Added API and tests for new GemmStridedBatched routine

3 files changed, 96 insertions, 79 deletions

diff --git a/scripts/generator/generator.py b/scripts/generator/generator.py
index 5fbce2c4..528e61dd 100755
--- a/scripts/generator/generator.py
+++ b/scripts/generator/generator.py
@@ -109,71 +109,72 @@ col = "height * width * channels"
 im2col_constants = ["channels", "height", "width", "kernel_h", "kernel_w", "pad_h", "pad_w", "stride_h", "stride_w", "dilation_h", "dilation_w"]
 ROUTINES = [
 [  # Level 1: vector-vector
-  Routine(False, True,  False, False, "1", "rotg",  T, [S,D],            [],                  [],                                                     [],         ["sa","sb","sc","ss"],        ["1","1","1","1"], [],       "",    "Generate givens plane rotation", "", []),
-  Routine(False, True,  False, False, "1", "rotmg", T, [S,D],            [],                  [],                                                     ["sy1"],    ["sd1","sd2","sx1","sparam"], ["1","1","1","1","1"], [],   "",    "Generate modified givens plane rotation", "", []),
-  Routine(False, True,  False, False, "1", "rot",   T, [S,D],            ["n"],               [],                                                     [],         ["x","y"],                    [xn,yn],       ["cos","sin"],"",    "Apply givens plane rotation", "", []),
-  Routine(False, True,  False, False, "1", "rotm",  T, [S,D],            ["n"],               [],                                                     [],         ["x","y","sparam"],           [xn,yn,"1"],   [],           "",    "Apply modified givens plane rotation", "", []),
-  Routine(True,  True,  False, False, "1", "swap",  T, [S,D,C,Z,H],      ["n"],               [],                                                     [],         ["x","y"],                    [xn,yn],       [],           "",    "Swap two vectors", "Interchanges _n_ elements of vectors _x_ and _y_.", []),
-  Routine(True,  True,  False, False, "1", "scal",  T, [S,D,C,Z,H],      ["n"],               [],                                                     [],         ["x"],                        [xn],          ["alpha"],    "",    "Vector scaling", "Multiplies _n_ elements of vector _x_ by a scalar constant _alpha_.", []),
-  Routine(True,  True,  False, False, "1", "copy",  T, [S,D,C,Z,H],      ["n"],               [],                                                     ["x"],      ["y"],                        [xn,yn],       [],           "",    "Vector copy", "Copies the contents of vector _x_ into vector _y_.", []),
-  Routine(True,  True,  False, False, "1", "axpy",  T, [S,D,C,Z,H],      ["n"],               [],                                                     ["x"],      ["y"],                        [xn,yn],       ["alpha"],    "",    "Vector-times-constant plus vector", "Performs the operation _y = alpha * x + y_, in which _x_ and _y_ are vectors and _alpha_ is a scalar constant.", []),
-  Routine(True,  True,  False, False, "1", "dot",   T, [S,D,H],          ["n"],               [],                                                     ["x","y"],  ["dot"],                      [xn,yn,"1"],   [],           "n",   "Dot product of two vectors", "Multiplies _n_ elements of the vectors _x_ and _y_ element-wise and accumulates the results. The sum is stored in the _dot_ buffer.", []),
-  Routine(True,  True,  False, False, "1", "dotu",  T, [C,Z],            ["n"],               [],                                                     ["x","y"],  ["dot"],                      [xn,yn,"1"],   [],           "n",   "Dot product of two complex vectors", "See the regular xDOT routine.", []),
-  Routine(True,  True,  False, False, "1", "dotc",  T, [C,Z],            ["n"],               [],                                                     ["x","y"],  ["dot"],                      [xn,yn,"1"],   [],           "n",   "Dot product of two complex vectors, one conjugated", "See the regular xDOT routine.", []),
-  Routine(True,  True,  False, False, "1", "nrm2",  T, [S,D,Sc,Dz,H],    ["n"],               [],                                                     ["x"],      ["nrm2"],                     [xn,"1"],      [],           "2*n", "Euclidian norm of a vector", "Accumulates the square of _n_ elements in the _x_ vector and takes the square root. The resulting L2 norm is stored in the _nrm2_ buffer.", []),
-  Routine(True,  True,  False, False, "1", "asum",  T, [S,D,Sc,Dz,H],    ["n"],               [],                                                     ["x"],      ["asum"],                     [xn,"1"],      [],           "n",   "Absolute sum of values in a vector", "Accumulates the absolute value of _n_ elements in the _x_ vector. The results are stored in the _asum_ buffer.", []),
-  Routine(True,  False, False, False, "1", "sum",   T, [S,D,Sc,Dz,H],    ["n"],               [],                                                     ["x"],      ["sum"],                      [xn,"1"],      [],           "n",   "Sum of values in a vector (non-BLAS function)", "Accumulates the values of _n_ elements in the _x_ vector. The results are stored in the _sum_ buffer. This routine is the non-absolute version of the xASUM BLAS routine.", []),
-  Routine(True,  True,  False, False, "1", "amax",  T, [iS,iD,iC,iZ,iH], ["n"],               [],                                                     ["x"],      ["imax"],                     [xn,"1"],      [],           "2*n", "Index of absolute maximum value in a vector", "Finds the index of the maximum of the absolute values in the _x_ vector. The resulting integer index is stored in the _imax_ buffer.", []),
-  Routine(True,  False, False, False, "1", "amin",  T, [iS,iD,iC,iZ,iH], ["n"],               [],                                                     ["x"],      ["imin"],                     [xn,"1"],      [],           "2*n", "Index of absolute minimum value in a vector (non-BLAS function)", "Finds the index of the minimum of the absolute values in the _x_ vector. The resulting integer index is stored in the _imin_ buffer.", []),
-  Routine(True,  False, False, False, "1", "max",   T, [iS,iD,iC,iZ,iH], ["n"],               [],                                                     ["x"],      ["imax"],                     [xn,"1"],      [],           "2*n", "Index of maximum value in a vector (non-BLAS function)", "Finds the index of the maximum of the values in the _x_ vector. The resulting integer index is stored in the _imax_ buffer. This routine is the non-absolute version of the IxAMAX BLAS routine.", []),
-  Routine(True,  False, False, False, "1", "min",   T, [iS,iD,iC,iZ,iH], ["n"],               [],                                                     ["x"],      ["imin"],                     [xn,"1"],      [],           "2*n", "Index of minimum value in a vector (non-BLAS function)", "Finds the index of the minimum of the values in the _x_ vector. The resulting integer index is stored in the _imin_ buffer. This routine is the non-absolute minimum version of the IxAMAX BLAS routine.", []),
+  Routine(False, True,  0, False, "1", "rotg",  T, [S,D],            [],                  [],                                                     [],         ["sa","sb","sc","ss"],        ["1","1","1","1"], [],       "",    "Generate givens plane rotation", "", []),
+  Routine(False, True,  0, False, "1", "rotmg", T, [S,D],            [],                  [],                                                     ["sy1"],    ["sd1","sd2","sx1","sparam"], ["1","1","1","1","1"], [],   "",    "Generate modified givens plane rotation", "", []),
