cudart.h

/*
 * Copyright © 2010-2013 Inria.  All rights reserved.
 * Copyright © 2010-2011 Université Bordeaux 1
 * Copyright © 2011 Cisco Systems, Inc.  All rights reserved.
 * See COPYING in top-level directory.
 */

#ifndef HWLOC_CUDART_H
#define HWLOC_CUDART_H

#include <hwloc.h>
#include <hwloc/autogen/config.h>
#include <hwloc/helper.h>
#ifdef HWLOC_LINUX_SYS
#include <hwloc/linux.h>
#endif

#include <cuda_runtime_api.h>


#ifdef __cplusplus
extern "C" {
#endif


static __hwloc_inline int
hwloc_cudart_get_device_pci_ids(hwloc_topology_t topology __hwloc_attribute_unused,
                                int idx, int *domain, int *bus, int *dev)
{
  cudaError_t cerr;
  struct cudaDeviceProp prop;

  cerr = cudaGetDeviceProperties(&prop, idx);
  if (cerr) {
    errno = ENOSYS;
    return -1;
  }

#ifdef CU_DEVICE_ATTRIBUTE_PCI_DOMAIN_ID
  *domain = prop.pciDomainID;
#else
  *domain = 0;
#endif

  *bus = prop.pciBusID;
  *dev = prop.pciDeviceID;

  return 0;
}

static __hwloc_inline int
hwloc_cudart_get_device_cpuset(hwloc_topology_t topology __hwloc_attribute_unused,
                               int idx, hwloc_cpuset_t set)
{
#ifdef HWLOC_LINUX_SYS
  /* If we're on Linux, use the sysfs mechanism to get the local cpus */
#define HWLOC_CUDART_DEVICE_SYSFS_PATH_MAX 128
  char path[HWLOC_CUDART_DEVICE_SYSFS_PATH_MAX];
  FILE *sysfile = NULL;
  int domain, bus, dev;

  if (hwloc_cudart_get_device_pci_ids(topology, idx, &domain, &bus, &dev))
    return -1;

  if (!hwloc_topology_is_thissystem(topology)) {
    errno = EINVAL;
    return -1;
  }

  sprintf(path, "/sys/bus/pci/devices/%04x:%02x:%02x.0/local_cpus", domain, bus, dev);
  sysfile = fopen(path, "r");
  if (!sysfile)
    return -1;

  hwloc_linux_parse_cpumap_file(sysfile, set);
  if (hwloc_bitmap_iszero(set))
    hwloc_bitmap_copy(set, hwloc_topology_get_complete_cpuset(topology));

  fclose(sysfile);
#else
  /* Non-Linux systems simply get a full cpuset */
  hwloc_bitmap_copy(set, hwloc_topology_get_complete_cpuset(topology));
#endif
  return 0;
}

static __hwloc_inline hwloc_obj_t
hwloc_cudart_get_device_pcidev(hwloc_topology_t topology, int idx)
{
  int domain, bus, dev;

  if (hwloc_cudart_get_device_pci_ids(topology, idx, &domain, &bus, &dev))
    return NULL;

  return hwloc_get_pcidev_by_busid(topology, domain, bus, dev, 0);
}

static __hwloc_inline hwloc_obj_t
hwloc_cudart_get_device_osdev_by_index(hwloc_topology_t topology, unsigned idx)
{
        hwloc_obj_t osdev = NULL;
        while ((osdev = hwloc_get_next_osdev(topology, osdev)) != NULL) {
                if (HWLOC_OBJ_OSDEV_COPROC == osdev->attr->osdev.type
                    && osdev->name
                    && !strncmp("cuda", osdev->name, 4)
                    && atoi(osdev->name + 4) == (int) idx)
                        return osdev;
        }
        return NULL;
}

#ifdef __cplusplus
} /* extern "C" */
#endif


#endif /* HWLOC_CUDART_H */