v3.14.3

Makefile.am

@@ -163,6 +163,7 @@ cpuminer_SOURCES = \
   algo/sha/sha256-hash-4way.c \
   algo/sha/sha512-hash-4way.c \
   algo/sha/hmac-sha256-hash.c \
+  algo/sha/hmac-sha256-hash-4way.c \
   algo/sha/sha2.c \
   algo/sha/sha256t-gate.c \
   algo/sha/sha256t-4way.c \
@@ -256,6 +257,7 @@ 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 \

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

README.txt

@@ -1,8 +1,8 @@
 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.
+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 
@@ -15,8 +15,8 @@ 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 +31,29 @@ 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*
+cpuminer-avx512.exe          "-march=skylake-avx512"  Skylake-X, Cascadelake-X
+cpuminer-zen.exe             "-march=znver1"          AMD Ryzen, Threadripper
+cpuminer-avx512-sha-vaes.exe "-march=icelake-client"  Icelake*
+
+* Haswell includes Broadwell, Skylake, Kabylake, Coffeelake & Cometlake. 
+Icelake is only available on some laptops. Mining with a laptop is not
+recommended. The icelake build is included in anticipation of Intel eventually
+releasing a desktop CPU with a microarchitecture newer than Skylake.
+
+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 instalation 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:
 

RELEASE_NOTES

@@ -65,6 +65,91 @@ If not what makes it happen or not happen?
 Change 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.
+Fixed phi AVX2 crash.
+
+v3.12.8.1
+
+Issue #261: Fixed yescryptr8g invalid shares.
+
+v3.12.8
+
+Yespower sha256 prehash made thread safe.
+
+Rewrote diff conversion functions from scratch to be simpler and use 
+long double (float80) and int128 arithmetic for improved accuracy and
+precision.
+
+Some code cleanup and assorted small changes.
+
+v3.12.7
+
+Issue #257: fixed a file descriptor leak which caused the CPU temperature
+and frequency query to report zeros after mining for a couple of hours.
+
+Issue #253: stale share reduction for yescrypt, sonoa.
+
 v3.12.6.1
 
 Issue #252: Fixed SSL mining (stratum+tcps://)
@@ -106,7 +191,7 @@ a specific algo name.
 v3.12.4.6
 
 Issue #246: fixed getwork repeated new block logs with same height. New work
-for the same block is now reported as "New work" instead of New block".
+for the same block is now reported as "New work" instead of "New block".
 Also added a check that work is new before generating "New work" log.
 
 Added target diff to getwork new block log.

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,

algo-gate-api.c

@@ -90,35 +90,172 @@ 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");
    return 0;
 };
-/*
-void null_hash_suw()
-{
-  applog(LOG_WARNING,"SWERR: null_hash_suw unsafe null function");
-};
-*/
+
 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->hash_suw                = (void*)&null_hash_suw;
    gate->get_new_work            = (void*)&std_get_new_work;
    gate->work_decode             = (void*)&std_le_work_decode;
    gate->decode_extra_data       = (void*)&do_nothing;
    gate->gen_merkle_root         = (void*)&sha256d_gen_merkle_root;
-   gate->stratum_gen_work        = (void*)&std_stratum_gen_work;
    gate->build_stratum_request   = (void*)&std_le_build_stratum_request;
    gate->malloc_txs_request      = (void*)&std_malloc_txs_request;
    gate->submit_getwork_result   = (void*)&std_le_submit_getwork_result;
@@ -186,6 +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;
@@ -232,11 +370,6 @@ 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_05_algo     ( gate ); break;
-     case ALGO_YESCRYPTR8:   register_yescryptr8_05_algo   ( gate ); break;
-     case ALGO_YESCRYPTR16:  register_yescryptr16_05_algo  ( gate ); break;
-     case ALGO_YESCRYPTR32:  register_yescryptr32_05_algo  ( gate ); break;
-*/
     case ALGO_YESCRYPT:      register_yescrypt_algo      ( gate ); break;
     case ALGO_YESCRYPTR8:    register_yescryptr8_algo    ( gate ); break;
     case ALGO_YESCRYPTR8G:   register_yescryptr8g_algo   ( gate ); break;
@@ -263,7 +396,6 @@ 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;   

algo-gate-api.h

@@ -75,7 +75,7 @@
 
 // my hack at creating a set data type using bit masks. Set inclusion,
 // exclusion union and intersection operations are provided for convenience. In // some cases it may be desireable to use boolean algebra directly on the
-// data to perfomr set operations. Sets can be represented as single
+// data to perform set operations. Sets can be represented as single
 // elements, a bitwise OR of multiple elements, a bitwise OR of multiple
 // set variables or constants, or combinations of the above.
 // Examples:
@@ -110,13 +110,13 @@ inline bool set_excl ( set_t a, set_t b ) { return (a & b) == 0; }
 
 typedef struct
 {
-// mandatory functions, must be overwritten
+// Mandatory functions, one of these is mandatory. If a generic scanhash
+// is used a custom hash function must be registered, with a custom scanhash
+// the custom 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* );
 
-// not used anywhere
-// optional unsafe, must be overwritten if algo uses function
-int ( *hash )     ( void*, const void*, uint32_t ) ;
-//void ( *hash_suw ) ( void*, const void* );
+int ( *hash )     ( void*, const void*, int );
 
 //optional, safe to use default in most cases
 
@@ -124,14 +124,11 @@ int ( *hash )     ( void*, const void*, uint32_t ) ;
 // threads.
 bool ( *miner_thread_init )     ( int );
 
-// Generate global blockheader from stratum data.
-void ( *stratum_gen_work )      ( struct stratum_ctx*, struct work* );
-
 // Get thread local copy of blockheader with unique nonce.
 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* );
@@ -166,7 +163,9 @@ bool ( *do_this_thread )        ( int );
 // After do_this_thread
 void ( *resync_threads )        ( struct work* );
 
+// No longer needed
 json_t* (*longpoll_rpc_call)      ( CURL*, int*, char* );
+
 set_t optimizations;
 int  ( *get_work_data_size )     ();
 int  ntime_index;
@@ -204,31 +203,74 @@ 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    ();
-//void null_hash_suw();
 
 // optional safe targets, default listed first unless noted.
 
 void std_get_new_work( struct work *work, struct work *g_work, int thr_id,
                        uint32_t* end_nonce_ptr );
 
-void std_stratum_gen_work( struct stratum_ctx *sctx, struct work *work );
-
 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 );
@@ -251,10 +293,6 @@ void std_build_block_header( struct work* g_work, uint32_t version,
 void std_build_extraheader( struct work *work, struct stratum_ctx *sctx );
 
 json_t* std_longpoll_rpc_call( CURL *curl, int *err, char *lp_url );
-//json_t* jr2_longpoll_rpc_call( CURL *curl, int *err );
-
-//bool std_stratum_handle_response( json_t *val );
-//bool jr2_stratum_handle_response( json_t *val );
 
 bool std_ready_to_mine( struct work* work, struct stratum_ctx* stratum,
                         int thr_id );
@@ -267,17 +305,12 @@ 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.
 bool register_algo( algo_gate_t *gate );
 
-// Overrides a common set of functions used by RPC2 and other RPC2-specific
-// init. Called by algo's register function before initializing algo-specific
-// functions and data.
-//bool register_json_rpc2( algo_gate_t *gate );
-
 // use this to call the hash function of an algo directly, ie util.c test.
 void exec_hash_function( int algo, void *output, const void *pdata );
 

algo/blake/blake-4way.c

@@ -48,7 +48,7 @@ int scanhash_blake_4way( struct work *work, uint32_t max_nonce,
       if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
       {
           pdata[19] = n+i;
-          submit_lane_solution( work, hash+(i<<3), mythr, i );
+          submit_solution( work, hash+(i<<3), mythr );
       }
       n += 4;
 
@@ -107,7 +107,7 @@ int scanhash_blake_8way( struct work *work, uint32_t max_nonce,
       if ( (hash+i)[7] <= HTarget && fulltest( hash+i, ptarget ) )
       {
           pdata[19] = n+i;
-          submit_lane_solution( work, hash+(i<<3), mythr, i );
+          submit_solution( work, hash+(i<<3), mythr );
       }
       n += 8;
 

algo/blake/blake2b-4way.c

@@ -45,7 +45,7 @@ int scanhash_blake2b_8way( struct work *work, uint32_t max_nonce,
           if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
           {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
           }
       }
       n += 8;
@@ -100,7 +100,7 @@ int scanhash_blake2b_4way( struct work *work, uint32_t max_nonce,
           if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
           {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
           }
       }
       n += 4;

algo/blake/blake2s-4way.c

@@ -49,7 +49,7 @@ int scanhash_blake2s_16way( struct work *work, uint32_t max_nonce,
          if ( likely( fulltest( lane_hash, ptarget ) && !opt_benchmark ) )
          {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
          }
       }
       n += 16;
@@ -104,7 +104,7 @@ int scanhash_blake2s_8way( struct work *work, uint32_t max_nonce,
          if ( likely( fulltest( lane_hash, ptarget ) && !opt_benchmark ) )
          {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
          }
       }
       n += 8;
@@ -157,7 +157,7 @@ int scanhash_blake2s_4way( struct work *work, uint32_t max_nonce,
          if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
          {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
               }
       }
       n += 4;

algo/blake/blakecoin-4way.c

@@ -49,7 +49,7 @@ int scanhash_blakecoin_4way( struct work *work, uint32_t max_nonce,
            && !opt_benchmark )
       {
           pdata[19] = n+i;
-          submit_lane_solution( work, hash+(i<<3), mythr, i );
+          submit_solution( work, hash+(i<<3), mythr );
       }
       n += 4;
 
@@ -108,7 +108,7 @@ int scanhash_blakecoin_8way( struct work *work, uint32_t max_nonce,
           && !opt_benchmark )
       {
           pdata[19] = n+i;
-          submit_lane_solution( work, hash+(i<<3), mythr, i );
+          submit_solution( work, hash+(i<<3), mythr );
       }
       n += 8;
    } while ( (n < max_nonce) && !work_restart[thr_id].restart );

algo/blake/decred-4way.c

@@ -62,7 +62,7 @@ int scanhash_decred_4way( struct work *work, uint32_t max_nonce,
       if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
       {
           pdata[DECRED_NONCE_INDEX] = n+i;
-          submit_lane_solution( work, hash+(i<<3), mythr, i );
+          submit_solution( work, hash+(i<<3), mythr );
       }
       n += 4;
   } while ( (n < max_nonce) && !work_restart[thr_id].restart );

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 );

algo/blake/pentablake-4way.c

@@ -105,7 +105,7 @@ int scanhash_pentablake_4way( struct work *work,
                   && fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
               {
                  pdata[19] = n + i;
-                 submit_lane_solution( work, hash+(i<<3), mythr, i );
+                 submit_solution( work, hash+(i<<3), mythr );
               }
               n += 4;
 

algo/bmw/bmw512-4way.c

@@ -46,7 +46,7 @@ int scanhash_bmw512_8way( struct work *work, uint32_t max_nonce,
           if ( fulltest( lane_hash, ptarget ) )
           {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
           }
       }
       n += 8;
@@ -99,7 +99,7 @@ int scanhash_bmw512_4way( struct work *work, uint32_t max_nonce,
           if ( fulltest( lane_hash, ptarget ) )
           {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
           }
       }
       n += 4;

algo/groestl/groestl-4way.c

@@ -53,7 +53,7 @@ int scanhash_groestl_4way( struct work *work, uint32_t max_nonce,
         if ( fulltest( hash+(lane<<3), ptarget) && !opt_benchmark )
         {
            pdata[19] = n + lane;
-           submit_lane_solution( work, hash+(lane<<3), mythr, lane );
+           submit_solution( work, hash+(lane<<3), mythr );
         }
         n += 4;
      } while ( ( n < last_nonce ) && !work_restart[thr_id].restart );

algo/groestl/myrgr-4way.c

@@ -143,7 +143,7 @@ int scanhash_myriad_8way( struct work *work, uint32_t max_nonce,
          if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
          {
             pdata[19] = n + lane;
-            submit_lane_solution( work, lane_hash, mythr, lane );
+            submit_solution( work, lane_hash, mythr );
          }
       }
       n += 8;
@@ -226,7 +226,7 @@ int scanhash_myriad_4way( struct work *work, uint32_t max_nonce,
          if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
          {
             pdata[19] = n + lane;
-            submit_lane_solution( work, lane_hash, mythr, lane );
+            submit_solution( work, lane_hash, mythr );
          }
       }
       n += 4;

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 )
@@ -155,11 +159,10 @@ 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 +174,7 @@ 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 );

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;

algo/jh/jha-4way.c

@@ -129,7 +129,7 @@ int scanhash_jha_4way( struct work *work, uint32_t max_nonce,
                  if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
                  {
                     pdata[19] = n+i;
-                    submit_lane_solution( work, lane_hash, mythr, i );
+                    submit_solution( work, lane_hash, mythr );
                  }
               }
               n += 4;

algo/keccak/keccak-4way.c

@@ -45,7 +45,7 @@ int scanhash_keccak_8way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) )
           {
               pdata[19] = bswap_32( n + lane );
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
           }
       }
       *noncev = _mm512_add_epi32( *noncev,
@@ -97,7 +97,7 @@ int scanhash_keccak_4way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ))
           {
               pdata[19] = bswap_32( n + lane );
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
           }
       }
       *noncev = _mm256_add_epi32( *noncev,

algo/keccak/sha3d-4way.c

@@ -52,7 +52,7 @@ int scanhash_sha3d_8way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) )
           {
               pdata[19] = bswap_32( n + lane );
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
           }
       }
       *noncev = _mm512_add_epi32( *noncev,
@@ -111,7 +111,7 @@ int scanhash_sha3d_4way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) )
           {
               pdata[19] = bswap_32( n + lane );
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
           }
       }
       *noncev = _mm256_add_epi32( *noncev,

algo/lyra2/allium-4way.c

@@ -245,7 +245,7 @@ int scanhash_allium_16way( struct work *work, uint32_t max_nonce,
      if ( unlikely( valid_hash( hash+(lane<<3), ptarget ) && !bench ) )
      {
          pdata[19] = bswap_32( n + lane );
-         submit_lane_solution( work, hash+(lane<<3), mythr, lane );
+         submit_solution( work, hash+(lane<<3), mythr );
      }
      *noncev = _mm512_add_epi32( *noncev, m512_const1_32( 16 ) );
      n += 16;
@@ -394,7 +394,7 @@ int scanhash_allium_8way( struct work *work, uint32_t max_nonce,
         if ( unlikely( valid_hash( lane_hash, ptarget ) && !bench ) )
         {
            pdata[19] = bswap_32( n + lane );
-           submit_lane_solution( work, lane_hash, mythr, lane );
+           submit_solution( work, lane_hash, mythr );
         }
      }
      n += 8;

algo/lyra2/lyra2-gate.c

@@ -215,9 +215,6 @@ 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 +222,6 @@ 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;

algo/lyra2/lyra2h-4way.c

@@ -76,7 +76,7 @@ int scanhash_lyra2h_4way( struct work *work, uint32_t max_nonce,
            && !opt_benchmark )
       {
           pdata[19] = n+i;         
-          submit_lane_solution( work, hash+(i<<3), mythr, i );
+          submit_solution( work, hash+(i<<3), mythr );
       }
       n += 4;
    } while (  (n < max_nonce-4) && !work_restart[thr_id].restart);

algo/lyra2/lyra2rev2-4way.c

@@ -200,7 +200,7 @@ int scanhash_lyra2rev2_16way( struct work *work, const uint32_t max_nonce,
          if ( likely( valid_hash( lane_hash, ptarget ) && !bench ) )
          {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
          }
       }
       *noncev = _mm512_add_epi32( *noncev, m512_const1_32( 16 ) );
@@ -342,7 +342,7 @@ int scanhash_lyra2rev2_8way( struct work *work, const uint32_t max_nonce,
          if ( likely( valid_hash( lane_hash, ptarget ) && !bench ) )
          {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
          }
       }
       *noncev = _mm256_add_epi32( *noncev, m256_const1_32( 8 ) );
@@ -469,7 +469,7 @@ int scanhash_lyra2rev2_4way( struct work *work, uint32_t max_nonce,
          if ( valid_hash( lane_hash, ptarget ) && !opt_benchmark )
          {
             pdata[19] = n + lane;         
-            submit_lane_solution( work, lane_hash, mythr, lane );
+            submit_solution( work, lane_hash, mythr );
          }
       }
       n += 4;

algo/lyra2/lyra2rev3-4way.c

@@ -165,7 +165,7 @@ int scanhash_lyra2rev3_16way( struct work *work, const uint32_t max_nonce,
          if ( likely( valid_hash( lane_hash, ptarget ) && !opt_benchmark ) )
          {
              pdata[19] = n + lane;
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
          }
       }
       n += 16;
@@ -284,7 +284,7 @@ int scanhash_lyra2rev3_8way( struct work *work, const uint32_t max_nonce,
          if ( likely( valid_hash( lane_hash, ptarget ) && !bench ) )
          {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
          }
       }
       *noncev = _mm256_add_epi32( *noncev, m256_const1_32( 8 ) );
@@ -386,7 +386,7 @@ int scanhash_lyra2rev3_4way( struct work *work, const uint32_t max_nonce,
          if ( valid_hash( lane_hash, ptarget ) && !opt_benchmark ) 
          {
               pdata[19] = bswap_32( n + lane );    
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
 	      }
       }
       *noncev = _mm_add_epi32( *noncev, m128_const1_32( 4 ) );

algo/lyra2/lyra2z-4way.c

@@ -124,7 +124,7 @@ int scanhash_lyra2z_16way( struct work *work, uint32_t max_nonce,
         if ( unlikely( valid_hash( lane_hash, ptarget ) && !bench ) )
         {
            pdata[19] = bswap_32( n + lane );
-           submit_lane_solution( work, lane_hash, mythr, lane );
+           submit_solution( work, lane_hash, mythr );
         }
       }
       *noncev = _mm512_add_epi32( *noncev, m512_const1_32( 16 ) );
@@ -222,7 +222,7 @@ int scanhash_lyra2z_8way( struct work *work, uint32_t max_nonce,
         if ( unlikely( valid_hash( lane_hash, ptarget ) && !bench ) )
         {
            pdata[19] = bswap_32( n + lane );
-           submit_lane_solution( work, lane_hash, mythr, lane );
+           submit_solution( work, lane_hash, mythr );
         }
       }
       *noncev = _mm256_add_epi32( *noncev, m256_const1_32( 8 ) );
@@ -301,7 +301,7 @@ int scanhash_lyra2z_4way( struct work *work, uint32_t max_nonce,
         if ( unlikely( valid_hash( lane_hash, ptarget ) && !bench ) )
         {
            pdata[19] = bswap_32( n + lane );
-           submit_lane_solution( work, lane_hash, mythr, lane );
+           submit_solution( work, lane_hash, mythr );
         }
       }
       *noncev = _mm_add_epi32( *noncev, m128_const1_32( 4 ) );

algo/lyra2/lyra2z330.c

@@ -68,7 +68,7 @@ bool lyra2z330_thread_init()
 
 bool register_lyra2z330_algo( algo_gate_t* gate )
 {
-  gate->optimizations = SSE42_OPT | AVX2_OPT;
+  gate->optimizations = SSE2_OPT | AVX2_OPT;
   gate->miner_thread_init = (void*)&lyra2z330_thread_init;
   gate->scanhash   = (void*)&scanhash_lyra2z330;
   gate->hash       = (void*)&lyra2z330_hash;

algo/lyra2/phi2-4way.c

@@ -302,7 +302,7 @@ int scanhash_phi2_8way( struct work *work, uint32_t max_nonce,
          if ( valid_hash( lane_hash, ptarget ) )
          {
             be32enc( pdata + 19, n + lane );
-            submit_lane_solution( work, lane_hash, mythr, lane );
+            submit_solution( work, lane_hash, mythr );
          }
       }
       n += 8;
@@ -483,7 +483,7 @@ int scanhash_phi2_4way( struct work *work, uint32_t max_nonce,
          if ( valid_hash( lane_hash, ptarget ) )
          {
             be32enc( pdata + 19, n + lane );
-            submit_lane_solution( work, lane_hash, mythr, lane );
+            submit_solution( work, lane_hash, mythr );
          }
       }
       edata[        19 ] += 4;

algo/nist5/nist5-4way.c

@@ -108,7 +108,7 @@ int scanhash_nist5_8way( struct work *work, uint32_t max_nonce,
            if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
            {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
            }
         }
         n += 8;
@@ -196,7 +196,7 @@ int scanhash_nist5_4way( struct work *work, uint32_t max_nonce,
            if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
            {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
            }
         }
         n += 4;

