diff mbox series

[v3,2/2] USB: serial: cp210x: Proper RTS control when buffers fill

Message ID ECCF8E73-91F3-4080-BE17-1714BC8818FB@silabs.com (mailing list archive)
State New, archived
Headers show
Series [v3,1/2] USB: serial: cp210x: Enable usb generic throttle/unthrottle | expand

Commit Message

Phu Luu June 26, 2020, 4:24 a.m. UTC
From: Brant Merryman <brant.merryman@silabs.com>

CP210x hardware disables auto-RTS but leaves auto-CTS when
in hardware flow control mode and UART on cp210x hardware
is disabled. This allows data to flow out, but new data
will not come into the port. When re-opening the port, if
auto-CTS is enabled on the cp210x, then auto-RTS must be
re-enabled in the driver.

Signed-off-by: Phu Luu <phu.luu@silabs.com>
Signed-off-by: Brant Merryman <brant.merryman@silabs.com>
---
06/09/2020: Patch v3 2/2 Modified based on feedback from Johan Hovold <johan@kernel.org>
12/18/2019: Patch v2 Broken into two patches and modified based on feedback from Johan Hovold <johan@kernel.org>
12/09/2019: Initial submission of patch "Proper RTS control when buffers fill"

 drivers/usb/serial/cp210x.c | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

Comments

Johan Hovold June 26, 2020, 7:42 a.m. UTC | #1
On Fri, Jun 26, 2020 at 04:24:20AM +0000, Phu Luu wrote:
> From: Brant Merryman <brant.merryman@silabs.com>
> 
> CP210x hardware disables auto-RTS but leaves auto-CTS when
> in hardware flow control mode and UART on cp210x hardware
> is disabled. This allows data to flow out, but new data
> will not come into the port. When re-opening the port, if
> auto-CTS is enabled on the cp210x, then auto-RTS must be
> re-enabled in the driver.
> 
> Signed-off-by: Phu Luu <phu.luu@silabs.com>
> Signed-off-by: Brant Merryman <brant.merryman@silabs.com>

Please revisit these tags as well.

> ---
> 06/09/2020: Patch v3 2/2 Modified based on feedback from Johan Hovold <johan@kernel.org>
> 12/18/2019: Patch v2 Broken into two patches and modified based on feedback from Johan Hovold <johan@kernel.org>
> 12/09/2019: Initial submission of patch "Proper RTS control when buffers fill"
> 
>  drivers/usb/serial/cp210x.c | 18 ++++++++++++++++++
>  1 file changed, 18 insertions(+)
> 
> diff --git a/drivers/usb/serial/cp210x.c b/drivers/usb/serial/cp210x.c
> index bcceb4ad8be0..091441b1c5b9 100644
> --- a/drivers/usb/serial/cp210x.c
> +++ b/drivers/usb/serial/cp210x.c
> @@ -917,6 +917,7 @@ static void cp210x_get_termios_port(struct usb_serial_port *port,
>  	u32 baud;
>  	u16 bits;
>  	u32 ctl_hs;
> +	u32 flow_repl;
>  
>  	cp210x_read_u32_reg(port, CP210X_GET_BAUDRATE, &baud);
>  
> @@ -1017,6 +1018,23 @@ static void cp210x_get_termios_port(struct usb_serial_port *port,
>  	ctl_hs = le32_to_cpu(flow_ctl.ulControlHandshake);
>  	if (ctl_hs & CP210X_SERIAL_CTS_HANDSHAKE) {
>  		dev_dbg(dev, "%s - flow control = CRTSCTS\n", __func__);
> +		/*
> +		 * When the port is closed, the CP210x hardware disables
> +		 * auto-RTS and RTS is deasserted but it leaves auto-CTS when
> +		 * in hardware flow control mode. This prevents new data from
> +		 * being received by the port. When re-opening the port, if
> +		 * auto-CTS is enabled on the cp210x, then auto-RTS must be
> +		 * re-enabled in the driver.
> +		 */

I already asked this of Brant, but could you please be more specific
about which state the RTS-line end up in when disabling the UART (e.g.
0x00: statically inactive)?

The reason I ask is that after open() returns, the tty layer would raise
RTS, which should again allow data to flow in in contrast to what the
comment and changelog text claims.

We still want to enable auto-RTS of course so the patch is otherwise
correct.

> +		flow_repl = le32_to_cpu(flow_ctl.ulFlowReplace);
> +		flow_repl &= ~CP210X_SERIAL_RTS_MASK;
> +		flow_repl |= CP210X_SERIAL_RTS_SHIFT(CP210X_SERIAL_RTS_FLOW_CTL);
> +		flow_ctl.ulFlowReplace = cpu_to_le32(flow_repl);
> +		cp210x_write_reg_block(port,
> +				CP210X_SET_FLOW,
> +				&flow_ctl,
> +				sizeof(flow_ctl));
> +
>  		cflag |= CRTSCTS;
>  	} else {
>  		dev_dbg(dev, "%s - flow control = NONE\n", __func__);

