On Thu, 21 Apr 2022 15:51:03 -0400 Peter Xu <peterx@redhat.com> wrote:

> > - * file-backed memory.  This should only be specified when we will completely
> > - * drop the page in the mm, either by truncation or unmapping of the vma.  By
> > - * default, the flag is not set.
> > - */
> > -#define  ZAP_FLAG_DROP_MARKER        ((__force zap_flags_t) BIT(0))
> > -
> >  #endif /* _LINUX_MM_H */
> > _
> > 
> 
> Andrew,
> 
> It may work, but I have worry that zap_flags_t should still be a common
> struct for mm not hugetlb specific, let's say it's still legal some .c file
> wants to reference it without hugetlb knowledge?
> 
> >From that POV, could we perhaps move these chunk into mm_types.h (which
> hugetlb.h includes anyway)?
> 

Sure.  ZAP_FLAG_DROP_MARKER isn't a type so how about we leave that in
mm.h?


--- a/include/linux/mm.h~mm-hugetlb-only-drop-uffd-wp-special-pte-if-required-fix-fix
+++ a/include/linux/mm.h
@@ -3431,8 +3431,6 @@ madvise_set_anon_name(struct mm_struct *
 }
 #endif
 
-typedef unsigned int __bitwise zap_flags_t;
-
 /*
  * Whether to drop the pte markers, for example, the uffd-wp information for
  * file-backed memory.  This should only be specified when we will completely
--- a/include/linux/mm_types.h~mm-hugetlb-only-drop-uffd-wp-special-pte-if-required-fix-fix
+++ a/include/linux/mm_types.h
@@ -850,4 +850,6 @@ enum fault_flag {
 	FAULT_FLAG_ORIG_PTE_VALID =	1 << 11,
 };
 
+typedef unsigned int __bitwise zap_flags_t;
+
 #endif /* _LINUX_MM_TYPES_H */
_