Similar to change cba5a10f, xf86ScreenSetCursor() would dereference ScreenPriv
without NULL checking it. If Option "SWCursor" is specified, ScreenPriv == NULL.

Without this fix, it is observed that setting Option "SWCursor" "on" on the
modesetting driver in a PRIME configuration will segfault the server.

It is important to return success rather than failure in the instance that
ScreenPriv == NULL and pCurs == NullCursor, because otherwise xf86SetCursor()
can fall into infinite recursion: xf86SetCursor(pCurs) calls
xf86ScreenSetCursor(pCurs), and if FALSE, calls xf86SetCursor(NullCursor). If
xf86ScreenSetCursor(NullCursor) returns FALSE, it calls
xf86SetCursor(NullCursor) again and this repeats forever.

Signed-off-by: Alex Goins <[email protected]>
---
 hw/xfree86/ramdac/xf86HWCurs.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/hw/xfree86/ramdac/xf86HWCurs.c b/hw/xfree86/ramdac/xf86HWCurs.c
index 2e4c9e5..366837c 100644
--- a/hw/xfree86/ramdac/xf86HWCurs.c
+++ b/hw/xfree86/ramdac/xf86HWCurs.c
@@ -181,9 +181,16 @@ xf86ScreenSetCursor(ScreenPtr pScreen, CursorPtr pCurs, 
int x, int y)
     xf86CursorScreenPtr ScreenPriv =
         (xf86CursorScreenPtr) dixLookupPrivate(&pScreen->devPrivates,
                                                xf86CursorScreenKey);
-    xf86CursorInfoPtr infoPtr = ScreenPriv->CursorInfoPtr;
+
+    xf86CursorInfoPtr infoPtr;
     unsigned char *bits;
 
+    if (!ScreenPriv) { /* NULL if Option "SWCursor" */
+        return (pCurs == NullCursor);
+    }
+
+    infoPtr = ScreenPriv->CursorInfoPtr;
+
     if (pCurs == NullCursor) {
         (*infoPtr->HideCursor) (infoPtr->pScrn);
         return TRUE;
-- 
2.7.4

_______________________________________________
[email protected]: X.Org development
Archives: http://lists.x.org/archives/xorg-devel
Info: https://lists.x.org/mailman/listinfo/xorg-devel

Reply via email to