+  Routine(False, True,  0, False, "1", "rot",   T, [S,D],            ["n"],               [],                                                     [],         ["x","y"],                    [xn,yn],       ["cos","sin"],"",    "Apply givens plane rotation", "", []),
+  Routine(False, True,  0, False, "1", "rotm",  T, [S,D],            ["n"],               [],                                                     [],         ["x","y","sparam"],           [xn,yn,"1"],   [],           "",    "Apply modified givens plane rotation", "", []),
+  Routine(True,  True,  0, False, "1", "swap",  T, [S,D,C,Z,H],      ["n"],               [],                                                     [],         ["x","y"],                    [xn,yn],       [],           "",    "Swap two vectors", "Interchanges _n_ elements of vectors _x_ and _y_.", []),
+  Routine(True,  True,  0, False, "1", "scal",  T, [S,D,C,Z,H],      ["n"],               [],                                                     [],         ["x"],                        [xn],          ["alpha"],    "",    "Vector scaling", "Multiplies _n_ elements of vector _x_ by a scalar constant _alpha_.", []),
+  Routine(True,  True,  0, False, "1", "copy",  T, [S,D,C,Z,H],      ["n"],               [],                                                     ["x"],      ["y"],                        [xn,yn],       [],           "",    "Vector copy", "Copies the contents of vector _x_ into vector _y_.", []),
+  Routine(True,  True,  0, False, "1", "axpy",  T, [S,D,C,Z,H],      ["n"],               [],                                                     ["x"],      ["y"],                        [xn,yn],       ["alpha"],    "",    "Vector-times-constant plus vector", "Performs the operation _y = alpha * x + y_, in which _x_ and _y_ are vectors and _alpha_ is a scalar constant.", []),
+  Routine(True,  True,  0, False, "1", "dot",   T, [S,D,H],          ["n"],               [],                                                     ["x","y"],  ["dot"],                      [xn,yn,"1"],   [],           "n",   "Dot product of two vectors", "Multiplies _n_ elements of the vectors _x_ and _y_ element-wise and accumulates the results. The sum is stored in the _dot_ buffer.", []),
+  Routine(True,  True,  0, False, "1", "dotu",  T, [C,Z],            ["n"],               [],                                                     ["x","y"],  ["dot"],                      [xn,yn,"1"],   [],           "n",   "Dot product of two complex vectors", "See the regular xDOT routine.", []),
+  Routine(True,  True,  0, False, "1", "dotc",  T, [C,Z],            ["n"],               [],                                                     ["x","y"],  ["dot"],                      [xn,yn,"1"],   [],           "n",   "Dot product of two complex vectors, one conjugated", "See the regular xDOT routine.", []),
+  Routine(True,  True,  0, False, "1", "nrm2",  T, [S,D,Sc,Dz,H],    ["n"],               [],                                                     ["x"],      ["nrm2"],                     [xn,"1"],      [],           "2*n", "Euclidian norm of a vector", "Accumulates the square of _n_ elements in the _x_ vector and takes the square root. The resulting L2 norm is stored in the _nrm2_ buffer.", []),
+  Routine(True,  True,  0, False, "1", "asum",  T, [S,D,Sc,Dz,H],    ["n"],               [],                                                     ["x"],      ["asum"],                     [xn,"1"],      [],           "n",   "Absolute sum of values in a vector", "Accumulates the absolute value of _n_ elements in the _x_ vector. The results are stored in the _asum_ buffer.", []),
+  Routine(True,  False, 0, False, "1", "sum",   T, [S,D,Sc,Dz,H],    ["n"],               [],                                                     ["x"],      ["sum"],                      [xn,"1"],      [],           "n",   "Sum of values in a vector (non-BLAS function)", "Accumulates the values of _n_ elements in the _x_ vector. The results are stored in the _sum_ buffer. This routine is the non-absolute version of the xASUM BLAS routine.", []),
+  Routine(True,  True,  0, False, "1", "amax",  T, [iS,iD,iC,iZ,iH], ["n"],               [],                                                     ["x"],      ["imax"],                     [xn,"1"],      [],           "2*n", "Index of absolute maximum value in a vector", "Finds the index of the maximum of the absolute values in the _x_ vector. The resulting integer index is stored in the _imax_ buffer.", []),
+  Routine(True,  False, 0, False, "1", "amin",  T, [iS,iD,iC,iZ,iH], ["n"],               [],                                                     ["x"],      ["imin"],                     [xn,"1"],      [],           "2*n", "Index of absolute minimum value in a vector (non-BLAS function)", "Finds the index of the minimum of the absolute values in the _x_ vector. The resulting integer index is stored in the _imin_ buffer.", []),
+  Routine(True,  False, 0, False, "1", "max",   T, [iS,iD,iC,iZ,iH], ["n"],               [],                                                     ["x"],      ["imax"],                     [xn,"1"],      [],           "2*n", "Index of maximum value in a vector (non-BLAS function)", "Finds the index of the maximum of the values in the _x_ vector. The resulting integer index is stored in the _imax_ buffer. This routine is the non-absolute version of the IxAMAX BLAS routine.", []),
+  Routine(True,  False, 0, False, "1", "min",   T, [iS,iD,iC,iZ,iH], ["n"],               [],                                                     ["x"],      ["imin"],                     [xn,"1"],      [],           "2*n", "Index of minimum value in a vector (non-BLAS function)", "Finds the index of the minimum of the values in the _x_ vector. The resulting integer index is stored in the _imin_ buffer. This routine is the non-absolute minimum version of the IxAMAX BLAS routine.", []),
 ],
 [  # Level 2: matrix-vector
-  Routine(True,  True,  False, False, "2a", "gemv",  T,  [S,D,C,Z,H],    ["m","n"],           ["layout","a_transpose"],                               ["a","x"],  ["y"],                        [amn,xmn,ynm], ["alpha","beta"], "",    "General matrix-vector multiplication", "Performs the operation _y = alpha * A * x + beta * y_, in which _x_ is an input vector, _y_ is an input and output vector, _A_ is an input matrix, and _alpha_ and _beta_ are scalars. The matrix _A_ can optionally be transposed before performing the operation.", [ald_m]),
-  Routine(True,  True,  False, False, "2a", "gbmv",  T,  [S,D,C,Z,H],    ["m","n","kl","ku"], ["layout","a_transpose"],                               ["a","x"],  ["y"],                        [amn,xmn,ynm], ["alpha","beta"], "",    "General banded matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is banded instead.", [ald_kl_ku_one]),
-  Routine(True,  True,  False, False, "2a", "hemv",  T,  [C,Z],          ["n"],               ["layout","triangle"],                                  ["a","x"],  ["y"],                        [an,xn,yn],    ["alpha","beta"], "",    "Hermitian matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is an Hermitian matrix instead.", [ald_n]),
-  Routine(True,  True,  False, False, "2a", "hbmv",  T,  [C,Z],          ["n","k"],           ["layout","triangle"],                                  ["a","x"],  ["y"],                        [an,xn,yn],    ["alpha","beta"], "",    "Hermitian banded matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is an Hermitian banded matrix instead.", [ald_k_one]),
-  Routine(True,  True,  False, False, "2a", "hpmv",  T,  [C,Z],          ["n"],               ["layout","triangle"],                                  ["ap","x"], ["y"],                        [apn,xn,yn],   ["alpha","beta"], "",    "Hermitian packed matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is an Hermitian packed matrix instead and represented as _AP_.", []),
-  Routine(True,  True,  False, False, "2a", "symv",  T,  [S,D,H],        ["n"],               ["layout","triangle"],                                  ["a","x"],  ["y"],                        [an,xn,yn],    ["alpha","beta"], "",    "Symmetric matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is symmetric instead.", [ald_n]),
-  Routine(True,  True,  False, False, "2a", "sbmv",  T,  [S,D,H],        ["n","k"],           ["layout","triangle"],                                  ["a","x"],  ["y"],                        [an,xn,yn],    ["alpha","beta"], "",    "Symmetric banded matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is symmetric and banded instead.", [ald_k_one]),
-  Routine(True,  True,  False, False, "2a", "spmv",  T,  [S,D,H],        ["n"],               ["layout","triangle"],                                  ["ap","x"], ["y"],                        [apn,xn,yn],   ["alpha","beta"], "",    "Symmetric packed matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is a symmetric packed matrix instead and represented as _AP_.", []),
-  Routine(True,  True,  False, False, "2a", "trmv",  T,  [S,D,C,Z,H],    ["n"],               ["layout","triangle","a_transpose","diagonal"],         ["a"],      ["x"],                        [an,xn],       [],               "n",   "Triangular matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is triangular instead.", [ald_n]),
-  Routine(True,  True,  False, False, "2a", "tbmv",  T,  [S,D,C,Z,H],    ["n","k"],           ["layout","triangle","a_transpose","diagonal"],         ["a"],      ["x"],                        [an,xn],       [],               "n",   "Triangular banded matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is triangular and banded instead.", [ald_k_one]),
-  Routine(True,  True,  False, False, "2a", "tpmv",  T,  [S,D,C,Z,H],    ["n"],               ["layout","triangle","a_transpose","diagonal"],         ["ap"],     ["x"],                        [apn,xn],      [],               "n",   "Triangular packed matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is a triangular packed matrix instead and repreented as _AP_.", []),
-  Routine(True,  True,  False, False, "2a", "trsv",  T,  [S,D,C,Z],      ["n"],               ["layout","triangle","a_transpose","diagonal"],         ["a"],      ["x"],                        [an,xn],       [],               "",    "Solves a triangular system of equations", "", []),
-  Routine(False, True,  False, False, "2a", "tbsv",  T,  [S,D,C,Z],      ["n","k"],           ["layout","triangle","a_transpose","diagonal"],         ["a"],      ["x"],                        [an,xn],       [],               "",    "Solves a banded triangular system of equations", "", [ald_k_one]),
-  Routine(False, True,  False, False, "2a", "tpsv",  T,  [S,D,C,Z],      ["n"],               ["layout","triangle","a_transpose","diagonal"],         ["ap"],     ["x"],                        [apn,xn],      [],               "",    "Solves a packed triangular system of equations", "", []),
+  Routine(True,  True,  0, False, "2a", "gemv",  T,  [S,D,C,Z,H],    ["m","n"],           ["layout","a_transpose"],                               ["a","x"],  ["y"],                        [amn,xmn,ynm], ["alpha","beta"], "",    "General matrix-vector multiplication", "Performs the operation _y = alpha * A * x + beta * y_, in which _x_ is an input vector, _y_ is an input and output vector, _A_ is an input matrix, and _alpha_ and _beta_ are scalars. The matrix _A_ can optionally be transposed before performing the operation.", [ald_m]),
+  Routine(True,  True,  0, False, "2a", "gbmv",  T,  [S,D,C,Z,H],    ["m","n","kl","ku"], ["layout","a_transpose"],                               ["a","x"],  ["y"],                        [amn,xmn,ynm], ["alpha","beta"], "",    "General banded matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is banded instead.", [ald_kl_ku_one]),
+  Routine(True,  True,  0, False, "2a", "hemv",  T,  [C,Z],          ["n"],               ["layout","triangle"],                                  ["a","x"],  ["y"],                        [an,xn,yn],    ["alpha","beta"], "",    "Hermitian matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is an Hermitian matrix instead.", [ald_n]),
+  Routine(True,  True,  0, False, "2a", "hbmv",  T,  [C,Z],          ["n","k"],           ["layout","triangle"],                                  ["a","x"],  ["y"],                        [an,xn,yn],    ["alpha","beta"], "",    "Hermitian banded matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is an Hermitian banded matrix instead.", [ald_k_one]),
+  Routine(True,  True,  0, False, "2a", "hpmv",  T,  [C,Z],          ["n"],               ["layout","triangle"],                                  ["ap","x"], ["y"],                        [apn,xn,yn],   ["alpha","beta"], "",    "Hermitian packed matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is an Hermitian packed matrix instead and represented as _AP_.", []),
+  Routine(True,  True,  0, False, "2a", "symv",  T,  [S,D,H],        ["n"],               ["layout","triangle"],                                  ["a","x"],  ["y"],                        [an,xn,yn],    ["alpha","beta"], "",    "Symmetric matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is symmetric instead.", [ald_n]),
+  Routine(True,  True,  0, False, "2a", "sbmv",  T,  [S,D,H],        ["n","k"],           ["layout","triangle"],                                  ["a","x"],  ["y"],                        [an,xn,yn],    ["alpha","beta"], "",    "Symmetric banded matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is symmetric and banded instead.", [ald_k_one]),
+  Routine(True,  True,  0, False, "2a", "spmv",  T,  [S,D,H],        ["n"],               ["layout","triangle"],                                  ["ap","x"], ["y"],                        [apn,xn,yn],   ["alpha","beta"], "",    "Symmetric packed matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is a symmetric packed matrix instead and represented as _AP_.", []),
+  Routine(True,  True,  0, False, "2a", "trmv",  T,  [S,D,C,Z,H],    ["n"],               ["layout","triangle","a_transpose","diagonal"],         ["a"],      ["x"],                        [an,xn],       [],               "n",   "Triangular matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is triangular instead.", [ald_n]),
+  Routine(True,  True,  0, False, "2a", "tbmv",  T,  [S,D,C,Z,H],    ["n","k"],           ["layout","triangle","a_transpose","diagonal"],         ["a"],      ["x"],                        [an,xn],       [],               "n",   "Triangular banded matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is triangular and banded instead.", [ald_k_one]),
+  Routine(True,  True,  0, False, "2a", "tpmv",  T,  [S,D,C,Z,H],    ["n"],               ["layout","triangle","a_transpose","diagonal"],         ["ap"],     ["x"],                        [apn,xn],      [],               "n",   "Triangular packed matrix-vector multiplication", "Same operation as xGEMV, but matrix _A_ is a triangular packed matrix instead and repreented as _AP_.", []),
+  Routine(True,  True,  0, False, "2a", "trsv",  T,  [S,D,C,Z],      ["n"],               ["layout","triangle","a_transpose","diagonal"],         ["a"],      ["x"],                        [an,xn],       [],               "",    "Solves a triangular system of equations", "", []),
