v26.1

RELEASE_NOTES

@@ -32,8 +32,6 @@ Requirements
 
 32 bit CPUs are not supported.
 
-Older CPUs are supported by open source cpuminer-multi by TPruvot but at reduced performance.
-
 Mining on mobile devices that meet the requirements is not recommended due to the risk of
 overheating and damaging the battery. Mining has unlimited demand, it will push any device
 to or beyond its limits. There is also a fire risk with overheated lithium batteries.
@@ -75,6 +73,10 @@ If not what makes it happen or not happen?
 Change Log
 ----------
 
+v26.1
+
+Fixed segfault in scrypt algo on some older CPUs.
+
 v25.7
 
 Fixed a bug calculating TTF longer than 1 year.

algo/hamsi/hamsi-hash-4way.c

@@ -1137,7 +1137,7 @@ do { \
   } \
 } while (0)
 
-// v3 ternary logic, 8 instructions, 2 local vars
+// v4 ternary logic, 8 instructions, 2 local vars
 #define SBOX( a, b, c, d ) \
 { \
   __m256i tb, td; \

algo/scrypt/scrypt.c

@@ -37,8 +37,8 @@
 
 #if defined(SIMD512)
   #define SCRYPT_THROUGHPUT 16
-#elif defined(__SHA__) || defined(__ARM_FEATURE_SHA2)
-  #define SCRYPT_THROUGHPUT 2
+//#elif defined(__SHA__) || defined(__ARM_FEATURE_SHA2)
+//  #define SCRYPT_THROUGHPUT 2
 #elif defined(__AVX2__)
   #define SCRYPT_THROUGHPUT 8
 #elif defined(__SSE2__) || defined(__ARM_NEON)
@@ -162,7 +162,7 @@ static inline void PBKDF2_SHA256_128_32(uint32_t *tstate, uint32_t *ostate,
 }
 
 #endif    // throughput 1
-          //
+          
 #if defined(__SHA__) || defined(__ARM_FEATURE_SHA2)
 
 static inline void HMAC_SHA256_80_init_SHA_2BUF( const uint32_t *key0, 
@@ -1230,7 +1230,8 @@ static int scrypt_N_1_1_256_sha_2buf( const uint32_t *input,
 
 #if ( SCRYPT_THROUGHPUT == 4 )
 
-#if defined(__SHA__)
+#if 0
+//#if defined(__SHA__) || defined(__ARM_FEATURE_SHA2)
 
 static int scrypt_N_1_1_256_4way_sha( const uint32_t *input, uint32_t *output,
            uint32_t *midstate, int N, int thrid )
@@ -1244,6 +1245,15 @@ static int scrypt_N_1_1_256_4way_sha( const uint32_t *input, uint32_t *output,
     memcpy( tstate+16, midstate, 32 );
     memcpy( tstate+24, midstate, 32 );
     
+    HMAC_SHA256_80_init_SHA_2BUF( input, input+20, tstate, tstate+8,
+                                  ostate, ostate+8 );
+    PBKDF2_SHA256_80_128_SHA_2BUF( tstate, tstate+8, ostate, ostate+8,
+                                   input, input+20,  W, W+32 );
+    HMAC_SHA256_80_init_SHA_2BUF( input+40, input+60, tstate+16, tstate+24,
+                                  ostate+16, ostate+24 );
+    PBKDF2_SHA256_80_128_SHA_2BUF( tstate+16, tstate+24, ostate+16, ostate+24,
+                                   input+40, input+60,  W+64, W+96 );
+/*    
     HMAC_SHA256_80_init(  input,     tstate,    ostate    );
     PBKDF2_SHA256_80_128( tstate,    ostate,    input,     W );
     HMAC_SHA256_80_init(  input +20, tstate+ 8, ostate+ 8 );
@@ -1252,7 +1262,7 @@ static int scrypt_N_1_1_256_4way_sha( const uint32_t *input, uint32_t *output,
     PBKDF2_SHA256_80_128( tstate+16, ostate+16, input +40, W+64 );
     HMAC_SHA256_80_init(  input +60, tstate+24, ostate+24 );
     PBKDF2_SHA256_80_128( tstate+24, ostate+24, input +60, W+96 );
-
+*/
 /*    
    // Working Linear single threaded SIMD
    scrypt_core_simd128( W,    V, N );
@@ -1278,11 +1288,16 @@ static int scrypt_N_1_1_256_4way_sha( const uint32_t *input, uint32_t *output,
 
    if ( work_restart[thrid].restart ) return 0;
 
+   PBKDF2_SHA256_128_32_SHA_2BUF( tstate,    tstate+ 8, ostate,    ostate+ 8,
+                                  W,    W+32, output,    output+ 8 );
+   PBKDF2_SHA256_128_32_SHA_2BUF( tstate+16, tstate+24, ostate+16, ostate+24,
+                                  W+64, W+96, output+16, output+24 );
+/*   
    PBKDF2_SHA256_128_32( tstate,    ostate,    W,    output    );
    PBKDF2_SHA256_128_32( tstate+ 8, ostate+ 8, W+32, output+ 8 );
    PBKDF2_SHA256_128_32( tstate+16, ostate+16, W+64, output+16 );
    PBKDF2_SHA256_128_32( tstate+24, ostate+24, W+96, output+24 );
-
+*/
    return 1;
 }
 
@@ -1390,13 +1405,13 @@ extern int scanhash_scrypt( struct work *work, uint32_t max_nonce,
          rc = scrypt_N_1_1_256_8way( data, hash, midstate, opt_param_n,
                                      thr_id );
 #elif ( SCRYPT_THROUGHPUT == 4 )
-  #if defined(__SHA__) || defined(__ARM_FEATURE_SHA2)
-         rc = scrypt_N_1_1_256_4way_sha( data, hash, midstate, opt_param_n,
-                                         thr_id );
-  #else
+//  #if defined(__SHA__) || defined(__ARM_FEATURE_SHA2)
+//         rc = scrypt_N_1_1_256_4way_sha( data, hash, midstate, opt_param_n,
+//                                         thr_id );
+//  #else
          rc = scrypt_N_1_1_256_4way( data, hash, midstate, opt_param_n,
                                      thr_id );
-  #endif
+//  #endif
 #elif ( SCRYPT_THROUGHPUT == 2 ) && ( defined(__SHA__) || defined(__ARM_FEATURE_SHA2) )
          rc = scrypt_N_1_1_256_sha_2buf( data, hash, midstate, opt_param_n,
                                          thr_id );
@@ -1444,11 +1459,6 @@ bool scrypt_miner_thread_init( int thr_id )
 
 bool register_scrypt_algo( algo_gate_t* gate )
 {
-#if defined(__SHA__) || defined(__ARM_FEATURE_SHA2)
-   gate->optimizations = SSE2_OPT | SSE42_OPT | AVX_OPT | SHA256_OPT | NEON_OPT;
-#else
-   gate->optimizations = SSE2_OPT | SSE42_OPT | AVX_OPT | AVX2_OPT | AVX512_OPT | NEON_OPT;
-#endif
    gate->miner_thread_init =(void*)&scrypt_miner_thread_init;
    gate->scanhash         = (void*)&scanhash_scrypt;
    opt_target_factor = 65536.0;
@@ -1469,7 +1479,7 @@ bool register_scrypt_algo( algo_gate_t* gate )
      case 8:  // AVX2
        if ( opt_param_n > 0x4000 )
          scratchbuf_size = opt_param_n * 3 * 128;  // 3 buf
-     else
+       else
          scratchbuf_size = opt_param_n * 2 * 128;  // 2 way
      break;
      case 4:  // SSE2, NEON
@@ -1477,8 +1487,9 @@ bool register_scrypt_algo( algo_gate_t* gate )
          scratchbuf_size = opt_param_n * 2 * 128;  // 2 buf
        else
          scratchbuf_size = opt_param_n * 4 * 128;  // 4 way
+     break;
      default:
-         scratchbuf_size = opt_param_n;  // 1 way
+         scratchbuf_size = opt_param_n * 128;  // 1 way
    }
 
    char t_units[4] = {0};

armbuild-all.sh

@@ -56,7 +56,7 @@ make clean || echo clean
 rm -f config.status
 CFLAGS="-O3 -march=armv8.2-a+crypto -Wall -flax-vector-conversions" ./configure  --with-curl
 make -j $(nproc)
-mv cpuminer cpuminer-armv8-crypto
+mv cpuminer cpuminer-armv8.2-crypto
 
 make clean || echo clean
 rm -f config.status

configure

@@ -1,6 +1,6 @@
 #! /bin/sh
 # Guess values for system-dependent variables and create Makefiles.
-# Generated by GNU Autoconf 2.71 for cpuminer-opt 25.7.
+# Generated by GNU Autoconf 2.71 for cpuminer-opt 26.1.
 #
 #
 # Copyright (C) 1992-1996, 1998-2017, 2020-2021 Free Software Foundation,
@@ -608,8 +608,8 @@ MAKEFLAGS=
 # Identity of this package.
 PACKAGE_NAME='cpuminer-opt'
 PACKAGE_TARNAME='cpuminer-opt'
-PACKAGE_VERSION='25.7'
-PACKAGE_STRING='cpuminer-opt 25.7'
+PACKAGE_VERSION='26.1'
+PACKAGE_STRING='cpuminer-opt 26.1'
 PACKAGE_BUGREPORT=''
 PACKAGE_URL=''
 
@@ -1359,7 +1359,7 @@ if test "$ac_init_help" = "long"; then
   # Omit some internal or obsolete options to make the list less imposing.
   # This message is too long to be a string in the A/UX 3.1 sh.
   cat <<_ACEOF
-\`configure' configures cpuminer-opt 25.7 to adapt to many kinds of systems.
+\`configure' configures cpuminer-opt 26.1 to adapt to many kinds of systems.
 
 Usage: $0 [OPTION]... [VAR=VALUE]...
 
@@ -1431,7 +1431,7 @@ fi
 
 if test -n "$ac_init_help"; then
   case $ac_init_help in
-     short | recursive ) echo "Configuration of cpuminer-opt 25.7:";;
+     short | recursive ) echo "Configuration of cpuminer-opt 26.1:";;
    esac
   cat <<\_ACEOF
 
@@ -1536,7 +1536,7 @@ fi
 test -n "$ac_init_help" && exit $ac_status
 if $ac_init_version; then
   cat <<\_ACEOF
-cpuminer-opt configure 25.7
+cpuminer-opt configure 26.1
 generated by GNU Autoconf 2.71
 
 Copyright (C) 2021 Free Software Foundation, Inc.
@@ -1983,7 +1983,7 @@ cat >config.log <<_ACEOF
 This file contains any messages produced by compilers while
 running configure, to aid debugging if configure makes a mistake.
 
-It was created by cpuminer-opt $as_me 25.7, which was
+It was created by cpuminer-opt $as_me 26.1, which was
 generated by GNU Autoconf 2.71.  Invocation command line was
 
   $ $0$ac_configure_args_raw
@@ -3591,7 +3591,7 @@ fi
 
 # Define the identity of the package.
  PACKAGE='cpuminer-opt'
- VERSION='25.7'
+ VERSION='26.1'
 
 
 printf "%s\n" "#define PACKAGE \"$PACKAGE\"" >>confdefs.h
@@ -5808,11 +5808,11 @@ if test x$ac_prog_cxx_stdcxx = xno
 then :
   { printf "%s\n" "$as_me:${as_lineno-$LINENO}: checking for $CXX option to enable C++11 features" >&5
 printf %s "checking for $CXX option to enable C++11 features... " >&6; }
-if test ${ac_cv_prog_cxx_cxx11+y}
+if test ${ac_cv_prog_cxx_11+y}
 then :
   printf %s "(cached) " >&6
 else $as_nop
-  ac_cv_prog_cxx_cxx11=no
+  ac_cv_prog_cxx_11=no
 ac_save_CXX=$CXX
 cat confdefs.h - <<_ACEOF >conftest.$ac_ext
 /* end confdefs.h.  */
@@ -5854,11 +5854,11 @@ if test x$ac_prog_cxx_stdcxx = xno
 then :
   { printf "%s\n" "$as_me:${as_lineno-$LINENO}: checking for $CXX option to enable C++98 features" >&5
 printf %s "checking for $CXX option to enable C++98 features... " >&6; }
-if test ${ac_cv_prog_cxx_cxx98+y}
+if test ${ac_cv_prog_cxx_98+y}
 then :
   printf %s "(cached) " >&6
 else $as_nop
-  ac_cv_prog_cxx_cxx98=no
+  ac_cv_prog_cxx_98=no
 ac_save_CXX=$CXX
 cat confdefs.h - <<_ACEOF >conftest.$ac_ext
 /* end confdefs.h.  */
@@ -7435,7 +7435,7 @@ cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1
 # report actual input values of CONFIG_FILES etc. instead of their
 # values after options handling.
 ac_log="
-This file was extended by cpuminer-opt $as_me 25.7, which was
+This file was extended by cpuminer-opt $as_me 26.1, which was
 generated by GNU Autoconf 2.71.  Invocation command line was
 
   CONFIG_FILES    = $CONFIG_FILES
@@ -7503,7 +7503,7 @@ ac_cs_config_escaped=`printf "%s\n" "$ac_cs_config" | sed "s/^ //; s/'/'\\\\\\\\
 cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1
 ac_cs_config='$ac_cs_config_escaped'
 ac_cs_version="\\
-cpuminer-opt config.status 25.7
+cpuminer-opt config.status 26.1
 configured by $0, generated by GNU Autoconf 2.71,
   with options \\"\$ac_cs_config\\"
 

configure.ac

@@ -1,4 +1,4 @@
-AC_INIT([cpuminer-opt], [25.7])
+AC_INIT([cpuminer-opt], [26.1])
 
 AC_PREREQ([2.59c])
 AC_CANONICAL_SYSTEM

configure~

@@ -1,6 +1,6 @@
 #! /bin/sh
 # Guess values for system-dependent variables and create Makefiles.
-# Generated by GNU Autoconf 2.72 for cpuminer-opt 25.7.
+# Generated by GNU Autoconf 2.72 for cpuminer-opt 26.1.
 #
 #
 # Copyright (C) 1992-1996, 1998-2017, 2020-2023 Free Software Foundation,
@@ -601,8 +601,8 @@ MAKEFLAGS=
 # Identity of this package.
 PACKAGE_NAME='cpuminer-opt'
 PACKAGE_TARNAME='cpuminer-opt'
-PACKAGE_VERSION='25.7'
-PACKAGE_STRING='cpuminer-opt 25.7'
+PACKAGE_VERSION='26.1'
+PACKAGE_STRING='cpuminer-opt 26.1'
 PACKAGE_BUGREPORT=''
 PACKAGE_URL=''
 
@@ -1352,7 +1352,7 @@ if test "$ac_init_help" = "long"; then
   # Omit some internal or obsolete options to make the list less imposing.
   # This message is too long to be a string in the A/UX 3.1 sh.
   cat <<_ACEOF
-'configure' configures cpuminer-opt 25.7 to adapt to many kinds of systems.
+'configure' configures cpuminer-opt 26.1 to adapt to many kinds of systems.
 
 Usage: $0 [OPTION]... [VAR=VALUE]...
 
@@ -1424,7 +1424,7 @@ fi
 
 if test -n "$ac_init_help"; then
   case $ac_init_help in
-     short | recursive ) echo "Configuration of cpuminer-opt 25.7:";;
+     short | recursive ) echo "Configuration of cpuminer-opt 26.1:";;
    esac
   cat <<\_ACEOF
 
@@ -1528,7 +1528,7 @@ fi
 test -n "$ac_init_help" && exit $ac_status
 if $ac_init_version; then
   cat <<\_ACEOF
-cpuminer-opt configure 25.7
+cpuminer-opt configure 26.1
 generated by GNU Autoconf 2.72
 
 Copyright (C) 2023 Free Software Foundation, Inc.
@@ -1949,7 +1949,7 @@ cat >config.log <<_ACEOF
 This file contains any messages produced by compilers while
 running configure, to aid debugging if configure makes a mistake.
 
-It was created by cpuminer-opt $as_me 25.7, which was
+It was created by cpuminer-opt $as_me 26.1, which was
 generated by GNU Autoconf 2.72.  Invocation command line was
 
   $ $0$ac_configure_args_raw
@@ -3768,7 +3768,7 @@ fi
 
 # Define the identity of the package.
  PACKAGE='cpuminer-opt'
- VERSION='25.7'
+ VERSION='26.1'
 
 
 printf "%s\n" "#define PACKAGE \"$PACKAGE\"" >>confdefs.h
@@ -7581,7 +7581,7 @@ cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1
 # report actual input values of CONFIG_FILES etc. instead of their
 # values after options handling.
 ac_log="
-This file was extended by cpuminer-opt $as_me 25.7, which was
+This file was extended by cpuminer-opt $as_me 26.1, which was
 generated by GNU Autoconf 2.72.  Invocation command line was
 
   CONFIG_FILES    = $CONFIG_FILES
@@ -7649,7 +7649,7 @@ ac_cs_config_escaped=`printf "%s\n" "$ac_cs_config" | sed "s/^ //; s/'/'\\\\\\\\
 cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1
 ac_cs_config='$ac_cs_config_escaped'
 ac_cs_version="\\
-cpuminer-opt config.status 25.7
+cpuminer-opt config.status 26.1
 configured by $0, generated by GNU Autoconf 2.72,
   with options \\"\$ac_cs_config\\"
 

cpu-miner.c

@@ -1232,7 +1232,7 @@ static int share_result( int result, struct work *work,
      sprintf( ares, "A%d", accepted_share_count );
      sprintf( bres, "B%d", solved_block_count );
      if ( reason )
-        stale = strstr( reason, "job" );
+        stale = strstr( reason, "job" ) || strstr( reason, "Job" );
      else if ( work )
         stale =  work->data[ algo_gate.ntime_index ]
              != g_work.data[ algo_gate.ntime_index ];

simd-utils/simd-128.h

@@ -642,6 +642,15 @@ static inline void v128_memcpy( v128_t *dst, const v128_t *src, const int n )
 #define v128_shuflr32(v)    _mm_shuffle_epi32( v, 0x39 )
 #define v128_shufll32(v)    _mm_shuffle_epi32( v, 0x93 )
 
+/* Zen6 AMD only
+// Reverse bits in bytes
+#if defined(__AVX512VL__) && defined(__AVX512BMM__)
+
+#define v128_bitrev8           _mm_vbitrevb_epi8
+
+#endif
+*/
+
 // Endian byte swap.
 
 #if defined(__SSSE3__)

simd-utils/simd-256.h

@@ -409,6 +409,15 @@ static inline __m256i mm256_shuflr128_x8( const __m256i v, const int c )
 { return _mm256_alignr_epi8( v, v, c ); }
 */
 
+/* Zen6 AMD only
+// Reverse bits in bytes
+#if defined(__AVX512VL__) && defined(__AVX512BMM__)
+
+#define mm256_bitrev8           _mm256_vbitrevb_epi8
+
+#endif
+*/
+
 // Reverse byte order in elements, endian bswap.
 #define mm256_bswap_64( v )     _mm256_shuffle_epi8( v, V256_BSWAP64 )
 

simd-utils/simd-512.h

@@ -257,6 +257,15 @@ static inline void memcpy_512( __m512i *dst, const __m512i *src, const int n )
 #endif
 */
 
+/* Zen6 AMD only
+// Reverse bits in bytes
+#if defined(__AVX512BMM__)
+
+#define mm512_bitrev8           _mm512_vbitrevb_epi8
+
+#endif
+*/
+
 //
 // Reverse byte order of packed elements, vectorized endian conversion.
 

simd-utils/simd-neon.h

@@ -431,8 +431,10 @@ static inline uint32x4_t v128_shuflr32( uint32x4_t v )
 static inline uint32x4_t v128_shufll32( uint32x4_t v )
 {   return vextq_u32( v, v, 3 ); }
 
-// reverse bits in bytes, nothing like it in x86_64
+/* not used
+// reverse bits in bytes, nothing like it in x86_64 until Zen6
 #define v128_bitrev8           vrbitq_u8
+*/
 
