All of lore.kernel.org
 help / color / mirror / Atom feed
From: Tianci Yin <tianci.yin-5C7GfCeVMHo@public.gmane.org>
To: amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org
Cc: "Deucher Alexander"
	<Alexander.Deucher-5C7GfCeVMHo@public.gmane.org>,
	"Tuikov Luben" <Luben.Tuikov-5C7GfCeVMHo@public.gmane.org>,
	"Christian König" <christian.koenig-5C7GfCeVMHo@public.gmane.org>,
	"Tianci Yin" <tianci.yin-5C7GfCeVMHo@public.gmane.org>
Subject: [PATCH 7/8] drm/amdgpu: reserve vram for memory training
Date: Fri, 11 Oct 2019 11:50:32 +0800	[thread overview]
Message-ID: <20191011035033.24935-7-tianci.yin@amd.com> (raw)
In-Reply-To: <20191011035033.24935-1-tianci.yin-5C7GfCeVMHo@public.gmane.org>

From: "Tianci.Yin" <tianci.yin@amd.com>

memory training using specific fixed vram segment, reserve these
segments before anyone may allocate it.

Change-Id: I1436755813a565608a2857a683f535377620a637
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Tianci.Yin <tianci.yin@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 96 +++++++++++++++++++++++++
 1 file changed, 96 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 9da6350a4ba2..42d0fcb98382 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -1667,6 +1667,93 @@ static int amdgpu_ttm_fw_reserve_vram_init(struct amdgpu_device *adev)
 					  &adev->fw_vram_usage.va);
 }
 
+/*
+ * Memoy training reservation functions
+ */
+
+/**
+ * amdgpu_ttm_training_reserve_vram_fini - free memory training reserved vram
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * free memory training reserved vram if it has been reserved.
+ */
+static int amdgpu_ttm_training_reserve_vram_fini(struct amdgpu_device *adev)
+{
+	struct psp_memory_training_context *ctx = &adev->psp.mem_train_ctx;
+
+	ctx->init = PSP_MEM_TRAIN_NOT_SUPPORT;
+	if (ctx->c2p_bo) {
+		amdgpu_bo_free_kernel(&ctx->c2p_bo, NULL, NULL);
+		ctx->c2p_bo = NULL;
+	}
+	if (ctx->p2c_bo) {
+		amdgpu_bo_free_kernel(&ctx->p2c_bo, NULL, NULL);
+		ctx->p2c_bo = NULL;
+	}
+
+	return 0;
+}
+
+/**
+ * amdgpu_ttm_training_reserve_vram_init - create bo vram reservation from memory training
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * create bo vram reservation from memory training.
+ */
+static int amdgpu_ttm_training_reserve_vram_init(struct amdgpu_device *adev)
+{
+	int ret;
+	struct psp_memory_training_context *ctx = &adev->psp.mem_train_ctx;
+
+	memset(ctx, 0, sizeof(*ctx));
+	if (!adev->fw_vram_usage.mem_train_support) {
+		DRM_DEBUG("memory training does not support!\n");
+		return 0;
+	}
+
+	ctx->c2p_train_data_offset = adev->fw_vram_usage.mem_train_fb_loc;
+	ctx->p2c_train_data_offset = (adev->gmc.mc_vram_size - GDDR6_MEM_TRAINING_OFFSET);
+	ctx->train_data_size = GDDR6_MEM_TRAINING_DATA_SIZE_IN_BYTES;
+
+	DRM_DEBUG("train_data_size:%llx,p2c_train_data_offset:%llx,c2p_train_data_offset:%llx.\n",
+		  ctx->train_data_size,
+		  ctx->p2c_train_data_offset,
+		  ctx->c2p_train_data_offset);
+
+	ret = amdgpu_bo_create_kernel_at(adev,
+					 ctx->p2c_train_data_offset,
+					 ctx->train_data_size,
+					 AMDGPU_GEM_DOMAIN_VRAM,
+					 &ctx->p2c_bo,
+					 NULL);
+	if (ret) {
+		DRM_ERROR("alloc p2c_bo failed(%d)!\n", ret);
+		ret = -ENOMEM;
+		goto err_out;
+	}
+
+	ret = amdgpu_bo_create_kernel_at(adev,
+					 ctx->c2p_train_data_offset,
+					 ctx->train_data_size,
+					 AMDGPU_GEM_DOMAIN_VRAM,
+					 &ctx->c2p_bo,
+					 NULL);
+	if (ret) {
+		DRM_ERROR("alloc c2p_bo failed(%d)!\n", ret);
+		ret = -ENOMEM;
+		goto err_out;
+	}
+
+	ctx->init = PSP_MEM_TRAIN_RESERVE_SUCCESS;
+	return 0;
+
+err_out:
+	amdgpu_ttm_training_reserve_vram_fini(adev);
+	return ret;
+}
+
 /**
  * amdgpu_ttm_init - Init the memory management (ttm) as well as various
  * gtt/vram related fields.
@@ -1740,6 +1827,14 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 		return r;
 	}
 