+  Routine(False, True,  0, False, "2a", "tbsv",  T,  [S,D,C,Z],      ["n","k"],           ["layout","triangle","a_transpose","diagonal"],         ["a"],      ["x"],                        [an,xn],       [],               "",    "Solves a banded triangular system of equations", "", [ald_k_one]),
+  Routine(False, True,  0, False, "2a", "tpsv",  T,  [S,D,C,Z],      ["n"],               ["layout","triangle","a_transpose","diagonal"],         ["ap"],     ["x"],                        [apn,xn],      [],               "",    "Solves a packed triangular system of equations", "", []),
   # Level 2: matrix update
-  Routine(True,  True,  False, False, "2b", "ger",   T,  [S,D,H],        ["m","n"],           ["layout"],                                             ["x","y"],  ["a"],                        [xm,yn,amn],   ["alpha"],        "",    "General rank-1 matrix update", "Performs the operation _A = alpha * x * y^T + A_, in which _x_ is an input vector, _y^T_ is the transpose of the input vector _y_, _A_ is the matrix to be updated, and _alpha_ is a scalar value.", [ald_m]),
-  Routine(True,  True,  False, False, "2b", "geru",  T,  [C,Z],          ["m","n"],           ["layout"],                                             ["x","y"],  ["a"],                        [xm,yn,amn],   ["alpha"],        "",    "General rank-1 complex matrix update", "Same operation as xGER, but with complex data-types.", [ald_m]),
-  Routine(True,  True,  False, False, "2b", "gerc",  T,  [C,Z],          ["m","n"],           ["layout"],                                             ["x","y"],  ["a"],                        [xm,yn,amn],   ["alpha"],        "",    "General rank-1 complex conjugated matrix update", "Same operation as xGERU, but the update is done based on the complex conjugate of the input vectors.", [ald_m]),
-  Routine(True,  True,  False, False, "2b", "her",   Tc, [Css,Zdd],      ["n"],               ["layout","triangle"],                                  ["x"],      ["a"],                        [xn,an],       ["alpha"],        "",    "Hermitian rank-1 matrix update", "Performs the operation _A = alpha * x * x^T + A_, in which x is an input vector, x^T is the transpose of this vector, _A_ is the triangular Hermetian matrix to be updated, and alpha is a scalar value.", [ald_n]),
-  Routine(True,  True,  False, False, "2b", "hpr",   Tc, [Css,Zdd],      ["n"],               ["layout","triangle"],                                  ["x"],      ["ap"],                       [xn,apn],      ["alpha"],        "",    "Hermitian packed rank-1 matrix update", "Same operation as xHER, but matrix _A_ is an Hermitian packed matrix instead and represented as _AP_.", []),
-  Routine(True,  True,  False, False, "2b", "her2",  T,  [C,Z],          ["n"],               ["layout","triangle"],                                  ["x","y"],  ["a"],                        [xn,yn,an],    ["alpha"],        "",    "Hermitian rank-2 matrix update", "Performs the operation _A = alpha * x * y^T + conj(alpha) * y * x^T + A_, in which _x_ is an input vector and _x^T_ its transpose, _y_ is an input vector and _y^T_ its transpose, _A_ is the triangular Hermetian matrix to be updated, _alpha_ is a scalar value and _conj(alpha)_ its complex conjugate.", [ald_n]),
-  Routine(True,  True,  False, False, "2b", "hpr2",  T,  [C,Z],          ["n"],               ["layout","triangle"],                                  ["x","y"],  ["ap"],                       [xn,yn,apn],   ["alpha"],        "",    "Hermitian packed rank-2 matrix update", "Same operation as xHER2, but matrix _A_ is an Hermitian packed matrix instead and represented as _AP_.", []),
-  Routine(True,  True,  False, False, "2b", "syr",   T,  [S,D,H],        ["n"],               ["layout","triangle"],                                  ["x"],      ["a"],                        [xn,an],       ["alpha"],        "",    "Symmetric rank-1 matrix update", "Same operation as xHER, but matrix A is a symmetric matrix instead.", [ald_n]),
-  Routine(True,  True,  False, False, "2b", "spr",   T,  [S,D,H],        ["n"],               ["layout","triangle"],                                  ["x"],      ["ap"],                       [xn,apn],      ["alpha"],        "",    "Symmetric packed rank-1 matrix update", "Same operation as xSPR, but matrix _A_ is a symmetric packed matrix instead and represented as _AP_.", []),
-  Routine(True,  True,  False, False, "2b", "syr2",  T,  [S,D,H],        ["n"],               ["layout","triangle"],                                  ["x","y"],  ["a"],                        [xn,yn,an],    ["alpha"],        "",    "Symmetric rank-2 matrix update", "Same operation as xHER2, but matrix _A_ is a symmetric matrix instead.", [ald_n]),
-  Routine(True,  True,  False, False, "2b", "spr2",  T,  [S,D,H],        ["n"],               ["layout","triangle"],                                  ["x","y"],  ["ap"],                       [xn,yn,apn],   ["alpha"],        "",    "Symmetric packed rank-2 matrix update", "Same operation as xSPR2, but matrix _A_ is a symmetric packed matrix instead and represented as _AP_.", []),
+  Routine(True,  True,  0, False, "2b", "ger",   T,  [S,D,H],        ["m","n"],           ["layout"],                                             ["x","y"],  ["a"],                        [xm,yn,amn],   ["alpha"],        "",    "General rank-1 matrix update", "Performs the operation _A = alpha * x * y^T + A_, in which _x_ is an input vector, _y^T_ is the transpose of the input vector _y_, _A_ is the matrix to be updated, and _alpha_ is a scalar value.", [ald_m]),
+  Routine(True,  True,  0, False, "2b", "geru",  T,  [C,Z],          ["m","n"],           ["layout"],                                             ["x","y"],  ["a"],                        [xm,yn,amn],   ["alpha"],        "",    "General rank-1 complex matrix update", "Same operation as xGER, but with complex data-types.", [ald_m]),
+  Routine(True,  True,  0, False, "2b", "gerc",  T,  [C,Z],          ["m","n"],           ["layout"],                                             ["x","y"],  ["a"],                        [xm,yn,amn],   ["alpha"],        "",    "General rank-1 complex conjugated matrix update", "Same operation as xGERU, but the update is done based on the complex conjugate of the input vectors.", [ald_m]),
+  Routine(True,  True,  0, False, "2b", "her",   Tc, [Css,Zdd],      ["n"],               ["layout","triangle"],                                  ["x"],      ["a"],                        [xn,an],       ["alpha"],        "",    "Hermitian rank-1 matrix update", "Performs the operation _A = alpha * x * x^T + A_, in which x is an input vector, x^T is the transpose of this vector, _A_ is the triangular Hermetian matrix to be updated, and alpha is a scalar value.", [ald_n]),
+  Routine(True,  True,  0, False, "2b", "hpr",   Tc, [Css,Zdd],      ["n"],               ["layout","triangle"],                                  ["x"],      ["ap"],                       [xn,apn],      ["alpha"],        "",    "Hermitian packed rank-1 matrix update", "Same operation as xHER, but matrix _A_ is an Hermitian packed matrix instead and represented as _AP_.", []),
+  Routine(True,  True,  0, False, "2b", "her2",  T,  [C,Z],          ["n"],               ["layout","triangle"],                                  ["x","y"],  ["a"],                        [xn,yn,an],    ["alpha"],        "",    "Hermitian rank-2 matrix update", "Performs the operation _A = alpha * x * y^T + conj(alpha) * y * x^T + A_, in which _x_ is an input vector and _x^T_ its transpose, _y_ is an input vector and _y^T_ its transpose, _A_ is the triangular Hermetian matrix to be updated, _alpha_ is a scalar value and _conj(alpha)_ its complex conjugate.", [ald_n]),
