blockifier: sha512 syscall

[dorimedini-starkware]

No description provided.

[reviewable-StarkWare]

This change is 

[dorimedini-starkware]

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

• blockifier: sha512 syscall #14072 : 2 dependent PRs (#14073 , #14205 ) 👈 (View in Graphite)
• release: upgrade cairo, native, proving-utils #14200
• main

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

[github-actions]

Artifacts upload workflows:

• Started at 2026-05-19T07:03:54Z
• Started at 2026-05-26T13:03:12Z
• Started at 2026-05-26T13:05:08Z
• Started at 2026-05-26T13:08:36Z

[cursor]

PR Summary

High Risk
Touches Starknet OS program hashes, virtual OS allowlist, and core syscall execution—consensus- and proving-sensitive; incorrect hashing or gas would be protocol-breaking.

Overview
Adds SHA-512 process-block syscall support end-to-end, mirroring the existing SHA-256 path: new Sierra libfuncs (sha512_*), blockifier VM/native execution via sha2::compress512, OS Cairo dispatch and execute_sha512_process_block, a sha512 non-selectable builtin with finalize_sha512 / RelocateSha512Segment / Sha512Finalize hints, and syscall gas (SHA512_PROCESS_BLOCK_GAS_COST ≈ 2.4M).

OS / proving: program_hash.json and allowed virtual OS program hash are updated; bytecode length expectations rise. Versioned constants gain Sha512ProcessBlock entries (non-zero OS step/builtin costs from 0.14.3 onward).

Tests & tooling: Contract test_sha512, blockifier syscall tests, flow/virtual OS coverage, allowed-libfunc lists (including pending-audit entries), and refreshed integration fixtures.

^{Reviewed by Cursor Bugbot for commit 7dd2592. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit be74220. Configure here.}

[Yoni-Starkware]

@Yoni-Starkware reviewed 19 files and all commit messages, and made 1 comment.
Reviewable status: 19 of 65 files reviewed, 1 unresolved discussion (waiting on dorimedini-starkware and liorgold2).

---

crates/blockifier/resources/blockifier_versioned_constants_0_14_3.json line 477 at r3 (raw file):

                    "bitwise_builtin": 2230
                }
            },

Not consistent with your slack message

Code quote:

            "Sha256ProcessBlock": {
                "n_steps": 1867,
                "n_memory_holes": 0,
                "builtin_instance_counter": {
                    "range_check_builtin": 65,
                    "bitwise_builtin": 1115
                }
            },
            "Sha512ProcessBlock": {
                "n_steps": 3734,
                "n_memory_holes": 0,
                "builtin_instance_counter": {
                    "range_check_builtin": 130,
                    "bitwise_builtin": 2230
                }
            },

[Yoni-Starkware]

Can you please add sha512 call in here? (virtual os e2e test)

sequencer/crates/starknet_os_flow_tests/src/virtual_os_test.rs

Line 291 in 1ebcb04

let multi_call_args = create_multicall_calldata(&[

@Yoni-Starkware reviewed 7 files and made 1 comment.
Reviewable status: 26 of 65 files reviewed, 1 unresolved discussion (waiting on dorimedini-starkware and liorgold2).

[Yoni-Starkware]

You should add SHA512_PROCESS_BLOCK_GAS_COST to constants_test.rs

@Yoni-Starkware reviewed 1 file and made 1 comment.
Reviewable status: 27 of 65 files reviewed, 1 unresolved discussion (waiting on dorimedini-starkware and liorgold2).

[Yoni-Starkware]

@Yoni-Starkware made 3 comments.
Reviewable status: 27 of 65 files reviewed, 4 unresolved discussions (waiting on dorimedini-starkware and liorgold2).

---

crates/starknet_os/src/hints/hint_implementation/execute_transactions/implementation.rs line 67 at r3 (raw file):

}

pub(crate) fn sha2_finalize(ctx: HintContext<'_>) -> OsHintResult {

(why not 256? nonblocking)

Code quote:

sha2_finalize(

---

crates/starknet_os/src/hints/hint_implementation/execute_transactions/implementation.rs line 68 at r3 (raw file):

pub(crate) fn sha2_finalize(ctx: HintContext<'_>) -> OsHintResult {
    let batch_size = &ctx.fetch_const(Const::ShaBatchSize)?.to_bigint();

Add a TODO to rename to Sha256BatchSize

Code quote:

