Merge branch 'iomap-4.9-dax' into for-next

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 */

fs/ext2/Kconfig

 config EXT2_FS
 	tristate "Second extended fs support"
+	select FS_IOMAP if FS_DAX
 	help
 	  Ext2 is a standard Linux file system for hard disks.
 

fs/ext2/ext2.h

@@ -814,6 +814,7 @@ extern const struct file_operations ext2_file_operations;
 /* inode.c */
 extern const struct address_space_operations ext2_aops;
 extern const struct address_space_operations ext2_nobh_aops;
+extern struct iomap_ops ext2_iomap_ops;
 
 /* namei.c */
 extern const struct inode_operations ext2_dir_inode_operations;

fs/ext2/file.c

@@ -22,11 +22,59 @@
 #include <linux/pagemap.h>
 #include <linux/dax.h>
 #include <linux/quotaops.h>
+#include <linux/iomap.h>
+#include <linux/uio.h>
 #include "ext2.h"
 #include "xattr.h"
 #include "acl.h"
 
 #ifdef CONFIG_FS_DAX
+static ssize_t ext2_dax_read_iter(struct kiocb *iocb, struct iov_iter *to)
+{
+	struct inode *inode = iocb->ki_filp->f_mapping->host;
+	ssize_t ret;
+
+	if (!iov_iter_count(to))
+		return 0; /* skip atime */
+
+	inode_lock_shared(inode);
+	ret = iomap_dax_rw(iocb, to, &ext2_iomap_ops);
+	inode_unlock_shared(inode);
+
+	file_accessed(iocb->ki_filp);
+	return ret;
+}
+
+static ssize_t ext2_dax_write_iter(struct kiocb *iocb, struct iov_iter *from)
+{
+	struct file *file = iocb->ki_filp;
+	struct inode *inode = file->f_mapping->host;
+	ssize_t ret;
+
+	inode_lock(inode);
+	ret = generic_write_checks(iocb, from);
+	if (ret <= 0)
+		goto out_unlock;
+	ret = file_remove_privs(file);
+	if (ret)
+		goto out_unlock;
+	ret = file_update_time(file);
+	if (ret)
+		goto out_unlock;
+
+	ret = iomap_dax_rw(iocb, from, &ext2_iomap_ops);
+	if (ret > 0 && iocb->ki_pos > i_size_read(inode)) {
+		i_size_write(inode, iocb->ki_pos);
+		mark_inode_dirty(inode);
+	}
+
+out_unlock:
+	inode_unlock(inode);
+	if (ret > 0)
+		ret = generic_write_sync(iocb, ret);
+	return ret;
+}
+
 /*
  * The lock ordering for ext2 DAX fault paths is:
  *
@@ -51,7 +99,7 @@ static int ext2_dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	}
 	down_read(&ei->dax_sem);
 
-	ret = dax_fault(vma, vmf, ext2_get_block);
+	ret = iomap_dax_fault(vma, vmf, &ext2_iomap_ops);
 
 	up_read(&ei->dax_sem);
 	if (vmf->flags & FAULT_FLAG_WRITE)
@@ -156,14 +204,28 @@ int ext2_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 	return ret;
 }
 
-/*
- * We have mostly NULL's here: the current defaults are ok for
- * the ext2 filesystem.
- */
+static ssize_t ext2_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
+{
+#ifdef CONFIG_FS_DAX
+	if (IS_DAX(iocb->ki_filp->f_mapping->host))
+		return ext2_dax_read_iter(iocb, to);
+#endif
+	return generic_file_read_iter(iocb, to);
+}
+
+static ssize_t ext2_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
+{
+#ifdef CONFIG_FS_DAX
+	if (IS_DAX(iocb->ki_filp->f_mapping->host))
+		return ext2_dax_write_iter(iocb, from);
+#endif
+	return generic_file_write_iter(iocb, from);
+}
+
 const struct file_operations ext2_file_operations = {
 	.llseek		= generic_file_llseek,
-	.read_iter	= generic_file_read_iter,
-	.write_iter	= generic_file_write_iter,
+	.read_iter	= ext2_file_read_iter,
+	.write_iter	= ext2_file_write_iter,
 	.unlocked_ioctl = ext2_ioctl,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl	= ext2_compat_ioctl,

