Update magma cholesky to use params type

theano/gpuarray/linalg.py

@@ -384,9 +384,6 @@ class GpuMagmaBase(COp):
                 magma_init()
                 ctx.is_magma_initialized = True
 
-    def get_params(self, node):
-        return node.inputs[0].type.context
-
 
 class GpuMagmaSVD(GpuMagmaBase):
     """Computes the svd of a matrix :math:`A` using magma library.
@@ -534,6 +531,8 @@ class GpuMagmaCholesky(GpuMagmaBase, CGpuKernelBase):
 
     """
     __props__ = ('lower', 'inplace')
+    check_input = False
+    params_type = ParamsType(lower=bool_t, inplace=bool_t, context=gpu_context_type)
 
     def __init__(self, lower=True, inplace=False):
         self.lower = lower
@@ -555,13 +554,8 @@ class GpuMagmaCholesky(GpuMagmaBase, CGpuKernelBase):
             raise TypeError("only `float32` is supported for now")
         return theano.Apply(self, [A], [A.type()])
 
-    def get_op_params(self):
-        params = []
-        if self.lower:
-            params.append(('LOWER', '1'))
-        if self.inplace:
-            params.append(('INPLACE', '1'))
-        return params
+    def get_params(self, node):
+        return self.params_type.get_params(self, context=node.inputs[0].type.context)
 
     def infer_shape(self, node, shapes):
         return [shapes[0]]

theano/gpuarray/magma_cholesky.c

@@ -39,7 +39,7 @@ setup_ext_cuda();
 #section support_code_struct
 
 int APPLY_SPECIFIC(magma_cholesky)(PyGpuArrayObject *A, PyGpuArrayObject **L,
-                                   PyGpuContextObject *c) {
+                                   PARAMS_TYPE* params) {
   const size_t *dims;
   size_t N, n2;
   magma_uplo_t ul;
@@ -50,37 +50,38 @@ int APPLY_SPECIFIC(magma_cholesky)(PyGpuArrayObject *A, PyGpuArrayObject **L,
                     "GpuMagmaCholesky: unsupported data type");
     return -1;
   }
+
+  // This is early to match the exit() in the fail label.
+  cuda_enter(params->context->ctx);
+
   if (!GpuArray_IS_C_CONTIGUOUS(&A->ga)) {
     PyErr_SetString(PyExc_ValueError,
                     "GpuMagmaCholesky: requires data to be C-contiguous");
-    return -1;
+    goto fail;
   }
   if (PyGpuArray_NDIM(A) != 2) {
     PyErr_SetString(PyExc_ValueError, "GpuMagmaCholesky: matrix rank error");
-    return -1;
+    goto fail;
   }
   dims = PyGpuArray_DIMS(A);
   if (dims[0] != dims[1]) {
     PyErr_SetString(PyExc_ValueError, "GpuMagmaCholesky: matrix is not square");
-    return -1;
+    goto fail;
   }
 
-  // This is early to match the exit() in the fail label.
-  cuda_enter(c->ctx);
-
-#ifdef INPLACE
-  Py_XDECREF(*L);
-  *L = A;
-  Py_INCREF(*L);
-#else
-  *L = theano_try_copy(*L, A);
-  if (*L == NULL) {
-    PyErr_SetString(
-        PyExc_RuntimeError,
-        "GpuMagmaCholesky: failed to allocate memory for the output");
-    goto fail;
+  if (params->inplace) {
+    Py_XDECREF(*L);
+    *L = A;
+    Py_INCREF(*L);
+  } else {
+    *L = theano_try_copy(*L, A);
+    if (*L == NULL) {
+      PyErr_SetString(
+          PyExc_RuntimeError,
+          "GpuMagmaCholesky: failed to allocate memory for the output");
+      goto fail;
+    }
   }
-#endif
 
