http://apple.doit.wisc.edu/mirrors/amug/linux/linuxppc/sources/tarballs/xscreensaver...

[xscreensaver] / utils / visual-gl.c

diff --git a/utils/visual-gl.c b/utils/visual-gl.c
new file mode 100644 (file)
index 0000000..fb83fe4
--- /dev/null
+++ b/utils/visual-gl.c
@@ -0,0 +1,99 @@
+/* xscreensaver, Copyright (c) 1999 by Jamie Zawinski <jwz@jwz.org>
+ *
+ * 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.  No representations are made about the suitability of this
+ * software for any purpose.  It is provided "as is" without express or 
+ * implied warranty.
+ */
+
+/* This file contains code for picking the best visual for GL programs by
+   actually asking the GL library to figure it out for us.  The code in
+   visual.c might do a good job of this on most systems, but not, in
+   particular, on SGIs.
+
+   Why?  Because with the SGI O2 X server is weird.
+
+   GL programs tend to work best on a visual that is *half* as deep as the
+   depth of the screen, since that way, they can do double-buffering.  So
+   generally, if the screen is 24 bits deep, but a 12-bit TrueColor visual
+   is available, then that's the visual you should use.
+
+   But on the server that runs on the O2 (a machine that has serious hardware
+   support for GL) the 12-bit PseudoColor visual looks awful (you get a black
+   and white, flickery image.)  On these machines, the visual you want turns
+   out to be 0x31 -- this is but one of the eight 15-bit TrueColor visuals
+   (yes, 8, and yes, 15) that O2s provide.  This is the only visual that works
+   properly -- as far as `xdpyinfo' is concerned, all of the 15-bit TrueColor
+   visuals are identical, but some flicker like mad, and some have deeply
+   weird artifacts (hidden surfaces show through!)  I suppose these other
+   visuals must be tied to some arcane hardware feature...
+
+   So the bottom line is, there exists information about visuals which is
+   available to GL, but which is not available via Xlib calls.  So the only
+   way to know which visual to use (other than impirically) is to actually
+   call GLX routines.
+ */
+
+#include "utils.h"
+#include "visual.h"
+
+#ifdef HAVE_GL
+# include <GL/gl.h>
+# include <GL/glx.h>
+#endif /* HAVE_GL */
+
+Visual *
+get_gl_visual (Screen *screen)
+{
+#ifdef HAVE_GL
+  XVisualInfo *vi = 0;
+  Display *dpy = DisplayOfScreen (screen);
+  int screen_num = screen_number (screen);
+  int attrs[20];
+  int i = 0;
+
+  attrs[i++] = GLX_RGBA;
+  attrs[i++] = GLX_RED_SIZE;     attrs[i++] = 1;
+  attrs[i++] = GLX_GREEN_SIZE;   attrs[i++] = 1;
+  attrs[i++] = GLX_BLUE_SIZE;    attrs[i++] = 1;
+  attrs[i++] = GLX_DEPTH_SIZE;   attrs[i++] = 1;
+  attrs[i++] = GLX_DOUBLEBUFFER;
+  attrs[i++] = 0;
+
+  vi = glXChooseVisual (dpy, screen_num, attrs);
+
+  if (!vi)                             /* Try without double-buffering. */
+    {
+      attrs[i-1] = 0;
+      vi = glXChooseVisual (dpy, screen_num, attrs);
+    }
+
+  if (!vi)                             /* Try mono. */
+    {
+      i = 0;
+      attrs[i++] = GLX_DOUBLEBUFFER;
+      attrs[i++] = 0;
+      vi = glXChooseVisual (dpy, screen_num, attrs);
+    }
+
+  if (!vi)                             /* Try mono without double-buffer. */
+    {
+      attrs[0] = 0;
+      vi = glXChooseVisual (dpy, screen_num, attrs);
+    }
+
+  if (!vi)
+    return 0;
+  else
+    {
+      Visual *v = vi->visual;
+      XFree (vi);
+      return v;
+    }
+#else  /* !HAVE_GL */
+  return 0;
+#endif /* !HAVE_GL */
+}