From 4652f96a6ffb4f7055290181c5234893d0096789 Mon Sep 17 00:00:00 2001
From: Rajesh Devaraj <rdevaraj@nvidia.com>
Date: Sat, 19 Mar 2022 18:13:02 +0000
Subject: [PATCH] gpu: nvgpu: add polling for back-to-back error reporting in
 av+l

When an error is reported to Safety_Services, it will be cleared
at FSI and reported to SEH (System Error Handler). Since MISC_EC
interface provides only one register for error reporting, there
is a need to poll the status of previously reported error before
reporting the next error. For this purpose, this patch adds logic
to perform polling using epl_get_misc_ec_err_status(), in AV+L.

JIRA NVGPU-8094
Bug 200729736

Change-Id: Ia01a2fc42a7ce586b7965a82c90027a9a2dd252b
Signed-off-by: Rajesh Devaraj <rdevaraj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2684141
Reviewed-by: Dinesh T <dt@nvidia.com>
Reviewed-by: Ankur Kishore <ankkishore@nvidia.com>
GVS: Gerrit_Virtual_Submit
---
 .../gpu/nvgpu/os/linux/cic/cic_report_err.c   | 52 ++++++++++++++++++-
 drivers/gpu/nvgpu/os/linux/module.c           |  4 ++
 2 files changed, 55 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/nvgpu/os/linux/cic/cic_report_err.c b/drivers/gpu/nvgpu/os/linux/cic/cic_report_err.c
index 9e3cffdc0..4ce110c94 100644
--- a/drivers/gpu/nvgpu/os/linux/cic/cic_report_err.c
+++ b/drivers/gpu/nvgpu/os/linux/cic/cic_report_err.c
@@ -19,6 +19,7 @@
 
 #ifdef CONFIG_NVGPU_ENABLE_MISC_EC
 #include <linux/tegra-epl.h>
+#include <nvgpu/timers.h>
 #include "os/linux/os_linux.h"
 #endif
 
@@ -26,11 +27,60 @@ struct gk20a;
 
 int nvgpu_cic_mon_report_err_safety_services(struct gk20a *g, u32 err_id)
 {
-	int ret = 0U;
+	int ret = 0;
 
 #ifdef CONFIG_NVGPU_ENABLE_MISC_EC
+	u32 ss_retries = 0U;
+	bool ss_status;
 	struct device *dev = dev_from_gk20a(g);
 
+	if (g->enable_polling == true) {
+		/**
+		 * Do polling, if the error to be reported is corrected one.
+		 */
+		if ((err_id & ERR_TYPE_MASK) == 0U) {
+			for (ss_retries = MAX_SS_RETRIES; ss_retries > 0U;
+					ss_retries--) {
+				ret = epl_get_misc_ec_err_status(dev,
+						MISC_EC_SW_ERR_CODE_0,
+						&ss_status);
+				if (ret == 0) {
+					if (ss_status == true) {
+						/**
+						 * Previously reported error is
+						 * cleared at Safety_Services.
+						 */
+						break;
+					} else {
+						nvgpu_udelay(SS_WAIT_DURATION_US);
+						continue;
+					}
+				} else if (ret == -ENODEV) {
+					nvgpu_err(g, "Error reporting is not "
+						"supported in this platform");
+					ret = 0;
+					return ret;
+				} else {
+					nvgpu_err(g, "Error reporting failed");
+					return ret;
+				}
+			}
+			if (ss_retries == 0U) {
+				nvgpu_err(g, "Error reporting failed: previous"
+						"error is not cleared");
+				ret = -1;
+				return ret;
+			}
+		}
+	}
+
+	/**
+	 * Enable polling immediately after reporting of first error from boot.
+	 */
+	if (g->enable_polling == false) {
+		g->enable_polling = true;
+	}
+
 	/**
 	 * MISC_EC_SW_ERR_CODE_0 register has been allocated for NvGPU
 	 * to report GPU HW errors to Safety_Services via MISC_EC interface.
diff --git a/drivers/gpu/nvgpu/os/linux/module.c b/drivers/gpu/nvgpu/os/linux/module.c
index 2c7e4d595..0fd9ca926 100644
--- a/drivers/gpu/nvgpu/os/linux/module.c
+++ b/drivers/gpu/nvgpu/os/linux/module.c
@@ -452,6 +452,10 @@ int gk20a_pm_finalize_poweron(struct device *dev)
 		}
 	}
 
+#ifdef CONFIG_NVGPU_ENABLE_MISC_EC
+	g->enable_polling = false;
+#endif
+
 	err = gk20a_restore_registers(g);
 	if (err)
 		goto done;