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 )

algo/quark/anime-4way.c

@@ -223,7 +223,7 @@ int scanhash_anime_8way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) )
           {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
           }
        }
        *noncev = _mm512_add_epi32( *noncev,
@@ -383,7 +383,7 @@ int scanhash_anime_4way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) )
           {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
           }
        }
        *noncev = _mm256_add_epi32( *noncev,

algo/quark/hmq1725-4way.c

@@ -596,7 +596,7 @@ int scanhash_hmq1725_8way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) )
           {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
           }
        }
        *noncev = _mm512_add_epi32( *noncev,
@@ -1018,7 +1018,7 @@ int scanhash_hmq1725_4way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) )
           {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
           }
        }
        *noncev = _mm256_add_epi32( *noncev,

algo/quark/quark-4way.c

@@ -235,7 +235,7 @@ int scanhash_quark_8way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) )
           {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
           }
        }
        *noncev = _mm512_add_epi32( *noncev,
@@ -408,7 +408,7 @@ int scanhash_quark_4way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) )
           {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
           }
        }
        *noncev = _mm256_add_epi32( *noncev,

algo/qubit/deep-2way.c

@@ -106,13 +106,13 @@ int scanhash_deep_2way( struct work *work,uint32_t max_nonce,
             if ( fulltest( hash, ptarget) && !opt_benchmark )
             {
                 pdata[19] = n;
-                submit_lane_solution( work, hash, mythr, 0 );
+                submit_solution( work, hash, mythr );
             }
             if ( !( (hash+8)[7] & mask ) )
             if ( fulltest( hash+8, ptarget) && !opt_benchmark )
             {
                pdata[19] = n+1;
-               submit_lane_solution( work, hash+8, mythr, 1 );
+               submit_solution( work, hash+8, mythr );
             }
             n += 2;
          } while ( ( n < max_nonce ) && !work_restart[thr_id].restart );

algo/qubit/qubit-2way.c

@@ -153,7 +153,7 @@ int scanhash_qubit_4way( struct work *work,uint32_t max_nonce,
         if ( likely( fulltest( hash+(lane<<3), ptarget) && !opt_benchmark ) )
         {
            pdata[19] = n + lane;
-           submit_lane_solution( work, hash+(lane<<3), mythr, lane );
+           submit_solution( work, hash+(lane<<3), mythr );
         }
         n += 4;
      } while ( ( n < max_nonce-4 ) && !work_restart[thr_id].restart );
@@ -255,13 +255,13 @@ int scanhash_qubit_2way( struct work *work,uint32_t max_nonce,
          if ( likely( fulltest( hash, ptarget) && !opt_benchmark ) )
          {
             pdata[19] = n;
-            submit_lane_solution( work, hash, mythr, 0 );
+            submit_solution( work, hash, mythr );
          }
          if ( unlikely( ( (hash+8))[7] <= Htarg ) )
          if ( likely( fulltest( hash+8, ptarget) && !opt_benchmark ) )
          {
             pdata[19] = n+1;
-            submit_lane_solution( work, hash+8, mythr, 1 );
+            submit_solution( work, hash+8, mythr );
          }
          n += 2;
      } while ( ( n < max_nonce ) && !work_restart[thr_id].restart );

algo/ripemd/lbry-4way.c

@@ -132,7 +132,7 @@ int scanhash_lbry_16way( struct work *work, uint32_t max_nonce,
          if ( likely( fulltest( lane_hash, ptarget ) && !opt_benchmark ) )
          {
             pdata[27] = n + i;
-            submit_lane_solution( work, lane_hash, mythr, i );
+            submit_solution( work, lane_hash, mythr );
          }
       }
       n += 16;
@@ -251,7 +251,7 @@ int scanhash_lbry_8way( struct work *work, uint32_t max_nonce,
          if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
          {
             pdata[27] = n + i;
-            submit_lane_solution( work, lane_hash, mythr, i );
+            submit_solution( work, lane_hash, mythr );
          }
       }
       n += 8;

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 );

algo/sha/hmac-sha256-hash-4way.c

@@ -0,0 +1,440 @@
+/*-
+ * Copyright 2005,2007,2009 Colin Percival
+ * Copywright 2020 JayDDee246@gmail.com
+ * 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.
+ */
+
+#include <sys/types.h>
+#include <stdint.h>
+#include <string.h>
+#include "hmac-sha256-hash-4way.h"
+#include "compat.h"
+
+// HMAC 4-way SSE2
+
+/**
+ * HMAC_SHA256_Buf(K, Klen, in, len, digest):
+ * Compute the HMAC-SHA256 of ${len} bytes from ${in} using the key ${K} of
+ * length ${Klen}, and write the result to ${digest}.
+ */
+void
+hmac_sha256_4way_full( void *digest, const void *K, size_t Klen,
+                       const void *in, size_t len )
+{
+   hmac_sha256_4way_context ctx;
+   hmac_sha256_4way_init( &ctx, K, Klen );
+   hmac_sha256_4way_update( &ctx, in, len );
+   hmac_sha256_4way_close( &ctx, digest );
+}
+
+/* Initialize an HMAC-SHA256 operation with the given key. */
+void
+hmac_sha256_4way_init( hmac_sha256_4way_context *ctx, const void *_K,
+                       size_t Klen )
+{
+	unsigned char pad[64*4] __attribute__ ((aligned (64)));
+	unsigned char khash[32*4] __attribute__ ((aligned (64)));
+	const unsigned char * K = _K;
+	size_t i;
+
+	/* If Klen > 64, the key is really SHA256(K). */
+	if ( Klen > 64 )
+   {
+		sha256_4way_init( &ctx->ictx );
+		sha256_4way_update( &ctx->ictx, K, Klen );
+		sha256_4way_close( &ctx->ictx, khash );
+		K = khash;
+		Klen = 32;
+	}
+
+	/* Inner SHA256 operation is SHA256(K xor [block of 0x36] || data). */
+   sha256_4way_init( &ctx->ictx );
+	memset( pad, 0x36, 64*4 );
+
+   for ( i = 0; i < Klen; i++ )
+		casti_m128i( pad, i ) = _mm_xor_si128( casti_m128i( pad, i ),
+                                             casti_m128i( K, i ) );
+
+   sha256_4way_update( &ctx->ictx, pad, 64 );
+
+	/* Outer SHA256 operation is SHA256(K xor [block of 0x5c] || hash). */
+	sha256_4way_init( &ctx->octx );
+	memset( pad, 0x5c, 64*4 );
+	for ( i = 0; i < Klen/4; i++ )
+		casti_m128i( pad, i ) = _mm_xor_si128( casti_m128i( pad, i ),
+                                             casti_m128i( K, i ) );
+	sha256_4way_update( &ctx->octx, pad, 64 );
+}
+
+/* Add bytes to the HMAC-SHA256 operation. */
+void
+hmac_sha256_4way_update( hmac_sha256_4way_context *ctx, const void *in,
+                         size_t len )
+{
+	/* Feed data to the inner SHA256 operation. */
+	sha256_4way_update( &ctx->ictx, in, len );
+}
+
+/* Finish an HMAC-SHA256 operation. */
+void
+hmac_sha256_4way_close( hmac_sha256_4way_context *ctx, void *digest )
+{
+	unsigned char ihash[32*4] __attribute__ ((aligned (64)));
+
+	/* Finish the inner SHA256 operation. */
+	sha256_4way_close( &ctx->ictx, ihash );
+
+	/* Feed the inner hash to the outer SHA256 operation. */
+	sha256_4way_update( &ctx->octx, ihash, 32 );
+
+	/* Finish the outer SHA256 operation. */
+	sha256_4way_close( &ctx->octx, digest );
+}
+
+/**
+ * PBKDF2_SHA256(passwd, passwdlen, salt, saltlen, c, buf, dkLen):
+ * Compute PBKDF2(passwd, salt, c, dkLen) using HMAC-SHA256 as the PRF, and
+ * write the output to buf.  The value dkLen must be at most 32 * (2^32 - 1).
+ */
+void
+pbkdf2_sha256_4way( uint8_t *buf, size_t dkLen,
+                    const uint8_t *passwd, size_t passwdlen,
+                    const uint8_t *salt, size_t saltlen, uint64_t c )
+{
+	hmac_sha256_4way_context PShctx, hctx;
+	uint8_t _ALIGN(128) T[32*4];
+	uint8_t _ALIGN(128) U[32*4];
+   __m128i ivec;
+   size_t i, clen;
+	uint64_t j;
+	int k;
+
+	/* Compute HMAC state after processing P and S. */
+	hmac_sha256_4way_init( &PShctx, passwd, passwdlen );
+	hmac_sha256_4way_update( &PShctx, salt, saltlen );
+
+	/* Iterate through the blocks. */
+	for ( i = 0; i * 32 < dkLen; i++ )
+   {
+		/* Generate INT(i + 1). */
+      ivec = _mm_set1_epi32( bswap_32( i+1 ) ); 
+
+		/* Compute U_1 = PRF(P, S || INT(i)). */
+		memcpy( &hctx, &PShctx, sizeof(hmac_sha256_4way_context) );
+		hmac_sha256_4way_update( &hctx, &ivec, 4 );
+		hmac_sha256_4way_close( &hctx, U );
+
+		/* T_i = U_1 ... */
+		memcpy( T, U, 32*4 );
+
+		for ( j = 2; j <= c; j++ )
+      {
+			/* Compute U_j. */
+			hmac_sha256_4way_init( &hctx, passwd, passwdlen );
+			hmac_sha256_4way_update( &hctx, U, 32 );
+			hmac_sha256_4way_close( &hctx, U );
+
+			/* ... xor U_j ... */
+			for ( k = 0; k < 8; k++ )
+				casti_m128i( T, k ) = _mm_xor_si128( casti_m128i( T, k ),
+                                                 casti_m128i( U, k ) );
+		}
+
+		/* Copy as many bytes as necessary into buf. */
+		clen = dkLen - i * 32;
+		if ( clen > 32 )
+			clen = 32;
+		memcpy( &buf[ i*32*4 ], T, clen*4 );
+	}
+}
+
+#if defined(__AVX2__)
+
+// HMAC 8-way AVX2
+
+void
+hmac_sha256_8way_full( void *digest, const void *K, size_t Klen,
+                       const void *in, size_t len )
+{
+   hmac_sha256_8way_context ctx;
+   hmac_sha256_8way_init( &ctx, K, Klen );
+   hmac_sha256_8way_update( &ctx, in, len );
+   hmac_sha256_8way_close( &ctx, digest );
+}
+
+/* Initialize an HMAC-SHA256 operation with the given key. */
+void
+hmac_sha256_8way_init( hmac_sha256_8way_context *ctx, const void *_K,
+                       size_t Klen )
+{
+   unsigned char pad[64*8] __attribute__ ((aligned (128)));
+   unsigned char khash[32*8] __attribute__ ((aligned (128)));
+   const unsigned char * K = _K;
+   size_t i;
+
+   /* If Klen > 64, the key is really SHA256(K). */
+   if ( Klen > 64 )
+   {
+      sha256_8way_init( &ctx->ictx );
+      sha256_8way_update( &ctx->ictx, K, Klen );
+      sha256_8way_close( &ctx->ictx, khash );
+      K = khash;
+      Klen = 32;
+   }
+
+   /* Inner SHA256 operation is SHA256(K xor [block of 0x36] || data). */
+   sha256_8way_init( &ctx->ictx );
+   memset( pad, 0x36, 64*8);
+
+   for ( i = 0; i < Klen/4; i++ )
+      casti_m256i( pad, i ) = _mm256_xor_si256( casti_m256i( pad, i ),
+                                                casti_m256i( K, i ) );
+
+   sha256_8way_update( &ctx->ictx, pad, 64 );
+
+   /* Outer SHA256 operation is SHA256(K xor [block of 0x5c] || hash). */
+   sha256_8way_init( &ctx->octx );
+   memset( pad, 0x5c, 64*8 );
+   for ( i = 0; i < Klen/4; i++ )
+      casti_m256i( pad, i ) = _mm256_xor_si256( casti_m256i( pad, i ),
+                                                casti_m256i( K, i ) );
+   sha256_8way_update( &ctx->octx, pad, 64 );
+}
+
+void
+hmac_sha256_8way_update( hmac_sha256_8way_context *ctx, const void *in,
+                         size_t len )
+{
+   /* Feed data to the inner SHA256 operation. */
+   sha256_8way_update( &ctx->ictx, in, len );
+}
+
+/* Finish an HMAC-SHA256 operation. */
+void
+hmac_sha256_8way_close( hmac_sha256_8way_context *ctx, void *digest )
+{
+   unsigned char ihash[32*8] __attribute__ ((aligned (128)));
+
+   /* Finish the inner SHA256 operation. */
+   sha256_8way_close( &ctx->ictx, ihash );
+
+   /* Feed the inner hash to the outer SHA256 operation. */
+   sha256_8way_update( &ctx->octx, ihash, 32 );
+
+   /* Finish the outer SHA256 operation. */
+   sha256_8way_close( &ctx->octx, digest );
+}
+
+/**
+ * PBKDF2_SHA256(passwd, passwdlen, salt, saltlen, c, buf, dkLen):
+ * Compute PBKDF2(passwd, salt, c, dkLen) using HMAC-SHA256 as the PRF, and
+ * write the output to buf.  The value dkLen must be at most 32 * (2^32 - 1).
+ */
+void
+pbkdf2_sha256_8way( uint8_t *buf, size_t dkLen, const uint8_t *passwd,
+                    size_t passwdlen, const uint8_t *salt, size_t saltlen,
+                    uint64_t c )
+{
+   hmac_sha256_8way_context PShctx, hctx;
+   uint8_t _ALIGN(128) T[32*8];
+   uint8_t _ALIGN(128) U[32*8];
+   size_t i, clen;
+   uint64_t j;
+   int k;
+
+   /* Compute HMAC state after processing P and S. */
+   hmac_sha256_8way_init( &PShctx, passwd, passwdlen );
+
+// saltlen can be odd number of bytes
+   hmac_sha256_8way_update( &PShctx, salt, saltlen );
+
+   /* Iterate through the blocks. */
+   for ( i = 0; i * 32 < dkLen; i++ )
+   {
+      __m256i ivec = _mm256_set1_epi32( bswap_32( i+1 ) );
+
+      /* Compute U_1 = PRF(P, S || INT(i)). */
+      memcpy( &hctx, &PShctx, sizeof(hmac_sha256_8way_context) );
+      hmac_sha256_8way_update( &hctx, &ivec, 4 );
+      hmac_sha256_8way_close( &hctx, U );
+
+      /* T_i = U_1 ... */
+      memcpy( T, U, 32*8 );
+
+      for ( j = 2; j <= c; j++ )
+      {
+         /* Compute U_j. */
+         hmac_sha256_8way_init( &hctx, passwd, passwdlen );
+         hmac_sha256_8way_update( &hctx, U, 32 );
+         hmac_sha256_8way_close( &hctx, U );
+
+         /* ... xor U_j ... */
+         for ( k = 0; k < 8; k++ )
+            casti_m256i( T, k ) = _mm256_xor_si256( casti_m256i( T, k ),
+                                                    casti_m256i( U, k ) );
+      }
+
+      /* Copy as many bytes as necessary into buf. */
+      clen = dkLen - i * 32;
+      if ( clen > 32 )
+         clen = 32;
+      memcpy( &buf[ i*32*8 ], T, clen*8 );
+   }
+}
+
+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+
+// HMAC 16-way AVX512
+
+void
+hmac_sha256_16way_full( void *digest, const void *K, size_t Klen,
+                        const void *in, size_t len )
+{
+   hmac_sha256_16way_context ctx;
+   hmac_sha256_16way_init( &ctx, K, Klen );
+   hmac_sha256_16way_update( &ctx, in, len );
+   hmac_sha256_16way_close( &ctx, digest );
+}
+
+void
+hmac_sha256_16way_init( hmac_sha256_16way_context *ctx, const void *_K,
+                       size_t Klen )
+{
+   unsigned char pad[64*16] __attribute__ ((aligned (128)));
+   unsigned char khash[32*16] __attribute__ ((aligned (128)));
+   const unsigned char * K = _K;
+   size_t i;
+
+   /* If Klen > 64, the key is really SHA256(K). */
+   if ( Klen > 64 )
+   {
+      sha256_16way_init( &ctx->ictx );
+      sha256_16way_update( &ctx->ictx, K, Klen );
+      sha256_16way_close( &ctx->ictx, khash );
+      K = khash;
+      Klen = 32;
+   }
+
+   /* Inner SHA256 operation is SHA256(K xor [block of 0x36] || data). */
+   sha256_16way_init( &ctx->ictx );
+   memset( pad, 0x36, 64*16 );
+
+   for ( i = 0; i < Klen; i++ )
+      casti_m512i( pad, i ) = _mm512_xor_si512( casti_m512i( pad, i ),
+                                                casti_m512i( K, i ) );
+   sha256_16way_update( &ctx->ictx, pad, 64 );
+
+   /* Outer SHA256 operation is SHA256(K xor [block of 0x5c] || hash). */
+   sha256_16way_init( &ctx->octx );
+   memset( pad, 0x5c, 64*16 );
+   for ( i = 0; i < Klen/4; i++ )
+      casti_m512i( pad, i ) = _mm512_xor_si512( casti_m512i( pad, i ),
+                                             casti_m512i( K, i ) );
+   sha256_16way_update( &ctx->octx, pad, 64 );
+}
+   
+void
+hmac_sha256_16way_update( hmac_sha256_16way_context *ctx, const void *in,
+                         size_t len )
+{
+   /* Feed data to the inner SHA256 operation. */
+   sha256_16way_update( &ctx->ictx, in, len );
+}
+
+/* Finish an HMAC-SHA256 operation. */
+void
+hmac_sha256_16way_close( hmac_sha256_16way_context *ctx, void *digest )
+{
+   unsigned char ihash[32*16] __attribute__ ((aligned (128)));
+
+   /* Finish the inner SHA256 operation. */
+   sha256_16way_close( &ctx->ictx, ihash );
+
+   /* Feed the inner hash to the outer SHA256 operation. */
+   sha256_16way_update( &ctx->octx, ihash, 32 );
+
+   /* Finish the outer SHA256 operation. */
+   sha256_16way_close( &ctx->octx, digest );
+}
+
+/**
+ * PBKDF2_SHA256(passwd, passwdlen, salt, saltlen, c, buf, dkLen):
+ * Compute PBKDF2(passwd, salt, c, dkLen) using HMAC-SHA256 as the PRF, and
+ * write the output to buf.  The value dkLen must be at most 32 * (2^32 - 1).
+ */
+void
+pbkdf2_sha256_16way( uint8_t *buf, size_t dkLen,
+                     const uint8_t *passwd, size_t passwdlen,
+                     const uint8_t *salt, size_t saltlen, uint64_t c )
+{
+   hmac_sha256_16way_context PShctx, hctx;
+   uint8_t _ALIGN(128) T[32*16];
+   uint8_t _ALIGN(128) U[32*16];
+   __m512i ivec;
+   size_t i, clen;
+   uint64_t j;
+   int k;
+
+   /* Compute HMAC state after processing P and S. */
+   hmac_sha256_16way_init( &PShctx, passwd, passwdlen );
+   hmac_sha256_16way_update( &PShctx, salt, saltlen );
+
+   /* Iterate through the blocks. */
+   for ( i = 0; i * 32 < dkLen; i++ )
+   {
+      /* Generate INT(i + 1). */
+      ivec = _mm512_set1_epi32( bswap_32( i+1 ) );
+
+      /* Compute U_1 = PRF(P, S || INT(i)). */
+      memcpy( &hctx, &PShctx, sizeof(hmac_sha256_16way_context) );
+      hmac_sha256_16way_update( &hctx, &ivec, 4 );
+      hmac_sha256_16way_close( &hctx, U );
+
+      /* T_i = U_1 ... */
+      memcpy( T, U, 32*16 );
+
+      for ( j = 2; j <= c; j++ )
+      {
+         /* Compute U_j. */
+         hmac_sha256_16way_init( &hctx, passwd, passwdlen );
+         hmac_sha256_16way_update( &hctx, U, 32 );
+         hmac_sha256_16way_close( &hctx, U );
+
+         /* ... xor U_j ... */
+         for ( k = 0; k < 8; k++ )
+            casti_m512i( T, k ) = _mm512_xor_si512( casti_m512i( T, k ),
+                                                    casti_m512i( U, k ) );
+      }
+
+      /* Copy as many bytes as necessary into buf. */
+      clen = dkLen - i * 32;
+      if ( clen > 32 )
+         clen = 32;
+      memcpy( &buf[ i*32*16 ], T, clen*16 );
+   }
+}
+
+#endif  // AVX512
+#endif  // AVX2
+

algo/sha/hmac-sha256-hash-4way.h

@@ -0,0 +1,107 @@
+/*-
+ * Copyright 2005,2007,2009 Colin Percival
+ * Copyright 2020 JayDDee@gmailcom
+ * 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/lib/libmd/sha256_Y.h,v 1.2 2006/01/17 15:35:56 phk Exp $
+ */
+
+#ifndef HMAC_SHA256_4WAY_H__
+#define HMAC_SHA256_4WAY_H__
+
+
+// Tested only 8-way with null pers
+
+#include <sys/types.h>
+#include <stdint.h>
+#include "simd-utils.h"
+#include "sha-hash-4way.h"
+
+typedef struct _hmac_sha256_4way_context
+{
+   sha256_4way_context ictx;
+   sha256_4way_context octx;
+} hmac_sha256_4way_context;
+
+//void SHA256_Buf( const void *, size_t len, uint8_t digest[32] );
+void hmac_sha256_4way_init( hmac_sha256_4way_context *, const void *, size_t );
+void hmac_sha256_4way_update( hmac_sha256_4way_context *, const void *,
+                              size_t );
+void hmac_sha256_4way_close( hmac_sha256_4way_context *, void* );
+void hmac_sha256_4way_full( void*, const void *, size_t Klen, const void *,
+                            size_t len );
+
+/**
+ * PBKDF2_SHA256(passwd, passwdlen, salt, saltlen, c, buf, dkLen):
+ * Compute PBKDF2(passwd, salt, c, dkLen) using HMAC-SHA256 as the PRF, and
+ * write the output to buf.  The value dkLen must be at most 32 * (2^32 - 1).
+ */
+void pbkdf2_sha256_4way( uint8_t *, size_t, const uint8_t *, size_t,
+                         const uint8_t *, size_t, uint64_t );
+
+#if defined(__AVX2__)
+
+typedef struct _hmac_sha256_8way_context
+{
+   sha256_8way_context ictx;
+   sha256_8way_context octx;
+} hmac_sha256_8way_context;
+
+//void SHA256_Buf( const void *, size_t len, uint8_t digest[32] );
+void hmac_sha256_8way_init( hmac_sha256_8way_context *, const void *, size_t );
+void hmac_sha256_8way_update( hmac_sha256_8way_context *, const void *,
+                              size_t );
+void hmac_sha256_8way_close( hmac_sha256_8way_context *, void* );
+void hmac_sha256_8way_full( void*, const void *, size_t Klen, const void *,
+                            size_t len );
+
+void pbkdf2_sha256_8way( uint8_t *, size_t, const uint8_t *, size_t,
+                        const uint8_t *, size_t, uint64_t );
+      
+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+
+typedef struct _hmac_sha256_16way_context
+{
+   sha256_16way_context ictx;
+   sha256_16way_context octx;
+} hmac_sha256_16way_context;
+
+//void SHA256_Buf( const void *, size_t len, uint8_t digest[32] );
+void hmac_sha256_16way_init( hmac_sha256_16way_context *,
+                             const void *, size_t );
+void hmac_sha256_16way_update( hmac_sha256_16way_context *, const void *,
+                              size_t );
+void hmac_sha256_16way_close( hmac_sha256_16way_context *, void* );
+void hmac_sha256_16way_full( void*, const void *, size_t Klen, const void *,
+                             size_t len );
+
+void pbkdf2_sha256_16way( uint8_t *, size_t, const uint8_t *, size_t,
+                          const uint8_t *, size_t, uint64_t );
+
+
+
+#endif   // AVX512
+#endif   // AVX2
+
+#endif // HMAC_SHA256_4WAY_H__

algo/sha/hmac-sha256-hash.c

@@ -81,16 +81,17 @@ HMAC_SHA256_Init( HMAC_SHA256_CTX *ctx, const void *_K, size_t Klen )
 