fs/ext2/inode.c

@@ -32,6 +32,7 @@
 #include <linux/buffer_head.h>
 #include <linux/mpage.h>
 #include <linux/fiemap.h>
+#include <linux/iomap.h>
 #include <linux/namei.h>
 #include <linux/uio.h>
 #include "ext2.h"
@@ -618,7 +619,7 @@ static void ext2_splice_branch(struct inode *inode,
  */
 static int ext2_get_blocks(struct inode *inode,
 			   sector_t iblock, unsigned long maxblocks,
-			   struct buffer_head *bh_result,
+			   u32 *bno, bool *new, bool *boundary,
 			   int create)
 {
 	int err = -EIO;
@@ -644,7 +645,6 @@ static int ext2_get_blocks(struct inode *inode,
 	/* Simplest case - block found, no allocation needed */
 	if (!partial) {
 		first_block = le32_to_cpu(chain[depth - 1].key);
-		clear_buffer_new(bh_result); /* What's this do? */
 		count++;
 		/*map more blocks*/
 		while (count < maxblocks && count <= blocks_to_boundary) {
@@ -699,7 +699,6 @@ static int ext2_get_blocks(struct inode *inode,
 			mutex_unlock(&ei->truncate_mutex);
 			if (err)
 				goto cleanup;
-			clear_buffer_new(bh_result);
 			goto got_it;
 		}
 	}
@@ -745,15 +744,16 @@ static int ext2_get_blocks(struct inode *inode,
 			mutex_unlock(&ei->truncate_mutex);
 			goto cleanup;
 		}
-	} else
-		set_buffer_new(bh_result);
+	} else {
+		*new = true;
+	}
 
 	ext2_splice_branch(inode, iblock, partial, indirect_blks, count);
 	mutex_unlock(&ei->truncate_mutex);
 got_it:
-	map_bh(bh_result, inode->i_sb, le32_to_cpu(chain[depth-1].key));
+	*bno = le32_to_cpu(chain[depth-1].key);
 	if (count > blocks_to_boundary)
-		set_buffer_boundary(bh_result);
+		*boundary = true;
 	err = count;
 	/* Clean up and exit */
 	partial = chain + depth - 1;	/* the whole chain */
@@ -765,19 +765,82 @@ static int ext2_get_blocks(struct inode *inode,
 	return err;
 }
 
-int ext2_get_block(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create)
+int ext2_get_block(struct inode *inode, sector_t iblock,
+		struct buffer_head *bh_result, int create)
 {
 	unsigned max_blocks = bh_result->b_size >> inode->i_blkbits;
-	int ret = ext2_get_blocks(inode, iblock, max_blocks,
-			      bh_result, create);
-	if (ret > 0) {
+	bool new = false, boundary = false;
+	u32 bno;
+	int ret;
+
+	ret = ext2_get_blocks(inode, iblock, max_blocks, &bno, &new, &boundary,
+			create);
+	if (ret <= 0)
+		return ret;
+
+	map_bh(bh_result, inode->i_sb, bno);
 	bh_result->b_size = (ret << inode->i_blkbits);
-		ret = 0;
-	}
+	if (new)
+		set_buffer_new(bh_result);
+	if (boundary)
+		set_buffer_boundary(bh_result);
+	return 0;
+
+}
+
+#ifdef CONFIG_FS_DAX
+static int ext2_iomap_begin(struct inode *inode, loff_t offset, loff_t length,
+		unsigned flags, struct iomap *iomap)
+{
+	unsigned int blkbits = inode->i_blkbits;
+	unsigned long first_block = offset >> blkbits;
+	unsigned long max_blocks = (length + (1 << blkbits) - 1) >> blkbits;
+	bool new = false, boundary = false;
+	u32 bno;
+	int ret;
+
+	ret = ext2_get_blocks(inode, first_block, max_blocks,
+			&bno, &new, &boundary, flags & IOMAP_WRITE);
+	if (ret < 0)
 		return ret;
 
+	iomap->flags = 0;
+	iomap->bdev = inode->i_sb->s_bdev;
+	iomap->offset = (u64)first_block << blkbits;
+
+	if (ret == 0) {
+		iomap->type = IOMAP_HOLE;
+		iomap->blkno = IOMAP_NULL_BLOCK;
+		iomap->length = 1 << blkbits;
+	} else {
+		iomap->type = IOMAP_MAPPED;
+		iomap->blkno = (sector_t)bno << (blkbits - 9);
+		iomap->length = (u64)ret << blkbits;
+		iomap->flags |= IOMAP_F_MERGED;
+	}
+
+	if (new)
+		iomap->flags |= IOMAP_F_NEW;
+	return 0;
 }
 
