[PATCH 1/1] USB: serial: pl2303: TA & TB alternate divider

Subsystems: the rest, usb serial subsystem, usb subsystem

STALE1971d

3 messages, 3 authors, 2021-03-16 · open the first message on its own page

[PATCH 1/1] USB: serial: pl2303: TA & TB alternate divider

From: <hidden>
Date: 2021-03-16 01:50:17

From: Michael Katzmann <redacted>

Use an alternate clock divider algorithm and bit ordering for the TA and
TB versions of the pl2303. It was discovered that these variants do not
produce the correct baud rates with the existing scheme.

see https://marc.info/?t=161392209800002&r=1&w=2

Signed-off-by: Michael G. Katzmann <redacted>
---
 drivers/usb/serial/pl2303.c | 45 +++++++++++++++++++++++++++++++++++++
 1 file changed, 45 insertions(+)
diff --git a/drivers/usb/serial/pl2303.c b/drivers/usb/serial/pl2303.c
index 7208966891d0..bf5828579918 100644
--- a/drivers/usb/serial/pl2303.c
+++ b/drivers/usb/serial/pl2303.c
@@ -188,6 +188,7 @@ struct pl2303_type_data {
 	unsigned long quirks;
 	unsigned int no_autoxonxoff:1;
 	unsigned int no_divisors:1;
+	unsigned int alt_divisors:1;
 };
 
 struct pl2303_serial_private {
@@ -217,10 +218,12 @@ static const struct pl2303_type_data pl2303_type_data[TYPE_COUNT] = {
 	[TYPE_TA] = {
 		.name			= "TA",
 		.max_baud_rate		= 6000000,
+		.alt_divisors		= true,
 	},
 	[TYPE_TB] = {
 		.name			= "TB",
 		.max_baud_rate		= 12000000,
+		.alt_divisors		= true,
 	},
 	[TYPE_HXD] = {
 		.name			= "HXD",
@@ -618,6 +621,46 @@ static speed_t pl2303_encode_baud_rate_divisor(unsigned char buf[4],
 	return baud;
 }
 
+static speed_t pl2303_encode_baud_rate_divisor_alt(unsigned char buf[4],
+                                                                speed_t baud)
+{
+        unsigned int baseline, mantissa, exponent;
+
+        /*
+         * Apparently, for the TA version the formula is:
+         *   baudrate = 12M * 32 / (mantissa * 2^exponent)
+         * where
+         *   mantissa = buf[10:0]
+         *   exponent = buf[15:13 16]
+         */
+        baseline = 12000000 * 32;
+        mantissa = baseline / baud;
+        if (mantissa == 0)
+                mantissa = 1;   /* Avoid dividing by zero if baud > 32*12M. */
+        exponent = 0;
+        while (mantissa >= 2048) {
+                if (exponent < 15) {
+                        mantissa >>= 1; /* divide by 2 */
+                        exponent++;
+                } else {
+                        /* Exponent is maxed. Trim mantissa and leave. */
+                        mantissa = 2047;
+                        break;
+                }
+        }
+
+        buf[3] = 0x80;
+        buf[2] = exponent & 0x01;
+        buf[1] = (exponent & ~0x01) << 4 | mantissa >> 8;
+        buf[0] = mantissa & 0xff;
+
+        /* Calculate and return the exact baud rate. */
+        baud = (baseline / mantissa) >> exponent;
+
+        return baud;
+}
+
+
 static void pl2303_encode_baud_rate(struct tty_struct *tty,
 					struct usb_serial_port *port,
 					u8 buf[4])
@@ -645,6 +688,8 @@ static void pl2303_encode_baud_rate(struct tty_struct *tty,
 
 	if (baud == baud_sup)
 		baud = pl2303_encode_baud_rate_direct(buf, baud);
+	else if (spriv->type->alt_divisors) 
+                baud = pl2303_encode_baud_rate_divisor_alt(buf, baud);
 	else
 		baud = pl2303_encode_baud_rate_divisor(buf, baud);
 
-- 
2.30.2

Re: [PATCH 1/1] USB: serial: pl2303: TA & TB alternate divider

From: Greg KH <hidden>
Date: 2021-03-16 07:20:42

On Mon, Mar 15, 2021 at 09:42:05PM -0400, michaelk@IEEE.org wrote:
From: Michael Katzmann <redacted>

Use an alternate clock divider algorithm and bit ordering for the TA and
TB versions of the pl2303. It was discovered that these variants do not
produce the correct baud rates with the existing scheme.

see https://marc.info/?t=161392209800002&r=1&w=2
Can you point to a lore.kernel.org thread instead?  We (kernel developer
community) have control over lore, but not over marc.info or any other
archive, so we never know if it will be around or not over time :)

thanks,

greg k-h

Re: [PATCH 1/1] USB: serial: pl2303: TA & TB alternate divider

From: Johan Hovold <johan@kernel.org>
Date: 2021-03-16 09:19:46

On Tue, Mar 16, 2021 at 08:19:57AM +0100, Greg Kroah-Hartman wrote:
On Mon, Mar 15, 2021 at 09:42:05PM -0400, michaelk@IEEE.org wrote:
quoted
From: Michael Katzmann <redacted>

Use an alternate clock divider algorithm and bit ordering for the TA and
TB versions of the pl2303. It was discovered that these variants do not
produce the correct baud rates with the existing scheme.

see https://marc.info/?t=161392209800002&r=1&w=2
Can you point to a lore.kernel.org thread instead?  We (kernel developer
community) have control over lore, but not over marc.info or any other
archive, so we never know if it will be around or not over time :)
I was gonna say that I could fix that up when applying, but turns out
there are some formatting issues with the patch.

Michael, it looks like you've used tabs instead of spaces for
indentation in some places. Try running scripts/checkpatch.pl on the
patch before submitting (which would have caught this).

Care to send a v2?

And the lore link would be:

	https://lore.kernel.org/r/3aee5708-7961-f464-8c5f-6685d96920d6@IEEE.org

Johan
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help