 	/* Inner SHA256 operation is SHA256(K xor [block of 0x36] || data). */
    SHA256_Init( &ctx->ictx );
-	memset( pad, 0x36, 64 );
-	for ( i = 0; i < Klen; i++ )
-		pad[i] ^= K[i];
+
+
+   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, 0x5c, 64);
-	for ( i = 0; i < Klen; i++ )
-		pad[i] ^= K[i];
+
+   for ( i = 0; i < Klen; i++ )  pad[i] = K[i] ^ 0x5c;
+   memset( pad + Klen, 0x5c, 64 - Klen );
 	SHA256_Update( &ctx->octx, pad, 64 );
 }
 
@@ -161,7 +162,13 @@ PBKDF2_SHA256( const uint8_t *passwd, size_t passwdlen, const uint8_t *salt,
 			HMAC_SHA256_Final( U, &hctx );
 
 			/* ... xor U_j ... */
-			for ( k = 0; k < 32; k++ )
+//         _mm256_xor_si256( *(__m256i*)T, *(__m256i*)U );
+//         _mm_xor_si128( ((__m128i*)T)[0], ((__m128i*)U)[0] );
+//         _mm_xor_si128( ((__m128i*)T)[1], ((__m128i*)U)[1] );
+
+//         for ( k = 0; k < 4; k++ )  T[k] ^= U[k];
+         
+         for ( k = 0; k < 32; k++ )
 				T[k] ^= U[k];
 		}
 

algo/sha/sha-hash-4way.h

@@ -58,6 +58,7 @@ void sha256_4way_init( sha256_4way_context *sc );
 void sha256_4way_update( sha256_4way_context *sc, const void *data,
                          size_t len );
 void sha256_4way_close( sha256_4way_context *sc, void *dst );
+void sha256_4way_full( void *dst, const void *data, size_t len );
 
 #endif  // SSE2
 
@@ -75,6 +76,7 @@ typedef struct {
 void sha256_8way_init( sha256_8way_context *sc );
 void sha256_8way_update( sha256_8way_context *sc, const void *data, size_t len );
 void sha256_8way_close( sha256_8way_context *sc, void *dst );
+void sha256_8way_full( void *dst, const void *data, size_t len );
 
 #endif  // AVX2
 
@@ -92,6 +94,7 @@ typedef struct {
 void sha256_16way_init( sha256_16way_context *sc );
 void sha256_16way_update( sha256_16way_context *sc, const void *data, size_t len );
 void sha256_16way_close( sha256_16way_context *sc, void *dst );
+void sha256_16way_full( void *dst, const void *data, size_t len );
 
 #endif // AVX512
 
@@ -110,6 +113,7 @@ void sha512_4way_init( sha512_4way_context *sc);
 void sha512_4way_update( sha512_4way_context *sc, const void *data,
                          size_t len );
 void sha512_4way_close( sha512_4way_context *sc, void *dst );
+void sha512_4way_full( void *dst, const void *data, size_t len );
 
 #endif  // AVX2
 
@@ -128,6 +132,7 @@ void sha512_8way_init( sha512_8way_context *sc);
 void sha512_8way_update( sha512_8way_context *sc, const void *data, 
                          size_t len );
 void sha512_8way_close( sha512_8way_context *sc, void *dst );
+void sha512_8way_full( void *dst, const void *data, size_t len );
 
 #endif  // AVX512
 

algo/sha/sha2.c

@@ -479,8 +479,8 @@ 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 +492,7 @@ 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 +522,8 @@ 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 +540,8 @@ 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 +553,7 @@ 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 +583,8 @@ 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 +612,11 @@ 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 +653,7 @@ 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 );
 

algo/sha/sha256-hash-4way.c

@@ -330,6 +330,14 @@ void sha256_4way_close( sha256_4way_context *sc, void *dst )
     mm128_block_bswap_32( dst, sc->val );
 }
 
+void sha256_4way_full( void *dst, const void *data, size_t len )
+{
+   sha256_4way_context ctx;
+   sha256_4way_init( &ctx );
+   sha256_4way_update( &ctx, data, len );
+   sha256_4way_close( &ctx, dst );
+}
+
 #if defined(__AVX2__)
 
 // SHA-256 8 way
@@ -498,6 +506,10 @@ void sha256_8way_init( sha256_8way_context *sc )
 */
 }
 
+
+// need to handle odd byte length for yespower.
+// Assume only last update is odd.
+
 void sha256_8way_update( sha256_8way_context *sc, const void *data, size_t len )
 {
    __m256i *vdata = (__m256i*)data;
@@ -564,6 +576,13 @@ void sha256_8way_close( sha256_8way_context *sc, void *dst )
     mm256_block_bswap_32( dst, sc->val );
 }
 
+void sha256_8way_full( void *dst, const void *data, size_t len )
+{
+   sha256_8way_context ctx;
+   sha256_8way_init( &ctx );
+   sha256_8way_update( &ctx, data, len );
+   sha256_8way_close( &ctx, dst );
+}
 
 #if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
 
@@ -791,6 +810,14 @@ void sha256_16way_close( sha256_16way_context *sc, void *dst )
     mm512_block_bswap_32( dst, sc->val );
 }
 
+void sha256_16way_full( void *dst, const void *data, size_t len )
+{
+   sha256_16way_context ctx;
+   sha256_16way_init( &ctx );
+   sha256_16way_update( &ctx, data, len );
+   sha256_16way_close( &ctx, dst );
+}
+
 #endif  // AVX512
 #endif  // __AVX2__
 #endif  // __SSE2__

algo/sha/sha256q-4way.c

@@ -85,7 +85,7 @@ int scanhash_sha256q_8way( struct work *work, uint32_t max_nonce,
 	         if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
             {
 	           pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
             }
 	      }
          n += 8;
@@ -173,7 +173,7 @@ int scanhash_sha256q_4way( struct work *work, uint32_t max_nonce,
             if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
             {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
             }
          }
          n += 4;

algo/sha/sha256t-4way.c

@@ -78,7 +78,7 @@ int scanhash_sha256t_8way( struct work *work, const uint32_t max_nonce,
             if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
             {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
 	         }
          }
          n += 8;
@@ -161,7 +161,7 @@ int scanhash_sha256t_4way( struct work *work, const uint32_t max_nonce,
             if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
             {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
 	         }
          } 
          n += 4;

algo/skein/skein-4way.c

@@ -65,7 +65,7 @@ int scanhash_skein_8way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) )
           {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
           }
        }
        *noncev = _mm512_add_epi32( *noncev,
@@ -162,7 +162,7 @@ int scanhash_skein_4way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) )
           {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
           }
        }
        *noncev = _mm256_add_epi32( *noncev,

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];
@@ -1099,6 +1099,7 @@ skein512_4way_final16( skein512_4way_context *sc,  void *out, const void *data )
    casti_m256i( out, 7 ) = h7;
 }
 
+// Broken for 80 bytes, use prehash.
 void
 skein256_4way_update(void *cc, const void *data, size_t len)
 {

algo/skein/skein2-4way.c

@@ -53,7 +53,7 @@ int scanhash_skein2_8way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) && !bench )
           {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
           }
        }
        *noncev = _mm512_add_epi32( *noncev,
@@ -115,7 +115,7 @@ int scanhash_skein2_4way( struct work *work, uint32_t max_nonce,
           if ( valid_hash( lane_hash, ptarget ) && !bench )
           {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
           }
        }
        *noncev = _mm256_add_epi32( *noncev,

algo/x11/c11-4way.c

@@ -279,7 +279,7 @@ int scanhash_c11_8way( struct work *work, uint32_t max_nonce,
              && fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
         {
            pdata[19] = n+i;
-           submit_lane_solution( work, hash+(i<<3), mythr, i );
+           submit_solution( work, hash+(i<<3), mythr );
         }
         n += 8;
      } while ( ( n < max_nonce ) && !work_restart[thr_id].restart );
@@ -459,7 +459,7 @@ int scanhash_c11_4way( struct work *work, uint32_t max_nonce,
             && fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
         {
            pdata[19] = n+i;
-           submit_lane_solution( work, hash+(i<<3), mythr, i );
+           submit_solution( work, hash+(i<<3), mythr );
         }
         n += 4;
      } while ( ( n < max_nonce ) && !work_restart[thr_id].restart );

algo/x11/timetravel-4way.c

@@ -221,7 +221,7 @@ int scanhash_timetravel_4way( struct work *work, uint32_t max_nonce,
           && !opt_benchmark )
       {
           pdata[19] = n+i;
-          submit_lane_solution( work, hash+(i<<3), mythr, i );
+          submit_solution( work, hash+(i<<3), mythr );
       }
       n += 4;
    } while ( ( n < max_nonce ) && !(*restart) );

algo/x11/timetravel10-4way.c

@@ -256,7 +256,7 @@ int scanhash_timetravel10_4way( struct work *work,
             && !opt_benchmark )
       {
          pdata[19] = n+i;
-         submit_lane_solution( work, hash+(i<<3), mythr, i );
+         submit_solution( work, hash+(i<<3), mythr );
       }
       n += 4;
    } while ( ( n < max_nonce ) && !(*restart) );

algo/x11/tribus-4way.c

@@ -128,7 +128,7 @@ int scanhash_tribus_8way( struct work *work, uint32_t max_nonce,
      if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
      {
           pdata[19] = n+i;
-          submit_lane_solution( work, hash+(i<<3), mythr, i );
+          submit_solution( work, hash+(i<<3), mythr );
      }
      n += 8;
    } while ( ( n < max_nonce-8 )  && !work_restart[thr_id].restart);
@@ -213,7 +213,7 @@ int scanhash_tribus_4way( struct work *work, uint32_t max_nonce,
      if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
      {
           pdata[19] = n+i;
-          submit_lane_solution( work, hash+(i<<3), mythr, i );
+          submit_solution( work, hash+(i<<3), mythr );
      }
      n += 4;
    } while ( ( n < max_nonce-4 )  && !work_restart[thr_id].restart);

algo/x11/x11-4way.c

@@ -279,7 +279,7 @@ int scanhash_x11_8way( struct work *work, uint32_t max_nonce,
               && fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
          {
              pdata[19] = n+i;
-             submit_lane_solution( work, hash+(i<<3), mythr, i );
+             submit_solution( work, hash+(i<<3), mythr );
          }
          n += 8;
      } while ( ( n < last_nonce ) && !work_restart[thr_id].restart );
@@ -469,7 +469,7 @@ int scanhash_x11_4way( struct work *work, uint32_t max_nonce,
                  && fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
             {
                pdata[19] = n+i;
-               submit_lane_solution( work, hash+(i<<3), mythr, i );
+               submit_solution( work, hash+(i<<3), mythr );
             }
             n += 4;
          } while ( ( n < max_nonce ) && !work_restart[thr_id].restart );

algo/x11/x11evo-4way.c

@@ -269,7 +269,7 @@ int scanhash_x11evo_4way( struct work* work, uint32_t max_nonce,
                  && fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
          {
             pdata[19] = n+i;
-            submit_lane_solution( work, hash+(i<<3), mythr, i );
+            submit_solution( work, hash+(i<<3), mythr );
          }
          n += 4;
      } while ( ( n < max_nonce ) && !work_restart[thr_id].restart );

algo/x11/x11gost-4way.c

@@ -312,7 +312,7 @@ int scanhash_x11gost_8way( struct work *work, uint32_t max_nonce,
               && fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
          {
              pdata[19] = n+i;
-             submit_lane_solution( work, hash+(i<<3), mythr, i );
+             submit_solution( work, hash+(i<<3), mythr );
          }
          n += 8;
      } while ( ( n < max_nonce ) && !work_restart[thr_id].restart );
@@ -498,7 +498,7 @@ int scanhash_x11gost_4way( struct work *work, uint32_t max_nonce,
              && fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
         {
            pdata[19] = n+i;
-           submit_lane_solution( work, hash+(i<<3), mythr, i );
+           submit_solution( work, hash+(i<<3), mythr );
         }
         n += 4;
      } while ( ( n < max_nonce ) && !work_restart[thr_id].restart );

algo/x12/x12-4way.c

@@ -170,6 +170,9 @@ void x12_8way_hash( void *state, const void *input )
      dintrlv_4x128_512( hash0, hash1, hash2, hash3, vhashA );
      dintrlv_4x128_512( hash4, hash5, hash6, hash7, vhashB );
 
+     update_final_echo( &ctx.echo, (BitSequence *)hash0,
+                       (const BitSequence *) hash0, 512 );
+     memcpy( &ctx.echo, &x12_8way_ctx.echo, sizeof(hashState_echo) );
      update_final_echo( &ctx.echo, (BitSequence *)hash1,
                        (const BitSequence *) hash1, 512 );
      memcpy( &ctx.echo, &x12_8way_ctx.echo, sizeof(hashState_echo) );
@@ -263,7 +266,7 @@ int scanhash_x12_8way( struct work *work, uint32_t max_nonce,
            if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
            {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
            }
         }
         n += 8;
@@ -363,6 +366,18 @@ void x12_4way_hash( void *state, const void *input )
      simd_2way_update_close( &ctx.simd, vhash, vhash, 512 );
      dintrlv_2x128( hash2, hash3, vhash, 512 );
 
+     update_final_echo( &ctx.echo, (BitSequence *)hash0,
+                       (const BitSequence *) hash0, 512 );
+     memcpy( &ctx.echo, &x12_4way_ctx.echo, sizeof(hashState_echo) );
+     update_final_echo( &ctx.echo, (BitSequence *)hash1,
+                       (const BitSequence *) hash1, 512 );
+     memcpy( &ctx.echo, &x12_4way_ctx.echo, sizeof(hashState_echo) );
+     update_final_echo( &ctx.echo, (BitSequence *)hash2,
+                       (const BitSequence *) hash2, 512 );
+     memcpy( &ctx.echo, &x12_4way_ctx.echo, sizeof(hashState_echo) );
+     update_final_echo( &ctx.echo, (BitSequence *)hash3,
+                       (const BitSequence *) hash3, 512 );
+     
      update_and_final_groestl( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
      memcpy( &ctx.groestl, &x12_4way_ctx.groestl, sizeof(hashState_groestl) );
      update_and_final_groestl( &ctx.groestl, (char*)hash1, (char*)hash1, 512 );
@@ -431,7 +446,7 @@ int scanhash_x12_4way( struct work *work, uint32_t max_nonce,
             if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
             {
                pdata[19] = n+i;
-               submit_lane_solution( work, hash+(i<<3), mythr, i );
+               submit_solution( work, hash+(i<<3), mythr );
             }
             n += 4;
          } while ( ( n < max_nonce ) && !work_restart[thr_id].restart );

algo/x13/phi1612-4way.c

@@ -208,7 +208,7 @@ int scanhash_phi1612_8way( struct work *work, uint32_t max_nonce,
         if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
         {
            pdata[19] = n+i;
-           submit_lane_solution( work, hash+(i<<3), mythr, i );
+           submit_solution( work, hash+(i<<3), mythr );
         }
         n += 8;
      } while ( ( n < max_nonce-8 ) && !work_restart[thr_id].restart );
@@ -251,8 +251,12 @@ void phi1612_4way_hash( void *state, const void *input )
      memcpy( &ctx, &phi1612_4way_ctx, sizeof(phi1612_4way_ctx) );
 
      // Skein parallel 4way
-     skein512_4way_update( &ctx.skein, input, 80 );
-     skein512_4way_close( &ctx.skein, vhash );
+
+// skein 4way is broken for 80 bytes
+//     skein512_4way_update( &ctx.skein, input, 80 );
+//     skein512_4way_close( &ctx.skein, vhash );
+     skein512_4way_prehash64( &ctx.skein, input );
+     skein512_4way_final16( &ctx.skein, vhash, input + (64*4) );
 
      // JH
      jh512_4way_update( &ctx.jh, vhash, 64 );
@@ -344,7 +348,7 @@ int scanhash_phi1612_4way( struct work *work, uint32_t max_nonce,
         if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
         {
            pdata[19] = n+i;
-           submit_lane_solution( work, hash+(i<<3), mythr, i );
+           submit_solution( work, hash+(i<<3), mythr );
         }
         n += 4;
      } while ( ( n < max_nonce ) && !work_restart[thr_id].restart );

algo/x13/skunk-4way.c

@@ -125,7 +125,7 @@ int scanhash_skunk_8way( struct work *work, uint32_t max_nonce,
       if ( unlikely( valid_hash( hash+(i<<3), ptarget ) && !bench ) )
       {
          pdata[19] = bswap_32( n+i );
-         submit_lane_solution( work, hash+(i<<3), mythr, i );
+         submit_solution( work, hash+(i<<3), mythr );
       }
       *noncev = _mm512_add_epi32( *noncev,
                                   m512_const1_64( 0x0000000800000000 ) );
@@ -237,7 +237,7 @@ int scanhash_skunk_4way( struct work *work, uint32_t max_nonce,
       if ( unlikely( valid_hash( hash+(i<<3), ptarget ) && !bench ) )
       {
          pdata[19] = bswap_32( n + i );
-         submit_lane_solution( work, hash+(i<<3), mythr, i );
+         submit_solution( work, hash+(i<<3), mythr );
       }
       *noncev = _mm256_add_epi32( *noncev,
                                   m256_const1_64( 0x0000000400000000 ) );

algo/x13/x13-4way.c

@@ -319,7 +319,7 @@ int scanhash_x13_8way( struct work *work, uint32_t max_nonce,
               && fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
          {
              pdata[19] = n+i;
-             submit_lane_solution( work, hash+(i<<3), mythr, i );
+             submit_solution( work, hash+(i<<3), mythr );
          }
          n += 8;
      } while ( ( n < last_nonce ) && !work_restart[thr_id].restart );
@@ -531,7 +531,7 @@ int scanhash_x13_4way( struct work *work, uint32_t max_nonce,
             if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
             {
                pdata[19] = n+i;
-               submit_lane_solution( work, hash+(i<<3), mythr, i );
+               submit_solution( work, hash+(i<<3), mythr );
             }
             n += 4;
          } while ( ( n < max_nonce )  && !work_restart[thr_id].restart );

algo/x13/x13bcd-4way.c

@@ -321,7 +321,7 @@ int scanhash_x13bcd_8way( struct work *work, uint32_t max_nonce,
         if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
         {
               pdata[19] = n+i;
-              submit_lane_solution( work, hash+(i<<3), mythr, i );
+              submit_solution( work, hash+(i<<3), mythr );
         }
         n += 8;
      } while ( ( n < last_nonce ) && !work_restart[thr_id].restart );
@@ -541,7 +541,7 @@ int scanhash_x13bcd_4way( struct work *work, uint32_t max_nonce,
         if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
         {
             pdata[19] = n+i;
-            submit_lane_solution( work, hash+(i<<3), mythr, i );
+            submit_solution( work, hash+(i<<3), mythr );
         }
         n += 4;
      } while ( ( n < last_nonce ) && !work_restart[thr_id].restart );

algo/x13/x13sm3-4way.c

@@ -246,7 +246,7 @@ int scanhash_x13sm3_4way( struct work *work, uint32_t max_nonce,
             if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
             {
                pdata[19] = n+i;
-              submit_lane_solution( work, hash+(i<<3), mythr, i );
+              submit_solution( work, hash+(i<<3), mythr );
             }
             n += 4;
          } while ( ( n < max_nonce ) && !work_restart[thr_id].restart );

algo/x14/polytimos-4way.c

@@ -129,7 +129,7 @@ int scanhash_polytimos_4way( struct work *work, uint32_t max_nonce,
       if( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
       {
          pdata[19] = n+i;
-         submit_lane_solution( work, hash+(i<<3), mythr, i );
+         submit_solution( work, hash+(i<<3), mythr );
       }
       n += 4;
 

algo/x14/veltor-4way.c

@@ -108,7 +108,7 @@ int scanhash_veltor_4way( struct work *work, uint32_t max_nonce,
          if ( (hash+(i<<3))[7] <= Htarg && fulltest( hash+(i<<3), ptarget ) )
          {
             pdata[19] = n+i;
-            submit_lane_solution( work, hash+(i<<3), mythr, i );
+            submit_solution( work, hash+(i<<3), mythr );
          }
          n += 4;
      } while ( ( n < max_nonce ) && !(*restart) );

algo/x14/x14-4way.c

@@ -324,7 +324,7 @@ int scanhash_x14_8way( struct work *work, uint32_t max_nonce,
             if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
             {
                 pdata[19] = n + lane;
-                submit_lane_solution( work, lane_hash, mythr, lane );
+                submit_solution( work, lane_hash, mythr );
             }
          }
          n += 8;
@@ -534,7 +534,7 @@ int scanhash_x14_4way( struct work *work, uint32_t max_nonce,
            if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
            {
                pdata[19] = n + lane;
-               submit_lane_solution( work, lane_hash, mythr, lane );
+               submit_solution( work, lane_hash, mythr );
            }
         }
         n += 4;

algo/x15/x15-4way.c

@@ -364,7 +364,7 @@ int scanhash_x15_8way( struct work *work, uint32_t max_nonce,
         if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
         {
            pdata[19] = n+i;
-           submit_lane_solution( work, hash, mythr, i );
+           submit_solution( work, hash, mythr );
         }
         n += 8;
      } while ( ( n < last_nonce ) && !work_restart[thr_id].restart );
@@ -592,7 +592,7 @@ int scanhash_x15_4way( struct work *work, uint32_t max_nonce,
          if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
          {
             pdata[19] = n+i;
-            submit_lane_solution( work, hash, mythr, i );
+            submit_solution( work, hash, mythr );
          }
          n += 4;
      } while ( ( n < last_nonce ) && !work_restart[thr_id].restart );

