New DimShuffle C-code fails on Windows

[ricardoV94]

This was first seen in https://github.com/pymc-devs/pymc/pull/5279

The following tests are failing on my Windows machine:

• test_elemwise.py::TestDimShuffle::test_infer_shape
• test_elemwise.py::TestDimShuffle::test_too_big_rank
• test_elemwise.py::TestDimShuffle::test_c_views

The first two tests which precede #701 pass before the relevant commit: https://github.com/aesara-devs/aesara/commit/e593b0ac57a0d56d4f6ffdd08d52c3be78ebf961 and fail after.

For sanity check, all tests in test_elemwise.py::TestBroadcast work fine in main.

Traceback

When running test_elemwise.py::TestDimShuffle::test_c_views:

C:\Users\ricar\miniconda3\envs\aesara-dev-custom\python.exe "C:\Program Files\JetBrains\PyCharm Community Edition 2020.1.1\plugins\python-ce\helpers\pycharm\_jb_pytest_runner.py" --target test_elemwise.py::TestDimShuffle.test_c_views
Launching pytest with arguments test_elemwise.py::TestDimShuffle::test_c_views in C:\Users\ricar\Documents\aesara\tests\tensor

============================= test session starts =============================
platform win32 -- Python 3.9.9, pytest-6.2.5, py-1.11.0, pluggy-1.0.0 -- C:\Users\ricar\miniconda3\envs\aesara-dev-custom\python.exe
cachedir: .pytest_cache
rootdir: C:\Users\ricar\Documents\aesara, configfile: setup.cfg
collecting ... collected 1 item

test_elemwise.py::TestDimShuffle::test_c_views Windows fatal exception: code 0xc0000374

Current thread 0x00002704 (most recent call first):
  File "C:\Users\ricar\Documents\aesara\aesara\link\c\basic.py", line 1747 in __call__
  File "C:\Users\ricar\Documents\aesara\tests\tensor\test_elemwise.py", line 135 in test_c_views
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\python.py", line 183 in pytest_pyfunc_call
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_callers.py", line 39 in _multicall
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_manager.py", line 80 in _hookexec
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_hooks.py", line 265 in __call__
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\python.py", line 1641 in runtest
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\runner.py", line 162 in pytest_runtest_call
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_callers.py", line 39 in _multicall
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_manager.py", line 80 in _hookexec
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_hooks.py", line 265 in __call__
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\runner.py", line 255 in <lambda>
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\runner.py", line 311 in from_call
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\runner.py", line 254 in call_runtest_hook
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\runner.py", line 215 in call_and_report
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\runner.py", line 126 in runtestprotocol
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\runner.py", line 109 in pytest_runtest_protocol
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_callers.py", line 39 in _multicall
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_manager.py", line 80 in _hookexec
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_hooks.py", line 265 in __call__
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\main.py", line 348 in pytest_runtestloop
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_callers.py", line 39 in _multicall
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_manager.py", line 80 in _hookexec
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_hooks.py", line 265 in __call__
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\main.py", line 323 in _main
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\main.py", line 269 in wrap_session
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\main.py", line 316 in pytest_cmdline_main
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_callers.py", line 39 in _multicall
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_manager.py", line 80 in _hookexec
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\pluggy\_hooks.py", line 265 in __call__
  File "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\lib\site-packages\_pytest\config\__init__.py", line 162 in main
  File "C:\Program Files\JetBrains\PyCharm Community Edition 2020.1.1\plugins\python-ce\helpers\pycharm\_jb_pytest_runner.py", line 43 in <module>

Process finished with exit code -1073740940 (0xC0000374)

Versions and main components

• Aesara version: main
• Aesara config (python -c "import aesara; print(aesara.config)")
• Python version: 3.9.9
• Operating system: Windows 10
• How did you install Aesara: used the following conda enviroment.yml:

name: aesara-dev-custom
channels:
- conda-forge
- defaults
dependencies:
 # base dependencies (see install guide for Windows)
- aesara=2.3.3
- pip
- python=3.9
# Extra stuff for dev, testing and docs build
- ipython
- pre-commit
- pytest

And then removed aesara with conda remove --force aesara to use the local branch

[ricardoV94]

Quick Google suggests it may be a reference count issue: https://stackoverflow.com/a/64960890

[brandonwillard]

What's the python -c "import aesara; print(aesara.config) output?

