AMD-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Mario Limonciello <superm1@kernel.org>
To: Eric Huang <jinhuieric.huang@amd.com>, amd-gfx@lists.freedesktop.org
Cc: Tishko.Araz@amd.com, Alexander.Deucher@amd.com
Subject: Re: [PATCH 1/3] drm/amdgpu: add new mes misc api for gfx11
Date: Fri, 21 Aug 2026 16:30:31 -0500	[thread overview]
Message-ID: <6a149891-1752-4d43-aa78-8573e1503837@kernel.org> (raw)
In-Reply-To: <20260819201443.282689-1-jinhuieric.huang@amd.com>



On 8/19/26 15:14, Eric Huang wrote:
> new api will be used to workaround a HW scheduler issue
> for 100% usage when gpu has no workload.
> 
> Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c | 23 +++++++++++++++++++++++
>   drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h |  4 ++++
>   drivers/gpu/drm/amd/amdgpu/mes_v11_0.c  | 10 +++++++++-
>   3 files changed, 36 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c
> index b96f94e5169f..421d937c1188 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c
> @@ -1140,6 +1140,29 @@ void amdgpu_mes_free_gang_ctx_index(struct amdgpu_mes *mes,
>   	amdgpu_mes_unlock(mes);
>   }
>   
> +int amdgpu_mes_notify_unmap_queue(struct amdgpu_device *adev)
> +{
> +	struct mes_misc_op_input op_input = {0};
> +	int r;
> +
> +	op_input.op = MES_MISC_NOTIFY_WORK_ON_UNMAPPED_QUEUE;
> +
> +	if (!adev->mes.funcs->misc_op) {
> +		dev_err(adev->dev, "mes notify unmap queue is not supported!\n");
> +		r = -EINVAL;
> +		goto error;
> +	}
> +
> +	amdgpu_mes_lock(&adev->mes);
> +	r = adev->mes.funcs->misc_op(&adev->mes, &op_input);
> +	amdgpu_mes_unlock(&adev->mes);
> +	if (r)
> +		dev_err(adev->dev, "failed to notify unmap queue.\n");
> +
> +error:
> +	return r;
> +}
> +
>   #if defined(CONFIG_DEBUG_FS)
>   
>   static int amdgpu_debugfs_mes_event_log_show(struct seq_file *m, void *unused)
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h
> index 977c057dcce8..0291ae14a1e7 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h
> @@ -366,6 +366,7 @@ enum mes_misc_opcode {
>   	MES_MISC_OP_WRM_REG_WR_WAIT,
>   	MES_MISC_OP_SET_SHADER_DEBUGGER,
>   	MES_MISC_OP_CHANGE_CONFIG,
> +	MES_MISC_OP_NOTIFY_WORK_ON_UNMAPPED_QUEUE
>   };
>   
>   struct mes_misc_op_input {
> @@ -643,4 +644,7 @@ int amdgpu_mes_alloc_gang_ctx_index(struct amdgpu_mes *mes,
>   				    uint32_t *index);
>   void amdgpu_mes_free_gang_ctx_index(struct amdgpu_mes *mes,
>   				    uint32_t index);
> +
> +int amdgpu_mes_notify_unmap_queue(struct amdgpu_device *adev);
> +
>   #endif /* __AMDGPU_MES_H__ */
> diff --git a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c
> index c57bcb9a98b1..df71e9447b35 100644
> --- a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c
> +++ b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c
> @@ -957,7 +957,15 @@ static int mes_v11_0_misc_op(struct amdgpu_mes *mes,
>   		misc_pkt.change_config.option.bits.limit_single_process =
>   				input->change_config.option.limit_single_process;
>   		break;
> -
> +	case MES_MISC_OP_NOTIFY_WORK_ON_UNMAPPED_QUEUE:
> +		if ((mes->adev->mes.sched_version & AMDGPU_MES_VERSION_MASK) < 0x70) {
> +			dev_warn_once(mes->adev->dev,
> +				      "MES FW version must be larger than 0x70 to support notify work on unmapped queue.\n");
> +			return 0;
> +		}
> +		misc_pkt.opcode = MESAPI_MISC__NOTIFY_WORK_ON_UNMAPPED_QUEUE;
> +		misc_pkt.queue_sch_level = AMD_PRIORITY_LEVEL_NORMAL;
> +		break;
>   	default:
>   		drm_err(adev_to_drm(mes->adev), "unsupported misc op (%d)\n", input->op);
>   		return -EINVAL;


      parent reply	other threads:[~2026-08-21 21:30 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-19 20:14 [PATCH 1/3] drm/amdgpu: add new mes misc api for gfx11 Eric Huang
2026-08-19 20:14 ` [PATCH 2/3] drm/amdkfd: workaround 100% gpu usage issue " Eric Huang
2026-08-21 21:31   ` Mario Limonciello
2026-08-19 20:14 ` [PATCH 3/3] drm/amdgpu/userq: create the same workaround of oversubscription timer Eric Huang
2026-08-21 21:32   ` Mario Limonciello
2026-08-21 21:30 ` Mario Limonciello [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=6a149891-1752-4d43-aa78-8573e1503837@kernel.org \
    --to=superm1@kernel.org \
    --cc=Alexander.Deucher@amd.com \
    --cc=Tishko.Araz@amd.com \
    --cc=amd-gfx@lists.freedesktop.org \
    --cc=jinhuieric.huang@amd.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox