2.6.0-test9-ruby VT_OPENQRY off by one?

[Wayne W. <wh...@Ma...>]

Hello,

I'm trying out 2.6.0-test9-ruby, and I noticed that XFree86 tries to comes
up on the last used vc (vt06) instead of the first unused vc (vt07).  I
think that there is an off by one error in the VT_OPENQRY ioctl and that
the patch below is appropriate.  I've included the full text of the
VT_OPENQRY ioctl, since I'm not sure if I understand everything fully.

Thanks,
Wayne

--- linux-2.6.0-test9-ruby/drivers/char/vt_ioctl.c~	2003-10-02 04:25:00.000000000 -0700
+++ linux-2.6.0-test9-ruby/drivers/char/vt_ioctl.c	2003-11-11 12:33:50.000000000 -0800
@@ -1020,21 +1020,21 @@
 	case VT_OPENQRY:
 	{
 		int j = vc->display_fg->first_vc;

 		for (i = 0; i < vc->display_fg->vc_count; ++i, j++) {
 			struct vc_data *tmp = find_vc(j);

 			if (!tmp || (tmp && !VT_IS_IN_USE(tmp)))
 				break;
 		}
-		ucval = i < vc->display_fg->vc_count ? (j) : -1;
+		ucval = i < vc->display_fg->vc_count ? (j+1) : -1;
 		goto setint;
 	}
 	/*
 	 * ioctl(fd, VT_ACTIVATE, num) will cause us to switch to vt # num,
 	 * unless we attempt to switch to the visible VT, just
 	 * to preserve sanity).
 	 */
 	case VT_ACTIVATE:
 	{
 		struct vc_data *tmp;