[ricardoV94]

What's the python -c "import aesara; print(aesara.config) output?

``` local floatX ({'float64', 'float16', 'float32'}) Doc: Default floating-point precision for python casts. Note: float16 support is experimental, use at your own risk. Value: float64 warn_float64 ({'warn', 'raise', 'pdb', 'ignore'}) Doc: Do an action when a tensor variable with float64 dtype is created. They can't be run on the GPU with the current(old) gpu back-end and are slow with gamer GPUs. Value: ignore pickle_test_value (>) Doc: Dump test values while pickling model. If True, test values will be dumped with model. Value: True cast_policy ({'custom', 'numpy+floatX'}) Doc: Rules for implicit type casting Value: custom deterministic ({'more', 'default'}) Doc: If `more`, sometimes we will select some implementation that are more deterministic, but slower. In particular, on the GPU, we will avoid using AtomicAdd. Sometimes we will still use non-deterministic implementation, e.g. when we do not have a GPU implementation that is deterministic. Also see the dnn.conv.algo* flags to cover more cases. Value: default device (cpu, opencl*, cuda*) Doc: Default device for computations. If cuda* or opencl*, change thedefault to try to move computation to the GPU. Do not use upper caseletters, only lower case even if NVIDIA uses capital letters. 'gpu' means let the driver select the gpu (needed for gpu in exclusive mode). 'gpuX' mean use the gpu number X. Value: cpu init_gpu_device (, opencl*, cuda*) Doc: Initialize the gpu device to use, works only if device=cpu. Unlike 'device', setting this option will NOT move computations, nor shared variables, to the specified GPU. It can be used to run GPU-specific tests on a particular GPU. Value: force_device (>) Doc: Raise an error if we can't use the specified device Value: False conv__assert_shape (>) Doc: If True, AbstractConv* ops will verify that user-provided shapes match the runtime shapes (debugging option, may slow down compilation) Value: False print_global_stats (>) Doc: Print some global statistics (time spent) at the end Value: False Doc: Context map for multi-gpu operation. Format is a semicolon-separated list of names and device names in the 'name->dev_name' format. An example that would map name 'test' to device 'cuda0' and name 'test2' to device 'opencl0:0' follows: "test->cuda0;test2->opencl0:0". Invalid context names are 'cpu', 'cuda*' and 'opencl*' Value: print_active_device (>) Doc: Print active device at when the GPU device is initialized. Value: True gpuarray__preallocate () Doc: If negative it disables the allocation cache. If between 0 and 1 it enables the allocation cache and preallocates that fraction of the total GPU memory. If 1 or greater it will preallocate that amount of memory (in megabytes). Value: 0.0 gpuarray__sched ({'single', 'multi', 'default'}) Doc: The sched parameter passed for context creation to pygpu. With CUDA, using "multi" is equivalent to using the parameter cudaDeviceScheduleBlockingSync. This is useful to lower the CPU overhead when waiting for GPU. One user found that it speeds up his other processes that was doing data augmentation. Value: default gpuarray__single_stream (>) Doc: If your computations are mostly lots of small elements, using single-stream will avoid the synchronization overhead and usually be faster. For larger elements it does not make a difference yet. In the future when true multi-stream is enabled in libgpuarray, this may change. If you want to make sure to have optimal performance, check both options. Value: True cuda__root () Doc: Location of the cuda installation Value: cuda__include_path () Doc: Location of the cuda includes Value: assert_no_cpu_op ({'warn', 'raise', 'pdb', 'ignore'}) Doc: Raise an error/warning if there is a CPU op in the computational graph. Value: ignore unpickle_function (>) Doc: Replace unpickled Aesara functions with None. This is useful to unpickle old graphs that pickled them when it shouldn't Value: True reoptimize_unpickled_function (>) Doc: Re-optimize the graph when an Aesara function is unpickled from the disk. Value: False dnn__conv__algo_fwd ({'winograd', 'fft_tiling', 'none', 'guess_on_shape_change', 'small', 'time_on_shape_change', 'large', 'guess_once', 'winograd_non_fused', 'time_once', 'fft'}) Doc: Default implementation to use for cuDNN forward convolution. Value: small dnn__conv__algo_bwd_data ({'winograd', 