/home/docs/checkouts/readthedocs.org/user_builds/advanced-micro-devices-rocrand/checkouts/docs-7.1.0/projects/rocrand/library/include/rocrand/rocrand_lfsr113.h Source File

/home/docs/checkouts/readthedocs.org/user_builds/advanced-micro-devices-rocrand/checkouts/docs-7.1.0/projects/rocrand/library/include/rocrand/rocrand_lfsr113.h Source File#

API library: /home/docs/checkouts/readthedocs.org/user_builds/advanced-micro-devices-rocrand/checkouts/docs-7.1.0/projects/rocrand/library/include/rocrand/rocrand_lfsr113.h Source File
 // Copyright (c) 2022-2025 Advanced Micro Devices, Inc. All rights reserved.
 //
 // Permission is hereby granted, free of charge, to any person obtaining a copy
 // of this software and associated documentation files (the "Software"), to deal
 // in the Software without restriction, including without limitation the rights
 // to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 // copies of the Software, and to permit persons to whom the Software is
 // furnished to do so, subject to the following conditions:
 //
 // The above copyright notice and this permission notice shall be included in
 // all copies or substantial portions of the Software.
 //
 // THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 // IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 // FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL THE
 // AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 // LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 // OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
 // THE SOFTWARE.
  
 #ifndef ROCRAND_LFSR113_H_
 #define ROCRAND_LFSR113_H_
  
 #include "rocrand/rocrand_lfsr113_precomputed.h"
  
 #include <hip/hip_runtime.h>
  
 #define ROCRAND_LFSR113_DEFAULT_SEED_X 2
 #define ROCRAND_LFSR113_DEFAULT_SEED_Y 8
 #define ROCRAND_LFSR113_DEFAULT_SEED_Z 16
 #define ROCRAND_LFSR113_DEFAULT_SEED_W 128 // end of group rocranddevice
  
 namespace rocrand_device
 {
 namespace detail
 {
  
 __forceinline__ __device__ __host__ void mul_mat_vec_inplace(const unsigned int* m, uint4* z)
 {
     unsigned int v[4]         = {z->x, z->y, z->z, z->w};
     unsigned int r[LFSR113_N] = {0};
     for(int ij = 0; ij < LFSR113_N * LFSR113_M; ij++)
     {
         const int          i = ij / LFSR113_M;
         const int          j = ij % LFSR113_M;
         const unsigned int b = (v[i] & (1U << j)) ? 0xffffffff : 0x0;
         for(int k = 0; k < LFSR113_N; k++)
         {
             r[k] ^= b & m[i * LFSR113_M * LFSR113_N + j * LFSR113_N + k];
         }
     }
     // Copy result into z
     z->x = r[0];
     z->y = r[1];
     z->z = r[2];
     z->w = r[3];
 }
 } // end namespace detail
  
 class lfsr113_engine
 {
 public:
     struct lfsr113_state
     {
         uint4 z;
         uint4 subsequence;
     };
  
     __forceinline__ __device__ __host__ lfsr113_engine(const uint4 seed
                                                        = {ROCRAND_LFSR113_DEFAULT_SEED_X,
                                                           ROCRAND_LFSR113_DEFAULT_SEED_Y,
                                                           ROCRAND_LFSR113_DEFAULT_SEED_Z,
                                                           ROCRAND_LFSR113_DEFAULT_SEED_W},
                                                        const unsigned int       subsequence = 0,
                                                        const unsigned long long offset      = 0)
     {
         this->seed(seed, subsequence, offset);
     }
  
     __forceinline__ __device__ __host__ void seed(uint4                    seed_value,
                                                   const unsigned long long subsequence,
                                                   const unsigned long long offset = 0)
     {
         m_state.subsequence = seed_value;
  
         reset_start_subsequence();
         discard_subsequence(subsequence);
         discard(offset);
     }
  
     __forceinline__ __device__ __host__ void discard()
     {
         discard_state();
     }
  
     __forceinline__ __device__ __host__ void discard(unsigned long long offset)
     {
 #ifdef __HIP_DEVICE_COMPILE__
         jump(offset, d_lfsr113_jump_matrices);
 #else
         jump(offset, h_lfsr113_jump_matrices);
 #endif
     }
  
     __forceinline__ __device__ __host__ void discard_subsequence(unsigned int subsequence)
     {
 // Discard n * 2^55 samples
 #ifdef __HIP_DEVICE_COMPILE__
         jump(subsequence, d_lfsr113_sequence_jump_matrices);
 #else
         jump(subsequence, h_lfsr113_sequence_jump_matrices);
 #endif
     }
  
