Skip to content

feat(intrinsics): add CTA barrier reductions #1138

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

Closed
almightychang wants to merge 3 commits into from
Closed

feat(intrinsics): add CTA barrier reductions #1138

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
5465e19
feat(intrinsics): add CTA barrier reductions
almightychang Aug 22, 2026
459dd93
test(examples): compile CTA barrier reductions
almightychang Aug 22, 2026
fc18327
docs(book): document CTA barrier reductions
almightychang Aug 22, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
1 change: 1 addition & 0 deletions crates/cuda-device/src/barrier.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,6 +105,7 @@ impl Barrier {
include!("generated/mbarrier_basic.rs");
include!("generated/mbarrier_extended.rs");
include!("generated/counted_barrier.rs");
include!("generated/barrier_reduction.rs");

// =============================================================================
// Barrier Arrive Operations
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/async_copy.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::async_copy` to keep existing paths stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/atomic.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::atomic` to keep existing paths stable.
Expand Down
160 changes: 160 additions & 0 deletions crates/cuda-device/src/generated/barrier_reduction.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,160 @@
/*
* SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
* SPDX-License-Identifier: Apache-2.0
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::barrier`.

/// Reduces predicates with AND while synchronizing all arriving CTA threads at an aligned numbered barrier.
///
/// # Safety
/// All participating threads must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_red_and_aligned_all(barrier_id: u32, predicate: bool) -> bool {
let _ = (barrier_id, predicate);
unreachable!("barrier_cta_red_and_aligned_all called outside CUDA kernel context")
}

/// Reduces predicates with AND while synchronizing the requested number of CTA threads at an aligned numbered barrier.
///
/// # Safety
/// All participating threads must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_red_and_aligned_count(
barrier_id: u32,
thread_count: u32,
predicate: bool,
) -> bool {
let _ = (barrier_id, predicate);
let _ = thread_count;
unreachable!("barrier_cta_red_and_aligned_count called outside CUDA kernel context")
}

/// Reduces predicates with AND while synchronizing all arriving CTA threads at a numbered barrier.
///
/// # Safety
/// All participating threads must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_red_and_all(barrier_id: u32, predicate: bool) -> bool {
let _ = (barrier_id, predicate);
unreachable!("barrier_cta_red_and_all called outside CUDA kernel context")
}

/// Reduces predicates with AND while synchronizing the requested number of CTA threads at a numbered barrier.
///
/// # Safety
/// All participating threads must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_red_and_count(
barrier_id: u32,
thread_count: u32,
predicate: bool,
) -> bool {
let _ = (barrier_id, predicate);
let _ = thread_count;
unreachable!("barrier_cta_red_and_count called outside CUDA kernel context")
}

/// Reduces predicates with OR while synchronizing all arriving CTA threads at an aligned numbered barrier.
///
/// # Safety
/// All participating threads must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_red_or_aligned_all(barrier_id: u32, predicate: bool) -> bool {
let _ = (barrier_id, predicate);
unreachable!("barrier_cta_red_or_aligned_all called outside CUDA kernel context")
}

/// Reduces predicates with OR while synchronizing the requested number of CTA threads at an aligned numbered barrier.
///
/// # Safety
/// All participating threads must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_red_or_aligned_count(
barrier_id: u32,
thread_count: u32,
predicate: bool,
) -> bool {
let _ = (barrier_id, predicate);
let _ = thread_count;
unreachable!("barrier_cta_red_or_aligned_count called outside CUDA kernel context")
}

/// Reduces predicates with OR while synchronizing all arriving CTA threads at a numbered barrier.
///
/// # Safety
/// All participating threads must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_red_or_all(barrier_id: u32, predicate: bool) -> bool {
let _ = (barrier_id, predicate);
unreachable!("barrier_cta_red_or_all called outside CUDA kernel context")
}

/// Reduces predicates with OR while synchronizing the requested number of CTA threads at a numbered barrier.
///
/// # Safety
/// All participating threads must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_red_or_count(
barrier_id: u32,
thread_count: u32,
predicate: bool,
) -> bool {
let _ = (barrier_id, predicate);
let _ = thread_count;
unreachable!("barrier_cta_red_or_count called outside CUDA kernel context")
}

/// Counts true predicates while synchronizing all arriving CTA threads at an aligned numbered barrier.
///
/// # Safety
/// All participating threads must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_red_popc_aligned_all(barrier_id: u32, predicate: bool) -> u32 {
let _ = (barrier_id, predicate);
unreachable!("barrier_cta_red_popc_aligned_all called outside CUDA kernel context")
}