ShaBatchSize

---

crates/starknet_os/src/hints/hint_implementation/execute_transactions/implementation.rs line 113 at r3 (raw file):

    let padding = calculate_sha512_padding(sha512_input_chunk_size_felts, number_of_missing_blocks);
    let sha512_ptr_end = ctx.get_ptr(Ids::Sha512PtrEnd)?;
    ctx.vm.load_data(sha512_ptr_end, &padding)?;

Please share this code with sha256

Code quote:

    let batch_size = &ctx.fetch_const(Const::Sha512BatchSize)?.to_bigint();
    let n = &ctx.get_integer(Ids::N)?.to_bigint();
    let number_of_missing_blocks = ((((-n) % batch_size) + batch_size) % batch_size)
        .to_u32()
        .expect("Failed to convert number of missing blocks to u32.");
    assert!(
        (0..N_MISSING_BLOCKS_BOUND).contains(&number_of_missing_blocks),
        "number_of_missing_blocks: {number_of_missing_blocks} is expected to be in the range [0, \
         {N_MISSING_BLOCKS_BOUND}). Got n: {n} and batch size: {batch_size}."
    );
    let sha512_input_chunk_size_felts =
        felt_to_usize(ctx.fetch_const(Const::Sha512InputChunkSize)?)?;
    assert!(
        (0..SHA256_INPUT_CHUNK_SIZE_BOUND).contains(&sha512_input_chunk_size_felts),
        "sha512_input_chunk_size_felts: {sha512_input_chunk_size_felts} is expected to be in the \
         range [0, {SHA256_INPUT_CHUNK_SIZE_BOUND})."
    );
    let padding = calculate_sha512_padding(sha512_input_chunk_size_felts, number_of_missing_blocks);
    let sha512_ptr_end = ctx.get_ptr(Ids::Sha512PtrEnd)?;
    ctx.vm.load_data(sha512_ptr_end, &padding)?;

[Yoni-Starkware]

@Yoni-Starkware made 1 comment.
Reviewable status: 27 of 65 files reviewed, 5 unresolved discussions (waiting on dorimedini-starkware and liorgold2).

---

crates/starknet_os/src/hints/hint_implementation/execute_transactions/implementation.rs line 107 at r3 (raw file):

        felt_to_usize(ctx.fetch_const(Const::Sha512InputChunkSize)?)?;
    assert!(
        (0..SHA256_INPUT_CHUNK_SIZE_BOUND).contains(&sha512_input_chunk_size_felts),

Should be the same? if so, please add an alias so it would be clear

Suggestion:

SHA512_INPUT_CHUNK_SIZE_BOUND

[Yoni-Starkware]

@Yoni-Starkware made 1 comment.
Reviewable status: 27 of 65 files reviewed, 6 unresolved discussions (waiting on dorimedini-starkware and liorgold2).

---

crates/starknet_os/src/hints/hint_implementation/execute_syscalls.rs line 65 at r3 (raw file):

    ctx.vm.add_relocation_rule(state_ptr, actual_out_state_ptr.into())?;
    Ok(())
}

Please share this code

Code quote:

pub(crate) fn relocate_sha512_segment<S: StateReader>(
    _hint_processor: &mut SnosHintProcessor<'_, S>,
    ctx: HintContext<'_>,
) -> OsHintResult {
    let state_ptr = ctx.get_nested_field_ptr(
        Ids::Response,
        CairoStruct::Sha512ProcessBlockResponsePtr,
        &["state_ptr"],
    )?;
    let actual_out_state_ptr = ctx.get_ptr(Ids::ActualOutState)?;

    // TODO(Nimrod): Use SHA256_STATE_SIZE_FELTS constant.
    let sha_state_size = 8;

    let data = ctx.vm.get_continuous_range(state_ptr, sha_state_size)?;
    ctx.vm.load_data(actual_out_state_ptr, &data)?;

    ctx.vm.add_relocation_rule(state_ptr, actual_out_state_ptr.into())?;
    Ok(())
}

[Yoni-Starkware]

@Yoni-Starkware made 2 comments.
Reviewable status: 27 of 65 files reviewed, 8 unresolved discussions (waiting on dorimedini-starkware and liorgold2).

