diff --git a/OpenCL/m06900_a0.cl b/OpenCL/m06900_a0.cl index 1b1563224..f313b80fc 100644 --- a/OpenCL/m06900_a0.cl +++ b/OpenCL/m06900_a0.cl @@ -7,6 +7,8 @@ #define _GOST_ +#define NEW_SIMD_CODE + #include "include/constants.h" #include "include/kernel_vendor.h" @@ -20,9 +22,7 @@ #include "OpenCL/common.c" #include "include/rp_kernel.h" #include "OpenCL/rp.c" - -#define COMPARE_S "OpenCL/check_single_comp4.c" -#define COMPARE_M "OpenCL/check_multi_comp4.c" +#include "OpenCL/simd.c" __constant u32 c_tables[4][256] = { @@ -304,7 +304,7 @@ __constant u32 c_tables[4][256] = #define _round(k1,k2,tbl) \ { \ - u32 t; \ + u32x t; \ t = (k1) + r; \ l ^= BOX (((t >> 0) & 0xff), 0, tbl) ^ \ BOX (((t >> 8) & 0xff), 1, tbl) ^ \ @@ -319,8 +319,8 @@ __constant u32 c_tables[4][256] = #define R(k,h,s,i,t) \ { \ - u32 r; \ - u32 l; \ + u32x r; \ + u32x l; \ r = h[i + 0]; \ l = h[i + 1]; \ _round (k[0], k[1], t); \ @@ -389,8 +389,8 @@ __constant u32 c_tables[4][256] = #define A(x) \ { \ - u32 l; \ - u32 r; \ + u32x l; \ + u32x r; \ l = x[0] ^ x[2]; \ r = x[1] ^ x[3]; \ x[0] = x[2]; \ @@ -405,8 +405,8 @@ __constant u32 c_tables[4][256] = #define AA(x) \ { \ - u32 l; \ - u32 r; \ + u32x l; \ + u32x r; \ l = x[0]; \ r = x[2]; \ x[0] = x[4]; \ @@ -664,8 +664,8 @@ __constant u32 c_tables[4][256] = #define PASS0(h,s,u,v,t) \ { \ - u32 k[8]; \ - u32 w[8]; \ + u32x k[8]; \ + u32x w[8]; \ X (w, u, v); \ P (k, w); \ R (k, h, s, 0, t); \ @@ -675,8 +675,8 @@ __constant u32 c_tables[4][256] = #define PASS2(h,s,u,v,t) \ { \ - u32 k[8]; \ - u32 w[8]; \ + u32x k[8]; \ + u32x w[8]; \ X (w, u, v); \ P (k, w); \ R (k, h, s, 2, t); \ @@ -687,8 +687,8 @@ __constant u32 c_tables[4][256] = #define PASS4(h,s,u,v,t) \ { \ - u32 k[8]; \ - u32 w[8]; \ + u32x k[8]; \ + u32x w[8]; \ X (w, u, v); \ P (k, w); \ R (k, h, s, 4, t); \ @@ -698,8 +698,8 @@ __constant u32 c_tables[4][256] = #define PASS6(h,s,u,v,t) \ { \ - u32 k[8]; \ - u32 w[8]; \ + u32x k[8]; \ + u32x w[8]; \ X (w, u, v); \ P (k, w); \ R (k, h, s, 6, t); \ @@ -757,41 +757,18 @@ __kernel void m06900_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu * loop */ - for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++) + for (u32 il_pos = 0; il_pos < rules_cnt; il_pos += VECT_SIZE) { - u32 w0[4]; - - w0[0] = pw_buf0[0]; - w0[1] = pw_buf0[1]; - w0[2] = pw_buf0[2]; - w0[3] = pw_buf0[3]; - - u32 w1[4]; - - w1[0] = pw_buf1[0]; - w1[1] = pw_buf1[1]; - w1[2] = pw_buf1[2]; - w1[3] = pw_buf1[3]; - - u32 w2[4]; - - w2[0] = 0; - w2[1] = 0; - w2[2] = 0; - w2[3] = 0; - - u32 w3[4]; - - w3[0] = 0; - w3[1] = 0; - w3[2] = 0; - w3[3] = 0; + u32x w0[4] = { 0 }; + u32x w1[4] = { 0 }; + u32x w2[4] = { 0 }; + u32x w3[4] = { 0 }; - const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len); + const u32 out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1); u32 w14 = out_len * 8; - u32 data[8]; + u32x data[8]; data[0] = w0[0]; data[1] = w0[1]; @@ -802,7 +779,7 @@ __kernel void m06900_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu data[6] = w1[2]; data[7] = w1[3]; - u32 state[16]; + u32x state[16]; state[ 0] = 0; state[ 1] = 0; @@ -821,8 +798,8 @@ __kernel void m06900_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu state[14] = data[6]; state[15] = data[7]; - u32 state_m[8]; - u32 data_m[8]; + u32x state_m[8]; + u32x data_m[8]; /* gost1 */ @@ -844,9 +821,9 @@ __kernel void m06900_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu data_m[6] = data[6]; data_m[7] = data[7]; - u32 tmp[8]; + u32x tmp[8]; - if (out_len > 0) + if (pw_len > 0) { PASS0 (state, tmp, state_m, data_m, s_tables); PASS2 (state, tmp, state_m, data_m, s_tables); @@ -936,12 +913,7 @@ __kernel void m06900_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu /* store */ - const u32 r0 = state[0]; - const u32 r1 = state[1]; - const u32 r2 = state[2]; - const u32 r3 = state[3]; - - #include COMPARE_M + COMPARE_M_SIMD (state[0], state[1], state[2], state[3]); } } @@ -1017,41 +989,18 @@ __kernel void m06900_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu * loop */ - for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++) + for (u32 il_pos = 0; il_pos < rules_cnt; il_pos += VECT_SIZE) { - u32 w0[4]; - - w0[0] = pw_buf0[0]; - w0[1] = pw_buf0[1]; - w0[2] = pw_buf0[2]; - w0[3] = pw_buf0[3]; - - u32 w1[4]; - - w1[0] = pw_buf1[0]; - w1[1] = pw_buf1[1]; - w1[2] = pw_buf1[2]; - w1[3] = pw_buf1[3]; - - u32 w2[4]; + u32x w0[4] = { 0 }; + u32x w1[4] = { 0 }; + u32x w2[4] = { 0 }; + u32x w3[4] = { 0 }; - w2[0] = 0; - w2[1] = 0; - w2[2] = 0; - w2[3] = 0; - - u32 w3[4]; - - w3[0] = 0; - w3[1] = 0; - w3[2] = 0; - w3[3] = 0; - - const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len); + const u32 out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1); u32 w14 = out_len * 8; - u32 data[8]; + u32x data[8]; data[0] = w0[0]; data[1] = w0[1]; @@ -1062,7 +1011,7 @@ __kernel void m06900_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu data[6] = w1[2]; data[7] = w1[3]; - u32 state[16]; + u32x state[16]; state[ 0] = 0; state[ 1] = 0; @@ -1081,8 +1030,8 @@ __kernel void m06900_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu state[14] = data[6]; state[15] = data[7]; - u32 state_m[8]; - u32 data_m[8]; + u32x state_m[8]; + u32x data_m[8]; /* gost1 */ @@ -1104,9 +1053,9 @@ __kernel void m06900_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu data_m[6] = data[6]; data_m[7] = data[7]; - u32 tmp[8]; + u32x tmp[8]; - if (out_len > 0) + if (pw_len > 0) { PASS0 (state, tmp, state_m, data_m, s_tables); PASS2 (state, tmp, state_m, data_m, s_tables); @@ -1196,12 +1145,7 @@ __kernel void m06900_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu /* store */ - const u32 r0 = state[0]; - const u32 r1 = state[1]; - const u32 r2 = state[2]; - const u32 r3 = state[3]; - - #include COMPARE_S + COMPARE_S_SIMD (state[0], state[1], state[2], state[3]); } }