8B70ULL, 0xA81A664BA2BFE8A1ULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    TMP = _mm_alignr_epi8(TMSG2, TMSG1, 4);
+    TMSG3 = _mm_add_epi32(TMSG3, TMP);
+    TMSG3 = _mm_sha256msg2_epu32(TMSG3, TMSG2);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+    TMSG1 = _mm_sha256msg1_epu32(TMSG1, TMSG2);
+
+    // Rounds 44-47
+    MSG = _mm_add_epi32(TMSG3, _mm_set_epi64x(0x106AA070F40E3585ULL, 0xD6990624D192E819ULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    TMP = _mm_alignr_epi8(TMSG3, TMSG2, 4);
+    TMSG0 = _mm_add_epi32(TMSG0, TMP);
+    TMSG0 = _mm_sha256msg2_epu32(TMSG0, TMSG3);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+    TMSG2 = _mm_sha256msg1_epu32(TMSG2, TMSG3);
+
+    // Rounds 48-51
+    MSG = _mm_add_epi32(TMSG0, _mm_set_epi64x(0x34B0BCB52748774CULL, 0x1E376C0819A4C116ULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    TMP = _mm_alignr_epi8(TMSG0, TMSG3, 4);
+    TMSG1 = _mm_add_epi32(TMSG1, TMP);
+    TMSG1 = _mm_sha256msg2_epu32(TMSG1, TMSG0);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+    TMSG3 = _mm_sha256msg1_epu32(TMSG3, TMSG0);
+
+    // Rounds 52-55
+    MSG = _mm_add_epi32(TMSG1, _mm_set_epi64x(0x682E6FF35B9CCA4FULL, 0x4ED8AA4A391C0CB3ULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    TMP = _mm_alignr_epi8(TMSG1, TMSG0, 4);
+    TMSG2 = _mm_add_epi32(TMSG2, TMP);
+    TMSG2 = _mm_sha256msg2_epu32(TMSG2, TMSG1);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+
+    // Rounds 56-59
+    MSG = _mm_add_epi32(TMSG2, _mm_set_epi64x(0x8CC7020884C87814ULL, 0x78A5636F748F82EEULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    TMP = _mm_alignr_epi8(TMSG2, TMSG1, 4);
+    TMSG3 = _mm_add_epi32(TMSG3, TMP);
+    TMSG3 = _mm_sha256msg2_epu32(TMSG3, TMSG2);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+
+    // Rounds 60-63
+    MSG = _mm_add_epi32(TMSG3, _mm_set_epi64x(0xC67178F2BEF9A3F7ULL, 0xA4506CEB90BEFFFAULL));
+    STATE1 = _mm_sha256rnds2_epu32(STATE1, STATE0, MSG);
+    MSG = _mm_shuffle_epi32(MSG, 0x0E);
+    STATE0 = _mm_sha256rnds2_epu32(STATE0, STATE1, MSG);
+
+    // Add values back to state
+    STATE0 = _mm_add_epi32(STATE0, ABEF_SAVE);
+    STATE1 = _mm_add_epi32(STATE1, CDGH_SAVE);
+
+    TMP = _mm_shuffle_epi32(STATE0, 0x1B); // FEBA
+    STATE1 = _mm_shuffle_epi32(STATE1, 0xB1); // DCHG
+    STATE0 = _mm_blend_epi16(TMP, STATE1, 0xF0); // DCBA
+    STATE1 = _mm_alignr_epi8(STATE1, TMP, 8); // ABEF
+
+    // Save state
+    _mm_store_si128((__m128i*) &state[0], STATE0);
+    _mm_store_si128((__m128i*) &state[4], STATE1);
+}
+
+
+#endif
--- a/algo/sha/sha256q-4way.c
+++ b/algo/sha/sha256q-4way.c
@@ -5,6 +5,79 @@
 #include <stdio.h>
 #include "sha-hash-4way.h"

+#if defined(SHA256T_16WAY)
+
+static __thread sha256_16way_context sha256_ctx16 __attribute__ ((aligned (64)));
+
+void sha256q_16way_hash( void* output, const void* input )
+{
+   uint32_t vhash[8*16] __attribute__ ((aligned (64)));
+   sha256_16way_context ctx;
+   memcpy( &ctx, &sha256_ctx16, sizeof ctx );
+
+   sha256_16way_update( &ctx, input + (64<<4), 16 );
+   sha256_16way_close( &ctx, vhash );
+
+   sha256_16way_init( &ctx );
+   sha256_16way_update( &ctx, vhash, 32 );
+   sha256_16way_close( &ctx, vhash );
+
+   sha256_16way_init( &ctx );
+   sha256_16way_update( &ctx, vhash, 32 );
+   sha256_16way_close( &ctx, vhash );
+
+   sha256_16way_init( &ctx );
+   sha256_16way_update( &ctx, vhash, 32 );
+   sha256_16way_close( &ctx, output );
+}
+
+int scanhash_sha256q_16way( struct work *work, const uint32_t max_nonce,
+                           uint64_t *hashes_done, struct thr_info *mythr )
+{
+   uint32_t vdata[20*16] __attribute__ ((aligned (64)));
+   uint32_t hash32[8*16] __attribute__ ((aligned (32)));
+   uint32_t lane_hash[8] __attribute__ ((aligned (32)));
+   uint32_t *hash32_d7 = &(hash32[7<<4]);
+   uint32_t *pdata = work->data;
+   const uint32_t *ptarget = work->target;
+   const uint32_t targ32_d7 = ptarget[7];
+   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 16;
+   uint32_t n = first_nonce;
+   __m512i  *noncev = (__m512i*)vdata + 19;   // aligned
+   const int thr_id = mythr->id;
+   const bool bench = opt_benchmark;
+
+   mm512_bswap32_intrlv80_16x32( vdata, pdata );
+   *noncev = _mm512_set_epi32( n+15, n+14, n+13, n+12, n+11, n+10, n+9, n+8,
+                               n+ 7, n+ 6, n+ 5, n+ 4, n+ 3, n+ 2, n+1, n );
+   sha256_16way_init( &sha256_ctx16 );
+   sha256_16way_update( &sha256_ctx16, vdata, 64 );
+
+   do
+   {
+     pdata[19] = n;
+     sha256q_16way_hash( hash32, vdata );
+     for ( int lane = 0; lane < 16; lane++ )
+     if ( unlikely( hash32_d7[ lane ] <= targ32_d7 ) )
+     {
+        extr_lane_16x32( lane_hash, hash32, lane, 256 );
+        if ( likely( valid_hash( lane_hash, ptarget ) && !bench ) )
+        {
+           pdata[19] = bswap_32( n + lane );
+           submit_solution( work, lane_hash, mythr );
+        }
+      }
+      *noncev = _mm512_add_epi32( *noncev, m512_const1_32( 16 ) );
+      n += 16;
+   } while ( (n < last_nonce) && !work_restart[thr_id].restart );
+   pdata[19] = n;
+   *hashes_done = n - first_nonce;
+   return 0;
+}
+
+#endif
+
 #if defined(SHA256T_8WAY)

 static __thread sha256_8way_context sha256_ctx8 __attribute__ ((aligned (64)));
@@ -31,68 +104,47 @@ void sha256q_8way_hash( void* output, const void* input )
   sha256_8way_close( &ctx, output );
 }

-int scanhash_sha256q_8way( struct work *work, uint32_t max_nonce,
-	                   uint64_t *hashes_done, struct thr_info *mythr )
+int scanhash_sha256q_8way( struct work *work, const uint32_t max_nonce,
+                           uint64_t *hashes_done, struct thr_info *mythr )
 {
-   uint32_t vdata[20*8] __attribute__ ((aligned (64)));
-   uint32_t hash[8*8] __attribute__ ((aligned (32)));
+   uint32_t vdata[20*8]  __attribute__ ((aligned (64)));
+   uint32_t hash32[8*8]    __attribute__ ((aligned (32)));
   uint32_t lane_hash[8] __attribute__ ((aligned (32)));
+   uint32_t *hash32_d7 = &(hash32[7<<3]);
   uint32_t *pdata = work->data;
-   uint32_t *ptarget = work->target;
-   const uint32_t Htarg = ptarget[7];
+   const uint32_t *ptarget = work->target;
+   const uint32_t targ32_d7 = ptarget[7];
   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 8;
   uint32_t n = first_nonce;
   __m256i  *noncev = (__m256i*)vdata + 19;   // aligned
-   int thr_id = mythr->id;  // thr_id arg is deprecated
+   const int thr_id = mythr->id;
+   const bool bench = opt_benchmark;

-   const uint64_t htmax[] = {          0,
-                                     0xF,
-                                    0xFF,
-                                   0xFFF,
-                                  0xFFFF,
-                              0x10000000 };
-   const uint32_t masks[] = {  0xFFFFFFFF,
-                               0xFFFFFFF0,
-                               0xFFFFFF00,
-                               0xFFFFF000,
-                               0xFFFF0000,
-                                        0 };
-
-   // Need big endian data
   mm256_bswap32_intrlv80_8x32( vdata, pdata );
+   *noncev = _mm256_set_epi32( n+7, n+6, n+5, n+4, n+3, n+2, n+1, n );
   sha256_8way_init( &sha256_ctx8 );
   sha256_8way_update( &sha256_ctx8, vdata, 64 );

-   for ( int m = 0; m < 6; m++ ) if ( Htarg <= htmax[m] )
+   do
   {
-      uint32_t mask = masks[m];
-      do
-      {
-         *noncev = mm256_bswap_32(
-		            _mm256_set_epi32( n+7, n+6, n+5, n+4, n+3, n+2, n+1, n ) );
-
-	      pdata[19] = n;
-         sha256q_8way_hash( hash, vdata );
-
-         uint32_t *hash7 = &(hash[7<<3]); 
-	 
-         for ( int lane = 0; lane < 8; lane++ )
-         if ( !( hash7[ lane ] & mask ) )
-         { 
-            // deinterleave hash for lane
-	         extr_lane_8x32( lane_hash, hash, lane, 256 );
-
-	         if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
-            {
-	           pdata[19] = n + lane;
-              submit_solution( work, lane_hash, mythr );
-            }
-	      }
-         n += 8;
-      } while ( (n < max_nonce-10) && !work_restart[thr_id].restart );
-      break;
-   }
-   *hashes_done = n - first_nonce + 1;
+     pdata[19] = n;
+     sha256q_8way_hash( hash32, vdata );
+     for ( int lane = 0; lane < 8; lane++ )
+     if ( unlikely( hash32_d7[ lane ] <= targ32_d7 ) )
+     {
+        extr_lane_8x32( lane_hash, hash32, lane, 256 );
+        if ( likely( valid_hash( lane_hash, ptarget ) && !bench ) )
+        {
+           pdata[19] = bswap_32( n + lane );
+           submit_solution( work, lane_hash, mythr );
+        }
+      }
+      *noncev = _mm256_add_epi32( *noncev, m256_const1_32( 8 ) );
+      n += 8;
+   } while ( (n < last_nonce) && !work_restart[thr_id].restart );
+   pdata[19] = n;
+   *hashes_done = n - first_nonce;
   return 0;
 }

--- a/algo/sha/sha256q.c
+++ b/algo/sha/sha256q.c
@@ -1,108 +1,74 @@
 #include "sha256t-gate.h"
-
-#if !defined(SHA256T_16WAY) && !defined(SHA256T_8WAY) && !defined(SHA256T_4WAY)
-
 #include <stdlib.h>
 #include <stdint.h>
 #include <string.h>
 #include <stdio.h>
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"

-static __thread SHA256_CTX sha256q_ctx __attribute__ ((aligned (64)));
+static __thread sph_sha256_context sha256q_ctx __attribute__ ((aligned (64)));

 void sha256q_midstate( const void* input )
 {
-    SHA256_Init( &sha256q_ctx );
-    SHA256_Update( &sha256q_ctx, input, 64 );
+   sph_sha256_init( &sha256q_ctx );
+   sph_sha256( &sha256q_ctx, input, 64 );
 }

-void sha256q_hash( void* output, const void* input )
+int sha256q_hash( void* output, const void* input )
 {
   uint32_t _ALIGN(64) hash[16];
   const int midlen = 64;            // bytes
   const int tail   = 80 - midlen;   // 16

-   SHA256_CTX ctx __attribute__ ((aligned (64)));
+   sph_sha256_context ctx __attribute__ ((aligned (64)));
   memcpy( &ctx, &sha256q_ctx, sizeof sha256q_ctx );

-   SHA256_Update( &ctx, input + midlen, tail );
-   SHA256_Final( (unsigned char*)hash, &ctx );
+   sph_sha256( &ctx, input + midlen, tail );
+   sph_sha256_close( &ctx, hash );

-   SHA256_Init( &ctx );
-   SHA256_Update( &ctx, hash, 32 );
-   SHA256_Final( (unsigned char*)hash, &ctx );
+   sph_sha256_init( &ctx );
+   sph_sha256( &ctx, hash, 32 );
+   sph_sha256_close( &ctx, hash );

-   SHA256_Init( &ctx );
-   SHA256_Update( &ctx, hash, 32 );
-   SHA256_Final( (unsigned char*)hash, &ctx );
+   sph_sha256_init( &ctx );
+   sph_sha256( &ctx, hash, 32 );
+   sph_sha256_close( &ctx, hash );

-   SHA256_Init( &ctx );
-   SHA256_Update( &ctx, hash, 32 );
-   SHA256_Final( (unsigned char*)hash, &ctx );
+   sph_sha256_init( &ctx );
+   sph_sha256( &ctx, hash, 32 );
+   sph_sha256_close( &ctx, output );

-   memcpy( output, hash, 32 );
+   return 1;
 }

 int scanhash_sha256q( struct work *work, uint32_t max_nonce,
                      uint64_t *hashes_done, struct thr_info *mythr )
 {
+   uint32_t edata[20] __attribute__((aligned(64)));
+   uint32_t hash[8] __attribute__((aligned(64)));
   uint32_t *pdata = work->data;
   uint32_t *ptarget = work->target;
-   uint32_t n = pdata[19] - 1;
   const uint32_t first_nonce = pdata[19];
-   const uint32_t Htarg = ptarget[7];
-#ifdef _MSC_VER
-   uint32_t __declspec(align(32)) hash64[8];
-#else
-   uint32_t hash64[8] __attribute__((aligned(32)));
-#endif
-   uint32_t endiandata[32];
-   int thr_id = mythr->id;  // thr_id arg is deprecated
+   const uint32_t last_nonce = max_nonce - 1;
+   uint32_t n = first_nonce;
+   const int thr_id = mythr->id;
+   const bool bench = opt_benchmark;

-   uint64_t htmax[] = {
-		0,
-		0xF,
-		0xFF,
-		0xFFF,
-		0xFFFF,
-		0x10000000
-	};
-   uint32_t masks[] = {
-		0xFFFFFFFF,
-		0xFFFFFFF0,
-		0xFFFFFF00,
-		0xFFFFF000,
-		0xFFFF0000,
-		0
-	};
+   mm128_bswap32_80( edata, pdata );
+   sha256q_midstate( edata );

-   // we need bigendian data...
-   casti_m128i( endiandata, 0 ) = mm128_bswap_32( casti_m128i( pdata, 0 ) );
-   casti_m128i( endiandata, 1 ) = mm128_bswap_32( casti_m128i( pdata, 1 ) );
-   casti_m128i( endiandata, 2 ) = mm128_bswap_32( casti_m128i( pdata, 2 ) );
-   casti_m128i( endiandata, 3 ) = mm128_bswap_32( casti_m128i( pdata, 3 ) );
-   casti_m128i( endiandata, 4 ) = mm128_bswap_32( casti_m128i( pdata, 4 ) );
-
-   sha256q_midstate( endiandata );
-
-   for ( int m = 0; m < 6; m++ )
+   do
   {
-      if ( Htarg <= htmax[m] )
+      edata[19] = n;
+      if ( likely( sha256q_hash( hash, edata ) ) )
+      if ( unlikely( valid_hash( hash, ptarget ) && !bench ) )
      {
-         uint32_t mask = masks[m];
-         do {
-            pdata[19] = ++n;
-            be32enc(&endiandata[19], n);
-            sha256q_hash( hash64, endiandata );
-            if ( !( hash64[7] & mask ) )
-            if ( fulltest( hash64, ptarget ) && !opt_benchmark )
-               submit_solution( work, hash64, mythr );
-         } while ( n < max_nonce && !work_restart[thr_id].restart );
-         break;
+         pdata[19] = bswap_32( n );
+         submit_solution( work, hash, mythr );
      }
-   }
-   *hashes_done = n - first_nonce + 1;
+      n++;
+   } while ( n < last_nonce && !work_restart[thr_id].restart );
+   *hashes_done = n - first_nonce;
   pdata[19] = n;
   return 0;
 }
-#endif
+
--- a/algo/sha/sha256t-4way.c
+++ b/algo/sha/sha256t-4way.c
@@ -5,6 +5,75 @@
 #include <stdio.h>
 #include "sha-hash-4way.h"

+#if defined(SHA256T_16WAY)
+
+static __thread sha256_16way_context sha256_ctx16 __attribute__ ((aligned (64)));
+
+void sha256t_16way_hash( void* output, const void* input )
+{
+   uint32_t vhash[8*16] __attribute__ ((aligned (64)));
+   sha256_16way_context ctx;
+   memcpy( &ctx, &sha256_ctx16, sizeof ctx );
+
+   sha256_16way_update( &ctx, input + (64<<4), 16 );
+   sha256_16way_close( &ctx, vhash );
+
+   sha256_16way_init( &ctx );
+   sha256_16way_update( &ctx, vhash, 32 );
+   sha256_16way_close( &ctx, vhash );
+
+   sha256_16way_init( &ctx );
+   sha256_16way_update( &ctx, vhash, 32 );
+   sha256_16way_close( &ctx, output );
+}
+
+int scanhash_sha256t_16way( struct work *work, const uint32_t max_nonce,
+                           uint64_t *hashes_done, struct thr_info *mythr )
+{
+   uint32_t vdata[20*16] __attribute__ ((aligned (64)));
+   uint32_t hash32[8*16] __attribute__ ((aligned (32)));
+   uint32_t lane_hash[8] __attribute__ ((aligned (32)));
+   uint32_t *hash32_d7 = &(hash32[7<<4]);
+   uint32_t *pdata = work->data;
+   const uint32_t *ptarget = work->target;
+   const uint32_t targ32_d7 = ptarget[7];
+   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 16;
+   uint32_t n = first_nonce;
+   __m512i  *noncev = (__m512i*)vdata + 19;   // aligned
+   const int thr_id = mythr->id;
+   const bool bench = opt_benchmark;
+
+   mm512_bswap32_intrlv80_16x32( vdata, pdata );
+   *noncev = _mm512_set_epi32( n+15, n+14, n+13, n+12, n+11, n+10, n+9, n+8,
+                               n+ 7, n+ 6, n+ 5, n+ 4, n+ 3, n+ 2, n+1, n );
+   sha256_16way_init( &sha256_ctx16 );
+   sha256_16way_update( &sha256_ctx16, vdata, 64 );
+
+   do
+   {
+     pdata[19] = n;
+     sha256t_16way_hash( hash32, vdata );
+     for ( int lane = 0; lane < 16; lane++ )
+     if ( unlikely( hash32_d7[ lane ] <= targ32_d7 ) )
+     {
+        extr_lane_16x32( lane_hash, hash32, lane, 256 );
+        if ( likely( valid_hash( lane_hash, ptarget ) && !bench ) )
+        {
+           pdata[19] = bswap_32( n + lane );
+           submit_solution( work, lane_hash, mythr );
+        }
+      }
+      *noncev = _mm512_add_epi32( *noncev, m512_const1_32( 16 ) );
+      n += 16;
+   } while ( (n < last_nonce) && !work_restart[thr_id].restart );
+   pdata[19] = n;
+   *hashes_done = n - first_nonce;
+   return 0;
+}
+
+#endif
+
 #if defined(SHA256T_8WAY)

 static __thread sha256_8way_context sha256_ctx8 __attribute__ ((aligned (64)));
@@ -31,61 +100,43 @@ int scanhash_sha256t_8way( struct work *work, const uint32_t max_nonce,
                           uint64_t *hashes_done, struct thr_info *mythr )
 {
   uint32_t vdata[20*8]  __attribute__ ((aligned (64)));
-   uint32_t hash[8*8]    __attribute__ ((aligned (32)));
+   uint32_t hash32[8*8]    __attribute__ ((aligned (32)));
   uint32_t lane_hash[8] __attribute__ ((aligned (32)));
-   uint32_t *hash7 = &(hash[7<<3]);
+   uint32_t *hash32_d7 = &(hash32[7<<3]);
   uint32_t *pdata = work->data;
   const uint32_t *ptarget = work->target;
-   const uint32_t Htarg = ptarget[7];
+   const uint32_t targ32_d7 = ptarget[7];
   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 8;
   uint32_t n = first_nonce;
   __m256i  *noncev = (__m256i*)vdata + 19;   // aligned
   const int thr_id = mythr->id;
+   const bool bench = opt_benchmark;

-   const uint64_t htmax[] = {          0,
-                                     0xF,
-                                    0xFF,
-                                   0xFFF,
-                                  0xFFFF,
-                              0x10000000 };
-   const uint32_t masks[] = {  0xFFFFFFFF,
-                               0xFFFFFFF0,
-                               0xFFFFFF00,
-                               0xFFFFF000,
-                               0xFFFF0000,
-                                        0 };
-
-
-   // Need big endian data
   mm256_bswap32_intrlv80_8x32( vdata, pdata );
+   *noncev = _mm256_set_epi32( n+7, n+6, n+5, n+4, n+3, n+2, n+1, n );
   sha256_8way_init( &sha256_ctx8 );
   sha256_8way_update( &sha256_ctx8, vdata, 64 );

-   for ( int m = 0; m < 6; m++ ) if ( Htarg <= htmax[m] )
+   do
   {
-      const uint32_t mask = masks[m];
-      do
-      {
-        *noncev = mm256_bswap_32( _mm256_set_epi32(
-                                          n+7,n+6,n+5,n+4,n+3,n+2,n+1,n ) );
-         pdata[19] = n;
-         sha256t_8way_hash( hash, vdata );
-         for ( int lane = 0; lane < 8; lane++ )
-         if ( !( hash7[ lane ] & mask ) )
-         {
-            // deinterleave hash for lane
-            extr_lane_8x32( lane_hash, hash, lane, 256 );
-            if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
-            {
-              pdata[19] = n + lane;
-              submit_solution( work, lane_hash, mythr );
-	         }
-         }
-         n += 8;
-      } while ( (n < max_nonce-10) && !work_restart[thr_id].restart );
-      break;
-   }
-   *hashes_done = n - first_nonce + 1;
+     pdata[19] = n;
+     sha256t_8way_hash( hash32, vdata );
+     for ( int lane = 0; lane < 8; lane++ )
+     if ( unlikely( hash32_d7[ lane ] <= targ32_d7 ) )
+     {
+        extr_lane_8x32( lane_hash, hash32, lane, 256 );
+        if ( likely( valid_hash( lane_hash, ptarget ) && !bench ) )
+        {
+           pdata[19] = bswap_32( n + lane );
+           submit_solution( work, lane_hash, mythr );
+        }
+      }
+      *noncev = _mm256_add_epi32( *noncev, m256_const1_32( 8 ) );
+      n += 8;
+   } while ( (n < last_nonce) && !work_restart[thr_id].restart );
+   pdata[19] = n;
+   *hashes_done = n - first_nonce;
   return 0;
 }

--- a/algo/sha/sha256t-gate.c
+++ b/algo/sha/sha256t-gate.c
@@ -2,38 +2,41 @@

 bool register_sha256t_algo( algo_gate_t* gate )
 {
-#if defined(SHA256T_8WAY)
-    gate->optimizations = SSE2_OPT | AVX2_OPT | SHA_OPT;
-    gate->scanhash   = (void*)&scanhash_sha256t_8way;
-    gate->hash       = (void*)&sha256t_8way_hash;
-#elif defined(SHA256T_4WAY)
-    gate->optimizations = SSE2_OPT | AVX2_OPT | SHA_OPT;
-    gate->scanhash   = (void*)&scanhash_sha256t_4way;
-    gate->hash       = (void*)&sha256t_4way_hash;
-#else
+    gate->optimizations = SSE2_OPT | AVX2_OPT | AVX512_OPT;
+#if defined(SHA256T_16WAY)
+    gate->scanhash   = (void*)&scanhash_sha256t_16way;
+    gate->hash       = (void*)&sha256t_16way_hash;
+#elif defined(__SHA__)
    gate->optimizations = SHA_OPT;
    gate->scanhash   = (void*)&scanhash_sha256t;
    gate->hash       = (void*)&sha256t_hash;
+#elif defined(SHA256T_8WAY)
+    gate->scanhash   = (void*)&scanhash_sha256t_8way;
+    gate->hash       = (void*)&sha256t_8way_hash;
+#else
+    gate->scanhash   = (void*)&scanhash_sha256t_4way;
+    gate->hash       = (void*)&sha256t_4way_hash;
 #endif
    return true;
 }

 bool register_sha256q_algo( algo_gate_t* gate )
 {
-#if defined(SHA256T_8WAY)
-    gate->optimizations = SSE2_OPT | AVX2_OPT | SHA_OPT;
-    gate->scanhash   = (void*)&scanhash_sha256q_8way;
-    gate->hash       = (void*)&sha256q_8way_hash;
-#elif defined(SHA256T_4WAY)
-    gate->optimizations = SSE2_OPT | AVX2_OPT | SHA_OPT;
-    gate->scanhash   = (void*)&scanhash_sha256q_4way;
-    gate->hash       = (void*)&sha256q_4way_hash;
-#else
+    gate->optimizations = SSE2_OPT | AVX2_OPT | AVX512_OPT;
+#if defined(SHA256T_16WAY)
+    gate->scanhash   = (void*)&scanhash_sha256q_16way;
+    gate->hash       = (void*)&sha256q_16way_hash;
+#elif defined(__SHA__)
    gate->optimizations = SHA_OPT;
    gate->scanhash   = (void*)&scanhash_sha256q;
    gate->hash       = (void*)&sha256q_hash;
+#elif defined(SHA256T_8WAY)
+    gate->scanhash   = (void*)&scanhash_sha256q_8way;
+    gate->hash       = (void*)&sha256q_8way_hash;
+#else
+    gate->scanhash   = (void*)&scanhash_sha256q_4way;
+    gate->hash       = (void*)&sha256q_4way_hash;
 #endif
    return true;
-
 }

--- a/algo/sha/sha256t-gate.h
+++ b/algo/sha/sha256t-gate.h
@@ -4,18 +4,27 @@
 #include <stdint.h>
 #include "algo-gate-api.h"

-// Override multi way on ryzen, SHA is better.
-#if !defined(__SHA__)
- #if defined(__AVX2__)
-  #define SHA256T_8WAY
- #elif defined(__SSE2__)
-  #define SHA256T_4WAY
- #endif
+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+  #define SHA256T_16WAY 1
+#elif defined(__AVX2__)
+  #define SHA256T_8WAY 1
+#else
+  #define SHA256T_4WAY 1
 #endif

 bool register_sha256t_algo( algo_gate_t* gate );
 bool register_sha256q_algo( algo_gate_t* gate );

+#if defined(SHA256T_16WAY)
+
+void sha256t_16way_hash( void *output, const void *input );
+int scanhash_sha256t_16way( struct work *work, uint32_t max_nonce,
+                           uint64_t *hashes_done, struct thr_info *mythr );
+void sha256q_16way_hash( void *output, const void *input );
+int scanhash_sha256q_16way( struct work *work, uint32_t max_nonce,
+                           uint64_t *hashes_done, struct thr_info *mythr );
+#endif
+
 #if defined(SHA256T_8WAY)

 void sha256t_8way_hash( void *output, const void *input );
@@ -36,10 +45,11 @@ int scanhash_sha256q_4way( struct work *work, uint32_t max_nonce,
                           uint64_t *hashes_done, struct thr_info *mythr );
 #endif

-void sha256t_hash( void *output, const void *input );
+
+int sha256t_hash( void *output, const void *input );
 int scanhash_sha256t( struct work *work, uint32_t max_nonce,
                      uint64_t *hashes_done, struct thr_info *mythr );
-void sha256q_hash( void *output, const void *input );
+int sha256q_hash( void *output, const void *input );
 int scanhash_sha256q( struct work *work, uint32_t max_nonce,
                      uint64_t *hashes_done, struct thr_info *mythr );

--- a/algo/sha/sha256t.c
+++ b/algo/sha/sha256t.c
@@ -1,105 +1,73 @@
 #include "sha256t-gate.h"
-
-#if !defined(SHA256T_16WAY) && !defined(SHA256T_8WAY) && !defined(SHA256T_4WAY)
-
 #include <stdlib.h>
 #include <stdint.h>
 #include <string.h>
 #include <stdio.h>
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"

-static __thread SHA256_CTX sha256t_ctx __attribute__ ((aligned (64)));
+// Only used on CPUs with SHA
+
+static __thread sph_sha256_context sha256t_ctx __attribute__ ((aligned (64)));

 void sha256t_midstate( const void* input )
 {
-    SHA256_Init( &sha256t_ctx );
-    SHA256_Update( &sha256t_ctx, input, 64 );
+   sph_sha256_init( &sha256t_ctx );
+   sph_sha256( &sha256t_ctx, input, 64 );
 }

-void sha256t_hash( void* output, const void* input )
+int sha256t_hash( void* output, const void* input )
 {
   uint32_t _ALIGN(64) hash[16];
   const int midlen = 64;            // bytes
   const int tail   = 80 - midlen;   // 16

-   SHA256_CTX ctx __attribute__ ((aligned (64)));
+   sph_sha256_context ctx __attribute__ ((aligned (64)));
   memcpy( &ctx, &sha256t_ctx, sizeof sha256t_ctx );

-   SHA256_Update( &ctx, input + midlen, tail );
-   SHA256_Final( (unsigned char*)hash, &ctx );
+   sph_sha256( &ctx, input + midlen, tail );
+   sph_sha256_close( &ctx, hash );

-   SHA256_Init( &ctx );
-   SHA256_Update( &ctx, hash, 32 );
-   SHA256_Final( (unsigned char*)hash, &ctx );
+   sph_sha256_init( &ctx );
+   sph_sha256( &ctx, hash, 32 );
+   sph_sha256_close( &ctx, hash );

-   SHA256_Init( &ctx );
-   SHA256_Update( &ctx, hash, 32 );
-   SHA256_Final( (unsigned char*)hash, &ctx );
+   sph_sha256_init( &ctx );
+   sph_sha256( &ctx, hash, 32 );
+   sph_sha256_close( &ctx, output );

-   memcpy( output, hash, 32 );
+   return 1;
 }

 int scanhash_sha256t( struct work *work, uint32_t max_nonce,
                      uint64_t *hashes_done, struct thr_info *mythr )
 {
+   uint32_t edata[20] __attribute__((aligned(64)));
+   uint32_t hash[8] __attribute__((aligned(64)));
   uint32_t *pdata = work->data;
   uint32_t *ptarget = work->target;
-   uint32_t n = pdata[19] - 1;
   const uint32_t first_nonce = pdata[19];
-   const uint32_t Htarg = ptarget[7];
-#ifdef _MSC_VER
-   uint32_t __declspec(align(32)) hash64[8];
-#else
-   uint32_t hash64[8] __attribute__((aligned(32)));
-#endif
-   uint32_t endiandata[32];
-   int thr_id = mythr->id;  // thr_id arg is deprecated
+   const uint32_t last_nonce = max_nonce - 1;
+   uint32_t n = first_nonce;
+   const int thr_id = mythr->id;
+   const bool bench = opt_benchmark;

-   uint64_t htmax[] = {
-		0,
-		0xF,
-		0xFF,
-		0xFFF,
-		0xFFFF,
-		0x10000000
-	};
-   uint32_t masks[] = {
-		0xFFFFFFFF,
-		0xFFFFFFF0,
-		0xFFFFFF00,
-		0xFFFFF000,
-		0xFFFF0000,
-		0
-	};
+   mm128_bswap32_80( edata, pdata );
+   sha256t_midstate( edata );

-   // we need bigendian data...
-   casti_m128i( endiandata, 0 ) = mm128_bswap_32( casti_m128i( pdata, 0 ) );
-   casti_m128i( endiandata, 1 ) = mm128_bswap_32( casti_m128i( pdata, 1 ) );
-   casti_m128i( endiandata, 2 ) = mm128_bswap_32( casti_m128i( pdata, 2 ) );
-   casti_m128i( endiandata, 3 ) = mm128_bswap_32( casti_m128i( pdata, 3 ) );
-   casti_m128i( endiandata, 4 ) = mm128_bswap_32( casti_m128i( pdata, 4 ) );
-
-   sha256t_midstate( endiandata );
-
-   for ( int m = 0; m < 6; m++ )
+   do
   {
-      if ( Htarg <= htmax[m] )
+      edata[19] = n;
+      if ( likely( sha256t_hash( hash, edata ) ) )
+      if ( unlikely( valid_hash( hash, ptarget ) && !bench ) )
      {
-         uint32_t mask = masks[m];
-         do {
-            pdata[19] = ++n;
-            be32enc(&endiandata[19], n);
-            sha256t_hash( hash64, endiandata );
-            if ( !(hash64[7] & mask) )
-            if ( fulltest( hash64, ptarget ) && !opt_benchmark )
-               submit_solution( work, hash64, mythr );
-         } while ( n < max_nonce && !work_restart[thr_id].restart );
-         break;
+         pdata[19] = bswap_32( n );
+         submit_solution( work, hash, mythr );
      }
-   }
-   *hashes_done = n - first_nonce + 1;
+      n++;
+   } while ( n < last_nonce && !work_restart[thr_id].restart );
+   *hashes_done = n - first_nonce;
   pdata[19] = n;
   return 0;
 }
-#endif
+

--- a/algo/sha/sph_sha2.c
+++ b/algo/sha/sph_sha2.c
@@ -71,8 +71,13 @@ static const sph_u32 H256[8] = {
 * of the compression function.
 */

-#if SPH_SMALL_FOOTPRINT_SHA2
+#if defined(__SHA__)

+#include "sha256-hash-opt.c"
+
+#else   // no SHA
+
+/*
 static const sph_u32 K[64] = {
 	SPH_C32(0x428A2F98), SPH_C32(0x71374491),
 	SPH_C32(0xB5C0FBCF), SPH_C32(0xE9B5DBA5),
@@ -107,6 +112,9 @@ static const sph_u32 K[64] = {
 	SPH_C32(0x90BEFFFA), SPH_C32(0xA4506CEB),
 	SPH_C32(0xBEF9A3F7), SPH_C32(0xC67178F2)
 };
+*/
+
+#if SPH_SMALL_FOOTPRINT_SHA2

 #define SHA2_MEXP1(in, pc)   do { \
 		W[pc] = in(pc); \
@@ -191,7 +199,7 @@ static const sph_u32 K[64] = {
 		(r)[7] = SPH_T32((r)[7] + H); \
 	} while (0)

-#else
+#else  // large footprint (default)

 #define SHA2_ROUND_BODY(in, r)   do { \
 		sph_u32 A, B, C, D, E, F, G, H, T1, T2; \
@@ -600,7 +608,7 @@ static const sph_u32 K[64] = {
 		(r)[7] = SPH_T32((r)[7] + H); \
 	} while (0)

-#endif
+#endif  // small footprint else

 /*
 * One round of SHA-224 / SHA-256. The data must be aligned for 32-bit access.
@@ -613,6 +621,9 @@ sha2_round(const unsigned char *data, sph_u32 r[8])
 #undef SHA2_IN
 }

+#endif   // SHA else
+
+
 /* see sph_sha2.h */
 void
 sph_sha224_init(void *cc)
@@ -653,7 +664,7 @@ void
 sph_sha224_close(void *cc, void *dst)
 {
 	sha224_close(cc, dst, 7);
-	sph_sha224_init(cc);
+//	sph_sha224_init(cc);
 }

 /* see sph_sha2.h */
@@ -661,7 +672,7 @@ void
 sph_sha224_addbits_and_close(void *cc, unsigned ub, unsigned n, void *dst)
 {
 	sha224_addbits_and_close(cc, ub, n, dst, 7);
-	sph_sha224_init(cc);
+//	sph_sha224_init(cc);
 }

 /* see sph_sha2.h */
@@ -677,14 +688,14 @@ void
 sph_sha256_addbits_and_close(void *cc, unsigned ub, unsigned n, void *dst)
 {
 	sha224_addbits_and_close(cc, ub, n, dst, 8);
-	sph_sha256_init(cc);
+//	sph_sha256_init(cc);
 }

 /* see sph_sha2.h */
-void
-sph_sha224_comp(const sph_u32 msg[16], sph_u32 val[8])
-{
-#define SHA2_IN(x)   msg[x]
-	SHA2_ROUND_BODY(SHA2_IN, val);
-#undef SHA2_IN
-}
+//void
+//sph_sha224_comp(const sph_u32 msg[16], sph_u32 val[8])
+//{
+//#define SHA2_IN(x)   msg[x]
+//	SHA2_ROUND_BODY(SHA2_IN, val);
+//#undef SHA2_IN
+//}
--- a/algo/sha/sph_sha2.h
+++ b/algo/sha/sph_sha2.h
@@ -73,7 +73,7 @@ typedef struct {
 	sph_u32 count_high, count_low;
 #endif
 #endif
-} sph_sha224_context;
+} sph_sha224_context __attribute__((aligned(64)));

 /**
 * This structure is a context for SHA-256 computations. It is identical
--- a/algo/shavite/shavite-hash-2way.c
+++ b/algo/shavite/shavite-hash-2way.c
@@ -23,6 +23,19 @@ static const uint32_t IV512[] =
   _mm256_blend_epi32( mm256_ror128_32( a ), \
                       mm256_ror128_32( b ), 0x88 )

+#if defined(__VAES__)
+
+#define mm256_aesenc_2x128( x, k ) \
+   _mm256_aesenc_epi128( x, _mm256_castsi128_si256( k ) )
+
+#else
+
+#define mm256_aesenc_2x128( x, k ) \
+   mm256_concat_128( _mm_aesenc_si128( mm128_extr_hi128_256( x ), k ), \
+                     _mm_aesenc_si128( mm128_extr_lo128_256( x ), k ) )
+
+#endif
+
 static void
 c512_2way( shavite512_2way_context *ctx, const void *msg )
 {
@@ -304,7 +317,7 @@ void shavite512_2way_close( shavite512_2way_context *ctx, void *dst )
    uint32_t vp = ctx->ptr>>5;

    // Terminating byte then zero pad
-    casti_m256i( buf, vp++ ) = m256_const2_64( 0, 0x0000000000000080 );
+    casti_m256i( buf, vp++ ) = m256_const1_i128( 0x0000000000000080 );

    // Zero pad full vectors up to count
    for ( ; vp < 6; vp++ )      
@@ -384,13 +397,13 @@ void shavite512_2way_update_close( shavite512_2way_context *ctx, void *dst,

   if ( vp == 0 )    // empty buf, xevan.
   { 
-      casti_m256i( buf, 0 ) = m256_const2_64( 0, 0x0000000000000080 );
+      casti_m256i( buf, 0 ) = m256_const1_i128( 0x0000000000000080 );
      memset_zero_256( (__m256i*)buf + 1, 5 );
      ctx->count0 = ctx->count1 = ctx->count2 = ctx->count3 = 0;
   }
   else     // half full buf, everyone else.
   {
-    casti_m256i( buf, vp++ ) = m256_const2_64( 0, 0x0000000000000080 );
+    casti_m256i( buf, vp++ ) = m256_const1_i128( 0x0000000000000080 );
      memset_zero_256( (__m256i*)buf + vp, 6 - vp );
   }

@@ -474,13 +487,13 @@ void shavite512_2way_full( shavite512_2way_context *ctx, void *dst,

   if ( vp == 0 )    // empty buf, xevan.
   {
-      casti_m256i( buf, 0 ) = m256_const2_64( 0, 0x0000000000000080 );
+      casti_m256i( buf, 0 ) = m256_const1_i128( 0x0000000000000080 );
      memset_zero_256( (__m256i*)buf + 1, 5 );
      ctx->count0 = ctx->count1 = ctx->count2 = ctx->count3 = 0;
   }
   else     // half full buf, everyone else.
   {
-    casti_m256i( buf, vp++ ) = m256_const2_64( 0, 0x0000000000000080 );
+    casti_m256i( buf, vp++ ) = m256_const1_i128( 0x0000000000000080 );
      memset_zero_256( (__m256i*)buf + vp, 6 - vp );
   }

--- a/algo/shavite/shavite-hash-4way.c
+++ b/algo/shavite/shavite-hash-4way.c
@@ -292,7 +292,7 @@ void shavite512_4way_close( shavite512_4way_context *ctx, void *dst )
    uint32_t vp = ctx->ptr>>6;

    // Terminating byte then zero pad
-    casti_m512i( buf, vp++ ) = m512_const2_64( 0, 0x0000000000000080 );
+    casti_m512i( buf, vp++ ) = m512_const1_i128( 0x0000000000000080 );

    // Zero pad full vectors up to count
    for ( ; vp < 6; vp++ )      
@@ -372,13 +372,13 @@ void shavite512_4way_update_close( shavite512_4way_context *ctx, void *dst,

   if ( vp == 0 )    // empty buf, xevan.
   { 
-      casti_m512i( buf, 0 ) = m512_const2_64( 0, 0x0000000000000080 );
+      casti_m512i( buf, 0 ) = m512_const1_i128( 0x0000000000000080 );
      memset_zero_512( (__m512i*)buf + 1, 5 );
      ctx->count0 = ctx->count1 = ctx->count2 = ctx->count3 = 0;
   }
   else     // half full buf, everyone else.
   {
-    casti_m512i( buf, vp++ ) = m512_const2_64( 0, 0x0000000000000080 );
+    casti_m512i( buf, vp++ ) = m512_const1_i128( 0x0000000000000080 );
      memset_zero_512( (__m512i*)buf + vp, 6 - vp );
   }

@@ -463,13 +463,13 @@ void shavite512_4way_full( shavite512_4way_context *ctx, void *dst,

   if ( vp == 0 )    // empty buf, xevan.
   {
-      casti_m512i( buf, 0 ) = m512_const2_64( 0, 0x0000000000000080 );
+      casti_m512i( buf, 0 ) = m512_const1_i128( 0x0000000000000080 );
      memset_zero_512( (__m512i*)buf + 1, 5 );
      ctx->count0 = ctx->count1 = ctx->count2 = ctx->count3 = 0;
   }
   else     // half full buf, everyone else.
   {
-    casti_m512i( buf, vp++ ) = m512_const2_64( 0, 0x0000000000000080 );
+    casti_m512i( buf, vp++ ) = m512_const1_i128( 0x0000000000000080 );
      memset_zero_512( (__m512i*)buf + vp, 6 - vp );
   }

--- a/algo/skein/skein-4way.c
+++ b/algo/skein/skein-4way.c
@@ -2,14 +2,8 @@
 #include <string.h>
 #include <stdint.h>
 #include "skein-hash-4way.h"
-
-// 8 way is faster than SHA on Icelake
-// SHA is faster than 4 way on Ryzen
-//
-#if defined(__SHA__)
-  #include <openssl/sha.h>
-#endif
 #include "algo/sha/sha-hash-4way.h"
+#include "algo/sha/sph_sha2.h"

 #if defined (SKEIN_8WAY)

@@ -93,7 +87,7 @@ void skeinhash_4way( void *state, const void *input )
     uint32_t hash1[16] __attribute__ ((aligned (64)));
     uint32_t hash2[16] __attribute__ ((aligned (64)));
     uint32_t hash3[16] __attribute__ ((aligned (64)));
-     SHA256_CTX ctx_sha256;
+     sph_sha256_context ctx_sha256;
 #else
     uint32_t vhash32[16*4] __attribute__ ((aligned (64)));
     sha256_4way_context ctx_sha256;
@@ -102,31 +96,29 @@ void skeinhash_4way( void *state, const void *input )
     skein512_4way_final16( &ctx_skein, vhash64, input + (64*4) );

 #if defined(__SHA__)      
+
     dintrlv_4x64( hash0, hash1, hash2, hash3, vhash64, 512 );
-
-     SHA256_Init( &ctx_sha256 );
-     SHA256_Update( &ctx_sha256, (unsigned char*)hash0, 64 );
-     SHA256_Final( (unsigned char*)hash0, &ctx_sha256 );
-
-     SHA256_Init( &ctx_sha256 );
-     SHA256_Update( &ctx_sha256, (unsigned char*)hash1, 64 );
-     SHA256_Final( (unsigned char*)hash1, &ctx_sha256 );
-
-     SHA256_Init( &ctx_sha256 );
-     SHA256_Update( &ctx_sha256, (unsigned char*)hash2, 64 );
-     SHA256_Final( (unsigned char*)hash2, &ctx_sha256 );
-
-     SHA256_Init( &ctx_sha256 );
-     SHA256_Update( &ctx_sha256, (unsigned char*)hash3, 64 );
-     SHA256_Final( (unsigned char*)hash3, &ctx_sha256 );
-
+     sph_sha256_init( &ctx_sha256 );
+     sph_sha256( &ctx_sha256, hash0, 64 );
+     sph_sha256_close( &ctx_sha256, hash0 );
+     sph_sha256_init( &ctx_sha256 );
+     sph_sha256( &ctx_sha256, hash1, 64 );
+     sph_sha256_close( &ctx_sha256, hash1 );
+     sph_sha256_init( &ctx_sha256 );
+     sph_sha256( &ctx_sha256, hash2, 64 );
+     sph_sha256_close( &ctx_sha256, hash2 );
+     sph_sha256_init( &ctx_sha256 );
+     sph_sha256( &ctx_sha256, hash3, 64 );
+     sph_sha256_close( &ctx_sha256, hash3 );
     intrlv_4x32( state, hash0, hash1, hash2, hash3, 256 );
-#else
-     rintrlv_4x64_4x32( vhash32, vhash64, 512 );

+#else
+
+     rintrlv_4x64_4x32( vhash32, vhash64, 512 );
     sha256_4way_init( &ctx_sha256 );
     sha256_4way_update( &ctx_sha256, vhash32, 64 );
     sha256_4way_close( &ctx_sha256, state );
+
 #endif
 }

--- a/algo/skein/skein-hash-4way.c
+++ b/algo/skein/skein-hash-4way.c
@@ -731,7 +731,7 @@ void skein512_8way_full( skein512_8way_context *sc, void *out, const void *data,
 void
 skein512_8way_prehash64( skein512_8way_context *sc, const void *data )
 {
-   __m512i *vdata = (__m512*)data;
+   __m512i *vdata = (__m512i*)data;
   __m512i *buf = sc->buf;
   buf[0] = vdata[0];
   buf[1] = vdata[1];
--- a/algo/skein/skein.c
+++ b/algo/skein/skein.c
@@ -5,21 +5,21 @@
 #include <string.h>
 #include <stdint.h>
 #include "sph_skein.h"
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"

 void skeinhash(void *state, const void *input)
 {
     uint32_t hash[16] __attribute__ ((aligned (64)));
     sph_skein512_context ctx_skein;
-     SHA256_CTX           ctx_sha256;
+     sph_sha256_context   ctx_sha256;

     sph_skein512_init( &ctx_skein );
     sph_skein512( &ctx_skein, input, 80 );
     sph_skein512_close( &ctx_skein, hash );

-     SHA256_Init( &ctx_sha256 );
-     SHA256_Update( &ctx_sha256, (unsigned char*)hash, 64 );
-     SHA256_Final( (unsigned char*) hash, &ctx_sha256 );
+     sph_sha256_init( &ctx_sha256 );
+     sph_sha256( &ctx_sha256, hash, 64 );
+     sph_sha256_close( &ctx_sha256, hash );

     memcpy(state, hash, 32);
 }
--- a/algo/swifftx/stdbool.h
+++ b/algo/swifftx/stdbool.h
@@ -1,47 +0,0 @@
-/*
- * Copyright (c) 2000 Jeroen Ruigrok van der Werven <asmodai@FreeBSD.org>
- * 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.
- *
- * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
- * ANY EXPRESS 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 AUTHOR OR 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.
- *
- * $FreeBSD: src/include/stdbool.h,v 1.6 2002/08/16 07:33:14 alfred Exp $
- */
-
-#ifndef _STDBOOL_H_
-#define _STDBOOL_H_     
-
-#define __bool_true_false_are_defined   1
-
-#ifndef __cplusplus
-
-#define false   0
-#define true    1
-
-//#define bool    _Bool
-//#if __STDC_VERSION__ < 199901L && __GNUC__ < 3
-//typedef int     _Bool;
-//#endif
-typedef int bool;
-
-#endif /* !__cplusplus */
-
-#endif /* !_STDBOOL_H_ */
--- a/algo/swifftx/swifftx.c.bak
+++ b/algo/swifftx/swifftx.c.bak
--- a/algo/x13/phi1612-4way.c
+++ b/algo/x13/phi1612-4way.c
@@ -7,7 +7,7 @@
 #include "algo/jh/jh-hash-4way.h"
 #include "algo/cubehash/cubehash_sse2.h"
 #include "algo/cubehash/cube-hash-2way.h"
-#include "algo/fugue/sph_fugue.h"
+#include "algo/fugue/fugue-aesni.h"
 #include "algo/gost/sph_gost.h"
 #include "algo/echo/aes_ni/hash_api.h"
 #if defined(__VAES__)
@@ -20,7 +20,7 @@ typedef struct {
    skein512_8way_context   skein;
    jh512_8way_context      jh;
    cube_4way_context       cube;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    sph_gost512_context     gost;
 #if defined(__VAES__)
    echo_4way_context       echo;
@@ -36,7 +36,7 @@ void init_phi1612_8way_ctx()
     skein512_8way_init( &phi1612_8way_ctx.skein );
     jh512_8way_init( &phi1612_8way_ctx.jh );
     cube_4way_init( &phi1612_8way_ctx.cube, 512, 16, 32 );
-     sph_fugue512_init( &phi1612_8way_ctx.fugue );
+     fugue512_Init( &phi1612_8way_ctx.fugue, 512 );
     sph_gost512_init( &phi1612_8way_ctx.gost );
 #if defined(__VAES__)
     echo_4way_init( &phi1612_8way_ctx.echo, 512 );
@@ -79,29 +79,14 @@ void phi1612_8way_hash( void *state, const void *input )
     dintrlv_4x128_512( hash4, hash5, hash6, hash7, vhash );

     // Fugue
-     sph_fugue512( &ctx.fugue, hash0, 64 );
-     sph_fugue512_close( &ctx.fugue, hash0 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash1, 64 );
-     sph_fugue512_close( &ctx.fugue, hash1 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash2, 64 );
-     sph_fugue512_close( &ctx.fugue, hash2 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash3, 64 );
-     sph_fugue512_close( &ctx.fugue, hash3 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash4, 64 );
-     sph_fugue512_close( &ctx.fugue, hash4 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash5, 64 );
-     sph_fugue512_close( &ctx.fugue, hash5 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash6, 64 );
-     sph_fugue512_close( &ctx.fugue, hash6 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash7, 64 );
-     sph_fugue512_close( &ctx.fugue, hash7 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash4, hash4, 64 );
+     fugue512_full( &ctx.fugue, hash5, hash5, 64 );
+     fugue512_full( &ctx.fugue, hash6, hash6, 64 );
+     fugue512_full( &ctx.fugue, hash7, hash7, 64 );

     // Gost
     sph_gost512( &ctx.gost, hash0, 64 );
@@ -223,7 +208,7 @@ typedef struct {
    skein512_4way_context   skein;
    jh512_4way_context      jh;
    cubehashParam           cube;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    sph_gost512_context     gost;
    hashState_echo          echo;
 } phi1612_4way_ctx_holder;
@@ -235,7 +220,6 @@ void init_phi1612_4way_ctx()
     skein512_4way_init( &phi1612_4way_ctx.skein );
     jh512_4way_init( &phi1612_4way_ctx.jh );
     cubehashInit( &phi1612_4way_ctx.cube, 512, 16, 32 );
-     sph_fugue512_init( &phi1612_4way_ctx.fugue );
     sph_gost512_init( &phi1612_4way_ctx.gost );
     init_echo( &phi1612_4way_ctx.echo, 512 );
 };
@@ -275,17 +259,10 @@ void phi1612_4way_hash( void *state, const void *input )
     cubehashUpdateDigest( &ctx.cube, (byte*)hash3, (const byte*) hash3, 64 );

     // Fugue
-     sph_fugue512( &ctx.fugue, hash0, 64 );
-     sph_fugue512_close( &ctx.fugue, hash0 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash1, 64 );
-     sph_fugue512_close( &ctx.fugue, hash1 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash2, 64 );
-     sph_fugue512_close( &ctx.fugue, hash2 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash3, 64 );
-     sph_fugue512_close( &ctx.fugue, hash3 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );

     // Gost
     sph_gost512( &ctx.gost, hash0, 64 );
--- a/algo/x13/phi1612.c
+++ b/algo/x13/phi1612.c
@@ -8,24 +8,28 @@
 #include <stdio.h>
 #include "algo/gost/sph_gost.h"
 #include "algo/echo/sph_echo.h"
-#include "algo/fugue/sph_fugue.h"
 #include "algo/cubehash/cubehash_sse2.h"
 #include "algo/skein/sph_skein.h"
 #include "algo/jh/sph_jh.h"
 #ifdef __AES__
  #include "algo/echo/aes_ni/hash_api.h"
+  #include "algo/fugue/fugue-aesni.h"
+#else
+  #include "algo/echo/sph_echo.h"
+  #include "algo/fugue/sph_fugue.h"
 #endif

 typedef struct {
     sph_skein512_context    skein;
     sph_jh512_context       jh;
     cubehashParam           cube;
-     sph_fugue512_context    fugue;
     sph_gost512_context     gost;
 #ifdef __AES__
     hashState_echo          echo;
+     hashState_fugue         fugue;
 #else
     sph_echo512_context     echo;
+     sph_fugue512_context    fugue;
 #endif
 } phi_ctx_holder;

@@ -38,12 +42,13 @@ void init_phi1612_ctx()
     sph_skein512_init( &phi_ctx.skein );
     sph_jh512_init( &phi_ctx.jh );
     cubehashInit( &phi_ctx.cube, 512, 16, 32 );
-     sph_fugue512_init( &phi_ctx.fugue );
     sph_gost512_init( &phi_ctx.gost );
 #ifdef __AES__
     init_echo( &phi_ctx.echo, 512 );
+     fugue512_Init( &phi_ctx.fugue, 512 );
 #else
     sph_echo512_init( &phi_ctx.echo );
+     sph_fugue512_init( &phi_ctx.fugue );
 #endif
 }

@@ -69,8 +74,13 @@ void phi1612_hash(void *output, const void *input)

     cubehashUpdateDigest( &ctx.cube, (byte*) hash, (const byte*)hash, 64 );

+#if defined(__AES__)
+     fugue512_Update( &ctx.fugue, hash, 512 ); 
+     fugue512_Final( &ctx.fugue, hash ); 
+#else
     sph_fugue512( &ctx.fugue, (const void*)hash, 64 );
     sph_fugue512_close( &ctx.fugue, (void*)hash );
+#endif

     sph_gost512( &ctx.gost, hash, 64 );
     sph_gost512_close( &ctx.gost, hash );
--- a/algo/x13/skunk-4way.c
+++ b/algo/x13/skunk-4way.c
@@ -5,7 +5,7 @@
 #include <stdio.h>
 #include "algo/skein/skein-hash-4way.h"
 #include "algo/gost/sph_gost.h"
-#include "algo/fugue/sph_fugue.h"
+#include "algo/fugue/fugue-aesni.h"
 #include "algo/cubehash/cubehash_sse2.h"
 #include "algo/cubehash/cube-hash-2way.h"

@@ -14,7 +14,7 @@
 typedef struct {
    skein512_8way_context skein;
    cube_4way_context     cube;
-    sph_fugue512_context  fugue;
+    hashState_fugue         fugue;
    sph_gost512_context   gost;
 } skunk_8way_ctx_holder;

@@ -46,29 +46,15 @@ void skunk_8way_hash( void *output, const void *input )
     cube_4way_init( &ctx.cube, 512, 16, 32 );           
     cube_4way_update_close( &ctx.cube, vhash, vhash, 64 );  
     dintrlv_4x128_512( hash4, hash5, hash6, hash7, vhash );
-     
-     sph_fugue512( &ctx.fugue, hash0, 64 );
-     sph_fugue512_close( &ctx.fugue, hash0 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash1, 64 );
-     sph_fugue512_close( &ctx.fugue, hash1 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash2, 64 );
-     sph_fugue512_close( &ctx.fugue, hash2 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash3, 64 );
-     sph_fugue512_close( &ctx.fugue, hash3 );
-     sph_fugue512( &ctx.fugue, hash4, 64 );
-     sph_fugue512_close( &ctx.fugue, hash4 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash5, 64 );
-     sph_fugue512_close( &ctx.fugue, hash5 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash6, 64 );
-     sph_fugue512_close( &ctx.fugue, hash6 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash7, 64 );
-     sph_fugue512_close( &ctx.fugue, hash7 );
+
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash4, hash4, 64 );
+     fugue512_full( &ctx.fugue, hash5, hash5, 64 );
+     fugue512_full( &ctx.fugue, hash6, hash6, 64 );
+     fugue512_full( &ctx.fugue, hash7, hash7, 64 );

     sph_gost512( &ctx.gost, hash0, 64 );
     sph_gost512_close( &ctx.gost, output );
@@ -140,7 +126,6 @@ bool skunk_8way_thread_init()
 {
   skein512_8way_init( &skunk_8way_ctx.skein );
   cube_4way_init( &skunk_8way_ctx.cube, 512, 16, 32 );
-   sph_fugue512_init( &skunk_8way_ctx.fugue );
   sph_gost512_init( &skunk_8way_ctx.gost );
   return true;
 }
@@ -150,7 +135,7 @@ bool skunk_8way_thread_init()
 typedef struct {
    skein512_4way_context skein;
    cubehashParam         cube;
-    sph_fugue512_context  fugue;
+    hashState_fugue       fugue;
    sph_gost512_context   gost;
 } skunk_4way_ctx_holder;

@@ -178,17 +163,10 @@ void skunk_4way_hash( void *output, const void *input )
     memcpy( &ctx.cube, &skunk_4way_ctx.cube, sizeof(cubehashParam) );
     cubehashUpdateDigest( &ctx.cube, (byte*)hash3, (const byte*) hash3, 64 );

-     sph_fugue512( &ctx.fugue, hash0, 64 );
-     sph_fugue512_close( &ctx.fugue, hash0 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash1, 64 );
-     sph_fugue512_close( &ctx.fugue, hash1 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash2, 64 );
-     sph_fugue512_close( &ctx.fugue, hash2 );
-     sph_fugue512_init( &ctx.fugue );
-     sph_fugue512( &ctx.fugue, hash3, 64 );
-     sph_fugue512_close( &ctx.fugue, hash3 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );

     sph_gost512( &ctx.gost, hash0, 64 );
     sph_gost512_close( &ctx.gost, hash0 );
@@ -252,7 +230,6 @@ bool skunk_4way_thread_init()
 {
   skein512_4way_init( &skunk_4way_ctx.skein );
   cubehashInit( &skunk_4way_ctx.cube, 512, 16, 32 );
-   sph_fugue512_init( &skunk_4way_ctx.fugue );
   sph_gost512_init( &skunk_4way_ctx.gost );
   return true;
 }
--- a/algo/x13/skunk-gate.c
+++ b/algo/x13/skunk-gate.c
@@ -2,7 +2,7 @@

 bool register_skunk_algo( algo_gate_t* gate )
 {
-   gate->optimizations = SSE2_OPT | AVX2_OPT | AVX512_OPT;
+   gate->optimizations = SSE2_OPT | AVX2_OPT | AVX512_OPT | AES_OPT;
 #if defined (SKUNK_8WAY)
   gate->miner_thread_init = (void*)&skunk_8way_thread_init;
   gate->scanhash = (void*)&scanhash_skunk_8way;
--- a/algo/x13/skunk.c
+++ b/algo/x13/skunk.c
@@ -8,13 +8,21 @@
 #include <stdio.h>
 #include "algo/gost/sph_gost.h"
 #include "algo/skein/sph_skein.h"
-#include "algo/fugue/sph_fugue.h"
 #include "algo/cubehash/cubehash_sse2.h"
+#if defined(__AES__)
+  #include "algo/fugue/fugue-aesni.h"
+#else
+  #include "algo/fugue/sph_fugue.h"
+#endif

 typedef struct {
    sph_skein512_context  skein;
    cubehashParam         cube;
+#if defined(__AES__)
+    hashState_fugue       fugue;
+#else
    sph_fugue512_context  fugue;
+#endif
    sph_gost512_context   gost;
 } skunk_ctx_holder;

@@ -32,8 +40,13 @@ void skunkhash( void *output, const void *input )

     cubehashUpdateDigest( &ctx.cube, (byte*) hash, (const byte*)hash, 64 );

+#if defined(__AES__)
+     fugue512_Update( &ctx.fugue, hash, 512 ); 
+     fugue512_Final( &ctx.fugue, hash ); 
+#else
     sph_fugue512( &ctx.fugue, hash, 64 );
     sph_fugue512_close( &ctx.fugue, hash );
+#endif

     sph_gost512( &ctx.gost, hash, 64 );
     sph_gost512_close( &ctx.gost, hash );
@@ -87,8 +100,12 @@ bool skunk_thread_init()
 {
   sph_skein512_init( &skunk_ctx.skein );
   cubehashInit( &skunk_ctx.cube, 512, 16, 32 );
-   sph_fugue512_init( &skunk_ctx.fugue );
-   sph_gost512_init( &skunk_ctx.gost );
+#if defined(__AES__)
+    fugue512_Init( &skunk_ctx.fugue, 512 );
+#else
+    sph_fugue512_init( &skunk_ctx.fugue );
+#endif
+    sph_gost512_init( &skunk_ctx.gost );
   return true;
 }
 #endif
--- a/algo/x13/x13-4way.c
+++ b/algo/x13/x13-4way.c
@@ -16,7 +16,7 @@
 #include "algo/simd/simd-hash-2way.h"
 #include "algo/echo/aes_ni/hash_api.h"
 #include "algo/hamsi/hamsi-hash-4way.h"
-#include "algo/fugue/sph_fugue.h"
+#include "algo/fugue/fugue-aesni.h"
 #if defined(__VAES__)
  #include "algo/groestl/groestl512-hash-4way.h"
  #include "algo/shavite/shavite-hash-4way.h"
@@ -35,7 +35,7 @@ typedef struct {
    cube_4way_context       cube;
    simd_4way_context       simd;
    hamsi512_8way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
 #if defined(__VAES__)
    groestl512_4way_context groestl;
    shavite512_4way_context shavite;
@@ -60,7 +60,7 @@ void init_x13_8way_ctx()
     cube_4way_init( &x13_8way_ctx.cube, 512, 16, 32 );
     simd_4way_init( &x13_8way_ctx.simd, 512 );
     hamsi512_8way_init( &x13_8way_ctx.hamsi );
-     sph_fugue512_init( &x13_8way_ctx.fugue );
+     fugue512_Init( &x13_8way_ctx.fugue, 512 );
 #if defined(__VAES__)
     groestl512_4way_init( &x13_8way_ctx.groestl, 64 );
     shavite512_4way_init( &x13_8way_ctx.shavite );
@@ -255,29 +255,29 @@ void x13_8way_hash( void *state, const void *input )
                       vhash );

     // 13 Fugue serial
-     sph_fugue512( &ctx.fugue, hash0, 64 );
-     sph_fugue512_close( &ctx.fugue, hash0 );
-     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash1, 64 );
-     sph_fugue512_close( &ctx.fugue, hash1 );
-     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash2, 64 );
-     sph_fugue512_close( &ctx.fugue, hash2 );
-     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash3, 64 );
-     sph_fugue512_close( &ctx.fugue, hash3 );
-     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash4, 64 );
-     sph_fugue512_close( &ctx.fugue, hash4 );
-     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash5, 64 );
-     sph_fugue512_close( &ctx.fugue, hash5 );
-     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash6, 64 );
-     sph_fugue512_close( &ctx.fugue, hash6 );
-     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash7, 64 );
-     sph_fugue512_close( &ctx.fugue, hash7 );
+     fugue512_Update( &ctx.fugue, hash0, 512 );
+     fugue512_Final( &ctx.fugue, hash0 ); 
+     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash1, 512 );
+     fugue512_Final( &ctx.fugue, hash1 );   
+     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash2, 512 );
+     fugue512_Final( &ctx.fugue, hash2 );   
+     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash3, 512 );
+     fugue512_Final( &ctx.fugue, hash3 );   
+     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash4, 512 );
+     fugue512_Final( &ctx.fugue, hash4 );   
+     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash5, 512 );
+     fugue512_Final( &ctx.fugue, hash5 );   
+     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash6, 512 );
+     fugue512_Final( &ctx.fugue, hash6 );   
+     memcpy( &ctx.fugue, &x13_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash7, 512 );
+     fugue512_Final( &ctx.fugue, hash7 );   
     
     memcpy( state,     hash0, 32 );
     memcpy( state+ 32, hash1, 32 );
@@ -344,7 +344,7 @@ typedef struct {
    simd_2way_context       simd;
    hashState_echo          echo;
    hamsi512_4way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
 } x13_4way_ctx_holder;

 x13_4way_ctx_holder x13_4way_ctx __attribute__ ((aligned (64)));
@@ -363,7 +363,7 @@ void init_x13_4way_ctx()
     simd_2way_init( &x13_4way_ctx.simd, 512 );
     init_echo( &x13_4way_ctx.echo, 512 );
     hamsi512_4way_init( &x13_4way_ctx.hamsi );
-     sph_fugue512_init( &x13_4way_ctx.fugue );
+     fugue512_Init( &x13_4way_ctx.fugue, 512 );
 };

 void x13_4way_hash( void *state, const void *input )
@@ -477,17 +477,17 @@ void x13_4way_hash( void *state, const void *input )
     dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );

     // 13 Fugue serial
-     sph_fugue512( &ctx.fugue, hash0, 64 );
-     sph_fugue512_close( &ctx.fugue, hash0 );
-     memcpy( &ctx.fugue, &x13_4way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash1, 64 );
-     sph_fugue512_close( &ctx.fugue, hash1 );
-     memcpy( &ctx.fugue, &x13_4way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash2, 64 );
-     sph_fugue512_close( &ctx.fugue, hash2 );
-     memcpy( &ctx.fugue, &x13_4way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash3, 64 );
-     sph_fugue512_close( &ctx.fugue, hash3 );
+     fugue512_Update( &ctx.fugue, hash0, 512 );
+     fugue512_Final( &ctx.fugue, hash0 );      
+     memcpy( &ctx.fugue, &x13_4way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash1, 512 );
+     fugue512_Final( &ctx.fugue, hash1 );       
+     memcpy( &ctx.fugue, &x13_4way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash2, 512 );
+     fugue512_Final( &ctx.fugue, hash2 );      
+     memcpy( &ctx.fugue, &x13_4way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash3, 512 );
+     fugue512_Final( &ctx.fugue, hash3 );   

     memcpy( state,    hash0, 32 );
     memcpy( state+32, hash1, 32 );
--- a/algo/x13/x13.c
+++ b/algo/x13/x13.c
@@ -13,7 +13,6 @@
 #include "algo/skein/sph_skein.h"
 #include "algo/shavite/sph_shavite.h"
 #include "algo/hamsi/sph_hamsi.h"
-#include "algo/fugue/sph_fugue.h"
 #include "algo/luffa/luffa_for_sse2.h"
 #include "algo/cubehash/cubehash_sse2.h"
 #include "algo/simd/nist.h"
@@ -21,9 +20,11 @@
 #if defined(__AES__)
  #include "algo/echo/aes_ni/hash_api.h"
  #include "algo/groestl/aes_ni/hash-groestl.h"
+  #include "algo/fugue/fugue-aesni.h"
 #else
  #include "algo/groestl/sph_groestl.h"
  #include "algo/echo/sph_echo.h"
+  #include "algo/fugue/sph_fugue.h"
 #endif

 typedef struct {
@@ -32,9 +33,11 @@ typedef struct {
 #if defined(__AES__)
   hashState_echo          echo;
   hashState_groestl       groestl;
+   hashState_fugue         fugue;
 #else
   sph_groestl512_context   groestl;
   sph_echo512_context      echo;
+   sph_fugue512_context    fugue;
 #endif
   sph_jh512_context       jh;
   sph_keccak512_context   keccak;
@@ -44,7 +47,6 @@ typedef struct {
   sph_shavite512_context  shavite;
   hashState_sd            simd;
   sph_hamsi512_context    hamsi;
-   sph_fugue512_context    fugue;
 } x13_ctx_holder;

 x13_ctx_holder x13_ctx;
@@ -56,9 +58,11 @@ void init_x13_ctx()
 #if defined(__AES__)
   init_groestl( &x13_ctx.groestl, 64 );
   init_echo( &x13_ctx.echo, 512 );
+   fugue512_Init( &x13_ctx.fugue, 512 );
 #else
   sph_groestl512_init( &x13_ctx.groestl );
   sph_echo512_init( &x13_ctx.echo );
+   sph_fugue512_init( &x13_ctx.fugue );
 #endif
   sph_skein512_init( &x13_ctx.skein );
   sph_jh512_init( &x13_ctx.jh );
@@ -68,7 +72,6 @@ void init_x13_ctx()
   sph_shavite512_init( &x13_ctx.shavite );
   init_sd( &x13_ctx.simd, 512 );
   sph_hamsi512_init( &x13_ctx.hamsi );
-   sph_fugue512_init( &x13_ctx.fugue );
 };

 void x13hash(void *output, const void *input)
@@ -84,11 +87,9 @@ void x13hash(void *output, const void *input)
    sph_bmw512_close( &ctx.bmw, hash );

 #if defined(__AES__)
-    init_groestl( &ctx.groestl, 64 );
    update_and_final_groestl( &ctx.groestl, (char*)hash,
                                      (const char*)hash, 512 );
 #else
-    sph_groestl512_init( &ctx.groestl );
    sph_groestl512( &ctx.groestl, hash, 64 );
    sph_groestl512_close( &ctx.groestl, hash );
 #endif
@@ -125,8 +126,13 @@ void x13hash(void *output, const void *input)
    sph_hamsi512( &ctx.hamsi, hash, 64 );
    sph_hamsi512_close( &ctx.hamsi, hash );

+#if defined(__AES__)
+    fugue512_Update( &ctx.fugue, hash, 512 );
+    fugue512_Final( &ctx.fugue, hash );  
+#else
    sph_fugue512( &ctx.fugue, hash, 64 );
-    sph_fugue512_close( &ctx.fugue, hash );
+    sph_fugue512_close( &ctx.fugue, hash ); 
+#endif

 	 memcpy( output, hash, 32 );
 }
--- a/algo/x13/x13bcd-4way.c
+++ b/algo/x13/x13bcd-4way.c
@@ -16,7 +16,7 @@
 #include "algo/echo/aes_ni/hash_api.h"
 #include "algo/sm3/sm3-hash-4way.h"
 #include "algo/hamsi/hamsi-hash-4way.h"
-#include "algo/fugue/sph_fugue.h"
+#include "algo/fugue/fugue-aesni.h"
 #if defined(__VAES__)
  #include "algo/groestl/groestl512-hash-4way.h"
  #include "algo/shavite/shavite-hash-4way.h"
@@ -35,7 +35,7 @@ typedef struct {
    simd_4way_context       simd;
    sm3_8way_ctx_t          sm3;
    hamsi512_8way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
 #if defined(__VAES__)
    groestl512_4way_context groestl;
    shavite512_4way_context shavite;
@@ -61,7 +61,7 @@ void init_x13bcd_8way_ctx()
     simd_4way_init( &x13bcd_8way_ctx.simd, 512 );
     sm3_8way_init( &x13bcd_8way_ctx.sm3 );
     hamsi512_8way_init( &x13bcd_8way_ctx.hamsi );
-     sph_fugue512_init( &x13bcd_8way_ctx.fugue );
+     fugue512_Init( &x13bcd_8way_ctx.fugue, 512 );
 #if defined(__VAES__)
     groestl512_4way_init( &x13bcd_8way_ctx.groestl, 64 );
     shavite512_4way_init( &x13bcd_8way_ctx.shavite );
@@ -257,36 +257,30 @@ void x13bcd_8way_hash( void *state, const void *input )
                       hash4, hash5, hash6, hash7, vhash );

     // Fugue serial
-     sph_fugue512( &ctx.fugue, hash0, 64 );
-     sph_fugue512_close( &ctx.fugue, state );
-     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue,
-                         sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash1, 64 );
-     sph_fugue512_close( &ctx.fugue, state+32 );
-     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue,
-                         sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash2, 64 );
-     sph_fugue512_close( &ctx.fugue, state+64 );
-     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue,
-                         sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash3, 64 );
-     sph_fugue512_close( &ctx.fugue, state+96 );
-     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue,
-                         sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash4, 64 );
-     sph_fugue512_close( &ctx.fugue, state+128 );
-     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue,
-                         sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash5, 64 );
-     sph_fugue512_close( &ctx.fugue, state+160 );
-     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue,
-                         sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash6, 64 );
-     sph_fugue512_close( &ctx.fugue, state+192 );
-     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue,
-                         sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash7, 64 );
-     sph_fugue512_close( &ctx.fugue, state+224 );
+     fugue512_Update( &ctx.fugue, hash0, 512 );
+     fugue512_Final( &ctx.fugue, state );
+     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash1, 512 );
+     fugue512_Final( &ctx.fugue, state+32 );
+     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash2, 512 );
+     fugue512_Final( &ctx.fugue, state+64 );
+     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash3, 512 );
+     fugue512_Final( &ctx.fugue, state+96 );
+     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash4, 512 );
+     fugue512_Final( &ctx.fugue, state+128 );
+     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash5, 512 );
+     fugue512_Final( &ctx.fugue, state+160 );
+     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash6, 512 );
+     fugue512_Final( &ctx.fugue, state+192 );
+     memcpy( &ctx.fugue, &x13bcd_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash7, 512 );
+     fugue512_Final( &ctx.fugue, state+224 );
+
 }

 int scanhash_x13bcd_8way( struct work *work, uint32_t max_nonce,
@@ -346,7 +340,7 @@ typedef struct {
    hashState_echo          echo;
    sm3_4way_ctx_t          sm3;
    hamsi512_4way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
 } x13bcd_4way_ctx_holder;

 x13bcd_4way_ctx_holder x13bcd_4way_ctx __attribute__ ((aligned (64)));
@@ -366,7 +360,7 @@ void init_x13bcd_4way_ctx()
     init_echo( &x13bcd_4way_ctx.echo, 512 );
     sm3_4way_init( &x13bcd_4way_ctx.sm3 );
     hamsi512_4way_init( &x13bcd_4way_ctx.hamsi );
-     sph_fugue512_init( &x13bcd_4way_ctx.fugue );
+     fugue512_Init( &x13bcd_4way_ctx.fugue, 512 );
 };

 void x13bcd_4way_hash( void *state, const void *input )
@@ -489,20 +483,17 @@ void x13bcd_4way_hash( void *state, const void *input )
     dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );

     // Fugue serial
-     sph_fugue512( &ctx.fugue, hash0, 64 );
-     sph_fugue512_close( &ctx.fugue, hash0 );
-     memcpy( &ctx.fugue, &x13bcd_4way_ctx.fugue,
-                         sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash1, 64 );
-     sph_fugue512_close( &ctx.fugue, hash1 );
-     memcpy( &ctx.fugue, &x13bcd_4way_ctx.fugue,
-                         sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash2, 64 );
-     sph_fugue512_close( &ctx.fugue, hash2 );
-     memcpy( &ctx.fugue, &x13bcd_4way_ctx.fugue,
-                         sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash3, 64 );
-     sph_fugue512_close( &ctx.fugue, hash3 );
+     fugue512_Update( &ctx.fugue, hash0, 512 );
+     fugue512_Final( &ctx.fugue, hash0 );
+     memcpy( &ctx.fugue, &x13bcd_4way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash1, 512 );
+     fugue512_Final( &ctx.fugue, hash1 );
+     memcpy( &ctx.fugue, &x13bcd_4way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash2, 512 );
+     fugue512_Final( &ctx.fugue, hash2 );
+     memcpy( &ctx.fugue, &x13bcd_4way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash3, 512 );
+     fugue512_Final( &ctx.fugue, hash3 );

     memcpy( state,    hash0, 32 );
     memcpy( state+32, hash1, 32 );
--- a/algo/x13/x13bcd.c
+++ b/algo/x13/x13bcd.c
@@ -14,16 +14,17 @@
 #include "algo/skein/sph_skein.h"
 #include "algo/shavite/sph_shavite.h"
 #include "algo/hamsi/sph_hamsi.h"
-#include "algo/fugue/sph_fugue.h"
 #include "algo/cubehash/cubehash_sse2.h"
 #include "algo/simd/nist.h"

 #if defined(__AES__)
  #include "algo/echo/aes_ni/hash_api.h"
  #include "algo/groestl/aes_ni/hash-groestl.h"
+  #include "algo/fugue/fugue-aesni.h"
 #else
  #include "algo/groestl/sph_groestl.h"
  #include "algo/echo/sph_echo.h"
+  #include "algo/fugue/sph_fugue.h"
 #endif

 typedef struct {
@@ -32,9 +33,11 @@ typedef struct {
 #if defined(__AES__)
   hashState_echo          echo;
   hashState_groestl       groestl;
+   hashState_fugue         fugue;
 #else
   sph_groestl512_context   groestl;
   sph_echo512_context      echo;
+   sph_fugue512_context    fugue;
 #endif
   sph_jh512_context       jh;
   sph_keccak512_context   keccak;
@@ -43,7 +46,6 @@ typedef struct {
   sph_shavite512_context  shavite;
   hashState_sd            simd;
   sph_hamsi512_context    hamsi;
-   sph_fugue512_context    fugue;
   sm3_ctx_t               sm3;
 } x13bcd_ctx_holder;

@@ -56,9 +58,11 @@ void init_x13bcd_ctx()
 #if defined(__AES__)
   init_groestl( &x13bcd_ctx.groestl, 64 );
   init_echo( &x13bcd_ctx.echo, 512 );
+   fugue512_Init( &x13bcd_ctx.fugue, 512 );
 #else
   sph_groestl512_init( &x13bcd_ctx.groestl );
   sph_echo512_init( &x13bcd_ctx.echo );
+   sph_fugue512_init( &x13bcd_ctx.fugue );
 #endif
   sph_skein512_init( &x13bcd_ctx.skein );
   sph_jh512_init( &x13bcd_ctx.jh );
@@ -68,7 +72,6 @@ void init_x13bcd_ctx()
   init_sd( &x13bcd_ctx.simd,512 );
   sm3_init( &x13bcd_ctx.sm3 );
   sph_hamsi512_init( &x13bcd_ctx.hamsi );
-   sph_fugue512_init( &x13bcd_ctx.fugue );
 };

 void x13bcd_hash(void *output, const void *input)
@@ -129,8 +132,13 @@ void x13bcd_hash(void *output, const void *input)
    sph_hamsi512( &ctx.hamsi, hash, 64 );
    sph_hamsi512_close( &ctx.hamsi, hash );

+#if defined(__AES__)
+    fugue512_Update( &ctx.fugue, hash, 512 );
+    fugue512_Final( &ctx.fugue, hash );
+#else
    sph_fugue512( &ctx.fugue, hash, 64 );
    sph_fugue512_close( &ctx.fugue, hash );
+#endif

    memcpy( output, hash, 32 );
 }
--- a/algo/x14/x14-4way.c
+++ b/algo/x14/x14-4way.c
@@ -17,7 +17,7 @@
 #include "algo/echo/aes_ni/hash_api.h"
 #include "algo/echo/sph_echo.h"
 #include "algo/hamsi/hamsi-hash-4way.h"
-#include "algo/fugue/sph_fugue.h"
+#include "algo/fugue/fugue-aesni.h"
 #include "algo/shabal/shabal-hash-4way.h"
 #if defined(__VAES__)
  #include "algo/groestl/groestl512-hash-4way.h"
@@ -37,7 +37,7 @@ typedef struct {
    cube_4way_context       cube;
    simd_4way_context       simd;
    hamsi512_8way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_8way_context  shabal;
 #if defined(__VAES__)
    groestl512_4way_context groestl;
@@ -63,7 +63,7 @@ void init_x14_8way_ctx()
     cube_4way_init( &x14_8way_ctx.cube, 512, 16, 32 );
     simd_4way_init( &x14_8way_ctx.simd, 512 );
     hamsi512_8way_init( &x14_8way_ctx.hamsi );
-     sph_fugue512_init( &x14_8way_ctx.fugue );
+     fugue512_Init( &x14_8way_ctx.fugue, 512 );
     shabal512_8way_init( &x14_8way_ctx.shabal );
 #if defined(__VAES__)
     groestl512_4way_init( &x14_8way_ctx.groestl, 64 );
@@ -259,29 +259,29 @@ void x14_8way_hash( void *state, const void *input )
                       vhash );

     // 13 Fugue serial
-     sph_fugue512( &ctx.fugue, hash0, 64 );
-     sph_fugue512_close( &ctx.fugue, hash0 );
-     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash1, 64 );
-     sph_fugue512_close( &ctx.fugue, hash1 );
-     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash2, 64 );
-     sph_fugue512_close( &ctx.fugue, hash2 );
-     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash3, 64 );
-     sph_fugue512_close( &ctx.fugue, hash3 );
-     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash4, 64 );
-     sph_fugue512_close( &ctx.fugue, hash4 );
-     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash5, 64 );
-     sph_fugue512_close( &ctx.fugue, hash5 );
-     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash6, 64 );
-     sph_fugue512_close( &ctx.fugue, hash6 );
-     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash7, 64 );
-     sph_fugue512_close( &ctx.fugue, hash7 );
+     fugue512_Update( &ctx.fugue, hash0, 512 );
+     fugue512_Final( &ctx.fugue, hash0 );
+     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash1, 512 );
+     fugue512_Final( &ctx.fugue, hash1 );
+     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash2, 512 );
+     fugue512_Final( &ctx.fugue, hash2 );
+     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash3, 512 );
+     fugue512_Final( &ctx.fugue, hash3 );
+     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash4, 512 );
+     fugue512_Final( &ctx.fugue, hash4 );
+     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash5, 512 );
+     fugue512_Final( &ctx.fugue, hash5 );
+     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash6, 512 );
+     fugue512_Final( &ctx.fugue, hash6 );
+     memcpy( &ctx.fugue, &x14_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash7, 512 );
+     fugue512_Final( &ctx.fugue, hash7 );

     // 14 Shabal, parallel 32 bit
     intrlv_8x32_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
@@ -348,7 +348,7 @@ typedef struct {
    simd_2way_context       simd;
    hashState_echo          echo;
    hamsi512_4way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_4way_context  shabal;
 } x14_4way_ctx_holder;

@@ -368,7 +368,7 @@ void init_x14_4way_ctx()
     simd_2way_init( &x14_4way_ctx.simd, 512 );
     init_echo( &x14_4way_ctx.echo, 512 );
     hamsi512_4way_init( &x14_4way_ctx.hamsi );
-     sph_fugue512_init( &x14_4way_ctx.fugue );
+     fugue512_Init( &x14_4way_ctx.fugue, 512 );
     shabal512_4way_init( &x14_4way_ctx.shabal );
 };

@@ -483,17 +483,17 @@ void x14_4way_hash( void *state, const void *input )
     dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );

     // 13 Fugue serial
-     sph_fugue512( &ctx.fugue, hash0, 64 );
-     sph_fugue512_close( &ctx.fugue, hash0 );
-     memcpy( &ctx.fugue, &x14_4way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash1, 64 );
-     sph_fugue512_close( &ctx.fugue, hash1 );
-     memcpy( &ctx.fugue, &x14_4way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash2, 64 );
-     sph_fugue512_close( &ctx.fugue, hash2 );
-     memcpy( &ctx.fugue, &x14_4way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash3, 64 );
-     sph_fugue512_close( &ctx.fugue, hash3 );
+     fugue512_Update( &ctx.fugue, hash0, 512 );
+     fugue512_Final( &ctx.fugue, hash0 );
+     memcpy( &ctx.fugue, &x14_4way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash1, 512 );
+     fugue512_Final( &ctx.fugue, hash1 );
+     memcpy( &ctx.fugue, &x14_4way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash2, 512 );
+     fugue512_Final( &ctx.fugue, hash2 );
+     memcpy( &ctx.fugue, &x14_4way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash3, 512 );
+     fugue512_Final( &ctx.fugue, hash3 );

     // 14 Shabal, parallel 32 bit
     intrlv_4x32( vhash, hash0, hash1, hash2, hash3, 512 );
--- a/algo/x14/x14.c
+++ b/algo/x14/x14.c
@@ -13,7 +13,6 @@
 #include "algo/skein/sph_skein.h"
 #include "algo/shavite/sph_shavite.h"
 #include "algo/hamsi/sph_hamsi.h"
-#include "algo/fugue/sph_fugue.h"
 #include "algo/shabal/sph_shabal.h"
 #include "algo/luffa/luffa_for_sse2.h"
 #include "algo/cubehash/cubehash_sse2.h"
@@ -21,9 +20,11 @@
 #if defined(__AES__)
  #include "algo/echo/aes_ni/hash_api.h"
  #include "algo/groestl/aes_ni/hash-groestl.h"
+  #include "algo/fugue/fugue-aesni.h"
 #else
  #include "algo/groestl/sph_groestl.h"
  #include "algo/echo/sph_echo.h"
+  #include "algo/fugue/sph_fugue.h"
 #endif

 typedef struct {
@@ -32,9 +33,11 @@ typedef struct {
 #if defined(__AES__)
   hashState_groestl       groestl;
   hashState_echo          echo;
+   hashState_fugue         fugue;
 #else
   sph_groestl512_context  groestl;
   sph_echo512_context     echo;
+   sph_fugue512_context    fugue;
 #endif
   sph_jh512_context       jh;
   sph_keccak512_context   keccak;
@@ -44,7 +47,6 @@ typedef struct {
   sph_shavite512_context  shavite;
   hashState_sd            simd;
   sph_hamsi512_context    hamsi;
-   sph_fugue512_context    fugue;
   sph_shabal512_context   shabal;
 } x14_ctx_holder;

@@ -57,9 +59,11 @@ void init_x14_ctx()
 #if defined(__AES__)
   init_groestl( &x14_ctx.groestl, 64 );
   init_echo( &x14_ctx.echo, 512 );
+   fugue512_Init( &x14_ctx.fugue, 512 );
 #else
   sph_groestl512_init( &x14_ctx.groestl );
   sph_echo512_init( &x14_ctx.echo );
+   sph_fugue512_init( &x14_ctx.fugue );
 #endif
   sph_skein512_init( &x14_ctx.skein );
   sph_jh512_init( &x14_ctx.jh );
@@ -69,7 +73,6 @@ void init_x14_ctx()
   sph_shavite512_init( &x14_ctx.shavite );
   init_sd( &x14_ctx.simd,512 );
   sph_hamsi512_init( &x14_ctx.hamsi );
-   sph_fugue512_init( &x14_ctx.fugue );
   sph_shabal512_init( &x14_ctx.shabal );
 };

@@ -125,8 +128,13 @@ void x14hash(void *output, const void *input)
    sph_hamsi512(&ctx.hamsi, hash, 64);
    sph_hamsi512_close(&ctx.hamsi, hash);

+#if defined(__AES__)
+    fugue512_Update( &ctx.fugue, hash, 512 );
+    fugue512_Final( &ctx.fugue, hash );
+#else
    sph_fugue512(&ctx.fugue, hash, 64);
    sph_fugue512_close(&ctx.fugue, hash);
+#endif

    sph_shabal512( &ctx.shabal, hash, 64 );
 	 sph_shabal512_close( &ctx.shabal, hash );
--- a/algo/x15/x15-4way.c
+++ b/algo/x15/x15-4way.c
@@ -17,7 +17,7 @@
 #include "algo/echo/aes_ni/hash_api.h"
 #include "algo/echo/sph_echo.h"
 #include "algo/hamsi/hamsi-hash-4way.h"
-#include "algo/fugue/sph_fugue.h"
+#include "algo/fugue/fugue-aesni.h"
 #include "algo/shabal/shabal-hash-4way.h"
 #include "algo/whirlpool/sph_whirlpool.h"
 #if defined(__VAES__)
@@ -38,7 +38,7 @@ typedef struct {
    cube_4way_context       cube;
    simd_4way_context       simd;
    hamsi512_8way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_8way_context  shabal;
    sph_whirlpool_context   whirlpool;
 #if defined(__VAES__)
@@ -65,7 +65,7 @@ void init_x15_8way_ctx()
     cube_4way_init( &x15_8way_ctx.cube, 512, 16, 32 );
     simd_4way_init( &x15_8way_ctx.simd, 512 );
     hamsi512_8way_init( &x15_8way_ctx.hamsi );
-     sph_fugue512_init( &x15_8way_ctx.fugue );
+     fugue512_Init( &x15_8way_ctx.fugue, 512 );
     shabal512_8way_init( &x15_8way_ctx.shabal );
     sph_whirlpool_init( &x15_8way_ctx.whirlpool );
 #if defined(__VAES__)
@@ -260,30 +260,29 @@ void x15_8way_hash( void *state, const void *input )
                       vhash );

     // 13 Fugue
-     sph_fugue512( &ctx.fugue, hash0, 64 );
-     sph_fugue512_close( &ctx.fugue, hash0 );
-     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash1, 64 );
-     sph_fugue512_close( &ctx.fugue, hash1 );
-     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash2, 64 );
-     sph_fugue512_close( &ctx.fugue, hash2 );
-     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash3, 64 );
-     sph_fugue512_close( &ctx.fugue, hash3 );
-     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash4, 64 );
-     sph_fugue512_close( &ctx.fugue, hash4 );
-     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash5, 64 );
-     sph_fugue512_close( &ctx.fugue, hash5 );
-     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash6, 64 );
-     sph_fugue512_close( &ctx.fugue, hash6 );
-     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash7, 64 );
-     sph_fugue512_close( &ctx.fugue, hash7 );
-
+     fugue512_Update( &ctx.fugue, hash0, 512 );
+     fugue512_Final( &ctx.fugue, hash0 );
+     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash1, 512 );
+     fugue512_Final( &ctx.fugue, hash1 );
+     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash2, 512 );
+     fugue512_Final( &ctx.fugue, hash2 );
+     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash3, 512 );
+     fugue512_Final( &ctx.fugue, hash3 );
+     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash4, 512 );
+     fugue512_Final( &ctx.fugue, hash4 );
+     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash5, 512 );
+     fugue512_Final( &ctx.fugue, hash5 );
+     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash6, 512 );
+     fugue512_Final( &ctx.fugue, hash6 );
+     memcpy( &ctx.fugue, &x15_8way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash7, 512 );
+     fugue512_Final( &ctx.fugue, hash7 );

     // 14 Shabal, parallel 32 bit
     intrlv_8x32_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
@@ -387,7 +386,7 @@ typedef struct {
    simd_2way_context       simd;
    hashState_echo          echo;
    hamsi512_4way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_4way_context  shabal;
    sph_whirlpool_context   whirlpool;
 } x15_4way_ctx_holder;
@@ -408,7 +407,7 @@ void init_x15_4way_ctx()
     simd_2way_init( &x15_4way_ctx.simd, 512 );
     init_echo( &x15_4way_ctx.echo, 512 );
     hamsi512_4way_init( &x15_4way_ctx.hamsi );
-     sph_fugue512_init( &x15_4way_ctx.fugue );
+     fugue512_Init( &x15_4way_ctx.fugue, 512 );
     shabal512_4way_init( &x15_4way_ctx.shabal );
     sph_whirlpool_init( &x15_4way_ctx.whirlpool );
 };
@@ -524,17 +523,17 @@ void x15_4way_hash( void *state, const void *input )
     dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );

     // 13 Fugue
-     sph_fugue512( &ctx.fugue, hash0, 64 );
-     sph_fugue512_close( &ctx.fugue, hash0 );
-     memcpy( &ctx.fugue, &x15_4way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash1, 64 );
-     sph_fugue512_close( &ctx.fugue, hash1 );
-     memcpy( &ctx.fugue, &x15_4way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash2, 64 );
-     sph_fugue512_close( &ctx.fugue, hash2 );
-     memcpy( &ctx.fugue, &x15_4way_ctx.fugue, sizeof(sph_fugue512_context) );
-     sph_fugue512( &ctx.fugue, hash3, 64 );
-     sph_fugue512_close( &ctx.fugue, hash3 );
+     fugue512_Update( &ctx.fugue, hash0, 512 );
+     fugue512_Final( &ctx.fugue, hash0 );
+     memcpy( &ctx.fugue, &x15_4way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash1, 512 );
+     fugue512_Final( &ctx.fugue, hash1 );
+     memcpy( &ctx.fugue, &x15_4way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash2, 512 );
+     fugue512_Final( &ctx.fugue, hash2 );
+     memcpy( &ctx.fugue, &x15_4way_ctx.fugue, sizeof(hashState_fugue) );
+     fugue512_Update( &ctx.fugue, hash3, 512 );
+     fugue512_Final( &ctx.fugue, hash3 );

     // 14 Shabal, parallel 32 bit
     intrlv_4x32( vhash, hash0, hash1, hash2, hash3, 512 );
--- a/algo/x15/x15.c
+++ b/algo/x15/x15.c
@@ -23,9 +23,11 @@
 #if defined(__AES__)
  #include "algo/echo/aes_ni/hash_api.h"
  #include "algo/groestl/aes_ni/hash-groestl.h"
+  #include "algo/fugue/fugue-aesni.h"
 #else
  #include "algo/groestl/sph_groestl.h"
  #include "algo/echo/sph_echo.h"
+  #include "algo/fugue/sph_fugue.h"
 #endif

 typedef struct {
@@ -34,9 +36,11 @@ typedef struct {
 #if defined(__AES__)
   hashState_echo          echo;
   hashState_groestl       groestl;
+   hashState_fugue         fugue;
 #else
   sph_groestl512_context   groestl;
   sph_echo512_context      echo;
+   sph_fugue512_context    fugue;
 #endif
   sph_jh512_context       jh;
   sph_keccak512_context   keccak;
@@ -46,7 +50,6 @@ typedef struct {
   sph_shavite512_context  shavite;
   hashState_sd            simd;
   sph_hamsi512_context    hamsi;
-   sph_fugue512_context    fugue;
   sph_shabal512_context   shabal;
   sph_whirlpool_context   whirlpool;
 } x15_ctx_holder;
@@ -60,9 +63,11 @@ void init_x15_ctx()
 #if defined(__AES__)
   init_groestl( &x15_ctx.groestl, 64 );
   init_echo( &x15_ctx.echo, 512 );
+   fugue512_Init( &x15_ctx.fugue, 512 );
 #else
   sph_groestl512_init( &x15_ctx.groestl );
   sph_echo512_init( &x15_ctx.echo );
+   sph_fugue512_init( &x15_ctx.fugue );
 #endif
   sph_skein512_init( &x15_ctx.skein );
   sph_jh512_init( &x15_ctx.jh );
@@ -72,7 +77,6 @@ void init_x15_ctx()
   sph_shavite512_init( &x15_ctx.shavite );
   init_sd( &x15_ctx.simd, 512 );
   sph_hamsi512_init( &x15_ctx.hamsi );
-   sph_fugue512_init( &x15_ctx.fugue );
   sph_shabal512_init( &x15_ctx.shabal );
   sph_whirlpool_init( &x15_ctx.whirlpool );
 };
@@ -131,8 +135,13 @@ void x15hash(void *output, const void *input)
    sph_hamsi512( &ctx.hamsi, hash, 64 );
    sph_hamsi512_close( &ctx.hamsi, hash );

+#if defined(__AES__)
+    fugue512_Update( &ctx.fugue, hash, 512 );
+    fugue512_Final( &ctx.fugue, hash );
+#else
    sph_fugue512( &ctx.fugue, hash, 64 );
    sph_fugue512_close( &ctx.fugue, hash );
+#endif

    sph_shabal512( &ctx.shabal, hash, 64 );
    sph_shabal512_close( &ctx.shabal, hash );
--- a/algo/x16/hex.c
+++ b/algo/x16/hex.c
@@ -6,30 +6,6 @@
 */
 #include "x16r-gate.h"

-#include <stdio.h>
-#include <stdlib.h>
-#include <string.h>
-#include "algo/blake/sph_blake.h"
-#include "algo/bmw/sph_bmw.h"
-#include "algo/groestl/sph_groestl.h"
-#include "algo/jh/sph_jh.h"
-#include "algo/keccak/sph_keccak.h"
-#include "algo/skein/sph_skein.h"
-#include "algo/shavite/sph_shavite.h"
-#include "algo/luffa/luffa_for_sse2.h"
-#include "algo/cubehash/cubehash_sse2.h"
-#include "algo/simd/nist.h"
-#include "algo/echo/sph_echo.h"
-#include "algo/hamsi/sph_hamsi.h"
-#include "algo/fugue/sph_fugue.h"
-#include "algo/shabal/sph_shabal.h"
-#include "algo/whirlpool/sph_whirlpool.h"
-#include <openssl/sha.h>
-#if defined(__AES__)
-  #include "algo/echo/aes_ni/hash_api.h"
-  #include "algo/groestl/aes_ni/hash-groestl.h"
-#endif
-
 static void hex_getAlgoString(const uint32_t* prevblock, char *output)
 {
   char *sptr = output;
@@ -47,34 +23,6 @@ static void hex_getAlgoString(const uint32_t* prevblock, char *output)
   *sptr = '\0';
 }

-/*
-union _hex_context_overlay
-{
-#if defined(__AES__)
-        hashState_echo          echo;
-        hashState_groestl       groestl;
-#else
-        sph_groestl512_context   groestl;
-        sph_echo512_context      echo;
-#endif
-        sph_blake512_context    blake;
-        sph_bmw512_context      bmw;
-        sph_skein512_context    skein;
-        sph_jh512_context       jh;
-        sph_keccak512_context   keccak;
-        hashState_luffa         luffa;
-        cubehashParam           cube;
-        shavite512_context      shavite;
-        hashState_sd            simd;
-        sph_hamsi512_context    hamsi;
-        sph_fugue512_context    fugue;
-        sph_shabal512_context   shabal;
-        sph_whirlpool_context   whirlpool;
-        SHA512_CTX              sha512;
-};
-typedef union _hex_context_overlay hex_context_overlay;
-*/
-
 static __thread x16r_context_overlay hex_ctx;

 int hex_hash( void* output, const void* input, int thrid )
@@ -187,8 +135,12 @@ int hex_hash( void* output, const void* input, int thrid )
            sph_hamsi512_close( &ctx.hamsi, hash );
         break;
         case FUGUE:
+#if defined(__AES__)
+             fugue512_full( &ctx.fugue, hash, in, size );
+#else
             sph_fugue512_full( &ctx.fugue, hash, in, size );
-         break;
+#endif
+	     break;
         case SHABAL:
            if ( i == 0 ) 
               sph_shabal512( &ctx.shabal, in+64, 16 );
@@ -209,9 +161,9 @@ int hex_hash( void* output, const void* input, int thrid )
                sph_whirlpool512_full( &ctx.whirlpool, hash, in,  size );
         break;
         case SHA_512:
-             SHA512_Init( &ctx.sha512 );
-             SHA512_Update( &ctx.sha512, in, size );
-             SHA512_Final( (unsigned char*) hash, &ctx.sha512 );
+             sph_sha512_init( &ctx.sha512 );
+             sph_sha512( &ctx.sha512, in, size );
+             sph_sha512_close( &ctx.sha512, hash );
         break;
      }

--- a/algo/x16/minotaur.c
+++ b/algo/x16/minotaur.c
@@ -0,0 +1,302 @@
+// Minotaur hash
+
+#include "algo-gate-api.h"
+#include <stdlib.h>
+#include <stdint.h>
+#include <string.h>
+#include <stdio.h>
+#include "algo/blake/sph_blake.h"
+#include "algo/bmw/sph_bmw.h"
+#include "algo/jh/sph_jh.h"
+#include "algo/keccak/sph_keccak.h"
+#include "algo/skein/sph_skein.h"
+#include "algo/shavite/sph_shavite.h"
+#include "algo/luffa/luffa_for_sse2.h"
+#include "algo/cubehash/cubehash_sse2.h"
+#include "algo/simd/nist.h"
+#include "algo/hamsi/sph_hamsi.h"
+#include "algo/shabal/sph_shabal.h"
+#include "algo/whirlpool/sph_whirlpool.h"
+#include "algo/sha/sph_sha2.h"
+#if defined(__AES__)
+  #include "algo/echo/aes_ni/hash_api.h"
+  #include "algo/groestl/aes_ni/hash-groestl.h"
+  #include "algo/fugue/fugue-aesni.h"
+#else
+  #include "algo/echo/sph_echo.h"
+  #include "algo/groestl/sph_groestl.h"
+  #include "algo/fugue/sph_fugue.h"
+#endif
+
+// Config
+#define MINOTAUR_ALGO_COUNT	16
+
+typedef struct TortureNode TortureNode;
+typedef struct TortureGarden TortureGarden;
+
+// Graph of hash algos plus SPH contexts
+struct TortureGarden
+{
+#if defined(__AES__)
+        hashState_echo          echo;
+        hashState_groestl       groestl;
+        hashState_fugue         fugue;
+#else
+        sph_echo512_context     echo;
+        sph_groestl512_context  groestl;
+        sph_fugue512_context    fugue;
+#endif
+        sph_blake512_context    blake;
+        sph_bmw512_context      bmw;
+        sph_skein512_context    skein;
+        sph_jh512_context       jh;
+        sph_keccak512_context   keccak;
+        hashState_luffa         luffa;
+        cubehashParam           cube;
+        shavite512_context      shavite;
+        hashState_sd            simd;
+        sph_hamsi512_context    hamsi;
+        sph_shabal512_context   shabal;
+        sph_whirlpool_context   whirlpool;
+        sph_sha512_context      sha512;
+
+    struct TortureNode {
+        unsigned int algo;
+        TortureNode *child[2];
+    } nodes[22];
+} __attribute__ ((aligned (64)));
+
+// Get a 64-byte hash for given 64-byte input, using given TortureGarden contexts and given algo index
+static void get_hash( void *output, const void *input, TortureGarden *garden,
+	              unsigned int algo )
+{    
+	unsigned char hash[64] __attribute__ ((aligned (64)));
+
+    switch (algo) {
+        case 0:
+            sph_blake512_init(&garden->blake);
+            sph_blake512(&garden->blake, input, 64);
+            sph_blake512_close(&garden->blake, hash);
+            break;
+        case 1:
+            sph_bmw512_init(&garden->bmw);
+            sph_bmw512(&garden->bmw, input, 64);
+            sph_bmw512_close(&garden->bmw, hash);        
+            break;
+        case 2:
+            cubehashInit( &garden->cube, 512, 16, 32 );
+            cubehashUpdateDigest( &garden->cube, (byte*)hash,
+			         (const byte*)input, 64 );
+            break;
+        case 3:
+#if defined(__AES__)
+            echo_full( &garden->echo, (BitSequence *)hash, 512,
+                              (const BitSequence *)input, 64 );
+#else
+            sph_echo512_init(&garden->echo);
+            sph_echo512(&garden->echo, input, 64);
+            sph_echo512_close(&garden->echo, hash);          
+#endif
+	    break;
+        case 4:
+#if defined(__AES__)
+            fugue512_full( &garden->fugue, hash, input, 64 );
+#else
+            sph_fugue512_full( &garden->fugue, hash, input, 64 );
+#endif
+	    break;
+        case 5:
+#if defined(__AES__)
+            groestl512_full( &garden->groestl, (char*)hash, (char*)input, 512 );
+#else
+            sph_groestl512_init(&garden->groestl);
+            sph_groestl512(&garden->groestl, input, 64);
+            sph_groestl512_close(&garden->groestl, hash);          
+#endif
+	    break;
+        case 6:
+            sph_hamsi512_init(&garden->hamsi);
+            sph_hamsi512(&garden->hamsi, input, 64);
+            sph_hamsi512_close(&garden->hamsi, hash);          
+            break;
+        case 7:
+            sph_sha512_init( &garden->sha512 );
+            sph_sha512( &garden->sha512, input, 64 );
+            sph_sha512_close( &garden->sha512, hash );
+            break;
+        case 8:
+            sph_jh512_init(&garden->jh);
+            sph_jh512(&garden->jh, input, 64);
+            sph_jh512_close(&garden->jh, hash);          
+            break;
+        case 9:
+            sph_keccak512_init(&garden->keccak);
+            sph_keccak512(&garden->keccak, input, 64);
+            sph_keccak512_close(&garden->keccak, hash);
+            break;
+        case 10:
+            init_luffa( &garden->luffa, 512 );
+            update_and_final_luffa( &garden->luffa, (BitSequence*)hash,
+                                    (const BitSequence*)input, 64 );
+            break;
+        case 11:
+            sph_shabal512_init(&garden->shabal);
+            sph_shabal512(&garden->shabal, input, 64);
+            sph_shabal512_close(&garden->shabal, hash);          
+            break;
+        case 12:
+            sph_shavite512_init(&garden->shavite);
+            sph_shavite512(&garden->shavite, input, 64);
+            sph_shavite512_close(&garden->shavite, hash);          
+            break;
+        case 13:
+            init_sd( &garden->simd, 512 );
+            update_final_sd( &garden->simd, (BitSequence *)hash,
+                              (const BitSequence*)input, 512 );
+            break;
+        case 14:
+            sph_skein512_init(&garden->skein);
+            sph_skein512(&garden->skein, input, 64);
+            sph_skein512_close(&garden->skein, hash);          
+            break;
+        case 15:
+            sph_whirlpool_init(&garden->whirlpool);
+            sph_whirlpool(&garden->whirlpool, input, 64);
+            sph_whirlpool_close(&garden->whirlpool, hash);          
+            break;
+    }
+
+    memcpy(output, hash, 64);
+}
+
+static __thread TortureGarden garden;
+
+bool initialize_torture_garden()
+{
+    // Create torture garden nodes. Note that both sides of 19 and 20 lead to 21, and 21 has no children (to make traversal complete).
+
+   garden.nodes[ 0].child[0] = &garden.nodes[ 1];
+   garden.nodes[ 0].child[1] = &garden.nodes[ 2];
+   garden.nodes[ 1].child[0] = &garden.nodes[ 3];
+   garden.nodes[ 1].child[1] = &garden.nodes[ 4];
+   garden.nodes[ 2].child[0] = &garden.nodes[ 5];
+   garden.nodes[ 2].child[1] = &garden.nodes[ 6];
+   garden.nodes[ 3].child[0] = &garden.nodes[ 7];
+   garden.nodes[ 3].child[1] = &garden.nodes[ 8];
+   garden.nodes[ 4].child[0] = &garden.nodes[ 9];
+   garden.nodes[ 4].child[1] = &garden.nodes[10];
+   garden.nodes[ 5].child[0] = &garden.nodes[11];
+   garden.nodes[ 5].child[1] = &garden.nodes[12];
+   garden.nodes[ 6].child[0] = &garden.nodes[13];
+   garden.nodes[ 6].child[1] = &garden.nodes[14];
+   garden.nodes[ 7].child[0] = &garden.nodes[15];
+   garden.nodes[ 7].child[1] = &garden.nodes[16];
+   garden.nodes[ 8].child[0] = &garden.nodes[15];
+   garden.nodes[ 8].child[1] = &garden.nodes[16];
+   garden.nodes[ 9].child[0] = &garden.nodes[15];
+   garden.nodes[ 9].child[1] = &garden.nodes[16];
+   garden.nodes[10].child[0] = &garden.nodes[15];
+   garden.nodes[10].child[1] = &garden.nodes[16];
+   garden.nodes[11].child[0] = &garden.nodes[17];
+   garden.nodes[11].child[1] = &garden.nodes[18];
+   garden.nodes[12].child[0] = &garden.nodes[17];
+   garden.nodes[12].child[1] = &garden.nodes[18];
+   garden.nodes[13].child[0] = &garden.nodes[17];
+   garden.nodes[13].child[1] = &garden.nodes[18];
+   garden.nodes[14].child[0] = &garden.nodes[17];
+   garden.nodes[14].child[1] = &garden.nodes[18];
+   garden.nodes[15].child[0] = &garden.nodes[19];
+   garden.nodes[15].child[1] = &garden.nodes[20];
+   garden.nodes[16].child[0] = &garden.nodes[19];
+   garden.nodes[16].child[1] = &garden.nodes[20];
+   garden.nodes[17].child[0] = &garden.nodes[19];
+   garden.nodes[17].child[1] = &garden.nodes[20];
+   garden.nodes[18].child[0] = &garden.nodes[19];
+   garden.nodes[18].child[1] = &garden.nodes[20];
+   garden.nodes[19].child[0] = &garden.nodes[21];
+   garden.nodes[19].child[1] = &garden.nodes[21];
+   garden.nodes[20].child[0] = &garden.nodes[21];
+   garden.nodes[20].child[1] = &garden.nodes[21];
+   garden.nodes[21].child[0] = NULL;
+   garden.nodes[21].child[1] = NULL;
+
+   return true;
+}
+
+// Produce a 32-byte hash from 80-byte input data
+int minotaur_hash( void *output, const void *input, int thr_id )
+{    
+    unsigned char hash[64] __attribute__ ((aligned (64)));
+
+    // Find initial sha512 hash
+    sph_sha512_init( &garden.sha512 );
+    sph_sha512( &garden.sha512, input, 80 );
+    sph_sha512_close( &garden.sha512, hash );
+
+    // algo 6 (Hamsi) is very slow. It's faster to skip hashing this nonce
+    // if Hamsi is needed but only the first and last functions are
+    // currently known. Abort if either is Hamsi.
+    if ( ( ( hash[ 0] % MINOTAUR_ALGO_COUNT ) == 6 )
+      || ( ( hash[21] % MINOTAUR_ALGO_COUNT ) == 6 ) )
+         return 0;
+
+    // Assign algos to torture garden nodes based on initial hash
+    for ( int i = 0; i < 22; i++ )
+        garden.nodes[i].algo = hash[i] % MINOTAUR_ALGO_COUNT;
+
+    // Send the initial hash through the torture garden
+    TortureNode *node = &garden.nodes[0];
+
+    while ( node )
+    {
+      get_hash( hash, hash, &garden, node->algo );
+      node = node->child[ hash[63] & 1 ];
+    }
+
+    memcpy( output, hash, 32 );
+    return 1;
+}
+
+int scanhash_minotaur( struct work *work, uint32_t max_nonce,
+                      uint64_t *hashes_done, struct thr_info *mythr )
+{
+   uint32_t edata[20] __attribute__((aligned(64)));
+   uint32_t hash[8] __attribute__((aligned(64)));
+   uint32_t *pdata = work->data;
+   uint32_t *ptarget = work->target;
+   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 1;
+   uint32_t n = first_nonce;
+   const int thr_id = mythr->id;
+   const bool bench = opt_benchmark;
+   uint64_t skipped = 0;
+
+   mm128_bswap32_80( edata, pdata );
+   do
+   {
+      edata[19] = n;
+      if ( likely( algo_gate.hash( hash, edata, thr_id ) ) )
+      {
+	 if ( unlikely( valid_hash( hash, ptarget ) && !bench ) )
+         {
+            pdata[19] = bswap_32( n );
+            submit_solution( work, hash, mythr );
+         }
+      }
+      else skipped++;
+      n++;
+   } while ( n < last_nonce && !work_restart[thr_id].restart );
+   *hashes_done = n - first_nonce - skipped;
+   pdata[19] = n;
+   return 0;
+}
+
+bool register_minotaur_algo( algo_gate_t* gate )
+{
+  gate->scanhash = (void*)&scanhash_minotaur;
+  gate->hash      = (void*)&minotaur_hash;
+  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT;
+  gate->miner_thread_init = (void*)&initialize_torture_garden;
+  return true;
+};
+
--- a/algo/x16/x16r-4way.c
+++ b/algo/x16/x16r-4way.c
@@ -16,8 +16,7 @@

 #if defined (X16R_8WAY)

-// Perform midstate prehash of hash functions with block size <= 64 bytes
-// and interleave 4x64 before nonce insertion for final hash.
+// Perform midstate prehash of hash functions with block size <= 72 bytes.