/// Counts true predicates while synchronizing the requested number of CTA threads at an aligned numbered barrier.
///
/// # Safety
/// All participating threads must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_red_popc_aligned_count(
barrier_id: u32,
thread_count: u32,
predicate: bool,
) -> u32 {
let _ = (barrier_id, predicate);
let _ = thread_count;
unreachable!("barrier_cta_red_popc_aligned_count called outside CUDA kernel context")
}

/// Counts true predicates while synchronizing all arriving CTA threads at a numbered barrier.
///
/// # Safety
/// All participating threads must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_red_popc_all(barrier_id: u32, predicate: bool) -> u32 {
let _ = (barrier_id, predicate);
unreachable!("barrier_cta_red_popc_all called outside CUDA kernel context")
}

/// Counts true predicates while synchronizing the requested number of CTA threads at a numbered barrier.
///
/// # Safety
/// All participating threads must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_red_popc_count(
barrier_id: u32,
thread_count: u32,
predicate: bool,
) -> u32 {
let _ = (barrier_id, predicate);
let _ = thread_count;
unreachable!("barrier_cta_red_popc_count called outside CUDA kernel context")
}
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/bf16.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::bf16` to keep existing paths stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/bf16x2.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::bf16x2` to keep existing paths stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/clc.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::clc` to keep its public API stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/cluster_barrier.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::cluster` to keep public paths stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/cluster_memory.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::cluster` to keep its public API stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/cluster_sreg.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// This file is included inside `cuda_device::cluster`.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/convert.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::convert`.
Expand Down
20 changes: 15 additions & 5 deletions crates/cuda-device/src/generated/counted_barrier.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,15 +4,15 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::barrier` to keep its public API stable.

/// Signals arrival of the requested number of CTA threads at a numbered barrier without waiting.
///
/// # Safety
/// Every participating thread must use a compatible barrier ID and expected thread count.
/// Every participating thread must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_arrive(barrier_id: u32, thread_count: u32) {
let _ = (barrier_id, thread_count);
Expand All @@ -22,7 +22,7 @@ pub unsafe fn barrier_cta_arrive(barrier_id: u32, thread_count: u32) {
/// Signals arrival of the requested number of CTA threads at an aligned numbered barrier without waiting.
///
/// # Safety
/// Every participating thread must use a compatible barrier ID and expected thread count.
/// Every participating thread must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_arrive_aligned(barrier_id: u32, thread_count: u32) {
let _ = (barrier_id, thread_count);
Expand All @@ -32,7 +32,7 @@ pub unsafe fn barrier_cta_arrive_aligned(barrier_id: u32, thread_count: u32) {
/// Synchronizes the requested number of CTA threads at a numbered barrier.
///
/// # Safety
/// Every participating thread must use a compatible barrier ID and expected thread count.
/// Every participating thread must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_sync(barrier_id: u32, thread_count: u32) {
let _ = (barrier_id, thread_count);
Expand All @@ -42,9 +42,19 @@ pub unsafe fn barrier_cta_sync(barrier_id: u32, thread_count: u32) {
/// Synchronizes the requested number of CTA threads at an aligned numbered barrier.
///
/// # Safety
/// Every participating thread must use a compatible barrier ID and expected thread count.
/// Every participating thread must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_sync_aligned(barrier_id: u32, thread_count: u32) {
let _ = (barrier_id, thread_count);
unreachable!("barrier_cta_sync_aligned called outside CUDA kernel context")
}

/// Synchronizes all arriving CTA threads at a numbered barrier.
///
/// # Safety
/// Every participating thread must use a compatible barrier protocol.
#[inline(never)]
pub unsafe fn barrier_cta_sync_all(barrier_id: u32) {
let _ = barrier_id;
unreachable!("barrier_cta_sync_all called outside CUDA kernel context")
}
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/debug_control.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::debug` to keep its public API stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/debug_sreg.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::debug` to keep existing paths stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/dotprod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::dotprod` to keep existing paths stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/f16.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::f16` to keep existing paths stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/f16x2.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::f16x2` to keep existing paths stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/f32x2.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::f32x2` to keep existing paths stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/fence.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// This file is included inside `cuda_device::fence` so existing paths stay stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/float.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::float`.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/grid_dependency.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::grid` to keep its public API stable.
Expand Down
2 changes: 1 addition & 1 deletion crates/cuda-device/src/generated/grid_sreg.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
*/

// @generated by cuda-intrinsics-gen 0.1.0. DO NOT EDIT.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 00372dfe2e27296fcd2a4505eff6e547153693b60a4fcc0d1599a4d589193226.
// Catalog schema/version: 46/0.1.0; intrinsic ABI: v1; catalog SHA-256: 9b5ce71ce97b9a58e53bc98541d3449020497dbf7b46b91c1da19faf5b68619f.
// LLVM source: 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e.

// Included inside `cuda_device::grid` to keep existing paths stable.
Expand Down
Loading