Rebase to e33e3c1 by rubujubi · Pull Request #385 · movement-network/aptos-core

rubujubi · 2026-06-25T03:09:42Z

Description

Downstream Aptos pre license change commits from last fork f7a17a6 (on movement) to e33e3c1 (on upstream) onto movement m1.

This PR is created by rebasing onto e33e3c1, and solve the merge conflicts. Mechanical conflicts as explained in the below section has been directly fixed in the merge commit.

Mechanical conflicts — rename/delete resolutions

Group	Count	Rule applied
`.github/{workflows,actions}` rename/delete	20	Movement renamed these upstream CI files to `.bak` (archived); upstream deleted them → kept Movement's `.bak`
`aptos-rosetta` removal	6	Movement removed the crate; upstream modified the files → honored deletion (`git rm`) + a `git rm -r crates/aptos-rosetta` sweep to catch any upstream re-adds
`my_first_dapp` example manifests	2	Upstream deleted the example; m1 had bumped deps → followed upstream deletion

Total: 28 rename/delete conflicts

The `values_impl::IntegerValue` type introduced a rather unnecessary wrapper for integer values, creating to conversions back and force to `Value`. Moreover, the newtype wrapper `Value` for `ValueImpl` was unnecessary as well and created overhead because it cannot be eliminated by the compiler when applied in iterators (as in `values.into_iter().map(Value).collect()`). Both have been removed.

This enables intermediate tools to use `$` in identifiers to separate from user identifiers, e.g. names for public struct functions. This is only supported from bytecode version 9 onwards. Co-authored-by: Teng Zhang <rahxephon89@163.com>

…__`. (aptos-labs#17854) * [compiler & language] Support builtin constant `__COMPILE_FOR_TESTING__`. Allow code to branch over a compile time constant `__COMPILE_FOR_TESTING__`. The non-reachable branch will be optimized (or at least turned into dead code.) The name is under discussion. We should make the right decision here since any future compile time constants should be following this naming convention. I see compile constants as an equivalent to Rust attributes, at least if it comes to conditional compilation. This falls into the categorie of builtin constants which we already have for the spec llanguage. At the same time, we would enable `MAX_U64` et. all as builtin constants in the language. Builtin constants can be shadowed by the user, so no harm to existing move programs would be implied. * Fully support MIN/MAX for all integer types. Also move unit test out of framework, because it may not compile with older compilers.

Co-authored-by: Igor <igor-aptos@users.noreply.github.com>

Pushed to one node and let it run overnight. Seems like compactions can keep up.

…os-labs#17867) * OrderBook use of iterators and utilities * don't use iterators

…7876)

…os-labs#17882)

…ced on the book (aptos-labs#17868)

…s-compile + echo-aware)

…s), keep Movement governed-gas/treasury + test helpers, dedup merged test fns

…ting match arms (merge artifact)

… testing feature

- thread traversal_context through initialize_governed_gas_pool to match exec_function's updated 7-arg signature - pass the dealer public key to DefaultDKG::generate_transcript in the rand network_messages test - switch that test from the removed WeightedConfig name to the concrete WeightedConfigBlstrs alias

… ReplayNetworkSelection variants

…o stake test helpers

…eric

blacksmith-sh · 2026-07-02T18:52:48Z

Found 11 test failures on Blacksmith runners:

Failures

Test	View Logs
`aptos-transactional-test-harness::tests/runner::aptos_test_harness/bug_bbb75fa.move`	View Logs
`aptos-transactional-test-harness::tests/runner::aptos_test_harness/call_function.move`	View Logs
`aptos-transactional-test-harness::tests/runner::aptos_test_harness/diamond_clicker.move`	View Logs
`aptos-transactional-test-harness::tests/runner::aptos_test_harness/` `diamond_clicker_noinline.move`	View Logs
`aptos-transactional-test-harness::tests/runner::aptos_test_harness/publish_module.move`	View Logs
`aptos-transactional-test-harness::tests/runner::aptos_test_harness/` `randomness_safety.move`	View Logs
`aptos-transactional-test-harness::tests/runner::aptos_test_harness/smoke_test.move`	View Logs
`aptos-transactional-test-harness::tests/runner::aptos_test_harness/string_fmt.move`	View Logs
`aptos-transactional-test-harness::tests/runner::aptos_test_harness/tool_attributes.move`	View Logs
`aptos-transactional-test-harness::tests/runner::aptos_test_harness/` `write_set_too_large.move`	View Logs
`e2e-move-tests/tests::cryptoalgebra::test_deep_type_tag`	View Logs