'fft_tiling', 'deterministic', 'none', 'guess_on_shape_change', 'time_on_shape_change', 'guess_once', 'winograd_non_fused', 'time_once', 'fft'}) Doc: Default implementation to use for cuDNN backward convolution to get the gradients of the convolution with regard to the inputs. Value: none dnn__conv__algo_bwd_filter ({'fft_tiling', 'deterministic', 'none', 'small', 'guess_on_shape_change', 'time_on_shape_change', 'guess_once', 'winograd_non_fused', 'time_once', 'fft'}) Doc: Default implementation to use for cuDNN backward convolution to get the gradients of the convolution with regard to the filters. Value: none dnn__conv__precision ({'float32', 'float16', 'float64', 'as_input_f32', 'as_input'}) Doc: Default data precision to use for the computation in cuDNN convolutions (defaults to the same dtype as the inputs of the convolutions, or float32 if inputs are float16). Value: as_input_f32 dnn__base_path () Doc: Install location of cuDNN. Value: dnn__include_path () Doc: Location of the cudnn header Value: dnn__library_path () Doc: Location of the cudnn link library. Value: dnn__bin_path () Doc: Location of the cuDNN load library (on non-windows platforms, this is the same as dnn__library_path) Value: dnn__enabled ({'no_check', 'False', 'True', 'auto'}) Doc: 'auto', use cuDNN if available, but silently fall back to not using it if not present. If True and cuDNN can not be used, raise an error. If False, disable cudnn even if present. If no_check, assume present and the version between header and library match (so less compilation at context init) Value: auto magma__include_path () Doc: Location of the magma header Value: magma__library_path () Doc: Location of the magma library Value: magma__enabled (>) Doc: If True, use magma for matrix computation. If False, disable magma Value: False Doc: Default compilation mode Value: Mode cxx () Doc: The C++ compiler to use. Currently only g++ is supported, but supporting additional compilers should not be too difficult. If it is empty, no C++ code is compiled. Value: "C:\Users\ricar\miniconda3\envs\aesara-dev-custom\Library\mingw-w64\bin\g++.exe" linker ({'c', 'py', 'cvm', 'c|py_nogc', 'cvm_nogc', 'vm', 'vm_nogc', 'c|py'}) Doc: Default linker used if the aesara flags mode is Mode Value: cvm allow_gc (>) Doc: Do we default to delete intermediate results during Aesara function calls? Doing so lowers the memory requirement, but asks that we reallocate memory at the next function call. This is implemented for the default linker, but may not work for all linkers. Value: True optimizer ({'o2', 'o4', 'o3', 'merge', 'None', 'fast_compile', 'unsafe', 'o1', 'fast_run'}) Doc: Default optimizer. If not None, will use this optimizer with the Mode Value: o4 optimizer_verbose (>) Doc: If True, we print all optimization being applied Value: False on_opt_error ({'warn', 'raise', 'pdb', 'ignore'}) Doc: What to do when an optimization crashes: warn and skip it, raise the exception, or fall into the pdb debugger. Value: warn nocleanup (>) Doc: Suppress the deletion of code files that did not compile cleanly Value: False on_unused_input ({'warn', 'raise', 'ignore'}) Doc: What to do if a variable in the 'inputs' list of aesara.function() is not used in the graph. Value: raise gcc__cxxflags () Doc: Extra compiler flags for gcc Value: cmodule__warn_no_version (>) Doc: If True, will print a warning when compiling one or more Op with C code that can't be cached because there is no c_code_cache_version() function associated to at least one of those Ops. Value: False cmodule__remove_gxx_opt (>) Doc: If True, will remove the -O* parameter passed to g++.This is useful to debug in gdb modules compiled by Aesara.The parameter -g is passed by default to g++ Value: False cmodule__compilation_warning (>) Doc: If True, will print compilation warnings. Value: False cmodule__preload_cache (>) Doc: If set to True, will preload the C module cache at import time Value: False cmodule__age_thresh_use () Doc: In seconds. The time after which Aesara won't reuse a compile c module. Value: 2073600 cmodule__debug (>) Doc: If True, define a DEBUG macro (if not exists) for any compiled C code. Value: False compile__wait () Doc: Time to wait before retrying to acquire the compile lock. Value: 5 compile__timeout () Doc: In seconds, time that a process will wait before deciding to override an existing lock. An override only happens when the existing