+	/*
+	 *The reserved vram for memory training must be pinned to the specified
+	 *place on the VRAM, so reserve it early.
+	 */
+	r = amdgpu_ttm_training_reserve_vram_init(adev);
+	if (r)
+		return r;
+
 	/* allocate memory as required for VGA
 	 * This is used for VGA emulation and pre-OS scanout buffers to
 	 * avoid display artifacts while transitioning between pre-OS
@@ -1825,6 +1920,7 @@ void amdgpu_ttm_fini(struct amdgpu_device *adev)
 		return;
 
 	amdgpu_ttm_debugfs_fini(adev);
+	amdgpu_ttm_training_reserve_vram_fini(adev);
 	amdgpu_ttm_fw_reserve_vram_fini(adev);
 	if (adev->mman.aper_base_kaddr)
 		iounmap(adev->mman.aper_base_kaddr);
-- 
2.17.1

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

  parent reply	other threads:[~2019-10-11  3:50 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-11  3:50 [PATCH 1/8] drm/amdgpu: update amdgpu_discovery to handle revision Tianci Yin
     [not found] ` <20191011035033.24935-1-tianci.yin-5C7GfCeVMHo@public.gmane.org>
2019-10-11  3:50   ` [PATCH 2/8] drm/amdgpu: add a generic fb accessing helper function Tianci Yin
     [not found]     ` <20191011035033.24935-2-tianci.yin-5C7GfCeVMHo@public.gmane.org>
2019-10-11 22:44       ` Tuikov, Luben
2019-10-11  3:50   ` [PATCH 3/8] drm/amdgpu: introduce psp_v11_0_is_sos_alive interface Tianci Yin
2019-10-11  3:50   ` [PATCH 4/8] drm/amdgpu: update atomfirmware header with memory training related members Tianci Yin
     [not found]     ` <20191011035033.24935-4-tianci.yin-5C7GfCeVMHo@public.gmane.org>
2019-10-11 22:53       ` Tuikov, Luben
     [not found]         ` <155db3ea-82ba-e3d7-8e2f-96df99772871-5C7GfCeVMHo@public.gmane.org>
2019-10-12  2:21           ` Alex Deucher
2019-10-11  3:50   ` [PATCH 5/8] drm/amdgpu/atomfirmware: add memory training related helper functions Tianci Yin
     [not found]     ` <20191011035033.24935-5-tianci.yin-5C7GfCeVMHo@public.gmane.org>
2019-10-11 23:09       ` Tuikov, Luben
2019-10-11  3:50   ` [PATCH 6/8] drm/amdgpu: add psp memory training callbacks and macro Tianci Yin
2019-10-11  3:50   ` Tianci Yin [this message]
     [not found]     ` <20191011035033.24935-7-tianci.yin-5C7GfCeVMHo@public.gmane.org>
2019-10-11 23:23       ` [PATCH 7/8] drm/amdgpu: reserve vram for memory training Tuikov, Luben
     [not found]         ` <9084e67e-adc2-b512-b593-ca218c17a366-5C7GfCeVMHo@public.gmane.org>
2019-10-12  2:26           ` Alex Deucher
2019-10-14  8:26           ` Koenig, Christian
     [not found]             ` <a93b3b8e-4df9-f6e2-95f7-3f0c0d8bebdc-5C7GfCeVMHo@public.gmane.org>
2019-10-14  8:35               ` Yin, Tianci (Rico)
2019-10-11  3:50   ` [PATCH 8/8] drm/amdgpu/psp: add psp memory training implementation Tianci Yin
     [not found]     ` <20191011035033.24935-8-tianci.yin-5C7GfCeVMHo@public.gmane.org>
2019-10-11 23:44       ` Tuikov, Luben
  -- strict thread matches above, loose matches on Subject: below --
2019-10-08 19:29 [PATCH 0/8] low latency memory training for navi Alex Deucher
     [not found] ` <20191008192934.5481-1-alexander.deucher-5C7GfCeVMHo@public.gmane.org>
2019-10-08 19:29   ` [PATCH 7/8] drm/amdgpu: reserve vram for memory training Alex Deucher
     [not found]     ` <20191008192934.5481-8-alexander.deucher-5C7GfCeVMHo@public.gmane.org>
2019-10-09  3:44       ` Tuikov, Luben
     [not found]         ` <a9d04519-0ec0-41f6-289f-be156caccf76-5C7GfCeVMHo@public.gmane.org>
2019-10-09 11:12           ` Yin, Tianci (Rico)
     [not found]             ` <SN1PR12MB2544F1BE7D37DEC4721AF95D95950-z7L1TMIYDg7bG+2ccTdVRgdYzm3356FpvxpqHgZTriW3zl9H0oFU5g@public.gmane.org>
2019-10-09 11:40               ` Christian König
     [not found]                 ` <045b7cfd-989f-7cff-310f-92d9780e73fe-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
2019-10-09 11:50                   ` Yin, Tianci (Rico)

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=20191011035033.24935-7-tianci.yin@amd.com \
    --to=tianci.yin-5c7gfcevmho@public.gmane.org \
    --cc=Alexander.Deucher-5C7GfCeVMHo@public.gmane.org \
    --cc=Luben.Tuikov-5C7GfCeVMHo@public.gmane.org \
    --cc=amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org \
    --cc=christian.koenig-5C7GfCeVMHo@public.gmane.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.