Make it easier to test whether functionality was built

[ezyang]

I am a library and I want to make use of an nnpack installation that I have found on a user's system. Can I use it? Well, it depends: nnpack has --convolution-only and --inference-only, which will remove certain symbols from the built objects. Right now, the only way to test if this has occurred is to, configure-style, see if you can actually link against the symbol or question. It would be nice if there were an easier way to test for this. A few possibilities:

• If the installed nnpack.h was adjusted to only contain the actual definitions that actually made it into the objects, you'd get a more understandable "nnp_blah not in scope" rather than a linking error way later in the build process.
• Add an autogenerated header file that defines preprocessor names for all things that were actually built

Even better: don't let users control what symbols are available by configuration options; or split nnpack into separate libraries, one per logical unit of functionality.

Real case of this happening: https://github.com/pytorch/pytorch/issues/4336

[Maratyszcza]

@ezyang --convolution-only and --inference-only are intended to reduce size for mobile builds, and are off by default. I assume that users who explicitly enable these options understand what they are doing.

Unfortunately, neither option that you suggested is easy to implement:

• Generating headers is currently not supported in confu, the configuration system I use in NNPACK development.
• Removing --convolution-only and --inference-only options would incur a significant increase in size of mobile builds.
• Splitting NNPACK into multiple libraries would be a maintenance nightmare because inference/training functions still share a lot of low-level code.

For PyTorch, I would suggest to build NNPACK (using its CMake scripts) by default, and only use user-installed version if explicitly instructed to do so.

[ezyang]

OK, well, we ended up fixing this in PyTorch's build system using check_symbol_exists(nnp_convolution_kernel_gradient "nnpack.h" NNPACK_HAS_INFERENCE) to check if the symbol actually existed, so maybe if there is ever an FAQ written for this sort of thing, we could add this there.

[Maratyszcza]

@ezyang The variable name is confusing:

• NNPACK always includes inference function for convolutional layers (nnp_convolution_inference)
• --inference-only disables building functions which are only used for training
• --convolution-only disables building functions for non-convolutional layers (i.e. ReLU, SoftMax, FC, MaxPool)

Thus, it would make more sense to do check_symbol_exists(nnp_convolution_kernel_gradient "nnpack.h" NNPACK_HAS_TRAINING)