---

crates/starknet_os/src/hints/hint_implementation/execute_transactions/utils.rs line 24 at r3 (raw file):

];

pub(crate) fn calculate_padding(

Suggestion:

calculate_sha256_padding(

---

crates/starknet_os/src/hints/hint_implementation/execute_transactions/utils.rs line 68 at r3 (raw file):

    }
    padding
}

Please share, at least parts of it

Code quote:

pub(crate) fn calculate_sha512_padding(
    sha512_input_chunk_size_felts: usize,
    number_of_missing_blocks: u32,
) -> Vec<MaybeRelocatable> {
    let message = vec![0_u64; sha512_input_chunk_size_felts];
    let flat_message = sha2::digest::generic_array::GenericArray::from_exact_iter(
        message.iter().flat_map(|v| v.to_be_bytes()),
    )
    .expect("Failed to create a dummy message for sha512_finalize.");
    let mut state = SHA512_IV;
    sha2::compress512(&mut state, &[flat_message]);
    let padding_to_repeat: Vec<u64> =
        [message, SHA512_IV.to_vec(), state.to_vec()].into_iter().flatten().collect();

    let mut padding = vec![];
    let padding_extension =
        padding_to_repeat.iter().map(|x| MaybeRelocatable::from(Felt::from(*x)));
    for _ in 0..number_of_missing_blocks {
        padding.extend(padding_extension.clone());
    }
    padding
}

[Yoni-Starkware]

@Yoni-Starkware reviewed 6 files and made 1 comment.
Reviewable status: 33 of 65 files reviewed, 9 unresolved discussions (waiting on dorimedini-starkware and liorgold2).

---

crates/starknet_os/src/hint_processor/snos_syscall_executor.rs line 385 at r3 (raw file):

        vm.load_data(temp_segment, state)?;
        Ok(temp_segment)
    }

Identical to the above, please share the code or rename the first func

Code quote:

    fn write_sha512_out_state(
        &mut self,
        state: &[MaybeRelocatable],
        vm: &mut VirtualMachine,
    ) -> Result<Relocatable, Self::Error> {
        let temp_segment = vm.add_temporary_segment();
        vm.load_data(temp_segment, state)?;
        Ok(temp_segment)
    }

[dorimedini-starkware]

@dorimedini-starkware made 6 comments.
Reviewable status: 30 of 68 files reviewed, 9 unresolved discussions (waiting on liorgold2 and Yoni-Starkware).

---

crates/blockifier/resources/blockifier_versioned_constants_0_14_3.json line 477 at r3 (raw file):

Previously, Yoni-Starkware (Yoni) wrote…

Not consistent with your slack message

Done.

---

crates/starknet_os/src/hint_processor/snos_syscall_executor.rs line 385 at r3 (raw file):

Previously, Yoni-Starkware (Yoni) wrote…

Identical to the above, please share the code or rename the first func

Done.

---

crates/starknet_os/src/hints/hint_implementation/execute_syscalls.rs line 65 at r3 (raw file):

Previously, Yoni-Starkware (Yoni) wrote…

Please share this code

Done.

---

crates/starknet_os/src/hints/hint_implementation/execute_transactions/implementation.rs line 107 at r3 (raw file):

Previously, Yoni-Starkware (Yoni) wrote…

Should be the same? if so, please add an alias so it would be clear

it's hard-coded as 100 in both cases, so just added another named constant.

---

crates/starknet_os/src/hints/hint_implementation/execute_transactions/implementation.rs line 113 at r3 (raw file):

Previously, Yoni-Starkware (Yoni) wrote…

Please share this code with sha256

Done.

---

crates/starknet_os/src/hints/hint_implementation/execute_transactions/utils.rs line 68 at r3 (raw file):

Previously, Yoni-Starkware (Yoni) wrote…

Please share, at least parts of it

Done.

[dorimedini-starkware]

@dorimedini-starkware made 3 comments and resolved 3 discussions.
Reviewable status: 30 of 68 files reviewed, 6 unresolved discussions (waiting on liorgold2 and Yoni-Starkware).

---

crates/starknet_os/src/hints/hint_implementation/execute_transactions/implementation.rs line 67 at r3 (raw file):

Previously, Yoni-Starkware (Yoni) wrote…

(why not 256? nonblocking)

done, next PR in stack

---

crates/starknet_os/src/hints/hint_implementation/execute_transactions/implementation.rs line 68 at r3 (raw file):

Previously, Yoni-Starkware (Yoni) wrote…

Add a TODO to rename to Sha256BatchSize

done, next PR in stack

---

crates/starknet_os/src/hints/hint_implementation/execute_transactions/utils.rs line 24 at r3 (raw file):

];

pub(crate) fn calculate_padding(

done, next PR in stack

[dorimedini-starkware]

done, and here already

@dorimedini-starkware made 1 comment.
Reviewable status: 30 of 69 files reviewed, 6 unresolved discussions (waiting on liorgold2 and Yoni-Starkware).

