wifi: rt2x00: introduce DMA busy check watchdog for rt2800

When I tried to fix the watchdog of rt2800, I found that sometimes
the watchdog can not reset the hung device. This is because the
queue is not completely stuck, it just becomes very slow. The MTK
vendor driver for the new chip MT7603/MT7612 has a DMA busy watchdog
to detect device hangs by checking DMA busy status. This watchdog
implementation is something similar to it. To reduce unnecessary
reset, we can check the INT_SOURCE_CSR register together as I found
that when the radio hung, the RX/TX coherent interrupt will always
stuck at triggered state.

The 'watchdog' module parameter has been extended to control all
watchdogs(0=disabled, 1=hang watchdog, 2=DMA watchdog, 3=both). This
new watchdog function is a slight schedule and it won't affect the
transmission speed. So we can turn on it by default. Due to the
INT_SOURCE_CSR register is invalid on rt2800 USB NICs, the DMA busy
watchdog will be automatically disabled for them.

Tested on MT7620 and RT5350.
Signed-off-by: Shiji Yang <yangshiji66@outlook.com>
Acked-by: Stanislaw Gruszka <stf_xl@wp.pl>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/TYAP286MB0315D7462CE08A119A99DE34BCA4A@TYAP286MB0315.JPNP286.PROD.OUTLOOK.COM

drivers/net/wireless/ralink/rt2x00/rt2800.h

@@ -3194,4 +3194,8 @@ enum rt2800_eeprom_word {
  */
 #define BCN_TBTT_OFFSET 64
 
+/* Watchdog type mask */
+#define RT2800_WATCHDOG_HANG		BIT(0)
+#define RT2800_WATCHDOG_DMA_BUSY	BIT(1)
+
 #endif /* RT2800_H */

drivers/net/wireless/ralink/rt2x00/rt2800lib.c

@@ -30,9 +30,10 @@
 #include "rt2800lib.h"
 #include "rt2800.h"
 
-static bool modparam_watchdog;
-module_param_named(watchdog, modparam_watchdog, bool, S_IRUGO);
-MODULE_PARM_DESC(watchdog, "Enable watchdog to detect tx/rx hangs and reset hardware if detected");
+static unsigned int modparam_watchdog = RT2800_WATCHDOG_DMA_BUSY;
+module_param_named(watchdog, modparam_watchdog, uint, 0444);
+MODULE_PARM_DESC(watchdog, "Enable watchdog to recover tx/rx hangs.\n"
+		 "\t\t(0=disabled, 1=hang watchdog, 2=DMA watchdog(default), 3=both)");
 
 /*
  * Register access.
@@ -1261,15 +1262,12 @@ static void rt2800_update_survey(struct rt2x00_dev *rt2x00dev)
 	chan_survey->time_ext_busy += rt2800_register_read(rt2x00dev, CH_BUSY_STA_SEC);
 }
 
-void rt2800_watchdog(struct rt2x00_dev *rt2x00dev)
+static bool rt2800_watchdog_hung(struct rt2x00_dev *rt2x00dev)
 {
 	struct data_queue *queue;
 	bool hung_tx = false;
 	bool hung_rx = false;
 
-	if (test_bit(DEVICE_STATE_SCANNING, &rt2x00dev->flags))
-		return;
-
 	rt2800_update_survey(rt2x00dev);
 
 	queue_for_each(rt2x00dev, queue) {
@@ -1297,18 +1295,72 @@ void rt2800_watchdog(struct rt2x00_dev *rt2x00dev)
 		}
 	}
 
+	if (!hung_tx && !hung_rx)
+		return false;
+
 	if (hung_tx)
 		rt2x00_warn(rt2x00dev, "Watchdog TX hung detected\n");
 
 	if (hung_rx)
 		rt2x00_warn(rt2x00dev, "Watchdog RX hung detected\n");
 
-	if (hung_tx || hung_rx) {
-		queue_for_each(rt2x00dev, queue)
-			queue->wd_count = 0;
+	queue_for_each(rt2x00dev, queue)
+		queue->wd_count = 0;
+
+	return true;
+}
+
+static bool rt2800_watchdog_dma_busy(struct rt2x00_dev *rt2x00dev)
+{
+	bool busy_rx, busy_tx;
+	u32 reg_cfg = rt2800_register_read(rt2x00dev, WPDMA_GLO_CFG);
+	u32 reg_int = rt2800_register_read(rt2x00dev, INT_SOURCE_CSR);
+
+	if (rt2x00_get_field32(reg_cfg, WPDMA_GLO_CFG_RX_DMA_BUSY) &&
+	    rt2x00_get_field32(reg_int, INT_SOURCE_CSR_RX_COHERENT))
+		rt2x00dev->rxdma_busy++;
+	else
+		rt2x00dev->rxdma_busy = 0;
 
+	if (rt2x00_get_field32(reg_cfg, WPDMA_GLO_CFG_TX_DMA_BUSY) &&
+	    rt2x00_get_field32(reg_int, INT_SOURCE_CSR_TX_COHERENT))
+		rt2x00dev->txdma_busy++;
+	else
+		rt2x00dev->txdma_busy = 0;
+
+	busy_rx = rt2x00dev->rxdma_busy > 30 ? true : false;
+	busy_tx = rt2x00dev->txdma_busy > 30 ? true : false;
+
+	if (!busy_rx && !busy_tx)
+		return false;
+
+	if (busy_rx)
+		rt2x00_warn(rt2x00dev, "Watchdog RX DMA busy detected\n");
+
+	if (busy_tx)
+		rt2x00_warn(rt2x00dev, "Watchdog TX DMA busy detected\n");
+
+	rt2x00dev->rxdma_busy = 0;
+	rt2x00dev->txdma_busy = 0;
+
+	return true;
+}
+
+void rt2800_watchdog(struct rt2x00_dev *rt2x00dev)
+{
+	bool reset = false;
+
+	if (test_bit(DEVICE_STATE_SCANNING, &rt2x00dev->flags))
+		return;
+
+	if (modparam_watchdog & RT2800_WATCHDOG_DMA_BUSY)
+		reset = rt2800_watchdog_dma_busy(rt2x00dev);
+
+	if (modparam_watchdog & RT2800_WATCHDOG_HANG)
+		reset = rt2800_watchdog_hung(rt2x00dev) || reset;
+
+	if (reset)
 		ieee80211_restart_hw(rt2x00dev->hw);
-	}
 }
 EXPORT_SYMBOL_GPL(rt2800_watchdog);
 
@@ -12013,6 +12065,9 @@ int rt2800_probe_hw(struct rt2x00_dev *rt2x00dev)
 		__set_bit(REQUIRE_TASKLET_CONTEXT, &rt2x00dev->cap_flags);
 	}
 
+	/* USB NICs don't support DMA watchdog as INT_SOURCE_CSR is invalid */
+	if (rt2x00_is_usb(rt2x00dev))
+		modparam_watchdog &= ~RT2800_WATCHDOG_DMA_BUSY;
 	if (modparam_watchdog) {
 		__set_bit(CAPABILITY_RESTART_HW, &rt2x00dev->cap_flags);
 		rt2x00dev->link.watchdog_interval = msecs_to_jiffies(100);

drivers/net/wireless/ralink/rt2x00/rt2x00.h

@@ -926,6 +926,9 @@ struct rt2x00_dev {
 	 */
 	u16 beacon_int;
 
+	/* Rx/Tx DMA busy watchdog counter */
+	u16 rxdma_busy, txdma_busy;
+
 	/**
 	 * Timestamp of last received beacon
 	 */