dax: use common 4k zero page for dax mmap reads

When servicing mmap() reads from file holes the current DAX code
allocates a page cache page of all zeroes and places the struct page
pointer in the mapping->page_tree radix tree.

This has three major drawbacks:

1) It consumes memory unnecessarily. For every 4k page that is read via
   a DAX mmap() over a hole, we allocate a new page cache page. This
   means that if you read 1GiB worth of pages, you end up using 1GiB of
   zeroed memory. This is easily visible by looking at the overall
   memory consumption of the system or by looking at /proc/[pid]/smaps:

	7f62e72b3000-7f63272b3000 rw-s 00000000 103:00 12   /root/dax/data
	Size:            1048576 kB
	Rss:             1048576 kB
	Pss:             1048576 kB
	Shared_Clean:          0 kB
	Shared_Dirty:          0 kB
	Private_Clean:   1048576 kB
	Private_Dirty:         0 kB
	Referenced:      1048576 kB
	Anonymous:             0 kB
	LazyFree:              0 kB
	AnonHugePages:         0 kB
	ShmemPmdMapped:        0 kB
	Shared_Hugetlb:        0 kB
	Private_Hugetlb:       0 kB
	Swap:                  0 kB
	SwapPss:               0 kB
	KernelPageSize:        4 kB
	MMUPageSize:           4 kB
	Locked:                0 kB

2) It is slower than using a common zero page because each page fault
   has more work to do. Instead of just inserting a common zero page we
   have to allocate a page cache page, zero it, and then insert it. Here
   are the average latencies of dax_load_hole() as measured by ftrace on
   a random test box:

    Old method, using zeroed page cache pages:	3.4 us
    New method, using the common 4k zero page:	0.8 us

   This was the average latency over 1 GiB of sequential reads done by
   this simple fio script:

     [global]
     size=1G
     filename=/root/dax/data
     fallocate=none
     [io]
     rw=read
     ioengine=mmap

3) The fact that we had to check for both DAX exceptional entries and
   for page cache pages in the radix tree made the DAX code more
   complex.

Solve these issues by following the lead of the DAX PMD code and using a
common 4k zero page instead.  As with the PMD code we will now insert a
DAX exceptional entry into the radix tree instead of a struct page
pointer which allows us to remove all the special casing in the DAX
code.

Note that we do still pretty aggressively check for regular pages in the
DAX radix tree, especially where we take action based on the bits set in
the page.  If we ever find a regular page in our radix tree now that
most likely means that someone besides DAX is inserting pages (which has
happened lots of times in the past), and we want to find that out early
and fail loudly.

This solution also removes the extra memory consumption.  Here is that
same /proc/[pid]/smaps after 1GiB of reading from a hole with the new
code:

	7f2054a74000-7f2094a74000 rw-s 00000000 103:00 12   /root/dax/data
	Size:            1048576 kB
	Rss:                   0 kB
	Pss:                   0 kB
	Shared_Clean:          0 kB
	Shared_Dirty:          0 kB
	Private_Clean:         0 kB
	Private_Dirty:         0 kB
	Referenced:            0 kB
	Anonymous:             0 kB
	LazyFree:              0 kB
	AnonHugePages:         0 kB
	ShmemPmdMapped:        0 kB
	Shared_Hugetlb:        0 kB
	Private_Hugetlb:       0 kB
	Swap:                  0 kB
	SwapPss:               0 kB
	KernelPageSize:        4 kB
	MMUPageSize:           4 kB
	Locked:                0 kB

Overall system memory consumption is similarly improved.

Another major change is that we remove dax_pfn_mkwrite() from our fault
flow, and instead rely on the page fault itself to make the PTE dirty
and writeable.  The following description from the patch adding the
vm_insert_mixed_mkwrite() call explains this a little more:

   "To be able to use the common 4k zero page in DAX we need to have our
    PTE fault path look more like our PMD fault path where a PTE entry
    can be marked as dirty and writeable as it is first inserted rather
    than waiting for a follow-up dax_pfn_mkwrite() =>
    finish_mkwrite_fault() call.