…r output (format-only, #365 logic unchanged)

ganymedio · 2026-07-07T20:01:47Z

@rubujubi does Movement support orderless / nonce transactions? This PR exposes them in the CLI via --replay-protection-type nonce on run/publish/simulate. Default is seqnum so existing behavior is safe, but this exposes an upstream feature that may not be enabled, so if orderless-txn / payload-v2 isn't actually live on Movement then --replay-protection-type nonce produces transactions the node will reject. Which I think is okay, just something to be aware of.

ganymedio · 2026-07-07T20:08:27Z

A couple things to flag in the SDK:

Dead fields on TransactionFactory (likely a half-carried port). The PR adds two fields — use_txn_payload_v2_format and use_replay_protection_nonce — and a build() branch that consumes them:

payload: if self.use_txn_payload_v2_format || self.use_replay_protection_nonce {
    payload.upgrade_payload_with_fn(...)   // + thread_rng() nonce
} else { port). The PR adds two fields — use_txn_payload_v2_format and use_replay_protection_nonce — and a build() branch that consumes them:
payload: if self.use_txn_payload_v2_format || self.use_replay_protection_nonce {
    payload.upgrade_payload_with_fn(...)   // + thread_rng() nonce
} else {
    payload
},

but there doesn't appear to be a setter for either field. They're hardcoded false in new(), never mutated, and the fields are private so no other crate can set them. The factory's only with_* methods are gas/expiration/chain-id. So the if branch is currently unreachable (always takes the else).

It's not a bug, more like dead weight, just wanted to check that upstream didn't add a builder setter (e.g. a with_payload_v2_format(...)) that maybe didn't get carried through the rebase? Not sure if a factory setter dropped in the merge, or are these fields intentionally inert for now?

get_transaction_expiration_time() can panic. It now does unimplemented!() for the new Absolute expiration variant:

TransactionExpiration::Absolute { .. } => unimplemented!(),

No callers currently, so it's latent, but figured I'd flag for awareness.

rubujubi · 2026-07-13T04:06:04Z

@rubujubi does Movement support orderless / nonce transactions? This PR exposes them in the CLI via --replay-protection-type nonce on run/publish/simulate. Default is seqnum so existing behavior is safe, but this exposes an upstream feature that may not be enabled, so if orderless-txn / payload-v2 isn't actually live on Movement then --replay-protection-type nonce produces transactions the node will reject. Which I think is okay, just something to be aware of.

Thanks, Movement does not support it yet. I think it is fine to leave it here since currently --replay-protection-type nonce will be rejected cleanly

seanyoung

I've gone through the merge conflicts and they all look good to me.

The governed gas pool merge moves things around but everything is still there. I will double check once the governed gas pool is ready for merging.

I have to say it painstaking work, I did my best but I hope I did not miss anything.

areshand · 2026-07-29T22:14:39Z

validated through ephemeral mainnet. The chain is health. let's go ahead and land it.

junxzm1990 and others added 30 commits October 11, 2025 16:07

[vm sint] update test cases for ty_interner

b007e16

[Compiler v2] support signed integers

abf1425

[compiler sint] update trans test results

c487e0f

[compiler sint]: increase coverage of edge test cases

c94c50c

[compiler sint] update test case results

455c4df

[vm] Add layout caches (aptos-labs#17788)

343997b

[txn-emitter] support for orderless txns

3b8ff98

use txn summaries api instead of per txn api call

7ca04d6

support for orderless txn fa transfer

ce0f691

[replay] parallelize on chunk level

b0b33ca

[lint] Added empty_if lint and tests

ef8d5ba

Improved tests.

047b4d8

map modifiers (aptos-labs#17821)

6319b6b

Co-authored-by: Igor <igor-aptos@users.noreply.github.com>

[Keyless] Small improvements to logging and timeout.

e0b5356

[Keyless] Use spawn_blocking for crypto operations.

bbea594

[Keyless] Make Firestore DB update async.

28dc629

[storage] lower compaction threads more

b188fc5

Pushed to one node and let it run overnight. Seems like compactions can keep up.

[Rest API] support for signed int

06d69ca

[txt arg]: update recorded format

4e73a99

OrderBook use of BigOrderedMap utilities to avoid double lookups (apt…

d7abce7

…os-labs#17867) * OrderBook use of iterators and utilities * don't use iterators

delete unused code (aptos-labs#17849)

535c6b5

[benchmark] Enable layout caches for external benchmark (aptos-labs#1…

095f861

…7876)

add script to automate building and running with PGO (aptos-labs#17831)

490e80c

Address security review feedback (aptos-labs#17874)

d4ee6a8

[vm] Strengthen indexed ref checks (aptos-labs#16386)

59e7c8c

[vm] Disable static checks for depth in favour of runtime checks (apt…

4bf49f7

…os-labs#17882)

[Orderbook] Ability to return order actions when a maker order is pla…

eb5bb71

…ced on the book (aptos-labs#17868)

rubujubi added 4 commits June 26, 2026 01:15

resolve(test-infra): adopt e33 transactional-test-runner output (cros…

23e3006

…s-compile + echo-aware)

resolve(framework): adopt e33 stake.move (BigOrderedMap + fee acquire…

55f20ce

…s), keep Movement governed-gas/treasury + test helpers, dedup merged test fns

fix(timed-features): remove duplicate FixTableNativesMemoryDoubleCoun…

70bdc35

…ting match arms (merge artifact)

fix(move-vm): route transactional-tests verifier cache bypass through…

8a43dbb

… testing feature

rubujubi force-pushed the move-rebase-e33 branch from 47317bf to 8a43dbb Compare June 26, 2026 05:18

apenzk and others added 5 commits July 1, 2026 09:09

resolve(consensus): union imports for e33 rebase (tier 1)

b54c09b

fix(cli): route Replay client through Movement rest_endpoint, qualify…

cc97e55

… ReplayNetworkSelection variants

fix(rebase): align downstream code with e33 API changes

7546f74

This comment has been minimized.

Sign in to view

rubujubi added 2 commits July 2, 2026 02:41

fix(framework): dedup ordered_map append tests and add fee acquires t…

ce1fa04

…o stake test helpers

fix(migration): pin unsharded storage layout after e33 default flip

bc4ab3e

This comment has been minimized.

Sign in to view

fix(move-vm): restore #354 verify-before-pack order in PackClosureGen…

f77b478

…eric

fix(move-vm-tests): regenerate closure_compare.exp for e33 test-runne…

f283762

…r output (format-only, #365 logic unchanged)

rubujubi marked this pull request as ready for review July 9, 2026 17:37

rubujubi requested review from 0xIcarus, Primata, areshand, ganymedio and musitdev as code owners July 9, 2026 17:37

rubujubi requested a review from seanyoung July 9, 2026 17:37

ganymedio mentioned this pull request Jul 13, 2026

Add Movement network and CLI support MoveIndustries/forklift#1

Merged

seanyoung approved these changes Jul 14, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Rebase to e33e3c1 - #385

Rebase to e33e3c1#385
rubujubi wants to merge 835 commits into
m1from
move-rebase-e33

rubujubi commented Jun 25, 2026 •

edited

Loading

Uh oh!

This comment has been minimized.

This comment has been minimized.

blacksmith-sh Bot commented Jul 2, 2026 •

edited

Loading

Uh oh!

ganymedio commented Jul 7, 2026 •

edited

Loading

Uh oh!

ganymedio commented Jul 7, 2026 •

edited

Loading

Uh oh!

rubujubi commented Jul 13, 2026

Uh oh!

seanyoung left a comment

Uh oh!

areshand commented Jul 29, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

20 participants

Uh oh!

Conversation

rubujubi commented Jun 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Mechanical conflicts — rename/delete resolutions

Uh oh!

This comment has been minimized.

This comment has been minimized.

blacksmith-sh Bot commented Jul 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Failures

Uh oh!

ganymedio commented Jul 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ganymedio commented Jul 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rubujubi commented Jul 13, 2026

Uh oh!

seanyoung left a comment

Choose a reason for hiding this comment

Uh oh!

areshand commented Jul 29, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

20 participants

rubujubi commented Jun 25, 2026 •

edited

Loading

blacksmith-sh Bot commented Jul 2, 2026 •

edited

Loading

ganymedio commented Jul 7, 2026 •

edited

Loading

ganymedio commented Jul 7, 2026 •

edited

Loading