with SNMPv3, follow optimized logic path with pw_len <= 64, update changes/readme entries

OpenCL/m25000-pure.cl

@@ -32,6 +32,11 @@
 #define SNMPV3_MAX_ENGINE_ELEMS     16  // 16 * 4 = 64 > 32, also has to be multiple of 64
 #define SNMPV3_MAX_PNUM_ELEMS       4   // 4 * 4 = 16 > 9
 
+#define SNMPV3_MAX_PW_LENGTH_OPT    64
+#define SNMPV3_TMP_ELEMS_OPT        ((SNMPV3_MAX_PW_LENGTH_OPT * SNMPV3_MAX_PW_LENGTH) / 4)
+                                    // (64 * 64) / 4 = 1024
+                                    // for pw length > 64 we use global memory reads
+
 typedef struct hmac_md5_tmp
 {
   u32 tmp_md5[SNMPV3_TMP_ELEMS];
@@ -189,10 +194,7 @@ KERNEL_FQ void m25000_loop (KERN_ATTR_TMPS_ESALT (hmac_md5_tmp_t, snmpv3_t))
 
   const int pw_len64 = pw_len * 64;
 
-  #define SNMPV3_TMP_ELEMS_OPT 1024 // 1024 = (64 max pw length * 64) / sizeof (u32)
-                                    // for pw length > 64 we use global memory reads
-
-  if (pw_len < 64)
+  if (pw_len <= SNMPV3_MAX_PW_LENGTH_OPT)
   {
     u32 tmp_shared[SNMPV3_TMP_ELEMS_OPT];
 

OpenCL/m25100-pure.cl

@@ -30,6 +30,11 @@
 #define SNMPV3_MAX_ENGINE_ELEMS     16  // 16 * 4 = 64 > 32, also has to be multiple of 64
 #define SNMPV3_MAX_PNUM_ELEMS       4   // 4 * 4 = 16 > 9
 
+#define SNMPV3_MAX_PW_LENGTH_OPT    64
+#define SNMPV3_TMP_ELEMS_OPT        ((SNMPV3_MAX_PW_LENGTH_OPT * SNMPV3_MAX_PW_LENGTH) / 4)
+                                    // (64 * 64) / 4 = 1024
+                                    // for pw length > 64 we use global memory reads
+
 typedef struct hmac_md5_tmp
 {
   u32 tmp[SNMPV3_TMP_ELEMS];
@@ -147,13 +152,10 @@ KERNEL_FQ void m25100_loop (KERN_ATTR_TMPS_ESALT (hmac_md5_tmp_t, snmpv3_t))
 
   const int pw_len64 = pw_len * 64;
 
-  #define SNMPV3_TMP_ELEMS_OPT 1024 // 1024 = (64 max pw length * 64) / sizeof (u32)
-                                    // for pw length > 64 we use global memory reads
-
-  u32 tmp[SNMPV3_TMP_ELEMS_OPT];
-
-  if (pw_len < 64)
+  if (pw_len <= SNMPV3_MAX_PW_LENGTH_OPT)
   {
+    u32 tmp[SNMPV3_TMP_ELEMS_OPT];
+
     for (int i = 0; i < pw_len64 / 4; i++)
     {
       tmp[i] = tmps[gid].tmp[i];

OpenCL/m25200-pure.cl

@@ -30,6 +30,11 @@
 #define SNMPV3_MAX_ENGINE_ELEMS     16  // 16 * 4 = 64 > 32, also has to be multiple of 64
 #define SNMPV3_MAX_PNUM_ELEMS       4   // 4 * 4 = 16 > 9
 
+#define SNMPV3_MAX_PW_LENGTH_OPT    64
+#define SNMPV3_TMP_ELEMS_OPT        ((SNMPV3_MAX_PW_LENGTH_OPT * SNMPV3_MAX_PW_LENGTH) / 4)
+                                    // (64 * 64) / 4 = 1024
+                                    // for pw length > 64 we use global memory reads
+
 typedef struct hmac_sha1_tmp
 {
   u32 tmp[SNMPV3_TMP_ELEMS];
@@ -149,13 +154,10 @@ KERNEL_FQ void m25200_loop (KERN_ATTR_TMPS_ESALT (hmac_sha1_tmp_t, snmpv3_t))
 
   const int pw_len64 = pw_len * 64;
 
