diff --git a/OpenCL/inc_hash_md4.cl b/OpenCL/inc_hash_md4.cl
index d02add1fb..425aa3e09 100644
--- a/OpenCL/inc_hash_md4.cl
+++ b/OpenCL/inc_hash_md4.cl
@@ -1057,29 +1057,24 @@ DECLSPEC void md4_hmac_final (md4_hmac_ctx_t *ctx)
 {
   md4_final (&ctx->ipad);
 
-  u32 t0[4];
-  u32 t1[4];
-  u32 t2[4];
-  u32 t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[0];
+  ctx->opad.w0[1] = ctx->ipad.h[1];
+  ctx->opad.w0[2] = ctx->ipad.h[2];
+  ctx->opad.w0[3] = ctx->ipad.h[3];
+  ctx->opad.w1[0] = 0;
+  ctx->opad.w1[1] = 0;
+  ctx->opad.w1[2] = 0;
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = ctx->ipad.h[0];
-  t0[1] = ctx->ipad.h[1];
-  t0[2] = ctx->ipad.h[2];
-  t0[3] = ctx->ipad.h[3];
-  t1[0] = 0;
-  t1[1] = 0;
-  t1[2] = 0;
-  t1[3] = 0;
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  md4_update_64 (&ctx->opad, t0, t1, t2, t3, 16);
+  ctx->opad.len += 16;
 
   md4_final (&ctx->opad);
 }
@@ -1695,29 +1690,24 @@ DECLSPEC void md4_hmac_final_vector (md4_hmac_ctx_vector_t *ctx)
 {
   md4_final_vector (&ctx->ipad);
 
-  u32x t0[4];
-  u32x t1[4];
-  u32x t2[4];
-  u32x t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[0];
+  ctx->opad.w0[1] = ctx->ipad.h[1];
+  ctx->opad.w0[2] = ctx->ipad.h[2];
+  ctx->opad.w0[3] = ctx->ipad.h[3];
+  ctx->opad.w1[0] = 0;
+  ctx->opad.w1[1] = 0;
+  ctx->opad.w1[2] = 0;
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = ctx->ipad.h[0];
-  t0[1] = ctx->ipad.h[1];
-  t0[2] = ctx->ipad.h[2];
-  t0[3] = ctx->ipad.h[3];
-  t1[0] = 0;
-  t1[1] = 0;
-  t1[2] = 0;
-  t1[3] = 0;
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  md4_update_vector_64 (&ctx->opad, t0, t1, t2, t3, 16);
+  ctx->opad.len += 16;
 
   md4_final_vector (&ctx->opad);
 }
diff --git a/OpenCL/inc_hash_md5.cl b/OpenCL/inc_hash_md5.cl
index b88bb32ea..7a7dbc3c3 100644
--- a/OpenCL/inc_hash_md5.cl
+++ b/OpenCL/inc_hash_md5.cl
@@ -1093,29 +1093,24 @@ DECLSPEC void md5_hmac_final (md5_hmac_ctx_t *ctx)
 {
   md5_final (&ctx->ipad);
 
-  u32 t0[4];
-  u32 t1[4];
-  u32 t2[4];
-  u32 t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[0];
+  ctx->opad.w0[1] = ctx->ipad.h[1];
+  ctx->opad.w0[2] = ctx->ipad.h[2];
+  ctx->opad.w0[3] = ctx->ipad.h[3];
+  ctx->opad.w1[0] = 0;
+  ctx->opad.w1[1] = 0;
+  ctx->opad.w1[2] = 0;
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = ctx->ipad.h[0];
-  t0[1] = ctx->ipad.h[1];
-  t0[2] = ctx->ipad.h[2];
-  t0[3] = ctx->ipad.h[3];
-  t1[0] = 0;
-  t1[1] = 0;
-  t1[2] = 0;
-  t1[3] = 0;
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  md5_update_64 (&ctx->opad, t0, t1, t2, t3, 16);
+  ctx->opad.len += 16;
 
   md5_final (&ctx->opad);
 }
@@ -1767,29 +1762,24 @@ DECLSPEC void md5_hmac_final_vector (md5_hmac_ctx_vector_t *ctx)
 {
   md5_final_vector (&ctx->ipad);
 
-  u32x t0[4];
-  u32x t1[4];
-  u32x t2[4];
-  u32x t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[0];
+  ctx->opad.w0[1] = ctx->ipad.h[1];
+  ctx->opad.w0[2] = ctx->ipad.h[2];
+  ctx->opad.w0[3] = ctx->ipad.h[3];
+  ctx->opad.w1[0] = 0;
+  ctx->opad.w1[1] = 0;
+  ctx->opad.w1[2] = 0;
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = ctx->ipad.h[0];
-  t0[1] = ctx->ipad.h[1];
-  t0[2] = ctx->ipad.h[2];
-  t0[3] = ctx->ipad.h[3];
-  t1[0] = 0;
-  t1[1] = 0;
-  t1[2] = 0;
-  t1[3] = 0;
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  md5_update_vector_64 (&ctx->opad, t0, t1, t2, t3, 16);
+  ctx->opad.len += 16;
 
   md5_final_vector (&ctx->opad);
 }
diff --git a/OpenCL/inc_hash_ripemd160.cl b/OpenCL/inc_hash_ripemd160.cl
index a02e30452..9434bf527 100644
--- a/OpenCL/inc_hash_ripemd160.cl
+++ b/OpenCL/inc_hash_ripemd160.cl
@@ -1191,29 +1191,24 @@ DECLSPEC void ripemd160_hmac_final (ripemd160_hmac_ctx_t *ctx)
 {
   ripemd160_final (&ctx->ipad);
 
-  u32 t0[4];
-  u32 t1[4];
-  u32 t2[4];
-  u32 t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[0];
+  ctx->opad.w0[1] = ctx->ipad.h[1];
+  ctx->opad.w0[2] = ctx->ipad.h[2];
+  ctx->opad.w0[3] = ctx->ipad.h[3];
+  ctx->opad.w1[0] = ctx->ipad.h[4];
+  ctx->opad.w1[1] = 0;
+  ctx->opad.w1[2] = 0;
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = ctx->ipad.h[0];
-  t0[1] = ctx->ipad.h[1];
-  t0[2] = ctx->ipad.h[2];
-  t0[3] = ctx->ipad.h[3];
-  t1[0] = ctx->ipad.h[4];
-  t1[1] = 0;
-  t1[2] = 0;
-  t1[3] = 0;
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  ripemd160_update_64 (&ctx->opad, t0, t1, t2, t3, 20);
+  ctx->opad.len += 20;
 
   ripemd160_final (&ctx->opad);
 }
@@ -1964,29 +1959,24 @@ DECLSPEC void ripemd160_hmac_final_vector (ripemd160_hmac_ctx_vector_t *ctx)
 {
   ripemd160_final_vector (&ctx->ipad);
 
