modif to try to make gpu work on windows.

theano/sandbox/cuda/cuda_ndarray.cu

@@ -7,7 +7,7 @@
 #include "cuda_ndarray.cuh"
 
 //If true, when there is a gpu malloc or free error, we print the size of allocated memory on the device.
-#define COMPUTE_GPU_MEM_USED false
+#define COMPUTE_GPU_MEM_USED 0
 
 /////////////////////////
 // Alloc and Free
@@ -420,7 +420,7 @@ PyObject* CudaNdarray_Zeros(PyObject* dummy, PyObject* shape)
 PyObject * CudaNdarray_Copy(CudaNdarray * self)
 {
     PyObject * rval = CudaNdarray_new_null();
-    if ((!rval) or (-1 == self->nd))
+    if ((!rval) || (-1 == self->nd))
     {
         return rval;
     }
@@ -1871,7 +1871,7 @@ filter(PyObject* __unsed_self, PyObject *args) // args = (data, broadcastable, s
 
     CudaNdarray * cnda = (CudaNdarray*)py_data;
 
-    if (strict or CudaNdarray_Check(py_data))
+    if (strict || CudaNdarray_Check(py_data))
     {
         //TODO: support non-strict "casting" from a vt to the broadcastable/type/size that we need.
         if (!CudaNdarray_Check(py_data)) 
@@ -1890,7 +1890,7 @@ filter(PyObject* __unsed_self, PyObject *args) // args = (data, broadcastable, s
         }
         for (int i = 0; i < cnda->nd; ++i)
         {
-            if ((CudaNdarray_HOST_DIMS(cnda)[i] > 1) and PyInt_AsLong(PyTuple_GetItem(broadcastable, Py_ssize_t(i))))
+            if ((CudaNdarray_HOST_DIMS(cnda)[i] > 1) && PyInt_AsLong(PyTuple_GetItem(broadcastable, Py_ssize_t(i))))
             {
                 PyErr_Format(PyExc_TypeError, "Non-unit size in broadcastable vt dimension %i", i);
                 Py_DECREF(py_data);
@@ -1913,7 +1913,7 @@ filter(PyObject* __unsed_self, PyObject *args) // args = (data, broadcastable, s
         }
         for (int i = 0; i < data->nd; ++i)
         {
-            if ((data->dimensions[i] > 1) and PyInt_AsLong(PyTuple_GetItem(broadcastable, Py_ssize_t(i))))
+            if ((data->dimensions[i] > 1) && PyInt_AsLong(PyTuple_GetItem(broadcastable, Py_ssize_t(i))))
             {
                 PyErr_Format(PyExc_TypeError, "Non-unit size in broadcastable dimension %i", i);
                 Py_DECREF(data);
@@ -2114,7 +2114,7 @@ CudaNdarray_is_c_contiguous(const CudaNdarray * self)
 {
     bool c_contiguous = true;
     int size = 1;
-    for (int i = self->nd-1; (i >= 0) and c_contiguous; --i)
+    for (int i = self->nd-1; (i >= 0) && c_contiguous; --i)
     {
         if (CudaNdarray_HOST_DIMS(self)[i] == 1)
             continue;
@@ -2763,7 +2763,7 @@ CudaNdarray_dimshuffle(CudaNdarray * self, unsigned int len, const int * pattern
         }
         else
         {
-            if ((dims_taken[pattern[i]]) or (pattern[i]>= self->nd))
+            if ((dims_taken[pattern[i]]) || (pattern[i]>= self->nd))
             {
                 PyErr_SetString(PyExc_ValueError, "invalid pattern for Cudandarray_dimshuffle");
                 free(newdims);

theano/sandbox/cuda/nvcc_compiler.py

 import sys, os, subprocess, logging
 from theano.gof.cmodule import (std_libs, std_lib_dirs, std_include_dirs, dlimport,
     get_lib_extension, local_bitwidth)
-from theano import config
 import distutils
 import commands
 
@@ -67,7 +66,8 @@ def nvcc_module_compile_str(module_name, src_code, location=None, include_dirs=[
     if preargs is None:
         preargs= []
     else: preargs = list(preargs)
-    preargs.append('-fPIC')
+    if sys.platform!='win32':
+        preargs.append('-fPIC')
     no_opt = False
     cuda_root = config.cuda.root
     include_dirs = std_include_dirs() + include_dirs + [os.path.split(__file__)[0]]
@@ -116,12 +116,18 @@ def nvcc_module_compile_str(module_name, src_code, location=None, include_dirs=[
     cmd = [nvcc_path, '-shared', '-g'] + preargs1
     if config.nvcc.compiler_bindir:
         cmd.extend(['--compiler-bindir', config.nvcc.compiler_bindir])
-    if local_bitwidth() == 64:
-        cmd.append('-m64')
-        cmd.extend(['-Xcompiler', ','.join(preargs2 +[ '-m64'])])
-    else:
-        cmd.append('-m32')
-        cmd.extend(['-Xcompiler', ','.join(preargs2 +[ '-m32'])])
+   
+    if sys.platform!='win32':
+        if local_bitwidth() == 64:
+            cmd.append('-m64')
+            preargs2.append('-m64')
+        else:
+            cmd.append('-m32')
+            preargs2.append('-m32')
+            
+    if len(preargs2)>0:
+        cmd.extend(['-Xcompiler', ','.join(preargs2)])
+
     if os.path.exists(os.path.join(config.cuda.root,'lib')):
         cmd.extend(['-Xlinker',','.join(['-rpath',os.path.join(config.cuda.root,'lib')])])
         if sys.platform != 'darwin':