algo/x16/minotaur.c

@@ -0,0 +1,258 @@
+// 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/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"
+#else
+  #include "algo/echo/sph_echo.h"
+  #include "algo/groestl/sph_groestl.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;
+#else
+        sph_echo512_context      echo;
+        sph_groestl512_context   groestl;
+#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;
+
+    struct TortureNode {
+        unsigned int algo;
+        TortureNode *childLeft;
+        TortureNode *childRight;
+    } 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:
+            sph_fugue512_init(&garden->fugue);
+            sph_fugue512(&garden->fugue, input, 64);
+            sph_fugue512_close(&garden->fugue, hash);          
+            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:
+            SHA512_Init( &garden->sha512 );
+            SHA512_Update( &garden->sha512, input, 64 );
+            SHA512_Final( (unsigned char*)hash, &garden->sha512 );
+            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;
+    }
+
+    // Output the hash
+    memcpy(output, hash, 64);
+}
+
+// Recursively traverse a given torture garden starting with a given hash and given node within the garden. The hash is overwritten with the final hash.
+static void traverse_garden( TortureGarden *garden, void *hash,
+	                     TortureNode *node )
+{
+    unsigned char partialHash[64] __attribute__ ((aligned (64)));
+    get_hash(partialHash, hash, garden, node->algo);
+
+    if ( partialHash[63] % 2 == 0 )
+    {   // Last byte of output hash is even
+        if ( node->childLeft != NULL )
+            traverse_garden( garden, partialHash, node->childLeft );
+    }
+    else
+    {   // Last byte of output hash is odd
+        if ( node->childRight != NULL )
+            traverse_garden( garden, partialHash, node->childRight );
+    }
+
+    memcpy( hash, partialHash, 64 );
+}
+
+// Associate child nodes with a parent node
+static inline void link_nodes( TortureNode *parent, TortureNode *childLeft,
+	                       TortureNode *childRight ) 
+{
+    parent->childLeft = childLeft;
+    parent->childRight = childRight;
+}
+
+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).
+    link_nodes(&garden.nodes[0], &garden.nodes[1], &garden.nodes[2]);
+    link_nodes(&garden.nodes[1], &garden.nodes[3], &garden.nodes[4]);
+    link_nodes(&garden.nodes[2], &garden.nodes[5], &garden.nodes[6]);
+    link_nodes(&garden.nodes[3], &garden.nodes[7], &garden.nodes[8]);
+    link_nodes(&garden.nodes[4], &garden.nodes[9], &garden.nodes[10]);
+    link_nodes(&garden.nodes[5], &garden.nodes[11], &garden.nodes[12]);
+    link_nodes(&garden.nodes[6], &garden.nodes[13], &garden.nodes[14]);
+    link_nodes(&garden.nodes[7], &garden.nodes[15], &garden.nodes[16]);
+    link_nodes(&garden.nodes[8], &garden.nodes[15], &garden.nodes[16]);
+    link_nodes(&garden.nodes[9], &garden.nodes[15], &garden.nodes[16]);
+    link_nodes(&garden.nodes[10], &garden.nodes[15], &garden.nodes[16]);
+    link_nodes(&garden.nodes[11], &garden.nodes[17], &garden.nodes[18]);
+    link_nodes(&garden.nodes[12], &garden.nodes[17], &garden.nodes[18]);
+    link_nodes(&garden.nodes[13], &garden.nodes[17], &garden.nodes[18]);
+    link_nodes(&garden.nodes[14], &garden.nodes[17], &garden.nodes[18]);
+    link_nodes(&garden.nodes[15], &garden.nodes[19], &garden.nodes[20]);
+    link_nodes(&garden.nodes[16], &garden.nodes[19], &garden.nodes[20]);
+    link_nodes(&garden.nodes[17], &garden.nodes[19], &garden.nodes[20]);
+    link_nodes(&garden.nodes[18], &garden.nodes[19], &garden.nodes[20]);
+    link_nodes(&garden.nodes[19], &garden.nodes[21], &garden.nodes[21]);
+    link_nodes(&garden.nodes[20], &garden.nodes[21], &garden.nodes[21]);
+    garden.nodes[21].childLeft = NULL;
+    garden.nodes[21].childRight = 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
+    SHA512_Init( &garden.sha512 );
+    SHA512_Update( &garden.sha512, input, 80 );
+    SHA512_Final( (unsigned char*) hash, &garden.sha512 );
+
+    // 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
+    traverse_garden( &garden, hash, &garden.nodes[0] );
+
+    memcpy( output, hash, 32 );
+
+    return 1;
+}
+
+bool register_minotaur_algo( algo_gate_t* gate )
+{
+  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;
+};
+

algo/x16/x16r-4way.c

@@ -505,7 +505,7 @@ int scanhash_x16r_8way( struct work *work, uint32_t max_nonce,
       if ( unlikely( valid_hash( hash + (i<<3), ptarget ) && !bench ) )
       {
          pdata[19] = bswap_32( n+i );
-         submit_lane_solution( work, hash+(i<<3), mythr, i );
+         submit_solution( work, hash+(i<<3), mythr );
       }
       *noncev = _mm512_add_epi32( *noncev,
                                   m512_const1_64( 0x0000000800000000 ) );
@@ -869,7 +869,7 @@ int scanhash_x16r_4way( struct work *work, uint32_t max_nonce,
       if ( unlikely( valid_hash( hash + (i<<3), ptarget ) && !bench ) )
       {
          pdata[19] = bswap_32( n+i );
-         submit_lane_solution( work, hash+(i<<3), mythr, i );
+         submit_solution( work, hash+(i<<3), mythr );
       }
       *noncev = _mm256_add_epi32( *noncev,
                                   m256_const1_64( 0x0000000400000000 ) );

algo/x16/x16r-gate.c

@@ -135,18 +135,16 @@ 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 +162,35 @@ 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 )

algo/x16/x16rt-4way.c

@@ -46,7 +46,7 @@ int scanhash_x16rt_8way( struct work *work, uint32_t max_nonce,
       if ( unlikely( valid_hash( hash + (i<<3), ptarget ) && !bench ) )
       {
          pdata[19] = bswap_32( n+i );
-         submit_lane_solution( work, hash+(i<<3), mythr, i );
+         submit_solution( work, hash+(i<<3), mythr );
       }
       *noncev = _mm512_add_epi32( *noncev,
                                   m512_const1_64( 0x0000000800000000 ) );
@@ -99,7 +99,7 @@ int scanhash_x16rt_4way( struct work *work, uint32_t max_nonce,
       if ( unlikely( valid_hash( hash + (i<<3), ptarget ) && !bench ) )
       {
          pdata[19] = bswap_32( n+i );
-         submit_lane_solution( work, hash+(i<<3), mythr, i );
+         submit_solution( work, hash+(i<<3), mythr );
       }
       *noncev = _mm256_add_epi32( *noncev,
                                   m256_const1_64( 0x0000000400000000 ) );

algo/x16/x16rv2-4way.c

@@ -678,7 +678,7 @@ int scanhash_x16rv2_8way( struct work *work, uint32_t max_nonce,
       if ( unlikely( valid_hash( hash + (i<<3), ptarget ) && !bench ) )
       {
          pdata[19] = bswap_32( n+i );
-         submit_lane_solution( work, hash+(i<<3), mythr, i );
+         submit_solution( work, hash+(i<<3), mythr );
       }
       *noncev = _mm512_add_epi32( *noncev,
                                   m512_const1_64( 0x0000000800000000 ) );
@@ -1131,7 +1131,7 @@ int scanhash_x16rv2_4way( struct work *work, uint32_t max_nonce,
       if ( unlikely( valid_hash( hash + (i<<3), ptarget ) && !bench ) )
       {
          pdata[19] = bswap_32( n+i );
-         submit_lane_solution( work, hash+(i<<3), mythr, i );
+         submit_solution( work, hash+(i<<3), mythr );
       }
       *noncev = _mm256_add_epi32( *noncev,
                                   m256_const1_64( 0x0000000400000000 ) );

algo/x16/x21s-4way.c

@@ -177,7 +177,7 @@ int scanhash_x21s_8way( struct work *work, uint32_t max_nonce,
          if ( likely( valid_hash( lane_hash, ptarget ) && !bench ) )
          {
              pdata[19] = bswap_32( n + lane );
-             submit_lane_solution( work, lane_hash, mythr, lane );
+             submit_solution( work, lane_hash, mythr );
          }
       }
       *noncev = _mm512_add_epi32( *noncev,
@@ -347,7 +347,7 @@ int scanhash_x21s_4way( struct work *work, uint32_t max_nonce,
       if ( unlikely( valid_hash( hash + (i<<3), ptarget ) && !bench ) )
       {
          pdata[19] = bswap_32( n+i );
-         submit_lane_solution( work, hash+(i<<3), mythr, i );
+         submit_solution( work, hash+(i<<3), mythr );
       }
       *noncev = _mm256_add_epi32( *noncev,
                                   m256_const1_64( 0x0000000400000000 ) );

algo/x17/sonoa-4way.c

@@ -58,7 +58,7 @@ union _sonoa_8way_context_overlay
 
 typedef union _sonoa_8way_context_overlay sonoa_8way_context_overlay;
 
-void sonoa_8way_hash( void *state, const void *input )
+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)));
@@ -186,6 +186,7 @@ void sonoa_8way_hash( void *state, const void *input )
 
 #endif
 
+     if ( work_restart[thr_id].restart ) return 0;
 // 2
 
      bmw512_8way_full( &ctx.bmw, vhash, vhash, 64 );
@@ -301,6 +302,7 @@ void sonoa_8way_hash( void *state, const void *input )
      hamsi512_8way_update( &ctx.hamsi, vhash, 64 );
      hamsi512_8way_close( &ctx.hamsi, vhash );
 
+     if ( work_restart[thr_id].restart ) return 0;
 // 3
 
      bmw512_8way_full( &ctx.bmw, vhash, vhash, 64 );
@@ -430,6 +432,7 @@ void sonoa_8way_hash( void *state, const void *input )
      sph_fugue512_full( &ctx.fugue, hash6, hash6, 64 );
      sph_fugue512_full( &ctx.fugue, hash7, hash7, 64 );
 
+     if ( work_restart[thr_id].restart ) return 0;
 // 4
 
      intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
@@ -627,6 +630,7 @@ void sonoa_8way_hash( void *state, const void *input )
 
 #endif
 
+     if ( work_restart[thr_id].restart ) return 0;
 // 5
 
      bmw512_8way_full( &ctx.bmw, vhash, vhash, 64 );
@@ -779,6 +783,7 @@ void sonoa_8way_hash( void *state, const void *input )
      sph_whirlpool512_full( &ctx.whirlpool, hash6, hash6, 64 );
      sph_whirlpool512_full( &ctx.whirlpool, hash7, hash7, 64 );
 
+     if ( work_restart[thr_id].restart ) return 0;
 // 6
 
      intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
@@ -947,6 +952,7 @@ void sonoa_8way_hash( void *state, const void *input )
      sph_whirlpool512_full( &ctx.whirlpool, hash6, hash6, 64 );
      sph_whirlpool512_full( &ctx.whirlpool, hash7, hash7, 64 );
 
+     if ( work_restart[thr_id].restart ) return 0;
 // 7
 
      intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
@@ -1108,50 +1114,8 @@ void sonoa_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_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
-   {
-      sonoa_8way_hash( hash, vdata );
-
-      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_lane_solution( work, lane_hash, mythr, lane );
-         }
-      }
-      *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(SONOA_4WAY)
@@ -1179,7 +1143,7 @@ union _sonoa_4way_context_overlay
 
 typedef union _sonoa_4way_context_overlay sonoa_4way_context_overlay;
 
-void sonoa_4way_hash( void *state, const void *input )
+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)));
@@ -1243,6 +1207,7 @@ void sonoa_4way_hash( void *state, const void *input )
      echo_full( &ctx.echo, (BitSequence *)hash3, 512,
                      (const BitSequence *)hash3, 64 );
      
+     if ( work_restart[thr_id].restart ) return 0;
 // 2
 
      intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
@@ -1302,6 +1267,7 @@ void sonoa_4way_hash( void *state, const void *input )
      hamsi512_4way_update( &ctx.hamsi, vhash, 64 );
      hamsi512_4way_close( &ctx.hamsi, vhash );
 
+     if ( work_restart[thr_id].restart ) return 0;
 // 3
 
      bmw512_4way_init( &ctx.bmw );
@@ -1366,6 +1332,7 @@ void sonoa_4way_hash( void *state, const void *input )
      sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
      sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );
 
+     if ( work_restart[thr_id].restart ) return 0;
 // 4
      intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
 
@@ -1462,6 +1429,7 @@ void sonoa_4way_hash( void *state, const void *input )
      shavite512_2way_init( &ctx.shavite );
      shavite512_2way_update_close( &ctx.shavite, vhashB, vhashB, 64 );
 
+     if ( work_restart[thr_id].restart ) return 0;
 // 5
      rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
 
@@ -1546,6 +1514,7 @@ void sonoa_4way_hash( void *state, const void *input )
      sph_whirlpool512_full( &ctx.whirlpool, hash2, hash2, 64 );
      sph_whirlpool512_full( &ctx.whirlpool, hash3, hash3, 64 );
 
+     if ( work_restart[thr_id].restart ) return 0;
 // 6
 
      intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
@@ -1638,6 +1607,7 @@ void sonoa_4way_hash( void *state, const void *input )
      sph_whirlpool512_full( &ctx.whirlpool, hash2, hash2, 64 );
      sph_whirlpool512_full( &ctx.whirlpool, hash3, hash3, 64 );
 
+     if ( work_restart[thr_id].restart ) return 0;    
 // 7
 
      intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
@@ -1728,49 +1698,8 @@ void sonoa_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_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
-     {
-        sonoa_4way_hash( hash, vdata );
-
-        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_lane_solution( work, lane_hash, mythr, lane );
-           }
-        }
-        *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

algo/x17/sonoa-gate.c

@@ -3,14 +3,13 @@
 bool register_sonoa_algo( algo_gate_t* gate )
 {
 #if defined (SONOA_8WAY)
-  gate->scanhash  = (void*)&scanhash_sonoa_8way;
+  gate->scanhash  = (void*)&scanhash_8way_64in_32out;
   gate->hash      = (void*)&sonoa_8way_hash;
 #elif defined (SONOA_4WAY)
-  gate->scanhash  = (void*)&scanhash_sonoa_4way;
+  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;
 #endif
   gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;

algo/x17/sonoa-gate.h

@@ -14,21 +14,15 @@ bool register_sonoa_algo( algo_gate_t* gate );
 
 #if defined(SONOA_8WAY)
 
-void sonoa_8way_hash( void *state, const void *input );
-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)
 
-void sonoa_4way_hash( void *state, const void *input );
-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
 
-void sonoa_hash( void *state, const void *input );
-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

algo/x17/sonoa.c

@@ -83,27 +83,27 @@ void init_sonoa_ctx()
         sph_haval256_5_init(&sonoa_ctx.haval);
 };
 
-void sonoa_hash( void *state, const void *input )
+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)));
-        memcpy( &ctx, &sonoa_ctx, sizeof(sonoa_ctx) );
+   sonoa_ctx_holder ctx __attribute__ ((aligned (64)));
+   memcpy( &ctx, &sonoa_ctx, sizeof(sonoa_ctx) );
 
-        sph_blake512(&ctx.blake, input, 80);
+   sph_blake512(&ctx.blake, input, 80);
 	sph_blake512_close(&ctx.blake, hash);
 
 	sph_bmw512(&ctx.bmw, hash, 64);
 	sph_bmw512_close(&ctx.bmw, hash);
 
 #if defined(__AES__)
-        update_and_final_groestl( &ctx.groestl, (char*)hash,
-                                  (const char*)hash, 512 );
+   update_and_final_groestl( &ctx.groestl, (char*)hash,
+                                     (const char*)hash, 512 );
 #else
-        sph_groestl512(&ctx.groestl, hash, 64);
-        sph_groestl512_close(&ctx.groestl, hash);
+   sph_groestl512(&ctx.groestl, hash, 64);
+   sph_groestl512_close(&ctx.groestl, hash);
 #endif
 
-	sph_skein512(&ctx.skein, hash, 64);
+   sph_skein512(&ctx.skein, hash, 64);
 	sph_skein512_close(&ctx.skein, hash);
 
 	sph_jh512(&ctx.jh, hash, 64);
@@ -112,484 +112,461 @@ void sonoa_hash( void *state, const void *input )
 	sph_keccak512(&ctx.keccak, hash, 64);
 	sph_keccak512_close(&ctx.keccak, hash);
 
-        update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
-                                (const BitSequence*)hash, 64 );
+   update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
+                                 (const BitSequence*)hash, 64 );
 
-        cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
-                              (const byte*)hash, 64 );
+   cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
+                                   (const byte*)hash, 64 );
 
 	sph_shavite512(&ctx.shavite, hash, 64);
 	sph_shavite512_close(&ctx.shavite, hash);
 
-        update_final_sd( &ctx.simd, (BitSequence *)hash,
+   update_final_sd( &ctx.simd, (BitSequence *)hash,
                          (const BitSequence *)hash, 512 );
 
 #if defined(__AES__)
-        update_final_echo ( &ctx.echo, (BitSequence *)hash,
+   update_final_echo ( &ctx.echo, (BitSequence *)hash,
                             (const BitSequence *)hash, 512 );
 #else
-        sph_echo512(&ctx.echo, hash, 64);
-        sph_echo512_close(&ctx.echo, hash);
+   sph_echo512(&ctx.echo, hash, 64);
+   sph_echo512_close(&ctx.echo, hash);
 #endif
 
+   if ( work_restart[thr_id].restart ) return 0;
 //
 
-        sph_bmw512_init( &ctx.bmw);
-        sph_bmw512(&ctx.bmw, hash, 64);
-        sph_bmw512_close(&ctx.bmw, hash);
+   sph_bmw512_init( &ctx.bmw);
+   sph_bmw512(&ctx.bmw, hash, 64);
+   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 );
+   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);
+   sph_groestl512_init(&ctx.groestl );
+   sph_groestl512(&ctx.groestl, hash, 64);
+   sph_groestl512_close(&ctx.groestl, hash);
 #endif
 
-        sph_skein512_init( &ctx.skein);
-        sph_skein512(&ctx.skein, hash, 64);
-        sph_skein512_close(&ctx.skein, hash);
+   sph_skein512_init( &ctx.skein);
+   sph_skein512(&ctx.skein, hash, 64);
+   sph_skein512_close(&ctx.skein, hash);
 
-        sph_jh512_init( &ctx.jh);
-        sph_jh512(&ctx.jh, hash, 64);
-        sph_jh512_close(&ctx.jh, hash);
+   sph_jh512_init( &ctx.jh);
+   sph_jh512(&ctx.jh, hash, 64);
+   sph_jh512_close(&ctx.jh, hash);
 
-        sph_keccak512_init( &ctx.keccak );
-        sph_keccak512(&ctx.keccak, hash, 64);
-        sph_keccak512_close(&ctx.keccak, hash);
+   sph_keccak512_init( &ctx.keccak );
+   sph_keccak512(&ctx.keccak, hash, 64);
+   sph_keccak512_close(&ctx.keccak, hash);
 
-        init_luffa( &ctx.luffa, 512 );
-        update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
-                                (const BitSequence*)hash, 64 );
+   init_luffa( &ctx.luffa, 512 );
+   update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
+                                 (const BitSequence*)hash, 64 );
 
-        cubehashInit( &ctx.cubehash, 512, 16, 32 );
-        cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
-                              (const byte*)hash, 64 );
+   cubehashInit( &ctx.cubehash, 512, 16, 32 );
+   cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
+                                   (const byte*)hash, 64 );
 
-        sph_shavite512_init( &ctx.shavite );
-        sph_shavite512(&ctx.shavite, hash, 64);
-        sph_shavite512_close(&ctx.shavite, hash);
+   sph_shavite512_init( &ctx.shavite );
+   sph_shavite512(&ctx.shavite, hash, 64);
+   sph_shavite512_close(&ctx.shavite, hash);
 
-        init_sd( &ctx.simd, 512 );
-        update_final_sd( &ctx.simd, (BitSequence *)hash,
+   init_sd( &ctx.simd, 512 );
+   update_final_sd( &ctx.simd, (BitSequence *)hash,
                          (const BitSequence *)hash, 512 );
 
 #if defined(__AES__)
-        init_echo( &ctx.echo, 512 );
-        update_final_echo ( &ctx.echo, (BitSequence *)hash,
+   init_echo( &ctx.echo, 512 );
+   update_final_echo ( &ctx.echo, (BitSequence *)hash,
                             (const BitSequence *)hash, 512 );
 #else
-        sph_echo512_init( &ctx.echo );
-        sph_echo512(&ctx.echo, hash, 64);
-        sph_echo512_close(&ctx.echo, hash);
+   sph_echo512_init( &ctx.echo );
+   sph_echo512(&ctx.echo, hash, 64);
+   sph_echo512_close(&ctx.echo, hash);
 #endif
 
-        sph_hamsi512(&ctx.hamsi, hash, 64);
-        sph_hamsi512_close(&ctx.hamsi, hash);
+   sph_hamsi512(&ctx.hamsi, hash, 64);
+   sph_hamsi512_close(&ctx.hamsi, hash);
 	
+   if ( work_restart[thr_id].restart ) return 0;
 //
 
-        sph_bmw512_init( &ctx.bmw);
-	sph_bmw512(&ctx.bmw, hash, 64);
-        sph_bmw512_close(&ctx.bmw, hash);
+   sph_bmw512_init( &ctx.bmw);
+   sph_bmw512(&ctx.bmw, hash, 64);
+   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 );
+   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);
+   sph_groestl512_init(&ctx.groestl );
+   sph_groestl512(&ctx.groestl, hash, 64);
+   sph_groestl512_close(&ctx.groestl, hash);
 #endif
 