[Yoni-Starkware]

@Yoni-Starkware reviewed 18 files and all commit messages, made 1 comment, and resolved 6 discussions.
Reviewable status: 48 of 72 files reviewed, 1 unresolved discussion (waiting on dorimedini-starkware and liorgold2).

---

crates/blockifier_test_utils/resources/feature_contracts/cairo1/test_contract.cairo line 684 at r6 (raw file):

    #[external(v0)]
    fn test_sha512(ref self: ContractState) {

Add a case with a non-trivial array

[dorimedini-starkware]

@dorimedini-starkware resolved 1 discussion.
Reviewable status: 48 of 72 files reviewed, all discussions resolved (waiting on liorgold2).

[Yoni-Starkware]

@Yoni-Starkware reviewed 7 files and all commit messages, and made 2 comments.
Reviewable status: 53 of 72 files reviewed, 1 unresolved discussion (waiting on dorimedini-starkware and liorgold2).

---

crates/blockifier/src/execution/syscalls/vm_syscall_utils.rs line 570 at r6 (raw file):

        Ok(())
    }
}

Duplication of the above. You can just rename Sha256 -> SharedSha and delete this one
nonblocking

Code quote:

// Sha512ProcessBlock syscall.
#[derive(Debug, Eq, PartialEq)]
pub struct Sha512ProcessBlockRequest {
    pub state_ptr: Relocatable,
    pub input_start: Relocatable,
}

impl SyscallRequest for Sha512ProcessBlockRequest {
    fn read(
        vm: &VirtualMachine,
        ptr: &mut Relocatable,
    ) -> SyscallBaseResult<Sha512ProcessBlockRequest> {
        let state_start = vm.get_relocatable(*ptr)?;
        *ptr = (*ptr + 1)?;
        let input_start = vm.get_relocatable(*ptr)?;
        *ptr = (*ptr + 1)?;
        Ok(Sha512ProcessBlockRequest { state_ptr: state_start, input_start })
    }
}

#[derive(Debug, Eq, PartialEq)]
pub struct Sha512ProcessBlockResponse {
    pub state_ptr: Relocatable,
}

impl SyscallResponse for Sha512ProcessBlockResponse {
    fn write(self, vm: &mut VirtualMachine, ptr: &mut Relocatable) -> WriteResponseResult {
        write_maybe_relocatable(vm, ptr, self.state_ptr)?;
        Ok(())
    }
}

[dorimedini-starkware]

@dorimedini-starkware made 1 comment and resolved 1 discussion.
Reviewable status: 52 of 72 files reviewed, all discussions resolved (waiting on liorgold2 and Yoni-Starkware).

---

crates/blockifier/src/execution/syscalls/vm_syscall_utils.rs line 570 at r6 (raw file):

Previously, Yoni-Starkware (Yoni) wrote…

Duplication of the above. You can just rename Sha256 -> SharedSha and delete this one
nonblocking

done in top PR

[Yoni-Starkware]

@Yoni-Starkware reviewed 4 files and all commit messages.
Reviewable status: 54 of 73 files reviewed, all discussions resolved (waiting on liorgold2).

[Yoni-Starkware]

@Yoni-Starkware reviewed 1 file.
Reviewable status: 55 of 74 files reviewed, all discussions resolved (waiting on liorgold2).

[Yoni-Starkware]

@Yoni-Starkware reviewed 1 file and all commit messages.
Reviewable status: 56 of 75 files reviewed, all discussions resolved (waiting on liorgold2).

[Yoni-Starkware]

@Yoni-Starkware made 1 comment.
Reviewable status: 56 of 75 files reviewed, all discussions resolved (waiting on liorgold2).

[dorimedini-starkware]

@dorimedini-starkware reviewed 56 files and all commit messages.
Reviewable status: 56 of 75 files reviewed, all discussions resolved (waiting on liorgold2).

[dorimedini-starkware]

@dorimedini-starkware partially reviewed 25 files.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on liorgold2).