+  Routine(True,  True,  0, False, "2b", "hpr2",  T,  [C,Z],          ["n"],               ["layout","triangle"],                                  ["x","y"],  ["ap"],                       [xn,yn,apn],   ["alpha"],        "",    "Hermitian packed rank-2 matrix update", "Same operation as xHER2, but matrix _A_ is an Hermitian packed matrix instead and represented as _AP_.", []),
+  Routine(True,  True,  0, False, "2b", "syr",   T,  [S,D,H],        ["n"],               ["layout","triangle"],                                  ["x"],      ["a"],                        [xn,an],       ["alpha"],        "",    "Symmetric rank-1 matrix update", "Same operation as xHER, but matrix A is a symmetric matrix instead.", [ald_n]),
+  Routine(True,  True,  0, False, "2b", "spr",   T,  [S,D,H],        ["n"],               ["layout","triangle"],                                  ["x"],      ["ap"],                       [xn,apn],      ["alpha"],        "",    "Symmetric packed rank-1 matrix update", "Same operation as xSPR, but matrix _A_ is a symmetric packed matrix instead and represented as _AP_.", []),
+  Routine(True,  True,  0, False, "2b", "syr2",  T,  [S,D,H],        ["n"],               ["layout","triangle"],                                  ["x","y"],  ["a"],                        [xn,yn,an],    ["alpha"],        "",    "Symmetric rank-2 matrix update", "Same operation as xHER2, but matrix _A_ is a symmetric matrix instead.", [ald_n]),
+  Routine(True,  True,  0, False, "2b", "spr2",  T,  [S,D,H],        ["n"],               ["layout","triangle"],                                  ["x","y"],  ["ap"],                       [xn,yn,apn],   ["alpha"],        "",    "Symmetric packed rank-2 matrix update", "Same operation as xSPR2, but matrix _A_ is a symmetric packed matrix instead and represented as _AP_.", []),
 ],
 [  # Level 3: matrix-matrix
-  Routine(True,  True,  False, True,  "3", "gemm",  T,  [S,D,C,Z,H],     ["m","n","k"],        ["layout","a_transpose","b_transpose"],                ["a","b"],  ["c"],                        [amk,bkn,cmn],   ["alpha","beta"], "",    "General matrix-matrix multiplication", "Performs the matrix product _C = alpha * A * B + beta * C_, in which _A_ (_m_ by _k_) and _B_ (_k_ by _n_) are two general rectangular input matrices, _C_ (_m_ by _n_) is the matrix to be updated, and _alpha_ and _beta_ are scalar values. The matrices _A_ and/or _B_ can optionally be transposed before performing the operation.", [ald_transa_m_k, bld_transb_k_n, cld_m]),
-  Routine(True,  True,  False, False, "3", "symm",  T,  [S,D,C,Z,H],     ["m","n"],            ["layout","side","triangle"],                          ["a","b"],  ["c"],                        [ammn,bmnn,cmn], ["alpha","beta"], "",    "Symmetric matrix-matrix multiplication", "Same operation as xGEMM, but _A_ is symmetric instead. In case of `side == kLeft`, _A_ is a symmetric _m_ by _m_ matrix and _C = alpha * A * B + beta * C_ is performed. Otherwise, in case of `side == kRight`, _A_ is a symmtric _n_ by _n_ matrix and _C = alpha * B * A + beta * C_ is performed.", [ald_side_m_n, bld_m, cld_m]),
-  Routine(True,  True,  False, False, "3", "hemm",  T,  [C,Z],           ["m","n"],            ["layout","side","triangle"],                          ["a","b"],  ["c"],                        [ammn,bmnn,cmn], ["alpha","beta"], "",    "Hermitian matrix-matrix multiplication", "Same operation as xSYMM, but _A_ is an Hermitian matrix instead.", [ald_side_m_n, bld_m, cld_m]),
-  Routine(True,  True,  False, False, "3", "syrk",  T,  [S,D,C,Z,H],     ["n","k"],            ["layout","triangle","a_transpose"],                   ["a"],      ["c"],                        [ank,cn],        ["alpha","beta"], "",    "Rank-K update of a symmetric matrix", "Performs the matrix product _C = alpha * A * A^T + beta * C_ or _C = alpha * A^T * A + beta * C_, in which _A_ is a general matrix and _A^T_ is its transpose, _C_ (_n_ by _n_) is the symmetric matrix to be updated, and _alpha_ and _beta_ are scalar values.", [ald_trans_n_k, cld_m]),
-  Routine(True,  True,  False, False, "3", "herk",  Tc, [Css,Zdd],       ["n","k"],            ["layout","triangle","a_transpose"],                   ["a"],      ["c"],                        [ank,cn],        ["alpha","beta"], "",    "Rank-K update of a hermitian matrix", "Same operation as xSYRK, but _C_ is an Hermitian matrix instead.", [ald_trans_n_k, cld_m]),
-  Routine(True,  True,  False, False, "3", "syr2k", T,  [S,D,C,Z,H],     ["n","k"],            ["layout","triangle","ab_transpose"],                  ["a","b"],  ["c"],                        [ankab,bnkab,cn],["alpha","beta"], "",    "Rank-2K update of a symmetric matrix", "Performs the matrix product _C = alpha * A * B^T + alpha * B * A^T + beta * C_ or _C = alpha * A^T * B + alpha * B^T * A + beta * C_, in which _A_ and _B_ are general matrices and _A^T_ and _B^T_ are their transposed versions, _C_ (_n_ by _n_) is the symmetric matrix to be updated, and _alpha_ and _beta_ are scalar values.", [ald_trans_n_k, bld_trans_n_k, cld_n]),
-  Routine(True,  True,  False, False, "3", "her2k", TU, [Ccs,Zzd],       ["n","k"],            ["layout","triangle","ab_transpose"],                  ["a","b"],  ["c"],                        [ankab,bnkab,cn],["alpha","beta"], "",    "Rank-2K update of a hermitian matrix", "Same operation as xSYR2K, but _C_ is an Hermitian matrix instead.", [ald_trans_n_k, bld_trans_n_k, cld_n]),
-  Routine(True,  True,  False, False, "3", "trmm",  T,  [S,D,C,Z,H],     ["m","n"],            ["layout","side","triangle","a_transpose","diagonal"], ["a"],      ["b"],                        [amns,bmn],      ["alpha"],        "",    "Triangular matrix-matrix multiplication", "Performs the matrix product _B = alpha * A * B_ or _B = alpha * B * A_, in which _A_ is a unit or non-unit triangular matrix, _B_ (_m_ by _n_) is the general matrix to be updated, and _alpha_ is a scalar value.", [ald_side_m_n, bld_m]),
-  Routine(True,  True,  False, False, "3", "trsm",  T,  [S,D,C,Z],       ["m","n"],            ["layout","side","triangle","a_transpose","diagonal"], ["a"],      ["b"],                        [amns,bmn],      ["alpha"],        "",    "Solves a triangular system of equations", "Solves the equation _A * X = alpha * B_ for the unknown _m_ by _n_ matrix X, in which _A_ is an _n_ by _n_ unit or non-unit triangular matrix and B is an _m_ by _n_ matrix. The matrix _B_ is overwritten by the solution _X_.", []),
+  Routine(True,  True,  0, True,  "3", "gemm",  T,  [S,D,C,Z,H],     ["m","n","k"],        ["layout","a_transpose","b_transpose"],                ["a","b"],  ["c"],                        [amk,bkn,cmn],   ["alpha","beta"], "",    "General matrix-matrix multiplication", "Performs the matrix product _C = alpha * A * B + beta * C_, in which _A_ (_m_ by _k_) and _B_ (_k_ by _n_) are two general rectangular input matrices, _C_ (_m_ by _n_) is the matrix to be updated, and _alpha_ and _beta_ are scalar values. The matrices _A_ and/or _B_ can optionally be transposed before performing the operation.", [ald_transa_m_k, bld_transb_k_n, cld_m]),