-        sph_skein512_init( &ctx.skein);
-        sph_skein512(&ctx.skein, hash, 64);
-        sph_skein512_close(&ctx.skein, hash);
+   sph_skein512_init( &ctx.skein);
+   sph_skein512(&ctx.skein, hash, 64);
+   sph_skein512_close(&ctx.skein, hash);
 
-        sph_jh512_init( &ctx.jh);
-        sph_jh512(&ctx.jh, hash, 64);
-        sph_jh512_close(&ctx.jh, hash);
+   sph_jh512_init( &ctx.jh);
+   sph_jh512(&ctx.jh, hash, 64);
+   sph_jh512_close(&ctx.jh, hash);
 
-        sph_keccak512_init( &ctx.keccak );
-        sph_keccak512(&ctx.keccak, hash, 64);
-        sph_keccak512_close(&ctx.keccak, hash);
+   sph_keccak512_init( &ctx.keccak );
+   sph_keccak512(&ctx.keccak, hash, 64);
+   sph_keccak512_close(&ctx.keccak, hash);
 
-        init_luffa( &ctx.luffa, 512 );
-        update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
-                                (const BitSequence*)hash, 64 );
+   init_luffa( &ctx.luffa, 512 );
+   update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
+                                 (const BitSequence*)hash, 64 );
 
-        cubehashInit( &ctx.cubehash, 512, 16, 32 );
-        cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
-                              (const byte*)hash, 64 );
+   cubehashInit( &ctx.cubehash, 512, 16, 32 );
+   cubehashUpdateDigest( &ctx.cubehash, (byte*)hash,
+                                  (const byte*)hash, 64 );
 
-        sph_shavite512_init( &ctx.shavite );
-        sph_shavite512(&ctx.shavite, hash, 64);
-        sph_shavite512_close(&ctx.shavite, hash);
+   sph_shavite512_init( &ctx.shavite );
+   sph_shavite512(&ctx.shavite, hash, 64);
+   sph_shavite512_close(&ctx.shavite, hash);
 
-        init_sd( &ctx.simd, 512 );
-        update_final_sd( &ctx.simd, (BitSequence *)hash,
+   init_sd( &ctx.simd, 512 );
+   update_final_sd( &ctx.simd, (BitSequence *)hash,
                          (const BitSequence *)hash, 512 );
 
 #if defined(__AES__)
-        init_echo( &ctx.echo, 512 );
-        update_final_echo ( &ctx.echo, (BitSequence *)hash,
+   init_echo( &ctx.echo, 512 );
+   update_final_echo ( &ctx.echo, (BitSequence *)hash,
                             (const BitSequence *)hash, 512 );
 #else
-        sph_echo512_init( &ctx.echo );
-        sph_echo512(&ctx.echo, hash, 64);
-        sph_echo512_close(&ctx.echo, hash);
+   sph_echo512_init( &ctx.echo );
+   sph_echo512(&ctx.echo, hash, 64);
+   sph_echo512_close(&ctx.echo, hash);
 #endif
 
-        sph_hamsi512_init( &ctx.hamsi );
-        sph_hamsi512(&ctx.hamsi, hash, 64);
-        sph_hamsi512_close(&ctx.hamsi, hash);
+   sph_hamsi512_init( &ctx.hamsi );
+   sph_hamsi512(&ctx.hamsi, hash, 64);
+   sph_hamsi512_close(&ctx.hamsi, hash);
 
-        sph_fugue512(&ctx.fugue, hash, 64);
-        sph_fugue512_close(&ctx.fugue, hash);
+   sph_fugue512(&ctx.fugue, hash, 64);
+   sph_fugue512_close(&ctx.fugue, hash);
 
+   if ( work_restart[thr_id].restart ) return 0;
 //
 
-        sph_bmw512_init( &ctx.bmw);
-        sph_bmw512(&ctx.bmw, hash, 64);
-        sph_bmw512_close(&ctx.bmw, hash);
+   sph_bmw512_init( &ctx.bmw);
+   sph_bmw512(&ctx.bmw, hash, 64);
+   sph_bmw512_close(&ctx.bmw, hash);
 
 #if defined(__AES__)
-        init_groestl( &ctx.groestl, 64 );
-        update_and_final_groestl( &ctx.groestl, (char*)hash,
+   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);
+   sph_groestl512_init(&ctx.groestl );
+   sph_groestl512(&ctx.groestl, hash, 64);
+   sph_groestl512_close(&ctx.groestl, hash);
 #endif
 
-        sph_skein512_init( &ctx.skein);
-        sph_skein512(&ctx.skein, hash, 64);
-        sph_skein512_close(&ctx.skein, hash);
+   sph_skein512_init( &ctx.skein);
+   sph_skein512(&ctx.skein, hash, 64);
+   sph_skein512_close(&ctx.skein, hash);
 
-        sph_jh512_init( &ctx.jh);
-        sph_jh512(&ctx.jh, hash, 64);
-        sph_jh512_close(&ctx.jh, hash);
+   sph_jh512_init( &ctx.jh);
+   sph_jh512(&ctx.jh, hash, 64);
+   sph_jh512_close(&ctx.jh, hash);
 
-        sph_keccak512_init( &ctx.keccak );
-        sph_keccak512(&ctx.keccak, hash, 64);
-        sph_keccak512_close(&ctx.keccak, hash);
+   sph_keccak512_init( &ctx.keccak );
+   sph_keccak512(&ctx.keccak, hash, 64);
+   sph_keccak512_close(&ctx.keccak, hash);
 
-        init_luffa( &ctx.luffa, 512 );
-        update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
+   init_luffa( &ctx.luffa, 512 );
+   update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
                                 (const BitSequence*)hash, 64 );
 
-        cubehashInit( &ctx.cubehash, 512, 16, 32 );
-        cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
+   cubehashInit( &ctx.cubehash, 512, 16, 32 );
+   cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
                               (const byte*)hash, 64 );
 
-        sph_shavite512_init( &ctx.shavite );
-        sph_shavite512(&ctx.shavite, hash, 64);
-        sph_shavite512_close(&ctx.shavite, hash);
+   sph_shavite512_init( &ctx.shavite );
+   sph_shavite512(&ctx.shavite, hash, 64);
+   sph_shavite512_close(&ctx.shavite, hash);
 
-        init_sd( &ctx.simd, 512 );
-        update_final_sd( &ctx.simd, (BitSequence *)hash,
+   init_sd( &ctx.simd, 512 );
+   update_final_sd( &ctx.simd, (BitSequence *)hash,
                          (const BitSequence *)hash, 512 );
 
 #if defined(__AES__)
-        init_echo( &ctx.echo, 512 );
-        update_final_echo ( &ctx.echo, (BitSequence *)hash,
+   init_echo( &ctx.echo, 512 );
+   update_final_echo ( &ctx.echo, (BitSequence *)hash,
                             (const BitSequence *)hash, 512 );
 #else
-        sph_echo512_init( &ctx.echo );
-        sph_echo512(&ctx.echo, hash, 64);
-        sph_echo512_close(&ctx.echo, hash);
+   sph_echo512_init( &ctx.echo );
+   sph_echo512(&ctx.echo, hash, 64);
+   sph_echo512_close(&ctx.echo, hash);
 #endif
 
-        sph_hamsi512_init( &ctx.hamsi );
-        sph_hamsi512(&ctx.hamsi, hash, 64);
-        sph_hamsi512_close(&ctx.hamsi, hash);
+   sph_hamsi512_init( &ctx.hamsi );
+   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);
+   sph_fugue512_init( &ctx.fugue );
+   sph_fugue512(&ctx.fugue, hash, 64);
+   sph_fugue512_close(&ctx.fugue, hash);
 
-        sph_shabal512(&ctx.shabal, hash, 64);
-        sph_shabal512_close(&ctx.shabal, hash);
+   sph_shabal512(&ctx.shabal, hash, 64);
+   sph_shabal512_close(&ctx.shabal, hash);
 
-        sph_hamsi512_init( &ctx.hamsi );
-        sph_hamsi512(&ctx.hamsi, hash, 64);
-        sph_hamsi512_close(&ctx.hamsi, hash);
+   sph_hamsi512_init( &ctx.hamsi );
+   sph_hamsi512(&ctx.hamsi, hash, 64);
+   sph_hamsi512_close(&ctx.hamsi, hash);
 
 #if defined(__AES__)
-        init_echo( &ctx.echo, 512 );
-        update_final_echo ( &ctx.echo, (BitSequence *)hash,
+   init_echo( &ctx.echo, 512 );
+   update_final_echo ( &ctx.echo, (BitSequence *)hash,
                             (const BitSequence *)hash, 512 );
 #else
-        sph_echo512_init( &ctx.echo );
-        sph_echo512(&ctx.echo, hash, 64);
-        sph_echo512_close(&ctx.echo, hash);
+   sph_echo512_init( &ctx.echo );
+   sph_echo512(&ctx.echo, hash, 64);
+   sph_echo512_close(&ctx.echo, hash);
 #endif
 
-        sph_shavite512_init( &ctx.shavite );
-        sph_shavite512(&ctx.shavite, hash, 64);
-        sph_shavite512_close(&ctx.shavite, hash);
+   sph_shavite512_init( &ctx.shavite );
+   sph_shavite512(&ctx.shavite, hash, 64);
+   sph_shavite512_close(&ctx.shavite, hash);
 
+   if ( work_restart[thr_id].restart ) return 0;
 //
 
-        sph_bmw512_init( &ctx.bmw);
-        sph_bmw512(&ctx.bmw, hash, 64);
-        sph_bmw512_close(&ctx.bmw, hash);
+   sph_bmw512_init( &ctx.bmw);
+   sph_bmw512(&ctx.bmw, hash, 64);
+   sph_bmw512_close(&ctx.bmw, hash);
 
-        sph_shabal512_init( &ctx.shabal );
+   sph_shabal512_init( &ctx.shabal );
 	sph_shabal512(&ctx.shabal, hash, 64);
-        sph_shabal512_close(&ctx.shabal, hash);
+   sph_shabal512_close(&ctx.shabal, hash);
 
 #if defined(__AES__)
-        init_groestl( &ctx.groestl, 64 );
-        update_and_final_groestl( &ctx.groestl, (char*)hash,
+   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);
+   sph_groestl512_init(&ctx.groestl );
+   sph_groestl512(&ctx.groestl, hash, 64);
+   sph_groestl512_close(&ctx.groestl, hash);
 #endif
 
-        sph_skein512_init( &ctx.skein);
-        sph_skein512(&ctx.skein, hash, 64);
-        sph_skein512_close(&ctx.skein, hash);
+   sph_skein512_init( &ctx.skein);
+   sph_skein512(&ctx.skein, hash, 64);
+   sph_skein512_close(&ctx.skein, hash);
 
-        sph_jh512_init( &ctx.jh);
-        sph_jh512(&ctx.jh, hash, 64);
-        sph_jh512_close(&ctx.jh, hash);
+   sph_jh512_init( &ctx.jh);
+   sph_jh512(&ctx.jh, hash, 64);
+   sph_jh512_close(&ctx.jh, hash);
 
-        sph_keccak512_init( &ctx.keccak );
-        sph_keccak512(&ctx.keccak, hash, 64);
-        sph_keccak512_close(&ctx.keccak, hash);
+   sph_keccak512_init( &ctx.keccak );
+   sph_keccak512(&ctx.keccak, hash, 64);
+   sph_keccak512_close(&ctx.keccak, hash);
 
-        init_luffa( &ctx.luffa, 512 );
-        update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
+   init_luffa( &ctx.luffa, 512 );
+   update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
                                 (const BitSequence*)hash, 64 );
 
-        cubehashInit( &ctx.cubehash, 512, 16, 32 );
-        cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
+   cubehashInit( &ctx.cubehash, 512, 16, 32 );
+   cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
                               (const byte*)hash, 64 );
 
-        sph_shavite512_init( &ctx.shavite );
-        sph_shavite512(&ctx.shavite, hash, 64);
-        sph_shavite512_close(&ctx.shavite, hash);
+   sph_shavite512_init( &ctx.shavite );
+   sph_shavite512(&ctx.shavite, hash, 64);
+   sph_shavite512_close(&ctx.shavite, hash);
 
-        init_sd( &ctx.simd, 512 );
-        update_final_sd( &ctx.simd, (BitSequence *)hash,
+   init_sd( &ctx.simd, 512 );
+   update_final_sd( &ctx.simd, (BitSequence *)hash,
                          (const BitSequence *)hash, 512 );
 
 #if defined(__AES__)
-        init_echo( &ctx.echo, 512 );
-        update_final_echo ( &ctx.echo, (BitSequence *)hash,
+   init_echo( &ctx.echo, 512 );
+   update_final_echo ( &ctx.echo, (BitSequence *)hash,
                             (const BitSequence *)hash, 512 );
 #else
-        sph_echo512_init( &ctx.echo );
-        sph_echo512(&ctx.echo, hash, 64);
-        sph_echo512_close(&ctx.echo, hash);
+   sph_echo512_init( &ctx.echo );
+   sph_echo512(&ctx.echo, hash, 64);
+   sph_echo512_close(&ctx.echo, hash);
 #endif
 
-        sph_hamsi512_init( &ctx.hamsi );
-        sph_hamsi512(&ctx.hamsi, hash, 64);
-        sph_hamsi512_close(&ctx.hamsi, hash);
+   sph_hamsi512_init( &ctx.hamsi );
+   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);
+   sph_fugue512_init( &ctx.fugue );
+   sph_fugue512(&ctx.fugue, hash, 64);
+   sph_fugue512_close(&ctx.fugue, hash);
 
-        sph_shabal512_init( &ctx.shabal );
-        sph_shabal512(&ctx.shabal, hash, 64);
-        sph_shabal512_close(&ctx.shabal, hash);
+   sph_shabal512_init( &ctx.shabal );
+   sph_shabal512(&ctx.shabal, hash, 64);
+   sph_shabal512_close(&ctx.shabal, hash);
 
-        sph_whirlpool(&ctx.whirlpool, hash, 64);
-        sph_whirlpool_close(&ctx.whirlpool, hash);
+   sph_whirlpool(&ctx.whirlpool, hash, 64);
+   sph_whirlpool_close(&ctx.whirlpool, hash);
 
+   if ( work_restart[thr_id].restart ) return 0;
 //
-        sph_bmw512_init( &ctx.bmw);
-        sph_bmw512(&ctx.bmw, hash, 64);
-        sph_bmw512_close(&ctx.bmw, hash);
+   sph_bmw512_init( &ctx.bmw);
+   sph_bmw512(&ctx.bmw, hash, 64);
+   sph_bmw512_close(&ctx.bmw, hash);
 
 #if defined(__AES__)
-        init_groestl( &ctx.groestl, 64 );
-        update_and_final_groestl( &ctx.groestl, (char*)hash,
+   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);
+   sph_groestl512_init(&ctx.groestl );
+   sph_groestl512(&ctx.groestl, hash, 64);
+   sph_groestl512_close(&ctx.groestl, hash);
 #endif
 
-        sph_skein512_init( &ctx.skein);
-        sph_skein512(&ctx.skein, hash, 64);
-        sph_skein512_close(&ctx.skein, hash);
+   sph_skein512_init( &ctx.skein);
+   sph_skein512(&ctx.skein, hash, 64);
+   sph_skein512_close(&ctx.skein, hash);
 
-        sph_jh512_init( &ctx.jh);
-        sph_jh512(&ctx.jh, hash, 64);
-        sph_jh512_close(&ctx.jh, hash);
+   sph_jh512_init( &ctx.jh);
+   sph_jh512(&ctx.jh, hash, 64);
+   sph_jh512_close(&ctx.jh, hash);
 
-        sph_keccak512_init( &ctx.keccak );
-        sph_keccak512(&ctx.keccak, hash, 64);
-        sph_keccak512_close(&ctx.keccak, hash);
+   sph_keccak512_init( &ctx.keccak );
+   sph_keccak512(&ctx.keccak, hash, 64);
+   sph_keccak512_close(&ctx.keccak, hash);
 
-        init_luffa( &ctx.luffa, 512 );
-        update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
+   init_luffa( &ctx.luffa, 512 );
+   update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
                                 (const BitSequence*)hash, 64 );
 
-        cubehashInit( &ctx.cubehash, 512, 16, 32 );
-        cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
+   cubehashInit( &ctx.cubehash, 512, 16, 32 );
+   cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
                               (const byte*)hash, 64 );
 
-        sph_shavite512_init( &ctx.shavite );
-        sph_shavite512(&ctx.shavite, hash, 64);
-        sph_shavite512_close(&ctx.shavite, hash);
+   sph_shavite512_init( &ctx.shavite );
+   sph_shavite512(&ctx.shavite, hash, 64);
+   sph_shavite512_close(&ctx.shavite, hash);
 
-        init_sd( &ctx.simd, 512 );
-        update_final_sd( &ctx.simd, (BitSequence *)hash,
+   init_sd( &ctx.simd, 512 );
+   update_final_sd( &ctx.simd, (BitSequence *)hash,
                          (const BitSequence *)hash, 512 );
 
 #if defined(__AES__)
-        init_echo( &ctx.echo, 512 );
-        update_final_echo ( &ctx.echo, (BitSequence *)hash,
+   init_echo( &ctx.echo, 512 );
+   update_final_echo ( &ctx.echo, (BitSequence *)hash,
                             (const BitSequence *)hash, 512 );
 #else
-        sph_echo512_init( &ctx.echo );
-        sph_echo512(&ctx.echo, hash, 64);
-        sph_echo512_close(&ctx.echo, hash);
+   sph_echo512_init( &ctx.echo );
+   sph_echo512(&ctx.echo, hash, 64);
+   sph_echo512_close(&ctx.echo, hash);
 #endif
 
-        sph_hamsi512_init( &ctx.hamsi );
-        sph_hamsi512(&ctx.hamsi, hash, 64);
-        sph_hamsi512_close(&ctx.hamsi, hash);
+   sph_hamsi512_init( &ctx.hamsi );
+   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);
+   sph_fugue512_init( &ctx.fugue );
+   sph_fugue512(&ctx.fugue, hash, 64);
+   sph_fugue512_close(&ctx.fugue, hash);
 
-        sph_shabal512_init( &ctx.shabal );
-        sph_shabal512(&ctx.shabal, hash, 64);
-        sph_shabal512_close(&ctx.shabal, hash);
+   sph_shabal512_init( &ctx.shabal );
+   sph_shabal512(&ctx.shabal, hash, 64);
+   sph_shabal512_close(&ctx.shabal, hash);
 
-        sph_whirlpool_init( &ctx.whirlpool );
-        sph_whirlpool(&ctx.whirlpool, hash, 64);
-        sph_whirlpool_close(&ctx.whirlpool, hash);
+   sph_whirlpool_init( &ctx.whirlpool );
+   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 );
+   SHA512_Update( &ctx.sha512, hash, 64 );
+   SHA512_Final( (unsigned char*) hash, &ctx.sha512 );
 
-        sph_whirlpool_init( &ctx.whirlpool );
-        sph_whirlpool(&ctx.whirlpool, hash, 64);
-        sph_whirlpool_close(&ctx.whirlpool, hash);
+   sph_whirlpool_init( &ctx.whirlpool );
+   sph_whirlpool(&ctx.whirlpool, hash, 64);
+   sph_whirlpool_close(&ctx.whirlpool, hash);
 
+   if ( work_restart[thr_id].restart ) return 0;
 //
 
-        sph_bmw512_init( &ctx.bmw);
-        sph_bmw512(&ctx.bmw, hash, 64);
-        sph_bmw512_close(&ctx.bmw, hash);
+   sph_bmw512_init( &ctx.bmw);
+   sph_bmw512(&ctx.bmw, hash, 64);
+   sph_bmw512_close(&ctx.bmw, hash);
 
 #if defined(__AES__)
-        init_groestl( &ctx.groestl, 64 );
-        update_and_final_groestl( &ctx.groestl, (char*)hash,
+   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);
+   sph_groestl512_init(&ctx.groestl );
+   sph_groestl512(&ctx.groestl, hash, 64);
+   sph_groestl512_close(&ctx.groestl, hash);
 #endif
 
-        sph_skein512_init( &ctx.skein);
-        sph_skein512(&ctx.skein, hash, 64);
-        sph_skein512_close(&ctx.skein, hash);
+   sph_skein512_init( &ctx.skein);
+   sph_skein512(&ctx.skein, hash, 64);
+   sph_skein512_close(&ctx.skein, hash);
 
-        sph_jh512_init( &ctx.jh);
-        sph_jh512(&ctx.jh, hash, 64);
-        sph_jh512_close(&ctx.jh, hash);
+   sph_jh512_init( &ctx.jh);
+   sph_jh512(&ctx.jh, hash, 64);
+   sph_jh512_close(&ctx.jh, hash);
 
-        sph_keccak512_init( &ctx.keccak );
-        sph_keccak512(&ctx.keccak, hash, 64);
-        sph_keccak512_close(&ctx.keccak, hash);
+   sph_keccak512_init( &ctx.keccak );
+   sph_keccak512(&ctx.keccak, hash, 64);
+   sph_keccak512_close(&ctx.keccak, hash);
 
-        init_luffa( &ctx.luffa, 512 );
-        update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
+   init_luffa( &ctx.luffa, 512 );
+   update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
                                 (const BitSequence*)hash, 64 );
 
-        cubehashInit( &ctx.cubehash, 512, 16, 32 );
-        cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
+   cubehashInit( &ctx.cubehash, 512, 16, 32 );
+   cubehashUpdateDigest( &ctx.cubehash, (byte*) hash,
                               (const byte*)hash, 64 );
 
-        sph_shavite512_init( &ctx.shavite );
-        sph_shavite512(&ctx.shavite, hash, 64);
-        sph_shavite512_close(&ctx.shavite, hash);
+   sph_shavite512_init( &ctx.shavite );
+   sph_shavite512(&ctx.shavite, hash, 64);
+   sph_shavite512_close(&ctx.shavite, hash);
 
-        init_sd( &ctx.simd, 512 );
-        update_final_sd( &ctx.simd, (BitSequence *)hash,
+   init_sd( &ctx.simd, 512 );
+   update_final_sd( &ctx.simd, (BitSequence *)hash,
                          (const BitSequence *)hash, 512 );
 
