Skip to contentSkip to conversation

powerpc: experimental of serial port driver

RFC In review Arnd Bergmann 4 participants 10 messages Dec 7, 2006 – Dec 9, 2006

lore
Share & tools
lore
#0 lore raw

RFC 1/1 powerpc: experimental of serial port driver

This can be used for serial ports that are connected to an OF platform bus. This is independent of whether they also get detected as legacy_serial devices during early boot, because the serial core makes sure it handles multiple registrations fine.

It should be generic enough to work for more than just 8250 style devices, but those are the only ones I tested so far.

drivers/serial/Makefile +1 -0
Index: linux-2.6/drivers/serial/Makefile
===================================================================
@@ -56,3 +56,4 @@ obj-$(CONFIG_SERIAL_SGI_IOC4) += ioc4_se
 obj-$(CONFIG_SERIAL_SGI_IOC3) += ioc3_serial.o
 obj-$(CONFIG_SERIAL_ATMEL) += atmel_serial.o
 obj-$(CONFIG_SERIAL_NETX) += netx-serial.o
+obj-$(CONFIG_SERIAL_OF_PLATFORM) += of_serial.o
drivers/serial/of_serial.c new file +141 -0

141 lines not on this page · raw message

drivers/serial/Kconfig +10 -0
Index: linux-2.6/drivers/serial/Kconfig
===================================================================
@@ -965,4 +965,14 @@ config SERIAL_NETX_CONSOLE
 	  If you have enabled the serial port on the Motorola IMX
 	  CPU you can make it the console by answering Y to this option.
 
+config SERIAL_OF_PLATFORM
+	tristate "Serial port on Open Firmware platform bus"
+	depends on PPC_OF
+	depends on SERIAL_8250
+	help
+	  If you have a PowerPC based system that has serial ports
+	  on a platform specific bus, you should enable this option.
+	  Currently, only 8250 compatible ports are supported, but
+	  others can easily be added.
+
 endmenu
#1 lore raw
Quoted 11 lines · Arnd Bergmann

This can be used for serial ports that are connected to an OF platform bus. This is independent of whether they also get detected as legacy_serial devices during early boot, because the serial core makes sure it handles multiple registrations fine.

It should be generic enough to work for more than just 8250 style devices, but those are the only ones I tested so far.

I'd rather call it 8250_of.c and have it be specific to 8250's.

Ben.

Quoted 178 lines

Index: linux-2.6/drivers/serial/Makefile
===================================================================

drivers/serial/Makefile +1 -0
@@ -56,3 +56,4 @@ obj-$(CONFIG_SERIAL_SGI_IOC4) += ioc4_se
 obj-$(CONFIG_SERIAL_SGI_IOC3) += ioc3_serial.o
 obj-$(CONFIG_SERIAL_ATMEL) += atmel_serial.o
 obj-$(CONFIG_SERIAL_NETX) += netx-serial.o
+obj-$(CONFIG_SERIAL_OF_PLATFORM) += of_serial.o

Index: linux-2.6/drivers/serial/of_serial.c
===================================================================

