Re: [Xen-devel] [PATCH 2 of 4] minios/console: console_input() weak reference

[Ian Campbell <Ian.Campbell@xxxxxxxxxx>]

On Tue, 2012-11-27 at 11:39 +0000, Samuel Thibault wrote:
> Ian Campbell, le Tue 27 Nov 2012 11:34:39 +0000, a Ãcrit :
> > On Mon, 2012-11-26 at 21:42 +0000, Samuel Thibault wrote:
> > > Andrew Cooper, le Mon 26 Nov 2012 16:24:10 +0000, a Ãcrit :
> > > > In exactly the same style as app_main() in kernel.c, create a weak
> > > > reference console_input() function for applications to override to
> > > > quickly gain access to the console.
> > > > 
> > > > Signed-off-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>
> > > 
> > > Acked-by: Samuel Thibault <samuel.thibault@xxxxxxxxxxxx>
> > 
> > Any reason not to make xencons_rx itself the weak function?
> 
> I'd avoid exposing the pt_regs structure. Maybe we could drop the
> parameter from the xencons_rx function itself indeed, it has no use.

Sounds like a good idea to me.

> > And by extension I suppose xencons_tx too, for parity if nothing else.
> 
> I actually wonder what the purpose of xencons_tx is.

I suppose if you had a shared ring serial console you could accept
input? Less useful if you are using HYPERVISOR_console_io from a domU I
suspect...

> > The existing console_input function (previously xencons_rx) is pretty
> > odd, it relies on something reusing a tx buffer for rx?
> 
> ? I don't see what you are referring to.

xencons_rx says /* Just repeat what's written */ and prints the contents
of the buffer (this is rx remember, so buffer ought to be an output
buffer, not input) while xencons_tx just says /* Do nothing, handled by
_rx */.


> 
> Samuel



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel