Still another array access performance issue (in debug)

Zig Version

0.11.0-dev.3105+e46d7a369

Steps to Reproduce and Observed Behavior

This is a follow-up to #13938 which was only fixed in release mode(which means I still have to rely on the workaround to be able to run my code in debug mode). The reproduction code is basically identical:

const std = @import("std");

const len = 32*32*32;

fn getIndex(i: u16) u16 {
	return i;
}

pub const Chunk = struct {
	blocks: [len]u16 = undefined,
};

pub noinline fn regenerateMainMesh(chunk: *Chunk) u32 {
	var sum: u32 = 0;
	var i: u16 = 0;
	while(i < len) : (i += 1) {
		sum += chunk.blocks[getIndex(i)]; // ← workaround: (&chunk.blocks)[...]
	}
	return sum;
}

pub fn main() void {
	var chunk: Chunk = Chunk{};
	for(&chunk.blocks, 0..) |*block, i| {
		block.* = @intCast(i);
	}
	const start = std.time.nanoTimestamp();
	const sum = regenerateMainMesh(&chunk);
	const end = std.time.nanoTimestamp();
	std.log.err("Time: {} Sum: {}", .{end - start, sum});
}

Running this in debug mode is taking 93 ms for iterating over a 32768 element array:

$ zig run test.zig
error: Time: 93325090 Sum: 536854528

Running this in a profiler, I can see that the bottleneck still is a memcpy that is called once every iteration and appears to be copying the entire array:

Expected Behavior

When applying the workaround

- sum += chunk.blocks[getIndex(i)]; // ← workaround: (&chunk.blocks)[...]
+ sum += (&chunk.blocks)[getIndex(i)]; // ← workaround: (&chunk.blocks)[...]

the performance is significantly better, taking only 0.4 ms instead of 93 ms:

$ zig run test.zig
error: Time: 357398 Sum: 536854528

I just updated my zig version and the problem got a lot worse.

EDIT: I updated again and it is now back to how it was originally.

Expand

This simplified example was not affected in the past:

const std = @import("std");

const len = 32*32*32;

var blocks: [len]u16 = undefined;

pub noinline fn regenerateMainMesh() u32 {
	var sum: u32 = 0;
	var i: u16 = 0;
	while(i < len) : (i += 1) {
		sum += blocks[i]; // ← workaround: (&blocks)[...]
	}
	return sum;
}

pub fn main() void {
	for(&blocks, 0..) |*block, i| {
		block.* = @intCast(u16, i);
	}
	const start = std.time.nanoTimestamp();
	const sum = regenerateMainMesh();
	const end = std.time.nanoTimestamp();
	std.log.err("Time: {} Sum: {}", .{end - start, sum});
}

But now it is also slow:

$ zig version
0.11.0-dev.3380+7e0a02ee2
$ zig run test.zig
error: Time: 98298061 Sum: 536854528

$ ~/Downloads/zig-old/zig version
0.11.0-dev.3132+465272921
$ ~/Downloads/zig-old/zig run test.zig
error: Time: 632556 Sum: 536854528

Fix landed in 2ab7893.

Revert "Don't assume a write if an operand is not in function paramet…

Fix reverted in 9d5a133.

Revert "Don't assume a write if an operand is not in function paramet…

AstGen: avoid intermediate loads during field/array access

fixed by #25154

Navigation Menu

Repository navigation

Still another array access performance issue (in debug) #15685

Zig Version

Steps to Reproduce and Observed Behavior

Expected Behavior

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Still another array access performance issue (in debug) #15685

Description

Zig Version

Steps to Reproduce and Observed Behavior

Expected Behavior

Activity

IntegratedQuantum commented on Jun 6, 2023

andrewrk commented on Jan 24, 2024

andrewrk commented on Jan 24, 2024

IntegratedQuantum commented on Sep 21, 2025

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions