Re: [PATCH] drm/amdgpu: Print node id in peer map failure message

[Felix Kuehling <felix.kuehling@amd.com>]

Am 2022-02-15 um 20:42 schrieb Harish Kasiviswanathan:
> Print alloc node, peer node and memory domain when peer map fails. This
> is more useful
>
> Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
> ---
>   drivers/gpu/drm/amd/amdkfd/kfd_chardev.c | 5 +++--
>   1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
> index 965af2a08bc0..4d62784c7cc3 100644
> --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
> +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
> @@ -1215,8 +1215,9 @@ static int kfd_ioctl_map_memory_to_gpu(struct file *filep,
>   			peer_pdd->dev->adev, (struct kgd_mem *)mem,
>   			peer_pdd->drm_priv, &table_freed);
>   		if (err) {
> -			pr_err("Failed to map to gpu %d/%d\n",
> -			       i, args->n_devices);
> +			pr_err("Failed to map peer:%d alloc:%d domain:%d\n",
> +			       peer_pdd->dev->id, dev->id,

Not sure if the dev->id is the most descriptive way to name the GPU 
because none of our user mode tools really expose that number. If you 
use dev_err instead of pr_err, you get a nicer way of printing one of 
the devices for free (probably the dev->adev->dev). For the peer you 
could manually print the PCI B:D.F.

Regards,
   Felix


> +			       ((struct kgd_mem *)mem)->domain);
>   			goto map_memory_to_gpu_failed;
>   		}
>   		args->n_success = i+1;