xenocara/xserver/hw/xfree86/os-support/sco/sco_iop.c

91 lines
3.1 KiB
C

/*
* Copyright 2001 by J. Kean Johnston <jkj@caldera.com>
*
* Permission to use, copy, modify, distribute, and sell this software and its
* documentation for any purpose is hereby granted without fee, provided that
* the above copyright notice appear in all copies and that both that
* copyright notice and this permission notice appear in supporting
* documentation, and that the name J. Kean Johnston not be used in
* advertising or publicity pertaining to distribution of the software without
* specific, written prior permission. J. Kean Johnston makes no
* representations about the suitability of this software for any purpose.
* It is provided "as is" without express or implied warranty.
*
* J. KEAN JOHNSTON DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
* INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
* EVENT SHALL J. KEAN JOHNSTON BE LIABLE FOR ANY SPECIAL, INDIRECT OR
* CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF
* USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR
* OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
* PERFORMANCE OF THIS SOFTWARE.
*/
#ifdef HAVE_XORG_CONFIG_H
#include <xorg-config.h>
#endif
#include <X11/X.h>
#include "compiler.h"
#define _NEED_SYSI86
#include "xf86.h"
#include "xf86Priv.h"
#include "xf86OSpriv.h"
#include "xf86_OSlib.h"
/***************************************************************************/
/* I/O Permissions section */
/***************************************************************************/
/*
* There is a right way and a wrong way of doing this. Unfortunately, we
* are forced to do it the wrong way. The right way is to be told the range
* or ranges of I/O ports the driver(s) need access to, in order to use the
* CONS_IOPERM ioctl() to grant access only to those ports we care about.
* This way we can guarantee some small level of stability because a driver
* does not have access to all ports (which would mean it could play with
* the PIT and thus affect scheduling times, or a whole slew of other
* nasty things). However, because XFree86 currently only enables or disables
* ALL port access, we need to run at IOPL 3, which basically means the
* X Server runs at the same level as the kernel. You can image why this is
* unsafe. Oh, and this is not a problem unique to OSR5, other OSes are
* affected by this as well.
*
* So, for the time being, we change our IOPL until such time as the XFree86
* architecture is changed to allow for tighter control of I/O ports. If and
* when it is, then the CONS_ADDIOP/DELIOP ioctl() should be used to enable
* or disable access to the desired ports.
*/
extern long sysi86 (int cmd, ...);
static Bool IOEnabled = FALSE;
_X_EXPORT Bool
xf86EnableIO(void)
{
if (IOEnabled)
return TRUE;
if (sysi86(SI86V86, V86SC_IOPL, PS_IOPL) < 0) {
xf86Msg(X_WARNING,"Failed to set IOPL for extended I/O\n");
return FALSE;
}
IOEnabled = TRUE;
return TRUE;
}
_X_EXPORT void
xf86DisableIO(void)
{
if (!IOEnabled)
return;
sysi86(SI86V86, V86SC_IOPL, 0);
IOEnabled = FALSE;
}