Avoid cctor check if we provably executed it

[MichalStrehovsky]

For this program and the S method:

public class C2
{
    [MethodImpl(MethodImplOptions.NoInlining)]
    public string S() => d + s;

    public static readonly string s;

    private string d = "d";

    static C2() => s = Helpers.NoInline("s");
}

Before:

G_M34299_IG01:  ;; offset=0x0000
       push     rbx
       sub      rsp, 32
                                                ;; size=5 bbWeight=1 PerfScore 1.25
G_M34299_IG02:  ;; offset=0x0005
       mov      rbx, gword ptr [rcx+0x08]
       lea      rax, [(reloc 0x420b68)]
       cmp      qword ptr [rax-0x08], 0
       jne      SHORT G_M34299_IG05
                                                ;; size=18 bbWeight=1 PerfScore 5.50
G_M34299_IG03:  ;; offset=0x0017
       mov      rdx, qword ptr [(reloc 0x425630)]
       mov      rdx, gword ptr [rdx+0x08]
       mov      rcx, rbx
       call     System.String:Concat(System.String,System.String):System.String
       nop
                                                ;; size=20 bbWeight=1 PerfScore 5.50
G_M34299_IG04:  ;; offset=0x002B
       add      rsp, 32
       pop      rbx
       ret
                                                ;; size=6 bbWeight=1 PerfScore 1.75
G_M34299_IG05:  ;; offset=0x0031
       call     CORINFO_HELP_READYTORUN_GCSTATIC_BASE
       jmp      SHORT G_M34299_IG03
                                                ;; size=7 bbWeight=0 PerfScore 0.00

After:

       sub      rsp, 40
                                                ;; size=4 bbWeight=1 PerfScore 0.25
G_M34299_IG02:  ;; offset=0x0004
       mov      rcx, gword ptr [rcx+0x08]
       mov      rdx, qword ptr [(reloc 0x425608)]
       mov      rdx, gword ptr [rdx+0x08]
       call     System.String:Concat(System.String,System.String):System.String
       nop
                                                ;; size=21 bbWeight=1 PerfScore 7.25
G_M34299_IG03:  ;; offset=0x0019
       add      rsp, 40
       ret

[azure-pipelines]

Azure Pipelines:
Successfully started running 3 pipeline(s).
13 pipeline(s) were filtered out due to trigger conditions.
There may be pipelines that require an authorized user to comment /azp run to run.

[dotnet-policy-service]

Tagging subscribers to this area: @agocke, @dotnet/ilc-contrib
See info in area-owners.md if you want to be subscribed.

[Copilot]

Pull request overview

This PR updates the AOT RyuJit JIT-EE interface to use initClass analysis (rather than a coarse “has lazy cctor” check) when reporting static field access requirements, enabling the JIT to elide redundant class-constructor checks when initialization is already provably satisfied (e.g., instance-method cases).

Changes:

• Refactors initClass in the shared JitInterface to add a FieldDesc/MethodDesc overload and a handle-based wrapper.
• Updates ILCompiler.RyuJit getFieldInfo to consult initClass and (a) set CORINFO_FLG_FIELD_INITCLASS only when needed and (b) choose relocatable static base access when init is not required.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
src/coreclr/tools/Common/JitInterface/CorInfoImpl.cs	Adds an overload to run initClass logic directly on FieldDesc/MethodDesc, with a wrapper for handle-based entrypoints.
src/coreclr/tools/aot/ILCompiler.RyuJit/JitInterface/CorInfoImpl.RyuJit.cs	Uses initClass results in getFieldInfo to avoid unnecessary cctor triggering / helper usage for static field accesses.

[MichalPetryka]

Worth noting that this can change behaviour for calling instance methods with null this, but ECMA explicitly permits skipping cctor then (still there could be code relying on it being executed).

[hez2010]

Worth noting that this can change behaviour for calling instance methods with null this, but ECMA explicitly permits skipping cctor then (still there could be code relying on it being executed).

CoreCLR already skips it and calling instance methods with a null this seems like a scenario that we don't want to support at the first place. So I don't see any problem with this optimization.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

[MichalStrehovsky]

/azp run runtime-nativeaot-outerloop

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).

[jkotas]

Nice size savings MichalStrehovsky/rt-sz#254 !

[MichalStrehovsky]

/azp run runtime-nativeaot-outerloop

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

[MichalStrehovsky]

/azp run runtime-nativeaot-outerloop

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).

[MichalStrehovsky]

Nice size savings MichalStrehovsky/rt-sz#254 !

Sigh, too bad this missed .NET 11. As usual, .NET 12 will already be the best .NET by the time .NET 11 ships.

This is now blocked on #132676 since we're hitting JIT assert in outerloop

[MichalStrehovsky]

/azp run runtime-nativeaot-outerloop

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

[MichalStrehovsky]

/ba-g networking errors to the tune of #130458 and #126867 that apparently nobody cares about