+static int
+ext2_iomap_end(struct inode *inode, loff_t offset, loff_t length,
+		ssize_t written, unsigned flags, struct iomap *iomap)
+{
+	if (iomap->type == IOMAP_MAPPED &&
+	    written < length &&
+	    (flags & IOMAP_WRITE))
+		ext2_write_failed(inode->i_mapping, offset + length);
+	return 0;
+}
+
+struct iomap_ops ext2_iomap_ops = {
+	.iomap_begin		= ext2_iomap_begin,
+	.iomap_end		= ext2_iomap_end,
+};
+#endif /* CONFIG_FS_DAX */
+
 int ext2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		u64 start, u64 len)
 {
@@ -863,10 +926,9 @@ ext2_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 	loff_t offset = iocb->ki_pos;
 	ssize_t ret;
 
-	if (IS_DAX(inode))
-		ret = dax_do_io(iocb, inode, iter, ext2_get_block, NULL,
-				DIO_LOCKING);
-	else
+	if (WARN_ON_ONCE(IS_DAX(inode)))
+		return -EIO;
+
 	ret = blockdev_direct_IO(iocb, inode, iter, ext2_get_block);
 	if (ret < 0 && iov_iter_rw(iter) == WRITE)
 		ext2_write_failed(mapping, offset + count);

fs/internal.h

@@ -12,6 +12,7 @@
 struct super_block;
 struct file_system_type;
 struct iomap;
+struct iomap_ops;
 struct linux_binprm;
 struct path;
 struct mount;
@@ -164,3 +165,13 @@ extern struct dentry_operations ns_dentry_operations;
 extern int do_vfs_ioctl(struct file *file, unsigned int fd, unsigned int cmd,
 		    unsigned long arg);
 extern long vfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
+
+/*
+ * iomap support:
+ */
+typedef loff_t (*iomap_actor_t)(struct inode *inode, loff_t pos, loff_t len,
+		void *data, struct iomap *iomap);
+
+loff_t iomap_apply(struct inode *inode, loff_t pos, loff_t length,
+		unsigned flags, struct iomap_ops *ops, void *data,
+		iomap_actor_t actor);

fs/iomap.c

@@ -27,9 +27,6 @@
 #include <linux/dax.h>
 #include "internal.h"
 
-typedef loff_t (*iomap_actor_t)(struct inode *inode, loff_t pos, loff_t len,
-		void *data, struct iomap *iomap);
-
 /*
  * Execute a iomap write on a segment of the mapping that spans a
  * contiguous range of pages that have identical block mapping state.
@@ -41,7 +38,7 @@ typedef loff_t (*iomap_actor_t)(struct inode *inode, loff_t pos, loff_t len,
  * resources they require in the iomap_begin call, and release them in the
  * iomap_end call.
  */
-static loff_t
+loff_t
 iomap_apply(struct inode *inode, loff_t pos, loff_t length, unsigned flags,
 		struct iomap_ops *ops, void *data, iomap_actor_t actor)
 {

fs/xfs/xfs_aops.c

@@ -200,7 +200,7 @@ xfs_setfilesize_trans_alloc(
  * Update on-disk file size now that data has been written to disk.
  */
 STATIC int
-xfs_setfilesize(
+__xfs_setfilesize(
 	struct xfs_inode	*ip,
 	struct xfs_trans	*tp,
 	xfs_off_t		offset,
@@ -225,6 +225,23 @@ xfs_setfilesize(
 	return xfs_trans_commit(tp);
 }
 
+int
+xfs_setfilesize(
+	struct xfs_inode	*ip,
+	xfs_off_t		offset,
+	size_t			size)
+{
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_trans	*tp;
+	int			error;
+
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_fsyncts, 0, 0, 0, &tp);
+	if (error)
+		return error;
+
+	return __xfs_setfilesize(ip, tp, offset, size);
+}
+
 STATIC int
 xfs_setfilesize_ioend(
 	struct xfs_ioend	*ioend,
@@ -247,7 +264,7 @@ xfs_setfilesize_ioend(
 		return error;
 	}
 
-	return xfs_setfilesize(ip, tp, ioend->io_offset, ioend->io_size);
+	return __xfs_setfilesize(ip, tp, ioend->io_offset, ioend->io_size);
 }
 
 /*
@@ -1336,13 +1353,12 @@ xfs_end_io_direct_write(
 {
 	struct inode		*inode = file_inode(iocb->ki_filp);
 	struct xfs_inode	*ip = XFS_I(inode);
-	struct xfs_mount	*mp = ip->i_mount;
 	uintptr_t		flags = (uintptr_t)private;
 	int			error = 0;
 
 	trace_xfs_end_io_direct_write(ip, offset, size);
 
