[PATCH 1/2] accel/habanalabs: add pcie reset prepare/done hooks

[PATCH 1/2] accel/habanalabs: add pcie reset prepare/done hooks

[Oded Gabbay]

From: farah kassabri <fkassabri@habana.ai>

When working on a bare-metal system, if FLR will happen the firmware
will handle it and driver will have no knowledge of it, and this will
cause two issues:

1.The driver will be in operational state while it should be in reset.
  This will cause the heartbeat mechanism to keep sending messages to FW
  while pci device is in reset. Eventually heartbeat will fail and
  the device will end up in non-operational state.

2. After FW handles the FLR, and due to the reset it'll go back to
   preboot stage, and driver need to perform hard reset in order to
   load the boot fit binary.

This patch will add reset_prepare hook that will set the device to
be in disabled state, so it'll be not operational, and also
reset_done hook which will be called after the actual FLR handling,
then it will perform hard reset.

Signed-off-by: farah kassabri <fkassabri@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 .../accel/habanalabs/common/habanalabs_drv.c  | 34 +++++++++++++++++++
 1 file changed, 34 insertions(+)

diff --git a/drivers/accel/habanalabs/common/habanalabs_drv.c b/drivers/accel/habanalabs/common/habanalabs_drv.c
index 306a5bc9bf89..35ae0ff347f5 100644
--- a/drivers/accel/habanalabs/common/habanalabs_drv.c
+++ b/drivers/accel/habanalabs/common/habanalabs_drv.c
@@ -670,6 +670,38 @@ static pci_ers_result_t hl_pci_err_slot_reset(struct pci_dev *pdev)
 	return PCI_ERS_RESULT_RECOVERED;
 }
 
+static void hl_pci_reset_prepare(struct pci_dev *pdev)
+{
+	struct hl_device *hdev;
+
+	hdev = pci_get_drvdata(pdev);
+	if (!hdev)
+		return;
+
+	hdev->disabled = true;
+}
+
+static void hl_pci_reset_done(struct pci_dev *pdev)
+{
+	struct hl_device *hdev;
+	u32 flags;
+
+	hdev = pci_get_drvdata(pdev);
+	if (!hdev)
+		return;
+
+	/*
+	 * Schedule a thread to trigger hard reset.
+	 * The reason for this handler, is for rare cases where the driver is up
+	 * and FLR occurs. This is valid only when working with no VM, so FW handles FLR
+	 * and resets the device. FW will go back preboot stage, so driver needs to perform
+	 * hard reset in order to load FW fit again.
+	 */
+	flags = HL_DRV_RESET_HARD | HL_DRV_RESET_BYPASS_REQ_TO_FW;
+
+	hl_device_reset(hdev, flags);
+}
+
 static const struct dev_pm_ops hl_pm_ops = {
 	.suspend = hl_pmops_suspend,
 	.resume = hl_pmops_resume,
@@ -679,6 +711,8 @@ static const struct pci_error_handlers hl_pci_err_handler = {
 	.error_detected = hl_pci_err_detected,
 	.slot_reset = hl_pci_err_slot_reset,
 	.resume = hl_pci_err_resume,
+	.reset_prepare = hl_pci_reset_prepare,
+	.reset_done = hl_pci_reset_done,
 };
 
 static struct pci_driver hl_pci_driver = {
-- 
2.34.1

[PATCH 2/2] accel/habanalabs: update device boot error check

[Oded Gabbay]

From: Farah Kassabri <fkassabri@habana.ai>

Use a predefined mask which set the device critical boot errors.
Driver will fail and stop its loading, only upon detecting at least
one of those errors defined in this mask.

Signed-off-by: Farah Kassabri <fkassabri@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/accel/habanalabs/common/firmware_if.c | 122 +++++-------------
 1 file changed, 34 insertions(+), 88 deletions(-)

diff --git a/drivers/accel/habanalabs/common/firmware_if.c b/drivers/accel/habanalabs/common/firmware_if.c
index 47e8384134aa..ec04234e18a5 100644
--- a/drivers/accel/habanalabs/common/firmware_if.c
+++ b/drivers/accel/habanalabs/common/firmware_if.c
@@ -646,39 +646,27 @@ int hl_fw_send_heartbeat(struct hl_device *hdev)
 	return rc;
 }
 
