Merge tag 'xfs-for-linus-4.9-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs

Pull xfs and iomap updates from Dave Chinner:
 "The main things in this update are the iomap-based DAX infrastructure,
  an XFS delalloc rework, and a chunk of fixes to how log recovery
  schedules writeback to prevent spurious corruption detections when
  recovery of certain items was not required.

  The other main chunk of code is some preparation for the upcoming
  reflink functionality. Most of it is generic and cleanups that stand
  alone, but they were ready and reviewed so are in this pull request.

  Speaking of reflink, I'm currently planning to send you another pull
  request next week containing all the new reflink functionality. I'm
  working through a similar process to the last cycle, where I sent the
  reverse mapping code in a separate request because of how large it
  was. The reflink code merge is even bigger than reverse mapping, so
  I'll be doing the same thing again....

  Summary for this update:

   - change of XFS mailing list to linux-xfs@vger.kernel.org

   - iomap-based DAX infrastructure w/ XFS and ext2 support

   - small iomap fixes and additions

   - more efficient XFS delayed allocation infrastructure based on iomap

   - a rework of log recovery writeback scheduling to ensure we don't
     fail recovery when trying to replay items that are already on disk

   - some preparation patches for upcoming reflink support

   - configurable error handling fixes and documentation

   - aio access time update race fixes for XFS and
     generic_file_read_iter"

* tag 'xfs-for-linus-4.9-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs: (40 commits)
  fs: update atime before I/O in generic_file_read_iter
  xfs: update atime before I/O in xfs_file_dio_aio_read
  ext2: fix possible integer truncation in ext2_iomap_begin
  xfs: log recovery tracepoints to track current lsn and buffer submission
  xfs: update metadata LSN in buffers during log recovery
  xfs: don't warn on buffers not being recovered due to LSN
  xfs: pass current lsn to log recovery buffer validation
  xfs: rework log recovery to submit buffers on LSN boundaries
  xfs: quiesce the filesystem after recovery on readonly mount
  xfs: remote attribute blocks aren't really userdata
  ext2: use iomap to implement DAX
  ext2: stop passing buffer_head to ext2_get_blocks
  xfs: use iomap to implement DAX
  xfs: refactor xfs_setfilesize
  xfs: take the ilock shared if possible in xfs_file_iomap_begin
  xfs: fix locking for DAX writes
  dax: provide an iomap based fault handler
  dax: provide an iomap based dax read/write path
  dax: don't pass buffer_head to copy_user_dax
  dax: don't pass buffer_head to dax_insert_mapping
  ...

1 files changed, 240 insertions, 12 deletions

diff --git a/fs/dax.c b/fs/dax.c
index 993dc6fe0416..cc025f82ef07 100644
--- a/fs/dax.c
+++ b/fs/dax.c
@@ -31,6 +31,8 @@
 #include <linux/vmstat.h>
 #include <linux/pfn_t.h>
 #include <linux/sizes.h>
+#include <linux/iomap.h>
+#include "internal.h"
 
 /*
  * We use lowest available bit in exceptional entry for locking, other two
@@ -580,14 +582,13 @@ static int dax_load_hole(struct address_space *mapping, void *entry,
 	return VM_FAULT_LOCKED;
 }
 
-static int copy_user_bh(struct page *to, struct inode *inode,
-		struct buffer_head *bh, unsigned long vaddr)
+static int copy_user_dax(struct block_device *bdev, sector_t sector, size_t size,
+		struct page *to, unsigned long vaddr)
 {
 	struct blk_dax_ctl dax = {
-		.sector = to_sector(bh, inode),
-		.size = bh->b_size,
+		.sector = sector,
+		.size = size,
 	};
-	struct block_device *bdev = bh->b_bdev;
 	void *vto;
 
 	if (dax_map_atomic(bdev, &dax) < 0)
@@ -790,14 +791,13 @@ int dax_writeback_mapping_range(struct address_space *mapping,
 EXPORT_SYMBOL_GPL(dax_writeback_mapping_range);
 
 static int dax_insert_mapping(struct address_space *mapping,
-			struct buffer_head *bh, void **entryp,
-			struct vm_area_struct *vma, struct vm_fault *vmf)
+		struct block_device *bdev, sector_t sector, size_t size,
+		void **entryp, struct vm_area_struct *vma, struct vm_fault *vmf)
 {
 	unsigned long vaddr = (unsigned long)vmf->virtual_address;
-	struct block_device *bdev = bh->b_bdev;
 	struct blk_dax_ctl dax = {
-		.sector = to_sector(bh, mapping->host),
-		.size = bh->b_size,
+		.sector = sector,
+		.size = size,
 	};
 	void *ret;
 	void *entry = *entryp;
@@ -868,7 +868,8 @@ int dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf,
 	if (vmf->cow_page) {
 		struct page *new_page = vmf->cow_page;
 		if (buffer_written(&bh))
-			error = copy_user_bh(new_page, inode, &bh, vaddr);
+			error = copy_user_dax(bh.b_bdev, to_sector(&bh, inode),
+					bh.b_size, new_page, vaddr);
 		else
 			clear_user_highpage(new_page, vaddr);
 		if (error)
@@ -898,7 +899,8 @@ int dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf,
 
