habanalabs: reset device upon FD close if not idle

If device is not idle after user closes the FD we must reset device
as next user that will try to open FD will encounter a non-functional
device.

Signed-off-by: Ofir Bitton <obitton@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>

4 files changed, 19 insertions, 12 deletions

diff --git a/drivers/misc/habanalabs/common/context.c b/drivers/misc/habanalabs/common/context.c
index 62d705889ca8..19b6b045219e 100644
--- a/drivers/misc/habanalabs/common/context.c
+++ b/drivers/misc/habanalabs/common/context.c
@@ -12,7 +12,6 @@
 static void hl_ctx_fini(struct hl_ctx *ctx)
 {
 	struct hl_device *hdev = ctx->hdev;
-	u64 idle_mask[HL_BUSY_ENGINES_MASK_EXT_SIZE] = {0};
 	int i;
 
 	/* Release all allocated pending cb's, those cb's were never
@@ -57,14 +56,6 @@ static void hl_ctx_fini(struct hl_ctx *ctx)
 
 		/* Scrub both SRAM and DRAM */
 		hdev->asic_funcs->scrub_device_mem(hdev, 0, 0);
-
-		if ((!hdev->pldm) && (hdev->pdev) &&
-				(!hdev->asic_funcs->is_device_idle(hdev,
-					idle_mask,
-					HL_BUSY_ENGINES_MASK_EXT_SIZE, NULL)))
-			dev_notice(hdev->dev,
-					"device not idle after user context is closed (0x%llx, 0x%llx)\n",
-						idle_mask[0], idle_mask[1]);
 	} else {
 		dev_dbg(hdev->dev, "closing kernel context\n");
 		hdev->asic_funcs->ctx_fini(ctx);
diff --git a/drivers/misc/habanalabs/common/device.c b/drivers/misc/habanalabs/common/device.c
index bc58a91bf50a..0056282cec94 100644
--- a/drivers/misc/habanalabs/common/device.c
+++ b/drivers/misc/habanalabs/common/device.c
@@ -51,6 +51,8 @@ bool hl_device_operational(struct hl_device *hdev,
 
 static void hpriv_release(struct kref *ref)
 {
+	u64 idle_mask[HL_BUSY_ENGINES_MASK_EXT_SIZE] = {0};
+	bool device_is_idle = true;
 	struct hl_fpriv *hpriv;
 	struct hl_device *hdev;
 
@@ -71,7 +73,19 @@ static void hpriv_release(struct kref *ref)
 
 	kfree(hpriv);
 
-	if (hdev->reset_upon_device_release)
+	if ((!hdev->pldm) && (hdev->pdev) &&
+			(!hdev->asic_funcs->is_device_idle(hdev,
+				idle_mask,
+				HL_BUSY_ENGINES_MASK_EXT_SIZE, NULL))) {
+		dev_err(hdev->dev,
+			"device not idle after user context is closed (0x%llx_%llx)\n",
+			idle_mask[1], idle_mask[0]);
+
+		device_is_idle = false;
+	}
+
+	if ((hdev->reset_if_device_not_idle && !device_is_idle)
+			|| hdev->reset_upon_device_release)
 		hl_device_reset(hdev, 0);
 }
 
@@ -1108,8 +1122,8 @@ kill_processes:
 	if (!hdev->asic_funcs->is_device_idle(hdev, idle_mask,
 			HL_BUSY_ENGINES_MASK_EXT_SIZE, NULL)) {
 		dev_err(hdev->dev,
-			"device is not idle (mask %#llx %#llx) after reset\n",
-			idle_mask[0], idle_mask[1]);
+			"device is not idle (mask 0x%llx_%llx) after reset\n",
+			idle_mask[1], idle_mask[0]);
 		rc = -EIO;
 		goto out_err;
 	}
diff --git a/drivers/misc/habanalabs/common/habanalabs.h b/drivers/misc/habanalabs/common/habanalabs.h
index 56d2f41f8893..bcb5bfdd7f20 100644
--- a/drivers/misc/habanalabs/common/habanalabs.h
+++ b/drivers/misc/habanalabs/common/habanalabs.h
@@ -2311,6 +2311,7 @@ struct hl_device {
 	u8				rl_enable;
 	u8				reset_on_preboot_fail;
 	u8				reset_upon_device_release;
+	u8				reset_if_device_not_idle;
 };
 
 
diff --git a/drivers/misc/habanalabs/common/habanalabs_drv.c b/drivers/misc/habanalabs/common/habanalabs_drv.c
index 137e7dc63d3b..b55dd1c55166 100644
--- a/drivers/misc/habanalabs/common/habanalabs_drv.c
+++ b/drivers/misc/habanalabs/common/habanalabs_drv.c
@@ -264,6 +264,7 @@ static void set_driver_behavior_per_device(struct hl_device *hdev)
 	hdev->bmc_enable = 1;
 	hdev->hard_reset_on_fw_events = 1;
 	hdev->reset_on_preboot_fail = 1;
+	hdev->reset_if_device_not_idle = 1;
 
 	hdev->reset_pcilink = 0;
 	hdev->axi_drain = 0;