habanalabs: add uapi to retrieve engines status

Currently, to get engines status, user needed to read debugfs file
with root permissions.

This new uapi allows user apace apps retrieve status, so for example,
in case of failure, status can be retrieved immediately by the
application itself which runs without root permissions.

Signed-off-by: Dani Liberman <dliberman@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>

2 files changed, 41 insertions, 2 deletions

diff --git a/drivers/misc/habanalabs/common/debugfs.c b/drivers/misc/habanalabs/common/debugfs.c
index 90c91c1b2c10..c297248748d3 100644
--- a/drivers/misc/habanalabs/common/debugfs.c
+++ b/drivers/misc/habanalabs/common/debugfs.c
@@ -17,7 +17,6 @@
 #define MMU_ASID_BUF_SIZE	10
 #define MMU_KBUF_SIZE		(MMU_ADDR_BUF_SIZE + MMU_ASID_BUF_SIZE)
 #define I2C_MAX_TRANSACTION_LEN	8
-#define ENGINES_DATA_MAX_SIZE	SZ_16K
 
 static struct dentry *hl_debug_root;
 
@@ -626,7 +625,7 @@ static int engines_show(struct seq_file *s, void *data)
 	}
 
 	eng_data.actual_size = 0;
-	eng_data.allocated_buf_size = ENGINES_DATA_MAX_SIZE;
+	eng_data.allocated_buf_size = HL_ENGINES_DATA_MAX_SIZE;
 	eng_data.buf = vmalloc(eng_data.allocated_buf_size);
 	if (!eng_data.buf)
 		return -ENOMEM;
diff --git a/drivers/misc/habanalabs/common/habanalabs_ioctl.c b/drivers/misc/habanalabs/common/habanalabs_ioctl.c
index 6a30bd98ab5e..ec55c66fedd6 100644
--- a/drivers/misc/habanalabs/common/habanalabs_ioctl.c
+++ b/drivers/misc/habanalabs/common/habanalabs_ioctl.c
@@ -14,6 +14,7 @@
 #include <linux/fs.h>
 #include <linux/uaccess.h>
 #include <linux/slab.h>
+#include <linux/vmalloc.h>
 
 static u32 hl_debug_struct_size[HL_DEBUG_OP_TIMESTAMP + 1] = {
 	[HL_DEBUG_OP_ETR] = sizeof(struct hl_debug_params_etr),
@@ -697,6 +698,42 @@ static int eventfd_unregister(struct hl_fpriv *hpriv, struct hl_info_args *args)
 	return 0;
 }
 
+static int engine_status_info(struct hl_fpriv *hpriv, struct hl_info_args *args)
+{
+	void __user *out = (void __user *) (uintptr_t) args->return_pointer;
+	u32 status_buf_size = args->return_size;
+	struct hl_device *hdev = hpriv->hdev;
+	struct engines_data eng_data;
+	int rc;
+
+	if ((status_buf_size < SZ_1K) || (status_buf_size > HL_ENGINES_DATA_MAX_SIZE) || (!out))
+		return -EINVAL;
+
+	eng_data.actual_size = 0;
+	eng_data.allocated_buf_size = status_buf_size;
+	eng_data.buf = vmalloc(status_buf_size);
+	if (!eng_data.buf)
+		return -ENOMEM;
+
+	hdev->asic_funcs->is_device_idle(hdev, NULL, 0, &eng_data);
+
+	if (eng_data.actual_size > eng_data.allocated_buf_size) {
+		dev_err(hdev->dev,
+			"Engines data size (%d Bytes) is bigger than allocated size (%u Bytes)\n",
+			eng_data.actual_size, status_buf_size);
+		vfree(eng_data.buf);
+		return -ENOMEM;
+	}
+
+	args->user_buffer_actual_size = eng_data.actual_size;
+	rc = copy_to_user(out, eng_data.buf, min_t(size_t, status_buf_size, eng_data.actual_size)) ?
+				-EFAULT : 0;
+
+	vfree(eng_data.buf);
+
+	return rc;
+}
+
 static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data,
 				struct device *dev)
 {
@@ -812,6 +849,9 @@ static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data,
 	case HL_INFO_UNREGISTER_EVENTFD:
 		return eventfd_unregister(hpriv, args);
 
+	case HL_INFO_ENGINE_STATUS:
+		return engine_status_info(hpriv, args);
+
 	default:
 		dev_err(dev, "Invalid request %d\n", args->op);
 		rc = -EINVAL;