 	/* Filesystem should not return unwritten buffers to us! */
 	WARN_ON_ONCE(buffer_unwritten(&bh) || buffer_new(&bh));
-	error = dax_insert_mapping(mapping, &bh, &entry, vma, vmf);
+	error = dax_insert_mapping(mapping, bh.b_bdev, to_sector(&bh, inode),
+			bh.b_size, &entry, vma, vmf);
  unlock_entry:
 	put_locked_mapping_entry(mapping, vmf->pgoff, entry);
  out:
@@ -1241,3 +1243,229 @@ int dax_truncate_page(struct inode *inode, loff_t from, get_block_t get_block)
 	return dax_zero_page_range(inode, from, length, get_block);
 }
 EXPORT_SYMBOL_GPL(dax_truncate_page);
+
+#ifdef CONFIG_FS_IOMAP
+static loff_t
+iomap_dax_actor(struct inode *inode, loff_t pos, loff_t length, void *data,
+		struct iomap *iomap)
+{
+	struct iov_iter *iter = data;
+	loff_t end = pos + length, done = 0;
+	ssize_t ret = 0;
+
+	if (iov_iter_rw(iter) == READ) {
+		end = min(end, i_size_read(inode));
+		if (pos >= end)
+			return 0;
+
+		if (iomap->type == IOMAP_HOLE || iomap->type == IOMAP_UNWRITTEN)
+			return iov_iter_zero(min(length, end - pos), iter);
+	}
+
+	if (WARN_ON_ONCE(iomap->type != IOMAP_MAPPED))
+		return -EIO;
+
+	while (pos < end) {
+		unsigned offset = pos & (PAGE_SIZE - 1);
+		struct blk_dax_ctl dax = { 0 };
+		ssize_t map_len;
+
+		dax.sector = iomap->blkno +
+			(((pos & PAGE_MASK) - iomap->offset) >> 9);
+		dax.size = (length + offset + PAGE_SIZE - 1) & PAGE_MASK;
+		map_len = dax_map_atomic(iomap->bdev, &dax);
+		if (map_len < 0) {
+			ret = map_len;
+			break;
+		}
+
+		dax.addr += offset;
+		map_len -= offset;
+		if (map_len > end - pos)
+			map_len = end - pos;
+
+		if (iov_iter_rw(iter) == WRITE)
+			map_len = copy_from_iter_pmem(dax.addr, map_len, iter);
+		else
+			map_len = copy_to_iter(dax.addr, map_len, iter);
+		dax_unmap_atomic(iomap->bdev, &dax);
+		if (map_len <= 0) {
+			ret = map_len ? map_len : -EFAULT;
+			break;
+		}
+
+		pos += map_len;
+		length -= map_len;
+		done += map_len;
+	}
+
+	return done ? done : ret;
+}
+
+/**
+ * iomap_dax_rw - Perform I/O to a DAX file
+ * @iocb:	The control block for this I/O
+ * @iter:	The addresses to do I/O from or to
+ * @ops:	iomap ops passed from the file system
+ *
+ * This function performs read and write operations to directly mapped
+ * persistent memory.  The callers needs to take care of read/write exclusion
+ * and evicting any page cache pages in the region under I/O.
+ */
+ssize_t
+iomap_dax_rw(struct kiocb *iocb, struct iov_iter *iter,
+		struct iomap_ops *ops)
+{
+	struct address_space *mapping = iocb->ki_filp->f_mapping;
+	struct inode *inode = mapping->host;
+	loff_t pos = iocb->ki_pos, ret = 0, done = 0;
+	unsigned flags = 0;
+
+	if (iov_iter_rw(iter) == WRITE)
+		flags |= IOMAP_WRITE;
+
+	/*
+	 * Yes, even DAX files can have page cache attached to them:  A zeroed
+	 * page is inserted into the pagecache when we have to serve a write
+	 * fault on a hole.  It should never be dirtied and can simply be
+	 * dropped from the pagecache once we get real data for the page.
+	 *
+	 * XXX: This is racy against mmap, and there's nothing we can do about
+	 * it. We'll eventually need to shift this down even further so that
+	 * we can check if we allocated blocks over a hole first.
+	 */
+	if (mapping->nrpages) {
+		ret = invalidate_inode_pages2_range(mapping,
+				pos >> PAGE_SHIFT,
+				(pos + iov_iter_count(iter) - 1) >> PAGE_SHIFT);
+		WARN_ON_ONCE(ret);
+	}
+
+	while (iov_iter_count(iter)) {
+		ret = iomap_apply(inode, pos, iov_iter_count(iter), flags, ops,
