[PATCH 2/3] Bluetooth: btqcomipc: deassert the BTSS reset before enabling its clock

[Andrei-Alexandru Bleortu <me@andrei-z.com>]

An unbind asserts the BTSS reset through devres, and a later probe then
enabled the LPO branch clock with the subsystem still in reset:
"gcc_btss_lpo_clk status stuck at 'off'" and the probe failed with
-EBUSY. Deassert the reset first; devres then also undoes them in the
right order (clock off, then reset).

Assisted-by: LLM
Signed-off-by: Andrei-Alexandru Bleortu <me@andrei-z.com>
---
 drivers/bluetooth/btqcomipc.c | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/drivers/bluetooth/btqcomipc.c b/drivers/bluetooth/btqcomipc.c
index 21dd3e812..8513c1780 100644
--- a/drivers/bluetooth/btqcomipc.c
+++ b/drivers/bluetooth/btqcomipc.c
@@ -1147,16 +1147,20 @@ static int btqcomipc_probe(struct platform_device *pdev)
 
 	of_node_put(ipc_np);
 
-	lpo_clk = devm_clk_get_enabled(dev, "lpo");
-	if (IS_ERR(lpo_clk))
-		return dev_err_probe(dev, PTR_ERR(lpo_clk),
-				     "Failed to get lpo clock\n");
-
+	/*
+	 * The LPO branch clock cannot turn on while the subsystem is held
+	 * in reset, which is how an unbind leaves it: deassert first.
+	 */
 	btss_reset = devm_reset_control_get_exclusive_deasserted(dev, NULL);
 	if (IS_ERR_OR_NULL(btss_reset))
 		return dev_err_probe(dev, PTR_ERR(btss_reset),
 				     "unable to deassert reset\n");
 
+	lpo_clk = devm_clk_get_enabled(dev, "lpo");
+	if (IS_ERR(lpo_clk))
+		return dev_err_probe(dev, PTR_ERR(lpo_clk),
+				     "Failed to get lpo clock\n");
+
 	desc->irq = platform_get_irq(pdev, 0);
 	if (desc->irq < 0)
 		return dev_err_probe(dev, desc->irq, "Failed to acquire IRQ\n");