Files
TheRock/rocm-systems/projects/rocshmem/include/gin_anvil/sdma_factory.h
T
derek.holloway 7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
Flatten subprojects into a single base monorepo
2026-09-02 16:34:47 -07:00

61 lines
2.5 KiB
C
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
/******************************************************************************
* Copyright (c) 2026, Advanced Micro Devices, Inc. All rights reserved.
*
* SPDX-License-Identifier: MIT
*****************************************************************************/
#ifndef GIN_ANVIL_SDMA_FACTORY_H_
#define GIN_ANVIL_SDMA_FACTORY_H_
#include <stddef.h>
#include <stdint.h>
#define GIN_ANVIL_SDMA_API __attribute__((visibility("default")))
#ifdef __cplusplus
extern "C" {
#endif
/**
* Opaque handle: standalone SDMA queue table (per local peer × channels) +
* sdmaDirty bitmask. Independent of PGAS runtime init; uses AnvilLib device paths.
*/
typedef struct gin_anvil_sdma_opaque* gin_anvil_sdma_handle_t;
/** @return 1 if SDMA Anvil path is compiled in and HIP reports at least one device, else 0 */
GIN_ANVIL_SDMA_API int gin_anvil_sdma_probe(void);
/**
* Bootstrap allgather: int allgather(void* ctx, void* buf, size_t bytes_per_rank)
* Each rank fills buf[myRank * elem_size ..] before the collective; result is full table.
*
* On entry, `buf` must point to `nRanks * sizeof(int)` bytes. The caller sets
* `buf[myRank] = my_device_id` before invoking `allgather`; after the collective,
* `buf[r]` is the HIP device ordinal for rank `r`.
*
* @param my_device_id HIP ordinal for this rank (typically hipGetDevice).
* @param num_channels SDMA channels per peer pair (clamped to [1,8]).
* @param out_gpu_handles Device pointer to array of nRanks*num_channels pointers to
* SdmaQueueDeviceHandle (layout: local_pe * num_channels + ch).
* @param out_sdma_dirty Device pointer to a single uint64_t dirty bitmask (device memory).
*/
GIN_ANVIL_SDMA_API int gin_anvil_sdma_create(
int nRanks, int myRank, int my_device_id,
int (*allgather)(void* ctx, void* buf, size_t bytes_per_rank), void* allgather_ctx,
int num_channels, gin_anvil_sdma_handle_t* out_handle, void** out_gpu_handles,
uint64_t** out_sdma_dirty);
GIN_ANVIL_SDMA_API void gin_anvil_sdma_destroy(gin_anvil_sdma_handle_t handle);
/** Fields stored in the opaque handle for RCCL plugin / device code */
GIN_ANVIL_SDMA_API int gin_anvil_sdma_get_n_ranks(gin_anvil_sdma_handle_t handle);
GIN_ANVIL_SDMA_API int gin_anvil_sdma_get_num_channels(gin_anvil_sdma_handle_t handle);
/** 1 = spread wavefronts across channels (NCCL_GIN_ANVIL_SDMA_SPREAD_CHANNELS, default on) */
GIN_ANVIL_SDMA_API int gin_anvil_sdma_get_channel_stride(gin_anvil_sdma_handle_t handle);
#ifdef __cplusplus
}
#endif
#endif // GIN_ANVIL_SDMA_FACTORY_H_