Compilation: Roslyn, IL, JIT, tiered compilation, ReadyToRun and Native AOT — .NET Internals — CLR, CTS, CLS, IL, JIT, Memory Management & Garbage Collection

3. Compilation: Roslyn, IL, JIT, tiered compilation, ReadyToRun and Native AOT

Basic★ Must knowLesson 3 of 11

csharp

public static int Add(int a, int b) => a + b;

compiles to IL (stack-based, CPU-independent):

text

.method public hidebysig static int32 Add(int32 a, int32 b) cil managed
{
  ldarg.0      // push a
  ldarg.1      // push b
  add          // pop two, push sum
  ret          // return top of stack
}

and at runtime the JIT turns it into native code for the actual CPU:

text

; x64
lea eax, [rcx+rdx]
ret

Technique	When code becomes native	Pros	Cons
JIT (default)	At first call of each method	Optimises for the actual CPU and runtime profile	Startup cost ("warm-up")
Tiered compilation (.NET Core 3+)	Tier 0 quick JIT first, tier 1 optimised for hot methods (after ~30 calls), with Dynamic PGO (on by default since .NET 8)	Fast startup and fast steady state	More complex
ReadyToRun (R2R)	Precompiled native code inside the DLL; JIT can still re-optimise	Faster startup, still flexible	Larger files
Native AOT (.NET 7+)	Whole app compiled to a native executable at publish	Instant start, small memory, no JIT, no runtime install	No dynamic code loading, reflection limits (trimming), longer builds
NGen (.NET Framework)	Native images in a machine cache	Faster startup	Fragile, Framework-only

bash

dotnet publish -c Release -r linux-x64 -p:PublishReadyToRun=true
dotnet publish -c Release -r win-x64 -p:PublishAot=true