added version to softmax

kernel_codegen.py

@@ -25,6 +25,8 @@ def nvcc_kernel(name, params, body):
 
 def code_version(version):
     """decorator to support version-based cache mechanism"""
+    if not isinstance(version, tuple):
+        raise TypeError('version must be tuple', version)
     def deco(f):
         f.code_version = version
         return f
@@ -32,7 +34,7 @@ def code_version(version):
 
 UNVERSIONED = ()
 
-@code_version(UNVERSIONED)
+@code_version((1,))
 def inline_reduce(N, buf, pos, count, manner_fn):
     """
     Return C++ code for a function that reduces a contiguous buffer.
@@ -103,7 +105,7 @@ def inline_reduce_prod(N, buf, pos, count):
     return inline_reduce(N, buf, pos, count, lambda a, b: "%s * %s"%(a,b))
 
 
-@code_version(UNVERSIONED + inline_reduce_max.code_version + inline_reduce_sum.code_version)
+@code_version((1,) + inline_reduce_max.code_version + inline_reduce_sum.code_version)
 def inline_softmax(N, buf, buf2, threadPos, threadCount):
     """
     :Precondition: buf and buf2 contain two identical copies of the input to softmax

nnet.py

@@ -301,8 +301,8 @@ class GpuSoftmax (Op):
     def make_node(self, x):
         return Apply(self, [x],[x.type()])
     def c_code_cache_version(self):
-        return ()
-        # reduce (1,) + device_softmax.code_version
+        #return ()
+        return (1,) + inline_softmax.code_version
     def c_code(self, node, nodename, (x,), (z,), sub):
         fail = sub['fail']
         return """