fix: upgrade ckb-vm to fix snapshot behavior

[driftluo]

What problem does this PR solve?

chunk run with the snapshot unable to support the binary replacement behavior of exec, which will cause abnormal behavior after resume.

The upgraded VM supports backing up the current code(executable pages) to the snapshot to fix the above problems. nervosnetwork/ckb-vm#218

and #3176 should keep its behavior until hardfork activated

Check List

Tests

• Unit test
• Integration test

Release note

Title Only: Include only the PR title in the release note.

[quake]

Suggest to add unit test to cover this PR.

…

On Mon, Nov 15, 2021, 18:04 Mohanson ***@***.***> wrote: ***@***.**** requested changes on this pull request. ------------------------------ In script/src/syscalls/load_cell_data.rs <#3177 (comment)>: > @@ -104,13 +107,15 @@ impl<'a, DL: CellDataProvider + 'a> LoadCellData<'a, DL> { .load_cell_data(cell) .ok_or(VMError::Unexpected)? .slice((content_offset as usize)..(content_end as usize)); - machine.memory_mut().init_pages( - addr, - memory_size, - FLAG_DIRTY | FLAG_EXECUTABLE | FLAG_FREEZED, - Some(data), - 0, - )?; + // To be consistent with the mainnet, add this flag after hardfork I made a mistake before. Any write operation to the ckb-vm memory will be marked as dirty, and the correct logic should be: machine.memory_mut().init_pages( addr, memory_size, FLAG_EXECUTABLE | FLAG_FREEZED, Some(data), 0, )?; if !self.dirty_flag_enable { let mut current_addr = addr; while current_addr < addr + memory_size { let page = current_addr / RISCV_PAGESIZE as u64; machine.memory_mut().clear_flag(page, FLAG_DIRTY)?; current_addr += RISCV_PAGESIZE as u64; } } — You are receiving this because your review was requested. Reply to this email directly, view it on GitHub <#3177 (review)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAACGHTTDGHO4O5A25UWTWLUMDLMVANCNFSM5IBHDRYA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

[yangby-cryptape]

I pushed #3178 to add more tests, I think that PR should be merged before this PR.

Notice

#3178 doesn't fix the snapshot behavior.

How to enable the test for resuming from snapshot?

• This PR and my PR has conflicts, so resolved the conflicts at first.

  • This PR add true for all LoadCellData::new(..) in tests, but I think test both true and false is better.

  • Find is_vm_version_1_and_syscalls_2_enabled is a bit complicated, I think make a method is better than copied those code twice.

• Then find the follow code in script/src/verify/tests/ckb_latest/features_since_v2021.rs:

      // TODO fix resume from snapshot
      let should_be_invalid_permission = script_version <= ScriptVersion::V1;

  Apply the follow changes:

  -    // TODO fix resume from snapshot
  -    let should_be_invalid_permission = script_version <= ScriptVersion::V1;
  +    let should_be_invalid_permission = script_version < ScriptVersion::V1;

  See details in test: add more tests to ckb-script for resuming from snapshot #3178 (comment).

• Make sure all tests are passed at last.

[zhangsoledad]

bors r= yangby-cryptape,mohanson,zhangsoledad

[bors]

Build succeeded:

• ci_benchmarks_ubuntu
• ci_cargo_deny_ubuntu
• ci_integration_tests_ubuntu
• ci_linters_ubuntu
• ci_quick_checks_ubuntu
• ci_unit_tests_ubuntu
• ci_wasm_build_ubuntu