Skip to content

UCT/IB/MLX5/GDAKI: Fixed retaining inactive primary context. #11276

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

Draft
rakhmets wants to merge 1 commit into
base: master
Choose a base branch
from
Draft

UCT/IB/MLX5/GDAKI: Fixed retaining inactive primary context. #11276

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
25 changes: 13 additions & 12 deletions src/uct/ib/mlx5/gdaki/gdaki.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,6 +19,7 @@
#include <uct/ib/base/ib_verbs.h>
#include <uct/ib/mlx5/rc/rc_mlx5.h>
#include <uct/cuda/cuda_copy/cuda_copy_md.h>
#include <uct/cuda/base/cuda_ctx.h>
#include <uct/cuda/base/cuda_util.h>

#include "gpunetio/common/doca_gpunetio_verbs_def.h"
Expand All @@ -33,6 +34,7 @@ typedef struct {
uct_rc_iface_common_config_t super;
uct_rc_mlx5_iface_common_config_t mlx5;
unsigned num_channels;
int retain_primary_ctx;
} uct_rc_gdaki_iface_config_t;

ucs_config_field_t uct_rc_gdaki_iface_config_table[] = {
Expand All @@ -50,6 +52,11 @@ ucs_config_field_t uct_rc_gdaki_iface_config_table[] = {
ucs_offsetof(uct_rc_gdaki_iface_config_t, num_channels),
UCS_CONFIG_TYPE_UINT},

{"RETAIN_PRIMARY_CTX", "n",
"Retain and use an inactive CUDA primary context for memory allocation",
ucs_offsetof(uct_rc_gdaki_iface_config_t, retain_primary_ctx),
UCS_CONFIG_TYPE_BOOL},

{NULL}
};

Expand Down Expand Up @@ -763,17 +770,13 @@ static UCS_CLASS_INIT_FUNC(uct_rc_gdaki_iface_t, uct_md_h tl_md,
return status;
}

status = UCT_CUDADRV_FUNC_LOG_ERR(
cuDevicePrimaryCtxRetain(&self->cuda_ctx, self->cuda_dev));
status = uct_cuda_ctx_primary_push(self->cuda_dev,
config->retain_primary_ctx,
UCS_LOG_LEVEL_ERROR);
if (status != UCS_OK) {
return status;
}

status = UCT_CUDADRV_FUNC_LOG_ERR(cuCtxPushCurrent(self->cuda_ctx));
if (status != UCS_OK) {
goto err_ctx_release;
}

status = uct_rc_gdaki_alloc(sizeof(uint64_t), sizeof(uint64_t),
(void**)&self->atomic_buff, &self->atomic_raw);
if (status != UCS_OK) {
Expand All @@ -797,19 +800,17 @@ static UCS_CLASS_INIT_FUNC(uct_rc_gdaki_iface_t, uct_md_h tl_md,
err_lock:
ibv_dereg_mr(self->atomic_mr);
err_atomic:
cuMemFree(self->atomic_raw);
(void)UCT_CUDADRV_FUNC_LOG_WARN(cuMemFree(self->atomic_raw));
err_ctx:
(void)UCT_CUDADRV_FUNC_LOG_WARN(cuCtxPopCurrent(NULL));
err_ctx_release:
(void)UCT_CUDADRV_FUNC_LOG_WARN(cuDevicePrimaryCtxRelease(self->cuda_dev));
uct_cuda_ctx_primary_pop_and_release(self->cuda_dev);
return status;
}

static UCS_CLASS_CLEANUP_FUNC(uct_rc_gdaki_iface_t)
{
pthread_mutex_destroy(&self->ep_init_lock);
ibv_dereg_mr(self->atomic_mr);
cuMemFree(self->atomic_raw);
(void)UCT_CUDADRV_FUNC_LOG_WARN(cuMemFree(self->atomic_raw));
(void)UCT_CUDADRV_FUNC_LOG_WARN(cuDevicePrimaryCtxRelease(self->cuda_dev));
}

Expand Down
Loading