blockifier: update blake estimation for empty input

[avivg-starkware]

No description provided.

[reviewable-StarkWare]

This change is 

[avivg-starkware]

• blockifier: update blake estimation for empty input #8323 👈 (View in Graphite)
• main

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

[github-actions]

Artifacts upload workflows:

• Started at 2025-07-30T13:35:10Z
• Started at 2025-07-30T14:05:44Z
• Started at 2025-07-30T14:13:27Z
• Started at 2025-07-30T14:15:45Z
• Started at 2025-07-30T14:41:05Z
• Started at 2025-07-31T06:44:20Z
• Started at 2025-07-31T06:45:24Z
• Started at 2025-07-31T08:26:56Z

[AvivYossef-starkware]

Reviewed 3 of 3 files at r2, 2 of 2 files at r3, all commit messages.
Reviewable status: all files reviewed, 3 unresolved discussions (waiting on @noaov1)

---

crates/blockifier/src/execution/execution_utils_test.rs line 34 at r2 (raw file):

#[test]
fn test_zero_inputs() {
    // TODO(AvivG): Re-check this case in VM — input 0 was previously invalid when this estimation

Can you remove it now?

Code quote:

// TODO(AvivG): Re-check this case in VM — input 0 was previously invalid when this estimation

---

crates/blockifier/src/execution/execution_utils.rs line 449 at r2 (raw file):

    if total_u32s == 0 {
        return blake_estimation::STEPS_EMPTY_INPUT;
    }

if total_u32s == 0 {
return blake_estimation::STEPS_EMPTY_INPUT;
}

Suggestion:

    if total_u32s == 0 {
        // The empty input case is a special case.  
        return blake_estimation::STEPS_EMPTY_INPUT;
    }   

---

crates/blockifier/src/execution/execution_utils.rs line 472 at r2 (raw file):

    let n_felts =
        n_big_felts.checked_add(n_small_felts).expect("Overflow computing total number of felts");
    let builtin_instance_counter = if n_felts == 0 {

Personally, I prefer match here (non-blocking)

---

crates/blockifier/src/execution/execution_utils.rs line 473 at r2 (raw file):

        n_big_felts.checked_add(n_small_felts).expect("Overflow computing total number of felts");
    let builtin_instance_counter = if n_felts == 0 {
        HashMap::new()

Suggestion:

    let builtin_instance_counter = if n_felts == 0 {
        // The empty case does not use builtins at all.
        HashMap::new()

---

crates/blockifier/src/execution/execution_utils.rs line 475 at r2 (raw file):

        HashMap::new()
    } else {
        // One `range_check` per input felt to validate its size + Overhead.

Suggestion:

// One `range_check` per input felt to validate its size + Overhead for the non empty case..

---

crates/blockifier/src/execution/execution_utils.rs line 488 at r2 (raw file):

/// This includes:
/// - ExecutionResources gas.
/// - Blake opcode gas.

What do you mean here?
Why is the Blake opcode gas separated from the others?

Code quote:

/// This includes:
/// - ExecutionResources gas.
/// - Blake opcode gas.

---

crates/blockifier/src/execution/execution_utils.rs line 514 at r2 (raw file):

             This breaks the assumption that builtin costs are identical between provers.",
        );
    }

I prefer something like it.

Suggestion:

assert!(
    vm_resources
        .builtin_instance_counter
        .keys()
        .all(|&k| k == BuiltinName::range_check),
    "Expected either empty builtins or only `range_check` builtin, got: {:?}. \
     This breaks the assumption that builtin costs are identical between provers.",
    vm_resources.builtin_instance_counter.keys().collect::<Vec<_>>()
);

[avivg-starkware]

Reviewable status: 1 of 3 files reviewed, 3 unresolved discussions (waiting on @AvivYossef-starkware and @noaov1)

---

crates/blockifier/src/execution/execution_utils.rs line 449 at r2 (raw file):

Previously, AvivYossef-starkware wrote…

if total_u32s == 0 {
return blake_estimation::STEPS_EMPTY_INPUT;
}

Added

---

crates/blockifier/src/execution/execution_utils.rs line 472 at r2 (raw file):

Previously, AvivYossef-starkware wrote…

Personally, I prefer match here (non-blocking)

This way?

---

crates/blockifier/src/execution/execution_utils.rs line 488 at r2 (raw file):

Previously, AvivYossef-starkware wrote…

What do you mean here?
Why is the Blake opcode gas separated from the others?

Is it clearer now? Does the modification to the comment answer your question?

---

crates/blockifier/src/execution/execution_utils.rs line 514 at r2 (raw file):

Previously, AvivYossef-starkware wrote…

I prefer something like it.

no prob, replaced

---

crates/blockifier/src/execution/execution_utils_test.rs line 34 at r2 (raw file):

Previously, AvivYossef-starkware wrote…

Can you remove it now?

definatly

---

crates/blockifier/src/execution/execution_utils.rs line 473 at r2 (raw file):

        n_big_felts.checked_add(n_small_felts).expect("Overflow computing total number of felts");
    let builtin_instance_counter = if n_felts == 0 {
        HashMap::new()

Added

---

crates/blockifier/src/execution/execution_utils.rs line 475 at r2 (raw file):

        HashMap::new()
    } else {
        // One `range_check` per input felt to validate its size + Overhead.

Changed.

[AvivYossef-starkware]

Reviewed 2 of 2 files at r4, all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @noaov1)

---

crates/blockifier/src/execution/execution_utils.rs line 472 at r2 (raw file):

Previously, avivg-starkware wrote…

This way?

Yes

---

crates/blockifier/src/execution/execution_utils.rs line 488 at r2 (raw file):

Previously, avivg-starkware wrote…

Is it clearer now? Does the modification to the comment answer your question?

Why is Blake's opcode not part of the ExecutionResources?
Is it a problem of the VM?

[avivg-starkware]

Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @noaov1)

---

crates/blockifier/src/execution/execution_utils.rs line 488 at r2 (raw file):

Previously, AvivYossef-starkware wrote…

Why is Blake's opcode not part of the ExecutionResources?
Is it a problem of the VM?

--- continued discussion via Slack

[noaov1]

Reviewed 1 of 2 files at r3, 2 of 2 files at r4, all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @avivg-starkware)