 void x16r_8way_prehash( void *vdata, void *pdata )
 {
@@ -34,6 +33,11 @@ void x16r_8way_prehash( void *vdata, void *pdata )
         jh512_8way_init( &x16r_ctx.jh );
         jh512_8way_update( &x16r_ctx.jh, vdata, 64 );
      break;
+      case KECCAK:
+         mm512_bswap32_intrlv80_8x64( vdata, pdata );
+         keccak512_8way_init( &x16r_ctx.keccak );
+         keccak512_8way_update( &x16r_ctx.keccak, vdata, 72 );
+      break;
      case SKEIN:
         mm512_bswap32_intrlv80_8x64( vdata, pdata );
         skein512_8way_init( &x16r_ctx.skein );
@@ -173,13 +177,13 @@ int x16r_8way_hash_generic( void* output, const void* input, int thrid )
                          hash7, vhash );
         break;
         case KECCAK:
-            keccak512_8way_init( &ctx.keccak );
-            if ( i == 0 )
-               keccak512_8way_update( &ctx.keccak, input, size );
+           if ( i == 0 )
+               keccak512_8way_update( &ctx.keccak, input + (72<<3), 8 );
            else
            {
               intrlv_8x64( vhash, in0, in1, in2, in3, in4, in5, in6, in7, 
                            size<<3 );
+               keccak512_8way_init( &ctx.keccak );
               keccak512_8way_update( &ctx.keccak, vhash, size );
            }
            keccak512_8way_close( &ctx.keccak, vhash );
@@ -347,14 +351,14 @@ int x16r_8way_hash_generic( void* output, const void* input, int thrid )
                          hash7, vhash );
         break;
         case FUGUE:
-             sph_fugue512_full( &ctx.fugue, hash0, in0, size );
-             sph_fugue512_full( &ctx.fugue, hash1, in1, size );
-             sph_fugue512_full( &ctx.fugue, hash2, in2, size );
-             sph_fugue512_full( &ctx.fugue, hash3, in3, size );
-             sph_fugue512_full( &ctx.fugue, hash4, in4, size );
-             sph_fugue512_full( &ctx.fugue, hash5, in5, size );
-             sph_fugue512_full( &ctx.fugue, hash6, in6, size );
-             sph_fugue512_full( &ctx.fugue, hash7, in7, size );
+             fugue512_full( &ctx.fugue, hash0, in0, size );
+             fugue512_full( &ctx.fugue, hash1, in1, size );
+             fugue512_full( &ctx.fugue, hash2, in2, size );
+             fugue512_full( &ctx.fugue, hash3, in3, size );
+             fugue512_full( &ctx.fugue, hash4, in4, size );
+             fugue512_full( &ctx.fugue, hash5, in5, size );
+             fugue512_full( &ctx.fugue, hash6, in6, size );
+             fugue512_full( &ctx.fugue, hash7, in7, size );
         break;
         case SHABAL:
             intrlv_8x32( vhash, in0, in1, in2, in3, in4, in5, in6, in7,
@@ -490,6 +494,7 @@ int scanhash_x16r_8way( struct work *work, uint32_t max_nonce,
   {
      x16_r_s_getAlgoString( (const uint8_t*)bedata1, x16r_hash_order );
      s_ntime = ntime;
+
      if ( opt_debug && !thr_id )
          applog( LOG_INFO, "hash order %s (%08x)", x16r_hash_order, ntime );
   }
@@ -533,6 +538,11 @@ void x16r_4way_prehash( void *vdata, void *pdata )
         jh512_4way_init( &x16r_ctx.jh );
         jh512_4way_update( &x16r_ctx.jh, vdata, 64 );
      break;
+      case KECCAK:
+         mm256_bswap32_intrlv80_4x64( vdata, pdata );
+         keccak512_4way_init( &x16r_ctx.keccak );
+         keccak512_4way_update( &x16r_ctx.keccak, vdata, 72 );
+      break;
      case SKEIN:
         mm256_bswap32_intrlv80_4x64( vdata, pdata );
         skein512_4way_prehash64( &x16r_ctx.skein, vdata );
@@ -619,11 +629,20 @@ int x16r_4way_hash_generic( void* output, const void* input, int thrid )
            dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );
         break;
         case GROESTL:
+#if defined(__VAES__)
+            intrlv_2x128( vhash, in0, in1, size<<3 );
+            groestl512_2way_full( &ctx.groestl, vhash, vhash, size );
+            dintrlv_2x128_512( hash0, hash1, vhash );
+            intrlv_2x128( vhash, in2, in3, size<<3 );
+            groestl512_2way_full( &ctx.groestl, vhash, vhash, size );
+            dintrlv_2x128_512( hash2, hash3, vhash );
+#else
            groestl512_full( &ctx.groestl, (char*)hash0, (char*)in0, size<<3 );
            groestl512_full( &ctx.groestl, (char*)hash1, (char*)in1, size<<3 );
            groestl512_full( &ctx.groestl, (char*)hash2, (char*)in2, size<<3 );
            groestl512_full( &ctx.groestl, (char*)hash3, (char*)in3, size<<3 );
-         break;
+#endif
+   	    break;
         case JH:
            if ( i == 0 )
               jh512_4way_update( &ctx.jh, input + (64<<2), 16 );
@@ -637,12 +656,12 @@ int x16r_4way_hash_generic( void* output, const void* input, int thrid )
            dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );
         break;
         case KECCAK:
-            keccak512_4way_init( &ctx.keccak );
-            if ( i == 0 )
-               keccak512_4way_update( &ctx.keccak, input, size );
+           if ( i == 0 )
+               keccak512_4way_update( &ctx.keccak, input + (72<<2), 8 );
            else
            {
               intrlv_4x64( vhash, in0, in1, in2, in3, size<<3 );
+               keccak512_4way_init( &ctx.keccak );
               keccak512_4way_update( &ctx.keccak, vhash, size );
            }
            keccak512_4way_close( &ctx.keccak, vhash );
@@ -711,11 +730,20 @@ int x16r_4way_hash_generic( void* output, const void* input, int thrid )
            }
         break;
         case SHAVITE:
+#if defined(__VAES__)
+            intrlv_2x128( vhash, in0, in1, size<<3 );
+            shavite512_2way_full( &ctx.shavite, vhash, vhash, size );
+            dintrlv_2x128_512( hash0, hash1, vhash );
+            intrlv_2x128( vhash, in2, in3, size<<3 );
+            shavite512_2way_full( &ctx.shavite, vhash, vhash, size );
+            dintrlv_2x128_512( hash2, hash3, vhash );
+#else
            shavite512_full( &ctx.shavite, hash0, in0, size );
            shavite512_full( &ctx.shavite, hash1, in1, size );
            shavite512_full( &ctx.shavite, hash2, in2, size );
            shavite512_full( &ctx.shavite, hash3, in3, size );
-         break;
+#endif
+   	    break;
         case SIMD:
            intrlv_2x128( vhash, in0, in1, size<<3 );
            simd512_2way_full( &ctx.simd, vhash, vhash, size );
@@ -725,6 +753,14 @@ int x16r_4way_hash_generic( void* output, const void* input, int thrid )
            dintrlv_2x128_512( hash2, hash3, vhash );
         break;
         case ECHO:
+#if defined(__VAES__)
+            intrlv_2x128( vhash, in0, in1, size<<3 );
+            echo_2way_full( &ctx.echo, vhash, 512, vhash, size );
+            dintrlv_2x128_512( hash0, hash1, vhash );
+            intrlv_2x128( vhash, in2, in3, size<<3 );
+            echo_2way_full( &ctx.echo, vhash, 512, vhash, size );
+            dintrlv_2x128_512( hash2, hash3, vhash );
+#else
            echo_full( &ctx.echo, (BitSequence *)hash0, 512,
                              (const BitSequence *)in0, size );
            echo_full( &ctx.echo, (BitSequence *)hash1, 512,
@@ -733,7 +769,8 @@ int x16r_4way_hash_generic( void* output, const void* input, int thrid )
                              (const BitSequence *)in2, size );
            echo_full( &ctx.echo, (BitSequence *)hash3, 512,
                              (const BitSequence *)in3, size );
-         break;
+#endif
+   	    break;
         case HAMSI:
            if ( i == 0 )
               hamsi512_4way_update( &ctx.hamsi, input + (64<<2), 16 );
@@ -747,10 +784,10 @@ int x16r_4way_hash_generic( void* output, const void* input, int thrid )
            dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );
         break;
         case FUGUE:
-             sph_fugue512_full( &ctx.fugue, hash0, in0, size );
-             sph_fugue512_full( &ctx.fugue, hash1, in1, size );
-             sph_fugue512_full( &ctx.fugue, hash2, in2, size );
-             sph_fugue512_full( &ctx.fugue, hash3, in3, size );
+             fugue512_full( &ctx.fugue, hash0, in0, size );
+             fugue512_full( &ctx.fugue, hash1, in1, size );
+             fugue512_full( &ctx.fugue, hash2, in2, size );
+             fugue512_full( &ctx.fugue, hash3, in3, size );
         break;
         case SHABAL:
             intrlv_4x32( vhash, in0, in1, in2, in3, size<<3 );
@@ -856,7 +893,7 @@ int scanhash_x16r_4way( struct work *work, uint32_t max_nonce,
      x16_r_s_getAlgoString( (const uint8_t*)bedata1, x16r_hash_order );
      s_ntime = ntime;
      if ( opt_debug && !thr_id )
-              applog( LOG_INFO, "hash order %s (%08x)", x16r_hash_order, ntime );
+         applog( LOG_INFO, "hash order %s (%08x)", x16r_hash_order, ntime );
   }

   x16r_4way_prehash( vdata, pdata );
--- a/algo/x16/x16r-gate.c
+++ b/algo/x16/x16r-gate.c
@@ -61,7 +61,8 @@ bool register_x16r_algo( algo_gate_t* gate )
  gate->scanhash  = (void*)&scanhash_x16r;
  gate->hash      = (void*)&x16r_hash;
 #endif
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
+  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT |
+	                VAES_OPT | VAES256_OPT;
  x16_r_s_getAlgoString = (void*)&x16r_getAlgoString;
  opt_target_factor = 256.0;
  return true;
@@ -79,7 +80,8 @@ bool register_x16rv2_algo( algo_gate_t* gate )
  gate->scanhash  = (void*)&scanhash_x16rv2;
  gate->hash      = (void*)&x16rv2_hash;
 #endif
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
+  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT |
+	                VAES_OPT | VAES256_OPT;
  x16_r_s_getAlgoString = (void*)&x16r_getAlgoString;
  opt_target_factor = 256.0;
  return true;
@@ -97,7 +99,8 @@ bool register_x16s_algo( algo_gate_t* gate )
  gate->scanhash  = (void*)&scanhash_x16r;
  gate->hash      = (void*)&x16r_hash;
 #endif
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
+  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT |
+	                VAES_OPT | VAES256_OPT;
  x16_r_s_getAlgoString = (void*)&x16s_getAlgoString;
  opt_target_factor = 256.0;
  return true;
@@ -135,18 +138,16 @@ void x16rt_getAlgoString( const uint32_t *timeHash, char *output)

 void veil_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
 {
+   uint32_t merkleroothash[8];
+   uint32_t witmerkleroothash[8];
+   uint32_t denom10[8];
+   uint32_t denom100[8];
+   uint32_t denom1000[8];
+   uint32_t denom10000[8];
+   int i;
   uchar merkle_tree[64] = { 0 };
-   size_t t;

   algo_gate.gen_merkle_root( merkle_tree, sctx );
-   // Increment extranonce2
-   for ( t = 0; t < sctx->xnonce2_size && !( ++sctx->job.xnonce2[t] ); t++ );
-
-   // Assemble block header
-//   algo_gate.build_block_header( g_work, le32dec( sctx->job.version ),
-//          (uint32_t*) sctx->job.prevhash, (uint32_t*) merkle_tree,
-//          le32dec( sctx->job.ntime ), le32dec(sctx->job.nbits) );
-   int i;

   memset( g_work->data, 0, sizeof(g_work->data) );
   g_work->data[0] = le32dec( sctx->job.version );
@@ -164,35 +165,35 @@ void veil_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
   g_work->data[31] = 0x00000280;

   for ( i = 0; i < 8; i++ )
-      g_work->merkleroothash[7 - i] = be32dec((uint32_t *)merkle_tree + i);
+      merkleroothash[7 - i] = be32dec((uint32_t *)merkle_tree + i);
   for ( i = 0; i < 8; i++ )
-      g_work->witmerkleroothash[7 - i] = be32dec((uint32_t *)merkle_tree + i);
+      witmerkleroothash[7 - i] = be32dec((uint32_t *)merkle_tree + i);
   for ( i = 0; i < 8; i++ )
-      g_work->denom10[i] =    le32dec((uint32_t *)sctx->job.denom10 + i);
+      denom10[i] =    le32dec((uint32_t *)sctx->job.denom10 + i);
   for ( i = 0; i < 8; i++ )
-      g_work->denom100[i] =   le32dec((uint32_t *)sctx->job.denom100 + i);
+      denom100[i] =   le32dec((uint32_t *)sctx->job.denom100 + i);
   for ( i = 0; i < 8; i++ )
-      g_work->denom1000[i] =  le32dec((uint32_t *)sctx->job.denom1000 + i);
+      denom1000[i] =  le32dec((uint32_t *)sctx->job.denom1000 + i);
   for ( i = 0; i < 8; i++ )
-      g_work->denom10000[i] = le32dec((uint32_t *)sctx->job.denom10000 + i);
+      denom10000[i] = le32dec((uint32_t *)sctx->job.denom10000 + i);

   uint32_t pofnhash[8];
   memset(pofnhash, 0x00, 32);

-   char denom10_str      [ 2 * sizeof( g_work->denom10 )           + 1 ];
-   char denom100_str     [ 2 * sizeof( g_work->denom100 )          + 1 ];
-   char denom1000_str    [ 2 * sizeof( g_work->denom1000 )         + 1 ];
-   char denom10000_str   [ 2 * sizeof( g_work->denom10000 )        + 1 ];
-   char merkleroot_str   [ 2 * sizeof( g_work->merkleroothash )    + 1 ];
-   char witmerkleroot_str[ 2 * sizeof( g_work->witmerkleroothash ) + 1 ];
+   char denom10_str      [ 2 * sizeof( denom10 )           + 1 ];
+   char denom100_str     [ 2 * sizeof( denom100 )          + 1 ];
+   char denom1000_str    [ 2 * sizeof( denom1000 )         + 1 ];
+   char denom10000_str   [ 2 * sizeof( denom10000 )        + 1 ];
+   char merkleroot_str   [ 2 * sizeof( merkleroothash )    + 1 ];
+   char witmerkleroot_str[ 2 * sizeof( witmerkleroothash ) + 1 ];
   char pofn_str         [ 2 * sizeof( pofnhash )                  + 1 ];

-   cbin2hex( denom10_str,       (char*) g_work->denom10,           32 );
-   cbin2hex( denom100_str,      (char*) g_work->denom100,          32 );
-   cbin2hex( denom1000_str,     (char*) g_work->denom1000,         32 );
-   cbin2hex( denom10000_str,    (char*) g_work->denom10000,        32 );
-   cbin2hex( merkleroot_str,    (char*) g_work->merkleroothash,    32 );
-   cbin2hex( witmerkleroot_str, (char*) g_work->witmerkleroothash, 32 );
+   cbin2hex( denom10_str,       (char*) denom10,           32 );
+   cbin2hex( denom100_str,      (char*) denom100,          32 );
+   cbin2hex( denom1000_str,     (char*) denom1000,         32 );
+   cbin2hex( denom10000_str,    (char*) denom10000,        32 );
+   cbin2hex( merkleroot_str,    (char*) merkleroothash,    32 );
+   cbin2hex( witmerkleroot_str, (char*) witmerkleroothash, 32 );
   cbin2hex( pofn_str,          (char*) pofnhash,                  32 );

   if ( true )
@@ -232,7 +233,8 @@ bool register_x16rt_algo( algo_gate_t* gate )
  gate->scanhash  = (void*)&scanhash_x16rt;
  gate->hash      = (void*)&x16r_hash;
 #endif
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
+  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT |
+	                VAES_OPT | VAES256_OPT;
  opt_target_factor = 256.0;
  return true;
 };
@@ -249,7 +251,8 @@ bool register_x16rt_veil_algo( algo_gate_t* gate )
  gate->scanhash  = (void*)&scanhash_x16rt;
  gate->hash      = (void*)&x16r_hash;
 #endif
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
+  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT |
+	                VAES_OPT | VAES256_OPT;
  gate->build_extraheader = (void*)&veil_build_extraheader;
  opt_target_factor = 256.0;
  return true;
@@ -279,22 +282,17 @@ bool register_x21s_algo( algo_gate_t* gate )
  gate->scanhash          = (void*)&scanhash_x21s_8way;
  gate->hash              = (void*)&x21s_8way_hash;
  gate->miner_thread_init = (void*)&x21s_8way_thread_init;
-  gate->optimizations     = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT
-                            | VAES_OPT;
 #elif defined (X16R_4WAY)
  gate->scanhash          = (void*)&scanhash_x21s_4way;
  gate->hash              = (void*)&x21s_4way_hash;
  gate->miner_thread_init = (void*)&x21s_4way_thread_init;
-  gate->optimizations     = SSE2_OPT | AES_OPT | AVX2_OPT | SHA_OPT
-                            | AVX512_OPT | VAES_OPT;
 #else
  gate->scanhash          = (void*)&scanhash_x21s;
  gate->hash              = (void*)&x21s_hash;
  gate->miner_thread_init = (void*)&x21s_thread_init;
-  gate->optimizations     = SSE2_OPT | AES_OPT | AVX2_OPT | SHA_OPT
-                          | AVX512_OPT | VAES_OPT;
 #endif
-//  gate->optimizations     = SSE2_OPT | AES_OPT | AVX2_OPT | SHA_OPT | AVX512_OPT;
+  gate->optimizations     = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT |
+	                    VAES_OPT | VAES256_OPT;
  x16_r_s_getAlgoString   = (void*)&x16s_getAlgoString;
  opt_target_factor = 256.0;
  return true;
--- a/algo/x16/x16r-gate.h
+++ b/algo/x16/x16r-gate.h
@@ -20,12 +20,16 @@
 #include "algo/fugue/sph_fugue.h"
 #include "algo/shabal/sph_shabal.h"
 #include "algo/whirlpool/sph_whirlpool.h"
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"
+
 #if defined(__AES__)
-  #include "algo/echo/aes_ni/hash_api.h"
-  #include "algo/groestl/aes_ni/hash-groestl.h"
+#include "algo/echo/aes_ni/hash_api.h"
+#include "algo/groestl/aes_ni/hash-groestl.h"
+#include "algo/fugue/fugue-aesni.h"
 #endif
+
 #if defined (__AVX2__)
+
 #include "algo/blake/blake-hash-4way.h"
 #include "algo/bmw/bmw-hash-4way.h"
 #include "algo/groestl/aes_ni/hash-groestl.h"
@@ -38,11 +42,14 @@
 #include "algo/hamsi/hamsi-hash-4way.h"
 #include "algo/shabal/shabal-hash-4way.h"
 #include "algo/sha/sha-hash-4way.h"
+
 #if defined(__VAES__)
-  #include "algo/groestl/groestl512-hash-4way.h"
-  #include "algo/shavite/shavite-hash-4way.h"
-  #include "algo/echo/echo-hash-4way.h"
+#include "algo/groestl/groestl512-hash-4way.h"
+#include "algo/shavite/shavite-hash-2way.h"
+#include "algo/shavite/shavite-hash-4way.h"
+#include "algo/echo/echo-hash-4way.h"
 #endif
+
 #endif // AVX2

 #if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
@@ -111,7 +118,7 @@ union _x16r_8way_context_overlay
    cubehashParam           cube;
    simd_4way_context       simd;
    hamsi512_8way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_8way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_8way_context     sha512;
@@ -144,18 +151,24 @@ union _x16r_4way_context_overlay
 {
    blake512_4way_context   blake;
    bmw512_4way_context     bmw;
-    hashState_echo          echo;
+#if defined(__VAES__)
+    groestl512_2way_context groestl;
+    shavite512_2way_context shavite;
+    echo_2way_context       echo;
+#else
    hashState_groestl       groestl;
+    shavite512_context      shavite;
+    hashState_echo          echo;
+#endif
    skein512_4way_context   skein;
    jh512_4way_context      jh;
    keccak512_4way_context  keccak;
    luffa_2way_context      luffa;
    hashState_luffa         luffa1;
    cubehashParam           cube;
-    shavite512_context      shavite;
    simd_2way_context       simd;
    hamsi512_4way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_4way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_4way_context     sha512;
@@ -180,9 +193,11 @@ union _x16r_context_overlay
 #if defined(__AES__)
        hashState_echo          echo;
        hashState_groestl       groestl;
+        hashState_fugue         fugue;
 #else
        sph_groestl512_context   groestl;
        sph_echo512_context      echo;
+        sph_fugue512_context    fugue;
 #endif
        sph_blake512_context    blake;
        sph_bmw512_context      bmw;
@@ -194,10 +209,9 @@ union _x16r_context_overlay
        shavite512_context      shavite;
        hashState_sd            simd;
        sph_hamsi512_context    hamsi;
-        sph_fugue512_context    fugue;
        sph_shabal512_context   shabal;
        sph_whirlpool_context   whirlpool;
-        SHA512_CTX              sha512;
+        sph_sha512_context      sha512;
 } __attribute__ ((aligned (64)));

 typedef union _x16r_context_overlay x16r_context_overlay;
--- a/algo/x16/x16r.c
+++ b/algo/x16/x16r.c
@@ -151,8 +151,12 @@ int x16r_hash_generic( void* output, const void* input, int thrid )
            sph_hamsi512_close( &ctx.hamsi, hash );
         break;
         case FUGUE:
-            sph_fugue512_full( &ctx.fugue, hash, in, size );
-         break;
+#if defined(__AES__)
+         fugue512_full( &ctx.fugue, hash, in, size );
+#else
+	 sph_fugue512_full( &ctx.fugue, hash, in, size );
+#endif
+	 break;
         case SHABAL:
            if ( i == 0 )
               sph_shabal512( &ctx.shabal, in+64, 16 );
@@ -173,9 +177,9 @@ int x16r_hash_generic( void* output, const void* input, int thrid )
               sph_whirlpool512_full( &ctx.whirlpool, hash, in, size );
         break;
         case SHA_512:
-            SHA512_Init( &ctx.sha512 );
-            SHA512_Update( &ctx.sha512, in, size );
-            SHA512_Final( (unsigned char*) hash, &ctx.sha512 );
+            sph_sha512_init( &ctx.sha512 );
+            sph_sha512( &ctx.sha512, in, size );
+            sph_sha512_close( &ctx.sha512, hash );
         break;
      }

--- a/algo/x16/x16rv2-4way.c
+++ b/algo/x16/x16rv2-4way.c
@@ -8,30 +8,7 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
-#include "algo/blake/blake-hash-4way.h"
-#include "algo/bmw/bmw-hash-4way.h"
-#include "algo/groestl/aes_ni/hash-groestl.h"
-#include "algo/groestl/aes_ni/hash-groestl.h"
-#include "algo/skein/skein-hash-4way.h"
-#include "algo/jh/jh-hash-4way.h"
-#include "algo/keccak/keccak-hash-4way.h"
-#include "algo/shavite/sph_shavite.h"
-#include "algo/luffa/luffa-hash-2way.h"
-#include "algo/cubehash/cubehash_sse2.h"
-#include "algo/cubehash/cube-hash-2way.h"
-#include "algo/simd/simd-hash-2way.h"
-#include "algo/echo/aes_ni/hash_api.h"
-#include "algo/hamsi/hamsi-hash-4way.h"
-#include "algo/fugue/sph_fugue.h"
-#include "algo/shabal/shabal-hash-4way.h"
-#include "algo/whirlpool/sph_whirlpool.h"
-#include "algo/sha/sha-hash-4way.h"
 #include "algo/tiger/sph_tiger.h"
-#if defined(__VAES__)
-  #include "algo/groestl/groestl512-hash-4way.h"
-  #include "algo/shavite/shavite-hash-4way.h"
-  #include "algo/echo/echo-hash-4way.h"
-#endif

 #if defined (X16RV2_8WAY)

@@ -46,7 +23,7 @@ union _x16rv2_8way_context_overlay
    cubehashParam           cube;
    simd_4way_context       simd;
    hamsi512_8way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_8way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_8way_context     sha512;
@@ -432,14 +409,14 @@ int x16rv2_8way_hash( void* output, const void* input, int thrid )
                          hash7, vhash );
         break;
         case FUGUE:
-            sph_fugue512_full( &ctx.fugue, hash0, in0, size );
-            sph_fugue512_full( &ctx.fugue, hash1, in1, size );
-            sph_fugue512_full( &ctx.fugue, hash2, in2, size );
-            sph_fugue512_full( &ctx.fugue, hash3, in3, size );
-            sph_fugue512_full( &ctx.fugue, hash4, in4, size );
-            sph_fugue512_full( &ctx.fugue, hash5, in5, size );
-            sph_fugue512_full( &ctx.fugue, hash6, in6, size );
-            sph_fugue512_full( &ctx.fugue, hash7, in7, size );
+            fugue512_full( &ctx.fugue, hash0, in0, size );
+            fugue512_full( &ctx.fugue, hash1, in1, size );
+            fugue512_full( &ctx.fugue, hash2, in2, size );
+            fugue512_full( &ctx.fugue, hash3, in3, size );
+            fugue512_full( &ctx.fugue, hash4, in4, size );
+            fugue512_full( &ctx.fugue, hash5, in5, size );
+            fugue512_full( &ctx.fugue, hash6, in6, size );
+            fugue512_full( &ctx.fugue, hash7, in7, size );
         break;
         case SHABAL:
            intrlv_8x32( vhash, in0, in1, in2, in3, in4, in5, in6, in7,
@@ -695,17 +672,23 @@ union _x16rv2_4way_context_overlay
 {
    blake512_4way_context   blake;
    bmw512_4way_context     bmw;
-    hashState_echo          echo;
+#if defined(__VAES__)
+    groestl512_2way_context groestl;
+    shavite512_2way_context shavite;
+    echo_2way_context       echo;
+#else
    hashState_groestl       groestl;
+    shavite512_context      shavite;
+    hashState_echo          echo;
+#endif
    skein512_4way_context   skein;
    jh512_4way_context      jh;
    keccak512_4way_context  keccak;
    luffa_2way_context      luffa;
    cubehashParam           cube;
-    shavite512_context      shavite;
    simd_2way_context       simd;
    hamsi512_4way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_4way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_4way_context     sha512;
@@ -768,10 +751,19 @@ int x16rv2_4way_hash( void* output, const void* input, int thrid )
            dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );
         break;
         case GROESTL:
+#if defined(__VAES__)
+            intrlv_2x128( vhash, in0, in1, size<<3 );
+            groestl512_2way_full( &ctx.groestl, vhash, vhash, size );
+            dintrlv_2x128_512( hash0, hash1, vhash );
+            intrlv_2x128( vhash, in2, in3, size<<3 );
+            groestl512_2way_full( &ctx.groestl, vhash, vhash, size );
+            dintrlv_2x128_512( hash2, hash3, vhash );
+#else
            groestl512_full( &ctx.groestl, (char*)hash0, (char*)in0, size<<3 );
            groestl512_full( &ctx.groestl, (char*)hash1, (char*)in1, size<<3 );
            groestl512_full( &ctx.groestl, (char*)hash2, (char*)in2, size<<3 );
            groestl512_full( &ctx.groestl, (char*)hash3, (char*)in3, size<<3 );
+#endif
         break;
         case JH:
            if ( i == 0 )
@@ -910,10 +902,19 @@ int x16rv2_4way_hash( void* output, const void* input, int thrid )
            }
         break;
         case SHAVITE:
+#if defined(__VAES__)
+            intrlv_2x128( vhash, in0, in1, size<<3 );
+            shavite512_2way_full( &ctx.shavite, vhash, vhash, size );
+            dintrlv_2x128_512( hash0, hash1, vhash );
+            intrlv_2x128( vhash, in2, in3, size<<3 );
+            shavite512_2way_full( &ctx.shavite, vhash, vhash, size );
+            dintrlv_2x128_512( hash2, hash3, vhash );
+#else
            shavite512_full( &ctx.shavite, hash0, in0, size );
            shavite512_full( &ctx.shavite, hash1, in1, size );
            shavite512_full( &ctx.shavite, hash2, in2, size );
            shavite512_full( &ctx.shavite, hash3, in3, size );
+#endif
         break;
         case SIMD:
            intrlv_2x128( vhash, in0, in1, size<<3 );
@@ -924,6 +925,14 @@ int x16rv2_4way_hash( void* output, const void* input, int thrid )
            dintrlv_2x128_512( hash2, hash3, vhash );
         break;
         case ECHO:
+#if defined(__VAES__)
+            intrlv_2x128( vhash, in0, in1, size<<3 );
+            echo_2way_full( &ctx.echo, vhash, 512, vhash, size );
+            dintrlv_2x128_512( hash0, hash1, vhash );
+            intrlv_2x128( vhash, in2, in3, size<<3 );
+            echo_2way_full( &ctx.echo, vhash, 512, vhash, size );
+            dintrlv_2x128_512( hash2, hash3, vhash );
+#else
            echo_full( &ctx.echo, (BitSequence *)hash0, 512,
                              (const BitSequence *)in0, size );
            echo_full( &ctx.echo, (BitSequence *)hash1, 512,
@@ -932,6 +941,7 @@ int x16rv2_4way_hash( void* output, const void* input, int thrid )
                              (const BitSequence *)in2, size );
            echo_full( &ctx.echo, (BitSequence *)hash3, 512,
                              (const BitSequence *)in3, size );
+#endif
         break;
         case HAMSI:
            if ( i == 0 )
@@ -946,10 +956,10 @@ int x16rv2_4way_hash( void* output, const void* input, int thrid )
            dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );
         break;
         case FUGUE:
-            sph_fugue512_full( &ctx.fugue, hash0, in0, size );
-            sph_fugue512_full( &ctx.fugue, hash1, in1, size );
-            sph_fugue512_full( &ctx.fugue, hash2, in2, size );
-            sph_fugue512_full( &ctx.fugue, hash3, in3, size );
+            fugue512_full( &ctx.fugue, hash0, in0, size );
+            fugue512_full( &ctx.fugue, hash1, in1, size );
+            fugue512_full( &ctx.fugue, hash2, in2, size );
+            fugue512_full( &ctx.fugue, hash3, in3, size );
         break;
         case SHABAL:
             intrlv_4x32( vhash, in0, in1, in2, in3, size<<3 );
--- a/algo/x16/x16rv2.c
+++ b/algo/x16/x16rv2.c
@@ -8,41 +8,18 @@

 #if !defined(X16R_8WAY) && !defined(X16R_4WAY)

-#include <stdio.h>
-#include <stdlib.h>
-#include <string.h>
-#include "algo/blake/sph_blake.h"
-#include "algo/bmw/sph_bmw.h"
-#include "algo/groestl/sph_groestl.h"
-#include "algo/jh/sph_jh.h"
-#include "algo/keccak/sph_keccak.h"
-#include "algo/skein/sph_skein.h"
-#include "algo/shavite/sph_shavite.h"
-#include "algo/luffa/luffa_for_sse2.h"
-#include "algo/cubehash/cubehash_sse2.h"
-#include "algo/simd/nist.h"
-#include "algo/echo/sph_echo.h"
-#include "algo/hamsi/sph_hamsi.h"
-#include "algo/fugue/sph_fugue.h"
-#include "algo/shabal/sph_shabal.h"
-#include "algo/whirlpool/sph_whirlpool.h"
-#include <openssl/sha.h>
 #include "algo/tiger/sph_tiger.h"
-#if defined(__AES__)
-  #include "algo/echo/aes_ni/hash_api.h"
-  #include "algo/groestl/aes_ni/hash-groestl.h"
-#endif
-
-static __thread uint32_t s_ntime = UINT32_MAX;

 union _x16rv2_context_overlay
 {
 #if defined(__AES__)
        hashState_echo          echo;
        hashState_groestl       groestl;
+        hashState_fugue         fugue;
 #else
        sph_groestl512_context   groestl;
        sph_echo512_context      echo;
+        sph_fugue512_context    fugue;
 #endif
        sph_blake512_context    blake;
        sph_bmw512_context      bmw;
@@ -54,10 +31,9 @@ union _x16rv2_context_overlay
        shavite512_context      shavite;
        hashState_sd            simd;
        sph_hamsi512_context    hamsi;
-        sph_fugue512_context    fugue;
        sph_shabal512_context   shabal;
        sph_whirlpool_context   whirlpool;
-        SHA512_CTX              sha512;
+        sph_sha512_context      sha512;
        sph_tiger_context       tiger;
 };
 typedef union _x16rv2_context_overlay x16rv2_context_overlay;
@@ -160,8 +136,12 @@ int x16rv2_hash( void* output, const void* input, int thrid )
             sph_hamsi512_close( &ctx.hamsi, hash );
         break;
         case FUGUE:
+#if defined(__AES__)
+             fugue512_full( &ctx.fugue, hash, in, size );
+#else
             sph_fugue512_full( &ctx.fugue, hash, in, size );
-         break;
+#endif
+	     break;
         case SHABAL:
             sph_shabal512_init( &ctx.shabal );
             sph_shabal512( &ctx.shabal, in, size );
@@ -175,9 +155,9 @@ int x16rv2_hash( void* output, const void* input, int thrid )
             sph_tiger( &ctx.tiger, in, size );
             sph_tiger_close( &ctx.tiger, hash );
             padtiger512( hash );
-             SHA512_Init( &ctx.sha512 );
-             SHA512_Update( &ctx.sha512, hash, 64 );
-             SHA512_Final( (unsigned char*) hash, &ctx.sha512 );
+             sph_sha512_init( &ctx.sha512 );
+             sph_sha512( &ctx.sha512, hash, 64 );
+             sph_sha512_close( &ctx.sha512, hash );
         break;
      }

--- a/algo/x16/x21s-4way.c
+++ b/algo/x16/x21s-4way.c
@@ -13,7 +13,7 @@
 #include "algo/gost/sph_gost.h"
 #include "algo/lyra2/lyra2.h"
 #if defined(__SHA__)
- #include <openssl/sha.h>
+  #include "algo/sha/sph_sha2.h"
 #endif

 #if defined (X21S_8WAY)
@@ -209,7 +209,7 @@ union _x21s_4way_context_overlay
    sph_tiger_context       tiger;
    sph_gost512_context     gost;
 #if defined(__SHA__)
-    SHA256_CTX              sha256;
+    sph_sha256_context      sha256;
 #else
    sha256_4way_context     sha256;
 #endif
@@ -275,23 +275,18 @@ int x21s_4way_hash( void* output, const void* input, int thrid )

 #if defined(__SHA__)

-   SHA256_Init( &ctx.sha256 );
-   SHA256_Update( &ctx.sha256, hash0, 64 );
-   SHA256_Final( (unsigned char*)hash0, &ctx.sha256 );
-   SHA256_Init( &ctx.sha256 );
-   SHA256_Update( &ctx.sha256, hash1, 64 );
-   SHA256_Final( (unsigned char*)hash1, &ctx.sha256 );
-   SHA256_Init( &ctx.sha256 );
-   SHA256_Update( &ctx.sha256, hash2, 64 );
-   SHA256_Final( (unsigned char*)hash2, &ctx.sha256 );
-   SHA256_Init( &ctx.sha256 );
-   SHA256_Update( &ctx.sha256, hash3, 64 );
-   SHA256_Final( (unsigned char*)hash3, &ctx.sha256 );
-
-   memcpy( output,    hash0, 32 );
-   memcpy( output+32, hash1, 32 );
-   memcpy( output+64, hash2, 32 );
-   memcpy( output+96, hash3, 32 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash0, 64 );
+   sph_sha256_close( &ctx.sha256, output );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash1, 64 );
+   sph_sha256_close( &ctx.sha256, output+32 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash2, 64 );
+   sph_sha256_close( &ctx.sha256, output+64 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash3, 64 );
+   sph_sha256_close( &ctx.sha256, output+96 );

 #else

--- a/algo/x16/x21s.c
+++ b/algo/x16/x21s.c
@@ -8,7 +8,7 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"
 #include "algo/haval/sph-haval.h"
 #include "algo/tiger/sph_tiger.h"
 #include "algo/gost/sph_gost.h"
@@ -23,7 +23,7 @@ union _x21s_context_overlay
        sph_haval256_5_context  haval;
        sph_tiger_context       tiger;
        sph_gost512_context     gost;
-        SHA256_CTX              sha256;
+        sph_sha256_context      sha256;
 };
 typedef union _x21s_context_overlay x21s_context_overlay;

@@ -50,9 +50,9 @@ int x21s_hash( void* output, const void* input, int thrid )
   sph_gost512 ( &ctx.gost, (const void*) hash, 64 );
   sph_gost512_close( &ctx.gost, (void*) hash );

-   SHA256_Init( &ctx.sha256 );
-   SHA256_Update( &ctx.sha256, hash, 64 );
-   SHA256_Final( (unsigned char*)hash, &ctx.sha256 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash, 64 );
+   sph_sha256_close( &ctx.sha256, hash );

   memcpy( output, hash, 32 );

--- a/algo/x17/sonoa-4way.c
+++ b/algo/x17/sonoa-4way.c
@@ -16,7 +16,7 @@
 #include "algo/simd/simd-hash-2way.h"
 #include "algo/echo/aes_ni/hash_api.h"
 #include "algo/hamsi/hamsi-hash-4way.h"
-#include "algo/fugue/sph_fugue.h"
+#include "algo/fugue/fugue-aesni.h"
 #include "algo/shabal/shabal-hash-4way.h"
 #include "algo/whirlpool/sph_whirlpool.h"
 #include "algo/haval/haval-hash-4way.h"
@@ -40,7 +40,7 @@ union _sonoa_8way_context_overlay
    cube_4way_context       cube;
    simd_4way_context       simd;
    hamsi512_8way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_8way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_8way_context     sha512;
@@ -58,7 +58,7 @@ union _sonoa_8way_context_overlay

 typedef union _sonoa_8way_context_overlay sonoa_8way_context_overlay;

-int sonoa_8way_hash( void *state, const void *input, int thrid )
+int sonoa_8way_hash( void *state, const void *input, int thr_id )
 {
     uint64_t vhash[8*8] __attribute__ ((aligned (128)));
     uint64_t vhashA[8*8] __attribute__ ((aligned (64)));
@@ -186,7 +186,7 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )

 #endif

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 2

     bmw512_8way_full( &ctx.bmw, vhash, vhash, 64 );
@@ -302,7 +302,7 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )
     hamsi512_8way_update( &ctx.hamsi, vhash, 64 );
     hamsi512_8way_close( &ctx.hamsi, vhash );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 3

     bmw512_8way_full( &ctx.bmw, vhash, vhash, 64 );
@@ -423,16 +423,16 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )
     dintrlv_8x64_512( hash0, hash1, hash2, hash3, hash4, hash5, hash6, hash7,
                       vhash );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
-     sph_fugue512_full( &ctx.fugue, hash4, hash4, 64 );
-     sph_fugue512_full( &ctx.fugue, hash5, hash5, 64 );
-     sph_fugue512_full( &ctx.fugue, hash6, hash6, 64 );
-     sph_fugue512_full( &ctx.fugue, hash7, hash7, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash4, hash4, 64 );
+     fugue512_full( &ctx.fugue, hash5, hash5, 64 );
+     fugue512_full( &ctx.fugue, hash6, hash6, 64 );
+     fugue512_full( &ctx.fugue, hash7, hash7, 64 );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 4

     intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
@@ -554,14 +554,14 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )
     dintrlv_8x64_512( hash0, hash1, hash2, hash3, hash4, hash5, hash6, hash7,
                       vhash );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
-     sph_fugue512_full( &ctx.fugue, hash4, hash4, 64 );
-     sph_fugue512_full( &ctx.fugue, hash5, hash5, 64 );
-     sph_fugue512_full( &ctx.fugue, hash6, hash6, 64 );
-     sph_fugue512_full( &ctx.fugue, hash7, hash7, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash4, hash4, 64 );
+     fugue512_full( &ctx.fugue, hash5, hash5, 64 );
+     fugue512_full( &ctx.fugue, hash6, hash6, 64 );
+     fugue512_full( &ctx.fugue, hash7, hash7, 64 );

     intrlv_8x32_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
                      hash7 );
@@ -630,7 +630,7 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )

 #endif

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 5

     bmw512_8way_full( &ctx.bmw, vhash, vhash, 64 );
@@ -755,14 +755,14 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )
     dintrlv_8x64_512( hash0, hash1, hash2, hash3, hash4, hash5, hash6, hash7,
                       vhash );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
-     sph_fugue512_full( &ctx.fugue, hash4, hash4, 64 );
-     sph_fugue512_full( &ctx.fugue, hash5, hash5, 64 );
-     sph_fugue512_full( &ctx.fugue, hash6, hash6, 64 );
-     sph_fugue512_full( &ctx.fugue, hash7, hash7, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash4, hash4, 64 );
+     fugue512_full( &ctx.fugue, hash5, hash5, 64 );
+     fugue512_full( &ctx.fugue, hash6, hash6, 64 );
+     fugue512_full( &ctx.fugue, hash7, hash7, 64 );

     intrlv_8x32_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
                      hash7 );
@@ -783,7 +783,7 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )
     sph_whirlpool512_full( &ctx.whirlpool, hash6, hash6, 64 );
     sph_whirlpool512_full( &ctx.whirlpool, hash7, hash7, 64 );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 6

     intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
@@ -905,14 +905,14 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )
     dintrlv_8x64_512( hash0, hash1, hash2, hash3, hash4, hash5, hash6, hash7,
                       vhash );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
-     sph_fugue512_full( &ctx.fugue, hash4, hash4, 64 );
-     sph_fugue512_full( &ctx.fugue, hash5, hash5, 64 );
-     sph_fugue512_full( &ctx.fugue, hash6, hash6, 64 );
-     sph_fugue512_full( &ctx.fugue, hash7, hash7, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash4, hash4, 64 );
+     fugue512_full( &ctx.fugue, hash5, hash5, 64 );
+     fugue512_full( &ctx.fugue, hash6, hash6, 64 );
+     fugue512_full( &ctx.fugue, hash7, hash7, 64 );

     intrlv_8x32_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
                      hash7 );
@@ -952,7 +952,7 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )
     sph_whirlpool512_full( &ctx.whirlpool, hash6, hash6, 64 );
     sph_whirlpool512_full( &ctx.whirlpool, hash7, hash7, 64 );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 7

     intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
@@ -1074,14 +1074,14 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )
     dintrlv_8x64_512( hash0, hash1, hash2, hash3, hash4, hash5, hash6, hash7,
                       vhash );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
-     sph_fugue512_full( &ctx.fugue, hash4, hash4, 64 );
-     sph_fugue512_full( &ctx.fugue, hash5, hash5, 64 );
-     sph_fugue512_full( &ctx.fugue, hash6, hash6, 64 );
-     sph_fugue512_full( &ctx.fugue, hash7, hash7, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash4, hash4, 64 );
+     fugue512_full( &ctx.fugue, hash5, hash5, 64 );
+     fugue512_full( &ctx.fugue, hash6, hash6, 64 );
+     fugue512_full( &ctx.fugue, hash7, hash7, 64 );

     intrlv_8x32_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
                      hash7 );
@@ -1117,49 +1117,6 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )

     return 1;
 }
-     
-int scanhash_sonoa_8way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr )
-{
-   uint32_t hash[8*16] __attribute__ ((aligned (128)));
-   uint32_t vdata[20*8] __attribute__ ((aligned (64)));
-   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
-   uint32_t *hashd7 = &(hash[7<<3]);
-   uint32_t *pdata = work->data;
-   const uint32_t *ptarget = work->target;
-   const uint32_t first_nonce = pdata[19];
-   const uint32_t last_nonce = max_nonce - 8;
-   __m512i  *noncev = (__m512i*)vdata + 9;   // aligned
-   uint32_t n = first_nonce;
-   const int thr_id = mythr->id;
-   const uint32_t targ32 = ptarget[7];
-
-   mm512_bswap32_intrlv80_8x64( vdata, pdata );
-   *noncev = mm512_intrlv_blend_32(
-              _mm512_set_epi32( n+7, 0, n+6, 0, n+5, 0, n+4, 0,
-                                n+3, 0, n+2, 0, n+1, 0, n,   0 ), *noncev );
-
-   do
-   {
-      if ( sonoa_8way_hash( hash, vdata, thr_id ) )
-      for ( int lane = 0; lane < 8; lane++ )
-      if unlikely( ( hashd7[ lane ] <= targ32 ) )
-      {
-         extr_lane_8x32( lane_hash, hash, lane, 256 );
-         if ( likely( valid_hash( lane_hash, ptarget ) && !opt_benchmark ) )
-         {
-            pdata[19] = bswap_32( n + lane );
-            submit_solution( work, lane_hash, mythr );
-         }
-      }
-      *noncev = _mm512_add_epi32( *noncev,
-                                  m512_const1_64( 0x0000000800000000 ) );
-      n += 8;
-   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
-   pdata[19] = n;
-   *hashes_done = n - first_nonce;
-   return 0;
-}

 #elif defined(SONOA_4WAY)

@@ -1167,7 +1124,13 @@ union _sonoa_4way_context_overlay
 {
    blake512_4way_context   blake;
    bmw512_4way_context     bmw;
+#if defined(__VAES__)
+    groestl512_2way_context groestl;
+    echo512_2way_context    echo;
+#else
    hashState_groestl       groestl;
+    hashState_echo          echo;
+#endif
    skein512_4way_context   skein;
    jh512_4way_context      jh;
    keccak512_4way_context  keccak;
@@ -1175,9 +1138,8 @@ union _sonoa_4way_context_overlay
    cube_2way_context       cube;
    shavite512_2way_context shavite;
    simd_2way_context       simd;
-    hashState_echo          echo;
    hamsi512_4way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_4way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_4way_context     sha512;
@@ -1186,7 +1148,7 @@ union _sonoa_4way_context_overlay

 typedef union _sonoa_4way_context_overlay sonoa_4way_context_overlay;

-int sonoa_4way_hash( void *state, const void *input, int thrid )
+int sonoa_4way_hash( void *state, const void *input, int thr_id )
 {
     uint64_t hash0[8] __attribute__ ((aligned (64)));
     uint64_t hash1[8] __attribute__ ((aligned (64)));
@@ -1205,6 +1167,17 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     bmw512_4way_update( &ctx.bmw, vhash, 64 );
     bmw512_4way_close( &ctx.bmw, vhash );

+#if defined(__VAES__)
+
+     rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
+
+     groestl512_2way_full( &ctx.groestl, vhashA, vhashA, 64 );
+     groestl512_2way_full( &ctx.groestl, vhashB, vhashB, 64 );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

     groestl512_full( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -1214,6 +1187,8 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     
     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif
+
     skein512_4way_full( &ctx.skein, vhash, vhash, 64 );

     jh512_4way_init( &ctx.jh );
@@ -1238,6 +1213,15 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     simd512_2way_full( &ctx.simd, vhashA, vhashA, 64 );
     simd512_2way_full( &ctx.simd, vhashB, vhashB, 64 );

+#if defined(__VAES__)
+
+     echo_2way_full( &ctx.echo, vhashA, 512, vhashA, 64 );
+     echo_2way_full( &ctx.echo, vhashB, 512, vhashB, 64 );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_2x128_512( hash0, hash1, vhashA );
     dintrlv_2x128_512( hash2, hash3, vhashB );

@@ -1249,16 +1233,29 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
                     (const BitSequence *)hash2, 64 );
     echo_full( &ctx.echo, (BitSequence *)hash3, 512,
                     (const BitSequence *)hash3, 64 );
-     
-     if ( work_restart[thrid].restart ) return 0;
-// 2

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif
+     
+     if ( work_restart[thr_id].restart ) return 0;
+// 2
+
     bmw512_4way_init( &ctx.bmw );
     bmw512_4way_update( &ctx.bmw, vhash, 64 );
     bmw512_4way_close( &ctx.bmw, vhash );

+#if defined(__VAES__)
+
+     rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 ); 
+
+     groestl512_2way_full( &ctx.groestl, vhashA, vhashA, 64 );
+     groestl512_2way_full( &ctx.groestl, vhashB, vhashB, 64 );
+ 
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

     groestl512_full( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -1268,6 +1265,8 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif     
+
     skein512_4way_full( &ctx.skein, vhash, vhash, 64 );

     jh512_4way_init( &ctx.jh );
@@ -1292,6 +1291,15 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     simd512_2way_full( &ctx.simd, vhashA, vhashA, 64 );
     simd512_2way_full( &ctx.simd, vhashB, vhashB, 64 );

+#if defined(__VAES__)
+
+     echo_2way_full( &ctx.echo, vhashA, 512, vhashA, 64 );
+     echo_2way_full( &ctx.echo, vhashB, 512, vhashB, 64 );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_2x128_512( hash0, hash1, vhashA );
     dintrlv_2x128_512( hash2, hash3, vhashB );

@@ -1306,17 +1314,30 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif
+
     hamsi512_4way_init( &ctx.hamsi );
     hamsi512_4way_update( &ctx.hamsi, vhash, 64 );
     hamsi512_4way_close( &ctx.hamsi, vhash );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 3

     bmw512_4way_init( &ctx.bmw );
     bmw512_4way_update( &ctx.bmw, vhash, 64 );
     bmw512_4way_close( &ctx.bmw, vhash );

+#if defined(__VAES__)
+
+     rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 ); 
+
+     groestl512_2way_full( &ctx.groestl, vhashA, vhashA, 64 );
+     groestl512_2way_full( &ctx.groestl, vhashB, vhashB, 64 );
+ 
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

     groestl512_full( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -1326,6 +1347,8 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif     
+
     skein512_4way_full( &ctx.skein, vhash, vhash, 64 );

     jh512_4way_init( &ctx.jh );
@@ -1350,6 +1373,15 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     simd512_2way_full( &ctx.simd, vhashA, vhashA, 64 );
     simd512_2way_full( &ctx.simd, vhashB, vhashB, 64 );

+#if defined(__VAES__)
+
+     echo_2way_full( &ctx.echo, vhashA, 512, vhashA, 64 );
+     echo_2way_full( &ctx.echo, vhashB, 512, vhashB, 64 );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_2x128_512( hash0, hash1, vhashA );
     dintrlv_2x128_512( hash2, hash3, vhashB );

@@ -1364,18 +1396,20 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif
+
     hamsi512_4way_init( &ctx.hamsi );
     hamsi512_4way_update( &ctx.hamsi, vhash, 64 );
     hamsi512_4way_close( &ctx.hamsi, vhash );

     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 4
     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

@@ -1383,6 +1417,17 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     bmw512_4way_update( &ctx.bmw, vhash, 64 );
     bmw512_4way_close( &ctx.bmw, vhash );

+#if defined(__VAES__)
+
+     rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 ); 
+
+     groestl512_2way_full( &ctx.groestl, vhashA, vhashA, 64 );
+     groestl512_2way_full( &ctx.groestl, vhashB, vhashB, 64 );
+ 
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

     groestl512_full( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -1392,6 +1437,8 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif     
+
     skein512_4way_full( &ctx.skein, vhash, vhash, 64 );

     jh512_4way_init( &ctx.jh );
@@ -1416,6 +1463,15 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     simd512_2way_full( &ctx.simd, vhashA, vhashA, 64 );
     simd512_2way_full( &ctx.simd, vhashB, vhashB, 64 );

+#if defined(__VAES__)
+
+     echo_2way_full( &ctx.echo, vhashA, 512, vhashA, 64 );
+     echo_2way_full( &ctx.echo, vhashB, 512, vhashB, 64 );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_2x128_512( hash0, hash1, vhashA );
     dintrlv_2x128_512( hash2, hash3, vhashB );

@@ -1430,16 +1486,18 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif
+
     hamsi512_4way_init( &ctx.hamsi );
     hamsi512_4way_update( &ctx.hamsi, vhash, 64 );
     hamsi512_4way_close( &ctx.hamsi, vhash );

     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );

     intrlv_4x32_512( vhash, hash0, hash1, hash2, hash3 );

@@ -1453,6 +1511,15 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     hamsi512_4way_update( &ctx.hamsi, vhashB, 64 );
     hamsi512_4way_close( &ctx.hamsi, vhash );

+#if defined(__VAES__)
+
+     rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
+
+     echo_2way_full( &ctx.echo, vhashA, 512, vhashA, 64 );
+     echo_2way_full( &ctx.echo, vhashB, 512, vhashB, 64 );
+
+#else
+
     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

     echo_full( &ctx.echo, (BitSequence *)hash0, 512,
@@ -1467,12 +1534,14 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     intrlv_2x128_512( vhashA, hash0, hash1 );
     intrlv_2x128_512( vhashB, hash2, hash3 );

+#endif
+
     shavite512_2way_init( &ctx.shavite );
     shavite512_2way_update_close( &ctx.shavite, vhashA, vhashA, 64 );
     shavite512_2way_init( &ctx.shavite );
     shavite512_2way_update_close( &ctx.shavite, vhashB, vhashB, 64 );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 5
     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );

@@ -1486,6 +1555,20 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     shabal512_4way_update( &ctx.shabal, vhashB, 64 );
     shabal512_4way_close( &ctx.shabal, vhash );

+#if defined(__VAES__)
+
+//     rintrlv_4x32_2x128( vhashA, vhashB, vhash, 512 ); 
+     dintrlv_4x32_512( hash0, hash1, hash2, hash3, vhash );
+     intrlv_2x128_512( vhashA, hash0, hash1 );
+     intrlv_2x128_512( vhashB, hash2, hash3 );
+     
+     groestl512_2way_full( &ctx.groestl, vhashA, vhashA, 64 );
+     groestl512_2way_full( &ctx.groestl, vhashB, vhashB, 64 );
+ 
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_4x32_512( hash0, hash1, hash2, hash3, vhash );

     groestl512_full( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -1495,6 +1578,8 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif     
+
     skein512_4way_full( &ctx.skein, vhash, vhash, 64 );

     jh512_4way_init( &ctx.jh );
@@ -1519,6 +1604,15 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     simd512_2way_full( &ctx.simd, vhashA, vhashA, 64 );
     simd512_2way_full( &ctx.simd, vhashB, vhashB, 64 );

+#if defined(__VAES__)
+
+     echo_2way_full( &ctx.echo, vhashA, 512, vhashA, 64 );
+     echo_2way_full( &ctx.echo, vhashB, 512, vhashB, 64 );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_2x128_512( hash0, hash1, vhashA );
     dintrlv_2x128_512( hash2, hash3, vhashB );

@@ -1533,16 +1627,18 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif
+
     hamsi512_4way_init( &ctx.hamsi );
     hamsi512_4way_update( &ctx.hamsi, vhash, 64 );
     hamsi512_4way_close( &ctx.hamsi, vhash );

     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );

     intrlv_4x32_512( vhash, hash0, hash1, hash2, hash3 );

@@ -1557,7 +1653,7 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     sph_whirlpool512_full( &ctx.whirlpool, hash2, hash2, 64 );
     sph_whirlpool512_full( &ctx.whirlpool, hash3, hash3, 64 );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 6

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
@@ -1566,6 +1662,17 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     bmw512_4way_update( &ctx.bmw, vhash, 64 );
     bmw512_4way_close( &ctx.bmw, vhash );

+#if defined(__VAES__)
+
+     rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 ); 
+
+     groestl512_2way_full( &ctx.groestl, vhashA, vhashA, 64 );
+     groestl512_2way_full( &ctx.groestl, vhashB, vhashB, 64 );
+ 
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

     groestl512_full( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -1575,6 +1682,8 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif     
+
     skein512_4way_full( &ctx.skein, vhash, vhash, 64 );

     jh512_4way_init( &ctx.jh );
@@ -1599,6 +1708,15 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     simd512_2way_full( &ctx.simd, vhashA, vhashA, 64 );
     simd512_2way_full( &ctx.simd, vhashB, vhashB, 64 );

+#if defined(__VAES__)
+
+     echo_2way_full( &ctx.echo, vhashA, 512, vhashA, 64 );
+     echo_2way_full( &ctx.echo, vhashB, 512, vhashB, 64 );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_2x128_512( hash0, hash1, vhashA );
     dintrlv_2x128_512( hash2, hash3, vhashB );

@@ -1613,16 +1731,18 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif
+
     hamsi512_4way_init( &ctx.hamsi );
     hamsi512_4way_update( &ctx.hamsi, vhash, 64 );
     hamsi512_4way_close( &ctx.hamsi, vhash );

     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );

     intrlv_4x32_512( vhash, hash0, hash1, hash2, hash3 );

@@ -1650,7 +1770,7 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     sph_whirlpool512_full( &ctx.whirlpool, hash2, hash2, 64 );
     sph_whirlpool512_full( &ctx.whirlpool, hash3, hash3, 64 );

-     if ( work_restart[thrid].restart ) return 0;    
+     if ( work_restart[thr_id].restart ) return 0;    
 // 7

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
@@ -1659,6 +1779,17 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     bmw512_4way_update( &ctx.bmw, vhash, 64 );
     bmw512_4way_close( &ctx.bmw, vhash );

+#if defined(__VAES__)
+
+     rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 ); 
+
+     groestl512_2way_full( &ctx.groestl, vhashA, vhashA, 64 );
+     groestl512_2way_full( &ctx.groestl, vhashB, vhashB, 64 );
+ 
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

     groestl512_full( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -1668,6 +1799,8 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif     
+
     skein512_4way_full( &ctx.skein, vhash, vhash, 64 );

     jh512_4way_init( &ctx.jh );
@@ -1692,6 +1825,15 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     simd512_2way_full( &ctx.simd, vhashA, vhashA, 64 );
     simd512_2way_full( &ctx.simd, vhashB, vhashB, 64 );

+#if defined(__VAES__)
+
+     echo_2way_full( &ctx.echo, vhashA, 512, vhashA, 64 );
+     echo_2way_full( &ctx.echo, vhashB, 512, vhashB, 64 );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_2x128_512( hash0, hash1, vhashA );
     dintrlv_2x128_512( hash2, hash3, vhashB );

@@ -1706,16 +1848,18 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif
+
     hamsi512_4way_init( &ctx.hamsi );
     hamsi512_4way_update( &ctx.hamsi, vhash, 64 );
     hamsi512_4way_close( &ctx.hamsi, vhash );

     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );

     intrlv_4x32_512( vhash, hash0, hash1, hash2, hash3 );

@@ -1745,46 +1889,4 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     return 1;
 }

-int scanhash_sonoa_4way( struct work *work, const uint32_t max_nonce,
-	            uint64_t *hashes_done, struct thr_info *mythr )
-{
-     uint32_t hash[4*16] __attribute__ ((aligned (64)));
-     uint32_t vdata[24*4] __attribute__ ((aligned (64)));
-     uint32_t lane_hash[8] __attribute__ ((aligned (32)));
-     uint32_t *hashd7 = &( hash[7<<2] );
-     uint32_t *pdata = work->data;
-     const uint32_t *ptarget = work->target;
-     const uint32_t first_nonce = pdata[19];
-     const uint32_t last_nonce = max_nonce - 4;
-     const uint32_t targ32 = ptarget[7];
-     uint32_t n = first_nonce;
-     __m256i  *noncev = (__m256i*)vdata + 9;  
-     const int thr_id = mythr->id;
-
-     mm256_bswap32_intrlv80_4x64( vdata, pdata );
-     *noncev = mm256_intrlv_blend_32(
-                   _mm256_set_epi32( n+3, 0, n+2, 0, n+1, 0, n, 0 ), *noncev );
-
-     do
-     {
-        if ( sonoa_4way_hash( hash, vdata, thr_id ) )
-        for ( int lane = 0; lane < 4; lane++ )
-        if ( unlikely( hashd7[ lane ] <= targ32 ) )
-        {
-           extr_lane_4x32( lane_hash, hash, lane, 256 );
-           if ( likely( valid_hash( lane_hash, ptarget ) && !opt_benchmark ) )
-           {
-              pdata[19] = bswap_32( n + lane );
-              submit_solution( work, lane_hash, mythr );
-           }
-        }
-        *noncev = _mm256_add_epi32( *noncev,
-                                    m256_const1_64( 0x0000000400000000 ) );
-        n += 4;
-     } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
-     pdata[19] = n;
-     *hashes_done = n - first_nonce;
-     return 0;
-}
-
 #endif
--- a/algo/x17/sonoa-gate.c
+++ b/algo/x17/sonoa-gate.c
@@ -3,17 +3,16 @@
 bool register_sonoa_algo( algo_gate_t* gate )
 {
 #if defined (SONOA_8WAY)
-  gate->scanhash  = (void*)&scanhash_sonoa_8way;
-//  gate->hash      = (void*)&sonoa_8way_hash;
+  gate->scanhash  = (void*)&scanhash_8way_64in_32out;
+  gate->hash      = (void*)&sonoa_8way_hash;
 #elif defined (SONOA_4WAY)
-  gate->scanhash  = (void*)&scanhash_sonoa_4way;
-//  gate->hash      = (void*)&sonoa_4way_hash;
+  gate->scanhash  = (void*)&scanhash_4way_64in_32out;
+  gate->hash      = (void*)&sonoa_4way_hash;
 #else
  init_sonoa_ctx();
-  gate->scanhash  = (void*)&scanhash_sonoa;
-//  gate->hash      = (void*)&sonoa_hash;
+  gate->hash      = (void*)&sonoa_hash;
 #endif
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
+  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT | VAES256_OPT;
  return true;
 };

--- a/algo/x17/sonoa-gate.h
+++ b/algo/x17/sonoa-gate.h
@@ -14,21 +14,15 @@ bool register_sonoa_algo( algo_gate_t* gate );

 #if defined(SONOA_8WAY)

-int sonoa_8way_hash( void *state, const void *input, int thrid );
-int scanhash_sonoa_8way( struct work *work, uint32_t max_nonce,
-                         uint64_t *hashes_done, struct thr_info *mythr );
+int sonoa_8way_hash( void *state, const void *input, int thr_id );

 #elif defined(SONOA_4WAY)

-int sonoa_4way_hash( void *state, const void *input, int thrid );
-int scanhash_sonoa_4way( struct work *work, uint32_t max_nonce,
-                         uint64_t *hashes_done, struct thr_info *mythr );
+int sonoa_4way_hash( void *state, const void *input, int thr_id );

 #else

-int sonoa_hash( void *state, const void *input, int thrid );
-int scanhash_sonoa( struct work *work, uint32_t max_nonce,
-                  uint64_t *hashes_done, struct thr_info *mythr );
+int sonoa_hash( void *state, const void *input, int thr_id );
 void init_sonoa_ctx();

 #endif
--- a/algo/x17/sonoa.c
+++ b/algo/x17/sonoa.c
@@ -14,20 +14,21 @@
 #include "algo/skein/sph_skein.h"
 #include "algo/shavite/sph_shavite.h"
 #include "algo/hamsi/sph_hamsi.h"
-#include "algo/fugue/sph_fugue.h"
 #include "algo/shabal/sph_shabal.h"
 #include "algo/whirlpool/sph_whirlpool.h"
 #include "algo/haval/sph-haval.h"
 #include "algo/luffa/luffa_for_sse2.h"
 #include "algo/cubehash/cubehash_sse2.h"
 #include "algo/simd/nist.h"
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"
 #if defined(__AES__)
  #include "algo/echo/aes_ni/hash_api.h"
  #include "algo/groestl/aes_ni/hash-groestl.h"
+  #include "algo/fugue/fugue-aesni.h"
 #else
  #include "algo/groestl/sph_groestl.h"
  #include "algo/echo/sph_echo.h"
+  #include "algo/fugue/sph_fugue.h"
 #endif

 typedef struct {
@@ -36,9 +37,11 @@ typedef struct {
 #if defined(__AES__)
        hashState_echo          echo;
        hashState_groestl       groestl;
+        hashState_fugue         fugue;
 #else
        sph_groestl512_context  groestl;
        sph_echo512_context     echo;
+        sph_fugue512_context    fugue;
 #endif
        sph_jh512_context       jh;
        sph_keccak512_context   keccak;
@@ -48,10 +51,9 @@ typedef struct {
        sph_shavite512_context  shavite;
        hashState_sd            simd;
        sph_hamsi512_context    hamsi;
-        sph_fugue512_context    fugue;
        sph_shabal512_context   shabal;
        sph_whirlpool_context   whirlpool;
-        SHA512_CTX              sha512;
+        sph_sha512_context      sha512;
        sph_haval256_5_context  haval;
 } sonoa_ctx_holder;

@@ -64,9 +66,11 @@ void init_sonoa_ctx()
 #if defined(__AES__)
        init_echo( &sonoa_ctx.echo, 512 );
        init_groestl( &sonoa_ctx.groestl, 64 );
+        fugue512_Init( &sonoa_ctx.fugue, 512 );
 #else
        sph_groestl512_init(&sonoa_ctx.groestl );
        sph_echo512_init( &sonoa_ctx.echo );
+        sph_fugue512_init( &sonoa_ctx.fugue );
 #endif
        sph_skein512_init( &sonoa_ctx.skein);
        sph_jh512_init( &sonoa_ctx.jh);
@@ -76,14 +80,13 @@ void init_sonoa_ctx()
        sph_shavite512_init( &sonoa_ctx.shavite );
        init_sd( &sonoa_ctx.simd, 512 );
        sph_hamsi512_init( &sonoa_ctx.hamsi );
-        sph_fugue512_init( &sonoa_ctx.fugue );
        sph_shabal512_init( &sonoa_ctx.shabal );
        sph_whirlpool_init( &sonoa_ctx.whirlpool );
-        SHA512_Init( &sonoa_ctx.sha512 );
+        sph_sha512_init( &sonoa_ctx.sha512 );
        sph_haval256_5_init(&sonoa_ctx.haval);
 };

-int sonoa_hash( void *state, const void *input, int thrid )
+int sonoa_hash( void *state, const void *input, int thr_id )
 {
 	uint8_t hash[128] __attribute__ ((aligned (64)));
   sonoa_ctx_holder ctx __attribute__ ((aligned (64)));
@@ -132,7 +135,7 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_echo512_close(&ctx.echo, hash);
 #endif

-   if ( work_restart[thrid].restart ) return 0;
+   if ( work_restart[thr_id].restart ) return 0;
 //

   sph_bmw512_init( &ctx.bmw);
@@ -190,7 +193,7 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_hamsi512(&ctx.hamsi, hash, 64);
   sph_hamsi512_close(&ctx.hamsi, hash);
 	
-   if ( work_restart[thrid].restart ) return 0;
+   if ( work_restart[thr_id].restart ) return 0;
 //

   sph_bmw512_init( &ctx.bmw);
@@ -249,10 +252,15 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_hamsi512(&ctx.hamsi, hash, 64);
   sph_hamsi512_close(&ctx.hamsi, hash);

+#if defined(__AES__)
+   fugue512_Update( &ctx.fugue, hash, 512 );
+   fugue512_Final( &ctx.fugue, hash ); 
+#else   
   sph_fugue512(&ctx.fugue, hash, 64);
   sph_fugue512_close(&ctx.fugue, hash);
+#endif

-   if ( work_restart[thrid].restart ) return 0;
+   if ( work_restart[thr_id].restart ) return 0;
 //

   sph_bmw512_init( &ctx.bmw);
@@ -311,9 +319,11 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_hamsi512(&ctx.hamsi, hash, 64);
   sph_hamsi512_close(&ctx.hamsi, hash);

-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512(&ctx.fugue, hash, 64);
-   sph_fugue512_close(&ctx.fugue, hash);
+#if defined(__AES__)
+    fugue512_full( &ctx.fugue, hash, hash, 64 );
+#else
+    sph_fugue512_full( &ctx.fugue, hash, hash, 64 );
+#endif

   sph_shabal512(&ctx.shabal, hash, 64);
   sph_shabal512_close(&ctx.shabal, hash);
@@ -336,7 +346,7 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_shavite512(&ctx.shavite, hash, 64);
   sph_shavite512_close(&ctx.shavite, hash);

-   if ( work_restart[thrid].restart ) return 0;
+   if ( work_restart[thr_id].restart ) return 0;
 //

   sph_bmw512_init( &ctx.bmw);
@@ -399,9 +409,11 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_hamsi512(&ctx.hamsi, hash, 64);
   sph_hamsi512_close(&ctx.hamsi, hash);

-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512(&ctx.fugue, hash, 64);
-   sph_fugue512_close(&ctx.fugue, hash);
+#if defined(__AES__)
+    fugue512_full( &ctx.fugue, hash, hash, 64 );
+#else
+    sph_fugue512_full( &ctx.fugue, hash, hash, 64 );
+#endif

   sph_shabal512_init( &ctx.shabal );
   sph_shabal512(&ctx.shabal, hash, 64);
@@ -410,7 +422,7 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_whirlpool(&ctx.whirlpool, hash, 64);
   sph_whirlpool_close(&ctx.whirlpool, hash);

-   if ( work_restart[thrid].restart ) return 0;
+   if ( work_restart[thr_id].restart ) return 0;
 //
   sph_bmw512_init( &ctx.bmw);
   sph_bmw512(&ctx.bmw, hash, 64);
@@ -468,9 +480,11 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_hamsi512(&ctx.hamsi, hash, 64);
   sph_hamsi512_close(&ctx.hamsi, hash);

-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512(&ctx.fugue, hash, 64);
-   sph_fugue512_close(&ctx.fugue, hash);
+#if defined(__AES__)
+    fugue512_full( &ctx.fugue, hash, hash, 64 );
+#else
+    sph_fugue512_full( &ctx.fugue, hash, hash, 64 );
+#endif

   sph_shabal512_init( &ctx.shabal );
   sph_shabal512(&ctx.shabal, hash, 64);
@@ -480,14 +494,14 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_whirlpool(&ctx.whirlpool, hash, 64);
   sph_whirlpool_close(&ctx.whirlpool, hash);

-   SHA512_Update( &ctx.sha512, hash, 64 );
-   SHA512_Final( (unsigned char*) hash, &ctx.sha512 );
+   sph_sha512( &ctx.sha512, hash, 64 );
+   sph_sha512_close( &ctx.sha512, hash );

   sph_whirlpool_init( &ctx.whirlpool );
   sph_whirlpool(&ctx.whirlpool, hash, 64);
   sph_whirlpool_close(&ctx.whirlpool, hash);

-   if ( work_restart[thrid].restart ) return 0;
+   if ( work_restart[thr_id].restart ) return 0;
 //

   sph_bmw512_init( &ctx.bmw);
@@ -546,9 +560,11 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_hamsi512(&ctx.hamsi, hash, 64);
   sph_hamsi512_close(&ctx.hamsi, hash);

-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512(&ctx.fugue, hash, 64);
-   sph_fugue512_close(&ctx.fugue, hash);
+#if defined(__AES__)
+    fugue512_full( &ctx.fugue, hash, hash, 64 );
+#else
+    sph_fugue512_full( &ctx.fugue, hash, hash, 64 );
+#endif

   sph_shabal512_init( &ctx.shabal );
   sph_shabal512(&ctx.shabal, hash, 64);
@@ -558,9 +574,9 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_whirlpool(&ctx.whirlpool, hash, 64);
   sph_whirlpool_close(&ctx.whirlpool, hash);

-   SHA512_Init( &ctx.sha512 );
-   SHA512_Update( &ctx.sha512, hash, 64 );
-   SHA512_Final( (unsigned char*) hash, &ctx.sha512 );
+   sph_sha512_init( &ctx.sha512 );
+   sph_sha512( &ctx.sha512, hash, 64 );
+   sph_sha512_close( &ctx.sha512, hash );

   sph_haval256_5(&ctx.haval,(const void*) hash, 64);
   sph_haval256_5_close(&ctx.haval, hash);
@@ -569,34 +585,4 @@ int sonoa_hash( void *state, const void *input, int thrid )
   return 1;
 }

-int scanhash_sonoa( struct work *work, uint32_t max_nonce,
-             uint64_t *hashes_done, struct thr_info *mythr)
-{
-   uint32_t edata[20] __attribute__((aligned(64)));
-   uint32_t hash64[8] __attribute__((aligned(64)));
-   uint32_t *pdata = work->data;
-   uint32_t *ptarget = work->target;
-   uint32_t n = pdata[19];
-   const uint32_t first_nonce = pdata[19];
-   const int thr_id = mythr->id;
-   const bool bench = opt_benchmark;
-
-   mm128_bswap32_80( edata, pdata );
-
-   do
-   {
-      edata[19] = n;
-      if ( sonoa_hash( hash64, edata, thr_id ) )
-      if ( unlikely( valid_hash( hash64, ptarget ) && !bench ) )
-      {
-         pdata[19] = bswap_32( n );
-         submit_solution( work, hash64, mythr );
-      }
-      n++;
-   } while ( n < max_nonce && !work_restart[thr_id].restart );
-   *hashes_done = n - first_nonce;
-   pdata[19] = n;
-   return 0;
-}
-
 #endif
--- a/algo/x17/x17-4way.c
+++ b/algo/x17/x17-4way.c
@@ -21,7 +21,7 @@
 #include "algo/simd/simd-hash-2way.h"
 #include "algo/echo/aes_ni/hash_api.h"
 #include "algo/hamsi/hamsi-hash-4way.h"
-#include "algo/fugue/sph_fugue.h"
+#include "algo/fugue/fugue-aesni.h"
 #include "algo/shabal/shabal-hash-4way.h"
 #include "algo/whirlpool/sph_whirlpool.h"
 #include "algo/haval/haval-hash-4way.h"
@@ -49,7 +49,7 @@ union _x17_8way_context_overlay
 #endif
    simd_4way_context       simd;
    hamsi512_8way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_8way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_8way_context     sha512;
@@ -57,7 +57,7 @@ union _x17_8way_context_overlay
 } __attribute__ ((aligned (64)));
 typedef union _x17_8way_context_overlay x17_8way_context_overlay;

-void x17_8way_hash( void *state, const void *input )
+int x17_8way_hash( void *state, const void *input, int thr_id )
 {
     uint64_t vhash[8*8] __attribute__ ((aligned (128)));
     uint64_t vhashA[8*8] __attribute__ ((aligned (64)));
@@ -190,14 +190,14 @@ void x17_8way_hash( void *state, const void *input )
     dintrlv_8x64_512( hash0, hash1, hash2, hash3, hash4, hash5, hash6, hash7,
                       vhash );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
-     sph_fugue512_full( &ctx.fugue, hash4, hash4, 64 );
-     sph_fugue512_full( &ctx.fugue, hash5, hash5, 64 );
-     sph_fugue512_full( &ctx.fugue, hash6, hash6, 64 );
-     sph_fugue512_full( &ctx.fugue, hash7, hash7, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash4, hash4, 64 );
+     fugue512_full( &ctx.fugue, hash5, hash5, 64 );
+     fugue512_full( &ctx.fugue, hash6, hash6, 64 );
+     fugue512_full( &ctx.fugue, hash7, hash7, 64 );

     intrlv_8x32_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
                      hash7 );
@@ -230,50 +230,8 @@ void x17_8way_hash( void *state, const void *input )
     haval256_5_8way_init( &ctx.haval );
     haval256_5_8way_update( &ctx.haval, vhashA, 64 );
     haval256_5_8way_close( &ctx.haval, state );
-}

-int scanhash_x17_8way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr )
-{
-   uint32_t hash32[8*8] __attribute__ ((aligned (128)));
-   uint32_t vdata[20*8] __attribute__ ((aligned (64)));
-   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
-   uint32_t *hash32_d7 = &(hash32[7*8]);
-   uint32_t *pdata = work->data;
-   const uint32_t *ptarget = work->target;
-   const uint32_t first_nonce = pdata[19];
-   const uint32_t last_nonce = max_nonce - 8;
-   __m512i  *noncev = (__m512i*)vdata + 9; 
-   uint32_t n = first_nonce;
-   const int thr_id = mythr->id;
-   const uint32_t targ32_d7 = ptarget[7];
-   const bool bench = opt_benchmark;
-
-   mm512_bswap32_intrlv80_8x64( vdata, pdata );
-   *noncev = mm512_intrlv_blend_32(
-              _mm512_set_epi32( n+7, 0, n+6, 0, n+5, 0, n+4, 0,
-                                n+3, 0, n+2, 0, n+1, 0, n,   0 ), *noncev );
-   do
-   {
-      x17_8way_hash( hash32, vdata );
-
-      for ( int lane = 0; lane < 8; lane++ )
-      if ( unlikely( ( hash32_d7[ lane ] <= targ32_d7 ) && !bench ) )
-      {
-         extr_lane_8x32( lane_hash, hash32, lane, 256 );
-         if ( likely( valid_hash( lane_hash, ptarget ) ) )
-         {
-            pdata[19] = bswap_32( n + lane );
-            submit_solution( work, lane_hash, mythr );
-         }
-      }
-      *noncev = _mm512_add_epi32( *noncev,
-                                  m512_const1_64( 0x0000000800000000 ) );
-      n += 8;
-   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
-   pdata[19] = n;
-   *hashes_done = n - first_nonce;
-   return 0;
+     return 1;
 }

