Rework and simplify initial versioning and tiering rules by davidwrighton · Pull Request #125243 · dotnet/runtime

davidwrighton · 2026-03-05T22:15:16Z

Remove CallCountingInfo::Disabled stage and associated methods (DisableCallCounting, IsCallCountingEnabled, CreateWithCallCountingDisabled)
Move initial optimization tier determination from per-method JIT-time logic to a config-time decision (TieredCompilation_DefaultTier) computed once during EEConfig::sync()
Store the default code version's optimization tier directly on MethodDescCodeData, replacing the indirect approach of disabling call counting to signal optimized tier
Remove WasTieringDisabledBeforeJitting flag from PrepareCodeConfig and MethodDescVersioningState
Simplify GetJitFlags by removing the special default-version fast path and using the unified optimization tier switch for all code versions
Add OptimizationTierUnknown sentinel to distinguish uninitialized state
Simplify FinalizeOptimizationTierForTier0Load to handle R2R loading based on the stored optimization tier

- Remove CallCountingInfo::Disabled stage and associated methods (DisableCallCounting, IsCallCountingEnabled, CreateWithCallCountingDisabled) - Move initial optimization tier determination from per-method JIT-time logic to a config-time decision (TieredCompilation_DefaultTier) computed once during EEConfig::sync() - Store the default code version's optimization tier directly on MethodDescCodeData, replacing the indirect approach of disabling call counting to signal optimized tier - Remove WasTieringDisabledBeforeJitting flag from PrepareCodeConfig and MethodDescVersioningState - Simplify GetJitFlags by removing the special default-version fast path and using the unified optimization tier switch for all code versions - Add OptimizationTierUnknown sentinel to distinguish uninitialized state - Simplify FinalizeOptimizationTierForTier0Load to handle R2R loading based on the stored optimization tier Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

dotnet-policy-service · 2026-03-05T22:16:34Z

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

Copilot

Pull request overview

This PR refactors CoreCLR tiered compilation’s initial tier selection and tiering signaling by moving “default tier” decisions into EEConfig::sync(), removing the call-counting “disabled” signaling path, and storing the default version tier on MethodDesc code data.

Changes:

Compute and expose a single TieredCompilation_DefaultTier value in EEConfig, and use it for initial tier selection.
Add OptimizationTierUnknown and a per-MethodDesc stored optimization tier, and route default-version tier reads/writes through it.
Remove the call-counting “Disabled” stage and related APIs/flags, simplifying tier-related logic in the JIT flag path and prestub tier finalization.

Reviewed changes

Copilot reviewed 10 out of 10 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
src/coreclr/vm/tieredcompilation.cpp	Switch initial tier selection to use `EEConfig`’s computed default tier.
src/coreclr/vm/prestub.cpp	Remove “disable call counting” signaling; revise tier finalization for R2R load based on stored tier.
src/coreclr/vm/method.hpp	Add stored optimization tier to `MethodDescCodeData`; remove `WasTieringDisabledBeforeJitting` plumbing.
src/coreclr/vm/method.cpp	Initialize and implement `MethodDesc` optimization tier storage helpers.
src/coreclr/vm/eeconfig.h	Add `TieredCompilation_DefaultTier()` accessor and config backing field.
src/coreclr/vm/eeconfig.cpp	Compute `tieredCompilation_DefaultTier` during `EEConfig::sync()`.
src/coreclr/vm/codeversion.h	Add `OptimizationTierUnknown` sentinel value.
src/coreclr/vm/codeversion.cpp	Use stored `MethodDesc` tier when available; set it for default versions via `MethodDesc`.
src/coreclr/vm/callcounting.h	Remove `CallCountingInfo::Disabled` stage and disabled-construction helpers.
src/coreclr/vm/callcounting.cpp	Remove disabled-stage construction/guards and remove disabled-callcounting APIs’ implementations.

Copilot · 2026-03-05T22:22:52Z

src/coreclr/vm/method.cpp

+    IfFailRet(EnsureCodeDataExists(NULL));
+
+    _ASSERTE(m_codeData != NULL);
+    if (InterlockedExchangeT(&m_codeData->OptimizationTier, tier) != NativeCodeVersion::OptimizationTierUnknown)


SetMethodDescOptimizationTier() uses InterlockedExchangeT, which overwrites any previously-set OptimizationTier even when returning S_FALSE. This breaks the intended “set once from Unknown” semantics and can race with other threads, potentially flipping a method’s tier unexpectedly. Use InterlockedCompareExchangeT to only store the tier when the current value is OptimizationTierUnknown (and return S_FALSE without modifying otherwise).

Suggested change

if (InterlockedExchangeT(&m_codeData->OptimizationTier, tier) != NativeCodeVersion::OptimizationTierUnknown)

if (InterlockedCompareExchangeT(&m_codeData->OptimizationTier, tier, NativeCodeVersion::OptimizationTierUnknown) != NativeCodeVersion::OptimizationTierUnknown)

I agree with copilot here. This is an indiscriminate exchange. Why don't we want InterlockedCompareExchange?

Mmm... the latest change does a simple set, but there is some merit to not wanting to march over stuff willy nilly. I'll think about that tomorrow.

Looking into this, I've concluded that the current logic, which simply sets the value is correct. We could in theory add asserts that its behaving as expected, but just doing the set unconditionally is acceptable.

src/coreclr/vm/prestub.cpp

src/coreclr/vm/eeconfig.cpp

src/coreclr/vm/callcounting.cpp

src/coreclr/vm/method.cpp

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Copilot

Pull request overview

Copilot reviewed 10 out of 10 changed files in this pull request and generated 4 comments.

Copilot · 2026-03-05T23:38:34Z

src/coreclr/vm/tieredcompilation.cpp

    NativeCodeVersion nativeCodeVersion = config->GetCodeVersion();
-    if (nativeCodeVersion.IsDefaultVersion() && !config->WasTieringDisabledBeforeJitting())
-    {
-        MethodDesc *methodDesc = nativeCodeVersion.GetMethodDesc();
-        if (!methodDesc->IsEligibleForTieredCompilation())
-        {
-            _ASSERTE(nativeCodeVersion.GetOptimizationTier() == NativeCodeVersion::OptimizationTierOptimized);
-            return flags;
-        }
-
-        _ASSERT(!methodDesc->RequestedAggressiveOptimization());
-
-        if (g_pConfig->TieredCompilation_QuickJit())
-        {
-            NativeCodeVersion::OptimizationTier currentTier = nativeCodeVersion.GetOptimizationTier();
-            if (currentTier == NativeCodeVersion::OptimizationTier::OptimizationTier0Instrumented)
-            {
-                flags.Set(CORJIT_FLAGS::CORJIT_FLAG_BBINSTR);
-                flags.Set(CORJIT_FLAGS::CORJIT_FLAG_TIER0);
-                return flags;
-            }
-
-            if (currentTier == NativeCodeVersion::OptimizationTier::OptimizationTier1Instrumented)
-            {
-                flags.Set(CORJIT_FLAGS::CORJIT_FLAG_BBINSTR);
-                flags.Set(CORJIT_FLAGS::CORJIT_FLAG_TIER1);
-                return flags;
-            }
-
-            _ASSERTE(!nativeCodeVersion.IsFinalTier());
-            flags.Set(CORJIT_FLAGS::CORJIT_FLAG_TIER0);
-            if (g_pConfig->TieredPGO() && g_pConfig->TieredPGO_InstrumentOnlyHotCode())
-            {
-                // If we plan to only instrument hot code we have to make an exception
-                // for cold methods with loops so if those self promote to OSR they need
-                // some profile to optimize, so here we allow JIT to enable instrumentation
-                // if current method has loops and is eligible for OSR.
-                flags.Set(CORJIT_FLAGS::CORJIT_FLAG_BBINSTR_IF_LOOPS);
-            }
-            return flags;
-        }
-
-        methodDesc->GetLoaderAllocator()->GetCallCountingManager()->DisableCallCounting(nativeCodeVersion);
-        nativeCodeVersion.SetOptimizationTier(NativeCodeVersion::OptimizationTierOptimized);
-        return flags;
-    }

    switch (nativeCodeVersion.GetOptimizationTier())
    {


With the default-version fast path removed, GetJitFlags now relies entirely on nativeCodeVersion.GetOptimizationTier() and the switch below. Some switch cases mutate the tier (e.g., tier0 -> optimized when QuickJit is off), but this PR also removes the old mechanism that disabled call counting to indicate “start optimized and never tier”. Please ensure the publishing/call-counting paths (notably CallCountingManager::SetCodeEntryPoint, which is invoked from CodeVersionManager::PublishNativeCodeVersion) correctly bypass call counting for optimized default versions; otherwise it can hit assertions that assume default versions are not final tiers.

My analysis indicates this was a bug, but I've fixed in the next revision.

Copilot · 2026-03-05T23:38:34Z

src/coreclr/vm/prestub.cpp

        NativeCodeVersion codeVersion = pConfig->GetCodeVersion();
-        if (codeVersion.IsDefaultVersion())
-        {
-            pConfig->GetMethodDesc()->GetLoaderAllocator()->GetCallCountingManager()->DisableCallCounting(codeVersion);
-            _ASSERTE(codeVersion.IsFinalTier());
-        }
-        else if (!codeVersion.IsFinalTier())
+        if (!codeVersion.IsFinalTier())
        {
            codeVersion.SetOptimizationTier(NativeCodeVersion::OptimizationTierOptimized);
        }
-        pConfig->SetWasTieringDisabledBeforeJitting();
        shouldTier = false;


When tiering is disabled here (preemptive caller GC mode / UnmanagedCallersOnly), the code may set the default code version’s tier to OptimizationTierOptimized while the method remains eligible for tiered compilation. Since CodeVersionManager::PublishNativeCodeVersion still calls CallCountingManager::SetCodeEntryPoint for published code, make sure the call counting logic correctly treats optimized default versions as non-call-counted/final (the old implementation relied on DisableCallCounting). Otherwise this can trigger assertions in SetCodeEntryPoint that assume default versions are not final tiers.

Nice catch copilot, this was a bug in the PR.

src/coreclr/vm/callcounting.cpp

src/coreclr/vm/method.cpp

….com:davidwrighton/runtime into rework_and_simplify_initial_versioning_rules

AaronRobinsonMSFT · 2026-03-05T23:31:44Z

src/coreclr/vm/eeconfig.cpp

+            // 1 - OptimizationTier0 as we don't want to instrument the initial version (will only instrument hot Tier0)
+            // 2 - OptimizationTier0Instrumented - instrument all ILOnly code


Suggested change

// 1 - OptimizationTier0 as we don't want to instrument the initial version (will only instrument hot Tier0)

// 2 - OptimizationTier0Instrumented - instrument all ILOnly code

// - OptimizationTier0 as we don't want to instrument the initial version (will only instrument hot Tier0)

// - OptimizationTier0Instrumented - instrument all ILOnly code

I assume the ordinals aren't real values and this is really just a numbered list?

Yes. I have some vague memories that during the experiments around this stuff it was actually a more flexible policy, but it never has had flexibility in the shipped product.

AaronRobinsonMSFT · 2026-03-05T23:53:10Z

src/coreclr/vm/method.cpp

    return S_OK;
 }
+
+HRESULT MethodDesc::SetMethodDescOptimizationTier(NativeCodeVersion::OptimizationTier tier)


How is this handled by MethodDesc::Reset(), does the reset logic also need to be updated?

EnC methods are not eligible for tiering, so that doesn't apply, and dynamic methods aren't supported either, which are the only use cases for MethodDesc::Reset See CodeVersionManager::IsMethodSupported for details.