+  Routine(True,  True,  0, False, "3", "symm",  T,  [S,D,C,Z,H],     ["m","n"],            ["layout","side","triangle"],                          ["a","b"],  ["c"],                        [ammn,bmnn,cmn], ["alpha","beta"], "",    "Symmetric matrix-matrix multiplication", "Same operation as xGEMM, but _A_ is symmetric instead. In case of `side == kLeft`, _A_ is a symmetric _m_ by _m_ matrix and _C = alpha * A * B + beta * C_ is performed. Otherwise, in case of `side == kRight`, _A_ is a symmtric _n_ by _n_ matrix and _C = alpha * B * A + beta * C_ is performed.", [ald_side_m_n, bld_m, cld_m]),
+  Routine(True,  True,  0, False, "3", "hemm",  T,  [C,Z],           ["m","n"],            ["layout","side","triangle"],                          ["a","b"],  ["c"],                        [ammn,bmnn,cmn], ["alpha","beta"], "",    "Hermitian matrix-matrix multiplication", "Same operation as xSYMM, but _A_ is an Hermitian matrix instead.", [ald_side_m_n, bld_m, cld_m]),
+  Routine(True,  True,  0, False, "3", "syrk",  T,  [S,D,C,Z,H],     ["n","k"],            ["layout","triangle","a_transpose"],                   ["a"],      ["c"],                        [ank,cn],        ["alpha","beta"], "",    "Rank-K update of a symmetric matrix", "Performs the matrix product _C = alpha * A * A^T + beta * C_ or _C = alpha * A^T * A + beta * C_, in which _A_ is a general matrix and _A^T_ is its transpose, _C_ (_n_ by _n_) is the symmetric matrix to be updated, and _alpha_ and _beta_ are scalar values.", [ald_trans_n_k, cld_m]),
+  Routine(True,  True,  0, False, "3", "herk",  Tc, [Css,Zdd],       ["n","k"],            ["layout","triangle","a_transpose"],                   ["a"],      ["c"],                        [ank,cn],        ["alpha","beta"], "",    "Rank-K update of a hermitian matrix", "Same operation as xSYRK, but _C_ is an Hermitian matrix instead.", [ald_trans_n_k, cld_m]),
+  Routine(True,  True,  0, False, "3", "syr2k", T,  [S,D,C,Z,H],     ["n","k"],            ["layout","triangle","ab_transpose"],                  ["a","b"],  ["c"],                        [ankab,bnkab,cn],["alpha","beta"], "",    "Rank-2K update of a symmetric matrix", "Performs the matrix product _C = alpha * A * B^T + alpha * B * A^T + beta * C_ or _C = alpha * A^T * B + alpha * B^T * A + beta * C_, in which _A_ and _B_ are general matrices and _A^T_ and _B^T_ are their transposed versions, _C_ (_n_ by _n_) is the symmetric matrix to be updated, and _alpha_ and _beta_ are scalar values.", [ald_trans_n_k, bld_trans_n_k, cld_n]),
+  Routine(True,  True,  0, False, "3", "her2k", TU, [Ccs,Zzd],       ["n","k"],            ["layout","triangle","ab_transpose"],                  ["a","b"],  ["c"],                        [ankab,bnkab,cn],["alpha","beta"], "",    "Rank-2K update of a hermitian matrix", "Same operation as xSYR2K, but _C_ is an Hermitian matrix instead.", [ald_trans_n_k, bld_trans_n_k, cld_n]),
+  Routine(True,  True,  0, False, "3", "trmm",  T,  [S,D,C,Z,H],     ["m","n"],            ["layout","side","triangle","a_transpose","diagonal"], ["a"],      ["b"],                        [amns,bmn],      ["alpha"],        "",    "Triangular matrix-matrix multiplication", "Performs the matrix product _B = alpha * A * B_ or _B = alpha * B * A_, in which _A_ is a unit or non-unit triangular matrix, _B_ (_m_ by _n_) is the general matrix to be updated, and _alpha_ is a scalar value.", [ald_side_m_n, bld_m]),
+  Routine(True,  True,  0, False, "3", "trsm",  T,  [S,D,C,Z],       ["m","n"],            ["layout","side","triangle","a_transpose","diagonal"], ["a"],      ["b"],                        [amns,bmn],      ["alpha"],        "",    "Solves a triangular system of equations", "Solves the equation _A * X = alpha * B_ for the unknown _m_ by _n_ matrix X, in which _A_ is an _n_ by _n_ unit or non-unit triangular matrix and B is an _m_ by _n_ matrix. The matrix _B_ is overwritten by the solution _X_.", []),
 ],
 [  # Level X: extra routines (not part of BLAS)
   # Special routines:
-  Routine(True,  True,  False, False, "x", "omatcopy", T, [S,D,C,Z,H],   ["m","n"],            ["layout","a_transpose"],                              ["a"],      ["b"],                        [amn,bnma],      ["alpha"],        "",    "Scaling and out-place transpose/copy (non-BLAS function)", "Performs scaling and out-of-place transposition/copying of matrices according to _B = alpha*op(A)_, in which _A_ is an input matrix (_m_ rows by _n_ columns), _B_ an output matrix, and _alpha_ a scalar value. The operation _op_ can be a normal matrix copy, a transposition or a conjugate transposition.", [ald_m, bld_n]),
-  Routine(True,  True,  False, False, "x", "im2col",   T, [S,D,C,Z,H],   im2col_constants,     [],                                                    ["im"],     ["col"],                      [im,col],        [""],             "",    "Im2col function (non-BLAS function)", "Performs the im2col algorithm, in which _im_ is the input matrix and _col_ is the output matrix.", []),
+  Routine(True,  True,  0, False, "x", "omatcopy", T, [S,D,C,Z,H],   ["m","n"],            ["layout","a_transpose"],                              ["a"],      ["b"],                        [amn,bnma],      ["alpha"],        "",    "Scaling and out-place transpose/copy (non-BLAS function)", "Performs scaling and out-of-place transposition/copying of matrices according to _B = alpha*op(A)_, in which _A_ is an input matrix (_m_ rows by _n_ columns), _B_ an output matrix, and _alpha_ a scalar value. The operation _op_ can be a normal matrix copy, a transposition or a conjugate transposition.", [ald_m, bld_n]),
+  Routine(True,  True,  0, False, "x", "im2col",   T, [S,D,C,Z,H],   im2col_constants,     [],                                                    ["im"],     ["col"],                      [im,col],        [""],             "",    "Im2col function (non-BLAS function)", "Performs the im2col algorithm, in which _im_ is the input matrix and _col_ is the output matrix.", []),
   # Batched routines:
-  Routine(True,  True,  True,  False, "x", "axpy",     T, [S,D,C,Z,H],   ["n"],                [],                                                    ["x"],      ["y"],                        [xn,yn],         ["alpha"],        "",    "Batched version of AXPY", "As AXPY, but multiple operations are batched together for better performance.", []),
-  Routine(True,  True,  True,  False, "x", "gemm",     T, [S,D,C,Z,H],   ["m","n","k"],        ["layout","a_transpose","b_transpose"],                ["a","b"],  ["c"],                        [amk,bkn,cmn],   ["alpha","beta"], "",    "Batched version of GEMM", "As GEMM, but multiple operations are batched together for better performance.", [ald_transa_m_k, bld_transb_k_n, cld_m]),
+  Routine(True,  True,  1, False, "x", "axpy",     T, [S,D,C,Z,H],   ["n"],                [],                                                    ["x"],      ["y"],                        [xn,yn],         ["alpha"],        "",    "Batched version of AXPY", "As AXPY, but multiple operations are batched together for better performance.", []),
+  Routine(True,  True,  1, False, "x", "gemm",     T, [S,D,C,Z,H],   ["m","n","k"],        ["layout","a_transpose","b_transpose"],                ["a","b"],  ["c"],                        [amk,bkn,cmn],   ["alpha","beta"], "",    "Batched version of GEMM", "As GEMM, but multiple operations are batched together for better performance.", [ald_transa_m_k, bld_transb_k_n, cld_m]),
+  Routine(True,  True,  2, False, "x", "gemm",     T, [S,D,C,Z,H],   ["m","n","k"],        ["layout","a_transpose","b_transpose"],                ["a","b"],  ["c"],                        [amk,bkn,cmn],   ["alpha","beta"], "",    "StridedBatched version of GEMM", "As GEMM, but multiple strided operations are batched together for better performance.", [ald_transa_m_k, bld_transb_k_n, cld_m]),
 ]]
 