 #elif defined(X17_4WAY)
@@ -282,7 +240,13 @@ union _x17_4way_context_overlay
 {
    blake512_4way_context   blake;
    bmw512_4way_context     bmw;
+#if defined(__VAES__)
+    groestl512_2way_context groestl;
+    echo512_2way_context    echo;
+#else
    hashState_groestl       groestl;
+    hashState_echo          echo;
+#endif
    skein512_4way_context   skein;
    jh512_4way_context      jh;
    keccak512_4way_context  keccak;
@@ -290,9 +254,8 @@ union _x17_4way_context_overlay
    cube_2way_context       cube;
    shavite512_2way_context shavite;
    simd_2way_context       simd;
-    hashState_echo          echo;
    hamsi512_4way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_4way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_4way_context     sha512;
@@ -300,7 +263,7 @@ union _x17_4way_context_overlay
 };  
 typedef union _x17_4way_context_overlay x17_4way_context_overlay;

-void x17_4way_hash( void *state, const void *input )
+int x17_4way_hash( void *state, const void *input, int thr_id )
 {
     uint64_t vhash[8*4] __attribute__ ((aligned (64)));
     uint64_t vhashA[8*4] __attribute__ ((aligned (64)));
@@ -317,6 +280,17 @@ void x17_4way_hash( void *state, const void *input )
     bmw512_4way_update( &ctx.bmw, vhash, 64 );
     bmw512_4way_close( &ctx.bmw, vhash );

+#if defined(__VAES__)
+
+     rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
+
+     groestl512_2way_full( &ctx.groestl, vhashA, vhashA, 64 );
+     groestl512_2way_full( &ctx.groestl, vhashB, vhashB, 64 );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+     
     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

     groestl512_full( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -326,6 +300,8 @@ void x17_4way_hash( void *state, const void *input )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif
+
     skein512_4way_full( &ctx.skein, vhash, vhash, 64 );

     jh512_4way_init( &ctx.jh );
@@ -350,6 +326,15 @@ void x17_4way_hash( void *state, const void *input )
     simd512_2way_full( &ctx.simd, vhashA, vhashA, 64 );
     simd512_2way_full( &ctx.simd, vhashB, vhashB, 64 );

+#if defined(__VAES__)
+
+     echo_2way_full( &ctx.echo, vhashA, 512, vhashA, 64 );
+     echo_2way_full( &ctx.echo, vhashB, 512, vhashB, 64 );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
     dintrlv_2x128_512( hash0, hash1, vhashA );
     dintrlv_2x128_512( hash2, hash3, vhashB );

@@ -364,16 +349,18 @@ void x17_4way_hash( void *state, const void *input )

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif
+
     hamsi512_4way_init( &ctx.hamsi );
     hamsi512_4way_update( &ctx.hamsi, vhash, 64 );
     hamsi512_4way_close( &ctx.hamsi, vhash );

     dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+     fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+     fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+     fugue512_full( &ctx.fugue, hash3, hash3, 64 );

     intrlv_4x32_512( vhash, hash0, hash1, hash2, hash3 );

@@ -399,49 +386,8 @@ void x17_4way_hash( void *state, const void *input )
     haval256_5_4way_init( &ctx.haval );
     haval256_5_4way_update( &ctx.haval, vhashB, 64 );
     haval256_5_4way_close( &ctx.haval, state );
-}

-int scanhash_x17_4way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr )
-{
-   uint32_t hash32[8*4] __attribute__ ((aligned (64)));
-   uint32_t vdata[20*4] __attribute__ ((aligned (64)));
-   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
-   uint32_t *hash32_d7 = &(hash32[ 7*4 ]);
-   uint32_t *pdata = work->data;
-   const uint32_t *ptarget = work->target;
-   const uint32_t first_nonce = pdata[19];
-   const uint32_t last_nonce = max_nonce - 4;
-   __m256i  *noncev = (__m256i*)vdata + 9;
-   uint32_t n = first_nonce;
-   const int thr_id = mythr->id;
-   const uint32_t targ32_d7 = ptarget[7];
-   const bool bench = opt_benchmark;
-
-   mm256_bswap32_intrlv80_4x64( vdata, pdata );
-   *noncev = mm256_intrlv_blend_32(
-                   _mm256_set_epi32( n+3, 0, n+2, 0, n+1, 0, n, 0 ), *noncev );
-   do
-   {
-      x17_4way_hash( hash32, vdata );
-
-      for ( int lane = 0; lane < 4; lane++ )
-      if ( unlikely( hash32_d7[ lane ] <= targ32_d7 && !bench ) )
-      {  
-         extr_lane_4x32( lane_hash, hash32, lane, 256 );
-         if ( valid_hash( lane_hash, ptarget ) )
-         {
-            pdata[19] = bswap_32( n + lane );
-            submit_solution( work, lane_hash, mythr );
-         }            
-      }
-      *noncev = _mm256_add_epi32( *noncev,
-                                  m256_const1_64( 0x0000000400000000 ) );
-      n += 4;
-   } while ( likely( ( n <= last_nonce ) && !work_restart[thr_id].restart ) );
-   pdata[19] = n;
-   *hashes_done = n - first_nonce;
-   return 0;
+     return 1;
 }

 #endif
--- a/algo/x17/x17-gate.c
+++ b/algo/x17/x17-gate.c
@@ -3,16 +3,15 @@
 bool register_x17_algo( algo_gate_t* gate )
 {
 #if defined (X17_8WAY)
-  gate->scanhash  = (void*)&scanhash_x17_8way;
+  gate->scanhash  = (void*)&scanhash_8way_64in_32out;
  gate->hash      = (void*)&x17_8way_hash;
 #elif defined (X17_4WAY)
-  gate->scanhash  = (void*)&scanhash_x17_4way;
+  gate->scanhash  = (void*)&scanhash_4way_64in_32out;
  gate->hash      = (void*)&x17_4way_hash;
 #else
-  gate->scanhash  = (void*)&scanhash_x17;
  gate->hash      = (void*)&x17_hash;
 #endif
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
+  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT | VAES256_OPT;
  return true;
 };

--- a/algo/x17/x17-gate.h
+++ b/algo/x17/x17-gate.h
@@ -14,20 +14,15 @@ bool register_x17_algo( algo_gate_t* gate );

 #if defined(X17_8WAY)

-void x17_8way_hash( void *state, const void *input );
-int scanhash_x17_8way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr );
+int x17_8way_hash( void *state, const void *input, int thr_id );
+
 #elif defined(X17_4WAY)

-void x17_4way_hash( void *state, const void *input );
-int scanhash_x17_4way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr );
+int x17_4way_hash( void *state, const void *input, int thr_id );

 #endif

-void x17_hash( void *state, const void *input );
-int scanhash_x17( struct work *work, uint32_t max_nonce,
-                  uint64_t *hashes_done, struct thr_info *mythr );
+int x17_hash( void *state, const void *input, int thr_id );

 #endif

--- a/algo/x17/x17.c
+++ b/algo/x17/x17.c
@@ -13,20 +13,21 @@
 #include "algo/skein/sph_skein.h"
 #include "algo/shavite/sph_shavite.h"
 #include "algo/hamsi/sph_hamsi.h"
-#include "algo/fugue/sph_fugue.h"
 #include "algo/shabal/sph_shabal.h"
 #include "algo/whirlpool/sph_whirlpool.h"
 #include "algo/haval/sph-haval.h"
 #include "algo/luffa/luffa_for_sse2.h" 
 #include "algo/cubehash/cubehash_sse2.h"
 #include "algo/simd/nist.h"
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"
 #if defined(__AES__)
+  #include "algo/fugue/fugue-aesni.h"
  #include "algo/echo/aes_ni/hash_api.h"
  #include "algo/groestl/aes_ni/hash-groestl.h"
 #else
  #include "algo/groestl/sph_groestl.h"
  #include "algo/echo/sph_echo.h"
+  #include "algo/fugue/sph_fugue.h"
 #endif

 union _x17_context_overlay
@@ -36,9 +37,11 @@ union _x17_context_overlay
 #if defined(__AES__)
        hashState_groestl       groestl;
        hashState_echo          echo;
+        hashState_fugue         fugue;
 #else
        sph_groestl512_context  groestl;
        sph_echo512_context     echo;
+        sph_fugue512_context    fugue;
 #endif
        sph_jh512_context       jh;
        sph_keccak512_context   keccak;
@@ -48,15 +51,14 @@ union _x17_context_overlay
        sph_shavite512_context  shavite;
        hashState_sd            simd;
        sph_hamsi512_context    hamsi;
-        sph_fugue512_context    fugue;
        sph_shabal512_context   shabal;
        sph_whirlpool_context   whirlpool;
-        SHA512_CTX              sha512;
+        sph_sha512_context      sha512;
        sph_haval256_5_context  haval;
 };
 typedef union _x17_context_overlay x17_context_overlay;

-void x17_hash(void *output, const void *input)
+int x17_hash(void *output, const void *input, int thr_id )
 {
 //    unsigned char hash[64 * 4] __attribute__((aligned(64))) = {0};
    unsigned char hash[64] __attribute__((aligned(64)));
@@ -122,9 +124,11 @@ void x17_hash(void *output, const void *input)
    sph_hamsi512_close( &ctx.hamsi, hash );

    // 13 Fugue
-    sph_fugue512_init( &ctx.fugue );
-    sph_fugue512(&ctx.fugue, hash, 64 );
-    sph_fugue512_close(&ctx.fugue, hash );
+#if defined(__AES__)
+    fugue512_full( &ctx.fugue, hash, hash, 64 );
+#else
+    sph_fugue512_full( &ctx.fugue, hash, hash, 64 );
+#endif

    // X14 Shabal
    sph_shabal512_init( &ctx.shabal );
@@ -136,43 +140,15 @@ void x17_hash(void *output, const void *input)
    sph_whirlpool( &ctx.whirlpool, hash, 64 );
    sph_whirlpool_close( &ctx.whirlpool, hash );

-    SHA512_Init( &ctx.sha512 );
-    SHA512_Update( &ctx.sha512, hash, 64 );
-    SHA512_Final( (unsigned char*)hash, &ctx.sha512 );
+    sph_sha512_init( &ctx.sha512 );
+    sph_sha512( &ctx.sha512, hash, 64 );
+    sph_sha512_close( &ctx.sha512, hash );

    sph_haval256_5_init(&ctx.haval);
    sph_haval256_5( &ctx.haval, (const void*)hash, 64 );
    sph_haval256_5_close( &ctx.haval, output );
-}

-int scanhash_x17( struct work *work, uint32_t max_nonce,
-	          uint64_t *hashes_done, struct thr_info *mythr)
-{
-   uint32_t edata[20] __attribute__((aligned(64)));
-   uint32_t hash64[8] __attribute__((aligned(64)));
-   uint32_t *pdata = work->data;
-   uint32_t *ptarget = work->target;
-   uint32_t n = pdata[19] - 1;
-   const uint32_t first_nonce = pdata[19];
-   const int thr_id = mythr->id;
-   const bool bench = opt_benchmark;
-
-   mm128_bswap32_80( edata, pdata );
-   
-   do
-   {
-      edata[19] = n;
-      x17_hash( hash64, edata );
-      if ( unlikely( valid_hash( hash64, ptarget ) && !bench ) )
-      {
-         pdata[19] = bswap_32( n );
-         submit_solution( work, hash64, mythr );
-      }
-      n++;
-   } while ( n < max_nonce && !work_restart[thr_id].restart );
-   *hashes_done = n - first_nonce;
-   pdata[19] = n;
-   return 0;
+    return 1;
 }

 #endif
--- a/algo/x17/xevan-4way.c
+++ b/algo/x17/xevan-4way.c
@@ -16,7 +16,7 @@
 #include "algo/simd/simd-hash-2way.h"
 #include "algo/echo/aes_ni/hash_api.h"
 #include "algo/hamsi/hamsi-hash-4way.h"
-#include "algo/fugue/sph_fugue.h"
+#include "algo/fugue/fugue-aesni.h"
 #include "algo/shabal/shabal-hash-4way.h"
 #include "algo/whirlpool/sph_whirlpool.h"
 #include "algo/sha/sha-hash-4way.h"
@@ -40,7 +40,7 @@ union _xevan_8way_context_overlay
   cube_4way_context       cube;
   simd_4way_context       simd;
   hamsi512_8way_context   hamsi;
-   sph_fugue512_context    fugue;
+   hashState_fugue         fugue;
   shabal512_8way_context  shabal;
   sph_whirlpool_context   whirlpool;
   sha512_8way_context     sha512;
@@ -57,7 +57,7 @@ union _xevan_8way_context_overlay
 } __attribute__ ((aligned (64)));
 typedef union _xevan_8way_context_overlay xevan_8way_context_overlay;

-void xevan_8way_hash( void *output, const void *input )
+int xevan_8way_hash( void *output, const void *input, int thr_id )
 {
     uint64_t vhash[16<<3] __attribute__ ((aligned (128)));
     uint64_t vhashA[16<<3] __attribute__ ((aligned (64)));
@@ -192,14 +192,14 @@ void xevan_8way_hash( void *output, const void *input )
     dintrlv_8x64( hash0, hash1, hash2, hash3, hash4, hash5, hash6, hash7,
                   vhash, dataLen<<3 );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash4, hash4, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash5, hash5, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash6, hash6, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash7, hash7, dataLen );
+     fugue512_full( &ctx.fugue, hash0, hash0, dataLen );
+     fugue512_full( &ctx.fugue, hash1, hash1, dataLen );
+     fugue512_full( &ctx.fugue, hash2, hash2, dataLen );
+     fugue512_full( &ctx.fugue, hash3, hash3, dataLen );
+     fugue512_full( &ctx.fugue, hash4, hash4, dataLen );
+     fugue512_full( &ctx.fugue, hash5, hash5, dataLen );
+     fugue512_full( &ctx.fugue, hash6, hash6, dataLen );
+     fugue512_full( &ctx.fugue, hash7, hash7, dataLen );

     intrlv_8x32( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
                  hash7, dataLen<<3 );
@@ -355,14 +355,14 @@ void xevan_8way_hash( void *output, const void *input )
     dintrlv_8x64( hash0, hash1, hash2, hash3, hash4, hash5, hash6, hash7,
                   vhash, dataLen<<3 );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash4, hash4, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash5, hash5, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash6, hash6, dataLen );
-     sph_fugue512_full( &ctx.fugue, hash7, hash7, dataLen );
+     fugue512_full( &ctx.fugue, hash0, hash0, dataLen );
+     fugue512_full( &ctx.fugue, hash1, hash1, dataLen );
+     fugue512_full( &ctx.fugue, hash2, hash2, dataLen );
+     fugue512_full( &ctx.fugue, hash3, hash3, dataLen );
+     fugue512_full( &ctx.fugue, hash4, hash4, dataLen );
+     fugue512_full( &ctx.fugue, hash5, hash5, dataLen );
+     fugue512_full( &ctx.fugue, hash6, hash6, dataLen );
+     fugue512_full( &ctx.fugue, hash7, hash7, dataLen );

     intrlv_8x32( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
                  hash7, dataLen<<3 );
@@ -395,50 +395,8 @@ void xevan_8way_hash( void *output, const void *input )
     haval256_5_8way_init( &ctx.haval );
     haval256_5_8way_update( &ctx.haval, vhashA, dataLen );
     haval256_5_8way_close( &ctx.haval, output );
-}

-int scanhash_xevan_8way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr )
-{
-   uint32_t hash[8*8] __attribute__ ((aligned (128)));
-   uint32_t vdata[20*8] __attribute__ ((aligned (64)));
-   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
-   uint32_t *hashd7 = &(hash[7*8]);
-   uint32_t *pdata = work->data;
-   const uint32_t *ptarget = work->target;
-   const uint32_t first_nonce = pdata[19];
-   const uint32_t last_nonce = max_nonce - 8;
-   __m512i  *noncev = (__m512i*)vdata + 9;
-   uint32_t n = first_nonce;
-   const int thr_id = mythr->id;
-   const uint32_t targ32 = ptarget[7];
-   const bool bench = opt_benchmark;
-
-   mm512_bswap32_intrlv80_8x64( vdata, pdata );
-   *noncev = mm512_intrlv_blend_32(
-              _mm512_set_epi32( n+7, 0, n+6, 0, n+5, 0, n+4, 0,
-                                n+3, 0, n+2, 0, n+1, 0, n,   0 ), *noncev );
-   do
-   {
-      xevan_8way_hash( hash, vdata );
-
-      for ( int lane = 0; lane < 8; lane++ )
-      if ( unlikely( ( hashd7[ lane ] <= targ32 ) && !bench ) )
-      {
-         extr_lane_8x32( lane_hash, hash, lane, 256 );
-         if ( likely( valid_hash( lane_hash, ptarget ) ) )
-         {
-            pdata[19] = bswap_32( n + lane );
-            submit_solution( work, lane_hash, mythr );
-         }
-      }
-      *noncev = _mm512_add_epi32( *noncev,
-                                  m512_const1_64( 0x0000000800000000 ) );
-      n += 8;
-   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
-   pdata[19] = n;
-   *hashes_done = n - first_nonce;
-   return 0;
+     return 1;
 }

 #elif defined(XEVAN_4WAY)
@@ -447,17 +405,22 @@ union _xevan_4way_context_overlay
 {
 	blake512_4way_context   blake;
        bmw512_4way_context     bmw;
-        hashState_groestl       groestl;
-        skein512_4way_context   skein;
+#if defined(__VAES__)
+        groestl512_2way_context groestl;
+        echo_2way_context       echo;
+#else
+	hashState_groestl       groestl;
+        hashState_echo          echo;
+#endif
+	skein512_4way_context   skein;
        jh512_4way_context      jh;
        keccak512_4way_context  keccak;
        luffa_2way_context      luffa;
        cube_2way_context       cube;
        shavite512_2way_context shavite;
        simd_2way_context       simd;
-        hashState_echo          echo;
        hamsi512_4way_context   hamsi;
-        sph_fugue512_context    fugue;
+        hashState_fugue         fugue;
        shabal512_4way_context  shabal;
        sph_whirlpool_context   whirlpool;
        sha512_4way_context     sha512;
@@ -465,7 +428,7 @@ union _xevan_4way_context_overlay
 };
 typedef union _xevan_4way_context_overlay xevan_4way_context_overlay;

-void xevan_4way_hash( void *output, const void *input )
+int xevan_4way_hash( void *output, const void *input, int thr_id )
 {
     uint64_t hash0[16] __attribute__ ((aligned (64)));
     uint64_t hash1[16] __attribute__ ((aligned (64)));
@@ -484,7 +447,17 @@ void xevan_4way_hash( void *output, const void *input )
     bmw512_4way_update( &ctx.bmw, vhash, dataLen );
     bmw512_4way_close( &ctx.bmw, vhash );

-     // Serial
+#if defined(__VAES__)
+
+     rintrlv_4x64_2x128( vhashA, vhashB, vhash, dataLen<<3 );
+
+     groestl512_2way_full( &ctx.groestl, vhashA, vhashA, dataLen );
+     groestl512_2way_full( &ctx.groestl, vhashB, vhashB, dataLen );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, dataLen<<3 );
+
+#else
+     
     dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );

     groestl512_full( &ctx.groestl, (char*)hash0, (char*)hash0, dataLen<<3 );
@@ -492,9 +465,10 @@ void xevan_4way_hash( void *output, const void *input )
     groestl512_full( &ctx.groestl, (char*)hash2, (char*)hash2, dataLen<<3 );
     groestl512_full( &ctx.groestl, (char*)hash3, (char*)hash3, dataLen<<3 );

-     // Parallel 4way
     intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );

+#endif
+
     skein512_4way_full( &ctx.skein, vhash, vhash, dataLen );

     jh512_4way_init( &ctx.jh );
@@ -519,6 +493,15 @@ void xevan_4way_hash( void *output, const void *input )
     simd512_2way_full( &ctx.simd, vhashA, vhashA, dataLen );
     simd512_2way_full( &ctx.simd, vhashB, vhashB, dataLen );

+#if defined(__VAES__)
+
+     echo_2way_full( &ctx.echo, vhashA, 512, vhashA, dataLen );
+     echo_2way_full( &ctx.echo, vhashB, 512, vhashB, dataLen );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, dataLen<<3 );
+
+#else
+     
     dintrlv_2x128( hash0, hash1, vhashA, dataLen<<3 );
     dintrlv_2x128( hash2, hash3, vhashB, dataLen<<3 );

@@ -531,19 +514,20 @@ void xevan_4way_hash( void *output, const void *input )
     echo_full( &ctx.echo, (BitSequence *)hash3, 512,
                     (const BitSequence *)hash3, dataLen );

-     // Parallel
     intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );

+#endif
+
     hamsi512_4way_init( &ctx.hamsi );
     hamsi512_4way_update( &ctx.hamsi, vhash, dataLen );
     hamsi512_4way_close( &ctx.hamsi, vhash );

     dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, dataLen );
+     fugue512_full( &ctx.fugue, hash1, hash1, dataLen );
+     fugue512_full( &ctx.fugue, hash2, hash2, dataLen );
+     fugue512_full( &ctx.fugue, hash3, hash3, dataLen );

     // Parallel 4way 32 bit
     intrlv_4x32( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
@@ -584,6 +568,17 @@ void xevan_4way_hash( void *output, const void *input )
     bmw512_4way_update( &ctx.bmw, vhash, dataLen );
     bmw512_4way_close( &ctx.bmw, vhash );

+#if defined(__VAES__)
+
+     rintrlv_4x64_2x128( vhashA, vhashB, vhash, dataLen<<3 );
+
+     groestl512_2way_full( &ctx.groestl, vhashA, vhashA, dataLen );
+     groestl512_2way_full( &ctx.groestl, vhashB, vhashB, dataLen );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, dataLen<<3 );
+
+#else
+
     dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );

     groestl512_full( &ctx.groestl, (char*)hash0, (char*)hash0, dataLen<<3 );
@@ -593,6 +588,8 @@ void xevan_4way_hash( void *output, const void *input )

     intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );

+#endif
+
     skein512_4way_full( &ctx.skein, vhash, vhash, dataLen );

     jh512_4way_init( &ctx.jh );
@@ -617,6 +614,15 @@ void xevan_4way_hash( void *output, const void *input )
     simd512_2way_full( &ctx.simd, vhashA, vhashA, dataLen );
     simd512_2way_full( &ctx.simd, vhashB, vhashB, dataLen );

+#if defined(__VAES__)
+
+     echo_2way_full( &ctx.echo, vhashA, 512, vhashA, dataLen );
+     echo_2way_full( &ctx.echo, vhashB, 512, vhashB, dataLen );
+
+     rintrlv_2x128_4x64( vhash, vhashA, vhashB, dataLen<<3 );
+
+#else
+
     dintrlv_2x128( hash0, hash1, vhashA, dataLen<<3 );
     dintrlv_2x128( hash2, hash3, vhashB, dataLen<<3 );

@@ -631,16 +637,18 @@ void xevan_4way_hash( void *output, const void *input )

     intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );

+#endif
+
     hamsi512_4way_init( &ctx.hamsi );
     hamsi512_4way_update( &ctx.hamsi, vhash, dataLen );
     hamsi512_4way_close( &ctx.hamsi, vhash );

     dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );

-     sph_fugue512_full( &ctx.fugue, hash0, hash0, 64 );
-     sph_fugue512_full( &ctx.fugue, hash1, hash1, 64 );
-     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
-     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+     fugue512_full( &ctx.fugue, hash0, hash0, dataLen );
+     fugue512_full( &ctx.fugue, hash1, hash1, dataLen );
+     fugue512_full( &ctx.fugue, hash2, hash2, dataLen );
+     fugue512_full( &ctx.fugue, hash3, hash3, dataLen );

     intrlv_4x32( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );

@@ -666,49 +674,8 @@ void xevan_4way_hash( void *output, const void *input )
     haval256_5_4way_init( &ctx.haval );
     haval256_5_4way_update( &ctx.haval, vhashA, dataLen );
     haval256_5_4way_close( &ctx.haval, output );
-}

-int scanhash_xevan_4way( struct work *work, uint32_t max_nonce,
-                         uint64_t *hashes_done, struct thr_info *mythr )
-{
-   uint32_t hash[16*4] __attribute__ ((aligned (128)));
-   uint32_t vdata[20*4] __attribute__ ((aligned (64)));
-   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
-   uint32_t *hashd7 = &(hash[7<<2]);
-   uint32_t *pdata = work->data;
-   uint32_t *ptarget = work->target;
-   int thr_id = mythr->id;
-   __m256i  *noncev = (__m256i*)vdata + 9; 
-   const uint32_t targ32 = ptarget[7];
-   const uint32_t first_nonce = pdata[19];
-   const uint32_t last_nonce = max_nonce - 4;
-   uint32_t n = first_nonce;
-   const bool bench = opt_benchmark;
-
-   if ( bench )  ptarget[7] = 0x0cff;
-
-   mm256_bswap32_intrlv80_4x64( vdata, pdata );
-   *noncev = mm256_intrlv_blend_32(
-                   _mm256_set_epi32( n+3, 0, n+2, 0, n+1, 0, n, 0 ), *noncev );
-   do {
-      xevan_4way_hash( hash, vdata );
-      for ( int lane = 0; lane < 4; lane++ )
-      if ( unlikely( hashd7[ lane ] <= targ32 ) && ! bench )
-      {
-         extr_lane_4x32( lane_hash, hash, lane, 256 );
-	      if ( valid_hash( lane_hash, ptarget ) )
-         {
-             pdata[19] = bswap_32( n + lane );
-             submit_solution( work, lane_hash, mythr );
-         }
-      }
-      *noncev = _mm256_add_epi32( *noncev,
-                                  m256_const1_64( 0x0000000400000000 ) );
-      n += 4;
-   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
-   pdata[19] = n;
-   *hashes_done = n - first_nonce;
-   return 0;
+     return 1;
 }

 #endif
--- a/algo/x17/xevan-gate.c
+++ b/algo/x17/xevan-gate.c
@@ -3,17 +3,16 @@
 bool register_xevan_algo( algo_gate_t* gate )
 {
 #if defined (XEVAN_8WAY)
-  gate->scanhash  = (void*)&scanhash_xevan_8way;
+  gate->scanhash  = (void*)&scanhash_8way_64in_32out;
  gate->hash      = (void*)&xevan_8way_hash;
 #elif defined (XEVAN_4WAY)
-  gate->scanhash  = (void*)&scanhash_xevan_4way;
+  gate->scanhash  = (void*)&scanhash_4way_64in_32out;
  gate->hash      = (void*)&xevan_4way_hash;
 #else
  init_xevan_ctx();
-  gate->scanhash  = (void*)&scanhash_xevan;
  gate->hash      = (void*)&xevan_hash;
 #endif
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
+  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT | VAES256_OPT;
  opt_target_factor = 256.0;
  return true;
 };
--- a/algo/x17/xevan-gate.h
+++ b/algo/x17/xevan-gate.h
@@ -14,26 +14,15 @@ bool register_xevan_algo( algo_gate_t* gate );

 #if defined(XEVAN_8WAY)

-void xevan_8way_hash( void *state, const void *input );
+int xevan_8way_hash( void *state, const void *input, int thr_id );

-int scanhash_xevan_8way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr );
 #elif defined(XEVAN_4WAY)

-void xevan_4way_hash( void *state, const void *input );
-
-int scanhash_xevan_4way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr );
-
-//void init_xevan_4way_ctx();
+int xevan_4way_hash( void *state, const void *input, int thr_id );

 #else

-void xevan_hash( void *state, const void *input );
-
-int scanhash_xevan( struct work *work, uint32_t max_nonce,
-                  uint64_t *hashes_done, struct thr_info *mythr );
-
+int xevan_hash( void *state, const void *input, int trh_id );
 void init_xevan_ctx();

 #endif
--- a/algo/x17/xevan.c
+++ b/algo/x17/xevan.c
@@ -15,19 +15,20 @@
 #include "algo/shavite/sph_shavite.h"
 #include "algo/luffa/luffa_for_sse2.h"
 #include "algo/hamsi/sph_hamsi.h"
-#include "algo/fugue/sph_fugue.h"
 #include "algo/shabal/sph_shabal.h"
 #include "algo/whirlpool/sph_whirlpool.h"
 #include "algo/haval/sph-haval.h"
 #include "algo/simd/nist.h"
 #include "algo/cubehash/cubehash_sse2.h"
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"
 #if defined(__AES__)
  #include "algo/groestl/aes_ni/hash-groestl.h"
  #include "algo/echo/aes_ni/hash_api.h"
+  #include "algo/fugue/fugue-aesni.h"
 #else
  #include "algo/groestl/sph_groestl.h"
  #include "algo/echo/sph_echo.h"
+  #include "algo/fugue/sph_fugue.h"
 #endif

 typedef struct {
@@ -41,17 +42,18 @@ typedef struct {
        sph_shavite512_context  shavite;
        hashState_sd            simd;
        sph_hamsi512_context    hamsi;
-        sph_fugue512_context    fugue;
        sph_shabal512_context   shabal;
        sph_whirlpool_context   whirlpool;
-        SHA512_CTX              sha512;
+        sph_sha512_context      sha512;
        sph_haval256_5_context  haval;
 #if defined(__AES__)
        hashState_echo          echo;
        hashState_groestl       groestl;
+        hashState_fugue         fugue;
 #else
 	sph_groestl512_context  groestl;
        sph_echo512_context     echo;
+        sph_fugue512_context    fugue;
 #endif
 } xevan_ctx_holder;

@@ -69,21 +71,22 @@ void init_xevan_ctx()
        sph_shavite512_init( &xevan_ctx.shavite );
        init_sd( &xevan_ctx.simd, 512 );
        sph_hamsi512_init( &xevan_ctx.hamsi );
-        sph_fugue512_init( &xevan_ctx.fugue );
        sph_shabal512_init( &xevan_ctx.shabal );
        sph_whirlpool_init( &xevan_ctx.whirlpool );
-        SHA512_Init( &xevan_ctx.sha512 );
+        sph_sha512_init( &xevan_ctx.sha512 );
        sph_haval256_5_init(&xevan_ctx.haval);
 #if defined(__AES__)
        init_groestl( &xevan_ctx.groestl, 64 );
        init_echo( &xevan_ctx.echo, 512 );
+        fugue512_Init( &xevan_ctx.fugue, 512 );
 #else
 	sph_groestl512_init( &xevan_ctx.groestl );
        sph_echo512_init( &xevan_ctx.echo );
+        sph_fugue512_init( &xevan_ctx.fugue );
 #endif
 };

-void xevan_hash(void *output, const void *input)
+int xevan_hash(void *output, const void *input, int thr_id )
 {
   uint32_t _ALIGN(64) hash[32]; // 128 bytes required
 	const int dataLen = 128;
@@ -92,92 +95,27 @@ void xevan_hash(void *output, const void *input)

   sph_blake512( &ctx.blake, input, 80 );
   sph_blake512_close( &ctx.blake, hash );
-	memset(&hash[16], 0, 64);
+   memset(&hash[16], 0, 64);

-	sph_bmw512(&ctx.bmw, hash, dataLen);
-	sph_bmw512_close(&ctx.bmw, hash);
+   sph_bmw512(&ctx.bmw, hash, dataLen);
+   sph_bmw512_close(&ctx.bmw, hash);

 #if defined(__AES__)
   update_and_final_groestl( &ctx.groestl, (char*)hash,
                                     (const char*)hash, dataLen*8 );
 #else
-	sph_groestl512(&ctx.groestl, hash, dataLen);
-	sph_groestl512_close(&ctx.groestl, hash);
-#endif
-
-	sph_skein512(&ctx.skein, hash, dataLen);
-	sph_skein512_close(&ctx.skein, hash);
-
-	sph_jh512(&ctx.jh, hash, dataLen);
-	sph_jh512_close(&ctx.jh, hash);
-
-	sph_keccak512(&ctx.keccak, hash, dataLen);
-	sph_keccak512_close(&ctx.keccak, hash);
-
-   update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
-                                 (const BitSequence*)hash, dataLen );
-
-   cubehashUpdateDigest( &ctx.cubehash, (byte*)hash,
-                                 (const byte*) hash, dataLen );
-
-	sph_shavite512(&ctx.shavite, hash, dataLen);
-	sph_shavite512_close(&ctx.shavite, hash);
-
-   update_final_sd( &ctx.simd, (BitSequence *)hash,
-                         (const BitSequence *)hash, dataLen*8 );
-
-#if defined(__AES__)
-   update_final_echo( &ctx.echo, (BitSequence *) hash,
-                           (const BitSequence *) hash, dataLen*8 );
-#else
-	sph_echo512(&ctx.echo, hash, dataLen);
-	sph_echo512_close(&ctx.echo, hash);
-#endif
-
-	sph_hamsi512(&ctx.hamsi, hash, dataLen);
-	sph_hamsi512_close(&ctx.hamsi, hash);
-
-	sph_fugue512(&ctx.fugue, hash, dataLen);
-	sph_fugue512_close(&ctx.fugue, hash);
-
-	sph_shabal512(&ctx.shabal, hash, dataLen);
-	sph_shabal512_close(&ctx.shabal, hash);
-
-	sph_whirlpool(&ctx.whirlpool, hash, dataLen);
-	sph_whirlpool_close(&ctx.whirlpool, hash);
-
-   SHA512_Update( &ctx.sha512, hash, dataLen );
-   SHA512_Final( (unsigned char*) hash, &ctx.sha512 );
-
-	sph_haval256_5(&ctx.haval,(const void*) hash, dataLen);
-	sph_haval256_5_close(&ctx.haval, hash);
-
-	memset(&hash[8], 0, dataLen - 32);
-
-   memcpy( &ctx, &xevan_ctx, sizeof(xevan_ctx) );
-
-	sph_blake512(&ctx.blake, hash, dataLen);
-	sph_blake512_close(&ctx.blake, hash);
-
-	sph_bmw512(&ctx.bmw, hash, dataLen);
-	sph_bmw512_close(&ctx.bmw, hash);
-
-#if defined(__AES__)
-   update_and_final_groestl( &ctx.groestl, (char*)hash,
-                              (const BitSequence*)hash, dataLen*8 );
-#else
-	sph_groestl512(&ctx.groestl, hash, dataLen);
+   sph_groestl512(&ctx.groestl, hash, dataLen);
   sph_groestl512_close(&ctx.groestl, hash);
 #endif

-	sph_skein512(&ctx.skein, hash, dataLen);
-	sph_skein512_close(&ctx.skein, hash);
+   sph_skein512(&ctx.skein, hash, dataLen);
+   sph_skein512_close(&ctx.skein, hash);

-	sph_jh512(&ctx.jh, hash, dataLen);
-	sph_jh512_close(&ctx.jh, hash);
+   sph_jh512(&ctx.jh, hash, dataLen);
+   sph_jh512_close(&ctx.jh, hash);

-	sph_keccak512(&ctx.keccak, hash, dataLen);
-	sph_keccak512_close(&ctx.keccak, hash);
+   sph_keccak512(&ctx.keccak, hash, dataLen);
+   sph_keccak512_close(&ctx.keccak, hash);

   update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
                                 (const BitSequence*)hash, dataLen );
@@ -185,8 +123,8 @@ void xevan_hash(void *output, const void *input)
   cubehashUpdateDigest( &ctx.cubehash, (byte*)hash,
                                 (const byte*) hash, dataLen );

-	sph_shavite512(&ctx.shavite, hash, dataLen);
-	sph_shavite512_close(&ctx.shavite, hash);
+   sph_shavite512(&ctx.shavite, hash, dataLen);
+   sph_shavite512_close(&ctx.shavite, hash);

   update_final_sd( &ctx.simd, (BitSequence *)hash,
                         (const BitSequence *)hash, dataLen*8 );
@@ -199,55 +137,102 @@ void xevan_hash(void *output, const void *input)
   sph_echo512_close(&ctx.echo, hash);
 #endif

-	sph_hamsi512(&ctx.hamsi, hash, dataLen);
-	sph_hamsi512_close(&ctx.hamsi, hash);
+   sph_hamsi512(&ctx.hamsi, hash, dataLen);
+   sph_hamsi512_close(&ctx.hamsi, hash);

-	sph_fugue512(&ctx.fugue, hash, dataLen);
-	sph_fugue512_close(&ctx.fugue, hash);
+#if defined(__AES__)
+   fugue512_Update( &ctx.fugue, hash, dataLen*8 );
+   fugue512_Final( &ctx.fugue, hash ); 
+#else
+   sph_fugue512(&ctx.fugue, hash, dataLen);
+   sph_fugue512_close(&ctx.fugue, hash);
+#endif

-	sph_shabal512(&ctx.shabal, hash, dataLen);
-	sph_shabal512_close(&ctx.shabal, hash);
+   sph_shabal512(&ctx.shabal, hash, dataLen);
+   sph_shabal512_close(&ctx.shabal, hash);

-	sph_whirlpool(&ctx.whirlpool, hash, dataLen);
-	sph_whirlpool_close(&ctx.whirlpool, hash);
+   sph_whirlpool(&ctx.whirlpool, hash, dataLen);
+   sph_whirlpool_close(&ctx.whirlpool, hash);