-  u32x t0[4];
-  u32x t1[4];
-  u32x t2[4];
-  u32x t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[0];
+  ctx->opad.w0[1] = ctx->ipad.h[1];
+  ctx->opad.w0[2] = ctx->ipad.h[2];
+  ctx->opad.w0[3] = ctx->ipad.h[3];
+  ctx->opad.w1[0] = ctx->ipad.h[4];
+  ctx->opad.w1[1] = 0;
+  ctx->opad.w1[2] = 0;
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = ctx->ipad.h[0];
-  t0[1] = ctx->ipad.h[1];
-  t0[2] = ctx->ipad.h[2];
-  t0[3] = ctx->ipad.h[3];
-  t1[0] = ctx->ipad.h[4];
-  t1[1] = 0;
-  t1[2] = 0;
-  t1[3] = 0;
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  ripemd160_update_vector_64 (&ctx->opad, t0, t1, t2, t3, 20);
+  ctx->opad.len += 20;
 
   ripemd160_final_vector (&ctx->opad);
 }
diff --git a/OpenCL/inc_hash_sha1.cl b/OpenCL/inc_hash_sha1.cl
index 4efea16f0..4638139db 100644
--- a/OpenCL/inc_hash_sha1.cl
+++ b/OpenCL/inc_hash_sha1.cl
@@ -1359,29 +1359,24 @@ DECLSPEC void sha1_hmac_final (sha1_hmac_ctx_t *ctx)
 {
   sha1_final (&ctx->ipad);
 
-  u32 t0[4];
-  u32 t1[4];
-  u32 t2[4];
-  u32 t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[0];
+  ctx->opad.w0[1] = ctx->ipad.h[1];
+  ctx->opad.w0[2] = ctx->ipad.h[2];
+  ctx->opad.w0[3] = ctx->ipad.h[3];
+  ctx->opad.w1[0] = ctx->ipad.h[4];
+  ctx->opad.w1[1] = 0;
+  ctx->opad.w1[2] = 0;
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = ctx->ipad.h[0];
-  t0[1] = ctx->ipad.h[1];
-  t0[2] = ctx->ipad.h[2];
-  t0[3] = ctx->ipad.h[3];
-  t1[0] = ctx->ipad.h[4];
-  t1[1] = 0;
-  t1[2] = 0;
-  t1[3] = 0;
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  sha1_update_64 (&ctx->opad, t0, t1, t2, t3, 20);
+  ctx->opad.len += 20;
 
   sha1_final (&ctx->opad);
 }
@@ -2148,29 +2143,24 @@ DECLSPEC void sha1_hmac_final_vector (sha1_hmac_ctx_vector_t *ctx)
 {
   sha1_final_vector (&ctx->ipad);
 
-  u32x t0[4];
-  u32x t1[4];
-  u32x t2[4];
-  u32x t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[0];
+  ctx->opad.w0[1] = ctx->ipad.h[1];
+  ctx->opad.w0[2] = ctx->ipad.h[2];
+  ctx->opad.w0[3] = ctx->ipad.h[3];
+  ctx->opad.w1[0] = ctx->ipad.h[4];
+  ctx->opad.w1[1] = 0;
+  ctx->opad.w1[2] = 0;
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = ctx->ipad.h[0];
-  t0[1] = ctx->ipad.h[1];
-  t0[2] = ctx->ipad.h[2];
-  t0[3] = ctx->ipad.h[3];
-  t1[0] = ctx->ipad.h[4];
-  t1[1] = 0;
-  t1[2] = 0;
-  t1[3] = 0;
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  sha1_update_vector_64 (&ctx->opad, t0, t1, t2, t3, 20);
+  ctx->opad.len += 20;
 
   sha1_final_vector (&ctx->opad);
 }
diff --git a/OpenCL/inc_hash_sha224.cl b/OpenCL/inc_hash_sha224.cl
index e8075b707..d64f6a118 100644
--- a/OpenCL/inc_hash_sha224.cl
+++ b/OpenCL/inc_hash_sha224.cl
@@ -1108,29 +1108,24 @@ DECLSPEC void sha224_hmac_final (sha224_hmac_ctx_t *ctx)
 {
   sha224_final (&ctx->ipad);
 
-  u32 t0[4];
-  u32 t1[4];
-  u32 t2[4];
-  u32 t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[0];
+  ctx->opad.w0[1] = ctx->ipad.h[1];
+  ctx->opad.w0[2] = ctx->ipad.h[2];
+  ctx->opad.w0[3] = ctx->ipad.h[3];
+  ctx->opad.w1[0] = ctx->ipad.h[4];
+  ctx->opad.w1[1] = ctx->ipad.h[5];
+  ctx->opad.w1[2] = ctx->ipad.h[6];
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = ctx->ipad.h[0];
-  t0[1] = ctx->ipad.h[1];
-  t0[2] = ctx->ipad.h[2];
-  t0[3] = ctx->ipad.h[3];
-  t1[0] = ctx->ipad.h[4];
-  t1[1] = ctx->ipad.h[5];
-  t1[2] = ctx->ipad.h[6];
-  t1[3] = 0;
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  sha224_update_64 (&ctx->opad, t0, t1, t2, t3, 28);
+  ctx->opad.len += 28;
 
   sha224_final (&ctx->opad);
 }
@@ -1823,29 +1818,24 @@ DECLSPEC void sha224_hmac_final_vector (sha224_hmac_ctx_vector_t *ctx)
 {
   sha224_final_vector (&ctx->ipad);
 
-  u32x t0[4];
-  u32x t1[4];
-  u32x t2[4];
-  u32x t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[0];
+  ctx->opad.w0[1] = ctx->ipad.h[1];
+  ctx->opad.w0[2] = ctx->ipad.h[2];
+  ctx->opad.w0[3] = ctx->ipad.h[3];
+  ctx->opad.w1[0] = ctx->ipad.h[4];
+  ctx->opad.w1[1] = ctx->ipad.h[5];
+  ctx->opad.w1[2] = ctx->ipad.h[6];
+  ctx->opad.w1[3] = 0:
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = ctx->ipad.h[0];
-  t0[1] = ctx->ipad.h[1];
-  t0[2] = ctx->ipad.h[2];
-  t0[3] = ctx->ipad.h[3];
-  t1[0] = ctx->ipad.h[4];
-  t1[1] = ctx->ipad.h[5];
-  t1[2] = ctx->ipad.h[6];
-  t1[3] = 0;
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  sha224_update_vector_64 (&ctx->opad, t0, t1, t2, t3, 28);
+  ctx->opad.len += 28;
 
   sha224_final_vector (&ctx->opad);
 }
diff --git a/OpenCL/inc_hash_sha256.cl b/OpenCL/inc_hash_sha256.cl
index ee0cb762e..f47140950 100644
--- a/OpenCL/inc_hash_sha256.cl
+++ b/OpenCL/inc_hash_sha256.cl
@@ -1108,29 +1108,24 @@ DECLSPEC void sha256_hmac_final (sha256_hmac_ctx_t *ctx)
 {
   sha256_final (&ctx->ipad);
 