    Right now we can rely on having a dax_pfn_mkwrite() call because we
    can distinguish between these two cases in do_wp_page():

            case 1: 4k zero page => writable DAX storage
            case 2: read-only DAX storage => writeable DAX storage

    This distinction is made by via vm_normal_page(). vm_normal_page()
    returns false for the common 4k zero page, though, just as it does
    for DAX ptes. Instead of special casing the DAX + 4k zero page case
    we will simplify our DAX PTE page fault sequence so that it matches
    our DAX PMD sequence, and get rid of the dax_pfn_mkwrite() helper.
    We will instead use dax_iomap_fault() to handle write-protection
    faults.

    This means that insert_pfn() needs to follow the lead of
    insert_pfn_pmd() and allow us to pass in a 'mkwrite' flag. If
    'mkwrite' is set insert_pfn() will do the work that was previously
    done by wp_page_reuse() as part of the dax_pfn_mkwrite() call path"

Link: http://lkml.kernel.org/r/20170724170616.25810-4-ross.zwisler@linux.intel.comSigned-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: "Darrick J. Wong" <darrick.wong@oracle.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Andreas Dilger <adilger.kernel@dilger.ca>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox <mawilcox@microsoft.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Documentation/filesystems/dax.txt

@@ -63,9 +63,8 @@ Filesystem support consists of
 - implementing an mmap file operation for DAX files which sets the
   VM_MIXEDMAP and VM_HUGEPAGE flags on the VMA, and setting the vm_ops to
   include handlers for fault, pmd_fault, page_mkwrite, pfn_mkwrite. These
-  handlers should probably call dax_iomap_fault() (for fault and page_mkwrite
-  handlers), dax_iomap_pmd_fault(), dax_pfn_mkwrite() passing the appropriate
-  iomap operations.
+  handlers should probably call dax_iomap_fault() passing the appropriate
+  fault size and iomap operations.
 - calling iomap_zero_range() passing appropriate iomap operations instead of
   block_truncate_page() for DAX files
 - ensuring that there is sufficient locking between reads, writes,

fs/ext2/file.c

@@ -107,29 +107,6 @@ static int ext2_dax_fault(struct vm_fault *vmf)
 	return ret;
 }
 
-static int ext2_dax_pfn_mkwrite(struct vm_fault *vmf)
-{
-	struct inode *inode = file_inode(vmf->vma->vm_file);
-	struct ext2_inode_info *ei = EXT2_I(inode);
-	loff_t size;
-	int ret;
-
-	sb_start_pagefault(inode->i_sb);
-	file_update_time(vmf->vma->vm_file);
-	down_read(&ei->dax_sem);
-
-	/* check that the faulting page hasn't raced with truncate */
-	size = (i_size_read(inode) + PAGE_SIZE - 1) >> PAGE_SHIFT;
-	if (vmf->pgoff >= size)
-		ret = VM_FAULT_SIGBUS;
-	else
-		ret = dax_pfn_mkwrite(vmf);
-
-	up_read(&ei->dax_sem);
-	sb_end_pagefault(inode->i_sb);
-	return ret;
-}
-
 static const struct vm_operations_struct ext2_dax_vm_ops = {
 	.fault		= ext2_dax_fault,
 	/*
@@ -138,7 +115,7 @@ static const struct vm_operations_struct ext2_dax_vm_ops = {
 	 * will always fail and fail back to regular faults.
 	 */
 	.page_mkwrite	= ext2_dax_fault,
-	.pfn_mkwrite	= ext2_dax_pfn_mkwrite,
+	.pfn_mkwrite	= ext2_dax_fault,
 };
 
 static int ext2_file_mmap(struct file *file, struct vm_area_struct *vma)

fs/ext4/file.c

@@ -311,41 +311,11 @@ static int ext4_dax_fault(struct vm_fault *vmf)
 	return ext4_dax_huge_fault(vmf, PE_SIZE_PTE);
 }
 