-static bool fw_report_boot_dev0(struct hl_device *hdev, u32 err_val,
-								u32 sts_val)
+static bool fw_report_boot_dev0(struct hl_device *hdev, u32 err_val, u32 sts_val)
 {
 	bool err_exists = false;
 
 	if (!(err_val & CPU_BOOT_ERR0_ENABLED))
 		return false;
 
-	if (err_val & CPU_BOOT_ERR0_DRAM_INIT_FAIL) {
-		dev_err(hdev->dev,
-			"Device boot error - DRAM initialization failed\n");
-		err_exists = true;
-	}
+	if (err_val & CPU_BOOT_ERR0_DRAM_INIT_FAIL)
+		dev_err(hdev->dev, "Device boot error - DRAM initialization failed\n");
 
-	if (err_val & CPU_BOOT_ERR0_FIT_CORRUPTED) {
+	if (err_val & CPU_BOOT_ERR0_FIT_CORRUPTED)
 		dev_err(hdev->dev, "Device boot error - FIT image corrupted\n");
-		err_exists = true;
-	}
 
-	if (err_val & CPU_BOOT_ERR0_TS_INIT_FAIL) {
-		dev_err(hdev->dev,
-			"Device boot error - Thermal Sensor initialization failed\n");
-		err_exists = true;
-	}
+	if (err_val & CPU_BOOT_ERR0_TS_INIT_FAIL)
+		dev_err(hdev->dev, "Device boot error - Thermal Sensor initialization failed\n");
 
 	if (err_val & CPU_BOOT_ERR0_BMC_WAIT_SKIPPED) {
 		if (hdev->bmc_enable) {
-			dev_err(hdev->dev,
-				"Device boot error - Skipped waiting for BMC\n");
-			err_exists = true;
+			dev_err(hdev->dev, "Device boot error - Skipped waiting for BMC\n");
 		} else {
-			dev_info(hdev->dev,
-				"Device boot message - Skipped waiting for BMC\n");
+			dev_info(hdev->dev, "Device boot message - Skipped waiting for BMC\n");
 			/* This is an info so we don't want it to disable the
 			 * device
 			 */
@@ -686,101 +674,59 @@ static bool fw_report_boot_dev0(struct hl_device *hdev, u32 err_val,
 		}
 	}
 
-	if (err_val & CPU_BOOT_ERR0_NIC_DATA_NOT_RDY) {
-		dev_err(hdev->dev,
-			"Device boot error - Serdes data from BMC not available\n");
-		err_exists = true;
-	}
+	if (err_val & CPU_BOOT_ERR0_NIC_DATA_NOT_RDY)
+		dev_err(hdev->dev, "Device boot error - Serdes data from BMC not available\n");
 
-	if (err_val & CPU_BOOT_ERR0_NIC_FW_FAIL) {
-		dev_err(hdev->dev,
-			"Device boot error - NIC F/W initialization failed\n");
-		err_exists = true;
-	}
+	if (err_val & CPU_BOOT_ERR0_NIC_FW_FAIL)
+		dev_err(hdev->dev, "Device boot error - NIC F/W initialization failed\n");
 
-	if (err_val & CPU_BOOT_ERR0_SECURITY_NOT_RDY) {
-		dev_err(hdev->dev,
-			"Device boot warning - security not ready\n");
-		err_exists = true;
-	}
+	if (err_val & CPU_BOOT_ERR0_SECURITY_NOT_RDY)
+		dev_err(hdev->dev, "Device boot warning - security not ready\n");
 
-	if (err_val & CPU_BOOT_ERR0_SECURITY_FAIL) {
+	if (err_val & CPU_BOOT_ERR0_SECURITY_FAIL)
 		dev_err(hdev->dev, "Device boot error - security failure\n");
-		err_exists = true;
-	}
 
-	if (err_val & CPU_BOOT_ERR0_EFUSE_FAIL) {
+	if (err_val & CPU_BOOT_ERR0_EFUSE_FAIL)
 		dev_err(hdev->dev, "Device boot error - eFuse failure\n");
-		err_exists = true;
-	}
 
-	if (err_val & CPU_BOOT_ERR0_SEC_IMG_VER_FAIL) {
+	if (err_val & CPU_BOOT_ERR0_SEC_IMG_VER_FAIL)
 		dev_err(hdev->dev, "Device boot error - Failed to load preboot secondary image\n");
-		err_exists = true;
-	}
 
