GPU Manager: Implement ReadBlockUnsafe and WriteBlockUnsafe

Implement Kepler Memory on both Linear and BlockLinear.

Do some corrections in conversion shader instructions.

shader_ir/decode: Miscellaneous fixes to half-float decompilation

kernel/wait_object: Make GetHighestPriorityReadyThread() a const member function

yuzu/bootmanager: Resolve constructor initializer list warnings

kernel/svc: Name supervisor call 0x36

core: Reorganize boot order

This call was added to the SVC handlers in the 8.0.0 kernel, so we can
finally give it a name.

kernel/thread: Remove unused guest_handle member variable

gl_sampler_cache: Port sampler cache to OpenGL

Implement Bindless Textures on Shader Decompiler and GL backend

This doesn't actually modify internal state of a wait object, so it can
be const qualified.

The default constructor will always run, even when not specified, so
this is redundant.

However, the context member can indeed be initialized in the constructor
initializer list.

This include isn't used anymore so it can be removed.

Resolves -Wreorder warnings. These will automatically be initialized to
nullptr anyways, so these were redundant.

shader_ir/decode: Reduce the severity of common assertions

gl_rasterizer: Apply just the needed state on Clear

CMakeLists: Define QT_USE_QSTRINGBUILDER for the Qt target

shader/memory: Implement STG and global memory flushing

yuzu/configure_hotkey: Minor changes

common/{lz4_compression, zstd_compression}: Add missing header guards

This detects when a GPU Memory Block is not continous within host cpu
memory.

Use ReadBlockUnsafe on TIC and TSC reading as memory is never flushed
from host GPU there.

service: Update service function tables

kernel/svc: Implement svcMapProcessCodeMemory/svcUnmapProcessCodeMemory

kernel/thread: Remove BoostPriority()

Correct Pitch in Fermi2D

Operations done before the main half float operation (like HAdd) were
managing a packed value instead of the unpacked one. Adding an unpacked
operation allows us to drop the per-operand MetaHalfArithmetic entry,
simplifying the code overall.

GLSL decompilation for HMergeH0 was wrong. This addresses that issue.