Add `run_exports` to `cudnn` about aggregate HOT 6 CLOSED

Copying this from the developer documentation (link below):

cuDNN API Compatibility

Beginning in cuDNN 7, the binary compatibility of patch and minor releases is maintained as follows:

• Any patch release x.y.z is forward- or backward-compatible with applications built against another cuDNN patch release x.y.w (i.e., of the same major and minor version number, but having w!=z)
• cuDNN minor releases beginning with cuDNN 7 are binary backward-compatible with applications built against the same or earlier patch release (i.e., an app built against cuDNN 7.x is binary compatible with cuDNN library 7.y, where y>=x)
• Applications compiled with a cuDNN version 7.y are not guaranteed to work with 7.x release when y > x.

ref: https://docs.nvidia.com/deeplearning/sdk/cudnn-developer-guide/index.html#programming-model

from aggregate.

FWIW we have added a global pinning of cudnn in conda-forge to address this issue for now. ( conda-forge/conda-forge-pinning-feedstock#310 )

from aggregate.

I agree, the cudnn package should have a run_exports. The package is due for an update to version 7.6.4. I'll add a note to myself to add run_exports when the package is updated.

from aggregate.

That would be great. Thanks Jonathan! 😄

from aggregate.

cudnn 7.6.4 packages are now available in defaults for win-64 and linux-64 and include a run_export for cudnn >=7.6,<8.0a0.

from aggregate.

cc @mike-wendt @raydouglass @kkraus14 (for awareness)

from aggregate.