-  u32 t0[4];
-  u32 t1[4];
-  u32 t2[4];
-  u32 t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[0];
+  ctx->opad.w0[1] = ctx->ipad.h[1];
+  ctx->opad.w0[2] = ctx->ipad.h[2];
+  ctx->opad.w0[3] = ctx->ipad.h[3];
+  ctx->opad.w1[0] = ctx->ipad.h[4];
+  ctx->opad.w1[1] = ctx->ipad.h[5];
+  ctx->opad.w1[2] = ctx->ipad.h[6];
+  ctx->opad.w1[3] = ctx->ipad.h[7];
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = ctx->ipad.h[0];
-  t0[1] = ctx->ipad.h[1];
-  t0[2] = ctx->ipad.h[2];
-  t0[3] = ctx->ipad.h[3];
-  t1[0] = ctx->ipad.h[4];
-  t1[1] = ctx->ipad.h[5];
-  t1[2] = ctx->ipad.h[6];
-  t1[3] = ctx->ipad.h[7];
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  sha256_update_64 (&ctx->opad, t0, t1, t2, t3, 32);
+  ctx->opad.len += 32;
 
   sha256_final (&ctx->opad);
 }
@@ -1823,29 +1818,24 @@ DECLSPEC void sha256_hmac_final_vector (sha256_hmac_ctx_vector_t *ctx)
 {
   sha256_final_vector (&ctx->ipad);
 
-  u32x t0[4];
-  u32x t1[4];
-  u32x t2[4];
-  u32x t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[0];
+  ctx->opad.w0[1] = ctx->ipad.h[1];
+  ctx->opad.w0[2] = ctx->ipad.h[2];
+  ctx->opad.w0[3] = ctx->ipad.h[3];
+  ctx->opad.w1[0] = ctx->ipad.h[4];
+  ctx->opad.w1[1] = ctx->ipad.h[5];
+  ctx->opad.w1[2] = ctx->ipad.h[6];
+  ctx->opad.w1[3] = ctx->ipad.h[7];
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = ctx->ipad.h[0];
-  t0[1] = ctx->ipad.h[1];
-  t0[2] = ctx->ipad.h[2];
-  t0[3] = ctx->ipad.h[3];
-  t1[0] = ctx->ipad.h[4];
-  t1[1] = ctx->ipad.h[5];
-  t1[2] = ctx->ipad.h[6];
-  t1[3] = ctx->ipad.h[7];
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  sha256_update_vector_64 (&ctx->opad, t0, t1, t2, t3, 32);
+  ctx->opad.len += 32;
 
   sha256_final_vector (&ctx->opad);
 }
diff --git a/OpenCL/inc_hash_sha384.cl b/OpenCL/inc_hash_sha384.cl
index b37e3a003..b490e756b 100644
--- a/OpenCL/inc_hash_sha384.cl
+++ b/OpenCL/inc_hash_sha384.cl
@@ -1744,49 +1744,40 @@ DECLSPEC void sha384_hmac_final (sha384_hmac_ctx_t *ctx)
 {
   sha384_final (&ctx->ipad);
 
-  u32 t0[4];
-  u32 t1[4];
-  u32 t2[4];
-  u32 t3[4];
-  u32 t4[4];
-  u32 t5[4];
-  u32 t6[4];
-  u32 t7[4];
+  ctx->opad.w0[0] = h32_from_64_S (ctx->ipad.h[0]);
+  ctx->opad.w0[1] = l32_from_64_S (ctx->ipad.h[0]);
+  ctx->opad.w0[2] = h32_from_64_S (ctx->ipad.h[1]);
+  ctx->opad.w0[3] = l32_from_64_S (ctx->ipad.h[1]);
+  ctx->opad.w1[0] = h32_from_64_S (ctx->ipad.h[2]);
+  ctx->opad.w1[1] = l32_from_64_S (ctx->ipad.h[2]);
+  ctx->opad.w1[2] = h32_from_64_S (ctx->ipad.h[3]);
+  ctx->opad.w1[3] = l32_from_64_S (ctx->ipad.h[3]);
+  ctx->opad.w2[0] = h32_from_64_S (ctx->ipad.h[4]);
+  ctx->opad.w2[1] = l32_from_64_S (ctx->ipad.h[4]);
+  ctx->opad.w2[2] = h32_from_64_S (ctx->ipad.h[5]);
+  ctx->opad.w2[3] = l32_from_64_S (ctx->ipad.h[5]);
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
+  ctx->opad.w4[0] = 0;
+  ctx->opad.w4[1] = 0;
+  ctx->opad.w4[2] = 0;
+  ctx->opad.w4[3] = 0;
+  ctx->opad.w5[0] = 0;
+  ctx->opad.w5[1] = 0;
+  ctx->opad.w5[2] = 0;
+  ctx->opad.w5[3] = 0;
+  ctx->opad.w6[0] = 0;
+  ctx->opad.w6[1] = 0;
+  ctx->opad.w6[2] = 0;
+  ctx->opad.w6[3] = 0;
+  ctx->opad.w7[0] = 0;
+  ctx->opad.w7[1] = 0;
+  ctx->opad.w7[2] = 0;
+  ctx->opad.w7[3] = 0;
 
