Pass entire mask to components

[andrewmilson]

This change is 

[andrewmilson]

• Create blake example that uses GKR for lookups #807
• Create MLE eval verifier component #805
• Create MLE eval prover component #804
• Add build trace functions for MLE eval component #803
• Remove InteractionElements and LookupValues #802
• Pass entire mask to components #801 👈
• Add MleCollection for accumulating MLEs #800
• Combine eq step constraints into single constraint #799
• Move all MLE eval related constraints into same module #798
• dev

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

Join @andrewmilson and the rest of your teammates on Graphite

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 90.00000% with 12 lines in your changes missing coverage. Please review.

Project coverage is 92.18%. Comparing base (1003bf2) to head (bc7f1e0).
Report is 1 commits behind head on dev.

Files	Patch %	Lines
crates/prover/src/core/prover/mod.rs	74.07%	2 Missing and 5 partials ⚠️
...rates/prover/src/constraint_framework/component.rs	94.23%	1 Missing and 2 partials ⚠️
crates/prover/src/core/pcs/utils.rs	88.23%	0 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##              dev     #801      +/-   ##
==========================================
- Coverage   92.25%   92.18%   -0.08%     
==========================================
  Files          91       90       -1     
  Lines       12484    12473      -11     
  Branches    12484    12473      -11     
==========================================
- Hits        11517    11498      -19     
- Misses        861      864       +3     
- Partials      106      111       +5     

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

[spapinistarkware]

Reviewed 20 of 20 files at r2, all commit messages.
Reviewable status: all files reviewed, 5 unresolved discussions (waiting on @andrewmilson and @shaharsamocha7)

---

crates/prover/src/constraint_framework/component.rs line 28 at r2 (raw file):

// TODO(andrew): Consider better location for this.
#[derive(Debug, Default)]
pub struct TreeColumnSpanProvider {

WDYT?
And maybe We should rename TreeColumnSpan to TraceLocation?

Suggestion:

TraceLocationAllocator

---

crates/prover/src/constraint_framework/component.rs line 30 at r2 (raw file):

pub struct TreeColumnSpanProvider {
    /// Mapping of tree index to next available column offset.
    next_tree_offsets: BTreeMap<usize, usize>,

Why not TreeVec?

Suggestion:

TreeVec<usize>

---

crates/prover/src/constraint_framework/component.rs line 35 at r2 (raw file):

impl TreeColumnSpanProvider {
    fn next_for_structure<T>(&mut self, structure: &TreeVec<ColumnVec<T>>) -> Vec<TreeColumnSpan> {
        structure

    When next_tree_offsets is TreeVec:

Suggestion:

impl TreeColumnSpanProvider {
    fn next_for_structure<T>(&mut self, structure: &TreeVec<ColumnVec<T>>) -> TreeVec<TreeColumnSpan> {
        self.next_tree_offsets.resize(...)
        self.next_tree_offsets.as_mut().zip(structure).map(|(off, cols)|{
        ...
        })
`

---

crates/prover/src/core/prover/mod.rs line 76 at r2 (raw file):

    let commitment_scheme_proof = commitment_scheme.prove_values(sample_points, channel);

    // TODO(spapini): Save clone.

I don't think there's a clone here, right? Can you remove?

---

crates/prover/src/core/prover/mod.rs line 155 at r2 (raw file):

    let mut composition_cols = mask.last().into_iter().flatten();

    let col0 = &**composition_cols.next().ok_or(InvalidOodsSampleStructure)?;

Can you do this with std::array::from_fn?

[andrewmilson]

Reviewable status: all files reviewed, 5 unresolved discussions (waiting on @shaharsamocha7 and @spapinistarkware)

---

crates/prover/src/constraint_framework/component.rs line 28 at r2 (raw file):

Previously, spapinistarkware (Shahar Papini) wrote…

WDYT?
And maybe We should rename TreeColumnSpan to TraceLocation?

Updated. Personally prefer TreeColumnSpan since the trace location is made up of multiple WDYT?

---

crates/prover/src/constraint_framework/component.rs line 30 at r2 (raw file):

Previously, spapinistarkware (Shahar Papini) wrote…

Why not TreeVec?

Done.

---

crates/prover/src/constraint_framework/component.rs line 35 at r2 (raw file):

Previously, spapinistarkware (Shahar Papini) wrote…

When next_tree_offsets is TreeVec:

Done. Seemed simpler to use regular iterators though because of needing the index.

---

crates/prover/src/core/prover/mod.rs line 76 at r2 (raw file):

Previously, spapinistarkware (Shahar Papini) wrote…

I don't think there's a clone here, right? Can you remove?

Done.

---

crates/prover/src/core/prover/mod.rs line 155 at r2 (raw file):

Previously, spapinistarkware (Shahar Papini) wrote…

Can you do this with std::array::from_fn?

Done.

[spapinistarkware]

Reviewed 8 of 8 files at r3, all commit messages.
Reviewable status: all files reviewed, 2 unresolved discussions (waiting on @andrewmilson and @shaharsamocha7)

---

crates/prover/src/constraint_framework/component.rs line 28 at r2 (raw file):

Previously, andrewmilson (Andrew Milson) wrote…

Updated. Personally prefer TreeColumnSpan since the trace location is made up of multiple WDYT?

I'm just afraid that new people looking at this code will think it's Jibrish...
I'm open to new names.
Maybe TraceSubspan ?

---

crates/prover/src/core/pcs/mod.rs line 28 at r3 (raw file):

}

// #[derive(Debug, Clone, PartialEq, Eq)]

Remove?

[andrewmilson]

Reviewable status: 16 of 22 files reviewed, 1 unresolved discussion (waiting on @shaharsamocha7 and @spapinistarkware)

---

crates/prover/src/constraint_framework/component.rs line 28 at r2 (raw file):

Previously, spapinistarkware (Shahar Papini) wrote…

I'm just afraid that new people looking at this code will think it's Jibrish...
I'm open to new names.
Maybe TraceSubspan ?

Done

---

crates/prover/src/core/pcs/mod.rs line 28 at r3 (raw file):

Previously, spapinistarkware (Shahar Papini) wrote…

Remove?

Done

[spapinistarkware]

Reviewed 6 of 6 files at r4, all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @shaharsamocha7)