Skip to content

P
pytensor
提交 c41a814a authored 作者: Pierre Luc Carrier's avatar Pierre Luc Carrier
浏览文件
操作

Bug fixes in op GpuImages2Neibs

上级 6949fa64
隐藏空白字符变更
内嵌 并排
正在显示 包含 66 行增加 和 56 行删除
theano/sandbox/gpuarray/neighbours.py
浏览文件 @ c41a814a
# This is work in progress
from theano import Op, Apply
import numpy
from theano import Op, Apply, config
from theano.gof import local_optimizer
from theano.sandbox.cuda import cuda_available
from theano.sandbox.cuda.nvcc_compiler import NVCC_compiler
from theano.sandbox.neighbours import Images2Neibs
import theano.tensor as T
try:
import pygpu
from pygpu import gpuarray, elemwise
except ImportError:
pass
if cuda_available:
from theano.sandbox.cuda import CudaNdarrayType
from theano.sandbox.cuda.basic_ops import host_from_gpu, gpu_from_host
from theano.sandbox.cuda.opt import register_opt as register_gpu_opt
from theano.sandbox.gpuarray.basic_ops import (as_gpuarray_variable,
host_from_gpu, gpu_from_host)
from theano.sandbox.gpuarray.opt import register_opt as register_gpu_opt
from theano.sandbox.gpuarray.type import GpuArrayType
class GpuImages2Neibs(Images2Neibs, Op):
......@@ -22,20 +28,22 @@ class GpuImages2Neibs(Images2Neibs, Op):
self.mode = mode
def make_node(self, ten4, neib_shape, neib_step):
assert ten4.dtype == 'float32'
if not isinstance(ten4.type, CudaNdarrayType):
raise TypeError('ten4 must be cudandarray', ten4)
assert ten4.dtype in ['int64', 'float32', 'float64']
assert ten4.ndim == 4
assert neib_shape.ndim == 1
assert neib_step.ndim == 1
ten4 = as_gpuarray_variable(ten4)
neib_shape = T.as_tensor_variable(neib_shape)
neib_step = T.as_tensor_variable(neib_step)
return Apply(self, [ten4, neib_shape, neib_step],
[CudaNdarrayType(broadcastable=(False, False),
dtype=ten4.type.dtype)()])
[GpuArrayType(broadcastable=(False, False),
dtype=ten4.type.dtype)()])
def c_code_cache_version(self):
return (8,)
return (9,)
def c_headers(self):
return ['cuda.h', '<compyte/extension.h>', '<numpy_compat.h>',
......@@ -48,6 +56,8 @@ class GpuImages2Neibs(Images2Neibs, Op):
return ['setup_ext_cuda();']
def c_support_code_apply(self, node, nodename):
dtype_ten4 = node.inputs[0].dtype
dtype_z = node.outputs[0].dtype
mode = self.mode
return """
//a version that use less register but don't work in all case.
......@@ -64,9 +74,9 @@ class GpuImages2Neibs(Images2Neibs, Op):
const int grid_d,
const int stride0, const int stride1,
const int stride2, const int stride3,
float * global_ten4,
npy_%(dtype_ten4)s * global_ten4,
const int out_s0, const int out_s1,
float * global_out
npy_%(dtype_z)s * global_out
)
{
const int wrap_centered_idx_shift_x = c/2;
......@@ -136,9 +146,9 @@ class GpuImages2Neibs(Images2Neibs, Op):
const int grid_d,
const int stride0, const int stride1,
const int stride2, const int stride3,
float * global_ten4,
npy_%(dtype_ten4)s * global_ten4,
const int out_s0, const int out_s1,
float * global_out
npy_%(dtype_z)s * global_out
)
{
const int wrap_centered_idx_shift_x = c/2;
......@@ -196,17 +206,24 @@ class GpuImages2Neibs(Images2Neibs, Op):
}
}
}
""" % locals()
def c_code(self, node, name, inp, out, sub):
dtype_ten4 = node.inputs[0].dtype
dtype_neib_shape = node.inputs[1].dtype
dtype_neib_step = node.inputs[2].dtype
dtype_z = node.outputs[0].dtype
itemsize_ten4 = numpy.dtype(dtype_ten4).itemsize
itemsize_z = numpy.dtype(dtype_z).itemsize
typecode_z = pygpu.gpuarray.dtype_to_typecode(node.outputs[0].dtype)
ten4, neib_shape, neib_step = inp
ten4, neib_shape, neib_step = inp
z, = out
fail = sub['fail']
mode = self.mode
if config.gpuarray.sync:
cnda_thread_sync = "GpuArray_sync(&%(zz)s->ga);" % dict(zz=zz)
else:
cnda_thread_sync = ""
return """
#ifndef CEIL_INTDIV
#define CEIL_INTDIV(a, b) ((a/b) + ((a %% b) ? 1: 0))
......@@ -221,7 +238,7 @@ class GpuImages2Neibs(Images2Neibs, Op):
PyErr_Format(PyExc_TypeError, "pvals wrong rank");
%(fail)s;
}
if (PyGpuArray_NDIM(%(neib_shape)s) != 1)
if (PyArray_NDIM(%(neib_shape)s) != 1)
{
PyErr_Format(PyExc_TypeError, "unis wrong rank");
%(fail)s;
......@@ -234,13 +251,13 @@ class GpuImages2Neibs(Images2Neibs, Op):
%(fail)s;
}
const int c = *(dtype_%(neib_shape)s*) PyArray_GETPTR1(
const int c = *(npy_%(dtype_neib_shape)s*) PyArray_GETPTR1(
%(neib_shape)s, 0);
const int d = *(dtype_%(neib_shape)s*) PyArray_GETPTR1(
const int d = *(npy_%(dtype_neib_shape)s*) PyArray_GETPTR1(
%(neib_shape)s, 1);
const npy_intp step_x = (npy_intp) *(dtype_%(neib_step)s*)
const npy_intp step_x = (npy_intp) *(npy_%(dtype_neib_step)s*)
PyArray_GETPTR1(%(neib_step)s, 0);
const npy_intp step_y = (npy_intp) *(dtype_%(neib_step)s*)
const npy_intp step_y = (npy_intp) *(npy_%(dtype_neib_step)s*)
PyArray_GETPTR1(%(neib_step)s, 1);
if ( "%(mode)s" == "wrap_centered") {
......@@ -315,7 +332,7 @@ class GpuImages2Neibs(Images2Neibs, Op):
|| (PyGpuArray_DIMS(%(z)s)[1] != z_dim1))
{
Py_XDECREF(%(z)s);
npy_intp dims[2];
size_t dims[2];
dims[0] = z_dim0;
dims[1] = z_dim1;
%(z)s = pygpu_empty(2, dims, %(typecode_z)s,
......@@ -337,14 +354,14 @@ class GpuImages2Neibs(Images2Neibs, Op):
const int nb_stack = PyGpuArray_DIMS(%(ten4)s)[1];
const int height = PyGpuArray_DIMS(%(ten4)s)[2];
const int width = PyGpuArray_DIMS(%(ten4)s)[3];
const int c = *(dtype_%(neib_shape)s*) PyArray_GETPTR1(
const int c = *(npy_%(dtype_neib_shape)s*) PyArray_GETPTR1(
%(neib_shape)s, 0);
const int d = *(dtype_%(neib_shape)s*) PyArray_GETPTR1(
const int d = *(npy_%(dtype_neib_shape)s*) PyArray_GETPTR1(
