v3.19.0

INSTALL_WINDOWS

@@ -40,7 +40,7 @@ $ mkdir $HOME/usr/lib
    version available in the repositories.
 
 Download the following source code packages from their respective and
-respected download locations, copy them to ~/usr/lib/ and uncompress them. 
+respected download locations, copy them to $HOME/usr/lib/ and uncompress them. 
 
 openssl: https://github.com/openssl/openssl/releases
 
@@ -149,85 +149,10 @@ Copy cpuminer.exe to the release directory, compress and copy the release direct
 
 Run cpuminer
 
-In a command windows change directories to the unzipped release folder. to get a list of all options:
+In a command windows change directories to the unzipped release folder. To get a list of all options:
 
 cpuminer.exe --help
 
 Command options are specific to where you mine. Refer to the pool's instructions on how to set them.
 
 
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-Create a link to the locally compiled version of gmp.h
-
-$ ln -s $LOCAL_LIB/gmp-version/gmp.h ./gmp.h
-
-Edit configure.ac to fix lipthread package name.
-
-sed -i 's/"-lpthread"/"-lpthreadGC2"/g' configure.ac
-
-
-7. Compile
-
-you can use the default compile if you intend to use cpuminer-opt on the
-same CPU and the virtual machine supports that architecture.
-
-./build.sh
-
-Otherwise you can compile manually while setting options in CFLAGS.
-
-Some common options:
-
-To compile for a specific CPU architecture:
-
-CFLAGS="-O3 -march=znver1 -Wall" ./configure --with-curl
-
-This will compile for AMD Ryzen.
-
-You can compile more generically for a set of specific CPU features
-if you know what features you want:
-
-CFLAGS="-O3 -maes -msse4.2 -Wall" ./configure --with-curl
-
-This will compile for an older CPU that does not have AVX.
-
-You can find several examples in build-allarch.sh
-
-If you have a CPU with more than 64 threads and Windows 7 or higher you
-can enable the CPU Groups feature:
-
--D_WIN32_WINNT==0x0601
-
-Once you have run configure successfully run make with n CPU threads:
-
-make -j n
-
-Copy cpuminer.exe to the release directory, compress and copy the release
-directory to a Windows system and run cpuminer.exe from the command line.
-
-Run cpuminer
-
-In a command windows change directories to the unzipped release folder.
-to get a list of all options:
-
-cpuminer.exe --help
-
-Command options are specific to where you mine. Refer to the pool's
-instructions on how to set them.

RELEASE_NOTES

@@ -65,7 +65,22 @@ If not what makes it happen or not happen?
 Change Log
 ----------
 
-v3.8.2
+v3.19.0
+
+Windows binaries now built with support for CPU groups, requires Windows 7.
+
+Changes to cpu-affinity:
+  - PR#346: Fixed incorrect CPU affinity on Windows built for CPU groups,
+  - added support for CPU affinity for up to 256 threads or CPUs,
+  - streamlined code for more efficient initialization of miner threads,
+  - precise affining of each miner thread to a specific CPU,
+  - added an option to disable CPU affinity with "--cpu-affinity 0"
+
+Faster sha256t with AVX512 & AVX2.
+
+Added stratum error count to stats log, reported only when non-zero.
+
+v3.18.2
 
 Issue #342, fixed Groestl AES on Windows, broken in v3.18.0.
 

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

@@ -62,6 +62,12 @@ void sha256_4way_transform_le( __m128i *state_out,  const __m128i *data,
                             const __m128i *state_in );
 void sha256_4way_transform_be( __m128i *state_out,  const __m128i *data,
                             const __m128i *state_in );
+void sha256_4way_prehash_3rounds( __m128i *state_mid, __m128i *X,
+                                   const __m128i *W, const __m128i *state_in );
+void sha256_4way_final_rounds( __m128i *state_out, const __m128i *data,
+        const __m128i *state_in, const __m128i *state_mid, const __m128i *X );
+int sha256_4way_transform_le_short( __m128i *state_out, const __m128i *data,
+                                     const __m128i *state_in );
 