 #if defined(__AES__)
-        init_echo( &ctx.echo, 512 );
-        update_final_echo ( &ctx.echo, (BitSequence *)hash,
+   init_echo( &ctx.echo, 512 );
+   update_final_echo ( &ctx.echo, (BitSequence *)hash,
                             (const BitSequence *)hash, 512 );
 #else
-        sph_echo512_init( &ctx.echo );
-        sph_echo512(&ctx.echo, hash, 64);
-        sph_echo512_close(&ctx.echo, hash);
+   sph_echo512_init( &ctx.echo );
+   sph_echo512(&ctx.echo, hash, 64);
+   sph_echo512_close(&ctx.echo, hash);
 #endif
 
-        sph_hamsi512_init( &ctx.hamsi );
-        sph_hamsi512(&ctx.hamsi, hash, 64);
-        sph_hamsi512_close(&ctx.hamsi, hash);
+   sph_hamsi512_init( &ctx.hamsi );
+   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);
+   sph_fugue512_init( &ctx.fugue );
+   sph_fugue512(&ctx.fugue, hash, 64);
+   sph_fugue512_close(&ctx.fugue, hash);
 
-        sph_shabal512_init( &ctx.shabal );
-        sph_shabal512(&ctx.shabal, hash, 64);
-        sph_shabal512_close(&ctx.shabal, hash);
+   sph_shabal512_init( &ctx.shabal );
+   sph_shabal512(&ctx.shabal, hash, 64);
+   sph_shabal512_close(&ctx.shabal, hash);
 
-        sph_whirlpool_init( &ctx.whirlpool );
-        sph_whirlpool(&ctx.whirlpool, hash, 64);
-        sph_whirlpool_close(&ctx.whirlpool, hash);
+   sph_whirlpool_init( &ctx.whirlpool );
+   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 );
+   SHA512_Init( &ctx.sha512 );
+   SHA512_Update( &ctx.sha512, hash, 64 );
+   SHA512_Final( (unsigned char*) hash, &ctx.sha512 );
 
-        sph_haval256_5(&ctx.haval,(const void*) hash, 64);
-        sph_haval256_5_close(&ctx.haval, hash);
+   sph_haval256_5(&ctx.haval,(const void*) hash, 64);
+   sph_haval256_5_close(&ctx.haval, hash);
 
    memcpy(state, hash, 32);
-}
-
-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;
-      sonoa_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

algo/x17/x17-4way.c

@@ -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)));
@@ -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_lane_solution( work, lane_hash, mythr, lane );
-         }
-      }
-      *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)
@@ -300,7 +258,7 @@ 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)));
@@ -399,49 +357,8 @@ 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_lane_solution( work, lane_hash, mythr, lane );
-         }            
-      }
-      *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

algo/x17/x17-gate.c

@@ -3,13 +3,12 @@
 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;

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
 

algo/x17/x17.c

@@ -56,7 +56,7 @@ union _x17_context_overlay
 };
 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)));
@@ -143,36 +143,8 @@ void x17_hash(void *output, const void *input)
     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

algo/x17/xevan-4way.c

@@ -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)));
@@ -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_lane_solution( work, lane_hash, mythr, lane );
-         }
-      }
-      *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)
@@ -465,7 +423,7 @@ 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)));
@@ -540,10 +498,10 @@ void xevan_4way_hash( void *output, const void *input )
 
      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 );
+     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 );
 
      // Parallel 4way 32 bit
      intrlv_4x32( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
@@ -637,10 +595,10 @@ void xevan_4way_hash( void *output, const void *input )
 
      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 );
+     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 );
 
      intrlv_4x32( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
 
@@ -666,49 +624,8 @@ 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_lane_solution( work, lane_hash, mythr, lane );
-         }
-      }
-      *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

algo/x17/xevan-gate.c

@@ -3,14 +3,13 @@
 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;

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

algo/x17/xevan.c

@@ -83,7 +83,7 @@ void init_xevan_ctx()
 #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;
@@ -218,36 +218,8 @@ void xevan_hash(void *output, const void *input)
 	sph_haval256_5_close(&ctx.haval, hash);
 
 	memcpy(output, hash, 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;
-
-   mm128_bswap32_80( edata, pdata );
-
-   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;
+   return 1;
 }
 
 #endif

algo/x22/x22i-4way.c

@@ -493,7 +493,7 @@ int scanhash_x22i_8way( struct work* work, uint32_t max_nonce,
          if ( likely( fulltest( lane_hash, ptarget ) && !opt_benchmark ) )
          {
             pdata[19] = n + lane;
-            submit_lane_solution( work, lane_hash, mythr, lane );
+            submit_solution( work, lane_hash, mythr );
          }
       }
       n += 8;

algo/x22/x25x-4way.c

@@ -625,7 +625,7 @@ int scanhash_x25x_8way( struct work* work, uint32_t max_nonce,
          if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
          {
               pdata[19] = n + lane;
-              submit_lane_solution( work, lane_hash, mythr, lane );
+              submit_solution( work, lane_hash, mythr );
          }
       }
       n += 8;

algo/yescrypt/yescrypt-platform.h

@@ -31,6 +31,7 @@
 #undef HUGEPAGE_SIZE
 #endif
 
+/*
 static __inline uint32_t
 le32dec(const void *pp)
 {
@@ -50,6 +51,7 @@ le32enc(void *pp, uint32_t x)
 	p[2] = (x >> 16) & 0xff;
 	p[3] = (x >> 24) & 0xff;
 }
+*/
 
 static void *
 alloc_region(yescrypt_region_t * region, size_t size)
@@ -154,7 +156,7 @@ int yescrypt_init_shared(yescrypt_shared_t * shared, const uint8_t * param, size
 	if (yescrypt_kdf(&dummy, shared1,
 	    param, paramlen, NULL, 0, N, r, p, 0,
 	    YESCRYPT_RW | YESCRYPT_PARALLEL_SMIX | __YESCRYPT_INIT_SHARED_1,
-	    salt, sizeof(salt)))
+	    salt, sizeof(salt), 0 ) )
 		goto out;
 
 	half1 = half2 = *shared;
@@ -166,19 +168,19 @@ int yescrypt_init_shared(yescrypt_shared_t * shared, const uint8_t * param, size
 	if (p > 1 && yescrypt_kdf(&half1, &half2.shared1,
 	    param, paramlen, salt, sizeof(salt), N, r, p, 0,
 	    YESCRYPT_RW | YESCRYPT_PARALLEL_SMIX | __YESCRYPT_INIT_SHARED_2,
-	    salt, sizeof(salt)))
+	    salt, sizeof(salt), 0 ))
 		goto out;
 
 	if (yescrypt_kdf(&half2, &half1.shared1,
 	    param, paramlen, salt, sizeof(salt), N, r, p, 0,
 	    YESCRYPT_RW | YESCRYPT_PARALLEL_SMIX | __YESCRYPT_INIT_SHARED_1,
-	    salt, sizeof(salt)))
+	    salt, sizeof(salt), 0))
 		goto out;
 
 	if (yescrypt_kdf(&half1, &half2.shared1,
 	    param, paramlen, salt, sizeof(salt), N, r, p, 0,
 	    YESCRYPT_RW | YESCRYPT_PARALLEL_SMIX | __YESCRYPT_INIT_SHARED_1,
-	    buf, buflen))
+	    buf, buflen, 0))
 		goto out;
 
 	shared->mask1 = mask;

algo/yescrypt/yescrypt-simd.c

@@ -1149,7 +1149,7 @@ yescrypt_kdf(const yescrypt_shared_t * shared, yescrypt_local_t * local,
     const uint8_t * passwd, size_t passwdlen,
     const uint8_t * salt, size_t saltlen,
     uint64_t N, uint32_t r, uint32_t p, uint32_t t, yescrypt_flags_t flags,
-    uint8_t * buf, size_t buflen)
+    uint8_t * buf, size_t buflen, int thrid )
 {
 	uint8_t _ALIGN(128) sha256[32];
 	yescrypt_region_t tmp;
@@ -1157,6 +1157,7 @@ yescrypt_kdf(const yescrypt_shared_t * shared, yescrypt_local_t * local,
 	size_t B_size, V_size, XY_size, need;
 	uint8_t * B, * S;
 	salsa20_blk_t * V, * XY;
+   int retval = 1;
 
 	/*
 	 * YESCRYPT_PARALLEL_SMIX is a no-op at p = 1 for its intended purpose,
@@ -1312,6 +1313,12 @@ yescrypt_kdf(const yescrypt_shared_t * shared, yescrypt_local_t * local,
 	/* 1: (B_0 ... B_{p-1}) <-- PBKDF2(P, S, 1, p * MFLen) */
 	PBKDF2_SHA256(passwd, passwdlen, salt, saltlen, 1, B, B_size);
 
+   if ( work_restart[thrid].restart ) 
+   { 
+     retval = 0; 
+     goto out;
+   }
+   
 	if (t || flags)
 		memcpy(sha256, B, sizeof(sha256));
 
@@ -1339,9 +1346,21 @@ yescrypt_kdf(const yescrypt_shared_t * shared, yescrypt_local_t * local,
 		}
 	}
 
+   if ( work_restart[thrid].restart )
+   {
+     retval = 0;
+     goto out;
+   }
+
 	/* 5: DK <-- PBKDF2(P, B, 1, dkLen) */
 	PBKDF2_SHA256(passwd, passwdlen, B, B_size, 1, buf, buflen);
 
+   if ( work_restart[thrid].restart ) 
+   { 
+     retval = 0; 
+     goto out;
+   }
+   
 	/*
 	 * Except when computing classic scrypt, allow all computation so far
 	 * to be performed on the client.  The final steps below match those of
@@ -1370,9 +1389,10 @@ yescrypt_kdf(const yescrypt_shared_t * shared, yescrypt_local_t * local,
 	   }
 	}
 
+out:   
 	if (free_region(&tmp))
 		return -1;
 
 	/* Success! */
-	return 0;
+	return retval;
 }

algo/yescrypt/yescrypt.c

@@ -106,7 +106,8 @@ static const uint8_t* decode64_uint32(uint32_t* dst, uint32_t dstbits, const uin
 }
 
 uint8_t* yescrypt_r(const yescrypt_shared_t* shared, yescrypt_local_t* local,
-    const uint8_t* passwd, size_t passwdlen, const uint8_t* setting, uint8_t* buf, size_t buflen)
+    const uint8_t* passwd, size_t passwdlen, const uint8_t* setting,
+    uint8_t* buf, size_t buflen, int thrid )
 {
 	uint8_t hash[HASH_SIZE];
 	const uint8_t * src, * salt;
@@ -210,7 +211,9 @@ uint8_t* yescrypt_r(const yescrypt_shared_t* shared, yescrypt_local_t* local,
 		return NULL;
 	}
 
-	if (yescrypt_kdf(shared, local, passwd, passwdlen, salt, saltlen, N, r, p, 0, flags, hash, sizeof(hash))) {
+	if ( yescrypt_kdf( shared, local, passwd, passwdlen, salt, saltlen, N, r, p,
+            0, flags, hash, sizeof(hash), thrid ) == -1 )
+   {
 		printf("died10 ...");
 		fflush(stdout);
 		return NULL;
@@ -237,7 +240,7 @@ uint8_t* yescrypt_r(const yescrypt_shared_t* shared, yescrypt_local_t* local,
 	return buf;
 }
 
-uint8_t* yescrypt(const uint8_t* passwd, const uint8_t* setting)
+uint8_t* yescrypt(const uint8_t* passwd, const uint8_t* setting, int thrid )
 {
 	static uint8_t buf[4 + 1 + 5 + 5 + BYTES2CHARS(32) + 1 + HASH_LEN + 1];
 	yescrypt_shared_t shared;
@@ -252,7 +255,7 @@ uint8_t* yescrypt(const uint8_t* passwd, const uint8_t* setting)
 		return NULL;
 	}
 	retval = yescrypt_r(&shared, &local,
-	    passwd, 80, setting, buf, sizeof(buf));
+	    passwd, 80, setting, buf, sizeof(buf), thrid );
 	//printf("hashse='%s'\n", (char *)retval);
 	if (yescrypt_free_local(&local)) {
 		yescrypt_free_shared(&shared);
@@ -329,7 +332,7 @@ uint8_t* yescrypt_gensalt(uint32_t N_log2, uint32_t r, uint32_t p, yescrypt_flag
 
 static int yescrypt_bsty(const uint8_t * passwd, size_t passwdlen,
     const uint8_t * salt, size_t saltlen, uint64_t N, uint32_t r, uint32_t p,
-    uint8_t * buf, size_t buflen)
+    uint8_t * buf, size_t buflen, int thrid )
 {
 	static __thread int initialized = 0;
 	static __thread yescrypt_shared_t shared;
@@ -349,7 +352,7 @@ static int yescrypt_bsty(const uint8_t * passwd, size_t passwdlen,
 	}
 	retval = yescrypt_kdf(&shared, &local,
 	    passwd, passwdlen, salt, saltlen, N, r, p, 0, YESCRYPT_FLAGS,
-	    buf, buflen);
+	    buf, buflen, thrid );
 #if 0
 	if (yescrypt_free_local(&local)) {
 		yescrypt_free_shared(&shared);
@@ -370,16 +373,16 @@ char *yescrypt_client_key = NULL;
 int yescrypt_client_key_len = 0;
 
 /* main hash 80 bytes input */
-void yescrypt_hash( const char *input, char *output, uint32_t len )
+int yescrypt_hash( const char *input, char *output, uint32_t len, int thrid )
 {
-   yescrypt_bsty( (uint8_t*)input, len, (uint8_t*)input, len, YESCRYPT_N,
-                  YESCRYPT_R, YESCRYPT_P, (uint8_t*)output, 32 );
+   return yescrypt_bsty( (uint8_t*)input, len, (uint8_t*)input, len, YESCRYPT_N,
+                  YESCRYPT_R, YESCRYPT_P, (uint8_t*)output, 32, thrid );
 }
 
 /* for util.c test */
-void yescrypthash(void *output, const void *input)
+int yescrypthash(void *output, const void *input, int thrid)
 {
-	yescrypt_hash((char*) input, (char*) output, 80);
+	return yescrypt_hash((char*) input, (char*) output, 80, thrid);
 }
 
 int scanhash_yescrypt( struct work *work, uint32_t max_nonce,
@@ -392,13 +395,13 @@ int scanhash_yescrypt( struct work *work, uint32_t max_nonce,
    const uint32_t first_nonce = pdata[19];
    const uint32_t last_nonce = max_nonce;
    uint32_t n = first_nonce;
-   int thr_id = mythr->id;  // thr_id arg is deprecated
+   int thr_id = mythr->id; 
 
    for ( int k = 0; k < 19; k++ )
       be32enc( &endiandata[k], pdata[k] );
    endiandata[19] = n;
    do {
-      yescrypt_hash((char*) endiandata, (char*) vhash, 80);
+      if ( yescrypt_hash((char*) endiandata, (char*) vhash, 80, thr_id ) )
       if unlikely( valid_hash( vhash, ptarget ) && !opt_benchmark )
       {
           be32enc( pdata+19, n );
@@ -442,7 +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 );

algo/yescrypt/yescrypt.h

@@ -38,12 +38,13 @@ extern "C" {
 #include <stdint.h>
 #include <stdlib.h> /* for size_t */
 #include <stdbool.h>
+#include "miner.h"
 
 //#define  __SSE4_1__
 
-void yescrypt_hash(const char* input, char* output, uint32_t len);
+int yescrypt_hash(const char* input, char* output, uint32_t len, int thrid );
 
-void yescrypthash(void *output, const void *input);
+int yescrypthash(void *output, const void *input, int thrid );
 
 /**
  * crypto_scrypt(passwd, passwdlen, salt, saltlen, N, r, p, buf, buflen):
@@ -301,7 +302,7 @@ extern int yescrypt_kdf(const yescrypt_shared_t * __shared,
     const uint8_t * __salt, size_t __saltlen,
     uint64_t __N, uint32_t __r, uint32_t __p, uint32_t __t,
     yescrypt_flags_t __flags,
-    uint8_t * __buf, size_t __buflen);
+    uint8_t * __buf, size_t __buflen, int thrid);
 
 /**
  * yescrypt_r(shared, local, passwd, passwdlen, setting, buf, buflen):
@@ -321,7 +322,7 @@ extern uint8_t * yescrypt_r(const yescrypt_shared_t * __shared,
     yescrypt_local_t * __local,
     const uint8_t * __passwd, size_t __passwdlen,
     const uint8_t * __setting,
-    uint8_t * __buf, size_t __buflen);
+    uint8_t * __buf, size_t __buflen, int thrid);
 
 /**
  * yescrypt(passwd, setting):
@@ -339,7 +340,7 @@ extern uint8_t * yescrypt_r(const yescrypt_shared_t * __shared,
  *
  * MT-unsafe.
  */
-extern uint8_t * yescrypt(const uint8_t * __passwd, const uint8_t * __setting);
+extern uint8_t * yescrypt(const uint8_t * __passwd, const uint8_t * __setting, int thrid );
 
 /**
  * yescrypt_gensalt_r(N_log2, r, p, flags, src, srclen, buf, buflen):

algo/yespower/yescrypt-r8g.c

@@ -51,6 +51,10 @@ int scanhash_yespower_r8g( struct work *work, uint32_t max_nonce,
        be32enc( &endiandata[ i], pdata[ i ]);
     endiandata[19] = n;
 
+// do sha256 prehash
+   SHA256_Init( &sha256_prehash_ctx );
+   SHA256_Update( &sha256_prehash_ctx, endiandata, 64 );
+    
     do {
        yespower_tls( (unsigned char *)endiandata, params.perslen,
                       &params, (yespower_binary_t*)hash, thr_id );