 // reverse byte order
 #define v128_bswap16(v)        (uint16x8_t)vrev16q_u8( (uint8x16_t)(v) )

sysinfos.c

@@ -16,8 +16,8 @@
 #include "miner.h"
 #include "simd-utils.h"
 
-// Missing on MinGW, MacOS
-#if defined(__aarch64__) && !defined(WIN32) && !defined(__APPLE__)
+// hwcap.h missing on MinGW, MacOS
+#if defined(__aarch64__) && !(defined(WIN32) || defined(__APPLE__)) 
 #define ARM_AUXV 
 #endif
 
@@ -191,6 +191,7 @@ static inline int cpu_fanpercent()
 #define CPU_INFO             (1)
 #define CACHE_TLB_DESCRIPTOR (2)
 #define EXTENDED_FEATURES    (7)
+#define EXTENDED_FEATURE_ID  (0x21)
 #define AVX10_FEATURES       (0x24)
 #define HIGHEST_EXT_FUNCTION (0x80000000)
 #define EXTENDED_CPU_INFO    (0x80000001)
@@ -254,8 +255,8 @@ static inline int cpu_fanpercent()
 #define AVX512_BF16_Flag         (1<< 5)
 #define AMX_FP16_Flag            (1<<21)
 #define AVX_IFMA_Flag            (1<<23)
-#define MOVRS_Flag               (1<<31)  // Both names are referenced in docs
 #define AVX10_MOVRS_Flag         (1<<31)
+#define MOVRS_Flag               (1<<31)  // Both names are referenced in docs
 // EDX
 #define AVX_VNNI_INT8_Flag       (1<< 4)
 #define AVX_NE_CONVERT_Flag      (1<< 5)
@@ -264,6 +265,10 @@ static inline int cpu_fanpercent()
 #define AVX10_Flag               (1<<19)
 #define APX_F_Flag               (1<<21)
 
+// EXTENDED_FEATURE_ID: EAX=0x21, ECX=0
+// EAX
+#define AVX512_BMM_Flag          (1<<23)   // Zen6 AMD only
+
 // AVX10_FEATURES: EAX=0x24, ECX=0
 // EBX
 #define AVX10_VERSION_mask        0xff      // bits [7:0]
@@ -735,7 +740,7 @@ static inline bool has_avx512()
 #endif
 }
 
-static inline bool has_vbmi()
+static inline bool has_avx512vbmi()
 {
 #if defined(__x86_64__)
    unsigned int cpu_info[4] = { 0 };
@@ -746,7 +751,7 @@ static inline bool has_vbmi()
 #endif
 }
 
-static inline bool has_vbmi2()
+static inline bool has_avx512vbmi2()
 {
 #if defined(__x86_64__)
    unsigned int cpu_info[4] = { 0 };
@@ -757,6 +762,18 @@ static inline bool has_vbmi2()
 #endif
 }
 
+// Zen6 AMD only
+static inline bool has_avx512bmm()
+{
+#if defined(__x86_64__)
+   unsigned int cpu_info[4] = { 0 };
+   cpuid( EXTENDED_FEATURE_ID, 0, cpu_info );
+   return cpu_info[ EAX_Reg ] & AVX512_BMM_Flag;
+#else
+   return false;
+#endif
+}
+
 static inline bool has_amx()
 {
 #if defined(__x86_64__)
@@ -817,10 +834,9 @@ static inline bool has_sveaes()
 static inline bool has_sha256()
 {
 #if defined(__x86_64__)
-      unsigned int cpu_info[4] = { 0 };
-      cpuid( EXTENDED_FEATURES, 0, cpu_info );
-      return cpu_info[ EBX_Reg ] & SHA_Flag;
-   return false;
+   unsigned int cpu_info[4] = { 0 };
+   cpuid( EXTENDED_FEATURES, 0, cpu_info );
+   return cpu_info[ EBX_Reg ] & SHA_Flag;
 #elif defined(__aarch64__) && defined(HWCAP_SHA2)
    // NEON SHA256
    unsigned int cpu_info[4] = { 0 };
@@ -851,6 +867,7 @@ static inline bool has_sha512()
 #endif
 }
 
+// ARM64 only
 static inline bool has_sha3()
 {
 #if defined(__aarch64__) && defined(HWCAP_SHA3)