   // magma matrix cholesky
   N = dims[0];
@@ -90,11 +91,12 @@ int APPLY_SPECIFIC(magma_cholesky)(PyGpuArrayObject *A, PyGpuArrayObject **L,
 // matrix order which requires copying data, we can compute cholesky
 // decomposition where we change parameters lower to upper and upper to
 // lower.
-#ifdef LOWER
-  ul = MagmaUpper;
-#else
-  ul = MagmaLower;
-#endif
+  if (params->lower) {
+    ul = MagmaUpper;
+  }
+  else {
+    ul = MagmaLower;
+  }
   magma_spotrf_gpu(ul, N, (float *)PyGpuArray_DEV_DATA(*L), N, &info);
   if (info > 0) {
     PyErr_Format(PyExc_RuntimeError, "GpuMagmaCholesky: the leading minor of "
@@ -109,23 +111,23 @@ int APPLY_SPECIFIC(magma_cholesky)(PyGpuArrayObject *A, PyGpuArrayObject **L,
     goto fail;
   }
 
-#ifdef LOWER
-  res = tril_kernel_scall(1, &n2, 0, n2, N, (*L)->ga.offset, (*L)->ga.data);
-  if (res != GA_NO_ERROR) {
-    PyErr_Format(PyExc_RuntimeError, "GpuMagmaCholesky: tril_kernel %s.",
-                 GpuKernel_error(&k_tril_kernel, res));
-    goto fail;
-  }
-#else
-  res = triu_kernel_scall(1, &n2, 0, n2, N, (*L)->ga.offset, (*L)->ga.data);
-  if (res != GA_NO_ERROR) {
-    PyErr_Format(PyExc_RuntimeError, "GpuMagmaCholesky: triu_kernel %s.",
-                 GpuKernel_error(&k_triu_kernel, res));
-    goto fail;
+  if (params->lower) {
+    res = tril_kernel_scall(1, &n2, 0, n2, N, (*L)->ga.offset, (*L)->ga.data);
+    if (res != GA_NO_ERROR) {
+      PyErr_Format(PyExc_RuntimeError, "GpuMagmaCholesky: tril_kernel %s.",
+                   GpuKernel_error(&k_tril_kernel, res));
+      goto fail;
+    }
+  } else {
+    res = triu_kernel_scall(1, &n2, 0, n2, N, (*L)->ga.offset, (*L)->ga.data);
+    if (res != GA_NO_ERROR) {
+      PyErr_Format(PyExc_RuntimeError, "GpuMagmaCholesky: triu_kernel %s.",
+                   GpuKernel_error(&k_triu_kernel, res));
+      goto fail;
+    }
   }
-#endif
   res = 0;
 fail:
-  cuda_exit(c->ctx);
+  cuda_exit(params->context->ctx);
   return res;
 }

theano/gpuarray/magma_inv.c

@@ -20,7 +20,6 @@ int APPLY_SPECIFIC(magma_inv)(PyGpuArrayObject *A, PyGpuArrayObject **A_inv,
 
   // This is early to match the exit() in the fail label.
   cuda_enter(params->context->ctx);
-  magma_init();
 
   if (!GpuArray_IS_C_CONTIGUOUS(&A->ga)) {
     PyErr_SetString(PyExc_ValueError,

theano/gpuarray/magma_svd.c

@@ -27,7 +27,6 @@ int APPLY_SPECIFIC(magma_svd)(PyGpuArrayObject *A,
 
   // This is early to match the exit() in the fail label.
   cuda_enter(params->context->ctx);
-  magma_init();
 
   if (!GpuArray_IS_C_CONTIGUOUS(&A->ga)) {
     PyErr_SetString(PyExc_ValueError,

theano/gpuarray/opt.py

@@ -74,7 +74,7 @@ from .subtensor import (GpuIncSubtensor, GpuSubtensor,
 from .opt_util import alpha_merge, output_merge, pad_dims, unpad_dims
 from .reduction import GpuMaxAndArgmax
 from .linalg import (GpuCusolverSolve, MATRIX_STRUCTURES_SOLVE, GpuCholesky,
-                     cusolver_available, GpuMagmaMatrixInverse, GpuMagmaSVD,
+                     cusolver_available, GpuMagmaMatrixInverse, gpu_svd,
                      GpuMagmaCholesky, GpuMagmaQR, GpuMagmaEigh)
 
 _logger = logging.getLogger("theano.gpuarray.opt")