1. Patchew
  2. linux
  3. View series

[PATCH] tty: serial: Work around warning backtrace in serial8250_set_defaults

Guenter Roeck posted 1 patch 1 year ago 
drivers/tty/serial/8250/8250_port.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
[PATCH] tty: serial: Work around warning backtrace in serial8250_set_defaults
Posted by Guenter Roeck 1 year ago 
Commit 7c7e6c8924e7 ("tty: serial: handle HAS_IOPORT dependencies")
triggers warning backtraces on a number of platforms which don't support
IO ports.

WARNING: CPU: 0 PID: 0 at drivers/tty/serial/8250/8250_port.c:470 serial8250_set_defaults+0x148/0x1d8
Unsupported UART type 0

The problem is seen because serial8250_set_defaults() is called for
all members of the serial8250_ports[] array even if that array is
not initialized.

Work around the problem by only displaying the warning if the port
type is not 0 (UPIO_PORT) or if iobase is set for the port.

Fixes: 7c7e6c8924e7 ("tty: serial: handle HAS_IOPORT dependencies")
Acked-by: Arnd Bergmann <arnd@arndb.de>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
---
A complete fix will require a sequence of patches, which will have to be
tested thoroughly and is thus not 6.13 material. This patch doesn't fix
the underlying problem, but it is good enough for 6.13, or at least not
worse than 6.12, while at the same time avoiding the warning backtrace.

 drivers/tty/serial/8250/8250_port.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/tty/serial/8250/8250_port.c b/drivers/tty/serial/8250/8250_port.c
index 4d63d80e78a9..649e74e9b52f 100644
--- a/drivers/tty/serial/8250/8250_port.c
+++ b/drivers/tty/serial/8250/8250_port.c
@@ -467,7 +467,8 @@ static void set_io_from_upio(struct uart_port *p)
 		break;
 #endif
 	default:
-		WARN(1, "Unsupported UART type %x\n", p->iotype);
+		WARN(p->iotype != UPIO_PORT || p->iobase,
+		     "Unsupported UART type %x\n", p->iotype);
 		p->serial_in = no_serial_in;
 		p->serial_out = no_serial_out;
 	}
-- 
2.45.2
Re: [PATCH] tty: serial: Work around warning backtrace in serial8250_set_defaults
Posted by Greg Kroah-Hartman 1 year ago 
On Thu, Dec 05, 2024 at 06:30:33AM -0800, Guenter Roeck wrote:
> Commit 7c7e6c8924e7 ("tty: serial: handle HAS_IOPORT dependencies")
> triggers warning backtraces on a number of platforms which don't support
> IO ports.
> 
> WARNING: CPU: 0 PID: 0 at drivers/tty/serial/8250/8250_port.c:470 serial8250_set_defaults+0x148/0x1d8
> Unsupported UART type 0
> 
> The problem is seen because serial8250_set_defaults() is called for
> all members of the serial8250_ports[] array even if that array is
> not initialized.
> 
> Work around the problem by only displaying the warning if the port
> type is not 0 (UPIO_PORT) or if iobase is set for the port.
> 
> Fixes: 7c7e6c8924e7 ("tty: serial: handle HAS_IOPORT dependencies")
> Acked-by: Arnd Bergmann <arnd@arndb.de>
> Cc: Niklas Schnelle <schnelle@linux.ibm.com>
> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
> ---
> A complete fix will require a sequence of patches, which will have to be
> tested thoroughly and is thus not 6.13 material. This patch doesn't fix
> the underlying problem, but it is good enough for 6.13, or at least not
> worse than 6.12, while at the same time avoiding the warning backtrace.

Thanks, I'll take this for now.

greg k-h
Re: [PATCH] tty: serial: Work around warning backtrace in serial8250_set_defaults
Posted by Stafford Horne 1 year ago 
On Thu, Dec 05, 2024 at 06:30:33AM -0800, Guenter Roeck wrote:
> Commit 7c7e6c8924e7 ("tty: serial: handle HAS_IOPORT dependencies")
> triggers warning backtraces on a number of platforms which don't support
> IO ports.
> 
> WARNING: CPU: 0 PID: 0 at drivers/tty/serial/8250/8250_port.c:470 serial8250_set_defaults+0x148/0x1d8
> Unsupported UART type 0
> 
> The problem is seen because serial8250_set_defaults() is called for
> all members of the serial8250_ports[] array even if that array is
> not initialized.
> 
> Work around the problem by only displaying the warning if the port
> type is not 0 (UPIO_PORT) or if iobase is set for the port.
> 
> Fixes: 7c7e6c8924e7 ("tty: serial: handle HAS_IOPORT dependencies")
> Acked-by: Arnd Bergmann <arnd@arndb.de>
> Cc: Niklas Schnelle <schnelle@linux.ibm.com>
> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
> ---
> A complete fix will require a sequence of patches, which will have to be
> tested thoroughly and is thus not 6.13 material. This patch doesn't fix
> the underlying problem, but it is good enough for 6.13, or at least not
> worse than 6.12, while at the same time avoiding the warning backtrace.