%(neib_shape)s, 1);
const npy_intp step_x = (npy_intp) *(dtype_%(neib_step)s*)
const npy_intp step_x = (npy_intp) *(npy_%(dtype_neib_step)s*)
PyArray_GETPTR1(%(neib_step)s, 0);
const npy_intp step_y = (npy_intp) *(dtype_%(neib_step)s*)
const npy_intp step_y = (npy_intp) *(npy_%(dtype_neib_step)s*)
PyArray_GETPTR1(%(neib_step)s, 1);
dim3 n_threads(d,c,1);
......@@ -371,9 +388,9 @@ class GpuImages2Neibs(Images2Neibs, Op):
int, int, int ,int,
int, int,
int, int, int, int,
float*,
npy_%(dtype_ten4)s*,
int, int,
float*);
npy_%(dtype_z)s*);
if(n_threads.x==d && n_threads.y==c){
f = k_multi_warp_less_%(name)s;
}else{
......@@ -386,19 +403,19 @@ class GpuImages2Neibs(Images2Neibs, Op):
height, width,
c, d, step_x, step_y,
grid_c, grid_d,
PyGpuArray_STRIDES(%(ten4)s)[0],
PyGpuArray_STRIDES(%(ten4)s)[1],
PyGpuArray_STRIDES(%(ten4)s)[2],
PyGpuArray_STRIDES(%(ten4)s)[3],
PyGpuArray_STRIDES(%(ten4)s)[0] / %(itemsize_ten4)s,
PyGpuArray_STRIDES(%(ten4)s)[1] / %(itemsize_ten4)s,
PyGpuArray_STRIDES(%(ten4)s)[2] / %(itemsize_ten4)s,
PyGpuArray_STRIDES(%(ten4)s)[3] / %(itemsize_ten4)s,
(npy_%(dtype_ten4)s*)(
((char *)cuda_get_ptr(%(ten4)s->ga.data)) +
%(ten4)s->ga.offset),
PyGpuArray_STRIDES(%(z)s)[0],
PyGpuArray_STRIDES(%(z)s)[1],
PyGpuArray_STRIDES(%(z)s)[0] / %(itemsize_z)s,
PyGpuArray_STRIDES(%(z)s)[1] / %(itemsize_z)s,
(npy_%(dtype_z)s*)(((char *)cuda_get_ptr(%(z)s->ga.data)) +
%(z)s->ga.offset),
%(z)s->ga.offset)
);
CNDA_THREAD_SYNC;
%(cnda_thread_sync)s
cudaError_t sts = cudaGetLastError();
if (cudaSuccess != sts)
{
......@@ -427,12 +444,11 @@ def gpu_images2neibs(ten4, neib_shape, neib_step=None, mode='valid'):
@local_optimizer([Images2Neibs])
def use_gpu_images2neibs(node):
if (type(node.op) is Images2Neibs and
node.inputs[0].dtype == 'float32' and
node.inputs[0].dtype in ['int64', 'float32', 'float64'] and
node.op.mode in ['valid', 'ignore_borders',
'wrap_centered']):
return [host_from_gpu(gpu_images2neibs(gpu_from_host(node.inputs[0]),
node.inputs[1], node.inputs[2],
mode=node.op.mode))]
if cuda_available:
register_gpu_opt()(use_gpu_images2neibs)
register_gpu_opt()(use_gpu_images2neibs)
theano/sandbox/gpuarray/tests/test_neighbours.py
浏览文件 @ c41a814a
# Skip test if cuda_ndarray is not available.
from nose.plugins.skip import SkipTest
import unittest
import theano.sandbox.cuda as cuda_ndarray
if cuda_ndarray.cuda_available == False:
raise SkipTest('Optional package cuda disabled')
# We let that import do the init of the back-end if needed.
from theano.sandbox.gpuarray.tests.test_basic_ops import (mode_with_gpu,
mode_without_gpu)
import theano.sandbox.test_neighbours
from theano.sandbox.gpuarray.neighbours import GpuImages2Neibs
if theano.config.mode == 'FAST_COMPILE':
mode_with_gpu = theano.compile.mode.get_mode('FAST_RUN').including('gpu')
else:
mode_with_gpu = theano.compile.mode.get_default_mode().including('gpu')
class T_GpuImages2Neibs(theano.sandbox.test_neighbours.T_Images2Neibs):
mode = mode_with_gpu
op = GpuImages2Neibs
dtypes = ['float32']
dtypes = ['int64', 'float32', 'float64']
if __name__ == '__main__':
unittest.main()
Markdown 格式
0% 或
您添加了 0 人 到此讨论。请谨慎行事。
请先完成此评论的编辑！
请 注册 或者 后发表评论