drivers/serial/of_serial.c new file +141 -0
@@ -0,0 +1,141 @@
+/*
+ *  Serial Port driver for Open Firmware platform devices
+ *
+ *    Copyright (C) 2006 Arnd Bergmann <[email protected]>, IBM Corp.
+ *
+ *  This program is free software; you can redistribute it and/or
+ *  modify it under the terms of the GNU General Public License
+ *  as published by the Free Software Foundation; either version
+ *  2 of the License, or (at your option) any later version.
+ *
+ */
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/serial_core.h>
+#include <linux/serial_8250.h>
+
+#include <asm/of_platform.h>
+#include <asm/prom.h>
+
+/*
+ * Fill a struct uart_port for a given device node
+ */
+static int __devinit of_platform_serial_setup(struct of_device *ofdev,
+					int type, struct uart_port *port)
+{
+	struct resource resource;
+	struct device_node *np = ofdev->node;
+	const unsigned int *clk, *spd;
+	int ret;
+
+	memset(port, 0, sizeof *port);
+	spd = get_property(np, "current-speed", NULL);
+	clk = get_property(np, "clock-frequency", NULL);
+	if (!clk) {
+		dev_warn(&ofdev->dev, "no clock-frequency property set\n");
+		return -ENODEV;
+	}
+
+	ret = of_address_to_resource(np, 0, &resource);
+	if (ret) {
+		dev_warn(&ofdev->dev, "invalid address\n");
+		return ret;
+	}
+
+	spin_lock_init(&port->lock);
+	port->mapbase = resource.start;
+	port->irq = irq_of_parse_and_map(np, 0);
+	port->iotype = UPIO_MEM;
+	port->type = type;
+	port->uartclk = *clk;
+	port->flags = UPF_SHARE_IRQ | UPF_BOOT_AUTOCONF | UPF_IOREMAP;
+	port->dev = &ofdev->dev;
+/*	if (spd) // FIXME: how to set?
+		port->speed = spd;
+*/
+	return 0;
+}
+
+/*
+ * Try to register a serial port
+ */
+static int __devinit of_platform_serial_probe(struct of_device *ofdev,
+						const struct of_device_id *id)
+{
+	struct uart_port port;
+	int port_type;
+	int ret;
+
+	port_type = (unsigned long)id->data;
+	ret = of_platform_serial_setup(ofdev, port_type, &port);
+	if (ret)
+		goto out;
+
+	switch (port_type) {
+	case PORT_UNKNOWN:
+		dev_info(&ofdev->dev, "Unknown serial port found, "
+			"attempting to use 8250 driver\n");
+		/* fallthrough */
+	case PORT_8250 ... PORT_MAX_8250:
+		ret = serial8250_register_port(&port);
+		break;
+	default:
+		/* need to add code for these */
+		ret = -ENODEV;
+		break;
+	}
+	if (ret < 0)
+		goto out;
+
+	ofdev->dev.driver_data = (void *)(unsigned long)ret;
+	return 0;
+out:
+	irq_dispose_mapping(port.irq);
+	return ret;
+}
+
+/*
+ * Release a line
+ */
+static int of_platform_serial_remove(struct of_device *ofdev)
+{
+	int line = (unsigned long)ofdev->dev.driver_data;
+	serial8250_unregister_port(line);
+	return 0;
+}
+
+/*
+ * A few common types, add more as needed.
+ */
+static struct of_device_id __devinitdata of_platform_serial_table[] = {
+	{ .type = "serial", .compatible = "ns8250",   .data = (void *)PORT_8250, },
+	{ .type = "serial", .compatible = "ns16450",  .data = (void *)PORT_16450, },
+	{ .type = "serial", .compatible = "ns16550",  .data = (void *)PORT_16550, },
+	{ .type = "serial", .compatible = "ns16750",  .data = (void *)PORT_16750, },
+	{ .type = "serial",			      .data = (void *)PORT_UNKNOWN, },
+	{ /* end of list */ },
+};
+
+static struct of_platform_driver __devinitdata of_platform_serial_driver = {
+	.owner = THIS_MODULE,
+	.name = "of_serial",
+	.probe = of_platform_serial_probe,
+	.remove = of_platform_serial_remove,
+	.match_table = of_platform_serial_table,
+};
+
+static int __init of_platform_serial_init(void)
+{
+	return of_register_platform_driver(&of_platform_serial_driver);
+}
+module_init(of_platform_serial_init);
+
+static void __exit of_platform_serial_exit(void)
+{
+	return of_unregister_platform_driver(&of_platform_serial_driver);
+};
+module_exit(of_platform_serial_exit);
+
+MODULE_AUTHOR("Arnd Bergmann <[email protected]>");
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Serial Port driver for Open Firmware platform devices");

Index: linux-2.6/drivers/serial/Kconfig
===================================================================

drivers/serial/Kconfig +10 -0
@@ -965,4 +965,14 @@ config SERIAL_NETX_CONSOLE
 	  If you have enabled the serial port on the Motorola IMX
 	  CPU you can make it the console by answering Y to this option.
 
+config SERIAL_OF_PLATFORM
+	tristate "Serial port on Open Firmware platform bus"
+	depends on PPC_OF
+	depends on SERIAL_8250
+	help
+	  If you have a PowerPC based system that has serial ports
+	  on a platform specific bus, you should enable this option.
+	  Currently, only 8250 compatible ports are supported, but
+	  others can easily be added.
+
 endmenu

_______________________________________________
Linuxppc-dev mailing list
[email protected]
https://ozlabs.org/mailman/listinfo/linuxppc-dev

#2 lore raw
Quoted 1 line · Benjamin Herrenschmidt

I'd rather call it 8250_of.c and have it be specific to 8250's.

