All of lore.kernel.org
 help / color / mirror / Atom feed
From: Hawking Zhang <Hawking.Zhang@amd.com>
To: <amd-gfx@lists.freedesktop.org>,
	John Clements <John.Clements@amd.com>,
	Dennis Li <Dennis.Li@amd.com>,
	Alex Deucher <alexander.deucher@amd.com>
Cc: Hawking Zhang <Hawking.Zhang@amd.com>
Subject: [PATCH 07/11] drm/amdgpu: move xgmi ras functions to xgmi_ras_funcs
Date: Fri, 2 Apr 2021 14:43:41 +0800	[thread overview]
Message-ID: <20210402064345.14093-7-Hawking.Zhang@amd.com> (raw)
In-Reply-To: <20210402064345.14093-1-Hawking.Zhang@amd.com>

xgmi ras is not managed by gpu driver when gpu is
connected to cpu through xgmi. move all xgmi ras
functions to xgmi_ras_funcs so gpu driver only
initializes xgmi ras functions when it manages
xgmi ras.

Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reivewed-by: Dennis Li <Dennis.Li@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c  | 16 ++++++++++++++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h  |  9 +++++++++
 drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c  |  4 +++-
 drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c | 21 ++++++++++++++-------
 drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h |  7 +------
 drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c    |  2 +-
 6 files changed, 42 insertions(+), 17 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
index b9d68fd2610c..082f9d03a94a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
@@ -403,14 +403,26 @@ int amdgpu_gmc_ras_late_init(struct amdgpu_device *adev)
 			return r;
 	}
 
-	return amdgpu_xgmi_ras_late_init(adev);
+	if (!adev->gmc.xgmi.connected_to_cpu)
+		adev->gmc.xgmi.ras_funcs = &xgmi_ras_funcs;
+
+	if (adev->gmc.xgmi.ras_funcs &&
+	    adev->gmc.xgmi.ras_funcs->ras_late_init) {
+		r = adev->gmc.xgmi.ras_funcs->ras_late_init(adev);
+		if (r)
+			return r;
+	}
+
+	return 0;
 }
 
 void amdgpu_gmc_ras_fini(struct amdgpu_device *adev)
 {
 	amdgpu_umc_ras_fini(adev);
 	amdgpu_mmhub_ras_fini(adev);
-	amdgpu_xgmi_ras_fini(adev);
+	if (adev->gmc.xgmi.ras_funcs &&
+	    adev->gmc.xgmi.ras_funcs->ras_fini)
+		adev->gmc.xgmi.ras_funcs->ras_fini(adev);
 }
 
 	/*
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
index 7e248a4e2fa3..cbb7735c6988 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
@@ -135,6 +135,14 @@ struct amdgpu_gmc_funcs {
 	unsigned int (*get_vbios_fb_size)(struct amdgpu_device *adev);
 };
 
+struct amdgpu_xgmi_ras_funcs {
+	int (*ras_late_init)(struct amdgpu_device *adev);
+	void (*ras_fini)(struct amdgpu_device *adev);
+	int (*query_ras_error_count)(struct amdgpu_device *adev,
+				     void *ras_error_status);
+	void (*reset_ras_error_count)(struct amdgpu_device *adev);
+};
+
 struct amdgpu_xgmi {
 	/* from psp */
 	u64 node_id;
@@ -151,6 +159,7 @@ struct amdgpu_xgmi {
 	struct ras_common_if *ras_if;
 	bool connected_to_cpu;
 	bool pending_reset;
+	const struct amdgpu_xgmi_ras_funcs *ras_funcs;
 };
 
 struct amdgpu_gmc {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index ac3f4c3266bc..172738cc99db 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -809,7 +809,9 @@ int amdgpu_ras_query_error_status(struct amdgpu_device *adev,
 			adev->nbio.ras_funcs->query_ras_error_count(adev, &err_data);
 		break;
 	case AMDGPU_RAS_BLOCK__XGMI_WAFL:
-		amdgpu_xgmi_query_ras_error_count(adev, &err_data);
+		if (adev->gmc.xgmi.ras_funcs &&
+		    adev->gmc.xgmi.ras_funcs->query_ras_error_count)
+			adev->gmc.xgmi.ras_funcs->query_ras_error_count(adev, &err_data);
 		break;
 	default:
 		break;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
index 5a1b598f4d30..8567d5d77346 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
@@ -628,7 +628,7 @@ int amdgpu_xgmi_remove_device(struct amdgpu_device *adev)
 	return psp_xgmi_terminate(&adev->psp);
 }
 
-int amdgpu_xgmi_ras_late_init(struct amdgpu_device *adev)
+static int amdgpu_xgmi_ras_late_init(struct amdgpu_device *adev)
 {
 	int r;
 	struct ras_ih_if ih_info = {
@@ -642,7 +642,7 @@ int amdgpu_xgmi_ras_late_init(struct amdgpu_device *adev)
 	    adev->gmc.xgmi.num_physical_nodes == 0)
 		return 0;
 
-	amdgpu_xgmi_reset_ras_error_count(adev);
+	adev->gmc.xgmi.ras_funcs->reset_ras_error_count(adev);
 
 	if (!adev->gmc.xgmi.ras_if) {
 		adev->gmc.xgmi.ras_if = kmalloc(sizeof(struct ras_common_if), GFP_KERNEL);
@@ -664,7 +664,7 @@ int amdgpu_xgmi_ras_late_init(struct amdgpu_device *adev)
 	return r;
 }
 
-void amdgpu_xgmi_ras_fini(struct amdgpu_device *adev)
+static void amdgpu_xgmi_ras_fini(struct amdgpu_device *adev)
 {
 	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__XGMI_WAFL) &&
 			adev->gmc.xgmi.ras_if) {
@@ -691,7 +691,7 @@ static void pcs_clear_status(struct amdgpu_device *adev, uint32_t pcs_status_reg
 	WREG32_PCIE(pcs_status_reg, 0);
 }
 
-void amdgpu_xgmi_reset_ras_error_count(struct amdgpu_device *adev)
+static void amdgpu_xgmi_reset_ras_error_count(struct amdgpu_device *adev)
 {
 	uint32_t i;
 
@@ -751,8 +751,8 @@ static int amdgpu_xgmi_query_pcs_error_status(struct amdgpu_device *adev,
 	return 0;
 }
 
-int amdgpu_xgmi_query_ras_error_count(struct amdgpu_device *adev,
-				      void *ras_error_status)
+static int amdgpu_xgmi_query_ras_error_count(struct amdgpu_device *adev,
+					     void *ras_error_status)
 {
 	struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status;
 	int i;
@@ -801,10 +801,17 @@ int amdgpu_xgmi_query_ras_error_count(struct amdgpu_device *adev,
 		break;
 	}
 
-	amdgpu_xgmi_reset_ras_error_count(adev);
+	adev->gmc.xgmi.ras_funcs->reset_ras_error_count(adev);
 
 	err_data->ue_count += ue_cnt;
 	err_data->ce_count += ce_cnt;
 
 	return 0;
 }
+
+const struct amdgpu_xgmi_ras_funcs xgmi_ras_funcs = {
+	.ras_late_init = amdgpu_xgmi_ras_late_init,
+	.ras_fini = amdgpu_xgmi_ras_fini,
+	.query_ras_error_count = amdgpu_xgmi_query_ras_error_count,
+	.reset_ras_error_count = amdgpu_xgmi_reset_ras_error_count,
+};
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h
index 148560d63554..12969c0830d5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h
@@ -50,6 +50,7 @@ struct amdgpu_pcs_ras_field {
 	uint32_t pcs_err_shift;
 };
 
+extern const struct amdgpu_xgmi_ras_funcs xgmi_ras_funcs;
 struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev);
 void amdgpu_put_xgmi_hive(struct amdgpu_hive_info *hive);
 int amdgpu_xgmi_update_topology(struct amdgpu_hive_info *hive, struct amdgpu_device *adev);
@@ -58,14 +59,8 @@ int amdgpu_xgmi_remove_device(struct amdgpu_device *adev);
 int amdgpu_xgmi_set_pstate(struct amdgpu_device *adev, int pstate);
 int amdgpu_xgmi_get_hops_count(struct amdgpu_device *adev,
 		struct amdgpu_device *peer_adev);
-int amdgpu_xgmi_ras_late_init(struct amdgpu_device *adev);
-void amdgpu_xgmi_ras_fini(struct amdgpu_device *adev);
 uint64_t amdgpu_xgmi_get_relative_phy_addr(struct amdgpu_device *adev,
 					   uint64_t addr);
-int amdgpu_xgmi_query_ras_error_count(struct amdgpu_device *adev,
-				      void *ras_error_status);
-void amdgpu_xgmi_reset_ras_error_count(struct amdgpu_device *adev);
-
 static inline bool amdgpu_xgmi_same_hive(struct amdgpu_device *adev,
 		struct amdgpu_device *bo_adev)
 {
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index 1567dd227f51..ce2b1a9125ea 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -1208,7 +1208,7 @@ static int gmc_v9_0_early_init(void *handle)
 		adev->gmc.xgmi.supported = true;
 		adev->gmc.xgmi.connected_to_cpu =
 			adev->smuio.funcs->is_host_gpu_xgmi_supported(adev);
-        }
+	}
 
 	adev->gmc.shared_aperture_start = 0x2000000000000000ULL;
 	adev->gmc.shared_aperture_end =
-- 
2.17.1

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

  parent reply	other threads:[~2021-04-02  6:44 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-02  6:43 [PATCH 01/11] drm/amdgpu: initialze ras caps per paltform config Hawking Zhang
2021-04-02  6:43 ` [PATCH 02/11] drm/amdgpu: create umc_v6_7_funcs for aldebaran Hawking Zhang
2021-04-02  6:43 ` [PATCH 03/11] drm/amdgpu: add helper funtion to query umc ras error Hawking Zhang
2021-04-02  6:43 ` [PATCH 04/11] drm/amdgpu: implement umc query error count callback Hawking Zhang
2021-04-02  6:43 ` [PATCH 05/11] drm/amdgpu: implement query_ras_error_address callback Hawking Zhang
2021-04-02  6:43 ` [PATCH 06/11] drm/amdgpu: split nbio callbacks into ras and non-ras ones Hawking Zhang
2021-04-02  6:43 ` Hawking Zhang [this message]
2021-04-02  6:43 ` [PATCH 08/11] drm/amdgpu: split umc callbacks to " Hawking Zhang
2021-04-02  6:43 ` [PATCH 09/11] drm/amdgpu: do not register df_mca interrupt in certain config Hawking Zhang
2021-04-02  6:43 ` [PATCH 10/11] drm/amdgpu: split mmhub callbacks into ras and non-ras ones Hawking Zhang
2021-04-02  6:43 ` [PATCH 11/11] drm/amdgpu: split gfx " Hawking Zhang
2021-04-02  6:47   ` Clements, John

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=20210402064345.14093-7-Hawking.Zhang@amd.com \
    --to=hawking.zhang@amd.com \
    --cc=Dennis.Li@amd.com \
    --cc=John.Clements@amd.com \
    --cc=alexander.deucher@amd.com \
    --cc=amd-gfx@lists.freedesktop.org \
    /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 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.