feat(blockifier): replace the aliases in the state diff

[yoavGrs]

No description provided.

[reviewable-StarkWare]

This change is 

[yoavGrs]

• feat(blockifier): return compressed state diff at finalizing #2767
• feat(blockifier): replace the aliases in the state diff #2766 👈 (View in Graphite)
• feat(blockifier): define alias compressor struct #2765
• feat(blockifier): pass alias contract address #2674 : 1 other dependent PR (#2675 )
• feat(blockifier): iterate over aliases on the state diff #2535
• feat(blockifier): aliases updater struct #2534
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

[github-actions]

Artifacts upload workflows:

• Started at 2024-12-18T15:09:08Z
• Started at 2024-12-19T14:38:04Z
• Started at 2024-12-19T15:12:47Z
• Started at 2024-12-19T16:00:30Z
• Started at 2024-12-22T09:04:13Z
• Started at 2024-12-23T08:23:10Z
• Started at 2024-12-23T09:07:43Z
• Started at 2024-12-23T09:30:13Z
• Started at 2024-12-23T12:57:04Z
• Started at 2024-12-23T13:28:16Z
• Started at 2024-12-23T13:29:07Z
• Started at 2024-12-23T13:33:26Z
• Started at 2024-12-23T14:33:12Z

[dorimedini-starkware]

Reviewed 2 of 2 files at r1, all commit messages.
Reviewable status: all files reviewed, 4 unresolved discussions (waiting on @nimrod-starkware and @yoavGrs)

---

crates/blockifier/src/state/stateful_compression_test.rs line 0 at r1 (raw file):
can you implement a decompress test util and use it to compare original and compressed+decompressed diff?

---

crates/blockifier/src/state/stateful_compression_test.rs line 271 at r1 (raw file):

            .into_iter()
            .collect(),
        declared_contracts: vec![(ClassHash(Felt::THREE), true)].into_iter().collect(),

put large values here? to make sure these are not compressed, but for the correct reason

Code quote:

        compiled_class_hashes: vec![(ClassHash(Felt::ONE), CompiledClassHash(Felt::ZERO))]
            .into_iter()
            .collect(),
        declared_contracts: vec![(ClassHash(Felt::THREE), true)].into_iter().collect(),

---

crates/blockifier/src/state/stateful_compression.rs line 161 at r1 (raw file):

            .class_hashes
            .insert(alias_compressor.compress_address(address)?, *class_hash);
    }

we always compress class hashes? @Yoni-Starkware @nimrod-starkware

Code quote:

    for (address, class_hash) in state_diff.class_hashes.iter() {
        compressed_state_diff
            .class_hashes
            .insert(alias_compressor.compress_address(address)?, *class_hash);
    }

---

crates/blockifier/src/state/stateful_compression.rs line 174 at r1 (raw file):

    compressed_state_diff.declared_contracts.extend(state_diff.declared_contracts.iter());

    Ok(compressed_state_diff)

for forward-compatibility (if fields are added to the state diff) better to use .. notation and explicitly replace specific compressed fields. WDYT?

Suggestion:

    let alias_compressor = AliasCompressor { state, alias_contract_address };

    let nonces: ... = state_diff.nonces.iter().map(|(address, nonce)| {
        (alias_compressor.compress_address(address)?, *nonce)
    }).collect();
    let class_hashes: ... = state_diff.class_hashes.iter().map(|(address, class_hash)| {
        (alias_compressor.compress_address(address)?, *class_hash)
    }).collect();
    let storage: ... = state_diff.storage.iter().map(|((address, key), value)| {
        (
            (
                alias_compressor.compress_address(address)?,
                alias_compressor.compress_storage_key(key, address)?,
            ),
            *value,
        )
    }).collect();

    Ok(StateMaps {
        nonces,
        class_hashes,
        storage,
        ..state_diff
    })

[yoavGrs]

Reviewable status: 0 of 2 files reviewed, 4 unresolved discussions (waiting on @dorimedini-starkware and @nimrod-starkware)

---

crates/blockifier/src/state/stateful_compression.rs line 174 at r1 (raw file):

Previously, dorimedini-starkware wrote…

for forward-compatibility (if fields are added to the state diff) better to use .. notation and explicitly replace specific compressed fields. WDYT?

Done.
(Without ? in closures)

---

crates/blockifier/src/state/stateful_compression_test.rs line 271 at r1 (raw file):

Previously, dorimedini-starkware wrote…

put large values here? to make sure these are not compressed, but for the correct reason

Done.

---

crates/blockifier/src/state/stateful_compression_test.rs line at r1 (raw file):

Previously, dorimedini-starkware wrote…

can you implement a decompress test util and use it to compare original and compressed+decompressed diff?

Done.

[dorimedini-starkware]

Reviewed 1 of 2 files at r2, 1 of 1 files at r3, all commit messages.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @nimrod-starkware and @yoavGrs)

[yoavGrs]

Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @nimrod-starkware and @Yoni-Starkware)

---

crates/blockifier/src/state/stateful_compression.rs line 161 at r1 (raw file):

Previously, dorimedini-starkware wrote…

we always compress class hashes? @Yoni-Starkware @nimrod-starkware

In this code, we compress only the contract address.

[dorimedini-starkware]

Reviewed 1 of 2 files at r4, 1 of 1 files at r5, all commit messages.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @nimrod-starkware and @Yoni-Starkware)

---

crates/blockifier/src/state/stateful_compression.rs line 161 at r1 (raw file):

Previously, yoavGrs wrote…

In this code, we compress only the contract address.

what I meant to ask: if only class hash changed, do we compress the contract address?

[dorimedini-starkware]

Reviewed 1 of 2 files at r6, 1 of 1 files at r7, all commit messages.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @nimrod-starkware and @Yoni-Starkware)

[yoavGrs]

Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @nimrod-starkware and @Yoni-Starkware)

---

crates/blockifier/src/state/stateful_compression.rs line 161 at r1 (raw file):

Previously, dorimedini-starkware wrote…

what I meant to ask: if only class hash changed, do we compress the contract address?

This is what Nimrod wrote to me:

for address in address_to_class_hash.keys() | address_to_nonce.keys() | storage_updates.keys():
    if address in storage_updates:
         // allocate aliases for storage keys.
    // allocate alias for address

[Yoni-Starkware]

Reviewable status: all files reviewed, 2 unresolved discussions (waiting on @nimrod-starkware and @yoavGrs)

---

crates/blockifier/src/state/stateful_compression.rs line 174 at r1 (raw file):

Previously, yoavGrs wrote…

Done.
(Without ? in closures)

@yoavGrs can you please follow Dori's suggestion re collect instead of new + for? (convention)

---

crates/blockifier/src/state/stateful_compression.rs line 183 at r7 (raw file):

    ) -> CompressionResult<ContractAddress> {
        if contract_address.0 >= MIN_VALUE_FOR_ALIAS_ALLOC {
            Ok(self.get_alias(StorageKey(contract_address.0))?.try_into()?)

StorageKey -> AliasKey

[dorimedini-starkware]

Reviewed 2 of 2 files at r9, all commit messages.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @nimrod-starkware and @yoavGrs)

---

crates/blockifier/src/state/stateful_compression.rs line 174 at r1 (raw file):

Previously, Yoni-Starkware (Yoni) wrote…

@yoavGrs can you please follow Dori's suggestion re collect instead of new + for? (convention)

@Yoni-Starkware issue is ? within the iterator closures.
@yoavGrs you can collect::<Result<_, _>>()? to solve this though (collect is smart enough to collect an Iterator<Result<_, _>> into a Result<_, _>)

[dorimedini-starkware]

Reviewed 2 of 2 files at r11, all commit messages.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @nimrod-starkware and @yoavGrs)

[yoavGrs]

Reviewable status: 0 of 2 files reviewed, 1 unresolved discussion (waiting on @dorimedini-starkware, @nimrod-starkware, and @Yoni-Starkware)

---

crates/blockifier/src/state/stateful_compression.rs line 174 at r1 (raw file):

Previously, dorimedini-starkware wrote…

@Yoni-Starkware issue is ? within the iterator closures.
@yoavGrs you can collect::<Result<_, _>>()? to solve this though (collect is smart enough to collect an Iterator<Result<_, _>> into a Result<_, _>)

Done!
I should return Result from the closures, to allow using ? within them.

---

crates/blockifier/src/state/stateful_compression.rs line 183 at r7 (raw file):

Previously, Yoni-Starkware (Yoni) wrote…

StorageKey -> AliasKey

Compilation error: can't use a type alias as a constructor

[dorimedini-starkware]

Reviewed 2 of 2 files at r13, all commit messages.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @nimrod-starkware and @Yoni-Starkware)

[Yoni-Starkware]

Reviewed 1 of 2 files at r13.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @nimrod-starkware)