gpu: nvgpu: vgpu: ecc sysfs support for vgpu

- fetch ecc info from RM server and create sysfs nodes
- new file ecc_vgpu.c for platform-independent code
- add 2 new commands: GET_ECC_INFO and GET_ECC_COUNTER_VALUE

JIRA EVLR-2590

Change-Id: I040a9fcd23326e432ca93e9a028319f9c1c570f0
Signed-off-by: Kyle Guo <kyleg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1777428
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

2 files changed, 22 insertions, 0 deletions

diff --git a/drivers/gpu/nvgpu/include/nvgpu/vgpu/tegra_vgpu.h b/drivers/gpu/nvgpu/include/nvgpu/vgpu/tegra_vgpu.h
index 1eb26303..39d68dd1 100644
--- a/drivers/gpu/nvgpu/include/nvgpu/vgpu/tegra_vgpu.h
+++ b/drivers/gpu/nvgpu/include/nvgpu/vgpu/tegra_vgpu.h
@@ -26,6 +26,7 @@
 #define __TEGRA_VGPU_H
 
 #include <nvgpu/types.h>
+#include <nvgpu/ecc.h>  /* For NVGPU_ECC_STAT_NAME_MAX_SIZE */
 
 enum {
         TEGRA_VGPU_MODULE_GPU = 0,
@@ -120,6 +121,8 @@ enum {
         TEGRA_VGPU_CMD_UPDATE_PC_SAMPLING = 81,
         TEGRA_VGPU_CMD_SUSPEND = 82,
         TEGRA_VGPU_CMD_RESUME = 83,
+        TEGRA_VGPU_CMD_GET_ECC_INFO = 84,
+        TEGRA_VGPU_CMD_GET_ECC_COUNTER_VALUE = 85,
 };
 
 struct tegra_vgpu_connect_params {
@@ -378,6 +381,20 @@ struct tegra_vgpu_channel_free_hwpm_ctx {
         u64 handle;
 };
 
+struct tegra_vgpu_ecc_info_params {
+        u32 ecc_stats_count;
+};
+
+struct tegra_vgpu_ecc_info_entry {
+        u32 ecc_id;
+        char name[NVGPU_ECC_STAT_NAME_MAX_SIZE];
+};
+
+struct tegra_vgpu_ecc_counter_params {
+        u32 ecc_id;
+        u32 value;
+};
+
 struct tegra_vgpu_gr_ctx_params {
         u64 gr_ctx_handle;
         u64 as_handle;
@@ -659,6 +676,8 @@ struct tegra_vgpu_cmd_msg {
                 struct tegra_vgpu_map_syncpt_params map_syncpt;
                 struct tegra_vgpu_tsg_bind_channel_ex_params tsg_bind_channel_ex;
                 struct tegra_vgpu_channel_update_pc_sampling update_pc_sampling;
+                struct tegra_vgpu_ecc_info_params ecc_info;
+                struct tegra_vgpu_ecc_counter_params ecc_counter;
                 char padding[192];
         } params;
 };
diff --git a/drivers/gpu/nvgpu/include/nvgpu/vgpu/vgpu.h b/drivers/gpu/nvgpu/include/nvgpu/vgpu/vgpu.h
index 1e851b8e..762bc229 100644
--- a/drivers/gpu/nvgpu/include/nvgpu/vgpu/vgpu.h
+++ b/drivers/gpu/nvgpu/include/nvgpu/vgpu/vgpu.h
@@ -37,11 +37,14 @@ struct gk20a;
 struct vm_gk20a;
 struct nvgpu_gr_ctx;
 struct nvgpu_cpu_time_correlation_sample;
+struct vgpu_ecc_stat;
 
 struct vgpu_priv_data {
         u64 virt_handle;
         struct nvgpu_thread intr_handler;
         struct tegra_vgpu_constants_params constants;
+        struct vgpu_ecc_stat *ecc_stats;
+        int ecc_stats_count;
 };
 
 struct vgpu_priv_data *vgpu_get_priv_data(struct gk20a *g);