 
@@ -223,10 +224,10 @@ def main(argv):
                     if i == 6:
                         body += cpp.wrapper_cublas(routine)
                     if i == 7:
-                        if not routine.batched:
+                        if routine.batched == 0:
                             body += cpp.clblast_netlib_c_h(routine)
                     if i == 8:
-                        if not routine.batched:
+                        if routine.batched == 0:
                             body += cpp.clblast_netlib_c_cc(routine)
                     if i == 9:
                         body += cpp.clblast_h(routine, cuda=True)
diff --git a/scripts/generator/generator/cpp.py b/scripts/generator/generator/cpp.py
index 3631e737..51ca047c 100644
--- a/scripts/generator/generator/cpp.py
+++ b/scripts/generator/generator/cpp.py
@@ -58,7 +58,7 @@ def clblast_cc(routine, cuda=False):
             result += "    auto queue_cpp = Queue(*queue);" + NL
         event = "nullptr" if cuda else "event"
         result += "    auto routine = X" + routine.plain_name() + "<" + routine.template.template + ">(queue_cpp, " + event + ");" + NL
-        if routine.batched:
+        if routine.batched == 1:
             result += "    " + (NL + "    ").join(routine.batched_transform_to_cpp()) + NL
         if routine.temp_buffer:
             null = "0" if cuda else "nullptr"
@@ -110,7 +110,7 @@ def clblast_c_cc(routine):
         template = "<" + flavour.template + ">" if routine.no_scalars() else ""
         indent = " " * (16 + routine.length() + len(template))
         result += routine.routine_header_c(flavour, 27, "") + " {" + NL
-        if routine.batched:
+        if routine.batched == 1:
             result += "  " + (NL + "  ").join(routine.batched_transform_to_complex(flavour)) + NL
         result += "  try {" + NL
         result += "    return static_cast<CLBlastStatusCode>(" + NL
@@ -388,7 +388,7 @@ def performance_test(routine, level_string):
         found = False
         for flavour in routine.flavours:
             if flavour.precision_name == precision:
-                extra_template_argument = "0, " if routine.name == "gemm" and not routine.batched else ""
+                extra_template_argument = "0, " if routine.name == "gemm" and routine.batched == 0 else ""
                 result += NL + "      clblast::RunClient<clblast::TestX" + routine.plain_name()
                 result += flavour.test_template(extra_template_argument)
                 result += ">(argc, argv); break;" + NL
@@ -410,7 +410,7 @@ def correctness_test(routine, level_string):
     result += "int main(int argc, char *argv[]) {" + NL
     result += "  auto errors = size_t{0};" + NL
     not_first = "false"
-    extra_template_arguments = ["1, ", "2, "] if routine.name == "gemm" and not routine.batched else [""]
+    extra_template_arguments = ["1, ", "2, "] if routine.name == "gemm" and routine.batched == 0 else [""]
     for extra_template_argument in extra_template_arguments:
         for flavour in routine.flavours:
             result += "  errors += clblast::RunTests<clblast::TestX" + routine.plain_name()
diff --git a/scripts/generator/generator/routine.py b/scripts/generator/generator/routine.py
index dd3c2ecb..f7c2a701 100644
--- a/scripts/generator/generator/routine.py
+++ b/scripts/generator/generator/routine.py
@@ -12,12 +12,12 @@ import generator.convert as convert
 
