Remove unrelated changes

Update allowlist_for_publicAPI.json

Update on "Fuse row-wise sharded linear matmul to increase perf."

Instead of looping through and performing a matmul separately, we can
just perform a single matmul to ensure we launch a single cuda kernel for this
operation.

Differential Revision: [D36743354](https://our.internmc.facebook.com/intern/diff/D36743354/)

Differential Revision: [D36743354](https://our.internmc.facebook.com/intern/diff/D36743354)

[ghstack-poisoned]

Merge branch 'master' of https://github.com/pytorch/pytorch into lagu…

…erre-polynomial-l

review comments

review comments

Move THPStorage definitions out of `torch/csrc/generic`

Move THPStorage definitions out of `torch/csrc/generic`

[ROCm] TestGradients: Enable grad and gradgrad

Signed-off-by: Jagadish Krishnamoorthy <jagdish.krishna@gmail.com>

Shifted Chebyshev polynomial of the second kind