Skip to content

fix comment #7272

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
a10y merged 1 commit into from
Apr 2, 2026
+66 −65
Merged

fix comment #7272

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
65 changes: 0 additions & 65 deletions vortex-array/src/arrays/patched/array.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,71 +27,6 @@ use crate::patches::Patches;
use crate::stats::ArrayStats;
use crate::validity::Validity;

/// An array that partially "patches" another array with new values.
///
/// # Background
///
/// This is meant to be the foundation of a fully data-parallel patching strategy, based on the
/// work published in ["G-ALP" from Hepkema et al.](https://ir.cwi.nl/pub/35205/35205.pdf)
///
/// Patching is common when an encoding almost completely covers an array save a few exceptions.
/// In that case, rather than avoid the encoding entirely, it's preferable to
///
/// * Replace unencodable values with fillers (zeros, frequent values, nulls, etc.)
/// * Wrap the array with a `PatchedArray` signaling that when the original array is executed,
/// some of the decoded values must be overwritten.
///
/// In Vortex, the FastLanes bit-packing encoding is often the terminal node in an encoding tree,
/// and FastLanes has an intrinsic chunking of 1024 elements. Thus, 1024 elements is pervasively
/// a useful unit of chunking throughout Vortex, and so we use 1024 as a chunk size here
/// as well.
///
/// # Details
///
/// To patch an array, we first divide it into a set of chunks of length 1024, and then within
/// each chunk, we assign each position to a lane. The number of lanes depends on the width of
/// the underlying type.
///
/// Thus, rather than sorting patch indices and values by their global offset, they are sorted
/// primarily by their chunk, and then subsequently by their lanes.
///
/// The Patched array layout has 4 children
///
/// * `inner`: the inner array is the one containing encoded values, including the filler values
/// that need to be patched over at execution time
/// * `lane_offsets`: this is an indexing buffer that allows you to see into ranges of the other
/// two children
/// * `indices`: An array of `u16` chunk indices, indicating where within the chunk should the value
/// be overwritten by the patch value
/// * `values`: The child array containing the patch values, which should be inserted over
/// the values of the `inner` at the locations provided by `indices`
///
/// `indices` and `values` are aligned and accessed together.
///
/// ```text
///
/// chunk 0 chunk 0 chunk 0 chunk 0 chunk 0 chunk 0
/// lane 0 lane 1 lane 2 lane 3 lane 4 lane 5
/// ┌────────────┬────────────┬────────────┬────────────┬────────────┬────────────┐
/// lane_offsets │ 0 │ 0 │ 2 │ 2 │ 3 │ 5 │ ...
/// └─────┬──────┴─────┬──────┴─────┬──────┴──────┬─────┴──────┬─────┴──────┬─────┘
/// │ │ │ │ │ │
/// │ │ │ │ │ │
/// ┌─────┴────────────┘ └──────┬──────┘ ┌──────┘ └─────┐
/// │ │ │ │
/// │ │ │ │
/// │ │ │ │
/// ▼────────────┬────────────┬────────────▼────────────▼────────────┬────────────▼
/// indices │ │ │ │ │ │ │
/// │ │ │ │ │ │ │
/// ├────────────┼────────────┼────────────┼────────────┼────────────┼────────────┤
/// values │ │ │ │ │ │ │
/// │ │ │ │ │ │ │
/// └────────────┴────────────┴────────────┴────────────┴────────────┴────────────┘
/// ```
///
/// It turns out that this layout is optimal for executing patching on GPUs, because the
/// `lane_offsets` allows each thread in a warp to seek to its patches in constant time.
/// The inner array containing the base unpatched values.
pub(super) const INNER_SLOT: usize = 0;
/// The lane offsets array for locating patches within lanes.
Expand Down
66 changes: 66 additions & 0 deletions vortex-array/src/arrays/patched/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,72 @@
// SPDX-License-Identifier: Apache-2.0
// SPDX-FileCopyrightText: Copyright the Vortex contributors

//! An array that partially "patches" another array with new values.
//!
//! # Background
//!
//! This is meant to be the foundation of a fully data-parallel patching strategy, based on the
//! work published in ["G-ALP" from Hepkema et al.](https://ir.cwi.nl/pub/35205/35205.pdf)
//!
//! Patching is common when an encoding almost completely covers an array save a few exceptions.
//! In that case, rather than avoid the encoding entirely, it's preferable to
//!
//! * Replace unencodable values with fillers (zeros, frequent values, nulls, etc.)
//! * Wrap the array with a `PatchedArray` signaling that when the original array is executed,
//! some of the decoded values must be overwritten.
//!
//! In Vortex, the FastLanes bit-packing encoding is often the terminal node in an encoding tree,
//! and FastLanes has an intrinsic chunking of 1024 elements. Thus, 1024 elements is pervasively
//! a useful unit of chunking throughout Vortex, and so we use 1024 as a chunk size here
//! as well.
//!
//! # Details
//!
//! To patch an array, we first divide it into a set of chunks of length 1024, and then within
//! each chunk, we assign each position to a lane. The number of lanes depends on the width of
//! the underlying type.
//!
//! Thus, rather than sorting patch indices and values by their global offset, they are sorted
//! primarily by their chunk, and then subsequently by their lanes.
//!
//! The Patched array layout has 4 children
//!
//! * `inner`: the inner array is the one containing encoded values, including the filler values
//! that need to be patched over at execution time
//! * `lane_offsets`: this is an indexing buffer that allows you to see into ranges of the other
//! two children
//! * `indices`: An array of `u16` chunk indices, indicating where within the chunk should the value
//! be overwritten by the patch value
//! * `values`: The child array containing the patch values, which should be inserted over
//! the values of the `inner` at the locations provided by `indices`
//!
//! `indices` and `values` are aligned and accessed together.
//!
//! ```text
//!
//! chunk 0 chunk 0 chunk 0 chunk 0 chunk 0 chunk 0
//! lane 0 lane 1 lane 2 lane 3 lane 4 lane 5
//! ┌────────────┬────────────┬────────────┬────────────┬────────────┬────────────┐
//! lane_offsets │ 0 │ 0 │ 2 │ 2 │ 3 │ 5 │ ...
//! └─────┬──────┴─────┬──────┴─────┬──────┴──────┬─────┴──────┬─────┴──────┬─────┘
//! │ │ │ │ │ │
//! │ │ │ │ │ │
//! ┌─────┴────────────┘ └──────┬──────┘ ┌──────┘ └─────┐
//! │ │ │ │
//! │ │ │ │
//! │ │ │ │
//! ▼────────────┬────────────┬────────────▼────────────▼────────────┬────────────▼
//! indices │ │ │ │ │ │ │
//! │ │ │ │ │ │ │
//! ├────────────┼────────────┼────────────┼────────────┼────────────┼────────────┤
//! values │ │ │ │ │ │ │
//! │ │ │ │ │ │ │
//! └────────────┴────────────┴────────────┴────────────┴────────────┴────────────┘
//! ```
//!
//! It turns out that this layout is optimal for executing patching on GPUs, because the
//! `lane_offsets` allows each thread in a warp to seek to its patches in constant time.

mod array;
mod compute;
mod vtable;
Expand Down
Loading