qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [PATCH] evdev_keycode_map


From: Daniel P. Berrange
Subject: Re: [Qemu-devel] [PATCH] evdev_keycode_map
Date: Wed, 10 Dec 2008 11:28:45 +0000
User-agent: Mutt/1.4.1i

On Wed, Dec 10, 2008 at 12:42:35AM +0200, Riku Voipio wrote:
> On Wed, Sep 24, 2008 at 06:10:27PM -0500, Dustin Kirkland wrote:
> > I'm attaching a patch that we have applied to the qemu implementation
> > inside of the kvm package in Ubuntu.
> 
> Anyone know why this patch is still lingering?
> 
> > --- kvm-72+dfsg/qemu/configure      2008-07-27 08:20:10.000000000 -0500
> > +++ kvm-72+dfsg/qemu/configure      2008-09-24 16:04:28.036791147 -0500
> > @@ -1198,9 +1198,9 @@ if test "$sdl1" = "yes" ; then
> >    echo "#define CONFIG_SDL 1" >> $config_h
> >    echo "CONFIG_SDL=yes" >> $config_mak
> >    if test "$target_softmmu" = "no" -o "$static" = "yes"; then
> > -    echo "SDL_LIBS=$sdl_static_libs" >> $config_mak
> > +    echo "SDL_LIBS=-lX11 $sdl_static_libs" >> $config_mak
> >    else
> > -    echo "SDL_LIBS=`$sdl_config --libs`" >> $config_mak
> > +    echo "SDL_LIBS=-lX11 `$sdl_config --libs`" >> $config_mak
> >    fi
> 
> These changes make the SDL code dependant on X ? This
> would make compiling windows less trivial..

It ought to be possible to just #ifdef out the bit of the code which 
checks for evdev on non-Linux, and just default to the existing 
mapping QEMU had.

> > +uint8_t _translate_keycode(int keycode)
> >  {
> > -   return x_keycode_to_pc_keycode[key];
> > +    if (keycode < 9)
> > +        keycode = 0;
> > +    else if (keycode < 97)
> > +        keycode -= 8; /* just an offset */
> > +    else if (keycode < 158) {
> > +        if (check_for_evdev())
> > +            keycode = evdev_keycode_to_pc_keycode[keycode - 97];
> > +        else
> > +            keycode = x_keycode_to_pc_keycode[keycode - 97];
> 
> This is suboptimal, for every keycode translation we have several
> rountrips to Xserver and back. Consider someone using a remote X
> server. save the result of check_for_evdev() and use it in subsequent
> calls.

Sub-optimal is an understatement :-) Easiest would be to  declare
another "static uint8_t *keymap", and at startup check for evdev
and initialize this to point to either the PC or evdev key table.

Daniel
-- 
|: Red Hat, Engineering, London   -o-   http://people.redhat.com/berrange/ :|
|: http://libvirt.org  -o-  http://virt-manager.org  -o-  http://ovirt.org :|
|: http://autobuild.org       -o-         http://search.cpan.org/~danberr/ :|
|: GnuPG: 7D3B9505  -o-  F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|




reply via email to

[Prev in Thread] Current Thread [Next in Thread]