Don't gratuitously disable Render when ShadowFBing

There's no reason to do this, and I am at a loss to explain why that was
ever done.  The first instance of this logic comes from xfree86 (over 11
years ago!):

http://cvsweb.xfree86.org/cvsweb/xc/programs/Xserver/hw/xfree86/drivers/ati/atiscreen.c.diff?r1=1.15&r2=1.16

Now, cfb never had Render support, so that might make some sense if
there was an intermediate uncommitted state where the driver used fb for
direct but cfb for shadowed.  But that's really the only plausible
explanation I can think of.

Signed-off-by: Adam Jackson <ajax@redhat.com>

1 files changed, 4 insertions, 14 deletions

diff --git a/src/atiscreen.c b/src/atiscreen.c
index b1fbd3c..62a4dd9 100644
--- a/src/atiscreen.c
+++ b/src/atiscreen.c
@@ -476,21 +476,11 @@ ATIScreenInit(SCREEN_INIT_ARGS_DECL)
         }
     }
 
-    /* If applicable, initialise RENDER extension */
+    /* initialise RENDER extension */
+    if (!fbPictureInit(pScreen, NULL, 0) && (serverGeneration == 1))
     {
-        if (pATI->OptionShadowFB)
-        {
-            if (serverGeneration == 1)
-                xf86DrvMsg(pScreenInfo->scrnIndex, X_WARNING,
-                    "RENDER extension not supported with a shadowed"
-                    " framebuffer.\n");
-        }
-        else if (!fbPictureInit(pScreen, NULL, 0) &&
-                 (serverGeneration == 1))
-        {
-            xf86DrvMsg(pScreenInfo->scrnIndex, X_WARNING,
-                "RENDER extension initialisation failed.\n");
-        }
+	xf86DrvMsg(pScreenInfo->scrnIndex, X_WARNING,
+	    "RENDER extension initialisation failed.\n");
     }
 
     xf86SetBlackWhitePixels(pScreen);