Xen devel - Jun 2006 - [PATCH] xenconsole: fix xencons=tty[0-9]+ param

Hi,
	If you boot with e.g. xencons=tty3, then tty1/tty2 aren''t valid
devices.

	Attached patch makes it always create valid devices on tty1-tty63, but
all but the one specified by xencons are dummy devices.

Cheers,
Mark.


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 30 Jun 2006, at 17:56, Mark McLoughlin wrote:
> 	If you boot with e.g. xencons=tty3, then tty1/tty2 aren''t valid
> devices.
>
> 	Attached patch makes it always create valid devices on tty1-tty63, but
> all but the one specified by xencons are dummy devices.
Is your change to ttyS correct for anything other than ttyS0? Looks 
like you''d incorrectly bind on ttyS0.

Also I see you made a predicate more complicated which is repeated many 
times. Could you hide it behind a suitably named macro? You may need to 
make it more complicated still to fix ttyS case...

  -- Keir


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Hi,

On Fri, 2006-06-30 at 18:31 +0100, Keir Fraser wrote:
> On 30 Jun 2006, at 17:56, Mark McLoughlin wrote:
> 
> > 	If you boot with e.g. xencons=tty3, then tty1/tty2 aren''t
valid
> > devices.
> >
> > 	Attached patch makes it always create valid devices on tty1-tty63,
but
> > all but the one specified by xencons are dummy devices.
> 
> Is your change to ttyS correct for anything other than ttyS0? Looks 
> like you''d incorrectly bind on ttyS0.
>
> Also I see you made a predicate more complicated which is repeated many 
> times. Could you hide it behind a suitably named macro? You may need to 
> make it more complicated still to fix ttyS case...
	Good points. How about this?

Thanks,
Mark.



_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

>> Is your change to ttyS correct for anything other than ttyS0? Looks
>> like you''d incorrectly bind on ttyS0.
>>
>> Also I see you made a predicate more complicated which is repeated 
>> many
>> times. Could you hide it behind a suitably named macro? You may need 
>> to
>> make it more complicated still to fix ttyS case...
>
> 	Good points. How about this?
>
All looks okay except the following change:

-	tty_register_device(xencons_driver, 0, NULL);
+	tty_register_device(xencons_driver, XC_SERIAL ? 0 : xc_num - 1, NULL);

First, it ought to be (xc_mode == XC_SERIAL) right? Secondly, is 
registering just one of the tty devices what you want to do? Do you not 
need to register the ''dummy'' ones for this trick to work?

  -- Keir
  


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On Sun, 2006-07-02 at 16:39 +0100, Keir Fraser wrote:
> >> Is your change to ttyS correct for anything other than ttyS0?
Looks
> >> like you''d incorrectly bind on ttyS0.
> >>
> >> Also I see you made a predicate more complicated which is repeated
> >> many
> >> times. Could you hide it behind a suitably named macro? You may
need
> >> to
> >> make it more complicated still to fix ttyS case...
> >
> > 	Good points. How about this?
> >
> 
> All looks okay except the following change:
> 
> -	tty_register_device(xencons_driver, 0, NULL);
> +	tty_register_device(xencons_driver, XC_SERIAL ? 0 : xc_num - 1, NULL);
> 
> First, it ought to be (xc_mode == XC_SERIAL) right? 
	Doh, yes.
> Secondly, is registering just one of the tty devices what you want to
> do? Do you not need to register the ''dummy'' ones for this
trick to
> work?
	The previous behaviour was to only register the real tty, so I didn''t
want to change that.

	It seems that tty_register_device() only registers the device with
devfs and sysfs. Since we don''t use devfs, I wasn''t seeing any
real side
effects in testing, but I guess if you''re using devfs you want the
dummy
devices registered too.

	Removing the TTY_NO_DEVFS flag so that tty_register_driver() registers
the devices automatically seems to be the best thing to do?

Cheers,
Mark.


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel