- Correct 'multiply' to 'multiple' in data alignment comment.
- Correct 'aligment' to 'alignment' in data alignment comment.
- Fix typo 'in in' to 'in' in DLManagedTensorVersioned comment.

Since it was deleted in dmlc/dmlc-core#693

…ng (#174)

* [RFC] Support DLPACK C Functions for Speed Exchange and Stream Handling

This PR adds support for three C functions to speedup DLPack exchange.
As of now, DLPack exchange relies on python functions such as tensor.__dlpack__().

While they works well for common cases, the general overhead of such exchange is
at the level of 0.2-0.3 us for very well optimized version, and can go up to
0.4-1 us for less optimized implementation.

For a function that takes three arguments f(a, b, c), assume we run DLPack
exchange for each argument, the general conversion overhead usually gets to
around 1us and sometimes to 3us.

While such overhead can be acceptable in many settings, in GPU applications
the extra 1-3us overhead can still be significant.

This PR proposes three functions for speed exchange DLPack tensors without
going through python interpreter.

- DLPackFromPyObject: exports a PyObject Tensor to DLManagedTensorVesioned
- DLPackToPyObject: DLManagedTensorVesioned converts to a PyObject Tensor
- DLPackTensorAllocator: Used to expose one package's tensor allocator to another package
  - This allows for example we implement libraries that allocates intermediate tensor
    based on the caller's specified Tensor Allocator.

Our preliminary results show that these functions, when incorporated correctly
via native extensions such as c/c++, can bring exchange cost to the level of
30ns - 80ns, giving us about one order of maginitude speedup. That means functions
like f(a, b, c) can finish at 0.2us-0.4us level, which is close to what native
cpp extension overhead do without exchange.

* Incorporate suggestions to use a global table
for all function pointers

* Update naming and specification according to feedbacks

* Add maximum version parameter

* Update to include WorkStream proposal, move version to API chain

* Update based on latest feedback

* Incorporate feedbacks from sberg

Co-authored-by: Sebastian Berg <sebastianb@nvidia.com>

* Update include/dlpack/dlpack.h

Co-authored-by: Sebastian Berg <sebastian@sipsolutions.net>

* Update include/dlpack/dlpack.h

Co-authored-by: Sebastian Berg <sebastian@sipsolutions.net>

* Move exchange api to common header

* Update include/dlpack/dlpack.h

Co-authored-by: Lisandro Dalcin <dalcinl@gmail.com>

---------

Co-authored-by: Sebastian Berg <sebastianb@nvidia.com>
Co-authored-by: Sebastian Berg <sebastian@sipsolutions.net>
Co-authored-by: Lisandro Dalcin <dalcinl@gmail.com>

* Enforce strides to be not null when ndim is nonzero

This PR updates the spec to enforce the strides to be not null
when the ndim is nonzero.
This is going to help consumers to handle cases more uniformly.

* Add more comments on the null rationale

* Remove mention of framework handling

* Update NULL note, bump minor