AMD-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Zhu Lingshan <lingshan.zhu@amd.com>
To: <Alexander.Deucher@amd.com>, <felix.kuehling@amd.com>
Cc: <Ray.Huang@amd.com>, <amd-gfx@lists.freedesktop.org>,
	Zhu Lingshan <lingshan.zhu@amd.com>
Subject: [RFC 06/10] drm/amdgpu: GFX11: set mqd->perfcount from mqd_prop
Date: Tue, 22 Sep 2026 17:34:32 +0800	[thread overview]
Message-ID: <20260922093436.596947-7-lingshan.zhu@amd.com> (raw)
In-Reply-To: <20260922093436.596947-1-lingshan.zhu@amd.com>

This commit implements a new callback set_mqd_perfcount
in struct amdgpu_mqd. This callback helps setting
mqd->compute_perfcount_enable from amdgpu_mqd_prop->perfcount_enable.

This is a per-IP callback and it is NULL if an IP does not
support PMC.

This commit also implements set_mqd_perfcount for GFX11.

Signed-off-by: Zhu Lingshan <lingshan.zhu@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h |  6 ++++++
 drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c  | 11 +++++++++++
 2 files changed, 17 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h
index 7a5499ec1171..f7ae3d5484da 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h
@@ -479,12 +479,18 @@ struct amdgpu_mqd_prop {
 	uint64_t ctx_save_area_addr;
 	uint32_t ctx_save_area_size;
 	uint32_t ctl_stack_size;
+	/* PMC compute_perfcount_enable that programmed into this queue's MQD */
+	bool perfcount_enable;
 };
 
 struct amdgpu_mqd {
 	unsigned mqd_size;
 	int (*init_mqd)(struct amdgpu_device *adev, void *mqd,
 			struct amdgpu_mqd_prop *p);
+	/* Per-IP callback to set compute_perfcount_enable,
+	 * NULL when this IP's MQD has no perfcount_enable field.
+	 */
+	void (*set_mqd_perfcount)(void *mqd, struct amdgpu_mqd_prop *p);
 };
 
 /*
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
index 254a72f26ec5..252a1096b4a0 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
@@ -4323,6 +4323,7 @@ static int gfx_v11_0_compute_mqd_init(struct amdgpu_device *adev, void *m,
 
 	mqd->header = 0xC0310800;
 	mqd->compute_pipelinestat_enable = 0x00000001;
+	mqd->compute_perfcount_enable = prop->perfcount_enable;
 	mqd->compute_static_thread_mgmt_se0 = 0xffffffff;
 	mqd->compute_static_thread_mgmt_se1 = 0xffffffff;
 	mqd->compute_static_thread_mgmt_se2 = 0xffffffff;
@@ -4459,6 +4460,14 @@ static int gfx_v11_0_compute_mqd_init(struct amdgpu_device *adev, void *m,
 	return 0;
 }
 
+static void gfx_v11_0_compute_mqd_set_perfcount(void *m,
+						struct amdgpu_mqd_prop *prop)
+{
+	struct v11_compute_mqd *mqd = m;
+
+	mqd->compute_perfcount_enable = prop->perfcount_enable;
+}
+
 static int gfx_v11_0_kiq_init_register(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
@@ -7263,6 +7272,8 @@ static void gfx_v11_0_set_mqd_funcs(struct amdgpu_device *adev)
 		sizeof(struct v11_compute_mqd);
 	adev->mqds[AMDGPU_HW_IP_COMPUTE].init_mqd =
 		gfx_v11_0_compute_mqd_init;
+	adev->mqds[AMDGPU_HW_IP_COMPUTE].set_mqd_perfcount =
+		gfx_v11_0_compute_mqd_set_perfcount;
 }
 
 static void gfx_v11_0_set_user_wgp_inactive_bitmap_per_sh(struct amdgpu_device *adev,
-- 
2.55.0


  parent reply	other threads:[~2026-09-22  9:34 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-22  9:34 [RFC 00/10] amdgpu: implement Performance Monitor Counter support Zhu Lingshan
2026-09-22  9:34 ` [RFC 01/10] drm/amdgpu: implement PMC ioctl framework Zhu Lingshan
2026-09-23 14:17   ` James Zhu
2026-09-24  3:38     ` Zhu, Lingshan
2026-09-24 15:49       ` James Zhu
2026-10-08  7:59         ` Zhu, Lingshan
2026-09-22  9:34 ` [RFC 02/10] drm/amdgpu: add device-wide PMC config Zhu Lingshan
2026-09-23 14:13   ` James Zhu
2026-09-24  3:23     ` Zhu, Lingshan
2026-09-24 15:40       ` James Zhu
2026-10-08  7:50         ` Zhu, Lingshan
2026-09-22  9:34 ` [RFC 03/10] drm/amdgpu: introduce PMC ioctl handler Zhu Lingshan
2026-09-22  9:34 ` [RFC 04/10] drm/amdgpu: implement reservation mechanism for PMC Zhu Lingshan
2026-09-22  9:34 ` [RFC 05/10] drm/amdgpu: set device-wide PMC enablement indicator Zhu Lingshan
2026-09-22  9:34 ` Zhu Lingshan [this message]
2026-09-22  9:34 ` [RFC 07/10] drm/amdgpu: GFX12: set mqd->perfcount from mqd_prop Zhu Lingshan
2026-09-22  9:34 ` [RFC 08/10] drm/amdgpu: GFX12.1: " Zhu Lingshan
2026-09-22  9:34 ` [RFC 09/10] drm/amdgpu: implement device PMC handler Zhu Lingshan
2026-09-22  9:34 ` [RFC 10/10] drm/amdgpu: set userq perf state before publish it Zhu Lingshan
2026-09-23 14:02 ` [RFC 00/10] amdgpu: implement Performance Monitor Counter support James Zhu
2026-09-23 15:02   ` Welton, Benjamin
2026-09-23 15:11     ` James Zhu
2026-09-24  3:40     ` Zhu, Lingshan

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=20260922093436.596947-7-lingshan.zhu@amd.com \
    --to=lingshan.zhu@amd.com \
    --cc=Alexander.Deucher@amd.com \
    --cc=Ray.Huang@amd.com \
    --cc=amd-gfx@lists.freedesktop.org \
    --cc=felix.kuehling@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