Add support for PowerPC VLE instruction set

[cryptwhoa]

This adds support for PowerPC VLE to the homegrown PPC (ie non-capstone) branch.

Note that some instructions in the sample ELFs (see below) seem undecoded; this is due to issue #6290, because we don't support figuring out what type of vector instructions are supported yet, so the ELF recognizer doesn't know to decode them as SPE instructions. I expect to add support for this in a future PR, but this is the VLE instruction set by itself. Note that at the time I started writing this (maybe it's been fixed since then? probably not though if #6290 is still alive), it's a limitation of PPC binaries in general.

There was some grossness in getting the auto-VLE detection to work in ELFs. The mechanism used to distinguish VLE instructions is through page bits in the MMU. ELFs denote VLE sections the section's flag value. Discussions on slack indicated that the best way forward was to pass data from the ELF view to a platform recognizer via Metadata. Unfortunately, this seemed to require duplicating the section parsing logic.

Note that binaryninja's ELF loader doesn't seem to have a clean way to set architectures of different sections. In theory, an ELF could have both VLE and non-VLE code sections in it, though I don't know how often these appear in practice. The auto-detection logic only sets the architecture to VLE if all of the executable sections have the VLE flag.

Currently, the PR focuses on big-endian 32-bit VLE. There are likely other architectures: VLE instruction decoding is always big endian, but the docs don't say anything about data endianness, so a little-endian 32-bit VLE is probably possible. There can also be 64-bit VLE, of big/little endian. When variants are added (SPE, Altivec, etc.), these architectures will start to balloon, for the whole matrix of:

• VLE vs. non-VLE
• 32-bit vs 64-bit
• big vs little endian
• SPE vs Altivec vs ???

so I'm punting on that complexity for #6290.

Datasheets:
VLEPEM.pdf (the main document for VLE)
VLE_addendum.pdf (documentation for a few more instructions)

Sample ELFs can be generated from a container here: https://github.com/AutomotiveDevOps/nxp-devkit-mpc57xx-docker. For convenience, here's a zipped up package of all the ELFs:
ppcvle_elfs.zip

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.

---

noone seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}