cavium: 8250 serial driver changes
As mentioned in '[PATCH 1/2] serial: Allow for replaceable I/O

functions in 8250 driver.', we are in the process of preparing kernel

support for the Cavium Networks OCTEON processor for inclusion in the

main-line kernel sources.
The OCTEON's UART differs from the existing uart_configs, so we add a

new uart_config and check for it in several places.
This patch depends on the aforementioned 'Allow for replaceable I/O

functions in 8250 driver' patch.
Since this patch is part of the Cavium OCTEON processor port, we don't

expect that it would be committed until the rest of the port is

accepted.  However we would like feedback so that it might be

adjusted if necessary.
Signed-off-by: Tomaso Paoletti <tpaoletti@caviumnetworks.com>

Signed-off-by: David Daney <ddaney@caviumnetworks.com>

---

 drivers/serial/8250.c       |   30 +++++++++++++++++++++++++++---

 include/linux/serial_core.h |    3 ++-

 include/linux/serial_reg.h  |    6 ++++++

 3 files changed, 35 insertions(+), 4 deletions(-)
diff --git a/drivers/serial/8250.c b/drivers/serial/8250.c

index 02771d6..2ef79e9 100644

--- a/drivers/serial/8250.c

+++ b/drivers/serial/8250.c

@@ -264,6 +264,13 @@ static const struct serial8250_config uart_config[] = {

 		.fcr		= UART_FCR_ENABLE_FIFO | UART_FCR_R_TRIG_10,

 		.flags		= UART_CAP_FIFO,

 	},

+	[PORT_OCTEON] = {

+		.name		= "OCTEON",

+		.fifo_size	= 64,

+		.tx_loadsz	= 64,

+		.fcr		= UART_FCR_ENABLE_FIFO | UART_FCR_R_TRIG_10,

+		.flags		= UART_CAP_FIFO,

+	},

 };
 #if defined (CONFIG_SERIAL_8250_AU1X00)

@@ -902,7 +909,14 @@ static void autoconfig_16550a(struct uart_8250_port *up)

 	unsigned char status1, status2;

 	unsigned int iersave;
-	up->port.type = PORT_16550A;

+#ifdef CONFIG_CPU_CAVIUM_OCTEON

+	/* UPF_FIXED_PORT indicates an internal UART.  */

+	if (up->port.flags & UPF_FIXED_PORT)

+		up->port.type = PORT_OCTEON;

+	else

+#endif

+		up->port.type = PORT_16550A;

+

 	up->capabilities |= UART...
[ message continues ]

Not nice. Please keep CPU specific ifdefs out of the 8250 core code. Can

you not set a port flag for UPF_BROKEN_OCTEON or similar to clean that up

and also make the other tests that need things doing (eg the always

calling IRQ code use port flags of a more generic nature ?)

--

This looks somewhat wrong, IMHO a device driver should not assume that

a CONFIG_CPU_* symbol is exclusive. You could have (maybe not now, but

in the future) a kernel that supports running on an Octeon as well

as some other Mips64 processor, and have UPF_FIXED_PORT uart on some

other machine, which will make the kernel think it is a PORT_OCTEON.
	Arnd <><

--