Hi Christophe, On 07/19/2012 10:40 AM, Christophe Fergeau wrote:
--- gtk/spice-widget.c | 38 ++++++++++++++++++++++++++++++++++++++ 1 file changed, 38 insertions(+)diff --git a/gtk/spice-widget.c b/gtk/spice-widget.c index 02bb089..e5c48ed 100644 --- a/gtk/spice-widget.c +++ b/gtk/spice-widget.c @@ -979,6 +979,33 @@ static gboolean expose_event(GtkWidget *widget, GdkEventExpose *expose) #endif +static gboolean check_for_ctrl_alt_del(SpiceDisplay *display, int scancode) +{ + /* we need to look at hardware scancodes for ctrl+alt+del detection + * since the symbolic keys sent by gtk+ might have been remapped on the + * client side, which means ctrl+alt+del on the client and ctrl+alt+del + * on the guest might not correspond to the same physical keyboard keys + */ + if (is_key_pressed(display, 0x1d) || is_key_pressed(display, 0x11d)) /* ctrl */ + if (is_key_pressed(display, 0x38) || is_key_pressed(display, 0x138)) /* alt */ + if ((scancode == 0x53) || (scancode == 0x153)) /* DEL or keypad . */ + return true; + + return false;
Nitpick, Are there definitions for those scancodes ? I see e.g. KEY_LEFTALT and KEY_RIGHTALT in gtk/keymaps.csv, with values 56 and 312 (0x38 and 0x138) in their lines. (feel free to ignore and leave the patch as is with the nice comments) _______________________________________________ Spice-devel mailing list Spice-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/spice-devel
