* [PATCH] drm/amdkfd: report the numa weight between host and device over xgmi
@ 2021-04-29 9:34 Jonathan Kim
2021-04-29 14:46 ` Felix Kuehling
0 siblings, 1 reply; 2+ messages in thread
From: Jonathan Kim @ 2021-04-29 9:34 UTC (permalink / raw)
To: amd-gfx; +Cc: Oak.Zeng, Jonathan Kim, Ramesh.Errabolu, Felix.Kuehling
GPUs connected to CPUs over xGMI are bidirectional so set weight by a
single hop both ways.
Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Tested-by: Ramesh Errabolu <ramesh.errabolu@amd.com>
---
drivers/gpu/drm/amd/amdkfd/kfd_crat.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
index eb30324393a8..ca90b710d76a 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
@@ -1789,6 +1789,7 @@ static int kfd_fill_gpu_direct_io_link_to_cpu(int *avail_size,
*/
sub_type_hdr->flags |= CRAT_IOLINK_FLAGS_BI_DIRECTIONAL;
sub_type_hdr->io_interface_type = CRAT_IOLINK_TYPE_XGMI;
+ sub_type_hdr->num_hops_xgmi = 1;
} else {
sub_type_hdr->io_interface_type = CRAT_IOLINK_TYPE_PCIEXPRESS;
}
--
2.17.1
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH] drm/amdkfd: report the numa weight between host and device over xgmi
2021-04-29 9:34 [PATCH] drm/amdkfd: report the numa weight between host and device over xgmi Jonathan Kim
@ 2021-04-29 14:46 ` Felix Kuehling
0 siblings, 0 replies; 2+ messages in thread
From: Felix Kuehling @ 2021-04-29 14:46 UTC (permalink / raw)
To: Jonathan Kim, amd-gfx; +Cc: Oak.Zeng, Ramesh.Errabolu
Am 2021-04-29 um 5:34 a.m. schrieb Jonathan Kim:
> GPUs connected to CPUs over xGMI are bidirectional so set weight by a
> single hop both ways.
>
> Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
> Tested-by: Ramesh Errabolu <ramesh.errabolu@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
> ---
> drivers/gpu/drm/amd/amdkfd/kfd_crat.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
> index eb30324393a8..ca90b710d76a 100644
> --- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
> +++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
> @@ -1789,6 +1789,7 @@ static int kfd_fill_gpu_direct_io_link_to_cpu(int *avail_size,
> */
> sub_type_hdr->flags |= CRAT_IOLINK_FLAGS_BI_DIRECTIONAL;
> sub_type_hdr->io_interface_type = CRAT_IOLINK_TYPE_XGMI;
> + sub_type_hdr->num_hops_xgmi = 1;
> } else {
> sub_type_hdr->io_interface_type = CRAT_IOLINK_TYPE_PCIEXPRESS;
> }
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2021-04-29 14:46 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-04-29 9:34 [PATCH] drm/amdkfd: report the numa weight between host and device over xgmi Jonathan Kim
2021-04-29 14:46 ` Felix Kuehling
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.