[PATCH] habanalabs: add more protection of device during reset

[PATCH] habanalabs: add more protection of device during reset

[Oded Gabbay]

Prevent accesses to the device (register read/write) from debugfs entries
during reset as that can cause the device to get stuck.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/misc/habanalabs/debugfs.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/drivers/misc/habanalabs/debugfs.c b/drivers/misc/habanalabs/debugfs.c
index 1cf75010a379..20413e350343 100644
--- a/drivers/misc/habanalabs/debugfs.c
+++ b/drivers/misc/habanalabs/debugfs.c
@@ -528,6 +528,12 @@ static int engines_show(struct seq_file *s, void *data)
 	struct hl_dbg_device_entry *dev_entry = entry->dev_entry;
 	struct hl_device *hdev = dev_entry->hdev;
 
+	if (atomic_read(&hdev->in_reset)) {
+		dev_warn_ratelimited(hdev->dev,
+				"Can't check device idle during reset\n");
+		return 0;
+	}
+
 	hdev->asic_funcs->is_device_idle(hdev, NULL, s);
 
 	return 0;
@@ -640,6 +646,11 @@ static ssize_t hl_data_read32(struct file *f, char __user *buf,
 	u32 val;
 	ssize_t rc;
 
+	if (atomic_read(&hdev->in_reset)) {
+		dev_warn_ratelimited(hdev->dev, "Can't read during reset\n");
+		return 0;
+	}
+
 	if (*ppos)
 		return 0;
 
@@ -669,6 +680,11 @@ static ssize_t hl_data_write32(struct file *f, const char __user *buf,
 	u32 value;
 	ssize_t rc;
 
+	if (atomic_read(&hdev->in_reset)) {
+		dev_warn_ratelimited(hdev->dev, "Can't write during reset\n");
+		return 0;
+	}
+
 	rc = kstrtouint_from_user(buf, count, 16, &value);
 	if (rc)
 		return rc;
-- 
2.17.1

RE: [PATCH] habanalabs: add more protection of device during reset

[Tomer Tayar]

On Mon, Nov 18, 2019 at 15:07 Oded Gabbay wrote:
> Prevent accesses to the device (register read/write) from debugfs entries
> during reset as that can cause the device to get stuck.
> 
> Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

Reviewed-by: Tomer Tayar <ttayar@habana.ai>