Pyqrack

Concurrent with the release of Qrack v.8.2.2, this release improves SIMD performance and disables `QEngine` under `QBdt` by default.

**Full Changelog**: https://github.com/unitaryfund/pyqrack/compare/v0.21.13...v1.4.2

In deciding to remove in_ket() and out_ket() from (Rust) "Qook," it became apparent that mirror circuit fidelity testing could still greatly benefit, in terms of reduced simulation overhead, from a method capable of returning single permutation basis eigenstate probabilities.

This release adds support for PermutationProb() in the Qrack shared library API, wrapped as prob_perm() in the PyQrack API, to allow single permutation basis eigenstate probability checks, (as well as probability checks of permutations of arbitrary qubit lists).

The environment variable `QRACK_QUNITMULTI_DEVICES` has been added, as basically a duplicate of the system that parses and logically switches `QRACK_QPAGER_DEVICES`, (_without_ `QRACK_QUNITMULTI_HOST_PTR`, for now or always). Unlike the `QPager` equivalent, `QUnitMulti` will override its environment variable device list by its constructor device list. (This might also ultimately make sense for `QPager`, analogously, but that change could be considered semantically "breaking.")

The logical environment variable interpretation of `QRACK_MAX_PAGE_QB` has been debugged in `QStabilizerHybrid`.

A race condition and a logical edge case were identified and fixed in `QPager`. Upon statistical testing, no indication of `QPager` "edge cases" remains, even under `QUnit`.

Semantically, the `QRACK_SEGMENT_GLOBAL_QB` environment variable has been replaced with `QRACK_MAX_PAGE_QB`. (Rather than specifying a common "segment" level across heterogeneous devices, it makes more sense to specify a common qubit count ceiling.)

This is a patch release for `QPager` debugging and optimization. The API has not (semantically) changed, but `QPager` passes sanity checks for default stack inclusion on `test_mirror_circuit` in the C++11 Qrack benchmark suite.