-   SHA512_Update( &ctx.sha512, hash, dataLen );
-   SHA512_Final( (unsigned char*) hash, &ctx.sha512 );
+   sph_sha512( &ctx.sha512, hash, dataLen );
+   sph_sha512_close( &ctx.sha512, hash );

-	sph_haval256_5(&ctx.haval,(const void*) hash, dataLen);
-	sph_haval256_5_close(&ctx.haval, hash);
+   sph_haval256_5(&ctx.haval,(const void*) hash, dataLen);
+   sph_haval256_5_close(&ctx.haval, hash);

-	memcpy(output, hash, 32);
-}
+   memset(&hash[8], 0, dataLen - 32);

-int scanhash_xevan( struct work *work, uint32_t max_nonce,
-             uint64_t *hashes_done, struct thr_info *mythr)
-{
-   uint32_t edata[20] __attribute__((aligned(64)));
-   uint32_t hash64[8] __attribute__((aligned(64)));
-   uint32_t *pdata = work->data;
-   uint32_t *ptarget = work->target;
-   uint32_t n = pdata[19];
-   const uint32_t first_nonce = pdata[19];
-   const int thr_id = mythr->id;
-   const bool bench = opt_benchmark;
+   memcpy( &ctx, &xevan_ctx, sizeof(xevan_ctx) );

-   mm128_bswap32_80( edata, pdata );
+   sph_blake512(&ctx.blake, hash, dataLen);
+   sph_blake512_close(&ctx.blake, hash);

-   do
-   {
-      edata[19] = n;
-      xevan_hash( hash64, edata );
-      if ( unlikely( valid_hash( hash64, ptarget ) && !bench ) )
-      {
-         pdata[19] = bswap_32( n );
-         submit_solution( work, hash64, mythr );
-      }
-      n++;
-   } while ( n < max_nonce && !work_restart[thr_id].restart );
-   pdata[19] = n;
-   *hashes_done = n - first_nonce;
-   return 0;
+   sph_bmw512(&ctx.bmw, hash, dataLen);
+   sph_bmw512_close(&ctx.bmw, hash);
+
+#if defined(__AES__)
+   update_and_final_groestl( &ctx.groestl, (char*)hash,
+                              (const BitSequence*)hash, dataLen*8 );
+#else
+   sph_groestl512(&ctx.groestl, hash, dataLen);
+   sph_groestl512_close(&ctx.groestl, hash);
+#endif
+
+   sph_skein512(&ctx.skein, hash, dataLen);
+   sph_skein512_close(&ctx.skein, hash);
+
+   sph_jh512(&ctx.jh, hash, dataLen);
+   sph_jh512_close(&ctx.jh, hash);
+
+   sph_keccak512(&ctx.keccak, hash, dataLen);
+   sph_keccak512_close(&ctx.keccak, hash);
+
+   update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
+                                 (const BitSequence*)hash, dataLen );
+
+   cubehashUpdateDigest( &ctx.cubehash, (byte*)hash,
+                                 (const byte*) hash, dataLen );
+
+   sph_shavite512(&ctx.shavite, hash, dataLen);
+   sph_shavite512_close(&ctx.shavite, hash);
+
+   update_final_sd( &ctx.simd, (BitSequence *)hash,
+                         (const BitSequence *)hash, dataLen*8 );
+
+#if defined(__AES__)
+   update_final_echo( &ctx.echo, (BitSequence *) hash,
+                           (const BitSequence *) hash, dataLen*8 );
+#else
+   sph_echo512(&ctx.echo, hash, dataLen);
+   sph_echo512_close(&ctx.echo, hash);
+#endif
+
+   sph_hamsi512(&ctx.hamsi, hash, dataLen);
+   sph_hamsi512_close(&ctx.hamsi, hash);
+
+#if defined(__AES__)
+   fugue512_Update( &ctx.fugue, hash, dataLen*8 );
+   fugue512_Final( &ctx.fugue, hash );   
+#else
+   sph_fugue512(&ctx.fugue, hash, dataLen);
+   sph_fugue512_close(&ctx.fugue, hash);
+#endif
+
+   sph_shabal512(&ctx.shabal, hash, dataLen);
+   sph_shabal512_close(&ctx.shabal, hash);
+
+   sph_whirlpool(&ctx.whirlpool, hash, dataLen);
+   sph_whirlpool_close(&ctx.whirlpool, hash);
+
+   sph_sha512( &ctx.sha512, hash, dataLen );
+   sph_sha512_close( &ctx.sha512, hash );
+
+   sph_haval256_5(&ctx.haval,(const void*) hash, dataLen);
+   sph_haval256_5_close(&ctx.haval, hash);
+
+   memcpy(output, hash, 32);
+
+   return 1;
 }

 #endif
--- a/algo/x20/x20r.c
+++ b/algo/x20/x20r.c
@@ -18,7 +18,7 @@
 #include "algo/radiogatun/sph_radiogatun.h"
 #include "algo/panama/sph_panama.h"
 #include "algo/gost/sph_gost.h"
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"
 #if defined(__AES__)
  #include "algo/echo/aes_ni/hash_api.h"
  #include "algo/groestl/aes_ni/hash-groestl.h"
@@ -56,7 +56,7 @@ union _x20r_context_overlay
    sph_fugue512_context     fugue;
    sph_shabal512_context    shabal;
    sph_whirlpool_context    whirlpool;
-    SHA512_CTX               sha512;
+    sph_sha512_context       sha512;
    sph_haval256_5_context   haval;
    sph_gost512_context      gost;
    sph_radiogatun64_context radiogatun;
@@ -68,28 +68,6 @@ void x20r_hash(void* output, const void* input)
 {
   uint32_t _ALIGN(128) hash[64/4];
   x20r_context_overlay ctx;
-/*
-	sph_blake512_context     ctx_blake;
-	sph_bmw512_context       ctx_bmw;
-	sph_groestl512_context   ctx_groestl;
-	sph_skein512_context     ctx_skein;
-	sph_jh512_context        ctx_jh;
-	sph_keccak512_context    ctx_keccak;
-	sph_luffa512_context     ctx_luffa;
-	sph_cubehash512_context  ctx_cubehash;
-	sph_shavite512_context   ctx_shavite;
-	sph_simd512_context      ctx_simd;
-	sph_echo512_context      ctx_echo;
-	sph_hamsi512_context     ctx_hamsi;
-	sph_fugue512_context     ctx_fugue;
-	sph_shabal512_context    ctx_shabal;
-	sph_whirlpool_context    ctx_whirlpool;
-	sph_sha512_context       ctx_sha512;
-	sph_haval256_5_context   ctx_haval;
-	sph_gost512_context      ctx_gost;
-	sph_radiogatun64_context ctx_radiogatun;
-	sph_panama_context       ctx_panama;
-*/
   void *in = (void*) input;
   int size = 80;

@@ -194,9 +172,9 @@ void x20r_hash(void* output, const void* input)
 		sph_whirlpool_close(&ctx.whirlpool, hash);
 		break;
 	   case SHA_512:
-                SHA512_Init( &ctx.sha512 );
-                SHA512_Update( &ctx.sha512, in, size );
-                SHA512_Final( (unsigned char*) hash, &ctx.sha512 );
+                sph_sha512_Init( &ctx.sha512 );
+                sph_sha512( &ctx.sha512, in, size );
+                sph_sha512_close( &ctx.sha512, hash );
 		break;
 	   case HAVAL:
 		sph_haval256_5_init(&ctx.haval);
--- a/algo/x22/x22i-4way.c
+++ b/algo/x22/x22i-4way.c
@@ -11,9 +11,9 @@
 #include "algo/shavite/shavite-hash-2way.h"
 #include "algo/shavite/sph_shavite.h"
 #include "algo/simd/simd-hash-2way.h"
-#include "algo/shavite/sph_shavite.h"
+#include "algo/shavite/shavite-hash-2way.h"
 #include "algo/hamsi/hamsi-hash-4way.h"
-#include "algo/fugue/sph_fugue.h"
+#include "algo/fugue/fugue-aesni.h"
 #include "algo/shabal/shabal-hash-4way.h"
 #include "algo/whirlpool/sph_whirlpool.h"
 #include "algo/sha/sha-hash-4way.h"
@@ -27,7 +27,9 @@
  #include "algo/shavite/shavite-hash-4way.h"
  #include "algo/echo/echo-hash-4way.h"
 #endif
-
+#if defined(__SHA__)
+  #include "algo/sha/sph_sha2.h"
+#endif

 #if defined(X22I_8WAY)

@@ -42,14 +44,18 @@ union _x22i_8way_ctx_overlay
    cube_4way_context       cube;
    simd_4way_context       simd;
    hamsi512_8way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_8way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_8way_context     sha512;
    haval256_5_8way_context haval;
    sph_tiger_context       tiger;
    sph_gost512_context     gost;
+#if defined(X22I_8WAY_SHA)
+    sph_sha256_context      sha256;
+#else
    sha256_8way_context     sha256;
+#endif
 #if defined(__VAES__)
    groestl512_4way_context groestl;
    shavite512_4way_context shavite;
@@ -225,30 +231,14 @@ int x22i_8way_hash( void *output, const void *input, int thrid )
   dintrlv_8x64_512( hash0, hash1, hash2, hash3,
                     hash4, hash5, hash6, hash7, vhash );
   
-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512( &ctx.fugue, hash0, 64 );
-   sph_fugue512_close( &ctx.fugue, hash0 );
-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512( &ctx.fugue, hash1, 64 );
-   sph_fugue512_close( &ctx.fugue, hash1 );
-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512( &ctx.fugue, hash2, 64 );
-   sph_fugue512_close( &ctx.fugue, hash2 );
-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512( &ctx.fugue, hash3, 64 );
-   sph_fugue512_close( &ctx.fugue, hash3 );
-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512( &ctx.fugue, hash4, 64 );
-   sph_fugue512_close( &ctx.fugue, hash4 );
-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512( &ctx.fugue, hash5, 64 );
-   sph_fugue512_close( &ctx.fugue, hash5 );
-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512( &ctx.fugue, hash6, 64 );
-   sph_fugue512_close( &ctx.fugue, hash6 );
-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512( &ctx.fugue, hash7, 64 );
-   sph_fugue512_close( &ctx.fugue, hash7 );
+   fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+   fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+   fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+   fugue512_full( &ctx.fugue, hash3, hash3, 64 );
+   fugue512_full( &ctx.fugue, hash4, hash4, 64 );
+   fugue512_full( &ctx.fugue, hash5, hash5, 64 );
+   fugue512_full( &ctx.fugue, hash6, hash6, 64 );
+   fugue512_full( &ctx.fugue, hash7, hash7, 64 );

   intrlv_8x32_512( vhash, hash0, hash1, hash2, hash3,
                           hash4, hash5, hash6, hash7 );
@@ -399,6 +389,35 @@ int x22i_8way_hash( void *output, const void *input, int thrid )
   sph_gost512 ( &ctx.gost, (const void*) hash7, 64 );
   sph_gost512_close( &ctx.gost, (void*) hash7 );

+#if defined(X22I_8WAY_SHA)
+
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash0, 64 );
+   sph_sha256_close( &ctx.sha256, output );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash1, 64 );
+   sph_sha256_close( &ctx.sha256, output+32 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash2, 64 );
+   sph_sha256_close( &ctx.sha256, output+64 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash3, 64 );
+   sph_sha256_close( &ctx.sha256, output+96 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash4, 64 );
+   sph_sha256_close( &ctx.sha256, output+128 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash5, 64 );
+   sph_sha256_close( &ctx.sha256, output+160 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash6, 64 );
+   sph_sha256_close( &ctx.sha256, output+192 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash7, 64 );
+   sph_sha256_close( &ctx.sha256, output+224 );
+   
+#else
+
   intrlv_8x32_512( vhash, hash0, hash1, hash2, hash3,
                           hash4, hash5, hash6, hash7 );

@@ -406,9 +425,55 @@ int x22i_8way_hash( void *output, const void *input, int thrid )
   sha256_8way_update( &ctx.sha256, vhash, 64 );
   sha256_8way_close( &ctx.sha256, output );

+#endif
+
   return 1;
 }

+#if defined(X22I_8WAY_SHA)
+
+int scanhash_x22i_8way_sha( struct work *work, uint32_t max_nonce,
+                       uint64_t *hashes_done, struct thr_info *mythr )
+{
+   uint32_t hash[8*8] __attribute__ ((aligned (128)));
+   uint32_t vdata[20*8] __attribute__ ((aligned (64)));
+   uint32_t *pdata = work->data;
+   uint32_t *ptarget = work->target;
+   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 8;
+   __m512i  *noncev = (__m512i*)vdata + 9;
+   uint32_t n = first_nonce;
+   const int thr_id = mythr->id;
+   const bool bench = opt_benchmark;
+
+   if ( bench )  ptarget[7] = 0x08ff;
+
+   InitializeSWIFFTX();
+
+   mm512_bswap32_intrlv80_8x64( vdata, pdata );
+   *noncev = mm512_intrlv_blend_32(
+              _mm512_set_epi32( n+7, 0, n+6, 0, n+5, 0, n+4, 0,
+                                n+3, 0, n+2, 0, n+1, 0, n,   0 ), *noncev );
+   do
+   {
+      if ( x22i_8way_hash( hash, vdata, thr_id ) )
+      for ( int i = 0; i < 8; i++ )
+      if ( unlikely( valid_hash( hash + (i<<3), ptarget ) && !bench ) )
+      {
+         pdata[19] = bswap_32( n+i );
+         submit_solution( work, hash+(i<<3), mythr );
+      }
+      *noncev = _mm512_add_epi32( *noncev,
+                                  m512_const1_64( 0x0000000800000000 ) );
+      n += 8;
+   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
+   pdata[19] = n;
+   *hashes_done = n - first_nonce;
+   return 0;
+}
+
+#else
+
 int scanhash_x22i_8way( struct work *work, uint32_t max_nonce,
                       uint64_t *hashes_done, struct thr_info *mythr )
 {
@@ -456,53 +521,7 @@ int scanhash_x22i_8way( struct work *work, uint32_t max_nonce,
   return 0;
 }

-/*
-int scanhash_x22i_8way( struct work* work, uint32_t max_nonce,
-                   uint64_t *hashes_done, struct thr_info *mythr )
-{
-   uint32_t hash[8*16] __attribute__ ((aligned (128)));
-   uint32_t vdata[24*8] __attribute__ ((aligned (64)));
-   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
-   uint32_t *hash7 = &(hash[7<<3]);
-   uint32_t *pdata = work->data;
-   uint32_t *ptarget = work->target;
-   const uint32_t first_nonce = pdata[19];
-   __m512i  *noncev = (__m512i*)vdata + 9;   // aligned
-   uint32_t n = first_nonce;
-   const uint32_t last_nonce = max_nonce - 8;
-   const int thr_id = mythr->id;
-   const uint32_t Htarg = ptarget[7];
-
-   if (opt_benchmark)
-      ((uint32_t*)ptarget)[7] = 0x08ff;
-
-   InitializeSWIFFTX();
-
-   mm512_bswap32_intrlv80_8x64( vdata, pdata );
-   do
-   {
-      *noncev = mm512_intrlv_blend_32( mm512_bswap_32(
-               _mm512_set_epi32( n+7, 0, n+6, 0, n+5, 0, n+4, 0,
-                                 n+3, 0, n+2, 0, n+1, 0, n, 0 ) ), *noncev );
-      x22i_8way_hash( hash, vdata );
-
-      for ( int lane = 0; lane < 8; lane++ )
-      if unlikely( ( hash7[ lane ] <= Htarg ) )
-      {
-         extr_lane_8x32( lane_hash, hash, lane, 256 );
-         if ( likely( fulltest( lane_hash, ptarget ) && !opt_benchmark ) )
-         {
-            pdata[19] = n + lane;
-            submit_solution( work, lane_hash, mythr );
-         }
-      }
-      n += 8;
-   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
-
-   *hashes_done = n - first_nonce;
-   return 0;
-}
-*/
+#endif

 #elif defined(X22I_4WAY)

@@ -510,24 +529,33 @@ union _x22i_4way_ctx_overlay
 {
    blake512_4way_context   blake;
    bmw512_4way_context     bmw;
+#if defined(__VAES__)
+    groestl512_2way_context groestl;
+    echo_2way_context       echo;
+#else
    hashState_groestl       groestl;
    hashState_echo          echo;
+#endif
+    shavite512_2way_context shavite;
    skein512_4way_context   skein;
    jh512_4way_context      jh;
    keccak512_4way_context  keccak;
    luffa_2way_context      luffa;
    cube_2way_context       cube;
-    shavite512_2way_context shavite;
    simd_2way_context       simd;
    hamsi512_4way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_4way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_4way_context     sha512;
    haval256_5_4way_context haval;
    sph_tiger_context       tiger;
    sph_gost512_context     gost;
+#if defined(X22I_4WAY_SHA)
+    sph_sha256_context      sha256;
+#else
    sha256_4way_context     sha256;
+#endif
 };
 typedef union _x22i_4way_ctx_overlay x22i_ctx_overlay;

@@ -551,15 +579,29 @@ int x22i_4way_hash( void *output, const void *input, int thrid )
   bmw512_4way_init( &ctx.bmw );
   bmw512_4way_update( &ctx.bmw, vhash, 64 );
   bmw512_4way_close( &ctx.bmw, vhash );
+
+#if defined(__VAES__)
+
+   rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
+
+   groestl512_2way_full( &ctx.groestl, vhashA, vhashA, 64 );
+   groestl512_2way_full( &ctx.groestl, vhashB, vhashB, 64 );
+
+   rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
   dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

-   groestl512_full( &ctx.groestl, (char*)hash0, (const char*)hash0, 512 );
-   groestl512_full( &ctx.groestl, (char*)hash1, (const char*)hash1, 512 );
-   groestl512_full( &ctx.groestl, (char*)hash2, (const char*)hash2, 512 );
-   groestl512_full( &ctx.groestl, (char*)hash3, (const char*)hash3, 512 );
+   groestl512_full( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
+   groestl512_full( &ctx.groestl, (char*)hash1, (char*)hash1, 512 );
+   groestl512_full( &ctx.groestl, (char*)hash2, (char*)hash2, 512 );
+   groestl512_full( &ctx.groestl, (char*)hash3, (char*)hash3, 512 );

   intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif
+
   skein512_4way_full( &ctx.skein, vhash, vhash, 64 );

   jh512_4way_init( &ctx.jh );
@@ -586,6 +628,15 @@ int x22i_4way_hash( void *output, const void *input, int thrid )
   simd512_2way_full( &ctx.simd, vhashA, vhashA, 64 );
   simd512_2way_full( &ctx.simd, vhashB, vhashB, 64 );

+#if defined(__VAES__)
+
+   echo_2way_full( &ctx.echo, vhashA, 512, vhashA, 64 );
+   echo_2way_full( &ctx.echo, vhashB, 512, vhashB, 64 );
+
+   rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else
+
   dintrlv_2x128_512( hash0, hash1, vhashA );
   dintrlv_2x128_512( hash2, hash3, vhashB );
   
@@ -600,6 +651,8 @@ int x22i_4way_hash( void *output, const void *input, int thrid )

   intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

+#endif
+
   if ( work_restart[thrid].restart ) return false;
   
   hamsi512_4way_init( &ctx.hamsi );
@@ -607,18 +660,10 @@ int x22i_4way_hash( void *output, const void *input, int thrid )
   hamsi512_4way_close( &ctx.hamsi, vhash );
   dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );

-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512( &ctx.fugue, hash0, 64 );
-   sph_fugue512_close( &ctx.fugue, hash0 );
-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512( &ctx.fugue, hash1, 64 );
-   sph_fugue512_close( &ctx.fugue, hash1 );
-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512( &ctx.fugue, hash2, 64 );
-   sph_fugue512_close( &ctx.fugue, hash2 );
-   sph_fugue512_init( &ctx.fugue );
-   sph_fugue512( &ctx.fugue, hash3, 64 );
-   sph_fugue512_close( &ctx.fugue, hash3 );
+   fugue512_full( &ctx.fugue, hash0, hash0, 64 );
+   fugue512_full( &ctx.fugue, hash1, hash1, 64 );
+   fugue512_full( &ctx.fugue, hash2, hash2, 64 );
+   fugue512_full( &ctx.fugue, hash3, hash3, 64 );

   intrlv_4x32_512( vhash, hash0, hash1, hash2, hash3 );

@@ -649,7 +694,7 @@ int x22i_4way_hash( void *output, const void *input, int thrid )

   if ( work_restart[thrid].restart ) return false;
   
-	ComputeSingleSWIFFTX((unsigned char*)hash0, (unsigned char*)hashA0);
+   ComputeSingleSWIFFTX((unsigned char*)hash0, (unsigned char*)hashA0);
   ComputeSingleSWIFFTX((unsigned char*)hash1, (unsigned char*)hashA1);
   ComputeSingleSWIFFTX((unsigned char*)hash2, (unsigned char*)hashA2);
   ComputeSingleSWIFFTX((unsigned char*)hash3, (unsigned char*)hashA3);
@@ -663,7 +708,7 @@ int x22i_4way_hash( void *output, const void *input, int thrid )
   haval256_5_4way_close( &ctx.haval, vhash );
   dintrlv_4x32_512( hash0, hash1, hash2, hash3, vhash );
     
-	memset( hashA0, 0, 64 );
+   memset( hashA0, 0, 64 );
   memset( hashA1, 0, 64 );
   memset( hashA2, 0, 64 );
   memset( hashA3, 0, 64 );
@@ -678,8 +723,8 @@ int x22i_4way_hash( void *output, const void *input, int thrid )
   sph_tiger (&ctx.tiger, (const void*) hash2, 64);
   sph_tiger_close(&ctx.tiger, (void*) hashA2);
   sph_tiger_init(&ctx.tiger);
-	sph_tiger (&ctx.tiger, (const void*) hash3, 64);
-	sph_tiger_close(&ctx.tiger, (void*) hashA3);
+   sph_tiger (&ctx.tiger, (const void*) hash3, 64);
+   sph_tiger_close(&ctx.tiger, (void*) hashA3);

   if ( work_restart[thrid].restart ) return false;

@@ -706,9 +751,26 @@ int x22i_4way_hash( void *output, const void *input, int thrid )
   sph_gost512_init(&ctx.gost);
   sph_gost512 (&ctx.gost, (const void*) hash2, 64);
   sph_gost512_close(&ctx.gost, (void*) hash2);
-	sph_gost512_init(&ctx.gost);
-	sph_gost512 (&ctx.gost, (const void*) hash3, 64);
-	sph_gost512_close(&ctx.gost, (void*) hash3);
+   sph_gost512_init(&ctx.gost);
+   sph_gost512 (&ctx.gost, (const void*) hash3, 64);
+   sph_gost512_close(&ctx.gost, (void*) hash3);
+
+#if defined(X22I_4WAY_SHA)
+
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash0, 64 );
+   sph_sha256_close( &ctx.sha256, output );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash1, 64 );
+   sph_sha256_close( &ctx.sha256, output+32 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash2, 64 );
+   sph_sha256_close( &ctx.sha256, output+64 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash3, 64 );
+   sph_sha256_close( &ctx.sha256, output+96 );
+
+#else

   intrlv_4x32_512( vhash, hash0, hash1, hash2, hash3 );

@@ -716,11 +778,56 @@ int x22i_4way_hash( void *output, const void *input, int thrid )
   sha256_4way_update( &ctx.sha256, vhash, 64 );
   sha256_4way_close( &ctx.sha256, output );

+#endif
+
   return 1;
 }

+#if defined(X22I_4WAY_SHA)
+
+int scanhash_x22i_4way_sha( struct work* work, uint32_t max_nonce,
+                            uint64_t *hashes_done, struct thr_info *mythr )
+{
+   uint32_t hash[8*4] __attribute__ ((aligned (64)));
+   uint32_t vdata[20*4] __attribute__ ((aligned (64)));
+   uint32_t *pdata = work->data;
+   uint32_t *ptarget = work->target;
+   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 4;
+   __m256i  *noncev = (__m256i*)vdata + 9;
+   uint32_t n = first_nonce;
+   const int thr_id = mythr->id;
+   const bool bench = opt_benchmark;
+
+   if ( bench ) ptarget[7] = 0x08ff;
+
+   InitializeSWIFFTX();
+
+   mm256_bswap32_intrlv80_4x64( vdata, pdata );
+   *noncev = mm256_intrlv_blend_32(
+                   _mm256_set_epi32( n+3, 0, n+2, 0, n+1, 0, n, 0 ), *noncev );
+   do
+   {
+      if ( x22i_4way_hash( hash, vdata, thr_id ) )
+      for ( int i = 0; i < 4; i++ )
+      if ( unlikely( valid_hash( hash + (i<<3), ptarget ) && !bench ) )
+      {
+         pdata[19] = bswap_32( n+i );
+         submit_solution( work, hash+(i<<3), mythr );
+      }
+      *noncev = _mm256_add_epi32( *noncev,
+                                  m256_const1_64( 0x0000000400000000 ) );
+      n += 4;
+   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
+   pdata[19] = n;
+   *hashes_done = n - first_nonce;
+   return 0;
+}
+
+#else
+
 int scanhash_x22i_4way( struct work* work, uint32_t max_nonce,
-                   uint64_t *hashes_done, struct thr_info *mythr )
+                        uint64_t *hashes_done, struct thr_info *mythr )
 {
   uint32_t hash[8*4] __attribute__ ((aligned (64)));
   uint32_t vdata[20*4] __attribute__ ((aligned (64)));
@@ -765,4 +872,6 @@ int scanhash_x22i_4way( struct work* work, uint32_t max_nonce,
   return 0;
 }

+#endif
+
 #endif  // X22I_4WAY
--- a/algo/x22/x22i-gate.c
+++ b/algo/x22/x22i-gate.c
@@ -7,21 +7,32 @@
 bool register_x22i_algo( algo_gate_t* gate )
 {
 #if defined (X22I_8WAY)
-  gate->scanhash  = (void*)&scanhash_x22i_8way;
-  gate->hash      = (void*)&x22i_8way_hash;
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT
-                      | AVX512_OPT | VAES_OPT;
-#elif defined (X22I_4WAY)
-  gate->scanhash  = (void*)&scanhash_x22i_4way;
-  gate->hash      = (void*)&x22i_4way_hash;
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | SHA_OPT
-                      | AVX512_OPT | VAES_OPT;
+
+#if defined(X22I_8WAY_SHA)
+  gate->scanhash  = (void*)&scanhash_x22i_8way_sha;
 #else
+  gate->scanhash  = (void*)&scanhash_x22i_8way;
+#endif
+  gate->hash      = (void*)&x22i_8way_hash;
+
+#elif defined (X22I_4WAY)
+
+#if defined(X22I_4WAY_SHA)
+  gate->scanhash  = (void*)&scanhash_x22i_4way_sha;
+#else
+  gate->scanhash  = (void*)&scanhash_x22i_4way;
+#endif
+  gate->hash      = (void*)&x22i_4way_hash;
+
+#else
+
  gate->scanhash  = (void*)&scanhash_x22i;
  gate->hash      = (void*)&x22i_hash;
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | SHA_OPT
-                      | AVX512_OPT | VAES_OPT;
+
 #endif
+
+  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | SHA_OPT
+                      | AVX512_OPT | VAES_OPT | VAES256_OPT;
  return true;
 };

@@ -30,21 +41,15 @@ bool register_x25x_algo( algo_gate_t* gate )
 #if defined (X25X_8WAY)
  gate->scanhash  = (void*)&scanhash_x25x_8way;
  gate->hash      = (void*)&x25x_8way_hash;
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT
-                      | AVX512_OPT | VAES_OPT;
 #elif defined (X25X_4WAY)
  gate->scanhash  = (void*)&scanhash_x25x_4way;
  gate->hash      = (void*)&x25x_4way_hash;
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | SHA_OPT
-                      | AVX512_OPT | VAES_OPT;
 #else
  gate->scanhash  = (void*)&scanhash_x25x;
  gate->hash      = (void*)&x25x_hash;
-  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | SHA_OPT
-                      | AVX512_OPT | VAES_OPT;
 #endif
-//  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | SHA_OPT;
-
+  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | SHA_OPT |
+                        AVX512_OPT | VAES_OPT | VAES256_OPT;
  return true;
 };

--- a/algo/x22/x22i-gate.h
+++ b/algo/x22/x22i-gate.h
@@ -12,19 +12,34 @@
  #define X22I_4WAY 1
 #endif

+#if defined(__SHA__)
+//  #define X22I_8WAY_SHA 1
+  #define X22I_4WAY_SHA 1
+#endif
+
 bool register_x22i_algo( algo_gate_t* gate );

 #if defined(X22I_8WAY)

 int x22i_8way_hash( void *state, const void *input, int thrid );
+#if defined(X22I_8WAY_SHA)
+int scanhash_x22i_8way_sha( struct work *work, uint32_t max_nonce,
+                        uint64_t *hashes_done, struct thr_info *mythr );
+#else
 int scanhash_x22i_8way( struct work *work, uint32_t max_nonce,
                        uint64_t *hashes_done, struct thr_info *mythr );
+#endif

 #elif defined(X22I_4WAY)

 int x22i_4way_hash( void *state, const void *input, int thrid );
+#if defined(X22I_4WAY_SHA)
+int scanhash_x22i_4way_sha( struct work *work, uint32_t max_nonce,
+                            uint64_t *hashes_done, struct thr_info *mythr );
+#else
 int scanhash_x22i_4way( struct work *work, uint32_t max_nonce,
                        uint64_t *hashes_done, struct thr_info *mythr );
+#endif

 #else

@@ -40,6 +55,11 @@ int scanhash_x22i( struct work *work, uint32_t max_nonce,
  #define X25X_4WAY 1
 #endif

+#if defined(__SHA__)
+//  #define X25X_8WAY_SHA 1
+  #define X25X_4WAY_SHA 1
+#endif
+
 bool register_x25i_algo( algo_gate_t* gate );

 #if defined(X25X_8WAY)
--- a/algo/x22/x22i.c
+++ b/algo/x22/x22i.c
@@ -7,9 +7,11 @@
 #if defined(__AES__)
  #include "algo/echo/aes_ni/hash_api.h"
  #include "algo/groestl/aes_ni/hash-groestl.h"
+  #include "algo/fugue/fugue-aesni.h"
 #else
  #include "algo/groestl/sph_groestl.h"
  #include "algo/echo/sph_echo.h"
+  #include "algo/fugue/sph_fugue.h"
 #endif
 #include "algo/skein/sph_skein.h"
 #include "algo/jh/sph_jh.h"
@@ -19,10 +21,9 @@
 #include "algo/shavite/sph_shavite.h"
 #include "algo/simd/nist.h"
 #include "algo/hamsi/sph_hamsi.h"
-#include "algo/fugue/sph_fugue.h"
 #include "algo/shabal/sph_shabal.h"
 #include "algo/whirlpool/sph_whirlpool.h"
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"
 #include "algo/haval/sph-haval.h"
 #include "algo/tiger/sph_tiger.h"
 #include "algo/lyra2/lyra2.h"
@@ -36,9 +37,11 @@ union _x22i_context_overlay
 #if defined(__AES__)
        hashState_groestl       groestl;
        hashState_echo          echo;
+        hashState_fugue         fugue;
 #else
        sph_groestl512_context  groestl;
        sph_echo512_context     echo;
+        sph_fugue512_context    fugue;
 #endif
        sph_jh512_context       jh;
        sph_keccak512_context   keccak;
@@ -48,14 +51,13 @@ union _x22i_context_overlay
        sph_shavite512_context  shavite;
        hashState_sd            simd;
        sph_hamsi512_context    hamsi;
-        sph_fugue512_context    fugue;
        sph_shabal512_context   shabal;
        sph_whirlpool_context   whirlpool;
-        SHA512_CTX              sha512;
+        sph_sha512_context      sha512;
        sph_haval256_5_context  haval;
        sph_tiger_context       tiger;
        sph_gost512_context     gost;
-        SHA256_CTX              sha256;
+        sph_sha256_context      sha256;
 };
 typedef union _x22i_context_overlay x22i_context_overlay;

@@ -65,13 +67,13 @@ int x22i_hash( void *output, const void *input, int thrid )
   unsigned char hash2[65]    __attribute__((aligned(64))) = {0};
   x22i_context_overlay ctx;

-	sph_blake512_init(&ctx.blake);
-	sph_blake512(&ctx.blake, input, 80);
-	sph_blake512_close(&ctx.blake, hash);
+   sph_blake512_init(&ctx.blake);
+   sph_blake512(&ctx.blake, input, 80);
+   sph_blake512_close(&ctx.blake, hash);

-	sph_bmw512_init(&ctx.bmw);
-	sph_bmw512(&ctx.bmw, (const void*) hash, 64);
-	sph_bmw512_close(&ctx.bmw, hash);
+   sph_bmw512_init(&ctx.bmw);
+   sph_bmw512(&ctx.bmw, (const void*) hash, 64);
+   sph_bmw512_close(&ctx.bmw, hash);

 #if defined(__AES__)
   init_groestl( &ctx.groestl, 64 );
@@ -83,17 +85,17 @@ int x22i_hash( void *output, const void *input, int thrid )
   sph_groestl512_close( &ctx.groestl, hash );
 #endif
   
-	sph_skein512_init(&ctx.skein);
-	sph_skein512(&ctx.skein, (const void*) hash, 64);
-	sph_skein512_close(&ctx.skein, hash);
+   sph_skein512_init(&ctx.skein);
+   sph_skein512(&ctx.skein, (const void*) hash, 64);
+   sph_skein512_close(&ctx.skein, hash);

-	sph_jh512_init(&ctx.jh);
-	sph_jh512(&ctx.jh, (const void*) hash, 64);
-	sph_jh512_close(&ctx.jh, hash);
+   sph_jh512_init(&ctx.jh);
+   sph_jh512(&ctx.jh, (const void*) hash, 64);
+   sph_jh512_close(&ctx.jh, hash);

-	sph_keccak512_init(&ctx.keccak);
-	sph_keccak512(&ctx.keccak, (const void*) hash, 64);
-	sph_keccak512_close(&ctx.keccak, hash);
+   sph_keccak512_init(&ctx.keccak);
+   sph_keccak512(&ctx.keccak, (const void*) hash, 64);
+   sph_keccak512_close(&ctx.keccak, hash);

   if ( work_restart[thrid].restart ) return 0;
   
@@ -105,9 +107,9 @@ int x22i_hash( void *output, const void *input, int thrid )
   cubehashUpdateDigest( &ctx.cube, (byte*) hash,
                              (const byte*)hash, 64 );

-	sph_shavite512_init(&ctx.shavite);
-	sph_shavite512(&ctx.shavite, (const void*) hash, 64);
-	sph_shavite512_close(&ctx.shavite, hash);
+   sph_shavite512_init(&ctx.shavite);
+   sph_shavite512(&ctx.shavite, (const void*) hash, 64);
+   sph_shavite512_close(&ctx.shavite, hash);

   init_sd( &ctx.simd, 512 );
   update_final_sd( &ctx.simd, (BitSequence*)hash,
@@ -125,52 +127,56 @@ int x22i_hash( void *output, const void *input, int thrid )

   if ( work_restart[thrid].restart ) return 0;
   
-	sph_hamsi512_init(&ctx.hamsi);
-	sph_hamsi512(&ctx.hamsi, (const void*) hash, 64);
-	sph_hamsi512_close(&ctx.hamsi, hash);
+   sph_hamsi512_init(&ctx.hamsi);
+   sph_hamsi512(&ctx.hamsi, (const void*) hash, 64);
+   sph_hamsi512_close(&ctx.hamsi, hash);

-	sph_fugue512_init(&ctx.fugue);
-	sph_fugue512(&ctx.fugue, (const void*) hash, 64);
-	sph_fugue512_close(&ctx.fugue, hash);
+#if defined(__AES__)
+   fugue512_full( &ctx.fugue, hash, hash, 64 );
+#else
+   sph_fugue512_init(&ctx.fugue);
+   sph_fugue512(&ctx.fugue, (const void*) hash, 64);
+   sph_fugue512_close(&ctx.fugue, hash);
+#endif

-	sph_shabal512_init(&ctx.shabal);
-	sph_shabal512(&ctx.shabal, (const void*) hash, 64);
-	sph_shabal512_close(&ctx.shabal, &hash[64]);
+   sph_shabal512_init(&ctx.shabal);
+   sph_shabal512(&ctx.shabal, (const void*) hash, 64);
+   sph_shabal512_close(&ctx.shabal, &hash[64]);

-	sph_whirlpool_init(&ctx.whirlpool);
-	sph_whirlpool (&ctx.whirlpool, (const void*) &hash[64], 64);
-	sph_whirlpool_close(&ctx.whirlpool, &hash[128]);
+   sph_whirlpool_init(&ctx.whirlpool);
+   sph_whirlpool (&ctx.whirlpool, (const void*) &hash[64], 64);
+   sph_whirlpool_close(&ctx.whirlpool, &hash[128]);

-   SHA512_Init( &ctx.sha512 );
-   SHA512_Update(  &ctx.sha512, (const void*) &hash[128], 64);
-   SHA512_Final( (void*) &hash[192], &ctx.sha512 );
-
-	ComputeSingleSWIFFTX((unsigned char*)hash, (unsigned char*)hash2);
+   sph_sha512_init( &ctx.sha512 );
+   sph_sha512( &ctx.sha512, &hash[128], 64 );
+   sph_sha512_close( &ctx.sha512, &hash[192] );
+   
+   ComputeSingleSWIFFTX((unsigned char*)hash, (unsigned char*)hash2);

   if ( work_restart[thrid].restart ) return 0;
   
-	memset(hash, 0, 64);
-	sph_haval256_5_init(&ctx.haval);
-	sph_haval256_5(&ctx.haval,(const void*) hash2, 64);
-	sph_haval256_5_close(&ctx.haval,hash);
+   memset(hash, 0, 64);
+   sph_haval256_5_init(&ctx.haval);
+   sph_haval256_5(&ctx.haval,(const void*) hash2, 64);
+   sph_haval256_5_close(&ctx.haval,hash);