-/*
- * Handle write fault for VM_MIXEDMAP mappings. Similarly to ext4_dax_fault()
- * handler we check for races agaist truncate. Note that since we cycle through
- * i_mmap_sem, we are sure that also any hole punching that began before we
- * were called is finished by now and so if it included part of the file we
- * are working on, our pte will get unmapped and the check for pte_same() in
- * wp_pfn_shared() fails. Thus fault gets retried and things work out as
- * desired.
- */
-static int ext4_dax_pfn_mkwrite(struct vm_fault *vmf)
-{
-	struct inode *inode = file_inode(vmf->vma->vm_file);
-	struct super_block *sb = inode->i_sb;
-	loff_t size;
-	int ret;
-
-	sb_start_pagefault(sb);
-	file_update_time(vmf->vma->vm_file);
-	down_read(&EXT4_I(inode)->i_mmap_sem);
-	size = (i_size_read(inode) + PAGE_SIZE - 1) >> PAGE_SHIFT;
-	if (vmf->pgoff >= size)
-		ret = VM_FAULT_SIGBUS;
-	else
-		ret = dax_pfn_mkwrite(vmf);
-	up_read(&EXT4_I(inode)->i_mmap_sem);
-	sb_end_pagefault(sb);
-
-	return ret;
-}
-
 static const struct vm_operations_struct ext4_dax_vm_ops = {
 	.fault		= ext4_dax_fault,
 	.huge_fault	= ext4_dax_huge_fault,
 	.page_mkwrite	= ext4_dax_fault,
-	.pfn_mkwrite	= ext4_dax_pfn_mkwrite,
+	.pfn_mkwrite	= ext4_dax_fault,
 };
 #else
 #define ext4_dax_vm_ops	ext4_file_vm_ops

fs/xfs/xfs_file.c

@@ -1130,7 +1130,7 @@ xfs_filemap_pfn_mkwrite(
 	if (vmf->pgoff >= size)
 		ret = VM_FAULT_SIGBUS;
 	else if (IS_DAX(inode))
-		ret = dax_pfn_mkwrite(vmf);
+		ret = dax_iomap_fault(vmf, PE_SIZE_PTE, &xfs_iomap_ops);
 	xfs_iunlock(ip, XFS_MMAPLOCK_SHARED);
 	sb_end_pagefault(inode->i_sb);
 	return ret;

include/linux/dax.h

@@ -91,18 +91,17 @@ bool dax_write_cache_enabled(struct dax_device *dax_dev);
 
 /*
  * We use lowest available bit in exceptional entry for locking, one bit for
- * the entry size (PMD) and two more to tell us if the entry is a huge zero
- * page (HZP) or an empty entry that is just used for locking.  In total four
- * special bits.
+ * the entry size (PMD) and two more to tell us if the entry is a zero page or
+ * an empty entry that is just used for locking.  In total four special bits.
  *
- * If the PMD bit isn't set the entry has size PAGE_SIZE, and if the HZP and
- * EMPTY bits aren't set the entry is a normal DAX entry with a filesystem
+ * If the PMD bit isn't set the entry has size PAGE_SIZE, and if the ZERO_PAGE
+ * and EMPTY bits aren't set the entry is a normal DAX entry with a filesystem
  * block allocation.
  */
 #define RADIX_DAX_SHIFT	(RADIX_TREE_EXCEPTIONAL_SHIFT + 4)
 #define RADIX_DAX_ENTRY_LOCK (1 << RADIX_TREE_EXCEPTIONAL_SHIFT)
 #define RADIX_DAX_PMD (1 << (RADIX_TREE_EXCEPTIONAL_SHIFT + 1))
-#define RADIX_DAX_HZP (1 << (RADIX_TREE_EXCEPTIONAL_SHIFT + 2))
+#define RADIX_DAX_ZERO_PAGE (1 << (RADIX_TREE_EXCEPTIONAL_SHIFT + 2))
 #define RADIX_DAX_EMPTY (1 << (RADIX_TREE_EXCEPTIONAL_SHIFT + 3))
 
 static inline unsigned long dax_radix_sector(void *entry)
@@ -153,7 +152,6 @@ static inline unsigned int dax_radix_order(void *entry)
 	return 0;
 }
 #endif
-int dax_pfn_mkwrite(struct vm_fault *vmf);
 
 static inline bool dax_mapping(struct address_space *mapping)
 {

include/trace/events/fs_dax.h

@@ -190,8 +190,6 @@ DEFINE_EVENT(dax_pte_fault_class, name, \
 
 DEFINE_PTE_FAULT_EVENT(dax_pte_fault);
 DEFINE_PTE_FAULT_EVENT(dax_pte_fault_done);
-DEFINE_PTE_FAULT_EVENT(dax_pfn_mkwrite_no_entry);
-DEFINE_PTE_FAULT_EVENT(dax_pfn_mkwrite);
 DEFINE_PTE_FAULT_EVENT(dax_load_hole);
 
 TRACE_EVENT(dax_insert_mapping,