-  t0[0] = h32_from_64_S (ctx->ipad.h[0]);
-  t0[1] = l32_from_64_S (ctx->ipad.h[0]);
-  t0[2] = h32_from_64_S (ctx->ipad.h[1]);
-  t0[3] = l32_from_64_S (ctx->ipad.h[1]);
-  t1[0] = h32_from_64_S (ctx->ipad.h[2]);
-  t1[1] = l32_from_64_S (ctx->ipad.h[2]);
-  t1[2] = h32_from_64_S (ctx->ipad.h[3]);
-  t1[3] = l32_from_64_S (ctx->ipad.h[3]);
-  t2[0] = h32_from_64_S (ctx->ipad.h[4]);
-  t2[1] = l32_from_64_S (ctx->ipad.h[4]);
-  t2[2] = h32_from_64_S (ctx->ipad.h[5]);
-  t2[3] = l32_from_64_S (ctx->ipad.h[5]);
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-  t4[0] = 0;
-  t4[1] = 0;
-  t4[2] = 0;
-  t4[3] = 0;
-  t5[0] = 0;
-  t5[1] = 0;
-  t5[2] = 0;
-  t5[3] = 0;
-  t6[0] = 0;
-  t6[1] = 0;
-  t6[2] = 0;
-  t6[3] = 0;
-  t7[0] = 0;
-  t7[1] = 0;
-  t7[2] = 0;
-  t7[3] = 0;
-
-  sha384_update_128 (&ctx->opad, t0, t1, t2, t3, t4, t5, t6, t7, 48);
+  ctx->opad.len += 48;
 
   sha384_final (&ctx->opad);
 }
@@ -2859,49 +2850,40 @@ DECLSPEC void sha384_hmac_final_vector (sha384_hmac_ctx_vector_t *ctx)
 {
   sha384_final_vector (&ctx->ipad);
 
-  u32x t0[4];
-  u32x t1[4];
-  u32x t2[4];
-  u32x t3[4];
-  u32x t4[4];
-  u32x t5[4];
-  u32x t6[4];
-  u32x t7[4];
+  ctx->opad.w0[0] = h32_from_64 (ctx->ipad.h[0]);
+  ctx->opad.w0[1] = l32_from_64 (ctx->ipad.h[0]);
+  ctx->opad.w0[2] = h32_from_64 (ctx->ipad.h[1]);
+  ctx->opad.w0[3] = l32_from_64 (ctx->ipad.h[1]);
+  ctx->opad.w1[0] = h32_from_64 (ctx->ipad.h[2]);
+  ctx->opad.w1[1] = l32_from_64 (ctx->ipad.h[2]);
+  ctx->opad.w1[2] = h32_from_64 (ctx->ipad.h[3]);
+  ctx->opad.w1[3] = l32_from_64 (ctx->ipad.h[3]);
+  ctx->opad.w2[0] = h32_from_64 (ctx->ipad.h[4]);
+  ctx->opad.w2[1] = l32_from_64 (ctx->ipad.h[4]);
+  ctx->opad.w2[2] = h32_from_64 (ctx->ipad.h[5]);
+  ctx->opad.w2[3] = l32_from_64 (ctx->ipad.h[5]);
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
+  ctx->opad.w4[0] = 0;
+  ctx->opad.w4[1] = 0;
+  ctx->opad.w4[2] = 0;
+  ctx->opad.w4[3] = 0;
+  ctx->opad.w5[0] = 0;
+  ctx->opad.w5[1] = 0;
+  ctx->opad.w5[2] = 0;
+  ctx->opad.w5[3] = 0;
+  ctx->opad.w6[0] = 0;
+  ctx->opad.w6[1] = 0;
+  ctx->opad.w6[2] = 0;
+  ctx->opad.w6[3] = 0;
+  ctx->opad.w7[0] = 0;
+  ctx->opad.w7[1] = 0;
+  ctx->opad.w7[2] = 0;
+  ctx->opad.w7[3] = 0;
 
-  t0[0] = h32_from_64 (ctx->ipad.h[0]);
-  t0[1] = l32_from_64 (ctx->ipad.h[0]);
-  t0[2] = h32_from_64 (ctx->ipad.h[1]);
-  t0[3] = l32_from_64 (ctx->ipad.h[1]);
-  t1[0] = h32_from_64 (ctx->ipad.h[2]);
-  t1[1] = l32_from_64 (ctx->ipad.h[2]);
-  t1[2] = h32_from_64 (ctx->ipad.h[3]);
-  t1[3] = l32_from_64 (ctx->ipad.h[3]);
-  t2[0] = h32_from_64 (ctx->ipad.h[4]);
-  t2[1] = l32_from_64 (ctx->ipad.h[4]);
-  t2[2] = h32_from_64 (ctx->ipad.h[5]);
-  t2[3] = l32_from_64 (ctx->ipad.h[5]);
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-  t4[0] = 0;
-  t4[1] = 0;
-  t4[2] = 0;
-  t4[3] = 0;
-  t5[0] = 0;
-  t5[1] = 0;
-  t5[2] = 0;
-  t5[3] = 0;
-  t6[0] = 0;
-  t6[1] = 0;
-  t6[2] = 0;
-  t6[3] = 0;
-  t7[0] = 0;
-  t7[1] = 0;
-  t7[2] = 0;
-  t7[3] = 0;
-
-  sha384_update_vector_128 (&ctx->opad, t0, t1, t2, t3, t4, t5, t6, t7, 48);
+  ctx->opad.len += 48;
 
   sha384_final_vector (&ctx->opad);
 }
diff --git a/OpenCL/inc_hash_sha512.cl b/OpenCL/inc_hash_sha512.cl
index 66b71c958..a33488ac4 100644
--- a/OpenCL/inc_hash_sha512.cl
+++ b/OpenCL/inc_hash_sha512.cl
@@ -1861,49 +1861,40 @@ DECLSPEC void sha512_hmac_final (sha512_hmac_ctx_t *ctx)
 {
   sha512_final (&ctx->ipad);
 
-  u32 t0[4];
-  u32 t1[4];
-  u32 t2[4];
-  u32 t3[4];
-  u32 t4[4];
-  u32 t5[4];
-  u32 t6[4];
-  u32 t7[4];
+  ctx->opad.w0[0] = h32_from_64_S (ctx->ipad.h[0]);
+  ctx->opad.w0[1] = l32_from_64_S (ctx->ipad.h[0]);
+  ctx->opad.w0[2] = h32_from_64_S (ctx->ipad.h[1]);
+  ctx->opad.w0[3] = l32_from_64_S (ctx->ipad.h[1]);
+  ctx->opad.w1[0] = h32_from_64_S (ctx->ipad.h[2]);
+  ctx->opad.w1[1] = l32_from_64_S (ctx->ipad.h[2]);
+  ctx->opad.w1[2] = h32_from_64_S (ctx->ipad.h[3]);
+  ctx->opad.w1[3] = l32_from_64_S (ctx->ipad.h[3]);
+  ctx->opad.w2[0] = h32_from_64_S (ctx->ipad.h[4]);
+  ctx->opad.w2[1] = l32_from_64_S (ctx->ipad.h[4]);
+  ctx->opad.w2[2] = h32_from_64_S (ctx->ipad.h[5]);
+  ctx->opad.w2[3] = l32_from_64_S (ctx->ipad.h[5]);
+  ctx->opad.w3[0] = h32_from_64_S (ctx->ipad.h[6]);
+  ctx->opad.w3[1] = l32_from_64_S (ctx->ipad.h[6]);
+  ctx->opad.w3[2] = h32_from_64_S (ctx->ipad.h[7]);
+  ctx->opad.w3[3] = l32_from_64_S (ctx->ipad.h[7]);
+  ctx->opad.w4[0] = 0;
+  ctx->opad.w4[1] = 0;
+  ctx->opad.w4[2] = 0;
+  ctx->opad.w4[3] = 0;
+  ctx->opad.w5[0] = 0;
+  ctx->opad.w5[1] = 0;
+  ctx->opad.w5[2] = 0;
+  ctx->opad.w5[3] = 0;
+  ctx->opad.w6[0] = 0;
+  ctx->opad.w6[1] = 0;
+  ctx->opad.w6[2] = 0;
+  ctx->opad.w6[3] = 0;
+  ctx->opad.w7[0] = 0;
+  ctx->opad.w7[1] = 0;
+  ctx->opad.w7[2] = 0;
+  ctx->opad.w7[3] = 0;
 
