Skip to content

Offload intrinsic #147936

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

Jump to bottom
Merged
bors merged 2 commits into from
Nov 26, 2025
Merged

Offload intrinsic #147936

bors merged 2 commits into from
Nov 26, 2025

Conversation

@Sa4dUs
Copy link
Contributor

@Sa4dUs Sa4dUs commented Oct 21, 2025 •
edited by ZuseZ4
Loading

This PR implements the minimal mechanisms required to run a small subset of arbitrary offload kernels without relying on hardcoded names or metadata.

  • offload(kernel, (..args)): an intrinsic that generates the necessary host-side LLVM-IR code.
  • rustc_offload_kernel: a builtin attribute that marks device kernels to be handled appropriately.

Example usage (pseudocode):

fn kernel(x: *mut [f64; 128]) {
    core::intrinsics::offload(kernel_1, (x,))
}

#[cfg(target_os = "linux")]
extern "C" {
    pub fn kernel_1(array_b: *mut [f64; 128]);
}

#[cfg(not(target_os = "linux"))]
#[rustc_offload_kernel]
extern "gpu-kernel" fn kernel_1(x: *mut [f64; 128]) {
    unsafe { (*x)[0] = 21.0 };
}

@rustbot rustbot added A-LLVM Area: Code generation parts specific to LLVM. Both correctness bugs and optimization-related issues. S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. T-libs Relevant to the library team, which will review and decide on the PR/issue. labels Oct 21, 2025
@ZuseZ4 ZuseZ4 self-assigned this Oct 21, 2025
@Sa4dUs Sa4dUs force-pushed the offload-intrinsic branch from 9118683 to 23722aa Compare October 21, 2025 19:45
@rust-log-analyzer

This comment has been minimized.

Sign in to view
@ZuseZ4 ZuseZ4 added the F-gpu_offload `#![feature(gpu_offload)]` label Oct 22, 2025
ZuseZ4
ZuseZ4 reviewed Oct 24, 2025
View reviewed changes
compiler/rustc_middle/src/ty/offload_meta.rs Outdated
}

pub fn from_ty<'tcx>(tcx: TyCtxt<'tcx>, ty: Ty<'tcx>) -> Self {
OffloadMetadata { payload_size: get_payload_size(tcx, ty), mode: TransferKind::Both }
Copy link
Member

@ZuseZ4 ZuseZ4 Oct 24, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If you already have the code here, I would add a small check for & or byVal (implies Mode ToGPU), vs &mut (implies Both).

In the future we would hope to analyze the & or byval case more, if we never read from it (before writing) then we could use a new mode 4, which allocates directly on the gpu.

@ZuseZ4 ZuseZ4 mentioned this pull request Oct 24, 2025
Open
5 tasks
@rust-log-analyzer

This comment has been minimized.

Sign in to view
@bors
Copy link
Collaborator

bors commented Nov 5, 2025

☔ The latest upstream changes (presumably #148507) made this pull request unmergeable. Please resolve the merge conflicts.

@Sa4dUs Sa4dUs force-pushed the offload-intrinsic branch from e0fd7be to 97a8e96 Compare November 7, 2025 15:37
@rust-log-analyzer

This comment has been minimized.

Sign in to view
@bors
Copy link
Collaborator

bors commented Nov 9, 2025

☔ The latest upstream changes (presumably #148721) made this pull request unmergeable. Please resolve the merge conflicts.

@rustbot rustbot added the A-attributes Area: Attributes (`#[…]`, `#![…]`) label Nov 11, 2025
ZuseZ4
ZuseZ4 reviewed Nov 14, 2025
View reviewed changes
ZuseZ4
ZuseZ4 reviewed Nov 14, 2025
View reviewed changes
ZuseZ4
ZuseZ4 reviewed Nov 14, 2025
View reviewed changes
@rust-log-analyzer

This comment has been minimized.

Sign in to view
@rust-log-analyzer

This comment has been minimized.

Sign in to view
@rust-log-analyzer

This comment has been minimized.

