44 …        const CeedScalar *restrict u, CeedScalar *restrict v, const CeedInt JJ, const CeedInt CC) {  in CeedTensorContract_Avx_Blocked()  argument
58 … for (CeedInt cc = 0; cc < CC / 4; cc++) vv[jj][cc] = loadu(&v[(a * J + j + jj) * C + c + cc * 4]);  in CeedTensorContract_Avx_Blocked()
69 … for (CeedInt cc = 0; cc < CC / 4; cc++) storeu(&v[(a * J + j + jj) * C + c + cc * 4], vv[jj][cc]);  in CeedTensorContract_Avx_Blocked()
81 … for (CeedInt cc = 0; cc < CC / 4; cc++) vv[jj][cc] = loadu(&v[(a * J + j + jj) * C + c + cc * 4]);  in CeedTensorContract_Avx_Blocked()
93 … for (CeedInt cc = 0; cc < CC / 4; cc++) storeu(&v[(a * J + j + jj) * C + c + cc * 4], vv[jj][cc]);  in CeedTensorContract_Avx_Blocked()
106 …        const CeedScalar *restrict u, CeedScalar *restrict v, const CeedInt JJ, const CeedInt CC) {  in CeedTensorContract_Avx_Remainder()  argument
123         for (CeedInt jj = 0; jj < JJ; jj++) vv[jj] = loadu(&v[(a * J + j + jj) * C + c]);  in CeedTensorContract_Avx_Remainder()
135         for (CeedInt jj = 0; jj < JJ; jj++) storeu(&v[(a * J + j + jj) * C + c], vv[jj]);  in CeedTensorContract_Avx_Remainder()
143 … for (CeedInt c = (C / CC) * CC; c < C; c++) v[(a * J + j) * C + c] += tq * u[(a * B + b) * C + c];  in CeedTensorContract_Avx_Remainder()
154 … CeedTransposeMode t_mode, const CeedInt add, const CeedScalar *restrict u, CeedScalar *restrict v,  in CeedTensorContract_Avx_Single()  argument
169         for (CeedInt jj = 0; jj < JJ / 4; jj++) vv[aa][jj] = loadu(&v[(a + aa) * J + j + jj * 4]);  in CeedTensorContract_Avx_Single()
182         for (CeedInt jj = 0; jj < JJ / 4; jj++) storeu(&v[(a + aa) * J + j + jj * 4], vv[aa][jj]);  in CeedTensorContract_Avx_Single()
193       for (CeedInt jj = 0; jj < JJ / 4; jj++) vv[aa][jj] = loadu(&v[(a + aa) * J + j + jj * 4]);  in CeedTensorContract_Avx_Single()
206       for (CeedInt jj = 0; jj < JJ / 4; jj++) storeu(&v[(a + aa) * J + j + jj * 4], vv[aa][jj]);  in CeedTensorContract_Avx_Single()
218       for (CeedInt aa = 0; aa < AA; aa++) vv[aa] = loadu(&v[(a + aa) * J + j]);  in CeedTensorContract_Avx_Single()
237       for (CeedInt aa = 0; aa < AA; aa++) storeu(&v[(a + aa) * J + j], vv[aa]);  in CeedTensorContract_Avx_Single()
245       for (CeedInt a = A_break; a < A; a++) v[a * J + j] += tq * u[a * B + b];  in CeedTensorContract_Avx_Single()
255 …eedTransposeMode t_mode, const CeedInt add, const CeedScalar *restrict u, CeedScalar *restrict v) {  in CeedTensorContract_Avx_Blocked_4_8()  argument
256   return CeedTensorContract_Avx_Blocked(contract, A, B, C, J, t, t_mode, add, u, v, 4, 8);  in CeedTensorContract_Avx_Blocked_4_8()
259 …eedTransposeMode t_mode, const CeedInt add, const CeedScalar *restrict u, CeedScalar *restrict v) {  in CeedTensorContract_Avx_Remainder_8_8()  argument
260   return CeedTensorContract_Avx_Remainder(contract, A, B, C, J, t, t_mode, add, u, v, 8, 8);  in CeedTensorContract_Avx_Remainder_8_8()
263 …eedTransposeMode t_mode, const CeedInt add, const CeedScalar *restrict u, CeedScalar *restrict v) {  in CeedTensorContract_Avx_Single_4_8()  argument
264   return CeedTensorContract_Avx_Single(contract, A, B, C, J, t, t_mode, add, u, v, 4, 8);  in CeedTensorContract_Avx_Single_4_8()
271 …eedTransposeMode t_mode, const CeedInt add, const CeedScalar *restrict u, CeedScalar *restrict v) {  in CeedTensorContractApply_Avx()  argument
275     for (CeedInt q = 0; q < A * J * C; q++) v[q] = (CeedScalar)0.0;  in CeedTensorContractApply_Avx()
280     CeedTensorContract_Avx_Single_4_8(contract, A, B, C, J, t, t_mode, true, u, v);  in CeedTensorContractApply_Avx()
283 …if (C >= blk_size) CeedTensorContract_Avx_Blocked_4_8(contract, A, B, C, J, t, t_mode, true, u, v);  in CeedTensorContractApply_Avx()
285 …f (C % blk_size) CeedTensorContract_Avx_Remainder_8_8(contract, A, B, C, J, t, t_mode, true, u, v);  in CeedTensorContractApply_Avx()