[PATCH ath-current] Revert "wifi: ath10k: avoid unnecessary wait for service ready message"

[Baochen Qiang]

This reverts commit 51a73f1b2e56b0324b4a3bb8cebc4221b5be4c7a.

Although this commit benefits QCA6174, it breaks QCA988x and
QCA9984 [1][2]. Since it is not likely to root cause/fix this
issue in a short time, revert it to get those chips back.

Compile tested only.

Fixes: 51a73f1b2e56 ("wifi: ath10k: avoid unnecessary wait for service ready message")
Link: https://lore.kernel.org/ath10k/6d41bc00602c33ffbf68781f563ff2e6c6915a3e.camel@gmail.com # [1]
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=220671 # [2]
Signed-off-by: Baochen Qiang <baochen.qiang at oss.qualcomm.com>
---
 drivers/net/wireless/ath/ath10k/wmi.c | 39 ++++++++++++++++++-----------------
 1 file changed, 20 insertions(+), 19 deletions(-)

diff --git a/drivers/net/wireless/ath/ath10k/wmi.c b/drivers/net/wireless/ath/ath10k/wmi.c
index b3b00d324075bd1b82c76f2d93cf4e2a79ead769..b4aad6604d6d9d95b918d0a077f5d1511bc249ad 100644
--- a/drivers/net/wireless/ath/ath10k/wmi.c
+++ b/drivers/net/wireless/ath/ath10k/wmi.c
@@ -1764,32 +1764,33 @@ void ath10k_wmi_put_wmi_channel(struct ath10k *ar, struct wmi_channel *ch,
 
 int ath10k_wmi_wait_for_service_ready(struct ath10k *ar)
 {
-	unsigned long timeout = jiffies + WMI_SERVICE_READY_TIMEOUT_HZ;
 	unsigned long time_left, i;
 
-	/* Sometimes the PCI HIF doesn't receive interrupt
-	 * for the service ready message even if the buffer
-	 * was completed. PCIe sniffer shows that it's
-	 * because the corresponding CE ring doesn't fires
-	 * it. Workaround here by polling CE rings. Since
-	 * the message could arrive at any time, continue
-	 * polling until timeout.
-	 */
-	do {
+	time_left = wait_for_completion_timeout(&ar->wmi.service_ready,
+						WMI_SERVICE_READY_TIMEOUT_HZ);
+	if (!time_left) {
+		/* Sometimes the PCI HIF doesn't receive interrupt
+		 * for the service ready message even if the buffer
+		 * was completed. PCIe sniffer shows that it's
+		 * because the corresponding CE ring doesn't fires
+		 * it. Workaround here by polling CE rings once.
+		 */
+		ath10k_warn(ar, "failed to receive service ready completion, polling..\n");
+
 		for (i = 0; i < CE_COUNT; i++)
 			ath10k_hif_send_complete_check(ar, i, 1);
 
-		/* The 100 ms granularity is a tradeoff considering scheduler
-		 * overhead and response latency
-		 */
 		time_left = wait_for_completion_timeout(&ar->wmi.service_ready,
-							msecs_to_jiffies(100));
-		if (time_left)
-			return 0;
-	} while (time_before(jiffies, timeout));
+							WMI_SERVICE_READY_TIMEOUT_HZ);
+		if (!time_left) {
+			ath10k_warn(ar, "polling timed out\n");
+			return -ETIMEDOUT;
+		}
+
+		ath10k_warn(ar, "service ready completion received, continuing normally\n");
+	}
 
-	ath10k_warn(ar, "failed to receive service ready completion\n");
-	return -ETIMEDOUT;
+	return 0;
 }
 
 int ath10k_wmi_wait_for_unified_ready(struct ath10k *ar)

---
base-commit: d5d9da878287cb30199defc6941b695f33f9236b
change-id: 20251024-ath10k-revert-polling-first-change-2add89f0fd15

Best regards,
-- 
Baochen Qiang <baochen.qiang at oss.qualcomm.com>