 class Routine:
     """Class holding routine-specific information (e.g. name, which arguments, which precisions)"""
-    def __init__(self, implemented, has_tests, batched, temp_buffer, level, name, template, flavours, sizes, options,
+    def __init__(self, implemented, has_tests, batched_strided, temp_buffer, level, name, template, flavours, sizes, options,
                  inputs, outputs, buffer_sizes, scalars, scratch,
                  description, details, requirements):
         self.implemented = implemented
         self.has_tests = has_tests
-        self.batched = batched
+        self.batched = batched_strided
         self.temp_buffer = temp_buffer
         self.level = level
         self.name = name
@@ -35,38 +35,42 @@ class Routine:
         self.requirements = requirements
 
     def lowercase_name(self):
-        postfix = "batched" if self.batched else ""
+        postfix = "strided" if self.batched == 2 else ""
+        postfix += "batched" if self.batched != 0 else ""
         return self.name + postfix
 
     def plain_name(self):
-        postfix = "Batched" if self.batched else ""
+        postfix = "Strided" if self.batched == 2 else ""
+        postfix += "Batched" if self.batched != 0 else ""
         return self.name + postfix
 
     def capitalized_name(self):
-        postfix = "Batched" if self.batched else ""
+        postfix = "Strided" if self.batched == 2 else ""
+        postfix += "Batched" if self.batched != 0 else ""
         return self.name.capitalize() + postfix
 
     def upper_name(self):
-        postfix = "BATCHED" if self.batched else ""
+        postfix = "STRIDED" if self.batched == 2 else ""
+        postfix += "BATCHED" if self.batched != 0 else ""
         return self.name.upper() + postfix
 
     def b_star(self):
-        return "*" if self.batched else ""
+        return "*" if self.batched == 1 else ""
 
     def b_s(self):
-        return "s" if self.batched else ""
+        return "s" if self.batched == 1 else ""
 
     def batch_count_def(self):
-        return ["const size_t batch_count"] if self.batched else []
+        return ["const size_t batch_count"] if self.batched != 0 else []
 
     def batch_count_list(self):
-        return ["batch_count"] if self.batched else []
+        return ["batch_count"] if self.batched != 0 else []
 
     def batch_count_type(self):
-        return ["const size_t"] if self.batched else []
+        return ["const size_t"] if self.batched != 0 else []
 
     def batch_count_doc(self):
-        return ["`const size_t batch_count`: Number of batches. This value must be positive."] if self.batched else []
+        return ["`const size_t batch_count`: Number of batches. This value must be positive."] if self.batched != 0 else []
 
     def batched_transform_to_cpp(self):
         result = []
@@ -230,6 +234,8 @@ class Routine:
             a = [name + "_buffer"]
             b = [name + "_offset" + self.b_s()]
             c = [name + "_" + self.postfix(name)] if (name not in self.buffers_without_ld_inc()) else []
+            if self.batched == 2:
+                c += [name + "_stride"]
             return [", ".join(a + b + c)]
         return []
 
@@ -239,6 +245,8 @@ class Routine:
             a = [name + "_buffer_bis"]
             b = [name + "_offset"]
             c = [name + "_" + self.postfix(name)] if name not in self.buffers_without_ld_inc() else []
+            if self.batched == 2:
+                c += [name + "_stride"]
             return [", ".join(a + b + c)]
         return []
 
@@ -258,6 +266,8 @@ class Routine:
             a = [prefix + "cl_mem " + name + "_buffer"]
             b = ["const size_t " + self.b_star() + name + "_offset" + self.b_s()]
             c = ["const size_t " + name + "_" + self.postfix(name)] if name not in self.buffers_without_ld_inc() else []
+            if self.batched == 2:
+                c += ["const size_t " + name + "_stride"]
             return [", ".join(a + b + c)]
         return []
 
@@ -307,8 +317,10 @@ class Routine:
         if name in self.inputs or name in self.outputs:
             buffer_type = "unsigned int" if (name in self.index_buffers()) else self.template.buffer_type
             a = ["Buffer<" + buffer_type + ">(" + name + "_buffer)"]
-            b = [name + "_offsets_cpp"] if self.batched else [name + "_offset"]
+            b = [name + "_offsets_cpp"] if self.batched == 1 else [name + "_offset"]
             c = [name + "_" + self.postfix(name)] if (name not in self.buffers_without_ld_inc()) else []
+            if self.batched == 2:
+                c += [name + "_stride"]
             return [", ".join(a + b + c)]
         return []
 
@@ -375,6 +387,8 @@ class Routine:
             a = [prefix + "cl_mem"]
             b = ["const size_t" + self.b_star()]
             c = ["const size_t"] if (name not in self.buffers_without_ld_inc()) else []
+            if self.batched == 2:
+                c += ["const size_t"]
             return [", ".join(a + b + c)]
         return []
 
@@ -391,13 +405,15 @@ class Routine:
             if name not in self.buffers_without_ld_inc():
                 c = ["`const size_t " + name + "_" + self.postfix(name) + "`: " +
                      inc_ld_description + "of the " + inout + " " + math_name + ". This value must be greater than 0."]
+            if self.batched == 2:
+                c += ["`const size_t " + name + "_stride`: The (fixed) stride between two batches of the " + name.upper() + " matrix."]
             return a + b + c
         return []
 
     def scalar(self, name):
         """Retrieves the name of a scalar (alpha/beta)"""
         if name in self.scalars:
-            if self.batched:
+            if self.batched == 1:
                 return [name + "s_cpp"]
             return [name]
         return []
@@ -418,11 +434,11 @@ class Routine:
         """Retrieves the use of a scalar (alpha/beta)"""
         if name in self.scalars:
             if name == "alpha":
-                if self.batched:
+                if self.batched == 1:
                     return ["alphas_cpp.data()"]
                 return [flavour.use_alpha()]
             elif name == "beta":
-                if self.batched:
+                if self.batched == 1:
                     return ["betas_cpp.data()"]
                 return [flavour.use_beta()]
             return [name]
@@ -866,7 +882,7 @@ class Routine:
         if self.name in self.routines_scalar_no_return():
             routine_name += "_sub"
             indent += "    "
-        if self.batched:
+        if self.batched != 0:
             routine_name += "batched"
         result = return_type + extra_qualifier + " cblas_" + flavour.name.lower() + routine_name + "("
         result += (",\n" + indent).join([a for a in self.arguments_def_netlib(flavour)]) + ")"