-	if (err_val & CPU_BOOT_ERR0_PLL_FAIL) {
+	if (err_val & CPU_BOOT_ERR0_PLL_FAIL)
 		dev_err(hdev->dev, "Device boot error - PLL failure\n");
-		err_exists = true;
-	}
 
-	if (err_val & CPU_BOOT_ERR0_TMP_THRESH_INIT_FAIL) {
+	if (err_val & CPU_BOOT_ERR0_TMP_THRESH_INIT_FAIL)
 		dev_err(hdev->dev, "Device boot error - Failed to set threshold for temperature sensor\n");
-		err_exists = true;
-	}
 
-	if (err_val & CPU_BOOT_ERR0_DEVICE_UNUSABLE_FAIL) {
-		/* Ignore this bit, don't prevent driver loading */
-		dev_dbg(hdev->dev, "device unusable status is set\n");
-		err_val &= ~CPU_BOOT_ERR0_DEVICE_UNUSABLE_FAIL;
-	}
+	if (err_val & CPU_BOOT_ERR0_DEVICE_UNUSABLE_FAIL)
+		dev_err(hdev->dev, "device unusable status is set\n");
 
-	if (err_val & CPU_BOOT_ERR0_BINNING_FAIL) {
+	if (err_val & CPU_BOOT_ERR0_BINNING_FAIL)
 		dev_err(hdev->dev, "Device boot error - binning failure\n");
-		err_exists = true;
-	}
 
 	if (sts_val & CPU_BOOT_DEV_STS0_ENABLED)
 		dev_dbg(hdev->dev, "Device status0 %#x\n", sts_val);
 
+	if (err_val & CPU_BOOT_ERR0_DRAM_SKIPPED)
+		dev_err(hdev->dev, "Device boot warning - Skipped DRAM initialization\n");
+
+	if (err_val & CPU_BOOT_ERR_ENG_ARC_MEM_SCRUB_FAIL)
+		dev_err(hdev->dev, "Device boot error - ARC memory scrub failed\n");
+
+	/* All warnings should go here in order not to reach the unknown error validation */
 	if (err_val & CPU_BOOT_ERR0_EEPROM_FAIL) {
 		dev_err(hdev->dev, "Device boot error - EEPROM failure detected\n");
 		err_exists = true;
 	}
 
-	/* All warnings should go here in order not to reach the unknown error validation */
-	if (err_val & CPU_BOOT_ERR0_DRAM_SKIPPED) {
-		dev_warn(hdev->dev,
-			"Device boot warning - Skipped DRAM initialization\n");
-		/* This is a warning so we don't want it to disable the
-		 * device
-		 */
-		err_val &= ~CPU_BOOT_ERR0_DRAM_SKIPPED;
-	}
-
-	if (err_val & CPU_BOOT_ERR0_PRI_IMG_VER_FAIL) {
-		dev_warn(hdev->dev,
-			"Device boot warning - Failed to load preboot primary image\n");
-		/* This is a warning so we don't want it to disable the
-		 * device as we have a secondary preboot image
-		 */
-		err_val &= ~CPU_BOOT_ERR0_PRI_IMG_VER_FAIL;
-	}
+	if (err_val & CPU_BOOT_ERR0_PRI_IMG_VER_FAIL)
+		dev_warn(hdev->dev, "Device boot warning - Failed to load preboot primary image\n");
 
-	if (err_val & CPU_BOOT_ERR0_TPM_FAIL) {
-		dev_warn(hdev->dev,
-			"Device boot warning - TPM failure\n");
-		/* This is a warning so we don't want it to disable the
-		 * device
-		 */
-		err_val &= ~CPU_BOOT_ERR0_TPM_FAIL;
-	}
+	if (err_val & CPU_BOOT_ERR0_TPM_FAIL)
+		dev_warn(hdev->dev, "Device boot warning - TPM failure\n");
 
-	if (!err_exists && (err_val & ~CPU_BOOT_ERR0_ENABLED)) {
-		dev_err(hdev->dev,
-			"Device boot error - unknown ERR0 error 0x%08x\n", err_val);
+	if (err_val & CPU_BOOT_ERR_FATAL_MASK)
 		err_exists = true;
-	}
 
 	/* return error only if it's in the predefined mask */
 	if (err_exists && ((err_val & ~CPU_BOOT_ERR0_ENABLED) &
-- 
2.34.1