diff mbox series

[v2,1/8] serial: qcom-geni: fix fifo polling timeout

Message ID 20240906131336.23625-2-johan+linaro@kernel.org (mailing list archive)
State Not Applicable
Headers show
Series serial: qcom-geni: fix console corruption | expand

Commit Message

Johan Hovold Sept. 6, 2024, 1:13 p.m. UTC
The qcom_geni_serial_poll_bit() can be used to wait for events like
command completion and is supposed to wait for the time it takes to
clear a full fifo before timing out.

As noted by Doug, the current implementation does not account for start,
stop and parity bits when determining the timeout. The helper also does
not currently account for the shift register and the two-word
intermediate transfer register.

A too short timeout can specifically lead to lost characters when
waiting for a transfer to complete as the transfer is cancelled on
timeout.

Instead of determining the poll timeout on every call, store the fifo
timeout when updating it in set_termios() and make sure to take the
shift and intermediate registers into account. Note that serial core has
already added a 20 ms margin to the fifo timeout.

Also note that the current uart_fifo_timeout() interface does
unnecessary calculations on every call and did not exist in earlier
kernels so only store its result once. This facilitates backports too as
earlier kernels can derive the timeout from uport->timeout, which has
since been removed.

Fixes: c4f528795d1a ("tty: serial: msm_geni_serial: Add serial driver support for GENI based QUP")
Cc: stable@vger.kernel.org	# 4.17
Reported-by: Douglas Anderson <dianders@chromium.org>
Tested-by: Nícolas F. R. A. Prado <nfraprado@collabora.com>
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
---
 drivers/tty/serial/qcom_geni_serial.c | 31 +++++++++++++++------------
 1 file changed, 17 insertions(+), 14 deletions(-)

Comments

Doug Anderson Sept. 11, 2024, 8:19 p.m. UTC | #1
Hi,

On Fri, Sep 6, 2024 at 6:15 AM Johan Hovold <johan+linaro@kernel.org> wrote:
>
> The qcom_geni_serial_poll_bit() can be used to wait for events like
> command completion and is supposed to wait for the time it takes to
> clear a full fifo before timing out.
>
> As noted by Doug, the current implementation does not account for start,
> stop and parity bits when determining the timeout. The helper also does
> not currently account for the shift register and the two-word
> intermediate transfer register.
>
> A too short timeout can specifically lead to lost characters when
> waiting for a transfer to complete as the transfer is cancelled on
> timeout.
>
> Instead of determining the poll timeout on every call, store the fifo
> timeout when updating it in set_termios() and make sure to take the
> shift and intermediate registers into account. Note that serial core has
> already added a 20 ms margin to the fifo timeout.
>
> Also note that the current uart_fifo_timeout() interface does
> unnecessary calculations on every call and did not exist in earlier
> kernels so only store its result once. This facilitates backports too as
> earlier kernels can derive the timeout from uport->timeout, which has
> since been removed.
>
> Fixes: c4f528795d1a ("tty: serial: msm_geni_serial: Add serial driver support for GENI based QUP")
> Cc: stable@vger.kernel.org      # 4.17
> Reported-by: Douglas Anderson <dianders@chromium.org>
> Tested-by: Nícolas F. R. A. Prado <nfraprado@collabora.com>
> Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
> ---
>  drivers/tty/serial/qcom_geni_serial.c | 31 +++++++++++++++------------
>  1 file changed, 17 insertions(+), 14 deletions(-)

Reviewed-by: Douglas Anderson <dianders@chromium.org>
diff mbox series

Patch

diff --git a/drivers/tty/serial/qcom_geni_serial.c b/drivers/tty/serial/qcom_geni_serial.c
index 69a632fefc41..309c0bddf26a 100644
--- a/drivers/tty/serial/qcom_geni_serial.c
+++ b/drivers/tty/serial/qcom_geni_serial.c
@@ -124,7 +124,7 @@  struct qcom_geni_serial_port {
 	dma_addr_t tx_dma_addr;
 	dma_addr_t rx_dma_addr;
 	bool setup;
-	unsigned int baud;
+	unsigned long poll_timeout_us;
 	unsigned long clk_rate;
 	void *rx_buf;
 	u32 loopback;
@@ -270,22 +270,13 @@  static bool qcom_geni_serial_poll_bit(struct uart_port *uport,
 {
 	u32 reg;
 	struct qcom_geni_serial_port *port;
-	unsigned int baud;
-	unsigned int fifo_bits;
 	unsigned long timeout_us = 20000;
 	struct qcom_geni_private_data *private_data = uport->private_data;
 
 	if (private_data->drv) {
 		port = to_dev_port(uport);
-		baud = port->baud;
-		if (!baud)
-			baud = 115200;
-		fifo_bits = port->tx_fifo_depth * port->tx_fifo_width;
-		/*
-		 * Total polling iterations based on FIFO worth of bytes to be
-		 * sent at current baud. Add a little fluff to the wait.
-		 */
-		timeout_us = ((fifo_bits * USEC_PER_SEC) / baud) + 500;
+		if (port->poll_timeout_us)
+			timeout_us = port->poll_timeout_us;
 	}
 
 	/*
@@ -1244,11 +1235,11 @@  static void qcom_geni_serial_set_termios(struct uart_port *uport,
 	unsigned long clk_rate;
 	u32 ver, sampling_rate;
 	unsigned int avg_bw_core;
+	unsigned long timeout;
 
 	qcom_geni_serial_stop_rx(uport);
 	/* baud rate */
 	baud = uart_get_baud_rate(uport, termios, old, 300, 4000000);
-	port->baud = baud;
 
 	sampling_rate = UART_OVERSAMPLING;
 	/* Sampling rate is halved for IP versions >= 2.5 */
@@ -1326,9 +1317,21 @@  static void qcom_geni_serial_set_termios(struct uart_port *uport,
 	else
 		tx_trans_cfg |= UART_CTS_MASK;
 
-	if (baud)
+	if (baud) {
 		uart_update_timeout(uport, termios->c_cflag, baud);
 
+		/*
+		 * Make sure that qcom_geni_serial_poll_bitfield() waits for
+		 * the FIFO, two-word intermediate transfer register and shift
+		 * register to clear.
+		 *
+		 * Note that uart_fifo_timeout() also adds a 20 ms margin.
+		 */
+		timeout = jiffies_to_usecs(uart_fifo_timeout(uport));
+		timeout += 3 * timeout / port->tx_fifo_depth;
+		WRITE_ONCE(port->poll_timeout_us, timeout);
+	}
+
 	if (!uart_console(uport))
 		writel(port->loopback,
 				uport->membase + SE_UART_LOOPBACK_CFG);