kern: fix minor assembly bugs, avoid unnecessary function call in KScheduler hotloop

kern: optimize and bring into line with N our pstate.i management

kern: audit (and fix) our hardware maintenance instructions to match official kernel

kern: implement 13.1.0 kernel changes

kern: simplify random bitmap selection to match latest Nintendo logic

kern: add (and use) generic KSystemControlBase

kern: add toggleable support for 40-bit physaddr caps

kern/test: add wip qemu-virt board support to mesosphere

kern: devirtualize KReadableEvent::Reset, KWorkerTask::DoWorkerTask

kern: enable lto for global instants (saves about a page in deduplicated code)

kern: devirtualize KAutoObjectWithList::GetId()

kern: fix reorder-init in KHandleTable

kern: devirtualize most things that are free to devirtualize (see #1672)

kern: use optimized red black tree finds for remaining holdouts

kern: devirtualize several KAutoObject functions

kern: devirtualize remaining vcall for class token/dyncast

kern: avoid constexpr init for many objects (avoids unnecessary memory clear) (#1668)

kern: use new AtomicRef, use Atomic<bool>

util: better match true std::atomic semantics

kern/util: use custom atomics wrapper to substantially improve codegen

kern: use 13.0.0 revised KLightLock logic

kern: remove spurious assertion

kern: update scheduler for 13.0.0 change, fix some ctz/clz bugs

kern: fix use of plr vs plr_heap, fix close/unlock order in ArbitrateLock

ams: improve offsetof style consistency

kern: remove need for explicit reserved member in KAutoObject

kern: devirtualize KAutoObject::DynamicCast<>() ...

This is an optimization that saves the most common type of virtual call in the kernel (DynamicCast)
by storing class token as a member, rather than getting it via virtual call every time.

This does not currently cost any memory space on 64-bit targets, due to pre-existing padding space.

This optimization can be turned off via a compile-time flag for accuracy.

kern/util: update structure layouts to match Nintendo (saves 0x10 per KThread/KSession)

kern: fix some lingering non-m_ member variables

kern: tidy up capability calc