 #endif  // SSE2
 
@@ -84,10 +90,12 @@ void sha256_8way_transform_le( __m256i *state_out, const __m256i *data,
 void sha256_8way_transform_be( __m256i *state_out, const __m256i *data,
                                const __m256i *state_in );
 
-void sha256_8way_prehash_3rounds( __m256i *state_mid, const __m256i *W,
-                             const __m256i *state_in );
+void sha256_8way_prehash_3rounds( __m256i *state_mid, __m256i *X,
+                                 const __m256i *W, const __m256i *state_in );
 void sha256_8way_final_rounds( __m256i *state_out, const __m256i *data,
-                          const __m256i *state_in, const __m256i *state_mid );
+        const __m256i *state_in, const __m256i *state_mid, const __m256i *X );
+int sha256_8way_transform_le_short( __m256i *state_out, const __m256i *data,
+                                     const __m256i *state_in );
 
 #endif  // AVX2
 
@@ -109,10 +117,13 @@ void sha256_16way_transform_le( __m512i *state_out, const __m512i *data,
                              const __m512i *state_in );
 void sha256_16way_transform_be( __m512i *state_out, const __m512i *data,
                              const __m512i *state_in );
-void sha256_16way_prehash_3rounds( __m512i *state_mid, const __m512i *W,
-                             const __m512i *state_in );
+void sha256_16way_prehash_3rounds( __m512i *state_mid, __m512i *X,
+                                  const __m512i *W, const __m512i *state_in );
 void sha256_16way_final_rounds( __m512i *state_out, const __m512i *data,
-                          const __m512i *state_in, const __m512i *state_mid );
+        const __m512i *state_in, const __m512i *state_mid, const __m512i *X );
+
+int sha256_16way_transform_le_short( __m512i *state_out, const __m512i *data,
+                                     const __m512i *state_in );
 
 #endif // AVX512
 

algo/sha/sha2.c

@@ -611,11 +611,11 @@ static inline int scanhash_sha256d_8way_pooler( struct work *work,
 
 #endif /* HAVE_SHA256_8WAY */
 
-int scanhash_sha256d_pooler( struct work *work,
-	uint32_t max_nonce, uint64_t *hashes_done, struct thr_info *mythr )
+int scanhash_sha256d_pooler( 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;
+   uint32_t *pdata = work->data;
+   uint32_t *ptarget = work->target;
 	uint32_t _ALIGN(128) data[64];
 	uint32_t _ALIGN(32) hash[8];
 	uint32_t _ALIGN(32) midstate[8];
@@ -626,12 +626,12 @@ int scanhash_sha256d_pooler( struct work *work,
    int thr_id = mythr->id;  // thr_id arg is deprecated
 
 #ifdef HAVE_SHA256_8WAY
-	if (sha256_use_8way())
-		return scanhash_sha256d_8way_pooler( work,	max_nonce, hashes_done, mythr );
+	if ( sha256_use_8way() )
+		return scanhash_sha256d_8way_pooler( work, max_nonce, hashes_done, mythr );
 #endif
 #ifdef HAVE_SHA256_4WAY
-	if (sha256_use_4way())
-		return scanhash_sha256d_4way_pooler( work,	max_nonce, hashes_done, mythr );
+	if ( sha256_use_4way() )
+		return scanhash_sha256d_4way_pooler( work, max_nonce, hashes_done, mythr );
 #endif
 
 	memcpy(data, pdata + 16, 64);
@@ -695,8 +695,11 @@ bool register_sha256d_algo( algo_gate_t* gate )
    gate->optimizations = SSE2_OPT | AVX2_OPT | AVX512_OPT;
 #if defined(SHA256D_16WAY)
    gate->scanhash = (void*)&scanhash_sha256d_16way;
+//#elif defined(SHA256D_8WAY)
+//   gate->scanhash = (void*)&scanhash_sha256d_8way;
 #else
    gate->scanhash = (void*)&scanhash_sha256d_pooler;
+//   gate->scanhash = (void*)&scanhash_sha256d_4way;
 #endif
    //   gate->hash     = (void*)&sha256d;
    return true;