[PATCH 2/2] drm/xe/i2c: Expose AMC Alert reason sysfs

[Badal Nilawar <badal.nilawar@intel.com>]

AMC raises an SMBUS alert before performing a power removal or
power-cycle operation. The xe driver then places the device into
vendor-specific wedge mode until the recovery is performed.

Expose a read-only xe_amc_alert_reason sysfs attribute to help users
identify the required recovery action.

Assisted-by: Claude:claude-opus-4.8
Signed-off-by: Badal Nilawar <badal.nilawar@intel.com>
---
 .../ABI/testing/sysfs-driver-intel-xe-amc     | 21 ++++++
 drivers/gpu/drm/xe/xe_amc.c                   | 65 ++++++++++++++++++-
 2 files changed, 84 insertions(+), 2 deletions(-)
 create mode 100644 Documentation/ABI/testing/sysfs-driver-intel-xe-amc

diff --git a/Documentation/ABI/testing/sysfs-driver-intel-xe-amc b/Documentation/ABI/testing/sysfs-driver-intel-xe-amc
new file mode 100644
index 000000000000..b3de933efe11
--- /dev/null
+++ b/Documentation/ABI/testing/sysfs-driver-intel-xe-amc
@@ -0,0 +1,21 @@
+What:		/sys/bus/pci/drivers/xe/.../xe_amc_alert_reason
+Date:		August 2026
+KernelVersion:	7.3
+Contact:	intel-xe@lists.freedesktop.org
+Description:
+		This file exposes the reason for the most recent Add-In
+		Management Controller (AMC) alert on Intel Xe platforms.
+
+		An AMC alert is delivered via an SMBUS interrupt and causes the
+		device to be wedged, requiring vendor-specific recovery. This
+		attribute is created when such an alert is handled and is
+		available to all users as read-only.
+
+		Read returns a single line containing one of the following
+		alert reasons:
+
+			Firmware Download
+			Thermal Trip
+			OOB Request
+			OOB Reset
+			Catastrophic
diff --git a/drivers/gpu/drm/xe/xe_amc.c b/drivers/gpu/drm/xe/xe_amc.c
index 8ecadee6eea3..ceb2c4d618fe 100644
--- a/drivers/gpu/drm/xe/xe_amc.c
+++ b/drivers/gpu/drm/xe/xe_amc.c
@@ -18,6 +18,7 @@
 #include "xe_device.h"
 #include "xe_i2c.h"
 #include "xe_mmio.h"
+#include "xe_printk.h"
 
 /**
  * DOC: Add-In Management Controller (AMC)
@@ -56,6 +57,8 @@ static const char * const amc_alert[] = {
 struct xe_amc {
 	struct xe_i2c *i2c;
 	struct work_struct work;
+	u8 alert_reason;
+	bool sysfs_created;
 };
 
 struct amc_header {
@@ -104,6 +107,54 @@ static const struct amc_request amc_get_alert_reason = {
 	},
 };
 
+/**
+ * DOC: AMC Alert Reason
+ *
+ * On Intel Xe platforms, AMC sends an alert notification via an SMBUS interrupt
+ * to notify events such as firmware download, thermal trip or a
+ * catastrophic error. See enum xe_amc_alert for the full list of reasons.
+ * Upon an AMC alert the device is wedged and requires vendor-specific recovery.
+ *
+ * The alert reason is exposed through
+ * /sys/bus/pci/drivers/xe/.../xe_amc_alert_reason
+ *
+ * See Documentation/ABI/testing/sysfs-driver-intel-xe-amc for the ABI
+ * specification.
+ */
+
+static ssize_t xe_amc_alert_reason_show(struct device *dev,
+					struct device_attribute *attr, char *buff)
+{
+	struct xe_device *xe = pdev_to_xe_device(to_pci_dev(dev));
+	struct xe_amc *amc = xe->i2c->amc;
+
+	return sysfs_emit(buff, "%s\n", amc_alert[amc->alert_reason]);
+}
+static DEVICE_ATTR_RO(xe_amc_alert_reason);
+
+static void xe_remove_amc_alert_sysfs(struct xe_device *xe)
+{
+	if (xe->i2c->amc->sysfs_created)
+		device_remove_file(xe->drm.dev, &dev_attr_xe_amc_alert_reason);
+}
+
+static void xe_create_amc_alert_sysfs(struct xe_device *xe)
+{
+	struct device *dev = xe->drm.dev;
+	int ret;
+
+	if (xe->i2c->amc->sysfs_created)
+		return;
+
+	ret = device_create_file(dev, &dev_attr_xe_amc_alert_reason);
+	if (ret)
+		goto failed;
+
+	xe->i2c->amc->sysfs_created = true;
+failed:
+	dev_err(dev, "Failed to create sysfs file for amc alert reason\n");
+}
+
 static void xe_amc_work(struct work_struct *work)
 {
 	const struct amc_request *request = &amc_get_alert_reason;
@@ -158,10 +209,16 @@ static void xe_amc_work(struct work_struct *work)
 	case AMC_ALERT_THERMAL_TRIP:
 	case AMC_ALERT_OOB_REQUEST:
 	case AMC_ALERT_OOB_RESET:
-	case AMC_ALERT_CATERR:
+	case AMC_ALERT_CATERR: {
+		struct xe_device *xe = i2c_client_to_xe_device(client);
+
 		dev_warn(amc->i2c->drm_dev, "AMC Alert: %s\n", amc_alert[alert_reason]);
-		xe_device_declare_wedged(i2c_client_to_xe_device(client));
+		amc->alert_reason = alert_reason;
+		xe_device_set_wedged_method(xe, DRM_WEDGE_RECOVERY_VENDOR);
+		xe_device_declare_wedged(xe);
+		xe_create_amc_alert_sysfs(xe);
 		break;
+	}
 	default:
 		dev_warn(amc->i2c->drm_dev, "unknown AMC alert: %d\n", alert_reason);
 		break;
@@ -190,8 +247,12 @@ int xe_amc_init(struct xe_i2c *i2c)
 
 void xe_amc_exit(struct xe_i2c *i2c)
 {
+	struct xe_device *xe;
+
 	if (i2c->amc) {
 		cancel_work_sync(&i2c->amc->work);
+		xe = i2c_client_to_xe_device(i2c->client[XE_I2C_CLIENT_AMC]);
+		xe_remove_amc_alert_sysfs(xe);
 		kfree(i2c->amc);
 	}
 }
-- 
2.54.0