Johan
Johan Hovold July 6, 2020, 10:18 a.m. UTC | #2
On Fri, Jun 26, 2020 at 09:42:06AM +0200, Johan Hovold wrote:
> On Fri, Jun 26, 2020 at 04:24:20AM +0000, Phu Luu wrote:
> > From: Brant Merryman <brant.merryman@silabs.com>
> > 
> > CP210x hardware disables auto-RTS but leaves auto-CTS when
> > in hardware flow control mode and UART on cp210x hardware
> > is disabled. This allows data to flow out, but new data
> > will not come into the port. When re-opening the port, if
> > auto-CTS is enabled on the cp210x, then auto-RTS must be
> > re-enabled in the driver.
> > 
> > Signed-off-by: Phu Luu <phu.luu@silabs.com>
> > Signed-off-by: Brant Merryman <brant.merryman@silabs.com>
> 
> Please revisit these tags as well.
> 
> > ---
> > 06/09/2020: Patch v3 2/2 Modified based on feedback from Johan Hovold <johan@kernel.org>
> > 12/18/2019: Patch v2 Broken into two patches and modified based on feedback from Johan Hovold <johan@kernel.org>
> > 12/09/2019: Initial submission of patch "Proper RTS control when buffers fill"
> > 
> >  drivers/usb/serial/cp210x.c | 18 ++++++++++++++++++
> >  1 file changed, 18 insertions(+)
> > 
> > diff --git a/drivers/usb/serial/cp210x.c b/drivers/usb/serial/cp210x.c
> > index bcceb4ad8be0..091441b1c5b9 100644
> > --- a/drivers/usb/serial/cp210x.c
> > +++ b/drivers/usb/serial/cp210x.c
> > @@ -917,6 +917,7 @@ static void cp210x_get_termios_port(struct usb_serial_port *port,
> >  	u32 baud;
> >  	u16 bits;
> >  	u32 ctl_hs;
> > +	u32 flow_repl;
> >  
> >  	cp210x_read_u32_reg(port, CP210X_GET_BAUDRATE, &baud);
> >  
> > @@ -1017,6 +1018,23 @@ static void cp210x_get_termios_port(struct usb_serial_port *port,
> >  	ctl_hs = le32_to_cpu(flow_ctl.ulControlHandshake);
> >  	if (ctl_hs & CP210X_SERIAL_CTS_HANDSHAKE) {
> >  		dev_dbg(dev, "%s - flow control = CRTSCTS\n", __func__);
> > +		/*
> > +		 * When the port is closed, the CP210x hardware disables
> > +		 * auto-RTS and RTS is deasserted but it leaves auto-CTS when
> > +		 * in hardware flow control mode. This prevents new data from
> > +		 * being received by the port. When re-opening the port, if
> > +		 * auto-CTS is enabled on the cp210x, then auto-RTS must be
> > +		 * re-enabled in the driver.
> > +		 */
> 
> I already asked this of Brant, but could you please be more specific
> about which state the RTS-line end up in when disabling the UART (e.g.
> 0x00: statically inactive)?
> 
> The reason I ask is that after open() returns, the tty layer would raise
> RTS, which should again allow data to flow in in contrast to what the
> comment and changelog text claims.

I had another reason to look at this driver so I could determine that
the patch description and comment were indeed incorrect; the tty layer
asserts RTS when reopening the port just fine and data flows in.

> We still want to enable auto-RTS of course so the patch is otherwise
> correct.

I've fixed up the commit message and comment, reordered the SOB tags
and added your Co-developed-by tag before applying both of these now:

	https://git.kernel.org/pub/scm/linux/kernel/git/johan/usb-serial.git/log/?h=usb-next

Johan
Johan Hovold Aug. 17, 2020, 9:55 a.m. UTC | #3
On Mon, Aug 17, 2020 at 09:24:05AM +0700, Phú Lưu An wrote:
> Hi Johan,
> 
> My customer asked me when this patch is applied to the community. So, could
> you please tell me when it appears in a various distributions?

You'll have to ask the distributions about that.

The patch in question is in Linus's tree since about two weeks and
should be picked up and backported to the stable trees soon:

	c7614ff9b73a ("USB: serial: cp210x: re-enable auto-RTS on open")

Johan
diff mbox series

Patch

diff --git a/drivers/usb/serial/cp210x.c b/drivers/usb/serial/cp210x.c
index bcceb4ad8be0..091441b1c5b9 100644
--- a/drivers/usb/serial/cp210x.c
+++ b/drivers/usb/serial/cp210x.c
@@ -917,6 +917,7 @@  static void cp210x_get_termios_port(struct usb_serial_port *port,
 	u32 baud;
 	u16 bits;
 	u32 ctl_hs;
+	u32 flow_repl;
 
 	cp210x_read_u32_reg(port, CP210X_GET_BAUDRATE, &baud);
 
@@ -1017,6 +1018,23 @@  static void cp210x_get_termios_port(struct usb_serial_port *port,
 	ctl_hs = le32_to_cpu(flow_ctl.ulControlHandshake);
 	if (ctl_hs & CP210X_SERIAL_CTS_HANDSHAKE) {
 		dev_dbg(dev, "%s - flow control = CRTSCTS\n", __func__);
+		/*
+		 * When the port is closed, the CP210x hardware disables
+		 * auto-RTS and RTS is deasserted but it leaves auto-CTS when
+		 * in hardware flow control mode. This prevents new data from
+		 * being received by the port. When re-opening the port, if
+		 * auto-CTS is enabled on the cp210x, then auto-RTS must be
+		 * re-enabled in the driver.
+		 */
+		flow_repl = le32_to_cpu(flow_ctl.ulFlowReplace);
+		flow_repl &= ~CP210X_SERIAL_RTS_MASK;
+		flow_repl |= CP210X_SERIAL_RTS_SHIFT(CP210X_SERIAL_RTS_FLOW_CTL);
+		flow_ctl.ulFlowReplace = cpu_to_le32(flow_repl);
+		cp210x_write_reg_block(port,
+				CP210X_SET_FLOW,
+				&flow_ctl,
+				sizeof(flow_ctl));
+
 		cflag |= CRTSCTS;
 	} else {
 		dev_dbg(dev, "%s - flow control = NONE\n", __func__);