Tensorflow: o bug de usar várias GPUs, relacionado a tf.Variable fixado à CPU

Informação do ambiente

Sistema operacional: Ubuntu 14.04

Versão instalada de CUDA e cuDNN: 7.5 e 4.0.7
(anexe o resultado de ls -l /path/to/cuda/lib/libcud* ):

Se instalado a partir de fontes, forneça o hash de confirmação: 4a4f2461533847dde239851ecebe5056088a828c

Passos para reproduzir

Execute o seguinte código

import tensorflow as tf

def main():
    a = tf.Variable(1)
    init_a = tf.initialize_all_variables()
    with tf.Session() as sess:
        sess.run(init_a)

    with tf.device("/gpu:0"):
        b = tf.constant(2)
        init_b = tf.initialize_all_variables()
    with tf.Session() as sess:
        sess.run(init_b)

    with tf.device("/cpu:0"):
        c = tf.Variable(2)
        init_c = tf.initialize_all_variables()
    with tf.Session() as sess:
        sess.run(init_c)

    with tf.device("/gpu:0"):
        d = tf.Variable(2)
        init_d = tf.initialize_all_variables()
    with tf.Session() as sess:
        sess.run(init_d)

if __name__ == '__main__':
    main()

Logs ou outra saída que seria útil

(Se os logs forem grandes, envie como anexo).

I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcublas.so locally
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcudnn.so locally
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcufft.so locally
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcuda.so locally
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcurand.so locally
I tensorflow/core/common_runtime/gpu/gpu_init.cc:102] Found device 0 with properties: 
name: GeForce GTX TITAN X
major: 5 minor: 2 memoryClockRate (GHz) 1.266
pciBusID 0000:05:00.0
Total memory: 12.00GiB
Free memory: 11.02GiB
I tensorflow/core/common_runtime/gpu/gpu_init.cc:102] Found device 1 with properties: 
name: GeForce GTX 980
major: 5 minor: 2 memoryClockRate (GHz) 1.2785
pciBusID 0000:09:00.0
Total memory: 4.00GiB
Free memory: 3.91GiB
I tensorflow/core/common_runtime/gpu/gpu_init.cc:59] cannot enable peer access from device ordinal 0 to device ordinal 1
I tensorflow/core/common_runtime/gpu/gpu_init.cc:59] cannot enable peer access from device ordinal 1 to device ordinal 0
I tensorflow/core/common_runtime/gpu/gpu_init.cc:126] DMA: 0 1 
I tensorflow/core/common_runtime/gpu/gpu_init.cc:136] 0:   Y N 
I tensorflow/core/common_runtime/gpu/gpu_init.cc:136] 1:   N Y 
I tensorflow/core/common_runtime/gpu/gpu_device.cc:756] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX TITAN X, pci bus id: 0000:05:00.0)
I tensorflow/core/common_runtime/gpu/gpu_device.cc:756] Creating TensorFlow device (/gpu:1) -> (device: 1, name: GeForce GTX 980, pci bus id: 0000:09:00.0)
I tensorflow/core/common_runtime/gpu/gpu_device.cc:756] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX TITAN X, pci bus id: 0000:05:00.0)
I tensorflow/core/common_runtime/gpu/gpu_device.cc:756] Creating TensorFlow device (/gpu:1) -> (device: 1, name: GeForce GTX 980, pci bus id: 0000:09:00.0)
I tensorflow/core/common_runtime/gpu/gpu_device.cc:756] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX TITAN X, pci bus id: 0000:05:00.0)
I tensorflow/core/common_runtime/gpu/gpu_device.cc:756] Creating TensorFlow device (/gpu:1) -> (device: 1, name: GeForce GTX 980, pci bus id: 0000:09:00.0)
I tensorflow/core/common_runtime/gpu/gpu_device.cc:756] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX TITAN X, pci bus id: 0000:05:00.0)
I tensorflow/core/common_runtime/gpu/gpu_device.cc:756] Creating TensorFlow device (/gpu:1) -> (device: 1, name: GeForce GTX 980, pci bus id: 0000:09:00.0)
Traceback (most recent call last):
  File "test_multi_gpu.py", line 30, in <module>
    main()
  File "test_multi_gpu.py", line 26, in main
    sess.run(init_d)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/client/session.py", line 332, in run
    run_metadata_ptr)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/client/session.py", line 572, in _run
    feed_dict_string, options, run_metadata)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/client/session.py", line 652, in _do_run
    target_list, options, run_metadata)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/client/session.py", line 672, in _do_call
    raise type(e)(node_def, op, message)
tensorflow.python.framework.errors.InvalidArgumentError: Cannot assign a device to node 'Variable_2': Could not satisfy explicit device specification '/device:GPU:0' because no supported kernel for GPU devices is available
     [[Node: Variable_2 = Variable[container="", dtype=DT_INT32, shape=[], shared_name="", _device="/device:GPU:0"]()]]
Caused by op u'Variable_2', defined at:
  File "test_multi_gpu.py", line 30, in <module>
    main()
  File "test_multi_gpu.py", line 23, in main
    d = tf.Variable(2)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variables.py", line 211, in __init__
    dtype=dtype)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variables.py", line 292, in _init_from_args
    name=name)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/state_ops.py", line 139, in variable_op
    container=container, shared_name=shared_name)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/gen_state_ops.py", line 351, in _variable
    name=name)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/op_def_library.py", line 693, in apply_op
    op_def=op_def)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/ops.py", line 2177, in create_op
    original_op=self._default_original_op, op_def=op_def)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/ops.py", line 1161, in __init__
    self._traceback = _extract_stack()

Também percebi que a documentação para Uso de GPUs não menciona tf.Variable, ela envolve apenas tf.constant e tf.matmul.

OK, encontrei a documentação de [Redes Neurais Convolucionais] (https://www.tensorflow.org/versions/r0.8/tutorials/deep_cnn/index.html),
citações:

All variables are pinned to the CPU and accessed via tf.get_variable() in order to share them in a multi-GPU version. See how-to on Sharing Variables.

Gostaria de perguntar que, uma vez que tf.Variables está fixado na CPU por tensorflow, poderíamos corrigir esse erro? Precisamos observar com muito cuidado para excluir a declaração tf.Variable fora do escopo with tf.device('/gpu:xx') ou usar netsted with tf.device(None) para tratá-la?