-  t0[0] = h32_from_64_S (ctx->ipad.h[0]);
-  t0[1] = l32_from_64_S (ctx->ipad.h[0]);
-  t0[2] = h32_from_64_S (ctx->ipad.h[1]);
-  t0[3] = l32_from_64_S (ctx->ipad.h[1]);
-  t1[0] = h32_from_64_S (ctx->ipad.h[2]);
-  t1[1] = l32_from_64_S (ctx->ipad.h[2]);
-  t1[2] = h32_from_64_S (ctx->ipad.h[3]);
-  t1[3] = l32_from_64_S (ctx->ipad.h[3]);
-  t2[0] = h32_from_64_S (ctx->ipad.h[4]);
-  t2[1] = l32_from_64_S (ctx->ipad.h[4]);
-  t2[2] = h32_from_64_S (ctx->ipad.h[5]);
-  t2[3] = l32_from_64_S (ctx->ipad.h[5]);
-  t3[0] = h32_from_64_S (ctx->ipad.h[6]);
-  t3[1] = l32_from_64_S (ctx->ipad.h[6]);
-  t3[2] = h32_from_64_S (ctx->ipad.h[7]);
-  t3[3] = l32_from_64_S (ctx->ipad.h[7]);
-  t4[0] = 0;
-  t4[1] = 0;
-  t4[2] = 0;
-  t4[3] = 0;
-  t5[0] = 0;
-  t5[1] = 0;
-  t5[2] = 0;
-  t5[3] = 0;
-  t6[0] = 0;
-  t6[1] = 0;
-  t6[2] = 0;
-  t6[3] = 0;
-  t7[0] = 0;
-  t7[1] = 0;
-  t7[2] = 0;
-  t7[3] = 0;
-
-  sha512_update_128 (&ctx->opad, t0, t1, t2, t3, t4, t5, t6, t7, 64);
+  ctx->opad.len += 64;
 
   sha512_final (&ctx->opad);
 }
@@ -2976,49 +2967,40 @@ DECLSPEC void sha512_hmac_final_vector (sha512_hmac_ctx_vector_t *ctx)
 {
   sha512_final_vector (&ctx->ipad);
 
-  u32x t0[4];
-  u32x t1[4];
-  u32x t2[4];
-  u32x t3[4];
-  u32x t4[4];
-  u32x t5[4];
-  u32x t6[4];
-  u32x t7[4];
+  ctx->opad.w0[0] = h32_from_64 (ctx->ipad.h[0]);
+  ctx->opad.w0[1] = l32_from_64 (ctx->ipad.h[0]);
+  ctx->opad.w0[2] = h32_from_64 (ctx->ipad.h[1]);
+  ctx->opad.w0[3] = l32_from_64 (ctx->ipad.h[1]);
+  ctx->opad.w1[0] = h32_from_64 (ctx->ipad.h[2]);
+  ctx->opad.w1[1] = l32_from_64 (ctx->ipad.h[2]);
+  ctx->opad.w1[2] = h32_from_64 (ctx->ipad.h[3]);
+  ctx->opad.w1[3] = l32_from_64 (ctx->ipad.h[3]);
+  ctx->opad.w2[0] = h32_from_64 (ctx->ipad.h[4]);
+  ctx->opad.w2[1] = l32_from_64 (ctx->ipad.h[4]);
+  ctx->opad.w2[2] = h32_from_64 (ctx->ipad.h[5]);
+  ctx->opad.w2[3] = l32_from_64 (ctx->ipad.h[5]);
+  ctx->opad.w3[0] = h32_from_64 (ctx->ipad.h[6]);
+  ctx->opad.w3[1] = l32_from_64 (ctx->ipad.h[6]);
+  ctx->opad.w3[2] = h32_from_64 (ctx->ipad.h[7]);
+  ctx->opad.w3[3] = l32_from_64 (ctx->ipad.h[7]);
+  ctx->opad.w4[0] = 0;
+  ctx->opad.w4[1] = 0;
+  ctx->opad.w4[2] = 0;
+  ctx->opad.w4[3] = 0;
+  ctx->opad.w5[0] = 0;
+  ctx->opad.w5[1] = 0;
+  ctx->opad.w5[2] = 0;
+  ctx->opad.w5[3] = 0;
+  ctx->opad.w6[0] = 0;
+  ctx->opad.w6[1] = 0;
+  ctx->opad.w6[2] = 0;
+  ctx->opad.w6[3] = 0;
+  ctx->opad.w7[0] = 0;
+  ctx->opad.w7[1] = 0;
+  ctx->opad.w7[2] = 0;
+  ctx->opad.w7[3] = 0;
 