algo/yespower/yespower-4way.c

@@ -0,0 +1,692 @@
+/*-
+ * Copyright 2009 Colin Percival
+ * Copyright 2013-2018 Alexander Peslyak
+ * 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.
+ *
+ * This file was originally written by Colin Percival as part of the Tarsnap
+ * online backup system.
+ *
+ * This is a proof-of-work focused fork of yescrypt, including reference and
+ * cut-down implementation of the obsolete yescrypt 0.5 (based off its first
+ * submission to PHC back in 2014) and a new proof-of-work specific variation
+ * known as yespower 1.0.  The former is intended as an upgrade for
+ * cryptocurrencies that already use yescrypt 0.5 and the latter may be used
+ * as a further upgrade (hard fork) by those and other cryptocurrencies.  The
+ * version of algorithm to use is requested through parameters, allowing for
+ * both algorithms to co-exist in client and miner implementations (such as in
+ * preparation for a hard-fork).
+ *
+ * This is the reference implementation.  Its purpose is to provide a simple
+ * human- and machine-readable specification that implementations intended
+ * for actual use should be tested against.  It is deliberately mostly not
+ * optimized, and it is not meant to be used in production.  Instead, use
+ * yespower-opt.c.
+ */
+/*
+#warning "This reference implementation is deliberately mostly not optimized. Use yespower-opt.c instead unless you're testing (against) the reference implementation on purpose."
+*/
+#include <errno.h>
+#include <stdint.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "algo/sha/hmac-sha256-hash-4way.h"
+//#include "sysendian.h"
+
+#include "yespower.h"
+
+
+#if defined(__AVX2__)
+
+
+static void blkcpy_8way( __m256i *dst, const __m256i *src, size_t count )
+{
+	do {
+		*dst++ = *src++;
+	} while (--count);
+}
+
+static void blkxor_8way( __m256i *dst, const __m256i *src, size_t count )
+{
+	do {
+		*dst++ ^= *src++;
+	} while (--count);
+}
+
+/**
+ * salsa20(B):
+ * Apply the Salsa20 core to the provided block.
+ */
+static void salsa20_8way( __m256i B[16], uint32_t rounds )
+{
+	__m256i x[16];
+	size_t i;
+
+	/* SIMD unshuffle */
+	for ( i = 0; i < 16; i++ )
+		x[i * 5 % 16] = B[i];
+
+	for ( i = 0; i < rounds; i += 2 )
+   {
+#define R( a, b, c ) mm256_rol_32( _mm256_add_epi32( a, b ), c )
+      /* Operate on columns */
+
+      x[ 4] = _mm256_xor_si256( x[ 4], R( x[ 0], x[12],  7 ) );
+      x[ 8] = _mm256_xor_si256( x[ 8], R( x[ 4], x[ 0],  9 ) );
+      x[12] = _mm256_xor_si256( x[12], R( x[ 8], x[ 4], 13 ) );
+      x[ 0] = _mm256_xor_si256( x[ 0], R( x[12], x[ 8], 18 ) );
+
+      x[ 9] = _mm256_xor_si256( x[ 9], R( x[ 5], x[ 1],  7 ) );
+      x[13] = _mm256_xor_si256( x[13], R( x[ 9], x[ 5],  9 ) );
+      x[ 1] = _mm256_xor_si256( x[ 1], R( x[13], x[ 9], 13 ) );
+      x[ 5] = _mm256_xor_si256( x[ 5], R( x[ 1], x[13], 18 ) );
+
+      x[14] = _mm256_xor_si256( x[14], R( x[10], x[ 6],  7 ) );
+      x[ 2] = _mm256_xor_si256( x[ 2], R( x[14], x[10],  9 ) );
+      x[ 6] = _mm256_xor_si256( x[ 6], R( x[ 2], x[14], 13 ) );
+      x[10] = _mm256_xor_si256( x[10], R( x[ 6], x[ 2], 18 ) );
+
+      x[ 3] = _mm256_xor_si256( x[ 3], R( x[15], x[11],  7 ) );
+      x[ 7] = _mm256_xor_si256( x[ 7], R( x[ 3], x[15],  9 ) );
+      x[11] = _mm256_xor_si256( x[11], R( x[ 7], x[ 3], 13 ) );
+      x[15] = _mm256_xor_si256( x[15], R( x[11], x[ 7], 18 ) );
+
+		/* Operate on rows */
+
+      x[ 1] = _mm256_xor_si256( x[ 1], R( x[ 0], x[ 3],  7 ) );
+      x[ 2] = _mm256_xor_si256( x[ 2], R( x[ 1], x[ 0],  9 ) );
+      x[ 3] = _mm256_xor_si256( x[ 3], R( x[ 2], x[ 1], 13 ) );
+      x[ 0] = _mm256_xor_si256( x[ 0], R( x[ 3], x[ 2], 18 ) );
+
+      x[ 6] = _mm256_xor_si256( x[ 6], R( x[ 5], x[ 4],  7 ) );
+      x[ 7] = _mm256_xor_si256( x[ 7], R( x[ 6], x[ 5],  9 ) );
+      x[ 4] = _mm256_xor_si256( x[ 4], R( x[ 7], x[ 6], 13 ) );
+      x[ 5] = _mm256_xor_si256( x[ 5], R( x[ 4], x[ 7], 18 ) );
+
+      x[11] = _mm256_xor_si256( x[11], R( x[10], x[ 9],  7 ) );
+      x[ 8] = _mm256_xor_si256( x[ 8], R( x[11], x[10],  9 ) );
+      x[ 9] = _mm256_xor_si256( x[ 9], R( x[ 8], x[11], 13 ) );
+      x[10] = _mm256_xor_si256( x[10], R( x[ 9], x[ 8], 18 ) );
+
+      x[12] = _mm256_xor_si256( x[12], R( x[15], x[14],  7 ) );
+      x[13] = _mm256_xor_si256( x[13], R( x[12], x[15],  9 ) );
+      x[14] = _mm256_xor_si256( x[14], R( x[13], x[12], 13 ) );
+      x[15] = _mm256_xor_si256( x[15], R( x[14], x[13], 18 ) );
+
+#undef R
+	}
+
+	/* SIMD shuffle */
+	for (i = 0; i < 16; i++)
+		B[i] = _mm256_add_epi32( B[i], x[i * 5 % 16] );
+}
+
+/**
+ * blockmix_salsa(B):
+ * Compute B = BlockMix_{salsa20, 1}(B).  The input B must be 128 bytes in
+ * length.
+ */
+static void blockmix_salsa_8way( __m256i *B, uint32_t rounds )
+{
+	__m256i X[16];
+	size_t i;
+
+	/* 1: X <-- B_{2r - 1} */
+	blkcpy_8way( X, &B[16], 16 );
+
+	/* 2: for i = 0 to 2r - 1 do */
+	for ( i = 0; i < 2; i++ )
+   {
+		/* 3: X <-- H(X xor B_i) */
+		blkxor_8way( X, &B[i * 16], 16 );
+		salsa20_8way( X, rounds );
+
+		/* 4: Y_i <-- X */
+		/* 6: B' <-- (Y_0, Y_2 ... Y_{2r-2}, Y_1, Y_3 ... Y_{2r-1}) */
+		blkcpy_8way( &B[i * 16], X, 16 );
+	}
+}
+
+/*
+ * These are tunable, but they must meet certain constraints and are part of
+ * what defines a yespower version.
+ */
+#define PWXsimple 2
+#define PWXgather 4
+/* Version 0.5 */
+#define PWXrounds_0_5 6
+#define Swidth_0_5 8
+/* Version 1.0 */
+#define PWXrounds_1_0 3
+#define Swidth_1_0 11
+
+/* Derived values.  Not tunable on their own. */
+#define PWXbytes (PWXgather * PWXsimple * 8)
+#define PWXwords (PWXbytes / sizeof(uint32_t))
+#define rmin ((PWXbytes + 127) / 128)
+
+/* Runtime derived values.  Not tunable on their own. */
+#define Swidth_to_Sbytes1(Swidth) ((1 << Swidth) * PWXsimple * 8)
+#define Swidth_to_Smask(Swidth) (((1 << Swidth) - 1) * PWXsimple * 8)
+
+typedef struct {
+   __m256i (*S0)[2], (*S1)[2], (*S2)[2];
+   __m256i *S;
+	yespower_version_t version;
+	uint32_t salsa20_rounds;
+	uint32_t PWXrounds, Swidth, Sbytes, Smask;
+	size_t w;
+} pwxform_8way_ctx_t __attribute__ ((aligned (128)));
+
+/**
+ * pwxform(B):
+ * Transform the provided block using the provided S-boxes.
+ */
+static void pwxform_8way( __m256i *B, pwxform_8way_ctx_t *ctx )
+{
+	__m256i (*X)[PWXsimple][2] = (__m256i (*)[PWXsimple][2])B;
+	__m256i (*S0)[2] = ctx->S0, (*S1)[2] = ctx->S1, (*S2)[2] = ctx->S2;
+	__m256i Smask = _mm256_set1_epi32( ctx->Smask );
+	size_t w = ctx->w;
+	size_t i, j, k;
+
+	/* 1: for i = 0 to PWXrounds - 1 do */
+	for ( i = 0; i < ctx->PWXrounds; i++ )
+   {
+		/* 2: for j = 0 to PWXgather - 1 do */
+		for ( j = 0; j < PWXgather; j++ )
+      {
+// Are these pointers or data?
+         __m256i xl = X[j][0][0];
+			__m256i xh = X[j][0][1];
+			__m256i (*p0)[2], (*p1)[2];
+
+			// 3: p0 <-- (lo(B_{j,0}) & Smask) / (PWXsimple * 8) 
+
+// playing with pointers
+/*
+         p0 = S0 + (xl & Smask) / sizeof(*S0);
+			// 4: p1 <-- (hi(B_{j,0}) & Smask) / (PWXsimple * 8) 
+			p1 = S1 + (xh & Smask) / sizeof(*S1);
+*/
+			/* 5: for k = 0 to PWXsimple - 1 do */
+			for ( k = 0; k < PWXsimple; k++ )
+         {
+
+// shift from 32 bit data to 64 bit data
+            __m256i x0, x1, s00, s01, s10, s11;
+            __m128i *p0k = (__m128i*)p0[k];
+            __m128i *p1k = (__m128i*)p1[k];
+
+
+           s00 = _mm256_add_epi64( _mm256_cvtepu32_epi64( p0k[0] ),
+                _mm256_slli_epi64( _mm256_cvtepu32_epi64( p0k[2] ), 32 ) );
+           s01 = _mm256_add_epi64( _mm256_cvtepu32_epi64( p0k[1] ),
+                _mm256_slli_epi64( _mm256_cvtepu32_epi64( p0k[3] ), 32 ) );
+           s10 = _mm256_add_epi64( _mm256_cvtepu32_epi64( p1k[0] ),
+                _mm256_slli_epi64( _mm256_cvtepu32_epi64( p1k[2] ), 32 ) );
+           s11 = _mm256_add_epi64( _mm256_cvtepu32_epi64( p1k[1] ),
+                _mm256_slli_epi64( _mm256_cvtepu32_epi64( p1k[3] ), 32 ) );
+
+            __m128i *xx = (__m128i*)X[j][k];
+            x0 = _mm256_mul_epu32( _mm256_cvtepu32_epi64( xx[0] ),
+                                   _mm256_cvtepu32_epi64( xx[2] ) );
+            x1 = _mm256_mul_epu32( _mm256_cvtepu32_epi64( xx[1] ),
+                                   _mm256_cvtepu32_epi64( xx[3] ) );
+
+            x0 = _mm256_add_epi64( x0, s00 );
+            x1 = _mm256_add_epi64( x1, s01 );
+            
+            x0 = _mm256_xor_si256( x0, s10 );
+            x1 = _mm256_xor_si256( x1, s11 );
+
+            X[j][k][0] = x0; 
+            X[j][k][1] = x1;                        
+			}
+
+			if ( ctx->version != YESPOWER_0_5 &&
+			    ( i == 0 || j < PWXgather / 2 ) )
+         {
+				if ( j & 1 )
+            {
+					for ( k = 0; k < PWXsimple; k++ )
+               {
+						S1[w][0] = X[j][k][0];
+						S1[w][1] = X[j][k][1];
+						w++;
+					}
+				}
+            else
+            {
+					for ( k = 0; k < PWXsimple; k++ )
+               {
+						S0[w + k][0] = X[j][k][0];
+						S0[w + k][1] = X[j][k][1];
+					}
+				}
+			}
+		}
+	}
+
+	if ( ctx->version != YESPOWER_0_5 )
+   {
+		/* 14: (S0, S1, S2) <-- (S2, S0, S1) */
+		ctx->S0 = S2;
+		ctx->S1 = S0;
+		ctx->S2 = S1;
+		/* 15: w <-- w mod 2^Swidth */
+		ctx->w = w & ( ( 1 << ctx->Swidth ) * PWXsimple - 1 );
+	}
+}
+
+/**
+ * blockmix_pwxform(B, ctx, r):
+ * Compute B = BlockMix_pwxform{salsa20, ctx, r}(B).  The input B must be
+ * 128r bytes in length.
+ */
+static void blockmix_pwxform_8way( uint32_t *B, pwxform_8way_ctx_t *ctx,
+                                   size_t r )
+{
+	__m256i X[PWXwords];
+	size_t r1, i;
+
+	/* Convert 128-byte blocks to PWXbytes blocks */
+	/* 1: r_1 <-- 128r / PWXbytes */
+	r1 = 128 * r / PWXbytes;
+
+	/* 2: X <-- B'_{r_1 - 1} */
+	blkcpy_8way( X, &B[ (r1 - 1) * PWXwords ], PWXwords );
+
+	/* 3: for i = 0 to r_1 - 1 do */
+	for ( i = 0; i < r1; i++ )
+   {
+		/* 4: if r_1 > 1 */
+		if ( r1 > 1 )
+      {
+			/* 5: X <-- X xor B'_i */
+			blkxor_8way( X, &B[ i * PWXwords ], PWXwords );
+		}
+
+		/* 7: X <-- pwxform(X) */
+		pwxform_8way( X, ctx );
+
+		/* 8: B'_i <-- X */
+		blkcpy_8way( &B[ i * PWXwords ], X, PWXwords );
+	}
+
+	/* 10: i <-- floor((r_1 - 1) * PWXbytes / 64) */
+	i = ( r1 - 1 ) * PWXbytes / 64;
+
+	/* 11: B_i <-- H(B_i) */
+	salsa20_8way( &B[i * 16], ctx->salsa20_rounds );
+
+#if 1 /* No-op with our current pwxform settings, but do it to make sure */
+	/* 12: for i = i + 1 to 2r - 1 do */
+	for ( i++; i < 2 * r; i++ )
+   {
+		/* 13: B_i <-- H(B_i xor B_{i-1}) */
+		blkxor_8way( &B[i * 16], &B[ (i - 1) * 16 ], 16 );
+		salsa20_8way( &B[i * 16], ctx->salsa20_rounds );
+	}
+#endif
+}
+
+// This looks a lot like data dependent addressing
+
+/**
+ * integerify(B, r):
+ * Return the result of parsing B_{2r-1} as a little-endian integer.
+ */
+static __m256i integerify8( const __m256i *B, size_t r )
+{
+/*
+ * Our 32-bit words are in host byte order.  Also, they are SIMD-shuffled, but
+ * we only care about the least significant 32 bits anyway.
+ */
+	const __m256i *X = &B[ (2 * r - 1) * 16 ];
+	return X[0];
+}
+
+/**
+ * p2floor(x):
+ * Largest power of 2 not greater than argument.
+ */
+static uint32_t p2floor8( uint32_t x )
+{
+	uint32_t y;
+	while ( ( y = x & (x - 1) ) )
+		x = y;
+	return x;
+}
+
+/**
+ * wrap(x, i):
+ * Wrap x to the range 0 to i-1.
+ */
+static uint32_t wrap8( uint32_t x, uint32_t i )
+{
+	uint32_t n = p2floor( i );
+	return ( x & (n - 1) ) + (i - n);
+}
+
+/**
+ * smix1(B, r, N, V, X, ctx):
+ * Compute first loop of B = SMix_r(B, N).  The input B must be 128r bytes in
+ * length; the temporary storage V must be 128rN bytes in length; the temporary
+ * storage X must be 128r bytes in length.
+ */
+static void smix1_8way( __m256i *B, size_t r, uint32_t N,
+                        __m256i *V, __m256i *X, pwxform_8way_ctx_t *ctx )
+{
+	size_t s = 32 * r;
+	uint32_t i, j;
+	size_t k;
+
+	/* 1: X <-- B */
+	for ( k = 0; k < 2 * r; k++ )
+		for ( i = 0; i < 16; i++ )
+			X[ k * 16 + i ] = B[ k * 16 + ( i * 5 % 16 ) ];
+
+	if ( ctx->version != YESPOWER_0_5 )
+   {
+		for ( k = 1; k < r; k++ )
+      {
+			blkcpy_8way( &X[k * 32], &X[ (k - 1) * 32 ], 32 );
+			blockmix_pwxform_8way( &X[k * 32], ctx, 1 );
+		}
+	}
+
+	/* 2: for i = 0 to N - 1 do */
+	for ( i = 0; i < N; i++ )
+   {
+		/* 3: V_i <-- X */
+		blkcpy_8way( &V[i * s], X, s );
+
+		if ( i > 1 )
+      {
+
+// is j int or vector? Integrify has data dependent addressing?
+
+         /* j <-- Wrap(Integerify(X), i) */
+//			j = wrap8( integerify8( X, r ), i );
+
+			/* X <-- X xor V_j */
+			blkxor_8way( X, &V[j * s], s );
+		}
+
+		/* 4: X <-- H(X) */
+		if ( V != ctx->S )
+			blockmix_pwxform_8way( X, ctx, r );
+		else
+			blockmix_salsa_8way( X, ctx->salsa20_rounds );
+	}
+
+	/* B' <-- X */
+	for ( k = 0; k < 2 * r; k++ )
+		for ( i = 0; i < 16; i++ )
+			B[ k * 16 + ( i * 5 % 16 ) ] = X[ k * 16 + i ];
+}
+
+/**
+ * smix2(B, r, N, Nloop, V, X, ctx):
+ * Compute second loop of B = SMix_r(B, N).  The input B must be 128r bytes in
+ * length; the temporary storage V must be 128rN bytes in length; the temporary
+ * storage X must be 128r bytes in length.  The value N must be a power of 2
+ * greater than 1.
+ */
+static void smix2_8way( __m256i *B, size_t r, uint32_t N, uint32_t Nloop,
+                        __m256i *V, __m256i *X, pwxform_8way_ctx_t *ctx )
+{
+	size_t s = 32 * r;
+	uint32_t i, j;
+	size_t k;
+
+	/* X <-- B */
+	for ( k = 0; k < 2 * r; k++ )
+		for ( i = 0; i < 16; i++ )
+			X[ k * 16 + i ] = B[ k * 16 + ( i * 5 % 16 ) ];
+
+	/* 6: for i = 0 to N - 1 do */
+	for ( i = 0; i < Nloop; i++ )
+   {
+		/* 7: j <-- Integerify(X) mod N */
+//		j = integerify8(X, r) & (N - 1);
+
+		/* 8.1: X <-- X xor V_j */
+		blkxor_8way( X, &V[j * s], s );
+		/* V_j <-- X */
+		if ( Nloop != 2 )
+			blkcpy_8way( &V[j * s], X, s );
+
+		/* 8.2: X <-- H(X) */
+		blockmix_pwxform_8way( X, ctx, r );
+	}
+
+	/* 10: B' <-- X */
+	for ( k = 0; k < 2 * r; k++ )
+		for ( i = 0; i < 16; i++ )
+			B[ k * 16 + ( i * 5 % 16 ) ] = X[ k * 16 + i ];
+}
+
+/**
+ * smix(B, r, N, p, t, V, X, ctx):
+ * Compute B = SMix_r(B, N).  The input B must be 128rp bytes in length; the
+ * temporary storage V must be 128rN bytes in length; the temporary storage
+ * X must be 128r bytes in length.  The value N must be a power of 2 and at
+ * least 16.
+ */
+static void smix_8way( __m256i *B, size_t r, uint32_t N,
+                       __m256i *V, __m256i *X, pwxform_8way_ctx_t *ctx)
+{
+	uint32_t Nloop_all = (N + 2) / 3; /* 1/3, round up */
+	uint32_t Nloop_rw = Nloop_all;
+
+	Nloop_all++; Nloop_all &= ~(uint32_t)1; /* round up to even */
+
+   if ( ctx->version == YESPOWER_0_5 )
+		Nloop_rw &= ~(uint32_t)1; /* round down to even */
+	else
+		Nloop_rw++; Nloop_rw &= ~(uint32_t)1; /* round up to even */
+
+	smix1_8way( B, 1, ctx->Sbytes / 128, ctx->S, X, ctx );
+	smix1_8way( B, r, N, V, X, ctx );
+	smix2_8way( B, r, N, Nloop_rw /* must be > 2 */, V, X, ctx );
+	smix2_8way( B, r, N, Nloop_all - Nloop_rw /* 0 or 2 */, V, X, ctx );
+}
+
+/**
+ * yespower(local, src, srclen, params, dst):
+ * Compute yespower(src[0 .. srclen - 1], N, r), to be checked for "< target".
+ *
+ * Return 0 on success; or -1 on error.
+ */
+int yespower_8way( yespower_local_t *local, const __m256i *src, size_t srclen,
+              const yespower_params_t *params, yespower_8way_binary_t *dst,
+              int thrid )
+{
+	yespower_version_t version = params->version;
+	uint32_t N = params->N;
+	uint32_t r = params->r;
+	const uint8_t *pers = params->pers;
+	size_t perslen = params->perslen;
+	int retval = -1;
+	size_t B_size, V_size;
+	uint32_t *B, *V, *X, *S;
+	pwxform_8way_ctx_t ctx;
+	__m256i sha256[8];
+
+	/* Sanity-check parameters */
+	if ( (version != YESPOWER_0_5 && version != YESPOWER_1_0 ) ||
+	    N < 1024 || N > 512 * 1024 || r < 8 || r > 32 ||
+	    (N & (N - 1)) != 0 || r < rmin ||
+	    (!pers && perslen) )
+   {
+		errno = EINVAL;
+		return -1;
+	}
+
+	/* Allocate memory */
+	B_size = (size_t)128 * r;
+	V_size = B_size * N;
+	if ((V = malloc(V_size)) == NULL)
+		return -1;
+	if ((B = malloc(B_size)) == NULL)
+		goto free_V;
+	if ((X = malloc(B_size)) == NULL)
+		goto free_B;
+	ctx.version = version;
+	if (version == YESPOWER_0_5) {
+		ctx.salsa20_rounds = 8;
+		ctx.PWXrounds = PWXrounds_0_5;
+		ctx.Swidth = Swidth_0_5;
+		ctx.Sbytes = 2 * Swidth_to_Sbytes1(ctx.Swidth);
+	} else {
+		ctx.salsa20_rounds = 2;
+		ctx.PWXrounds = PWXrounds_1_0;
+		ctx.Swidth = Swidth_1_0;
+		ctx.Sbytes = 3 * Swidth_to_Sbytes1(ctx.Swidth);
+	}
+	if ((S = malloc(ctx.Sbytes)) == NULL)
+		goto free_X;
+	ctx.S = S;
+	ctx.S0 = (__m256i (*)[2])S;
+	ctx.S1 = ctx.S0 + (1 << ctx.Swidth) * PWXsimple;
+	ctx.S2 = ctx.S1 + (1 << ctx.Swidth) * PWXsimple;
+	ctx.Smask = Swidth_to_Smask(ctx.Swidth);
+	ctx.w = 0;
+
+   // do prehash
+	sha256_8way_full( sha256, src, srclen );
+
+
+  // need flexible size, use malloc;
+   __m256i vpers[128];
+
+	if ( version != YESPOWER_0_5 && perslen )
+      for ( int i = 0; i < perslen/4 + 1; i++ )
+         vpers[i] = _mm256_set1_epi32( pers[i] );
+
+	/* 1: (B_0 ... B_{p-1}) <-- PBKDF2(P, S, 1, p * MFLen) */
+	pbkdf2_sha256_8way( B, B_size, sha256, sizeof(sha256), vpers, perslen, 1 );
+
+	blkcpy_8way( sha256, B, sizeof(sha256) / sizeof(sha256[0] ) );
+
+	/* 3: B_i <-- MF(B_i, N) */
+	smix_8way( B, r, N, V, X, &ctx );
+
+	if ( version == YESPOWER_0_5 )
+   {
+		/* 5: DK <-- PBKDF2(P, B, 1, dkLen) */
+		pbkdf2_sha256_8way( dst, sizeof(*dst), sha256, sizeof(sha256),
+                          B, B_size, 1 );
+
+		if ( pers )
+      {
+			hmac_sha256_8way_full( dst, sizeof(*dst), vpers, perslen, sha256 );
+			sha256_8way_full( dst, sha256, sizeof(sha256) );
+		}
+	}
+   else
+		hmac_sha256_8way_full( dst, B + B_size - 64, 64, sha256, sizeof(sha256) );
+
+	/* Success! */
+	retval = 1;
+
+	/* Free memory */
+	free(S);
+free_X:
+	free(X);
+free_B:
+	free(B);
+free_V:
+	free(V);
+
+	return retval;
+}
+
+int yespower_8way_tls( const __m256i *src, size_t srclen,
+    const yespower_params_t *params, yespower_8way_binary_t *dst, int trhid )
+{
+/* The reference implementation doesn't use thread-local storage */
+	return yespower_8way( NULL, src, srclen, params, dst, trhid );
+}
+
+int yespower_init_local8( yespower_local_t *local )
+{
+/* The reference implementation doesn't use the local structure */
+	local->base = local->aligned = NULL;
+	local->base_size = local->aligned_size = 0;
+	return 0;
+}
+
+int yespower_free_local8( yespower_local_t *local )
+{
+/* The reference implementation frees its memory in yespower() */
+	(void)local; /* unused */
+	return 0;
+}
+
+int yespower_8way_hash( const char *input, char *output, uint32_t len,
+                        int thrid )
+{
+   return yespower_8way_tls( input, len, &yespower_params,
+           (yespower_binary_t*)output, thrid );
+}
+
+int scanhash_yespower_8way( struct work *work, uint32_t max_nonce,
+                            uint64_t *hashes_done, struct thr_info *mythr )
+{
+   uint32_t _ALIGN(128) hash[8*8];
+   uint32_t _ALIGN(128) vdata[20*8];
+   uint32_t _ALIGN(128) endiandata[20];
+   uint32_t *pdata = work->data;
+   uint32_t *ptarget = work->target;
+   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce;
+   uint32_t n = first_nonce;
+   const int thr_id = mythr->id;
+
+   for ( int k = 0; k < 19; k++ )
+      be32enc( &endiandata[k], pdata[k] );
+   endiandata[19] = n;
+
+// do sha256 prehash
+   SHA256_Init( &sha256_prehash_ctx );
+   SHA256_Update( &sha256_prehash_ctx, endiandata, 64 );
+
+   do {
+      if ( yespower_hash( vdata, hash, 80, thr_id ) )
+      if unlikely( valid_hash( hash, ptarget ) && !opt_benchmark )
+      {
+          be32enc( pdata+19, n );
+          submit_solution( work, hash, mythr );
+      }
+      endiandata[19] = ++n;
+   } while ( n < last_nonce && !work_restart[thr_id].restart );
+   *hashes_done = n - first_nonce;
+   pdata[19] = n;
+   return 0;
+}
+
+#endif  // AVX2

algo/yespower/yespower-gate.c

@@ -30,7 +30,10 @@
 
 #include "algo-gate-api.h"
 
