mm: prefer fault_around_pages to fault_around_bytes

All use of this value is now at page granularity, so specify the variable
as such too.  This simplifies the logic.

We maintain the debugfs entry to ensure that there are no user-visible
changes.

Link: https://lkml.kernel.org/r/4995bad07fe9baa51c786fa0d81819dddfb57654.1679089214.git.lstoakes@gmail.com
Signed-off-by: Lorenzo Stoakes <lstoakes@gmail.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

diff --git a/mm/memory.c b/mm/memory.c
index 9e6608b5026187ab866172b0c3422edb4ad79b94..0ee13d3de88cd1bab6979b5860e99bb1de9a801f 100644 (file)
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -4405,13 +4405,13 @@ vm_fault_t finish_fault(struct vm_fault *vmf)
        return ret;
 }
 
-static unsigned long fault_around_bytes __read_mostly =
-       rounddown_pow_of_two(65536);
+static unsigned long fault_around_pages __read_mostly =
+       65536 >> PAGE_SHIFT;
 
 #ifdef CONFIG_DEBUG_FS
 static int fault_around_bytes_get(void *data, u64 *val)
 {
-       *val = fault_around_bytes;
+       *val = fault_around_pages << PAGE_SHIFT;
        return 0;
 }
 
@@ -4423,10 +4423,13 @@ static int fault_around_bytes_set(void *data, u64 val)
 {
        if (val / PAGE_SIZE > PTRS_PER_PTE)
                return -EINVAL;
-       if (val > PAGE_SIZE)
-               fault_around_bytes = rounddown_pow_of_two(val);
-       else
-               fault_around_bytes = PAGE_SIZE; /* rounddown_pow_of_two(0) is undefined */
+
+       /*
+        * The minimum value is 1 page, however this results in no fault-around
+        * at all. See should_fault_around().
+        */
+       fault_around_pages = max(rounddown_pow_of_two(val) >> PAGE_SHIFT, 1UL);
+
        return 0;
 }
 DEFINE_DEBUGFS_ATTRIBUTE(fault_around_bytes_fops,
@@ -4452,18 +4455,18 @@ late_initcall(fault_around_debugfs);
  * This function doesn't cross VMA or page table boundaries, in order to call
  * map_pages() and acquire a PTE lock only once.
  *
- * fault_around_bytes defines how many bytes we'll try to map.
+ * fault_around_pages defines how many pages we'll try to map.
  * do_fault_around() expects it to be set to a power of two less than or equal
  * to PTRS_PER_PTE.
  *
  * The virtual address of the area that we map is naturally aligned to
- * fault_around_bytes rounded down to the machine page size
+ * fault_around_pages * PAGE_SIZE rounded down to the machine page size
  * (and therefore to page order).  This way it's easier to guarantee
  * that we don't cross page table boundaries.
  */
 static vm_fault_t do_fault_around(struct vm_fault *vmf)
 {
-       pgoff_t nr_pages = READ_ONCE(fault_around_bytes) >> PAGE_SHIFT;
+       pgoff_t nr_pages = READ_ONCE(fault_around_pages);
        pgoff_t pte_off = pte_index(vmf->address);
        /* The page offset of vmf->address within the VMA. */
        pgoff_t vma_off = vmf->pgoff - vmf->vma->vm_pgoff;
@@ -4498,7 +4501,8 @@ static inline bool should_fault_around(struct vm_fault *vmf)
        if (uffd_disable_fault_around(vmf->vma))
                return false;
 
-       return fault_around_bytes >> PAGE_SHIFT > 1;
+       /* A single page implies no faulting 'around' at all. */
+       return fault_around_pages > 1;
 }
 
 static vm_fault_t do_read_fault(struct vm_fault *vmf)