-  t0[0] = h32_from_64 (ctx->ipad.h[0]);
-  t0[1] = l32_from_64 (ctx->ipad.h[0]);
-  t0[2] = h32_from_64 (ctx->ipad.h[1]);
-  t0[3] = l32_from_64 (ctx->ipad.h[1]);
-  t1[0] = h32_from_64 (ctx->ipad.h[2]);
-  t1[1] = l32_from_64 (ctx->ipad.h[2]);
-  t1[2] = h32_from_64 (ctx->ipad.h[3]);
-  t1[3] = l32_from_64 (ctx->ipad.h[3]);
-  t2[0] = h32_from_64 (ctx->ipad.h[4]);
-  t2[1] = l32_from_64 (ctx->ipad.h[4]);
-  t2[2] = h32_from_64 (ctx->ipad.h[5]);
-  t2[3] = l32_from_64 (ctx->ipad.h[5]);
-  t3[0] = h32_from_64 (ctx->ipad.h[6]);
-  t3[1] = l32_from_64 (ctx->ipad.h[6]);
-  t3[2] = h32_from_64 (ctx->ipad.h[7]);
-  t3[3] = l32_from_64 (ctx->ipad.h[7]);
-  t4[0] = 0;
-  t4[1] = 0;
-  t4[2] = 0;
-  t4[3] = 0;
-  t5[0] = 0;
-  t5[1] = 0;
-  t5[2] = 0;
-  t5[3] = 0;
-  t6[0] = 0;
-  t6[1] = 0;
-  t6[2] = 0;
-  t6[3] = 0;
-  t7[0] = 0;
-  t7[1] = 0;
-  t7[2] = 0;
-  t7[3] = 0;
-
-  sha512_update_vector_128 (&ctx->opad, t0, t1, t2, t3, t4, t5, t6, t7, 64);
+  ctx->opad.len += 64;
 
   sha512_final_vector (&ctx->opad);
 }
diff --git a/OpenCL/inc_hash_streebog256.cl b/OpenCL/inc_hash_streebog256.cl
index 7f64d4bb6..13b74beda 100644
--- a/OpenCL/inc_hash_streebog256.cl
+++ b/OpenCL/inc_hash_streebog256.cl
@@ -1227,29 +1227,24 @@ DECLSPEC void streebog256_hmac_final (streebog256_hmac_ctx_t *ctx)
 {
   streebog256_final (&ctx->ipad);
 
-  u32 t0[4];
-  u32 t1[4];
-  u32 t2[4];
-  u32 t3[4];
+  ctx->opad.w0[0] = h32_from_64_S (ctx->ipad.h[3]);
+  ctx->opad.w0[1] = l32_from_64_S (ctx->ipad.h[3]);
+  ctx->opad.w0[2] = h32_from_64_S (ctx->ipad.h[2]);
+  ctx->opad.w0[3] = l32_from_64_S (ctx->ipad.h[2]);
+  ctx->opad.w1[0] = h32_from_64_S (ctx->ipad.h[1]);
+  ctx->opad.w1[1] = l32_from_64_S (ctx->ipad.h[1]);
+  ctx->opad.w1[2] = h32_from_64_S (ctx->ipad.h[0]);
+  ctx->opad.w1[3] = l32_from_64_S (ctx->ipad.h[0]);
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = h32_from_64_S (ctx->ipad.h[3]);
-  t0[1] = l32_from_64_S (ctx->ipad.h[3]);
-  t0[2] = h32_from_64_S (ctx->ipad.h[2]);
-  t0[3] = l32_from_64_S (ctx->ipad.h[2]);
-  t1[0] = h32_from_64_S (ctx->ipad.h[1]);
-  t1[1] = l32_from_64_S (ctx->ipad.h[1]);
-  t1[2] = h32_from_64_S (ctx->ipad.h[0]);
-  t1[3] = l32_from_64_S (ctx->ipad.h[0]);
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  streebog256_update_64 (&ctx->opad, t0, t1, t2, t3, 32);
+  ctx->opad.len += 32;
 
   streebog256_final (&ctx->opad);
 }
@@ -1822,29 +1817,24 @@ DECLSPEC void streebog256_hmac_final_vector (streebog256_hmac_ctx_vector_t *ctx)
 {
   streebog256_final_vector (&ctx->ipad);
 
-  u32x t0[4];
-  u32x t1[4];
-  u32x t2[4];
-  u32x t3[4];
+  ctx->opad.w0[0] = h32_from_64 (ctx->ipad.h[3]);
+  ctx->opad.w0[1] = l32_from_64 (ctx->ipad.h[3]);
+  ctx->opad.w0[2] = h32_from_64 (ctx->ipad.h[2]);
+  ctx->opad.w0[3] = l32_from_64 (ctx->ipad.h[2]);
+  ctx->opad.w1[0] = h32_from_64 (ctx->ipad.h[1]);
+  ctx->opad.w1[1] = l32_from_64 (ctx->ipad.h[1]);
+  ctx->opad.w1[2] = h32_from_64 (ctx->ipad.h[0]);
+  ctx->opad.w1[3] = l32_from_64 (ctx->ipad.h[0]);
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
-  t0[0] = h32_from_64 (ctx->ipad.h[3]);
-  t0[1] = l32_from_64 (ctx->ipad.h[3]);
-  t0[2] = h32_from_64 (ctx->ipad.h[2]);
-  t0[3] = l32_from_64 (ctx->ipad.h[2]);
-  t1[0] = h32_from_64 (ctx->ipad.h[1]);
-  t1[1] = l32_from_64 (ctx->ipad.h[1]);
-  t1[2] = h32_from_64 (ctx->ipad.h[0]);
-  t1[3] = l32_from_64 (ctx->ipad.h[0]);
-  t2[0] = 0;
-  t2[1] = 0;
-  t2[2] = 0;
-  t2[3] = 0;
-  t3[0] = 0;
-  t3[1] = 0;
-  t3[2] = 0;
-  t3[3] = 0;
-
-  streebog256_update_vector_64 (&ctx->opad, t0, t1, t2, t3, 32);
+  ctx->opad.len += 32;
 
   streebog256_final_vector (&ctx->opad);
 }
diff --git a/OpenCL/inc_hash_streebog512.cl b/OpenCL/inc_hash_streebog512.cl
index 5ef89421e..4018ee8e3 100644
--- a/OpenCL/inc_hash_streebog512.cl
+++ b/OpenCL/inc_hash_streebog512.cl
@@ -1227,29 +1227,43 @@ DECLSPEC void streebog512_hmac_final (streebog512_hmac_ctx_t *ctx)
 {
   streebog512_final (&ctx->ipad);
 
-  u32 t0[4];
-  u32 t1[4];
-  u32 t2[4];
-  u32 t3[4];
+  ctx->opad.w0[0] = h32_from_64_S (ctx->ipad.h[7]);
+  ctx->opad.w0[1] = l32_from_64_S (ctx->ipad.h[7]);
+  ctx->opad.w0[2] = h32_from_64_S (ctx->ipad.h[6]);
+  ctx->opad.w0[3] = l32_from_64_S (ctx->ipad.h[6]);
+  ctx->opad.w1[0] = h32_from_64_S (ctx->ipad.h[5]);
+  ctx->opad.w1[1] = l32_from_64_S (ctx->ipad.h[5]);
+  ctx->opad.w1[2] = h32_from_64_S (ctx->ipad.h[4]);
+  ctx->opad.w1[3] = l32_from_64_S (ctx->ipad.h[4]);
+  ctx->opad.w2[0] = h32_from_64_S (ctx->ipad.h[3]);
+  ctx->opad.w2[1] = l32_from_64_S (ctx->ipad.h[3]);
+  ctx->opad.w2[2] = h32_from_64_S (ctx->ipad.h[2]);
+  ctx->opad.w2[3] = l32_from_64_S (ctx->ipad.h[2]);
+  ctx->opad.w3[0] = h32_from_64_S (ctx->ipad.h[1]);
+  ctx->opad.w3[1] = l32_from_64_S (ctx->ipad.h[1]);
+  ctx->opad.w3[2] = h32_from_64_S (ctx->ipad.h[0]);
+  ctx->opad.w3[3] = l32_from_64_S (ctx->ipad.h[0]);
 
-  t0[0] = h32_from_64_S (ctx->ipad.h[7]);
-  t0[1] = l32_from_64_S (ctx->ipad.h[7]);
-  t0[2] = h32_from_64_S (ctx->ipad.h[6]);
-  t0[3] = l32_from_64_S (ctx->ipad.h[6]);
-  t1[0] = h32_from_64_S (ctx->ipad.h[5]);
-  t1[1] = l32_from_64_S (ctx->ipad.h[5]);
-  t1[2] = h32_from_64_S (ctx->ipad.h[4]);
-  t1[3] = l32_from_64_S (ctx->ipad.h[4]);
-  t2[0] = h32_from_64_S (ctx->ipad.h[3]);
-  t2[1] = l32_from_64_S (ctx->ipad.h[3]);
-  t2[2] = h32_from_64_S (ctx->ipad.h[2]);
-  t2[3] = l32_from_64_S (ctx->ipad.h[2]);
-  t3[0] = h32_from_64_S (ctx->ipad.h[1]);
-  t3[1] = l32_from_64_S (ctx->ipad.h[1]);
-  t3[2] = h32_from_64_S (ctx->ipad.h[0]);
-  t3[3] = l32_from_64_S (ctx->ipad.h[0]);
+  ctx->opad.len = 0;
 
-  streebog512_update_64 (&ctx->opad, t0, t1, t2, t3, 64);
+  streebog512_transform (&ctx->opad, ctx->opad.w0, ctx->opad.w1, ctx->opad.w2, ctx->opad.w3);
+
+  ctx->opad.w0[0] = 0;
+  ctx->opad.w0[1] = 0;
+  ctx->opad.w0[2] = 0;
+  ctx->opad.w0[3] = 0;
+  ctx->opad.w1[0] = 0;
+  ctx->opad.w1[1] = 0;
+  ctx->opad.w1[2] = 0;
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
   streebog512_final (&ctx->opad);
 }
@@ -1822,29 +1836,43 @@ DECLSPEC void streebog512_hmac_final_vector (streebog512_hmac_ctx_vector_t *ctx)
 {
   streebog512_final_vector (&ctx->ipad);
 
-  u32x t0[4];
-  u32x t1[4];
-  u32x t2[4];
-  u32x t3[4];
+  ctx->opad.w0[0] = h32_from_64 (ctx->ipad.h[7]);
+  ctx->opad.w0[1] = l32_from_64 (ctx->ipad.h[7]);
+  ctx->opad.w0[2] = h32_from_64 (ctx->ipad.h[6]);
+  ctx->opad.w0[3] = l32_from_64 (ctx->ipad.h[6]);
+  ctx->opad.w1[0] = h32_from_64 (ctx->ipad.h[5]);
+  ctx->opad.w1[1] = l32_from_64 (ctx->ipad.h[5]);
+  ctx->opad.w1[2] = h32_from_64 (ctx->ipad.h[4]);
+  ctx->opad.w1[3] = l32_from_64 (ctx->ipad.h[4]);
+  ctx->opad.w2[0] = h32_from_64 (ctx->ipad.h[3]);
+  ctx->opad.w2[1] = l32_from_64 (ctx->ipad.h[3]);
+  ctx->opad.w2[2] = h32_from_64 (ctx->ipad.h[2]);
+  ctx->opad.w2[3] = l32_from_64 (ctx->ipad.h[2]);
+  ctx->opad.w3[0] = h32_from_64 (ctx->ipad.h[1]);
+  ctx->opad.w3[1] = l32_from_64 (ctx->ipad.h[1]);
+  ctx->opad.w3[2] = h32_from_64 (ctx->ipad.h[0]);
+  ctx->opad.w3[3] = l32_from_64 (ctx->ipad.h[0]);
 
-  t0[0] = h32_from_64 (ctx->ipad.h[7]);
-  t0[1] = l32_from_64 (ctx->ipad.h[7]);
-  t0[2] = h32_from_64 (ctx->ipad.h[6]);
-  t0[3] = l32_from_64 (ctx->ipad.h[6]);
-  t1[0] = h32_from_64 (ctx->ipad.h[5]);
-  t1[1] = l32_from_64 (ctx->ipad.h[5]);
-  t1[2] = h32_from_64 (ctx->ipad.h[4]);
-  t1[3] = l32_from_64 (ctx->ipad.h[4]);
-  t2[0] = h32_from_64 (ctx->ipad.h[3]);
-  t2[1] = l32_from_64 (ctx->ipad.h[3]);
-  t2[2] = h32_from_64 (ctx->ipad.h[2]);
-  t2[3] = l32_from_64 (ctx->ipad.h[2]);
-  t3[0] = h32_from_64 (ctx->ipad.h[1]);
-  t3[1] = l32_from_64 (ctx->ipad.h[1]);
-  t3[2] = h32_from_64 (ctx->ipad.h[0]);
-  t3[3] = l32_from_64 (ctx->ipad.h[0]);
+  ctx->opad.len = 0;
 
-  streebog512_update_vector_64 (&ctx->opad, t0, t1, t2, t3, 64);
+  streebog512_transform_vector (&ctx->opad, ctx->opad.w0, ctx->opad.w1, ctx->opad.w2, ctx->opad.w3);
+
+  ctx->opad.w0[0] = 0;
+  ctx->opad.w0[1] = 0;
+  ctx->opad.w0[2] = 0;
+  ctx->opad.w0[3] = 0;
+  ctx->opad.w1[0] = 0;
+  ctx->opad.w1[1] = 0;
+  ctx->opad.w1[2] = 0;
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
   streebog512_final_vector (&ctx->opad);
 }
diff --git a/OpenCL/inc_hash_whirlpool.cl b/OpenCL/inc_hash_whirlpool.cl
index 6e9c1b8ab..e233a0cf9 100644
--- a/OpenCL/inc_hash_whirlpool.cl
+++ b/OpenCL/inc_hash_whirlpool.cl
@@ -2272,29 +2272,43 @@ DECLSPEC void whirlpool_hmac_final (whirlpool_hmac_ctx_t *ctx)
 {
   whirlpool_final (&ctx->ipad);
 
