On Tue, Jul 22, 2025 at 12:35:42PM +0100, Simon Horman wrote:

On Mon, Jul 21, 2025 at 07:32:26PM +0800, Dong Yibo wrote:

quoted

Initialize basic mbx function.

Signed-off-by: Dong Yibo <dong100@mucse.com>

...

quoted

diff --git a/drivers/net/ethernet/mucse/rnpgbe/rnpgbe_mbx.c b/drivers/net/ethernet/mucse/rnpgbe/rnpgbe_mbx.c

...

quoted

+/**
+ * mucse_obtain_mbx_lock_pf - obtain mailbox lock
+ * @hw: pointer to the HW structure
+ * @mbx_id: Id of vf/fw to obtain
+ *
+ * This function maybe used in an irq handler.
+ *
+ * @return: 0 if we obtained the mailbox lock
+ **/
+static int mucse_obtain_mbx_lock_pf(struct mucse_hw *hw, enum MBX_ID mbx_id)
+{
+	struct mucse_mbx_info *mbx = &hw->mbx;
+	int try_cnt = 5000, ret;
+	u32 reg;
+
+	reg = (mbx_id == MBX_FW) ? PF2FW_MBOX_CTRL(mbx) :
+				   PF2VF_MBOX_CTRL(mbx, mbx_id);
+	while (try_cnt-- > 0) {
+		/* Take ownership of the buffer */
+		mbx_wr32(hw, reg, MBOX_PF_HOLD);
+		/* force write back before check */
+		wmb();
+		if (mbx_rd32(hw, reg) & MBOX_PF_HOLD)
+			return 0;
+		udelay(100);
+	}
+	return ret;

ret is declared, and returned here.
But it is never initialised.

Perhaps it is appropriate to return an error value here,
and update the kernel doc for this function accordingly.

Flagged by W=1 builds with Clang 20.1.8, and Smatch.

quoted

+}

...

Got it, I will fix this.
Maybe my clang (10.0.0) is too old, I will update it and 
try W=1 again.
Thanks for your feedback.