-	memset(hash2, 0, 64);
-	sph_tiger_init(&ctx.tiger);
-	sph_tiger (&ctx.tiger, (const void*) hash, 64);
-	sph_tiger_close(&ctx.tiger, (void*) hash2);
+   memset(hash2, 0, 64);
+   sph_tiger_init(&ctx.tiger);
+   sph_tiger (&ctx.tiger, (const void*) hash, 64);
+   sph_tiger_close(&ctx.tiger, (void*) hash2);

-	memset(hash, 0, 64);
-	LYRA2RE((void*) hash, 32, (const void*) hash2, 32, (const void*) hash2, 32, 1, 4, 4);
+   memset(hash, 0, 64);
+   LYRA2RE((void*) hash, 32, (const void*) hash2, 32, (const void*) hash2, 32, 1, 4, 4);

-	sph_gost512_init(&ctx.gost);
-	sph_gost512 (&ctx.gost, (const void*) hash, 64);
-	sph_gost512_close(&ctx.gost, (void*) hash);
+   sph_gost512_init(&ctx.gost);
+   sph_gost512 (&ctx.gost, (const void*) hash, 64);
+   sph_gost512_close(&ctx.gost, (void*) hash);

-   SHA256_Init( &ctx.sha256 );
-   SHA256_Update(  &ctx.sha256, (const void*) hash, 64 );
-   SHA256_Final( (unsigned char*) hash, &ctx.sha256 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash, 64 );
+   sph_sha256_close( &ctx.sha256, hash );

-	memcpy(output, hash, 32);
+   memcpy(output, hash, 32);

   return 1;
 }
--- a/algo/x22/x25x-4way.c
+++ b/algo/x22/x25x-4way.c
@@ -15,10 +15,11 @@
 #include "algo/cubehash/cubehash_sse2.h"
 #include "algo/luffa/luffa-hash-2way.h"
 #include "algo/cubehash/cube-hash-2way.h"
+#include "algo/shavite/shavite-hash-2way.h"
 #include "algo/shavite/sph_shavite.h"
 #include "algo/simd/nist.h"
 #include "algo/simd/simd-hash-2way.h"
-#include "algo/fugue/sph_fugue.h"
+#include "algo/fugue/fugue-aesni.h"
 #include "algo/whirlpool/sph_whirlpool.h"
 #include "algo/tiger/sph_tiger.h"
 #include "algo/lyra2/lyra2.h"
@@ -31,6 +32,9 @@
  #include "algo/shavite/shavite-hash-4way.h"
  #include "algo/echo/echo-hash-4way.h"
 #endif
+#if defined(__SHA__)
+  #include "algo/sha/sph_sha2.h"
+#endif

 void x25x_shuffle( void *hash )
 {
@@ -72,14 +76,18 @@ union _x25x_8way_ctx_overlay
    cube_4way_context       cube;
    simd_4way_context       simd;
    hamsi512_8way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_8way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_8way_context     sha512;
    haval256_5_8way_context haval;
    sph_tiger_context       tiger;
    sph_gost512_context     gost;
+#if defined(X25X_8WAY_SHA)
+    sph_sha256_context      sha256;
+#else
    sha256_8way_context     sha256;
+#endif
    panama_8way_context     panama;
    blake2s_8way_state      blake2s;
 #if defined(__VAES__)
@@ -215,9 +223,9 @@ int x25x_8way_hash( void *output, const void *input, int thrid )

 #else

-	sph_shavite512_init(&ctx.shavite);
-	sph_shavite512(&ctx.shavite, (const void*) hash0[7], 64);
-	sph_shavite512_close(&ctx.shavite, hash0[8]);
+   sph_shavite512_init(&ctx.shavite);
+   sph_shavite512(&ctx.shavite, (const void*) hash0[7], 64);
+   sph_shavite512_close(&ctx.shavite, hash0[8]);
   sph_shavite512_init(&ctx.shavite);
   sph_shavite512(&ctx.shavite, (const void*) hash1[7], 64);
   sph_shavite512_close(&ctx.shavite, hash1[8]);
@@ -303,30 +311,15 @@ int x25x_8way_hash( void *output, const void *input, int thrid )
   dintrlv_8x64_512( hash0[11], hash1[11], hash2[11], hash3[11],
                     hash4[11], hash5[11], hash6[11], hash7[11], vhash );
   
-	sph_fugue512_init(&ctx.fugue);
-	sph_fugue512(&ctx.fugue, (const void*) hash0[11], 64);
-	sph_fugue512_close(&ctx.fugue, hash0[12]);
-   sph_fugue512_init(&ctx.fugue);
-   sph_fugue512(&ctx.fugue, (const void*) hash1[11], 64);
-   sph_fugue512_close(&ctx.fugue, hash1[12]);
-   sph_fugue512_init(&ctx.fugue);
-   sph_fugue512(&ctx.fugue, (const void*) hash2[11], 64);
-   sph_fugue512_close(&ctx.fugue, hash2[12]);
-   sph_fugue512_init(&ctx.fugue);
-   sph_fugue512(&ctx.fugue, (const void*) hash3[11], 64);
-   sph_fugue512_close(&ctx.fugue, hash3[12]);
-   sph_fugue512_init(&ctx.fugue);
-   sph_fugue512(&ctx.fugue, (const void*) hash4[11], 64);
-   sph_fugue512_close(&ctx.fugue, hash4[12]);
-   sph_fugue512_init(&ctx.fugue);
-   sph_fugue512(&ctx.fugue, (const void*) hash5[11], 64);
-   sph_fugue512_close(&ctx.fugue, hash5[12]);
-   sph_fugue512_init(&ctx.fugue);
-   sph_fugue512(&ctx.fugue, (const void*) hash6[11], 64);
-   sph_fugue512_close(&ctx.fugue, hash6[12]);
-   sph_fugue512_init(&ctx.fugue);
-   sph_fugue512(&ctx.fugue, (const void*) hash7[11], 64);
-   sph_fugue512_close(&ctx.fugue, hash7[12]);
+   fugue512_full( &ctx.fugue, hash0[12], hash0[11], 64 );
+   fugue512_full( &ctx.fugue, hash1[12], hash1[11], 64 );
+   fugue512_full( &ctx.fugue, hash2[12], hash2[11], 64 );
+   fugue512_full( &ctx.fugue, hash3[12], hash3[11], 64 );
+   fugue512_full( &ctx.fugue, hash4[12], hash4[11], 64 );
+   fugue512_full( &ctx.fugue, hash5[12], hash5[11], 64 );
+   fugue512_full( &ctx.fugue, hash6[12], hash6[11], 64 );
+   fugue512_full( &ctx.fugue, hash7[12], hash7[11], 64 );
+
   intrlv_8x32_512( vhash, hash0[12], hash1[12], hash2[12], hash3[12],
                           hash4[12], hash5[12], hash6[12], hash7[12] );

@@ -336,9 +329,9 @@ int x25x_8way_hash( void *output, const void *input, int thrid )
   dintrlv_8x32_512( hash0[13], hash1[13], hash2[13], hash3[13],
                     hash4[13], hash5[13], hash6[13], hash7[13], vhash );

-	sph_whirlpool_init(&ctx.whirlpool);
-	sph_whirlpool (&ctx.whirlpool, (const void*) hash0[13], 64);
-	sph_whirlpool_close(&ctx.whirlpool, hash0[14]);
+   sph_whirlpool_init(&ctx.whirlpool);
+   sph_whirlpool (&ctx.whirlpool, (const void*) hash0[13], 64);
+   sph_whirlpool_close(&ctx.whirlpool, hash0[14]);
   sph_whirlpool_init(&ctx.whirlpool);
   sph_whirlpool (&ctx.whirlpool, (const void*) hash1[13], 64);
   sph_whirlpool_close(&ctx.whirlpool, hash1[14]);
@@ -387,9 +380,9 @@ int x25x_8way_hash( void *output, const void *input, int thrid )
   dintrlv_8x32_512( hash0[17], hash1[17], hash2[17], hash3[17],
                     hash4[17], hash5[17], hash6[17], hash7[17], vhash );

-	sph_tiger_init(&ctx.tiger);
-	sph_tiger (&ctx.tiger, (const void*) hash0[17], 64);
-	sph_tiger_close(&ctx.tiger, (void*) hash0[18]);
+   sph_tiger_init(&ctx.tiger);
+   sph_tiger (&ctx.tiger, (const void*) hash0[17], 64);
+   sph_tiger_close(&ctx.tiger, (void*) hash0[18]);
   sph_tiger_init(&ctx.tiger);
   sph_tiger (&ctx.tiger, (const void*) hash1[17], 64);
   sph_tiger_close(&ctx.tiger, (void*) hash1[18]);
@@ -427,9 +420,9 @@ int x25x_8way_hash( void *output, const void *input, int thrid )
   LYRA2X_2WAY( vhash, 32, vhash, 32, 1, 4, 4 );
   dintrlv_2x256( hash6[19], hash7[19], vhash, 256 );

-	sph_gost512_init(&ctx.gost);
-	sph_gost512 (&ctx.gost, (const void*) hash0[19], 64);
-	sph_gost512_close(&ctx.gost, (void*) hash0[20]);
+   sph_gost512_init(&ctx.gost);
+   sph_gost512 (&ctx.gost, (const void*) hash0[19], 64);
+   sph_gost512_close(&ctx.gost, (void*) hash0[20]);
   sph_gost512_init(&ctx.gost);
   sph_gost512 (&ctx.gost, (const void*) hash1[19], 64);
   sph_gost512_close(&ctx.gost, (void*) hash1[20]);
@@ -451,6 +444,39 @@ int x25x_8way_hash( void *output, const void *input, int thrid )
   sph_gost512_init(&ctx.gost);
   sph_gost512 (&ctx.gost, (const void*) hash7[19], 64);
   sph_gost512_close(&ctx.gost, (void*) hash7[20]);
+
+#if defined(X25X_8WAY_SHA)
+
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash0[20], 64 );
+   sph_sha256_close( &ctx.sha256, hash0[21] );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash1[20], 64 );
+   sph_sha256_close( &ctx.sha256, hash1[21] );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash2[20], 64 );
+   sph_sha256_close( &ctx.sha256, hash2[21] );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash3[20], 64 );
+   sph_sha256_close( &ctx.sha256, hash3[21] );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash4[20], 64 );
+   sph_sha256_close( &ctx.sha256, hash4[21] );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash5[20], 64 );
+   sph_sha256_close( &ctx.sha256, hash5[21] );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash6[20], 64 );
+   sph_sha256_close( &ctx.sha256, hash6[21] );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash7[20], 64 );
+   sph_sha256_close( &ctx.sha256, hash7[21] );
+
+   intrlv_8x32_512( vhash, hash0[21], hash1[21], hash2[21], hash3[21],
+                           hash4[21], hash5[21], hash6[21], hash7[21] );
+   
+#else
+
   intrlv_8x32_512( vhashA, hash0[20], hash1[20], hash2[20], hash3[20],
                            hash4[20], hash5[20], hash6[20], hash7[20] );

@@ -460,6 +486,8 @@ int x25x_8way_hash( void *output, const void *input, int thrid )
   dintrlv_8x32_512( hash0[21], hash1[21], hash2[21], hash3[21],
                     hash4[21], hash5[21], hash6[21], hash7[21], vhash );

+#endif
+
   panama_8way_init( &ctx.panama );
   panama_8way_update( &ctx.panama, vhash, 64 );
   panama_8way_close( &ctx.panama, vhash );
@@ -589,77 +617,39 @@ int scanhash_x25x_8way( struct work *work, uint32_t max_nonce,
   return 0;
 }

-/*
-int scanhash_x25x_8way( struct work* work, uint32_t max_nonce,
-                   uint64_t *hashes_done, struct thr_info *mythr )
-{
-   uint32_t hash[8*16] __attribute__ ((aligned (128)));
-   uint32_t vdata[24*8] __attribute__ ((aligned (64)));
-   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
-   uint32_t *hash7 = &(hash[7<<3]);
-   uint32_t *pdata = work->data;
-   uint32_t *ptarget = work->target;
-   const uint32_t first_nonce = pdata[19];
-   __m512i  *noncev = (__m512i*)vdata + 9;   // aligned
-   uint32_t n = first_nonce;
-   const uint32_t last_nonce = max_nonce - 4;
-   const int thr_id = mythr->id;
-   const uint32_t Htarg = ptarget[7];
-
-   if (opt_benchmark)
-      ((uint32_t*)ptarget)[7] = 0x08ff;
-
-   InitializeSWIFFTX();
-
-   mm512_bswap32_intrlv80_8x64( vdata, pdata );
-   do
-   {
-      *noncev = mm512_intrlv_blend_32( mm512_bswap_32(
-              _mm512_set_epi32( n+7, 0, n+6, 0, n+5, 0, n+4, 0,
-                                n+3, 0, n+2, 0, n+1, 0, n,   0 ) ), *noncev );
-      x25x_8way_hash( hash, vdata );
-
-      for ( int lane = 0; lane < 8; lane++ ) if ( hash7[lane] <= Htarg )
-      {
-         extr_lane_8x32( lane_hash, hash, lane, 256 );
-         if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
-         {
-              pdata[19] = n + lane;
-              submit_solution( work, lane_hash, mythr );
-         }
-      }
-      n += 8;
-   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
-
-   *hashes_done = n - first_nonce;
-   return 0;
-}
-*/
-
 #elif defined(X25X_4WAY)

 union _x25x_4way_ctx_overlay
 {
    blake512_4way_context   blake;
    bmw512_4way_context     bmw;
+#if defined(__VAES__)
+    groestl512_2way_context groestl;
+    echo_2way_context       echo;
+#else
    hashState_groestl       groestl;
    hashState_echo          echo;
+#endif
    skein512_4way_context   skein;
    jh512_4way_context      jh;
    keccak512_4way_context  keccak;
-    hashState_luffa         luffa;
-    cubehashParam           cube;
-    sph_shavite512_context  shavite;
-    hashState_sd            simd;
+    luffa_2way_context      luffa;
+    cube_2way_context       cube;
+    shavite512_2way_context shavite;
+    simd_2way_context       simd;
    hamsi512_4way_context   hamsi;
-    sph_fugue512_context    fugue;
+    hashState_fugue         fugue;
    shabal512_4way_context  shabal;
    sph_whirlpool_context   whirlpool;
    sha512_4way_context     sha512;
    haval256_5_4way_context haval;
    sph_tiger_context       tiger;
    sph_gost512_context     gost;
+#if defined(X25X_4WAY_SHA)
+    sph_sha256_context      sha256;
+#else
    sha256_4way_context     sha256;
+#endif
    panama_4way_context     panama;
    blake2s_4way_state      blake2s;
 };
@@ -673,6 +663,8 @@ int x25x_4way_hash( void *output, const void *input, int thrid )
   unsigned char hash2[25][64] __attribute__((aligned(64))) = {0};
   unsigned char hash3[25][64] __attribute__((aligned(64))) = {0};
   unsigned char vhashX[24][64*4] __attribute__ ((aligned (64)));
+   uint64_t vhashA[8*4] __attribute__ ((aligned (64)));
+   uint64_t vhashB[8*4] __attribute__ ((aligned (64)));
   x25x_4way_ctx_overlay ctx __attribute__ ((aligned (64)));

   blake512_4way_full( &ctx.blake, vhash, input, 80 );
@@ -683,11 +675,25 @@ int x25x_4way_hash( void *output, const void *input, int thrid )
   bmw512_4way_close( &ctx.bmw, vhash );
   dintrlv_4x64_512( hash0[1], hash1[1], hash2[1], hash3[1], vhash );

+#if defined(__VAES__)
+
+   rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
+
+   groestl512_2way_full( &ctx.groestl, vhashA, vhashA, 64 );
+   groestl512_2way_full( &ctx.groestl, vhashB, vhashB, 64 );
+
+   dintrlv_2x128_512( hash0[2], hash1[2], vhashA );
+   dintrlv_2x128_512( hash2[2], hash3[2], vhashB );
+
+#else
+
   groestl512_full( &ctx.groestl, (char*)hash0[2], (const char*)hash0[1], 512 );
   groestl512_full( &ctx.groestl, (char*)hash1[2], (const char*)hash1[1], 512 );
   groestl512_full( &ctx.groestl, (char*)hash2[2], (const char*)hash2[1], 512 );
   groestl512_full( &ctx.groestl, (char*)hash3[2], (const char*)hash3[1], 512 );

+#endif
+
   intrlv_4x64_512( vhash, hash0[2], hash1[2], hash2[2], hash3[2] );
   skein512_4way_full( &ctx.skein, vhash, vhash, 64 );
   dintrlv_4x64_512( hash0[3], hash1[3], hash2[3], hash3[3], vhash );
@@ -704,41 +710,38 @@ int x25x_4way_hash( void *output, const void *input, int thrid )
   keccak512_4way_close( &ctx.keccak, vhash );
   dintrlv_4x64_512( hash0[5], hash1[5], hash2[5], hash3[5], vhash );

-   luffa_full( &ctx.luffa, (BitSequence*)hash0[6], 512,
-                     (const BitSequence*)hash0[5], 64 );
-   luffa_full( &ctx.luffa, (BitSequence*)hash1[6], 512,
-                     (const BitSequence*)hash1[5], 64 );
-   luffa_full( &ctx.luffa, (BitSequence*)hash2[6], 512,
-                     (const BitSequence*)hash2[5], 64 );
-   luffa_full( &ctx.luffa, (BitSequence*)hash3[6], 512,
-                     (const BitSequence*)hash3[5], 64 );
+   rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );

-   cubehash_full( &ctx.cube, (byte*)hash0[7], 512, (const byte*)hash0[6], 64 );
-   cubehash_full( &ctx.cube, (byte*)hash1[7], 512, (const byte*)hash1[6], 64 );
-   cubehash_full( &ctx.cube, (byte*)hash2[7], 512, (const byte*)hash2[6], 64 );
-   cubehash_full( &ctx.cube, (byte*)hash3[7], 512, (const byte*)hash3[6], 64 );
+   luffa512_2way_full( &ctx.luffa, vhashA, vhashA, 64 );
+   luffa512_2way_full( &ctx.luffa, vhashB, vhashB, 64 );
+   dintrlv_2x128_512( hash0[6], hash1[6], vhashA );
+   dintrlv_2x128_512( hash2[6], hash3[6], vhashB );
   
-   sph_shavite512_init(&ctx.shavite);
-   sph_shavite512(&ctx.shavite, (const void*) hash0[7], 64);
-   sph_shavite512_close(&ctx.shavite, hash0[8]);
-   sph_shavite512_init(&ctx.shavite);
-   sph_shavite512(&ctx.shavite, (const void*) hash1[7], 64);
-   sph_shavite512_close(&ctx.shavite, hash1[8]);
-   sph_shavite512_init(&ctx.shavite);
-   sph_shavite512(&ctx.shavite, (const void*) hash2[7], 64);
-   sph_shavite512_close(&ctx.shavite, hash2[8]);
-   sph_shavite512_init(&ctx.shavite);
-   sph_shavite512(&ctx.shavite, (const void*) hash3[7], 64);
-   sph_shavite512_close(&ctx.shavite, hash3[8]);
+   cube_2way_full( &ctx.cube, vhashA, 512, vhashA, 64 );
+   cube_2way_full( &ctx.cube, vhashB, 512, vhashB, 64 );
+   dintrlv_2x128_512( hash0[7], hash1[7], vhashA );
+   dintrlv_2x128_512( hash2[7], hash3[7], vhashB );

-   simd_full( &ctx.simd, (BitSequence*)hash0[9],
-                   (const BitSequence*)hash0[8], 512 );
-   simd_full( &ctx.simd, (BitSequence*)hash1[9],
-                   (const BitSequence*)hash1[8], 512 );
-   simd_full( &ctx.simd, (BitSequence*)hash2[9],
-                   (const BitSequence*)hash2[8], 512 );
-   simd_full( &ctx.simd, (BitSequence*)hash3[9],
-                   (const BitSequence*)hash3[8], 512 );
+   shavite512_2way_full( &ctx.shavite, vhashA, vhashA, 64 );
+   shavite512_2way_full( &ctx.shavite, vhashB, vhashB, 64 );
+   dintrlv_2x128_512( hash0[8], hash1[8], vhashA );
+   dintrlv_2x128_512( hash2[8], hash3[8], vhashB );
+
+   simd512_2way_full( &ctx.simd, vhashA, vhashA, 64 );
+   simd512_2way_full( &ctx.simd, vhashB, vhashB, 64 );
+   dintrlv_2x128_512( hash0[9], hash1[9], vhashA );
+   dintrlv_2x128_512( hash2[9], hash3[9], vhashB );
+
+#if defined(__VAES__)
+
+   echo_2way_full( &ctx.echo, vhashA, 512, vhashA, 64 );
+   echo_2way_full( &ctx.echo, vhashB, 512, vhashB, 64 );
+   dintrlv_2x128_512( hash0[10], hash1[10], vhashA );
+   dintrlv_2x128_512( hash2[10], hash3[10], vhashB );
+
+   rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
+
+#else

   echo_full( &ctx.echo, (BitSequence *)hash0[10], 512,
                   (const BitSequence *)hash0[ 9], 64 );
@@ -751,6 +754,8 @@ int x25x_4way_hash( void *output, const void *input, int thrid )

   intrlv_4x64_512( vhash, hash0[10], hash1[10], hash2[10], hash3[10] );

+#endif
+
   if ( work_restart[thrid].restart ) return 0;
   
   hamsi512_4way_init( &ctx.hamsi );
@@ -758,18 +763,10 @@ int x25x_4way_hash( void *output, const void *input, int thrid )
   hamsi512_4way_close( &ctx.hamsi, vhash );
   dintrlv_4x64_512( hash0[11], hash1[11], hash2[11], hash3[11], vhash );

-   sph_fugue512_init(&ctx.fugue);
-   sph_fugue512(&ctx.fugue, (const void*) hash0[11], 64);
-   sph_fugue512_close(&ctx.fugue, hash0[12]);
-   sph_fugue512_init(&ctx.fugue);
-   sph_fugue512(&ctx.fugue, (const void*) hash1[11], 64);
-   sph_fugue512_close(&ctx.fugue, hash1[12]);
-   sph_fugue512_init(&ctx.fugue);
-   sph_fugue512(&ctx.fugue, (const void*) hash2[11], 64);
-   sph_fugue512_close(&ctx.fugue, hash2[12]);
-   sph_fugue512_init(&ctx.fugue);
-   sph_fugue512(&ctx.fugue, (const void*) hash3[11], 64);
-   sph_fugue512_close(&ctx.fugue, hash3[12]);
+   fugue512_full( &ctx.fugue, hash0[12], hash0[11], 64 );
+   fugue512_full( &ctx.fugue, hash1[12], hash1[11], 64 );
+   fugue512_full( &ctx.fugue, hash2[12], hash2[11], 64 );
+   fugue512_full( &ctx.fugue, hash3[12], hash3[11], 64 );

   intrlv_4x32_512( vhash, hash0[12], hash1[12], hash2[12], hash3[12] );

@@ -849,6 +846,25 @@ int x25x_4way_hash( void *output, const void *input, int thrid )
   sph_gost512 (&ctx.gost, (const void*) hash3[19], 64);
   sph_gost512_close(&ctx.gost, (void*) hash3[20]);

+#if defined(X25X_4WAY_SHA)
+
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash0[20], 64 );
+   sph_sha256_close( &ctx.sha256, hash0[21] );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash1[20], 64 );
+   sph_sha256_close( &ctx.sha256, hash1[21] );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash2[20], 64 );
+   sph_sha256_close( &ctx.sha256, hash2[21] );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, hash3[20], 64 );
+   sph_sha256_close( &ctx.sha256, hash3[21] );
+
+   intrlv_4x32_512( vhash, hash0[21], hash1[21], hash2[21], hash3[21] );
+
+#else   
+
   intrlv_4x32_512( vhashX[0], hash0[20], hash1[20], hash2[20], hash3[20] );
   memset( vhash, 0, 64*4 );

@@ -857,6 +873,8 @@ int x25x_4way_hash( void *output, const void *input, int thrid )
   sha256_4way_close( &ctx.sha256, vhash );
   dintrlv_4x32_512( hash0[21], hash1[21], hash2[21], hash3[21], vhash );

+#endif
+
   panama_4way_init( &ctx.panama );
   panama_4way_update( &ctx.panama, vhash, 64 );
   panama_4way_close( &ctx.panama, vhash );
--- a/algo/x22/x25x.c
+++ b/algo/x22/x25x.c
@@ -7,9 +7,11 @@
 #if defined(__AES__)
  #include "algo/echo/aes_ni/hash_api.h"
  #include "algo/groestl/aes_ni/hash-groestl.h"
+  #include "algo/fugue/fugue-aesni.h"
 #else
  #include "algo/groestl/sph_groestl.h"
  #include "algo/echo/sph_echo.h"
+  #include "algo/fugue/sph_fugue.h"
 #endif
 #include "algo/skein/sph_skein.h"
 #include "algo/jh/sph_jh.h"
@@ -19,10 +21,9 @@
 #include "algo/shavite/sph_shavite.h"
 #include "algo/simd/nist.h"
 #include "algo/hamsi/sph_hamsi.h"
-#include "algo/fugue/sph_fugue.h"
 #include "algo/shabal/sph_shabal.h"
 #include "algo/whirlpool/sph_whirlpool.h"
-#include <openssl/sha.h>
+#include "algo/sha/sph_sha2.h"
 #include "algo/haval/sph-haval.h"
 #include "algo/tiger/sph_tiger.h"
 #include "algo/lyra2/lyra2.h"
@@ -39,9 +40,11 @@ union _x25x_context_overlay
 #if defined(__AES__)
        hashState_groestl       groestl;
        hashState_echo          echo;
+        hashState_fugue         fugue;
 #else
        sph_groestl512_context  groestl;
        sph_echo512_context     echo;
+        sph_fugue512_context    fugue;
 #endif
        sph_jh512_context       jh;
        sph_keccak512_context   keccak;
@@ -51,14 +54,13 @@ union _x25x_context_overlay
        sph_shavite512_context  shavite;
        hashState_sd            simd;
        sph_hamsi512_context    hamsi;
-        sph_fugue512_context    fugue;
        sph_shabal512_context   shabal;
        sph_whirlpool_context   whirlpool;
-        SHA512_CTX              sha512;
+        sph_sha512_context      sha512;
        sph_haval256_5_context  haval;
        sph_tiger_context       tiger;
        sph_gost512_context     gost;
-        SHA256_CTX              sha256;
+        sph_sha256_context      sha256;
        sph_panama_context      panama;
        blake2s_state           blake2s;
 };
@@ -69,13 +71,13 @@ int x25x_hash( void *output, const void *input, int thrid )
   unsigned char hash[25][64] __attribute__((aligned(64))) = {0};
   x25x_context_overlay ctx;

-	sph_blake512_init(&ctx.blake);
-	sph_blake512(&ctx.blake, input, 80);
-	sph_blake512_close(&ctx.blake, &hash[0] );
+   sph_blake512_init(&ctx.blake);
+   sph_blake512(&ctx.blake, input, 80);
+   sph_blake512_close(&ctx.blake, &hash[0] );

-	sph_bmw512_init(&ctx.bmw);
-	sph_bmw512(&ctx.bmw, (const void*) &hash[0], 64);
-	sph_bmw512_close(&ctx.bmw, &hash[1]);
+   sph_bmw512_init(&ctx.bmw);
+   sph_bmw512(&ctx.bmw, (const void*) &hash[0], 64);
+   sph_bmw512_close(&ctx.bmw, &hash[1]);

 #if defined(__AES__)
   init_groestl( &ctx.groestl, 64 );
@@ -87,17 +89,17 @@ int x25x_hash( void *output, const void *input, int thrid )
   sph_groestl512_close( &ctx.groestl, &hash[2] );
 #endif
   
-	sph_skein512_init(&ctx.skein);
-	sph_skein512(&ctx.skein, (const void*) &hash[2], 64);
-	sph_skein512_close(&ctx.skein, &hash[3]);
+   sph_skein512_init(&ctx.skein);
+   sph_skein512(&ctx.skein, (const void*) &hash[2], 64);
+   sph_skein512_close(&ctx.skein, &hash[3]);

-	sph_jh512_init(&ctx.jh);
-	sph_jh512(&ctx.jh, (const void*) &hash[3], 64);
-	sph_jh512_close(&ctx.jh, &hash[4]);
+   sph_jh512_init(&ctx.jh);
+   sph_jh512(&ctx.jh, (const void*) &hash[3], 64);
+   sph_jh512_close(&ctx.jh, &hash[4]);

-	sph_keccak512_init(&ctx.keccak);
-	sph_keccak512(&ctx.keccak, (const void*) &hash[4], 64);
-	sph_keccak512_close(&ctx.keccak, &hash[5]);
+   sph_keccak512_init(&ctx.keccak);
+   sph_keccak512(&ctx.keccak, (const void*) &hash[4], 64);
+   sph_keccak512_close(&ctx.keccak, &hash[5]);

   if ( work_restart[thrid].restart ) return 0;
   
@@ -109,9 +111,9 @@ int x25x_hash( void *output, const void *input, int thrid )
   cubehashUpdateDigest( &ctx.cube, (byte*) &hash[7],
                              (const byte*)&hash[6], 64 );

-	sph_shavite512_init(&ctx.shavite);
-	sph_shavite512(&ctx.shavite, (const void*) &hash[7], 64);
-	sph_shavite512_close(&ctx.shavite, &hash[8]);
+   sph_shavite512_init(&ctx.shavite);
+   sph_shavite512(&ctx.shavite, (const void*) &hash[7], 64);
+   sph_shavite512_close(&ctx.shavite, &hash[8]);

   init_sd( &ctx.simd, 512 );
   update_final_sd( &ctx.simd, (BitSequence*)&hash[9],
@@ -130,47 +132,51 @@ int x25x_hash( void *output, const void *input, int thrid )
   if ( work_restart[thrid].restart ) return 0;

   sph_hamsi512_init(&ctx.hamsi);
-	sph_hamsi512(&ctx.hamsi, (const void*) &hash[10], 64);
-	sph_hamsi512_close(&ctx.hamsi, &hash[11]);
+   sph_hamsi512(&ctx.hamsi, (const void*) &hash[10], 64);
+   sph_hamsi512_close(&ctx.hamsi, &hash[11]);

-	sph_fugue512_init(&ctx.fugue);
-	sph_fugue512(&ctx.fugue, (const void*) &hash[11], 64);
-	sph_fugue512_close(&ctx.fugue, &hash[12]);
+#if defined(__AES__)
+   fugue512_full( &ctx.fugue, &hash[12], &hash[11], 64 );
+#else
+   sph_fugue512_init(&ctx.fugue);
+   sph_fugue512(&ctx.fugue, (const void*) &hash[11], 64);
+   sph_fugue512_close(&ctx.fugue, &hash[12]);
+#endif

-	sph_shabal512_init(&ctx.shabal);
-	sph_shabal512(&ctx.shabal, (const void*) &hash[12], 64);
-	sph_shabal512_close(&ctx.shabal, &hash[13]);
+   sph_shabal512_init(&ctx.shabal);
+   sph_shabal512(&ctx.shabal, (const void*) &hash[12], 64);
+   sph_shabal512_close(&ctx.shabal, &hash[13]);

-	sph_whirlpool_init(&ctx.whirlpool);
-	sph_whirlpool (&ctx.whirlpool, (const void*) &hash[13], 64);
-	sph_whirlpool_close(&ctx.whirlpool, &hash[14]);
+   sph_whirlpool_init(&ctx.whirlpool);
+   sph_whirlpool (&ctx.whirlpool, (const void*) &hash[13], 64);
+   sph_whirlpool_close(&ctx.whirlpool, &hash[14]);

-   SHA512_Init( &ctx.sha512 );
-   SHA512_Update(  &ctx.sha512, (const void*) &hash[14], 64);
-   SHA512_Final( (void*) &hash[15], &ctx.sha512 );
+   sph_sha512_init( &ctx.sha512 );
+   sph_sha512( &ctx.sha512, &hash[14], 64 );
+   sph_sha512_close( &ctx.sha512, &hash[15] );

   ComputeSingleSWIFFTX((unsigned char*)&hash[12], (unsigned char*)&hash[16]);

-	sph_haval256_5_init(&ctx.haval);
-	sph_haval256_5(&ctx.haval,(const void*) &hash[16], 64);
-	sph_haval256_5_close(&ctx.haval,&hash[17]);
+   sph_haval256_5_init(&ctx.haval);
+   sph_haval256_5(&ctx.haval,(const void*) &hash[16], 64);
+   sph_haval256_5_close(&ctx.haval,&hash[17]);

   if ( work_restart[thrid].restart ) return 0;
   
-	sph_tiger_init(&ctx.tiger);
-	sph_tiger (&ctx.tiger, (const void*) &hash[17], 64);
-	sph_tiger_close(&ctx.tiger, (void*) &hash[18]);
+   sph_tiger_init(&ctx.tiger);
+   sph_tiger (&ctx.tiger, (const void*) &hash[17], 64);
+   sph_tiger_close(&ctx.tiger, (void*) &hash[18]);

-	LYRA2RE( (void*)&hash[19], 32, (const void*)&hash[18], 32,
+   LYRA2RE( (void*)&hash[19], 32, (const void*)&hash[18], 32,
            (const void*)&hash[18], 32, 1, 4, 4 );

-	sph_gost512_init(&ctx.gost);
-	sph_gost512 (&ctx.gost, (const void*) &hash[19], 64);
-	sph_gost512_close(&ctx.gost, (void*) &hash[20]);
+   sph_gost512_init(&ctx.gost);
+   sph_gost512 (&ctx.gost, (const void*) &hash[19], 64);
+   sph_gost512_close(&ctx.gost, (void*) &hash[20]);

-   SHA256_Init( &ctx.sha256 );
-   SHA256_Update(  &ctx.sha256, (const void*) &hash[20], 64 );
-   SHA256_Final( (unsigned char*) &hash[21], &ctx.sha256 );
+   sph_sha256_init( &ctx.sha256 );
+   sph_sha256( &ctx.sha256, &hash[20], 64 );
+   sph_sha256_close( &ctx.sha256, &hash[21] );

   sph_panama_init(&ctx.panama);
   sph_panama (&ctx.panama, (const void*) &hash[21], 64 );
--- a/algo/yescrypt/yescrypt-simd.c
+++ b/algo/yescrypt/yescrypt-simd.c
@@ -1302,10 +1302,7 @@ yescrypt_kdf(const yescrypt_shared_t * shared, yescrypt_local_t * local,
 		S = (uint8_t *)XY + XY_size;

 	if (t || flags) {
-		SHA256_CTX ctx;
-		SHA256_Init(&ctx);
-		SHA256_Update(&ctx, passwd, passwdlen);
-		SHA256_Final(sha256, &ctx);
+		SHA256_Buf( passwd, passwdlen, sha256 );
 		passwd = sha256;
 		passwdlen = sizeof(sha256);
 	}
@@ -1382,10 +1379,7 @@ yescrypt_kdf(const yescrypt_shared_t * shared, yescrypt_local_t * local,
 	   }
 	   /* Compute StoredKey */
 	   {
-		SHA256_CTX ctx;
-		SHA256_Init(&ctx);
-		SHA256_Update(&ctx, sha256, sizeof(sha256));
-		SHA256_Final(buf, &ctx);
+		SHA256_Buf( sha256, sizeof(sha256), buf );
 	   }
 	}

--- a/algo/yescrypt/yescrypt.c
+++ b/algo/yescrypt/yescrypt.c
@@ -445,7 +445,7 @@ bool register_yescrypt_algo( algo_gate_t* gate )

   YESCRYPT_P = 1;

-   applog( LOG_NOTICE,"Yescrypt parameters: N= %d, R= %d.", YESCRYPT_N,
+   applog( LOG_NOTICE,"Yescrypt parameters: N= %d, R= %d", YESCRYPT_N,
                                                            YESCRYPT_R );
   if ( yescrypt_client_key )
     applog( LOG_NOTICE,"Key= \"%s\"\n", yescrypt_client_key );
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
Jay D Dee	40089428c5	v3.15.7	2021-03-08 22:44:44 -05:00
Jay D Dee	dc6b007a18	v3.15.6	2021-02-12 15:16:53 -05:00
Jay D Dee	06bfaa1249	v3.15.5	2020-12-21 13:25:33 -05:00
Jay D Dee	6566e99a13	v3.15.4	2020-12-15 13:15:02 -05:00
Jay D Dee	ccfccbadd5	v3.15.3	2020-12-10 18:23:49 -05:00
Jay D Dee	45ecd0de14	v3.15.2	2020-11-15 17:57:06 -05:00
Jay D Dee	4fa8fcea8b	v3.15.1	2020-11-09 13:19:05 -05:00
Jay D Dee	c85fb3842b	v3.15.0	2020-10-02 10:48:37 -04:00
Jay D Dee	cdd587537e	v3.14.3	2020-06-18 17:30:26 -04:00
Jay D Dee	51a1d91abd	v3.14.2	2020-05-30 21:20:44 -04:00
Jay D Dee	13563e2598	v3.14.1	2020-05-21 13:00:29 -04:00
Jay D Dee	9571f85d53	v3.14.0	2020-05-20 13:56:35 -04:00
Jay D Dee	0e69756634	v3.13.2-segwit-test	2020-05-18 18:17:27 -04:00
Jay D Dee	9653bca1e2	v3.13.1.1	2020-05-17 19:21:37 -04:00
Jay D Dee	1c0719e8a4	v3.13.1	2020-05-10 21:34:55 -04:00
Jay D Dee	8b4b4dc613	v3.13.0.1	2020-05-07 17:57:04 -04:00
Jay D Dee	e76feaced8	v3.13.0	2020-05-06 00:53:43 -04:00