Didn't we already do that once?

Signature · 1 line
dwmw2
#3 lore raw
Quoted 5 lines · David Woodhouse
Quoted 1 line · Benjamin Herrenschmidt

I'd rather call it 8250_of.c and have it be specific to 8250's.

Didn't we already do that once?

You submitted a patch once to create a file with the same name, but different intend, see
http://ozlabs.org/pipermail/linuxppc-dev/2005-August/019464.html

I can't tell if it was ever merged, but it doesn't look like it. My driver is only meaningful for platforms that use the new style of_platform_device autoprobing. It will make the serial ports show up correctly in sysfs, even if they were not detected by the legacy_serial code we use for 8250 style console ports. This should be the same thing we do for 8250 PCI ports when the 8250_pci driver.

	Arnd <><
#9 lore raw
In reply to David Woodhouse
Quoted 4 lines · David Woodhouse
Quoted 1 line · Benjamin Herrenschmidt

I'd rather call it 8250_of.c and have it be specific to 8250's.

Didn't we already do that once?

Not really, right now, we use a platform driver (though it should work, again, I'm not completely sure why we need that at all).

legacy_serial.c should be detecting all 8250 type ports at boot, init them, setup a udbg console for them, and later on, register them as platform devices with the 8250 driver. The problem right now is that the code in there will only do that for known busses, which doesn't include opb. So we should just fix it for opb and things should work.

Ben.

#4 lore raw
In reply to Arnd Bergmann
Quoted 8 lines

I can't tell if it was ever merged, but it doesn't look like it. My driver is only meaningful for platforms that use the new style of_platform_device autoprobing. It will make the serial ports show up correctly in sysfs, even if they were not detected by the legacy_serial code we use for 8250 style console ports. This should be the same thing we do for 8250 PCI ports when the 8250_pci driver.

We sill need to fixup legacy_serial to detect them on opb right ?

Ben.

#5 lore raw
Quoted 1 line · Benjamin Herrenschmidt

We sill need to fixup legacy_serial to detect them on opb right ?

Yes, I have another patch for that. Currently, the port registers as compatible="ns16750", which is different from all the others in legacy_serial. Do you think it should be compatible="nc16750\0ns16550" instead, so we can use the same as the others? It actually _is_ compatible to 16550 anyway.

	Arnd <><
#6 lore raw
Quoted 8 lines · Arnd Bergmann
Quoted 1 line · Benjamin Herrenschmidt

We sill need to fixup legacy_serial to detect them on opb right ?

Yes, I have another patch for that. Currently, the port registers as compatible="ns16750", which is different from all the others in legacy_serial. Do you think it should be compatible="nc16750\0ns16550" instead, so we can use the same as the others? It actually _is_ compatible to 16550 anyway.

Both strings then.

Ben.

#7 lore raw

Hello.

Quoted 8 lines · Benjamin Herrenschmidt
Quoted 6 lines
Quoted 1 line

We sill need to fixup legacy_serial to detect them on opb right ?

Yes, I have another patch for that. Currently, the port registers
as compatible="ns16750", which is different from all the others
in legacy_serial. Do you think it should be compatible="nc16750\0ns16550" instead, so we can use the same as the others? It actually _is_ compatible to 16550 anyway.

Both strings then.

Then it should be something like "nc16750\0ns16550\0ns16450\08250".

Quoted 1 line

Ben.

WBR, Sergei

#8 lore raw
Quoted 1 line · Sergei Shtylyov

Then it should be something like "nc16750\0ns16550\0ns16450\08250".

Those aren't particularly interesting distinctions -- they're dead easy to probe for. The really interesting one is the type of SuperIO chip it's compatible with -- which specifies how you get baud rates above 115200. For some chips (SMSC iirc) you have a 'magic multiplier'; by setting the divisor to 0x8001 or 0x8002 you get 230400/460800 baud resp. For NatSemi type chips you have baud rate divisors in an alternate bank, and you get a higher baud_base if you use those but if you use the registers in bank zero it reverts to 115200, etc.

Signature · 1 line
dwmw2