wlcore: fix runtime pm imbalance in wlcore_irq_locked
authorDinghao Liu <dinghao.liu@zju.edu.cn>
Fri, 22 May 2020 04:49:04 +0000 (12:49 +0800)
committerKalle Valo <kvalo@codeaurora.org>
Fri, 29 May 2020 17:38:15 +0000 (20:38 +0300)
When wlcore_fw_status() returns an error code, a pairing
runtime PM usage counter decrement is needed to keep the
counter balanced. It's the same for all error paths after
wlcore_fw_status().

Signed-off-by: Dinghao Liu <dinghao.liu@zju.edu.cn>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20200522044906.29564-1-dinghao.liu@zju.edu.cn
drivers/net/wireless/ti/wlcore/main.c

index 0dcad49..de6c8a7 100644 (file)
@@ -548,7 +548,7 @@ static int wlcore_irq_locked(struct wl1271 *wl)
 
                ret = wlcore_fw_status(wl, wl->fw_status);
                if (ret < 0)
-                       goto out;
+                       goto err_ret;
 
                wlcore_hw_tx_immediate_compl(wl);
 
@@ -565,7 +565,7 @@ static int wlcore_irq_locked(struct wl1271 *wl)
                        ret = -EIO;
 
                        /* restarting the chip. ignore any other interrupt. */
-                       goto out;
+                       goto err_ret;
                }
 
                if (unlikely(intr & WL1271_ACX_SW_INTR_WATCHDOG)) {
@@ -575,7 +575,7 @@ static int wlcore_irq_locked(struct wl1271 *wl)
                        ret = -EIO;
 
                        /* restarting the chip. ignore any other interrupt. */
-                       goto out;
+                       goto err_ret;
                }
 
                if (likely(intr & WL1271_ACX_INTR_DATA)) {
@@ -583,7 +583,7 @@ static int wlcore_irq_locked(struct wl1271 *wl)
 
                        ret = wlcore_rx(wl, wl->fw_status);
                        if (ret < 0)
-                               goto out;
+                               goto err_ret;
 
                        /* Check if any tx blocks were freed */
                        spin_lock_irqsave(&wl->wl_lock, flags);
@@ -596,7 +596,7 @@ static int wlcore_irq_locked(struct wl1271 *wl)
                                 */
                                ret = wlcore_tx_work_locked(wl);
                                if (ret < 0)
-                                       goto out;
+                                       goto err_ret;
                        } else {
                                spin_unlock_irqrestore(&wl->wl_lock, flags);
                        }
@@ -604,7 +604,7 @@ static int wlcore_irq_locked(struct wl1271 *wl)
                        /* check for tx results */
                        ret = wlcore_hw_tx_delayed_compl(wl);
                        if (ret < 0)
-                               goto out;
+                               goto err_ret;
 
                        /* Make sure the deferred queues don't get too long */
                        defer_count = skb_queue_len(&wl->deferred_tx_queue) +
@@ -617,14 +617,14 @@ static int wlcore_irq_locked(struct wl1271 *wl)
                        wl1271_debug(DEBUG_IRQ, "WL1271_ACX_INTR_EVENT_A");
                        ret = wl1271_event_handle(wl, 0);
                        if (ret < 0)
-                               goto out;
+                               goto err_ret;
                }
 
                if (intr & WL1271_ACX_INTR_EVENT_B) {
                        wl1271_debug(DEBUG_IRQ, "WL1271_ACX_INTR_EVENT_B");
                        ret = wl1271_event_handle(wl, 1);
                        if (ret < 0)
-                               goto out;
+                               goto err_ret;
                }
 
                if (intr & WL1271_ACX_INTR_INIT_COMPLETE)
@@ -635,6 +635,7 @@ static int wlcore_irq_locked(struct wl1271 *wl)
                        wl1271_debug(DEBUG_IRQ, "WL1271_ACX_INTR_HW_AVAILABLE");
        }
 
+err_ret:
        pm_runtime_mark_last_busy(wl->dev);
        pm_runtime_put_autosuspend(wl->dev);