mm: fix vm_fault_t cast in VM_FAULT_GET_HINDEX()

Symmetrically to VM_FAULT_SET_HINDEX(), we need a force-cast in
VM_FAULT_GET_HINDEX() to tell sparse that this is intentional.

Sparse complains about the current code when building a kernel with
CONFIG_MEMORY_FAILURE:

  arch/x86/mm/fault.c:1058:53: warning: restricted vm_fault_t degrades to integer

Link: http://lkml.kernel.org/r/[email protected]
Fixes: 3d35390 ("mm: create the new vm_fault_t type")
Signed-off-by: Jann Horn <[email protected]>
Reviewed-by: Andrew Morton <[email protected]>
Cc: Souptick Joarder <[email protected]>
Cc: Matthew Wilcox <[email protected]>
Cc: Vlastimil Babka <[email protected]>
Cc: "Kirill A. Shutemov" <[email protected]>
Cc: Rik van Riel <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

include/linux/mm_types.h

@@ -671,7 +671,7 @@ enum vm_fault_reason {
 
 /* Encode hstate index for a hwpoisoned large page */
 #define VM_FAULT_SET_HINDEX(x) ((__force vm_fault_t)((x) << 16))
-#define VM_FAULT_GET_HINDEX(x) (((x) >> 16) & 0xf)
+#define VM_FAULT_GET_HINDEX(x) (((__force unsigned int)(x) >> 16) & 0xf)
 
 #define VM_FAULT_ERROR (VM_FAULT_OOM | VM_FAULT_SIGBUS |	\
 			VM_FAULT_SIGSEGV | VM_FAULT_HWPOISON |	\