This works on OpenRISC where I was seeing the issue.  Has someone comitted to
work on the root cause?

Tested-by: Stafford Horne <shorne@gmail.com>

>  drivers/tty/serial/8250/8250_port.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/tty/serial/8250/8250_port.c b/drivers/tty/serial/8250/8250_port.c
> index 4d63d80e78a9..649e74e9b52f 100644
> --- a/drivers/tty/serial/8250/8250_port.c
> +++ b/drivers/tty/serial/8250/8250_port.c
> @@ -467,7 +467,8 @@ static void set_io_from_upio(struct uart_port *p)
>  		break;
>  #endif
>  	default:
> -		WARN(1, "Unsupported UART type %x\n", p->iotype);
> +		WARN(p->iotype != UPIO_PORT || p->iobase,
> +		     "Unsupported UART type %x\n", p->iotype);
>  		p->serial_in = no_serial_in;
>  		p->serial_out = no_serial_out;
>  	}
> -- 
> 2.45.2
> 
>
Re: [PATCH] tty: serial: Work around warning backtrace in serial8250_set_defaults
Posted by Guenter Roeck 1 year ago 
On Thu, Dec 05, 2024 at 07:12:42PM +0000, Stafford Horne wrote:
> On Thu, Dec 05, 2024 at 06:30:33AM -0800, Guenter Roeck wrote:
> > Commit 7c7e6c8924e7 ("tty: serial: handle HAS_IOPORT dependencies")
> > triggers warning backtraces on a number of platforms which don't support
> > IO ports.
> > 
> > WARNING: CPU: 0 PID: 0 at drivers/tty/serial/8250/8250_port.c:470 serial8250_set_defaults+0x148/0x1d8
> > Unsupported UART type 0
> > 
> > The problem is seen because serial8250_set_defaults() is called for
> > all members of the serial8250_ports[] array even if that array is
> > not initialized.
> > 
> > Work around the problem by only displaying the warning if the port
> > type is not 0 (UPIO_PORT) or if iobase is set for the port.
> > 
> > Fixes: 7c7e6c8924e7 ("tty: serial: handle HAS_IOPORT dependencies")
> > Acked-by: Arnd Bergmann <arnd@arndb.de>
> > Cc: Niklas Schnelle <schnelle@linux.ibm.com>
> > Signed-off-by: Guenter Roeck <linux@roeck-us.net>
> > ---
> > A complete fix will require a sequence of patches, which will have to be
> > tested thoroughly and is thus not 6.13 material. This patch doesn't fix
> > the underlying problem, but it is good enough for 6.13, or at least not
> > worse than 6.12, while at the same time avoiding the warning backtrace.
> 
> This works on OpenRISC where I was seeing the issue.  Has someone comitted to
> work on the root cause?
> 
Yes, Arnd is working on it.

Guenter

> Tested-by: Stafford Horne <shorne@gmail.com>
> 
> >  drivers/tty/serial/8250/8250_port.c | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/tty/serial/8250/8250_port.c b/drivers/tty/serial/8250/8250_port.c
> > index 4d63d80e78a9..649e74e9b52f 100644
> > --- a/drivers/tty/serial/8250/8250_port.c
> > +++ b/drivers/tty/serial/8250/8250_port.c
> > @@ -467,7 +467,8 @@ static void set_io_from_upio(struct uart_port *p)
> >  		break;
> >  #endif
> >  	default:
> > -		WARN(1, "Unsupported UART type %x\n", p->iotype);
> > +		WARN(p->iotype != UPIO_PORT || p->iobase,
> > +		     "Unsupported UART type %x\n", p->iotype);
> >  		p->serial_in = no_serial_in;
> >  		p->serial_out = no_serial_out;
> >  	}
> > -- 
> > 2.45.2
> > 
> >

Patchew 2.1-devel-b67e4c2

© 2016 - 2025 Red Hat, Inc.