Fix crash reported by Arjun Jain on the mailing list with gpu reduction 1111 and strided inputs.

theano/sandbox/cuda/basic_ops.py

@@ -1572,6 +1572,9 @@ class GpuCAReduce(GpuOp):
             }
             n_threads.z -= 1;
 
+            //Maximum for Fermi GPU on that dimensions.
+            n_threads.z = std::min(n_threads.z, (unsigned)64);
+
             dim3 n_blocks(1,1,1);
             %(makecall)s
         }
@@ -1602,7 +1605,7 @@ class GpuCAReduce(GpuOp):
         """ % locals()
 
     def c_code_cache_version_apply(self, node):
-        version = [7]  # the version corresponding to the c code in this Op
+        version = [8]  # the version corresponding to the c code in this Op
 
         # now we insert versions for the ops on which we depend...
         scalar_node = Apply(self.scalar_op,

theano/sandbox/cuda/tests/test_basic_ops.py

@@ -114,7 +114,7 @@ def test_careduce():
                                ((4100,4,3,2),[1,3]),((4,4100,3,2),[1,3]),((4,3,4100,2),[1,3]),((4,3,2,4100),[1,3]),#0101
                                ((4100,4,3,2),[0,2,3]),((4,4100,3,2),[0,2,3]),((4,3,4100,2),[0,2,3]),#((4,3,2,4100),[0,2,3]),#1011
                                ((4100,4,3,2),[1,2,3]),((4,4100,3,2),[1,2,3]),((4,3,4100,2),[1,2,3]),((4,3,2,4100),[1,2,3]),#0111
-                               ((4100,2,3,4),[0,1,2,3]),((2,4100,3,4),[0,1,2,3]),((2,3,4100,4),[0,1,2,3]),((2,3,4,4100),[0,1,2,3]),#1111
+                               ((4100,2,3,4),[0,1,2,3]),((2,4100,3,4),[0,1,2,3]),((2,3,4100,4),[0,1,2,3]),((2,3,4,4100),[0,1,2,3]),((128,1,3,3), [0,1,2,3]),#1111
 
 
                                #test pattern implemented by reshape
@@ -197,7 +197,9 @@ def test_careduce():
         for shape, pattern in [#((5,),[0]),
                                ((5,4),[0,1]),((5,4),[0]),
                                ((5,4,3),[0]),((5,4,3),[0,1]),((5,4,3),[2]),((5,4,3),[0,1,2]),
-                               ((5,4,3,2),[0,1,2,3]), ((5,4,3,2),[0,2,3])]:
+                               ((5,4,3,2),[0,1,2,3]), ((5,4,3,2),[0,2,3]),
+                               ((128,1,3,3),[0,1,2,3]),
+        ]:
             op = careduce_op(scalar_op, axis=pattern)
             pat = tensor_pattern_to_gpu_pattern(shape, pattern)
 
@@ -229,7 +231,9 @@ def test_careduce():
                                ((5,4),[0,1]),((5,4),[0]),
                                ((5,4,3),[0]),((5,4,3),[0,1]),
                                ((5,4,3),[2]),((5,4,3),[0,1,2]),
-                               ((5,4,3,2),[0,1,2,3]), ((5,4,3,2),[0,2,3])]:
+                               ((5,4,3,2),[0,1,2,3]), ((5,4,3,2),[0,2,3]),
+                               ((128,1,3,3),[0,1,2,3]),
+        ]:
             op = careduce_op(scalar_op, axis=pattern)
             pat = tensor_pattern_to_gpu_pattern(shape, pattern)