LBFGS Logistic regression accelerated with OneDAL (builds against Random Forest PR #6364)

[rgesteve]

We are excited to review your PR.

So we can do the best job, please check:

• There's a descriptive title that will make sense to other developers some time from now.
• There's associated issues (N/A)
• Your change description explains what the change does, why you chose your approach, and anything else that reviewers should know. (please see below)
• You have included any necessary tests in the same PR.

Provides an accelerated (under certain circumstances) path to do logistic regression with the LBFGS optimizer. This assumes the algorithm is running on a suitable platform (non-ARM CPU) and is enabled by setting the environment variable MLNET_BACKEND to "ONEDAL" (a few extra checks are done against the shape of the dataset to ensure acceleration). Beyond this "knob", no change is required in user code.

This change builds on the infrastructure provided by PR #6364 in terms of OneDAL build and runtime support.

[rgesteve]

Closed in favor of #6521