lock is held by the same owner *and* has not been 'refreshed' by this owner for more than this period. Refreshes are done every half timeout period for running processes. Value: 120 ctc__root () Doc: Directory which contains the root of Baidu CTC library. It is assumed that the compiled library is either inside the build, lib or lib64 subdirectory, and the header inside the include directory. Value: tensor__cmp_sloppy () Doc: Relax aesara.tensor.math._allclose (0) not at all, (1) a bit, (2) more Value: 0 tensor__local_elemwise_fusion (>) Doc: Enable or not in fast_run mode(fast_run optimization) the elemwise fusion optimization Value: True lib__amblibm (>) Doc: Use amd's amdlibm numerical library Value: False tensor__insert_inplace_optimizer_validate_nb () Doc: -1: auto, if graph have less then 500 nodes 1, else 10 Value: -1 traceback__limit () Doc: The number of stack to trace. -1 mean all. Value: 8 traceback__compile_limit () Doc: The number of stack to trace to keep during compilation. -1 mean all. If greater then 0, will also make us save Aesara internal stack trace. Value: 0 experimental__unpickle_gpu_on_cpu (>) Doc: Allow unpickling of pickled GpuArrays as numpy.ndarrays.This is useful, if you want to open a GpuArray without having cuda installed.If you have cuda installed, this will force unpickling tobe done on the cpu to numpy.ndarray.Please be aware that this may get you access to the data,however, trying to unpicke gpu functions will not succeed.This flag is experimental and may be removed any time, whengpu<>cpu transparency is solved. Value: False experimental__local_alloc_elemwise (>) Doc: DEPRECATED: If True, enable the experimental optimization local_alloc_elemwise. Generates error if not True. Use optimizer_excluding=local_alloc_elemwise to disable. Value: True experimental__local_alloc_elemwise_assert (>) Doc: When the local_alloc_elemwise is applied, add an assert to highlight shape errors. Value: True warn__ignore_bug_before ({'1.0', '0.9', '0.4', 'None', '0.8.2', '0.3', '1.0.3', '0.5', 'all', '1.0.2', '0.10', '0.7', '0.4.1', '0.6', '0.8.1', '1.0.4', '1.0.1', '1.0.5', '0.8'}) Doc: If 'None', we warn about all Aesara bugs found by default. If 'all', we don't warn about Aesara bugs found by default. If a version, we print only the warnings relative to Aesara bugs found after that version. Warning for specific bugs can be configured with specific [warn] flags. Value: 0.9 exception_verbosity ({'high', 'low'}) Doc: If 'low', the text of exceptions will generally refer to apply nodes with short names such as Elemwise{add_no_inplace}. If 'high', some exceptions will also refer to apply nodes with long descriptions like: A. Elemwise{add_no_inplace} B. log_likelihood_v_given_h C. log_likelihood_h Value: low print_test_value (>) Doc: If 'True', the __eval__ of an Aesara variable will return its test_value when this is available. This has the practical conseguence that, e.g., in debugging `my_var` will print the same as `my_var.tag.test_value` when a test value is defined. Value: False compute_test_value ({'raise', 'ignore', 'off', 'pdb', 'warn'}) Doc: If 'True', Aesara will run each op at graph build time, using Constants, SharedVariables and the tag 'test_value' as inputs to the function. This helps the user track down problems in the graph before it gets optimized. Value: off compute_test_value_opt ({'raise', 'ignore', 'off', 'pdb', 'warn'}) Doc: For debugging Aesara optimization only. Same as compute_test_value, but is used during Aesara optimization Value: off check_input (>) Doc: Specify if types should check their input in their C code. It can be used to speed up compilation, reduce overhead (particularly for scalars) and reduce the number of generated C files. Value: True NanGuardMode__nan_is_error (>) Doc: Default value for nan_is_error Value: True NanGuardMode__inf_is_error (>) Doc: Default value for inf_is_error Value: True NanGuardMode__big_is_error (>) Doc: Default value for big_is_error Value: True NanGuardMode__action ({'warn', 'raise', 'pdb'}) Doc: What NanGuardMode does when it finds a problem Value: raise DebugMode__patience () Doc: Optimize graph this many times to detect inconsistency Value: 10 DebugMode__check_c (>) Doc: Run C implementations where possible Value: True DebugMode__check_py (>) Doc: Run Python implementations where possible Value: True DebugMode__check_finite (>) Doc: True -> complain about NaN/Inf results Value: True DebugMode__check_strides () Doc: Check that Python- and C-produced ndarrays have same strides. On difference: (0) - ignore, (1) warn, or (2) raise error Value: 0 DebugMode__warn_input_not_reused (>) Doc: Generate a warning when destroy_map or view_map says that an op works inplace, but the op did not reuse the input for its output. Value: True DebugMode__check_preallocated_output () Doc: Test thunks with pre-allocated memory as output storage. This is a list of strings separated by ":". Valid values are: "initial" (initial storage in storage map, happens with Scan),"previous" (previously-returned memory), "c_contiguous", "f_contiguous", "strided" (positive and negative strides), "wrong_size" (larger and smaller dimensions), and "ALL" (all of the above). Value: DebugMode__check_preallocated_output_ndim () Doc: When testing with "strided" preallocated output memory, test all combinations of strides over that number of (inner-most) dimensions. You may want to reduce that number to reduce memory or time usage, but it is advised to keep a minimum of 2. Value: 4 profiling__time_thunks (>) Doc: Time individual thunks when profiling Value: True profiling__n_apply () Doc: Number of Apply instances to print by default Value: 20 profiling__n_ops () Doc: Number of Ops to print by default Value: 20 profiling__output_line_width () Doc: Max line width for the profiling output Value: 512 profiling__min_memory_size () Doc: For the memory profile, do not print Apply nodes if the size of their outputs (in bytes) is lower than this threshold Value: 1024 profiling__min_peak_memory (>) Doc: The min peak memory usage of the order Value: False profiling__destination () Doc: File destination of the profiling output Value: stderr profiling__debugprint (>) Doc: Do a debugprint of the profiled functions Value: False profiling__ignore_first_call (>) Doc: Do we ignore the first call of an Aesara function. Value: False on_shape_error ({'warn', 'raise'}) Doc: warn: print a warning and use the default value. raise: raise an error Value: warn openmp (>) Doc: Allow (or not) parallel computation on the CPU with OpenMP. This is the default value used when creating an Op that supports OpenMP parallelization. It is preferable to define it via the Aesara configuration file ~/.aesararc or with the environment variable AESARA_FLAGS. Parallelization is only done for some operations that implement it, and even for operations that implement parallelism, each operation is free to respect this flag or not. You can control the number of threads used with the environment variable OMP_NUM_THREADS. If it is set to 1, we disable openmp in Aesara by default. Value: False openmp_elemwise_minsize () Doc: If OpenMP is enabled, this is the minimum size of vectors for which the openmp parallelization is enabled in element wise ops. Value: 200000 optimizer_excluding () Doc: When using the default mode, we will remove optimizer with these tags. Separate tags with ':'. Value: optimizer_including () Doc: When using the default mode, we will add optimizer with these tags. Separate tags with ':'. Value: optimizer_requiring () Doc: When using the default mode, we will require optimizer with these tags. Separate tags with ':'. Value: optdb__position_cutoff () Doc: Where to stop eariler during optimization. It represent the position of the optimizer where to stop. Value: inf optdb__max_use_ratio () Doc: A ratio that prevent infinite loop in EquilibriumOptimizer. Value: 8.0 cycle_detection ({'regular', 'fast'}) Doc: If cycle_detection is set to regular, most inplaces are allowed,but it is slower. If cycle_detection is set to faster, less inplacesare allowed, but it makes the compilation faster.The interaction of which one give the lower peak memory usage iscomplicated and not predictable, so if you are close to the peakmemory usage, triyng both could give you a small gain. Value: regular check_stack_trace ({'off', 'raise', 'warn', 'log'}) Doc: A flag for checking the stack trace during the optimization process. default (off): does not check the stack trace of any optimization log: inserts a dummy stack trace that identifies the optimizationthat inserted the variable that had an empty stack trace.warn: prints a warning if a stack trace is missing and also a dummystack trace is inserted that indicates which optimization insertedthe variable that had an empty stack trace.raise: raises an exception if a stack trace is missing Value: off metaopt__verbose () Doc: 0 for silent, 1 for only warnings, 2 for full output withtimings and selected implementation Value: 0 metaopt__optimizer_excluding () Doc: exclude optimizers with these tags. Separate tags with ':'. Value: metaopt__optimizer_including () Doc: include optimizers with these tags. Separate tags with ':'. Value: profile (>) Doc: If VM should collect profile information Value: False profile_optimizer (>) Doc: If VM should collect optimizer profile information Value: False profile_memory (>) Doc: If VM should collect memory profile information and print it Value: False Doc: Useful only for the vm linkers. When lazy is None, auto detect if lazy evaluation is needed and use the appropriate version. If lazy is True/False, force the version used between Loop/LoopGC and Stack. Value: None cache_optimizations (>) Doc: WARNING: work in progress, does not work yet. Specify if the optimization cache should be used. This cache will any optimized graph and its optimization. Actually slow downs a lot the first optimization, and could possibly still contains some bugs. Use at your own risks. Value: False unittests__rseed () Doc: Seed to use for randomized unit tests. Special value 'random' means using a seed of None. Value: 666 warn__round (>) Doc: Warn when using `tensor.round` with the default mode. Round changed its default from `half_away_from_zero` to `half_to_even` to have the same default as NumPy. Value: False compiledir_format () Doc: Format string for platform-dependent compiled module subdirectory (relative to base_compiledir). Available keys: aesara_version, device, gxx_version, hostname, numpy_version, platform, processor, python_bitwidth, python_int_bitwidth, python_version, short_platform. Defaults to compiledir_%(short_platform)s-%(processor)s-%(python_versi on)s-%(python_bitwidth)s. Value: compiledir_%(short_platform)s-%(processor)s-%(python_version)s-%(python_bitwidth)s Doc: platform-independent root directory for compiled modules Value: C:\Users\ricar\AppData\Local\Aesara Doc: platform-dependent cache directory for compiled modules Value: C:\Users\ricar\AppData\Local\Aesara\compiledir_Windows-10-10.0.19041-SP0-AMD64_Family_23_Model_17_Stepping_0_AuthenticAMD-3.9.9-64 Doc: Directory to cache pre-compiled kernels for the gpuarray backend. Value: C:\Users\ricar\AppData\Local\Aesara\compiledir_Windows-10-10.0.19041-SP0-AMD64_Family_23_Model_17_Stepping_0_AuthenticAMD-3.9.9-64\gpuarray_kernels blas__ldflags () Doc: lib[s] to include for [Fortran] level-3 blas implementation Value: blas__check_openmp (>) Doc: Check for openmp library conflict. WARNING: Setting this to False leaves you open to wrong results in blas-related operations. Value: True scan__allow_gc (>) Doc: Allow/disallow gc inside of Scan (default: False) Value: False scan__allow_output_prealloc (>) Doc: Allow/disallow memory preallocation for outputs inside of scan (default: True) Value: True ```

[twiecki]

Any ideas on a path forward here?

[brandonwillard]

Any ideas on a path forward here?

Someone with a good Windows development setup that can reproduce the issue needs to start debugging it. My first assumption is that this is just another reference count problem.

Hopefully, there's just a bug in the new implementation and one can find it by manually tracking the reference counts (e.g. print them all throughout the DimShuffle C code and look for cases where the count is 0 and the variable is still being actively used or passed off to be used). The fix would then be the addition of strategically placed Py_INCREFs and/or Py_DECREFs.

Worst case, the issue could be caused by a CPython version/implementation discrepancy and the above might fix things for Windows but break things in Linux (or introduce a memory leak).

Regardless, someone needs to do some simple debugging (and not forget to aesara-cache clear between changes).

[brandonwillard]

The issue seems to be the use of PyDimMem_FREE here. Changing it to free(reshape_shape.ptr) or free(_reshape_shape) fixes the issue.

[twiecki]

:+1: how painful was that to find?

[brandonwillard]

+1 how painful was that to find?

It took literally five minutes to find it after about an hour of building a Windows VM, setting up a dev environment, finding out how nearly impossible it is to get gdb working with a Conda m2w64-toolchain setup, etc., etc.