Sign in to view
@Sa4dUs Sa4dUs marked this pull request as ready for review November 16, 2025 10:27
@rustbot
Copy link
Collaborator

rustbot commented Nov 16, 2025

Some changes occurred to the intrinsics. Make sure the CTFE / Miri interpreter
gets adapted for the changes, if necessary.

cc @rust-lang/miri, @RalfJung, @oli-obk, @lcnr

@rustbot rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Nov 16, 2025
@bors
Copy link
Collaborator

bors commented Nov 18, 2025

☔ The latest upstream changes (presumably #148151) made this pull request unmergeable. Please resolve the merge conflicts.

@rustbot
Copy link
Collaborator

rustbot commented Nov 22, 2025

This PR was rebased onto a different main commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.

@rust-log-analyzer

This comment has been minimized.

Sign in to view
@Sa4dUs Sa4dUs force-pushed the offload-intrinsic branch 2 times, most recently from c39e4e5 to ce5970c Compare November 22, 2025 18:19
@ZuseZ4
Copy link
Member

ZuseZ4 commented Nov 25, 2025

This removes a good amount of the hacks from my first MVP, further improvements can land in a follow-up PR.
Verified to work on an MI 250X.

@bors r+

@bors
Copy link
Collaborator

bors commented Nov 25, 2025

📌 Commit f39ec47 has been approved by ZuseZ4

It is now in the queue for this repository.

@bors bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Nov 25, 2025
@jhpratt jhpratt mentioned this pull request Nov 26, 2025
Closed
@Zalathar Zalathar mentioned this pull request Nov 26, 2025
Merged
@bors bors merged commit 2b150f2 into rust-lang:main Nov 26, 2025
11 checks passed
@rustbot rustbot added this to the 1.93.0 milestone Nov 26, 2025
@Zalathar
Copy link
Member

Zalathar commented Nov 27, 2025

Perf results from rollup:

Some regressions in the large-workspace secondary benchmark, perhaps due to the additional metadata.

Zalathar
Zalathar reviewed Nov 27, 2025
View reviewed changes
compiler/rustc_codegen_llvm/src/back/write.rs
Comment on lines 707 to 708
// For now we only support up to 10 kernels named kernel_0 ... kernel_9, a follow-up PR is
// introducing a proper offload intrinsic to solve this limitation.
Copy link
Member

@Zalathar Zalathar Nov 27, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks like this comment is now outdated as of this PR.

Copy link
Member

@ZuseZ4 ZuseZ4 Nov 27, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks. We have a few follow-up PRs where I'll add it.

@ZuseZ4
Copy link
Member

ZuseZ4 commented Nov 27, 2025 •
edited
Loading

@Zalathar We add metadata and do work if and only if you set -Zoffload=Enable flag, which isn't done in that benchmark.
if cgcx.target_is_like_gpu && config.offload.contains(&config::Offload::Enable)

We also introduce one more intrinsic, so I guess the match arm against intrinsic get's slightly larger in compiler/rustc_codegen_llvm/src/intrinsic.rs. Maybe the large-workspace has a lot of intrinsics, so the slightly larger code size for the match arm here causes a perf impact?

@Mark-Simulacrum Mark-Simulacrum added the perf-regression Performance regression. label Dec 3, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@RalfJung RalfJung RalfJung left review comments

@Zalathar Zalathar Zalathar left review comments

@ZuseZ4 ZuseZ4 ZuseZ4 left review comments

Assignees

@ZuseZ4 ZuseZ4

Labels

A-attributes Area: Attributes (`#[…]`, `#![…]`) A-LLVM Area: Code generation parts specific to LLVM. Both correctness bugs and optimization-related issues. A-rustc-dev-guide Area: rustc-dev-guide F-gpu_offload `#![feature(gpu_offload)]` perf-regression Performance regression. S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. T-libs Relevant to the library team, which will review and decide on the PR/issue.

Projects

None yet

Milestone

1.93.0

Development

Successfully merging this pull request may close these issues.

8 participants

@Sa4dUs @rust-log-analyzer @bors @rustbot @ZuseZ4 @Zalathar @RalfJung @Mark-Simulacrum