AaronRobinsonMSFT · 2026-03-06T00:11:59Z

src/coreclr/vm/method.cpp

+    IfFailRet(EnsureCodeDataExists(NULL));
+
+    _ASSERTE(m_codeData != NULL);
+    if (InterlockedExchangeT(&m_codeData->OptimizationTier, tier) != NativeCodeVersion::OptimizationTierUnknown)


I agree with copilot here. This is an indiscriminate exchange. Why don't we want InterlockedCompareExchange?

- Remove redundant IsDefaultVersion() guard before IsFinalTier() check in CallCountingManager::SetCodeEntryPoint - Query optimization tier from CodeVersion instead of MethodDesc in FinalizeOptimizationTierForTier0Load - Clean up comment formatting in EEConfig::sync Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot

Pull request overview

Copilot reviewed 10 out of 10 changed files in this pull request and generated 2 comments.

src/coreclr/vm/eeconfig.h

Copilot · 2026-03-06T20:18:59Z

src/coreclr/vm/eeconfig.cpp

        }
    }
+#else // !FEATURE_TIERED_COMPILATION
+    tieredCompilation_DefaultTier = NativeCodeVersion::OptimizationTierOptimized;


In the !FEATURE_TIERED_COMPILATION branch, tieredCompilation_DefaultTier (a DWORD) is assigned an enum value without an explicit cast, while other assignments in this method cast to DWORD. Using a consistent explicit cast here helps avoid compiler warnings (and matches the nearby code style).

Suggested change

tieredCompilation_DefaultTier = NativeCodeVersion::OptimizationTierOptimized;

tieredCompilation_DefaultTier = (DWORD)NativeCodeVersion::OptimizationTierOptimized;

davidwrighton requested review from AaronRobinsonMSFT, Copilot and rcj1 March 5, 2026 22:15

github-actions bot added the area-VM-coreclr label Mar 5, 2026

dotnet-policy-service bot assigned davidwrighton Mar 5, 2026

Copilot started reviewing on behalf of davidwrighton March 5, 2026 22:16 View session

Copilot AI reviewed Mar 5, 2026

View reviewed changes

Update src/coreclr/vm/method.cpp

c639ff0

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Copilot AI review requested due to automatic review settings March 5, 2026 23:31

Copilot started reviewing on behalf of davidwrighton March 5, 2026 23:32 View session

Copilot AI reviewed Mar 5, 2026

View reviewed changes

davidwrighton added 2 commits March 5, 2026 16:08

Feedback

12130e6

Merge branch 'rework_and_simplify_initial_versioning_rules' of github…

12209cf

….com:davidwrighton/runtime into rework_and_simplify_initial_versioning_rules

AaronRobinsonMSFT reviewed Mar 6, 2026

View reviewed changes

build-analysis bot mentioned this pull request Mar 6, 2026

LibraryImportGenerator.Unit.Tests crashing on linux-x64 mono interpreter #100800

Open

Copilot AI review requested due to automatic review settings March 6, 2026 20:12

Copilot started reviewing on behalf of davidwrighton March 6, 2026 20:14 View session

Copilot AI reviewed Mar 6, 2026

View reviewed changes

	if (InterlockedExchangeT(&m_codeData->OptimizationTier, tier) != NativeCodeVersion::OptimizationTierUnknown)
	if (InterlockedCompareExchangeT(&m_codeData->OptimizationTier, tier, NativeCodeVersion::OptimizationTierUnknown) != NativeCodeVersion::OptimizationTierUnknown)

		// 1 - OptimizationTier0 as we don't want to instrument the initial version (will only instrument hot Tier0)
		// 2 - OptimizationTier0Instrumented - instrument all ILOnly code

	tieredCompilation_DefaultTier = NativeCodeVersion::OptimizationTierOptimized;
	tieredCompilation_DefaultTier = (DWORD)NativeCodeVersion::OptimizationTierOptimized;

Conversation

davidwrighton commented Mar 5, 2026

Uh oh!

dotnet-policy-service bot commented Mar 5, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

davidwrighton Mar 6, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Copilot AI Mar 6, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

davidwrighton Mar 6, 2026 •

edited

Loading