Add interaction_element_ids to Component trait.

[alonh5]

This change is 

[alonh5]

• Add wide fib lookup step constraint. #646
• Extract constraint code into functions. #644
• Use SecureField in interaction. #642
• Modify mask_points for interaction. #639
• Modify trace_log_degree_bounds for interaction. #638
• Add wide fib lookup constraint. #637
• Add interaction to proof. #636
• Split prove function. #634
• Create ComponentTraceWriter trait. #633
• Add interaction_element_ids to Component trait. #631 👈
• Remove wide fibonacci boundary constraint. #582 : 1 other dependent PR (#589 )
• Separate cpu and avx wide fib. #573 : 1 other dependent PR (#574 )
• Get lookup columns. #572
• Configure wide fib. #562
• dev

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

Join @alonh5 and the rest of your teammates on Graphite

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 0% with 11 lines in your changes are missing coverage. Please review.

Project coverage is 92.52%. Comparing base (8d16c06) to head (a976f53).

Files	Patch %	Lines
crates/prover/src/core/air/air_ext.rs	0.00%	8 Missing ⚠️
crates/prover/src/core/prover/mod.rs	0.00%	3 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##              dev     #631      +/-   ##
==========================================
- Coverage   92.64%   92.52%   -0.12%     
==========================================
  Files          69       69              
  Lines        8995     9006      +11     
  Branches     8995     9006      +11     
==========================================
  Hits         8333     8333              
- Misses        590      601      +11     
  Partials       72       72              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[shaharsamocha7]

Reviewed 6 of 6 files at r1, all commit messages.
Reviewable status: all files reviewed, 2 unresolved discussions (waiting on @alonh5 and @andrewmilson)

---

crates/prover/src/core/mod.rs line 63 at r1 (raw file):

}

pub struct InteractionElements(Vec<(String, BaseField)>);

This should be SecureField

Code quote:

BaseField

---

crates/prover/src/core/air/air_ext.rs line 53 at r1 (raw file):

                .dedup()
                .map(|id| {
                    let element = channel.draw_felt().0 .0;

Consider to draw all the elements together.
I think we have draw_felts function
Each channel hash could give you more then one element

Code quote:

channel.draw_felt().0

[alonh5]

Reviewable status: all files reviewed, 2 unresolved discussions (waiting on @andrewmilson and @shaharsamocha7)

---

crates/prover/src/core/mod.rs line 63 at r1 (raw file):

Previously, shaharsamocha7 wrote…

This should be SecureField

Right. I did this on purpose to add a constraint first more easily, and I plan to change it to a SecureField on top of this stack (there's also a TODO for it).

---

crates/prover/src/core/air/air_ext.rs line 53 at r1 (raw file):

Previously, shaharsamocha7 wrote…

Consider to draw all the elements together.
I think we have draw_felts function
Each channel hash could give you more then one element

Done.

[shaharsamocha7]

Please wait for @andrewmilson review

Reviewed 1 of 1 files at r2, 2 of 2 files at r3, all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @andrewmilson)

[andrewmilson]

Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @alonh5)

---

crates/prover/src/core/mod.rs line 63 at r3 (raw file):

}

pub struct InteractionElements(Vec<(String, BaseField)>);

Shouldn't this be SecureField?

Code quote:

BaseField

[alonh5]

Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @andrewmilson)

---

crates/prover/src/core/mod.rs line 63 at r3 (raw file):

Previously, andrewmilson (Andrew Milson) wrote…

Shouldn't this be SecureField?

Right. I did this on purpose to add a constraint first more easily, and I plan to change it to a SecureField on top of this stack (there's also a TODO for it).

[andrewmilson]

Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @alonh5)

---

crates/prover/src/core/air/mod.rs line 44 at r3 (raw file):

    ) -> ColumnVec<Vec<CirclePoint<SecureField>>>;

    // Returns the ids of the interaction elements used by the component.

Suggestion:

///

[alonh5]

Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @andrewmilson)

---

crates/prover/src/core/air/mod.rs line 44 at r3 (raw file):

    ) -> ColumnVec<Vec<CirclePoint<SecureField>>>;

    // Returns the ids of the interaction elements used by the component.

Done.

[andrewmilson]

Reviewed 2 of 6 files at r1, 2 of 2 files at r3, all commit messages.
Reviewable status: 5 of 6 files reviewed, 1 unresolved discussion (waiting on @alonh5 and @shaharsamocha7)

---

crates/prover/src/core/air/air_ext.rs line 55 at r4 (raw file):

        let elements = channel.draw_felts(ids.len()).into_iter().map(|e| e.0 .0);
        InteractionElements(zip_eq(ids, elements).collect_vec())
    }

Optional: could use BTreeSet<String> for interaction_element_ids

Suggestion:

    fn interaction_elements(&self, channel: &mut Blake2sChannel) -> InteractionElements {
        let mut ids = BTreeSet::new();
        
        for component in self.components() {
            ids = &ids | &component.interaction_element_ids();
        }
        
        let elements = channel.draw_felts(ids.len()).into_iter().map(|e| e.0 .0);
        InteractionElements(zip_eq(ids, elements).collect_vec())
    }

[alonh5]

Reviewable status: 4 of 6 files reviewed, 1 unresolved discussion (waiting on @andrewmilson and @shaharsamocha7)

---

crates/prover/src/core/air/air_ext.rs line 55 at r4 (raw file):

Previously, andrewmilson (Andrew Milson) wrote…

Optional: could use BTreeSet<String> for interaction_element_ids

Done.