42 …TensorContract_Avx_Blocked(CeedTensorContract contract, CeedInt A, CeedInt B, CeedInt C, CeedInt J,  in CeedTensorContract_Avx_Blocked()  argument
55       for (CeedInt c = 0; c < (C / CC) * CC; c += CC) {  in CeedTensorContract_Avx_Blocked()
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()
64               fmadd(vv[jj][cc], tqv, loadu(&u[(a * B + b) * 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()
77       for (CeedInt c = 0; c < (C / CC) * CC; c += 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()
88               fmadd(vv[jj][cc], tqv, loadu(&u[(a * B + b) * 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()
104 …nsorContract_Avx_Remainder(CeedTensorContract contract, CeedInt A, CeedInt B, CeedInt C, CeedInt J,  in CeedTensorContract_Avx_Remainder()  argument
118     for (CeedInt c = (C / CC) * CC; c < C; c += 4) {  in CeedTensorContract_Avx_Remainder()
123         for (CeedInt jj = 0; jj < JJ; jj++) vv[jj] = loadu(&v[(a * J + j + jj) * C + c]);  in CeedTensorContract_Avx_Remainder()
127           if (C - c == 1) tqu = set(0.0, 0.0, 0.0, u[(a * B + b) * C + c + 0]);  in CeedTensorContract_Avx_Remainder()
128 …  else if (C - c == 2) tqu = set(0.0, 0.0, u[(a * B + b) * C + c + 1], u[(a * B + b) * C + c + 0]);  in CeedTensorContract_Avx_Remainder()
129 …else if (C - c == 3) tqu = set(0.0, u[(a * B + b) * C + c + 2], u[(a * B + b) * C + c + 1], u[(a *…  in CeedTensorContract_Avx_Remainder()
130           else tqu = loadu(&u[(a * B + b) * 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()
153 …tract_Avx_Single(CeedTensorContract contract, CeedInt A, CeedInt B, CeedInt C, CeedInt J, const Ce…  in CeedTensorContract_Avx_Single()  argument
254 …_Avx_Blocked_4_8(CeedTensorContract contract, CeedInt A, CeedInt B, CeedInt C, CeedInt J, const Ce…  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()
258 …vx_Remainder_8_8(CeedTensorContract contract, CeedInt A, CeedInt B, CeedInt C, CeedInt J, const Ce…  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()
262 …t_Avx_Single_4_8(CeedTensorContract contract, CeedInt A, CeedInt B, CeedInt C, CeedInt J, const Ce…  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()
270 static int CeedTensorContractApply_Avx(CeedTensorContract contract, CeedInt A, CeedInt B, CeedInt C…  in CeedTensorContractApply_Avx()  argument
275     for (CeedInt q = 0; q < A * J * C; q++) v[q] = (CeedScalar)0.0;  in CeedTensorContractApply_Avx()
278   if (C == 1) {  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 …if (C % blk_size) CeedTensorContract_Avx_Remainder_8_8(contract, A, B, C, J, t, t_mode, true, u, v…  in CeedTensorContractApply_Avx()