thp: madvise(MADV_NOHUGEPAGE)

Add madvise MADV_NOHUGEPAGE to mark regions that are not important to be
hugepage backed.  Return -EINVAL if the vma is not of an anonymous type,
or the feature isn't built into the kernel.  Never silently return
success.

Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

2 files changed, 33 insertions, 12 deletions

diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index f4f6041176a4..fce667c0281d 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -16,6 +16,7 @@
 #include <linux/kthread.h>
 #include <linux/khugepaged.h>
 #include <linux/freezer.h>
+#include <linux/mman.h>
 #include <asm/tlb.h>
 #include <asm/pgalloc.h>
 #include "internal.h"
@@ -1388,18 +1389,36 @@ out:
         return ret;
 }
 
-int hugepage_madvise(unsigned long *vm_flags)
+int hugepage_madvise(unsigned long *vm_flags, int advice)
 {
-        /*
-         * Be somewhat over-protective like KSM for now!
-         */
-        if (*vm_flags & (VM_HUGEPAGE | VM_SHARED  | VM_MAYSHARE   |
-                         VM_PFNMAP   | VM_IO      | VM_DONTEXPAND |
-                         VM_RESERVED | VM_HUGETLB | VM_INSERTPAGE |
-                         VM_MIXEDMAP | VM_SAO))
-                return -EINVAL;
-
-        *vm_flags |= VM_HUGEPAGE;
+        switch (advice) {
+        case MADV_HUGEPAGE:
+                /*
+                 * Be somewhat over-protective like KSM for now!
+                 */
+                if (*vm_flags & (VM_HUGEPAGE |
+                                 VM_SHARED   | VM_MAYSHARE   |
+                                 VM_PFNMAP   | VM_IO      | VM_DONTEXPAND |
+                                 VM_RESERVED | VM_HUGETLB | VM_INSERTPAGE |
+                                 VM_MIXEDMAP | VM_SAO))
+                        return -EINVAL;
+                *vm_flags &= ~VM_NOHUGEPAGE;
+                *vm_flags |= VM_HUGEPAGE;
+                break;
+        case MADV_NOHUGEPAGE:
+                /*
+                 * Be somewhat over-protective like KSM for now!
+                 */
+                if (*vm_flags & (VM_NOHUGEPAGE |
+                                 VM_SHARED   | VM_MAYSHARE   |
+                                 VM_PFNMAP   | VM_IO      | VM_DONTEXPAND |
+                                 VM_RESERVED | VM_HUGETLB | VM_INSERTPAGE |
+                                 VM_MIXEDMAP | VM_SAO))
+                        return -EINVAL;
+                *vm_flags &= ~VM_HUGEPAGE;
+                *vm_flags |= VM_NOHUGEPAGE;
+                break;
+        }
 
         return 0;
 }
diff --git a/mm/madvise.c b/mm/madvise.c
index ecde40a401c1..bbac126e03ed 100644
--- a/mm/madvise.c
+++ b/mm/madvise.c
@@ -72,7 +72,8 @@ static long madvise_behavior(struct vm_area_struct * vma,
                         goto out;
                 break;
         case MADV_HUGEPAGE:
-                error = hugepage_madvise(&new_flags);
+        case MADV_NOHUGEPAGE:
+                error = hugepage_madvise(&new_flags, behavior);
                 if (error)
                         goto out;
                 break;
@@ -290,6 +291,7 @@ madvise_behavior_valid(int behavior)
 #endif
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
         case MADV_HUGEPAGE:
+        case MADV_NOHUGEPAGE:
 #endif
                 return 1;