* Optimize compilation with Ninja, notest, test=windows_ci, test=windows_op

* no cache on windows ci, notest, test=windows_ci, test=windows_op

* delete /Zc:inline compiled in NVCC, notest, test=windows_ci, test=windows_op

* fix test_warpctc_op, notest, test=windows_ci

* remove test code, test=develop

* add float64 input to ctc_loss

* modified error message of  warpctc

* update repo and tag of warpctc

* add test for warpctc with float64 input

* modified warpctc.cmake to make sure build always

* resolved sample code bug of warpctc

* add core.ops in warpctc dygraph

* fix a bug of test

tweak the interface of cache_third_party function - expose the SOURCE_DIR for each external library (#21899)

* modify the repo address of eigen and warpctc

* fix the eigen not work on windows

* fix the eigen and warpctc can't recompile

test=develop

* add support to gcc8, add docker env test=develop

* fix the random compilation failure on windows 

* remove unused FLAGS_warpctc_dir

test=develop

* remove FLAGS_warpctc_dir

test=develop

test=develop

* "fix cuda9 error"

* "change commit id"

* "remote git tag"

* Update source files.

* Update headers

* Update

* Update

* Update

* Update

* Fix a CMake dependency

* It will sightly faster compile and make warpctc compile
well on CUDA 9 and GCC 5.5

* Add Seq2BatchFunctor, which will be used in WarpCTCOp.

* Implement WrapCTCFunctor and WrapCTCKernel.

* Add unittest of warpctc_op.

* Modify the check_output inferface in python unittest framework to allow check a subset of outputs.

* Use absolute offset lod in warpctc_op and related functors.

* Refine the comments of warpctc_op.

* The new python unittest supports checking a subset of the outputs, so revoke the previous change.

* Rename the transform from LoDTensor to Tensor with shape [max_sequence_length, num_sequences, sequence_width] to PaddingSequenceFunctor.

* Update to the newest codes.

* Rename the PaddingSequenceFunctor to PaddingLoDTensorFunctor and remove the computation of dimensions out of the functos.

* Support the build for multiple architectures at one cmake command (iOS).

* Update the documentations.

Recover the mistakenly deleted WARPCTC variable in cmake.

So that they can be found when paddle is built as a add_subdirectory project in another project.

Fix case-sensitiveness