-  #define SNMPV3_TMP_ELEMS_OPT 1024 // 1024 = (64 max pw length * 64) / sizeof (u32)
-                                    // for pw length > 64 we use global memory reads
-
-  u32 tmp[SNMPV3_TMP_ELEMS_OPT];
-
-  if (pw_len < 64)
+  if (pw_len <= SNMPV3_MAX_PW_LENGTH_OPT)
   {
+    u32 tmp[SNMPV3_TMP_ELEMS_OPT];
+
     for (int i = 0; i < pw_len64 / 4; i++)
     {
       tmp[i] = tmps[gid].tmp[i];

OpenCL/m26700-pure.cl

@@ -30,6 +30,11 @@
 #define SNMPV3_MAX_ENGINE_ELEMS     16  // 16 * 4 = 64 > 32, also has to be multiple of 64
 #define SNMPV3_MAX_PNUM_ELEMS       4   // 4 * 4 = 16 > 9
 
+#define SNMPV3_MAX_PW_LENGTH_OPT    64
+#define SNMPV3_TMP_ELEMS_OPT        ((SNMPV3_MAX_PW_LENGTH_OPT * SNMPV3_MAX_PW_LENGTH) / 4)
+                                    // (64 * 64) / 4 = 1024
+                                    // for pw length > 64 we use global memory reads
+
 typedef struct hmac_sha224_tmp
 {
   u32 tmp[SNMPV3_TMP_ELEMS];
@@ -155,10 +160,7 @@ KERNEL_FQ void m26700_loop (KERN_ATTR_TMPS_ESALT (hmac_sha224_tmp_t, snmpv3_t))
 
   const int pw_len64 = pw_len * 64;
 
-  #define SNMPV3_TMP_ELEMS_OPT 1024 // 1024 = (64 max pw length * 64) / sizeof (u32)
-                                    // for pw length > 64 we use global memory reads
-
-  if (pw_len < 64)
+  if (pw_len <= SNMPV3_MAX_PW_LENGTH_OPT)
   {
     u32 tmp[SNMPV3_TMP_ELEMS_OPT];
 

OpenCL/m26800-pure.cl

@@ -30,6 +30,11 @@
 #define SNMPV3_MAX_ENGINE_ELEMS     16  // 16 * 4 = 64 > 32, also has to be multiple of 64
 #define SNMPV3_MAX_PNUM_ELEMS       4   // 4 * 4 = 16 > 9
 
+#define SNMPV3_MAX_PW_LENGTH_OPT    64
+#define SNMPV3_TMP_ELEMS_OPT        ((SNMPV3_MAX_PW_LENGTH_OPT * SNMPV3_MAX_PW_LENGTH) / 4)
+                                    // (64 * 64) / 4 = 1024
+                                    // for pw length > 64 we use global memory reads
+
 typedef struct hmac_sha256_tmp
 {
   u32 tmp[SNMPV3_TMP_ELEMS];
@@ -155,10 +160,7 @@ KERNEL_FQ void m26800_loop (KERN_ATTR_TMPS_ESALT (hmac_sha256_tmp_t, snmpv3_t))
 
   const int pw_len64 = pw_len * 64;
 
-  #define SNMPV3_TMP_ELEMS_OPT 1024 // 1024 = (64 max pw length * 64) / sizeof (u32)
-                                    // for pw length > 64 we use global memory reads
-
-  if (pw_len < 64)
+  if (pw_len <= SNMPV3_MAX_PW_LENGTH_OPT)
   {
     u32 tmp[SNMPV3_TMP_ELEMS_OPT];
 

docs/changes.txt

@@ -56,6 +56,8 @@
 - Added hash-mode: SNMPv3 HMAC-SHA1-96
 - Added hash-mode: SNMPv3 HMAC-SHA224-128
 - Added hash-mode: SNMPv3 HMAC-SHA256-192
+- Added hash-mode: SNMPv3 HMAC-SHA384-256
+- Added hash-mode: SNMPv3 HMAC-SHA512-384
 
 * changes v6.2.2 -> v6.2.3
 

docs/readme.txt

@@ -158,8 +158,10 @@ NVIDIA GPUs require "NVIDIA Driver" (440.64 or later) and "CUDA Toolkit" (9.0 or
 - SNMPv3 HMAC-MD5-96/HMAC-SHA1-96
 - SNMPv3 HMAC-MD5-96
 - SNMPv3 HMAC-SHA1-96
-- SNMPv3 HMAC-SHA256-192
 - SNMPv3 HMAC-SHA224-128
+- SNMPv3 HMAC-SHA256-192
+- SNMPv3 HMAC-SHA384-256
+- SNMPv3 HMAC-SHA512-384
 - WPA-EAPOL-PBKDF2
 - WPA-EAPOL-PMK
 - WPA-PBKDF2-PMKID+EAPOL