[PATCH 055/109] drm/amdgpu: Add bad page preloading and pre-reservation support

From: YiPeng Chai <[email protected]>

Add bad page preloading and pre-reservation support.

v2:
  Add modification to ensure amdgpu_sw_init does not fail
  and remove deferred ras instance release.

Signed-off-by: YiPeng Chai <[email protected]>
Reviewed-by: Hawking Zhang <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c    | 1 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c          | 7 ++++++-
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c          | 4 ++++
 drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.c | 8 ++------
 drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.h | 2 ++
 5 files changed, 15 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
index a404d8aa13eee..c03d3c40b4587 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
@@ -2840,6 +2840,7 @@ static int amdgpu_discovery_set_ras_ip_blocks(struct amdgpu_device *adev)
 	case IP_VERSION(13, 0, 6):
 	case IP_VERSION(13, 0, 12):
 	case IP_VERSION(13, 0, 14):
+	case IP_VERSION(15, 0, 8):
 		amdgpu_device_ip_block_add(adev, &ras_v1_0_ip_block);
 		break;
 	default:
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index 1cd12eeee12b6..d3872e4787460 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -3590,7 +3590,10 @@ int amdgpu_ras_init(struct amdgpu_device *adev)
 
 	amdgpu_ras_check_supported(adev);
 
-	if (!adev->ras_enabled || adev->asic_type == CHIP_VEGA10) {
+	amdgpu_ras_mgr_sw_init(adev);
+
+	if (!con->uniras_enabled &&
+	    (!adev->ras_enabled || adev->asic_type == CHIP_VEGA10)) {
 		/* set gfx block ras context feature for VEGA20 Gaming
 		 * send ras disable cmd to ras ta during ras late init.
 		 */
@@ -3980,6 +3983,8 @@ int amdgpu_ras_fini(struct amdgpu_device *adev)
 	struct amdgpu_ras_block_object *obj = NULL;
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 
+	amdgpu_ras_mgr_sw_fini(adev);
+
 	if (!adev->ras_enabled || !con)
 		return 0;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 5bcb6a966941d..8edf1791fa4db 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -56,6 +56,7 @@
 #include "amdgpu_amdkfd.h"
 #include "amdgpu_sdma.h"
 #include "amdgpu_ras.h"
+#include "amdgpu_ras_mgr.h"
 #include "amdgpu_hmm.h"
 #include "amdgpu_atomfirmware.h"
 #include "amdgpu_res_cursor.h"
@@ -2158,6 +2159,9 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	if (r)
 		return r;
 
+	/* RAS loads and reserves bad pages */
+	amdgpu_ras_mgr_early_init_service(adev);
+
 	if (adev->mman.resv_region[AMDGPU_RESV_MEM_TRAIN].size) {
 		struct psp_memory_training_context *ctx =
 					&adev->psp.mem_train_ctx;
diff --git a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.c b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.c
index 128a9bc1f2f09..525abb7e05474 100644
--- a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.c
+++ b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.c
@@ -277,9 +277,8 @@ static struct ras_core_context *amdgpu_ras_mgr_create_ras_core(struct amdgpu_dev
 	return ras_core_create(&init_config);
 }
 
-static int amdgpu_ras_mgr_sw_init(struct amdgpu_ip_block *ip_block)
+int amdgpu_ras_mgr_sw_init(struct amdgpu_device *adev)
 {
-	struct amdgpu_device *adev = ip_block->adev;
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 	struct amdgpu_ras_mgr *ras_mgr;
 	int ret = 0;
@@ -346,9 +345,8 @@ static int amdgpu_ras_mgr_sw_init(struct amdgpu_ip_block *ip_block)
 	return ret;
 }
 
-static int amdgpu_ras_mgr_sw_fini(struct amdgpu_ip_block *ip_block)
+int amdgpu_ras_mgr_sw_fini(struct amdgpu_device *adev)
 {
-	struct amdgpu_device *adev = ip_block->adev;
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 	struct amdgpu_ras_mgr *ras_mgr;
 
@@ -464,8 +462,6 @@ struct amdgpu_ras_mgr *amdgpu_ras_mgr_get_context(struct amdgpu_device *adev)
 
 static const struct amd_ip_funcs __maybe_unused ras_v1_0_ip_funcs = {
 	.name = "ras_v1_0",
-	.sw_init = amdgpu_ras_mgr_sw_init,
-	.sw_fini = amdgpu_ras_mgr_sw_fini,
 	.hw_init = amdgpu_ras_mgr_hw_init,
 	.hw_fini = amdgpu_ras_mgr_hw_fini,
 };
diff --git a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.h b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.h
index cb98606306932..32416c9d6086e 100644
--- a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.h
+++ b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.h
@@ -61,6 +61,8 @@ struct amdgpu_ras_mgr {
 
 extern const struct amdgpu_ip_block_version ras_v1_0_ip_block;
 
+int amdgpu_ras_mgr_sw_init(struct amdgpu_device *adev);
+int amdgpu_ras_mgr_sw_fini(struct amdgpu_device *adev);
 int amdgpu_ras_mgr_early_init_service(struct amdgpu_device *adev);
 struct amdgpu_ras_mgr *amdgpu_ras_mgr_get_context(
 			struct amdgpu_device *adev);
-- 
2.55.0