Merge tag 'amd-drm-next-6.14-2024-12-18' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.14-2024-12-18:

amdgpu:
- RAS updates
- ISP updates
- SDMA queue reset support
- Rework DPM powergating interfaces
- Documentation updates and cleanups
- Panel replay fixes
- DCN 3.5 updates
- DP tunneling fixes
- Use a pm notifier to more gracefully handle VRAM eviction on suspend or hibernate
- Add debugfs interfaces for forcing scheduling to specific engine instances
- GG 9.5 updates
- IH 4.4 updates
- Make missing optional firmware less noisy
- PSP 13.x updates
- SMU 13.x updates
- VCN 5.x updates
- JPEG 5.x updates
- Misc cleanups
- GC 12.x updates
- DRM panic support
- DC FAMS updates
- DSC fixes
- job handling fixes

amdkfd:
- GG 9.5 updates
- Logging improvements
- Misc cleanups
- Various Optimizations

Signed-off-by: Dave Airlie <airlied@redhat.com>

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241218201758.2580723-1-alexander.deucher@amd.com

1 files changed, 31 insertions, 0 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h
index ce4179db2a6d..a4a7e61817aa 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h
@@ -54,6 +54,22 @@
 
 /* Page retirement tag */
 #define UMC_ECC_NEW_DETECTED_TAG       0x1
+/*
+ * a flag to indicate v2 of channel index stored in eeprom
+ *
+ * v1 (legacy way): store channel index within a umc instance in eeprom
+ *    range in UMC v12: 0 ~ 7
+ * v2: store global channel index in eeprom
+ *    range in UMC v12: 0 ~ 127
+ *
+ * NOTE: it's better to store it in eeprom_table_record.mem_channel,
+ * but there is only 8 bits in mem_channel, and the channel number may
+ * increase in the future, we decide to save it in
+ * eeprom_table_record.retired_page. retired_page is useless in v2,
+ * we depend on eeprom_table_record.address instead of retired_page in v2.
+ * Only 48 bits are saved on eeprom, use bit 47 here.
+ */
+#define UMC_CHANNEL_IDX_V2	BIT_ULL(47)
 
 typedef int (*umc_func)(struct amdgpu_device *adev, uint32_t node_inst,
 			uint32_t umc_inst, uint32_t ch_inst, void *data);
@@ -70,6 +86,13 @@ struct amdgpu_umc_ras {
 			enum amdgpu_mca_error_type type, void *ras_error_status);
 	int (*update_ecc_status)(struct amdgpu_device *adev,
 			uint64_t status, uint64_t ipid, uint64_t addr);
+	int (*convert_ras_err_addr)(struct amdgpu_device *adev,
+			struct ras_err_data *err_data,
+			struct ta_ras_query_address_input *addr_in,
+			struct ta_ras_query_address_output *addr_out,
+			bool dump_addr);
+	uint32_t (*get_die_id_from_pa)(struct amdgpu_device *adev,
+			uint64_t mca_addr, uint64_t retired_page);
 };
 
 struct amdgpu_umc_funcs {
@@ -134,4 +157,12 @@ int amdgpu_umc_logs_ecc_err(struct amdgpu_device *adev,
 
 void amdgpu_umc_handle_bad_pages(struct amdgpu_device *adev,
 			void *ras_error_status);
+int amdgpu_umc_pages_in_a_row(struct amdgpu_device *adev,
+			struct ras_err_data *err_data, uint64_t pa_addr);
+int amdgpu_umc_lookup_bad_pages_in_a_row(struct amdgpu_device *adev,
+			uint64_t pa_addr, uint64_t *pfns, int len);
+int amdgpu_umc_mca_to_addr(struct amdgpu_device *adev,
+			uint64_t err_addr, uint32_t ch, uint32_t umc,
+			uint32_t node, uint32_t socket,
+			struct ta_ras_query_address_output *addr_out, bool dump_addr);
 #endif