drm/amdgpu: add ras support for gfx of aldebaran

add edc counter/status reset and query functions for gfx block of
aldebaran.

v2: change to clear edc counter explicitly
aldebaran hardware will not clear edc counter after driver reading them,
so driver should clear them explicitly.
Signed-off-by: Dennis Li <Dennis.Li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h

@@ -30,6 +30,7 @@
 #include "clearstate_defs.h"
 #include "amdgpu_ring.h"
 #include "amdgpu_rlc.h"
+#include "soc15.h"
 
 /* GFX current status */
 #define AMDGPU_GFX_NORMAL_MODE			0x00000000L

drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c

@@ -2113,6 +2113,19 @@ static const struct amdgpu_gfx_funcs gfx_v9_4_gfx_funcs = {
 	.query_ras_error_status = &gfx_v9_4_query_ras_error_status,
 };
 
+static const struct amdgpu_gfx_funcs gfx_v9_4_2_gfx_funcs = {
+	.get_gpu_clock_counter = &gfx_v9_0_get_gpu_clock_counter,
+	.select_se_sh = &gfx_v9_0_select_se_sh,
+	.read_wave_data = &gfx_v9_0_read_wave_data,
+	.read_wave_sgprs = &gfx_v9_0_read_wave_sgprs,
+	.read_wave_vgprs = &gfx_v9_0_read_wave_vgprs,
+	.select_me_pipe_q = &gfx_v9_0_select_me_pipe_q,
+	.ras_error_inject = &gfx_v9_4_2_ras_error_inject,
+	.query_ras_error_count = &gfx_v9_4_2_query_ras_error_count,
+	.reset_ras_error_count = &gfx_v9_4_2_reset_ras_error_count,
+	.query_ras_error_status = &gfx_v9_4_2_query_ras_error_status,
+};
+
 static int gfx_v9_0_gpu_early_init(struct amdgpu_device *adev)
 {
 	u32 gb_addr_config;
@@ -2185,6 +2198,7 @@ static int gfx_v9_0_gpu_early_init(struct amdgpu_device *adev)
 		gb_addr_config |= 0x22010042;
 		break;
 	case CHIP_ALDEBARAN:
+		adev->gfx.funcs = &gfx_v9_4_2_gfx_funcs;
 		adev->gfx.config.max_hw_contexts = 8;
 		adev->gfx.config.sc_prim_fifo_size_frontend = 0x20;
 		adev->gfx.config.sc_prim_fifo_size_backend = 0x100;

drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.h

@@ -29,4 +29,10 @@ void gfx_v9_4_2_debug_trap_config_init(struct amdgpu_device *adev,
 void gfx_v9_4_2_init_golden_registers(struct amdgpu_device *adev,
 				      uint32_t die_id);
 void gfx_v9_4_2_set_power_brake_sequence(struct amdgpu_device *adev);
+
+void gfx_v9_4_2_reset_ras_error_count(struct amdgpu_device *adev);
+int gfx_v9_4_2_ras_error_inject(struct amdgpu_device *adev, void *inject_if);
+void gfx_v9_4_2_query_ras_error_status(struct amdgpu_device *adev);
+int gfx_v9_4_2_query_ras_error_count(struct amdgpu_device *adev,
+				   void *ras_error_status);
 #endif /* __GFX_V9_4_2_H__ */

drivers/gpu/drm/amd/amdgpu/soc15.h

@@ -49,6 +49,13 @@ struct soc15_reg_rlcg {
 	u32	reg;
 };
 
+struct soc15_reg {
+	uint32_t hwip;
+	uint32_t inst;
+	uint32_t seg;
+	uint32_t reg_offset;
+};
+
 struct soc15_reg_entry {
 	uint32_t hwip;
 	uint32_t inst;
@@ -88,6 +95,10 @@ struct soc15_ras_field_entry {
 
 #define SOC15_REG_FIELD(reg, field) reg##__##field##_MASK, reg##__##field##__SHIFT
 
+#define SOC15_REG_FIELD_VAL(val, mask, shift)	(((val) & mask) >> shift)
+
+#define SOC15_RAS_REG_FIELD_VAL(val, entry, field) SOC15_REG_FIELD_VAL(val, entry.field##_count_mask, entry.field##_count_shift)
+
 void soc15_grbm_select(struct amdgpu_device *adev,
 		    u32 me, u32 pipe, u32 queue, u32 vmid);
 void soc15_set_virt_ops(struct amdgpu_device *adev);

drivers/gpu/drm/amd/amdgpu/soc15_common.h

@@ -118,6 +118,24 @@
 		}	\
 	} while (0)
 
+#define WREG32_SOC15_RLC_SHADOW_EX(prefix, ip, inst, reg, value) \
+	do {							\
+		uint32_t target_reg = adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg;\
+		if (amdgpu_sriov_fullaccess(adev)) {    \
+			uint32_t r2 = adev->reg_offset[GC_HWIP][0][prefix##SCRATCH_REG1_BASE_IDX] + prefix##SCRATCH_REG2;	\
+			uint32_t r3 = adev->reg_offset[GC_HWIP][0][prefix##SCRATCH_REG1_BASE_IDX] + prefix##SCRATCH_REG3;	\
+			uint32_t grbm_cntl = adev->reg_offset[GC_HWIP][0][prefix##GRBM_GFX_CNTL_BASE_IDX] + prefix##GRBM_GFX_CNTL;   \
+			uint32_t grbm_idx = adev->reg_offset[GC_HWIP][0][prefix##GRBM_GFX_INDEX_BASE_IDX] + prefix##GRBM_GFX_INDEX;   \
+			if (target_reg == grbm_cntl) \
+				WREG32(r2, value);	\
+			else if (target_reg == grbm_idx) \
+				WREG32(r3, value);	\
+			WREG32(target_reg, value);	\
+		} else {	\
+			WREG32(target_reg, value); \
+		}	\
+	} while (0)
+
 #define WREG32_SOC15_RLC(ip, inst, reg, value) \
 	do {							\
 			uint32_t target_reg = adev->reg_offset[GC_HWIP][0][reg##_BASE_IDX] + reg;\