-  u32 t0[4];
-  u32 t1[4];
-  u32 t2[4];
-  u32 t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[ 0];
+  ctx->opad.w0[1] = ctx->ipad.h[ 1];
+  ctx->opad.w0[2] = ctx->ipad.h[ 2];
+  ctx->opad.w0[3] = ctx->ipad.h[ 3];
+  ctx->opad.w1[0] = ctx->ipad.h[ 4];
+  ctx->opad.w1[1] = ctx->ipad.h[ 5];
+  ctx->opad.w1[2] = ctx->ipad.h[ 6];
+  ctx->opad.w1[3] = ctx->ipad.h[ 7];
+  ctx->opad.w2[0] = ctx->ipad.h[ 8];
+  ctx->opad.w2[1] = ctx->ipad.h[ 9];
+  ctx->opad.w2[2] = ctx->ipad.h[10];
+  ctx->opad.w2[3] = ctx->ipad.h[11];
+  ctx->opad.w3[0] = ctx->ipad.h[12];
+  ctx->opad.w3[1] = ctx->ipad.h[13];
+  ctx->opad.w3[2] = ctx->ipad.h[14];
+  ctx->opad.w3[3] = ctx->ipad.h[15];
 
-  t0[0] = ctx->ipad.h[ 0];
-  t0[1] = ctx->ipad.h[ 1];
-  t0[2] = ctx->ipad.h[ 2];
-  t0[3] = ctx->ipad.h[ 3];
-  t1[0] = ctx->ipad.h[ 4];
-  t1[1] = ctx->ipad.h[ 5];
-  t1[2] = ctx->ipad.h[ 6];
-  t1[3] = ctx->ipad.h[ 7];
-  t2[0] = ctx->ipad.h[ 8];
-  t2[1] = ctx->ipad.h[ 9];
-  t2[2] = ctx->ipad.h[10];
-  t2[3] = ctx->ipad.h[11];
-  t3[0] = ctx->ipad.h[12];
-  t3[1] = ctx->ipad.h[13];
-  t3[2] = ctx->ipad.h[14];
-  t3[3] = ctx->ipad.h[15];
+  ctx->opad.len += 64;
 
-  whirlpool_update_64 (&ctx->opad, t0, t1, t2, t3, 64);
+  whirlpool_transform (ctx->opad.w0, ctx->opad.w1, ctx->opad.w2, ctx->opad.w3, ctx->opad.h, ctx->opad.s_Ch, ctx->opad.s_Cl);
+
+  ctx->opad.w0[0] = 0;
+  ctx->opad.w0[1] = 0;
+  ctx->opad.w0[2] = 0;
+  ctx->opad.w0[3] = 0;
+  ctx->opad.w1[0] = 0;
+  ctx->opad.w1[1] = 0;
+  ctx->opad.w1[2] = 0;
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
   whirlpool_final (&ctx->opad);
 }
@@ -3046,29 +3060,43 @@ DECLSPEC void whirlpool_hmac_final_vector (whirlpool_hmac_ctx_vector_t *ctx)
 {
   whirlpool_final_vector (&ctx->ipad);
 
-  u32x t0[4];
-  u32x t1[4];
-  u32x t2[4];
-  u32x t3[4];
+  ctx->opad.w0[0] = ctx->ipad.h[ 0];
+  ctx->opad.w0[1] = ctx->ipad.h[ 1];
+  ctx->opad.w0[2] = ctx->ipad.h[ 2];
+  ctx->opad.w0[3] = ctx->ipad.h[ 3];
+  ctx->opad.w1[0] = ctx->ipad.h[ 4];
+  ctx->opad.w1[1] = ctx->ipad.h[ 5];
+  ctx->opad.w1[2] = ctx->ipad.h[ 6];
+  ctx->opad.w1[3] = ctx->ipad.h[ 7];
+  ctx->opad.w2[0] = ctx->ipad.h[ 8];
+  ctx->opad.w2[1] = ctx->ipad.h[ 9];
+  ctx->opad.w2[2] = ctx->ipad.h[10];
+  ctx->opad.w2[3] = ctx->ipad.h[11];
+  ctx->opad.w3[0] = ctx->ipad.h[12];
+  ctx->opad.w3[1] = ctx->ipad.h[13];
+  ctx->opad.w3[2] = ctx->ipad.h[14];
+  ctx->opad.w3[3] = ctx->ipad.h[15];
 
-  t0[0] = ctx->ipad.h[ 0];
-  t0[1] = ctx->ipad.h[ 1];
-  t0[2] = ctx->ipad.h[ 2];
-  t0[3] = ctx->ipad.h[ 3];
-  t1[0] = ctx->ipad.h[ 4];
-  t1[1] = ctx->ipad.h[ 5];
-  t1[2] = ctx->ipad.h[ 6];
-  t1[3] = ctx->ipad.h[ 7];
-  t2[0] = ctx->ipad.h[ 8];
-  t2[1] = ctx->ipad.h[ 9];
-  t2[2] = ctx->ipad.h[10];
-  t2[3] = ctx->ipad.h[11];
-  t3[0] = ctx->ipad.h[12];
-  t3[1] = ctx->ipad.h[13];
-  t3[2] = ctx->ipad.h[14];
-  t3[3] = ctx->ipad.h[15];
+  ctx->opad.len += 64;
 
-  whirlpool_update_vector_64 (&ctx->opad, t0, t1, t2, t3, 64);
+  whirlpool_transform_vector (ctx->opad.w0, ctx->opad.w1, ctx->opad.w2, ctx->opad.w3, ctx->opad.h, ctx->opad.s_Ch, ctx->opad.s_Cl);
+
+  ctx->opad.w0[0] = 0;
+  ctx->opad.w0[1] = 0;
+  ctx->opad.w0[2] = 0;
+  ctx->opad.w0[3] = 0;
+  ctx->opad.w1[0] = 0;
+  ctx->opad.w1[1] = 0;
+  ctx->opad.w1[2] = 0;
+  ctx->opad.w1[3] = 0;
+  ctx->opad.w2[0] = 0;
+  ctx->opad.w2[1] = 0;
+  ctx->opad.w2[2] = 0;
+  ctx->opad.w2[3] = 0;
+  ctx->opad.w3[0] = 0;
+  ctx->opad.w3[1] = 0;
+  ctx->opad.w3[2] = 0;
+  ctx->opad.w3[3] = 0;
 
   whirlpool_final_vector (&ctx->opad);
 }