-static yespower_params_t yespower_params;
+yespower_params_t yespower_params;
+
+//SHA256_CTX sha256_prehash_ctx;
+__thread SHA256_CTX sha256_prehash_ctx;
 
 // YESPOWER
 
@@ -55,6 +58,11 @@ int scanhash_yespower( struct work *work, uint32_t max_nonce,
    for ( int k = 0; k < 19; k++ )
       be32enc( &endiandata[k], pdata[k] );
    endiandata[19] = n;
+
+// do sha256 prehash
+   SHA256_Init( &sha256_prehash_ctx );
+   SHA256_Update( &sha256_prehash_ctx, endiandata, 64 );
+
    do {
       if ( yespower_hash( (char*)endiandata, (char*)vhash, 80, thr_id ) )
       if unlikely( valid_hash( vhash, ptarget ) && !opt_benchmark )
@@ -86,11 +94,16 @@ int scanhash_yespower_b2b( struct work *work, uint32_t max_nonce,
    const uint32_t first_nonce = pdata[19];
    uint32_t n = first_nonce;
    const uint32_t last_nonce = max_nonce;
-   const int thr_id = mythr->id;  // thr_id arg is deprecated
+   const int thr_id = mythr->id;
 
    for ( int k = 0; k < 19; k++ )
       be32enc( &endiandata[k], pdata[k] );
    endiandata[19] = n;
+
+// do sha256 prehash
+   SHA256_Init( &sha256_prehash_ctx );
+   SHA256_Update( &sha256_prehash_ctx, endiandata, 64 );
+
    do {
       if (yespower_b2b_hash( (char*) endiandata, (char*) vhash, 80, thr_id ) )
       if unlikely( valid_hash( vhash, ptarget ) && !opt_benchmark )
@@ -126,7 +139,7 @@ bool register_yespower_algo( algo_gate_t* gate )
      yespower_params.perslen = 0;
   }
 
-  applog( LOG_NOTICE,"Yespower parameters: N= %d, R= %d.", yespower_params.N,
+  applog( LOG_NOTICE,"Yespower parameters: N= %d, R= %d", yespower_params.N,
                                                            yespower_params.r );
   if ( yespower_params.pers )
      applog( LOG_NOTICE,"Key= \"%s\"\n", yespower_params.pers );
@@ -152,7 +165,7 @@ bool register_yespowerr16_algo( algo_gate_t* gate )
   return true;
  };
 
-/* not used
+/* not used, doesn't work
 bool register_yescrypt_05_algo( algo_gate_t* gate )
 {
    gate->optimizations = SSE2_OPT | SHA_OPT;
@@ -166,6 +179,40 @@ bool register_yescrypt_05_algo( algo_gate_t* gate )
    return true;
 }
 
+bool register_yescrypt_05_algo( algo_gate_t* gate )
+{
+   gate->optimizations = SSE2_OPT | SHA_OPT;
+   gate->scanhash   = (void*)&scanhash_yespower;
+   yespower_params.version = YESPOWER_0_5;
+
+   if ( opt_param_n )  yespower_params.N = opt_param_n;
+   else                yespower_params.N = 2048;
+
+   if ( opt_param_r )  yespower_params.r = opt_param_r;
+   else                yespower_params.r = 8;
+
+   if ( opt_param_key )
+   {
+     yespower_params.pers = opt_param_key;
+     yespower_params.perslen = strlen( opt_param_key );
+   }
+   else
+   {
+     yespower_params.pers = NULL;
+     yespower_params.perslen = 0;
+   }
+
+//   YESCRYPT_P = 1;
+
+   applog( LOG_NOTICE,"Yescrypt parameters: N= %d, R= %d.",
+                                      yespower_params.N, yespower_params.r );
+   if ( yespower_params.pers )
+     applog( LOG_NOTICE,"Key= \"%s\"\n", yespower_params.pers );
+
+   return true;
+}
+
+
 bool register_yescryptr8_05_algo( algo_gate_t* gate )
 {
    gate->optimizations = SSE2_OPT | SHA_OPT;
@@ -217,12 +264,12 @@ bool register_power2b_algo( algo_gate_t* gate )
   yespower_params.pers = "Now I am become Death, the destroyer of worlds";
   yespower_params.perslen = 46;
 
-  applog( LOG_NOTICE,"yespower-b2b parameters: N= %d, R= %d.", yespower_params.N,
+  applog( LOG_NOTICE,"yespower-b2b parameters: N= %d, R= %d", yespower_params.N,
                                                            yespower_params.r );
   applog( LOG_NOTICE,"Key= \"%s\"", yespower_params.pers );
   applog( LOG_NOTICE,"Key length= %d\n", yespower_params.perslen );
 
-  gate->optimizations = SSE2_OPT | SHA_OPT;
+  gate->optimizations = SSE2_OPT;
   gate->scanhash      = (void*)&scanhash_yespower_b2b;
   gate->hash          = (void*)&yespower_b2b_hash;
   opt_target_factor = 65536.0;

algo/yespower/yespower-opt.c

@@ -96,6 +96,8 @@
 #include <stdlib.h>
 #include <string.h>
 #include "algo/sha/hmac-sha256-hash.h"
+#include "algo/sha/hmac-sha256-hash-4way.h"
+
 #include "yespower.h"
 #include "yespower-platform.c"
 
@@ -1038,12 +1040,13 @@ int yespower(yespower_local_t *local,
 	salsa20_blk_t *V, *XY;
 	pwxform_ctx_t ctx;
 	uint8_t sha256[32];
+   SHA256_CTX sha256_ctx;
 
 	/* Sanity-check parameters */
-	if ((version != YESPOWER_0_5 && version != YESPOWER_1_0) ||
-	    N < 1024 || N > 512 * 1024 || r < 8 || r > 32 ||
-	    (N & (N - 1)) != 0 ||
-	    (!pers && perslen)) {
+	if ( (version != YESPOWER_0_5 && version != YESPOWER_1_0)
+      || N < 1024 || N > 512 * 1024 || r < 8 || r > 32
+      || (N & (N - 1)) != 0 || ( !pers && perslen ) )
+   {
 		errno = EINVAL;
 		return -1;
 	}
@@ -1051,20 +1054,22 @@ int yespower(yespower_local_t *local,
 	/* Allocate memory */
 	B_size = (size_t)128 * r;
 	V_size = B_size * N;
-	if (version == YESPOWER_0_5) {
+	if ( version == YESPOWER_0_5 )
+   {
 		XY_size = B_size * 2;
 		Swidth = Swidth_0_5;
-		ctx.Sbytes = 2 * Swidth_to_Sbytes1(Swidth);
+		ctx.Sbytes = 2 * Swidth_to_Sbytes1( Swidth );
 	} else {
 		XY_size = B_size + 64;
 		Swidth = Swidth_1_0;
-		ctx.Sbytes = 3 * Swidth_to_Sbytes1(Swidth);
+		ctx.Sbytes = 3 * Swidth_to_Sbytes1( Swidth );
 	}
 	need = B_size + V_size + XY_size + ctx.Sbytes;
-	if (local->aligned_size < need) {
-		if (free_region(local))
+	if ( local->aligned_size < need )
+   {
+		if ( free_region( local ) )
 			return -1;
-		if (!alloc_region(local, need))
+		if ( !alloc_region( local, need ) )
 			return -1;
 	}
 	B = (uint8_t *)local->aligned;
@@ -1072,48 +1077,81 @@ int yespower(yespower_local_t *local,
 	XY = (salsa20_blk_t *)((uint8_t *)V + V_size);
 	S = (uint8_t *)XY + XY_size;
 	ctx.S0 = S;
-	ctx.S1 = S + Swidth_to_Sbytes1(Swidth);
+	ctx.S1 = S + Swidth_to_Sbytes1( Swidth );
 
-	SHA256_Buf(src, srclen, sha256);
 
-	if (version == YESPOWER_0_5) {
-		PBKDF2_SHA256(sha256, sizeof(sha256), src, srclen, 1,
-		    B, B_size);
+// copy prehash, do tail   
+   memcpy( &sha256_ctx, &sha256_prehash_ctx, sizeof sha256_ctx );
+   SHA256_Update( &sha256_ctx, src+64, srclen-64 );
+   SHA256_Final( sha256, &sha256_ctx );
 
-      if ( work_restart[thrid].restart ) return false;
+//   SHA256_Buf(src, srclen, sha256);
+
+	if ( version == YESPOWER_0_5 )
+   {
+      PBKDF2_SHA256( sha256, sizeof(sha256), src, srclen, 1, B, B_size );
+
+      if ( work_restart[thrid].restart ) return 0;
    
-      memcpy(sha256, B, sizeof(sha256));
-		smix(B, r, N, V, XY, &ctx);
+      memcpy( sha256, B, sizeof(sha256) );
+		smix( B, r, N, V, XY, &ctx );
 
-      if ( work_restart[thrid].restart ) return false;
+      if ( work_restart[thrid].restart ) return 0;
 
-      PBKDF2_SHA256(sha256, sizeof(sha256), B, B_size, 1,
-		    (uint8_t *)dst, sizeof(*dst));
+      PBKDF2_SHA256( sha256, sizeof(sha256), B, B_size, 1, (uint8_t *)dst,
+                     sizeof(*dst) );
 
-		if (pers) {
-			HMAC_SHA256_Buf(dst, sizeof(*dst), pers, perslen,
-			    sha256);
+      if ( work_restart[thrid].restart ) return 0;
 
-         if ( work_restart[thrid].restart ) return false;
+      if ( pers )
+      {
+         src = pers;
+         srclen = perslen;
+      }
+      else
+         srclen = 0;
+      
+      HMAC_SHA256_CTX ctx;
+      HMAC_SHA256_Init( &ctx, dst, sizeof(*dst) );
+      HMAC_SHA256_Update( &ctx, src, srclen );
+      HMAC_SHA256_Final( sha256, &ctx );
 
-         SHA256_Buf(sha256, sizeof(sha256), (uint8_t *)dst);
+//      SHA256_CTX ctx;
+      SHA256_Init( &sha256_ctx );
+      SHA256_Update( &sha256_ctx, sha256, sizeof(sha256) );
+      SHA256_Final( (unsigned char*)dst, &sha256_ctx );
+
+
+/*
+      if ( pers )
+      {
+			HMAC_SHA256_Buf( dst, sizeof(*dst), pers, perslen, sha256 );
+         SHA256_Buf( sha256, sizeof(sha256), (uint8_t *)dst );
 		}
-	} else {
-		ctx.S2 = S + 2 * Swidth_to_Sbytes1(Swidth);
+*/
+   }
+   else
+   {
+		ctx.S2 = S + 2 * Swidth_to_Sbytes1( Swidth );
 		ctx.w = 0;
 
-		if (pers) {
+		if ( pers )
+      {
 			src = pers;
 			srclen = perslen;
-		} else {
-			srclen = 0;
 		}
+      else
+			srclen = 0;
 
-		PBKDF2_SHA256(sha256, sizeof(sha256), src, srclen, 1, B, 128);
-		memcpy(sha256, B, sizeof(sha256));
-		smix_1_0(B, r, N, V, XY, &ctx);
-		HMAC_SHA256_Buf(B + B_size - 64, 64,
-		    sha256, sizeof(sha256), (uint8_t *)dst);
+		PBKDF2_SHA256( sha256, sizeof(sha256), src, srclen, 1, B, 128 );
+		memcpy( sha256, B, sizeof(sha256) );
+
+      if ( work_restart[thrid].restart ) return 0;
+
+      smix_1_0( B, r, N, V, XY, &ctx );
+
+      HMAC_SHA256_Buf( B + B_size - 64, 64, sha256, sizeof(sha256),
+                       (uint8_t *)dst );
 	}
 
 	/* Success! */

algo/yespower/yespower-ref.c

@@ -453,9 +453,8 @@ static void smix(uint32_t *B, size_t r, uint32_t N,
  *
  * Return 0 on success; or -1 on error.
  */
-int yespower(yespower_local_t *local,
-    const uint8_t *src, size_t srclen,
-    const yespower_params_t *params, yespower_binary_t *dst)
+int yespower( yespower_local_t *local, const uint8_t *src, size_t srclen,
+    const yespower_params_t *params, yespower_binary_t *dst, int thrid ) 
 {
 	yespower_version_t version = params->version;
 	uint32_t N = params->N;
@@ -534,17 +533,16 @@ int yespower(yespower_local_t *local,
 
 		if (pers) {
 			HMAC_SHA256_Buf(dst, sizeof(*dst), pers, perslen,
-               return true;
 			    (uint8_t *)sha256);
 			SHA256_Buf(sha256, sizeof(sha256), (uint8_t *)dst);
 		}
 	} else {
-		HMAC_SHA256_Buf_P((uint8_t *)B + B_size - 64, 64,
+		HMAC_SHA256_Buf((uint8_t *)B + B_size - 64, 64,
 		    sha256, sizeof(sha256), (uint8_t *)dst);
 	}
 
 	/* Success! */
-	retval = 0;
+	retval = 1;
 
 	/* Free memory */
 	free(S);
@@ -559,10 +557,10 @@ free_V:
 }
 
 int yespower_tls(const uint8_t *src, size_t srclen,
-    const yespower_params_t *params, yespower_binary_t *dst)
+    const yespower_params_t *params, yespower_binary_t *dst, int thrid )
 {
 /* The reference implementation doesn't use thread-local storage */
-	return yespower(NULL, src, srclen, params, dst);
+	return yespower(NULL, src, srclen, params, dst, thrid );
 }
 
 int yespower_init_local(yespower_local_t *local)

algo/yespower/yespower.h

@@ -33,6 +33,8 @@
 #include <stdint.h>
 #include <stdlib.h> /* for size_t */
 #include "miner.h"
+#include "simd-utils.h"
+#include <openssl/sha.h>
 
 #ifdef __cplusplus
 extern "C" {
@@ -74,6 +76,11 @@ typedef struct {
 	unsigned char uc[32];
 } yespower_binary_t __attribute__ ((aligned (64)));
 
+extern yespower_params_t yespower_params;
+
+//SHA256_CTX sha256_prehash_ctx;
+extern __thread SHA256_CTX sha256_prehash_ctx;
+
 /**
  * yespower_init_local(local):
  * Initialize the thread-local (RAM) data structure.  Actual memory allocation
@@ -131,6 +138,24 @@ extern int yespower_tls(const uint8_t *src, size_t srclen,
 extern int yespower_b2b_tls(const uint8_t *src, size_t srclen,
     const yespower_params_t *params, yespower_binary_t *dst, int thr_id);
 
+
+#if defined(__AVX2__)
+
+typedef struct
+{
+   __m256i uc[8];
+} yespower_8way_binary_t __attribute__ ((aligned (128)));
+
+extern int yespower_8way( yespower_local_t *local, const __m256i *src,
+                          size_t srclen, const yespower_params_t *params,
+                          yespower_8way_binary_t *dst, int thrid );
+
+
+extern int yespower_8way_tls( const __m256i *src, size_t srclen,
+    const yespower_params_t *params, yespower_8way_binary_t *dst, int thr_id );
+
+#endif // AVX2
+
 #ifdef __cplusplus
 }
 #endif

build-allarch.sh

@@ -4,13 +4,20 @@
 # during develpment. However the information contained may provide compilation
 # tips to users.
 
-rm cpuminer-avx512 cpuminer-avx2 cpuminer-aes-avx cpuminer-aes-sse42 cpuminer-sse42 cpuminer-ssse3 cpuminer-sse2 cpuminer-zen  > /dev/null
+rm cpuminer-avx512-sha-vaes cpuminer-avx512 cpuminer-avx2 cpuminer-aes-avx cpuminer-aes-sse42 cpuminer-sse42 cpuminer-ssse3 cpuminer-sse2 cpuminer-zen  > /dev/null
 
 make distclean || echo clean
 rm -f config.status
 ./autogen.sh || echo done
-CFLAGS="-O3 -march=skylake-avx512 -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -march=icelake-client -Wall -fno-common" ./configure --with-curl
+make -j 8
+strip -s cpuminer.exe
+mv cpuminer.exe cpuminer-avx512-sha-vaes.exe
+strip -s cpuminer
+mv cpuminer cpuminer-avx512-sha-vaes
+
+CFLAGS="-O3 -march=skylake-avx512 -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-avx512.exe
 strip -s cpuminer
@@ -19,8 +26,8 @@ mv cpuminer cpuminer-avx512
 make clean || echo clean
 rm -f config.status
 # GCC 9 doesn't include AES with core-avx2
-CFLAGS="-O3 -march=core-avx2 -maes -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -march=core-avx2 -maes -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-avx2.exe
 strip -s cpuminer
@@ -28,17 +35,17 @@ mv cpuminer cpuminer-avx2
 
 make clean || echo clean
 rm -f config.status
-CFLAGS="-O3 -march=corei7-avx -maes -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -march=corei7-avx -maes -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-avx.exe
 strip -s cpuminer
-mv cpuminer cpuminer-aes-avx
+mv cpuminer cpuminer-avx
 
 make clean || echo clean
 rm -f config.status
-CFLAGS="-O3 -maes -msse4.2 -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -maes -msse4.2 -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-aes-sse42.exe
 strip -s cpuminer
@@ -46,8 +53,8 @@ mv cpuminer cpuminer-aes-sse42
 
 #make clean || echo clean
 #rm -f config.status
-#CFLAGS="-O3 -march=corei7 -Wall" ./configure --with-curl
-#make -j 16
+#CFLAGS="-O3 -march=corei7 -Wall -fno-common" ./configure --with-curl
+#make -j 8
 #strip -s cpuminer.exe
 #mv cpuminer.exe cpuminer-sse42.exe
 #strip -s cpuminer
@@ -55,8 +62,8 @@ mv cpuminer cpuminer-aes-sse42
 
 #make clean || echo clean
 #rm -f config.status
-#CFLAGS="-O3 -march=core2 -Wall" ./configure --with-curl
-#make -j 16
+#CFLAGS="-O3 -march=core2 -Wall -fno-common" ./configure --with-curl
+#make -j 8
 #strip -s cpuminer.exe
 #mv cpuminer.exe cpuminer-ssse3.exe
 #strip -s cpuminer
@@ -64,8 +71,8 @@ mv cpuminer cpuminer-aes-sse42
 
 make clean || echo clean
 rm -f config.status
-CFLAGS="-O3 -msse2 -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -msse2 -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-sse2.exe
 strip -s cpuminer
@@ -73,8 +80,8 @@ mv cpuminer cpuminer-sse2
 
 make clean || echo done
 rm -f config.status
-CFLAGS="-O3 -march=znver1 -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -march=znver1 -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-zen.exe
 strip -s cpuminer
@@ -82,8 +89,8 @@ mv cpuminer cpuminer-zen
 
 make clean || echo done
 rm -f config.status
-CFLAGS="-O3 -march=native -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -march=native -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 strip -s cpuminer
 

build-no-common.sh

@@ -0,0 +1,27 @@
+#!/bin/bash
+
+#if [ "$OS" = "Windows_NT" ]; then
+#    ./mingw64.sh
+#    exit 0
+#fi
+
+# Linux build
+
+make distclean || echo clean
+
+rm -f config.status
+./autogen.sh || echo done
+
+# Ubuntu 10.04 (gcc 4.4)
+# extracflags="-O3 -march=native -Wall -D_REENTRANT -funroll-loops -fvariable-expansion-in-unroller -fmerge-all-constants -fbranch-target-load-optimize2 -fsched2-use-superblocks -falign-loops=16 -falign-functions=16 -falign-jumps=16 -falign-labels=16"
+
+# Debian 7.7 / Ubuntu 14.04 (gcc 4.7+)
+#extracflags="$extracflags -Ofast -flto -fuse-linker-plugin -ftree-loop-if-convert-stores"
+
+#CFLAGS="-O3 -march=native -Wall" ./configure --with-curl --with-crypto=$HOME/usr
+CFLAGS="-O3 -march=native -Wall -fno-common" ./configure --with-curl
+#CFLAGS="-O3 -march=native -Wall" CXXFLAGS="$CFLAGS -std=gnu++11" ./configure --with-curl
+
+make -j 4
+
+strip -s cpuminer

clean-all.sh

@@ -3,8 +3,8 @@
 # imake clean and rm all the targetted executables.
 # tips to users.
 
-rm cpuminer-avx512 cpuminer-avx2 cpuminer-aes-avx cpuminer-aes-sse42 cpuminer-sse2 cpuminer-zen  > /dev/null
+rm cpuminer-avx512-sha-vaes cpuminer-avx512 cpuminer-avx2 cpuminer-avx cpuminer-aes-sse42 cpuminer-sse2 cpuminer-zen  > /dev/null
 
-rm cpuminer-avx512.exe cpuminer-avx2.exe cpuminer-aes-avx.exe cpuminer-aes-sse42.exe cpuminer-sse2.exe cpuminer-zen.exe  > /dev/null
+rm cpuminer-avx512-sha-vaes.exe cpuminer-avx512.exe cpuminer-avx2.exe cpuminer-avx.exe cpuminer-aes-sse42.exe cpuminer-sse2.exe cpuminer-zen.exe  > /dev/null
 
 make distclean > /dev/null