fix(gnovm): embedded type identity — struct field name + interface method-set flatten

[ltzmaxwell]

• Embedded type-alias fields take the field name from the alias as written (type Int = int; struct{ Int } → Int, not int), derived from the source AST expr across all three construction paths; covers qualified, pointer, interface, rune/byte, and alias-to-defined-type embeds.
• Embedded interfaces flatten into their method set at construction (flattenInterfaceMethods), so interface identity is the method set, not the embed/alias spelling: interface{ SAlias } == interface{ Stringer } == interface{ S() string }, matching Go.
• Cross-package unexported methods keep their origin-package identity via a stamped FieldType.PkgPath (new wire field): sealed interfaces stay sealed through flattening, and unexported selection/satisfaction gate on the origin package, not the enclosing interface's.
• Consensus-breaking (coordinated upgrade / fresh genesis, same release class as fix(gnovm): match Go's call-time dispatch for interface-bound method values #5737): TypeIDs of alias-embedding structs and embed-carrying interfaces change, and the runtime now assumes flattened interfaces — decoding pre-flattening persisted state is a hard error; regenesis (tx replay) or a re-flatten migration is required (see gnovm/adr/pr5739_interface_method_set_flattening.md).
• Tests: alias2–alias6 + iface_embed_* filetests (identity, cross-package sealing, access gate, conflict rejection), flatten/TypeID/hard-error unit tests, codec parity for the new wire field; all verified against Go.

[Gno2D2]

🛠 PR Checks Summary

All Automated Checks passed. ✅

Manual Checks (for Reviewers):

• IGNORE the bot requirements for this PR (force green CI check)

Read More

🤖 This bot helps streamline PR reviews by verifying automated checks and providing guidance for contributors and reviewers.

✅ Automated Checks (for Contributors):

🟢 Maintainers must be able to edit this pull request (more info)

☑️ Contributor Actions:

1. Fix any issues flagged by automated checks.
2. Follow the Contributor Checklist to ensure your PR is ready for review.
   • Add new tests, or document why they are unnecessary.
   • Provide clear examples/screenshots, if necessary.
   • Update documentation, if required.
   • Ensure no breaking changes, or include BREAKING CHANGE notes.
   • Link related issues/PRs, where applicable.

☑️ Reviewer Actions:

1. Complete manual checks for the PR, including the guidelines and additional checks if applicable.

📚 Resources:

• Report a bug with the bot.
• View the bot’s configuration file.

Debug

Automated Checks

Maintainers must be able to edit this pull request (more info)

If

🟢 Condition met
└── 🟢 And
    ├── 🟢 The base branch matches this pattern: ^master$
    └── 🟢 The pull request was created from a fork (head branch repo: ltzmaxwell/gno)

Then

🟢 Requirement satisfied
└── 🟢 Maintainer can modify this pull request

Manual Checks

**IGNORE** the bot requirements for this PR (force green CI check)

If

🟢 Condition met
└── 🟢 On every pull request

Can be checked by

• Any user with comment edit permission

[codecov]

Codecov Report

❌ Patch coverage is 74.28571% with 9 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
gnovm/pkg/gnolang/types.go	70.96%	6 Missing and 3 partials ⚠️

📢 Thoughts on this report? Let us know!

[davd-gzl]

Verified on 155f1a7 against a side-by-side Go run and current master:

                 struct{ Int }     interface{ SAlias }
                 vs struct{ int }   vs interface{ Stringer }
  Go             distinct  ✓        identical ✓
  master (pre)   identical ✗        identical ✓
  PR 155f1a7     distinct  ✓        distinct  ✗  <- new divergence

Full review: https://github.com/samouraiworld/gno-agent-workspace/blob/main/reviews/pr/5xxx/5739-preserve-embedded-alias-name/1-155f1a7/review_claude-opus-4-8_davd-gzl.md

[ltzmaxwell]

Why flatten is the correct fix — verified on master

Master names an embedded-interface entry from the resolved type (fillEmbeddedName: case *DeclaredType: ft.Name = ct.Name) and never flattens, so the embed's name is part of the interface TypeID. Verified on origin/master (ad09cb7):

comparison	master	Go
interface{ Stringer } vs interface{ Str() string }	distinct ✗	identical
interface{ SAlias } vs interface{ Stringer } (type SAlias = Stringer)	identical	identical

No naming policy — resolved or spelled — can make interface{ Stringer } equal interface{ Str() string }: the embed is still a single named entry (Stringer) rather than its method (Str). Go defines interface identity as the flattened method set — embedding contributes methods, not a name. The only fix that matches Go is to flatten the embed into its methods, removing the embed name from identity. That's why this PR takes the flatten path. (Recorded in gnovm/adr/pr5739_interface_method_set_flattening.md.)

[ltzmaxwell]

880040d (+ c8c7329) resolves the ADR's deferred rollout decision: the runtime now assumes flattened interfaces — the legacy embedded-interface branches in FindEmbeddedFieldType/VerifyImplementedBy are dropped.

• Invariant: every construction path flattens (flattenInterfaceMethods), so Methods holds only concrete methods (FuncType entries). An InterfaceKind entry is only possible by decoding pre-flattening persisted bytes.
• Compatibility: tolerating those bytes was a half-measure — flattening already moved interface identity (TypeIDs), so resolving against a legacy type would "work" while equality, type-keyed store entries, and hashes stay silently split from post-change construction. Loud-fail beats silent-wrong.
• Persistence: pre-flattening state is unsupported. A network carrying it needs fresh genesis (tx replay rebuilds all state flattened) or a re-flatten migration (flatten every persisted InterfaceType, remap old→new TypeIDs). Same release class as fix(gnovm): match Go's call-time dispatch for interface-bound method values #5737: consensus-breaking, coordinated upgrade.
• Enforcement: panicUnflattened — ungated at TypeID() (once per instance; the production choke), under -tags debugAssert at the hot per-lookup sites.

Details: gnovm/adr/pr5739_interface_method_set_flattening.md.

[ltzmaxwell]

This is the one ungated legacy-state check, placed here deliberately: store bytes are external input, and fillType sees every stored type — type entries via GetTypeSafe, and object-carried anonymous types via fillTypesOfValue (only DeclaredType collapses to RefType on persist; unnamed types nest inline, so a GetTypeSafe-only check would miss them). An embed entry can only come from bytes persisted before flattening, so it fails at decode, before any semantics run.

The interior sites (FindEmbeddedFieldType / VerifyImplementedBy / TypeID) assume the invariant on a validated store and only assert under -tags debugAssert.

[ltzmaxwell]

@thehowl — several changes landed after your approval; a quick second look would be appreciated. Summary:

• Your two comments applied: the unexported gate now keys on im.originPkg(...) (0c7c632, with iface_embed_xpkg_access.gno pinning the compile-time reject + the origin-package positive direction), and slices.ContainsFunc (7302974).
• ADR rollout question resolved — runtime now assumes flattened interfaces (880040d → a43ef22): the legacy embedded-interface branches in FindEmbeddedFieldType/VerifyImplementedBy are dropped. Enforcement is split by concern: one ungated check at the decode boundary (fillType — store bytes are external input; see the inline comment there), while interior sites (FindEmbeddedFieldType/VerifyImplementedBy/TypeID) assert under -tags debugAssert only. Rationale in the updated ADR; PR body now declares the release class (consensus-breaking, coordinated upgrade / fresh genesis, same as fix(gnovm): match Go's call-time dispatch for interface-bound method values #5737).
• Coverage gaps closed: alias5.gno (rune/byte builtin-alias embeds), and TestDoOpStructType_EmbedNames + alias7.gno for the runtime struct-op path — previously unreached by any filetest (verified by instrumentation).