+				iter, iomap_dax_actor);
+		if (ret <= 0)
+			break;
+		pos += ret;
+		done += ret;
+	}
+
+	iocb->ki_pos += done;
+	return done ? done : ret;
+}
+EXPORT_SYMBOL_GPL(iomap_dax_rw);
+
+/**
+ * iomap_dax_fault - handle a page fault on a DAX file
+ * @vma: The virtual memory area where the fault occurred
+ * @vmf: The description of the fault
+ * @ops: iomap ops passed from the file system
+ *
+ * When a page fault occurs, filesystems may call this helper in their fault
+ * or mkwrite handler for DAX files. Assumes the caller has done all the
+ * necessary locking for the page fault to proceed successfully.
+ */
+int iomap_dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf,
+			struct iomap_ops *ops)
+{
+	struct address_space *mapping = vma->vm_file->f_mapping;
+	struct inode *inode = mapping->host;
+	unsigned long vaddr = (unsigned long)vmf->virtual_address;
+	loff_t pos = (loff_t)vmf->pgoff << PAGE_SHIFT;
+	sector_t sector;
+	struct iomap iomap = { 0 };
+	unsigned flags = 0;
+	int error, major = 0;
+	void *entry;
+
+	/*
+	 * Check whether offset isn't beyond end of file now. Caller is supposed
+	 * to hold locks serializing us with truncate / punch hole so this is
+	 * a reliable test.
+	 */
+	if (pos >= i_size_read(inode))
+		return VM_FAULT_SIGBUS;
+
+	entry = grab_mapping_entry(mapping, vmf->pgoff);
+	if (IS_ERR(entry)) {
+		error = PTR_ERR(entry);
+		goto out;
+	}
+
+	if ((vmf->flags & FAULT_FLAG_WRITE) && !vmf->cow_page)
+		flags |= IOMAP_WRITE;
+
+	/*
+	 * Note that we don't bother to use iomap_apply here: DAX required
+	 * the file system block size to be equal the page size, which means
+	 * that we never have to deal with more than a single extent here.
+	 */
+	error = ops->iomap_begin(inode, pos, PAGE_SIZE, flags, &iomap);
+	if (error)
+		goto unlock_entry;
+	if (WARN_ON_ONCE(iomap.offset + iomap.length < pos + PAGE_SIZE)) {
+		error = -EIO;		/* fs corruption? */
+		goto unlock_entry;
+	}
+
+	sector = iomap.blkno + (((pos & PAGE_MASK) - iomap.offset) >> 9);
+
+	if (vmf->cow_page) {
+		switch (iomap.type) {
+		case IOMAP_HOLE:
+		case IOMAP_UNWRITTEN:
+			clear_user_highpage(vmf->cow_page, vaddr);
+			break;
+		case IOMAP_MAPPED:
+			error = copy_user_dax(iomap.bdev, sector, PAGE_SIZE,
+					vmf->cow_page, vaddr);
+			break;
+		default:
+			WARN_ON_ONCE(1);
+			error = -EIO;
+			break;
+		}
+
+		if (error)
+			goto unlock_entry;
+		if (!radix_tree_exceptional_entry(entry)) {
+			vmf->page = entry;
+			return VM_FAULT_LOCKED;
+		}
+		vmf->entry = entry;
+		return VM_FAULT_DAX_LOCKED;
+	}
+
+	switch (iomap.type) {
+	case IOMAP_MAPPED:
+		if (iomap.flags & IOMAP_F_NEW) {
+			count_vm_event(PGMAJFAULT);
+			mem_cgroup_count_vm_event(vma->vm_mm, PGMAJFAULT);
+			major = VM_FAULT_MAJOR;
+		}
+		error = dax_insert_mapping(mapping, iomap.bdev, sector,
+				PAGE_SIZE, &entry, vma, vmf);
+		break;
+	case IOMAP_UNWRITTEN:
+	case IOMAP_HOLE:
+		if (!(vmf->flags & FAULT_FLAG_WRITE))
+			return dax_load_hole(mapping, entry, vmf);
+		/*FALLTHRU*/
+	default:
+		WARN_ON_ONCE(1);
+		error = -EIO;
+		break;
+	}
+
+ unlock_entry:
+	put_locked_mapping_entry(mapping, vmf->pgoff, entry);
+ out:
+	if (error == -ENOMEM)
+		return VM_FAULT_OOM | major;
+	/* -EBUSY is fine, somebody else faulted on the same PTE */
+	if (error < 0 && error != -EBUSY)
+		return VM_FAULT_SIGBUS | major;
+	return VM_FAULT_NOPAGE | major;
+}
+EXPORT_SYMBOL_GPL(iomap_dax_fault);
+#endif /* CONFIG_FS_IOMAP */