     __forceinline__ __device__ __host__ unsigned int operator()()
     {
         return next();
     }
  
     __forceinline__ __device__ __host__ unsigned int next()
     {
         unsigned int b;
  
         b           = (((m_state.z.x << 6) ^ m_state.z.x) >> 13);
         m_state.z.x = (((m_state.z.x & 4294967294U) << 18) ^ b);
  
         b           = (((m_state.z.y << 2) ^ m_state.z.y) >> 27);
         m_state.z.y = (((m_state.z.y & 4294967288U) << 2) ^ b);
  
         b           = (((m_state.z.z << 13) ^ m_state.z.z) >> 21);
         m_state.z.z = (((m_state.z.z & 4294967280U) << 7) ^ b);
  
         b           = (((m_state.z.w << 3) ^ m_state.z.w) >> 12);
         m_state.z.w = (((m_state.z.w & 4294967168U) << 13) ^ b);
  
         return (m_state.z.x ^ m_state.z.y ^ m_state.z.z ^ m_state.z.w);
     }
  
 protected:
     __forceinline__ __device__ __host__ void reset_start_subsequence()
     {
         m_state.z.x = m_state.subsequence.x;
         m_state.z.y = m_state.subsequence.y;
         m_state.z.z = m_state.subsequence.z;
         m_state.z.w = m_state.subsequence.w;
     }
  
     // Advances the internal state to the next state.
     __forceinline__ __device__ __host__ void discard_state()
     {
         this->next();
     }
  
     __forceinline__ __device__ __host__ void
         jump(unsigned long long v,
              const unsigned int (&jump_matrices)[LFSR113_JUMP_MATRICES][LFSR113_SIZE])
     {
         // x~(n + v) = (A^v mod m)x~n mod m
         // The matrix (A^v mod m) can be precomputed for selected values of v.
         //
         // For LFSR113_JUMP_LOG2 = 2
         // lfsr113_jump_matrices contains precomputed matrices:
         //   A^1, A^4, A^16...
         //
         // For LFSR113_JUMP_LOG2 = 2 and LFSR113_SEQUENCE_JUMP_LOG2 = 55
         // lfsr113_sequence_jump_matrices contains precomputed matrices:
         //   A^(1 * 2^55), A^(4 * 2^55), A^(16 * 2^55)...
         //
         // Intermediate powers can be calculated as multiplication of the powers above.
  
         unsigned int mi = 0;
         while(v > 0)
         {
             const unsigned int is = static_cast<unsigned int>(v) & ((1 << LFSR113_JUMP_LOG2) - 1);
             for(unsigned int i = 0; i < is; i++)
             {
                 detail::mul_mat_vec_inplace(jump_matrices[mi], &m_state.z);
             }
             mi++;
             v >>= LFSR113_JUMP_LOG2;
         }
     }
  
 protected:
     lfsr113_state m_state;
  
 }; // lfsr113_engine class
  
 } // end namespace rocrand_device
  
 typedef rocrand_device::lfsr113_engine rocrand_state_lfsr113;
  
 __forceinline__ __device__ __host__
 void rocrand_init(const uint4 seed, const unsigned int subsequence, rocrand_state_lfsr113* state)
 {
     *state = rocrand_state_lfsr113(seed, subsequence);
 }
  
 __forceinline__ __device__ __host__
 void rocrand_init(const uint4              seed,
                   const unsigned int       subsequence,
                   const unsigned long long offset,
                   rocrand_state_lfsr113*   state)
 {
     *state = rocrand_state_lfsr113(seed, subsequence, offset);
 }
  
 __forceinline__ __device__ __host__
 unsigned int rocrand(rocrand_state_lfsr113* state)
 {
     return state->next();
 }
  
 __forceinline__ __device__ __host__
 void skipahead(unsigned long long offset, rocrand_state_lfsr113* state)
 {
     return state->discard(offset);
 }
  
 __forceinline__ __device__ __host__
 void skipahead_subsequence(unsigned int subsequence, rocrand_state_lfsr113* state)
 {
     return state->discard_subsequence(subsequence);
 }
  
 __forceinline__ __device__ __host__
 void skipahead_sequence(unsigned int sequence, rocrand_state_lfsr113* state)
 {
     return state->discard_subsequence(sequence);
 }
  // end of group rocranddevice
  
 #endif // ROCRAND_LFSR113_H_