-	if (XFS_FORCED_SHUTDOWN(mp))
+	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
 		return -EIO;
 
 	if (size <= 0)
@@ -1380,14 +1396,9 @@ xfs_end_io_direct_write(
 
 		error = xfs_iomap_write_unwritten(ip, offset, size);
 	} else if (flags & XFS_DIO_FLAG_APPEND) {
-		struct xfs_trans *tp;
-
 		trace_xfs_end_io_direct_write_append(ip, offset, size);
 
-		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_fsyncts, 0, 0, 0,
-				&tp);
-		if (!error)
-			error = xfs_setfilesize(ip, tp, offset, size);
+		error = xfs_setfilesize(ip, offset, size);
 	}
 
 	return error;

fs/xfs/xfs_aops.h

@@ -62,6 +62,7 @@ int	xfs_get_blocks_dax_fault(struct inode *inode, sector_t offset,
 
 int	xfs_end_io_direct_write(struct kiocb *iocb, loff_t offset,
 		ssize_t size, void *private);
+int	xfs_setfilesize(struct xfs_inode *ip, xfs_off_t offset, size_t size);
 
 extern void xfs_count_page_state(struct page *, int *, int *);
 extern struct block_device *xfs_find_bdev_for_inode(struct inode *);

fs/xfs/xfs_file.c

@@ -333,10 +333,7 @@ xfs_file_dax_read(
 	struct kiocb		*iocb,
 	struct iov_iter		*to)
 {
-	struct address_space	*mapping = iocb->ki_filp->f_mapping;
-	struct inode		*inode = mapping->host;
-	struct xfs_inode	*ip = XFS_I(inode);
-	struct iov_iter		data = *to;
+	struct xfs_inode	*ip = XFS_I(iocb->ki_filp->f_mapping->host);
 	size_t			count = iov_iter_count(to);
 	ssize_t			ret = 0;
 
@@ -346,11 +343,7 @@ xfs_file_dax_read(
 		return 0; /* skip atime */
 
 	xfs_rw_ilock(ip, XFS_IOLOCK_SHARED);
-	ret = dax_do_io(iocb, inode, &data, xfs_get_blocks_direct, NULL, 0);
-	if (ret > 0) {
-		iocb->ki_pos += ret;
-		iov_iter_advance(to, ret);
-	}
+	ret = iomap_dax_rw(iocb, to, &xfs_iomap_ops);
 	xfs_rw_iunlock(ip, XFS_IOLOCK_SHARED);
 
 	file_accessed(iocb->ki_filp);
@@ -712,70 +705,32 @@ xfs_file_dax_write(
 	struct kiocb		*iocb,
 	struct iov_iter		*from)
 {
-	struct address_space	*mapping = iocb->ki_filp->f_mapping;
-	struct inode		*inode = mapping->host;
+	struct inode		*inode = iocb->ki_filp->f_mapping->host;
 	struct xfs_inode	*ip = XFS_I(inode);
-	struct xfs_mount	*mp = ip->i_mount;
-	ssize_t			ret = 0;
-	int			unaligned_io = 0;
-	int			iolock;
-	struct iov_iter		data;
+	int			iolock = XFS_IOLOCK_EXCL;
+	ssize_t			ret, error = 0;
+	size_t			count;
+	loff_t			pos;
 
-	/* "unaligned" here means not aligned to a filesystem block */
-	if ((iocb->ki_pos & mp->m_blockmask) ||
-	    ((iocb->ki_pos + iov_iter_count(from)) & mp->m_blockmask)) {
-		unaligned_io = 1;
-		iolock = XFS_IOLOCK_EXCL;
-	} else if (mapping->nrpages) {
-		iolock = XFS_IOLOCK_EXCL;
-	} else {
-		iolock = XFS_IOLOCK_SHARED;
-	}
 	xfs_rw_ilock(ip, iolock);
-
 	ret = xfs_file_aio_write_checks(iocb, from, &iolock);
 	if (ret)
 		goto out;
 
-	/*
-	 * 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. dax_do_io() should really do this invalidation internally as
-	 * it will know if we've allocated over a holei for this specific IO and
-	 * if so it needs to update the mapping tree and invalidate existing
-	 * PTEs over the newly allocated range. Remove this invalidation when
-	 * dax_do_io() is fixed up.
-	 */
-	if (mapping->nrpages) {
-		loff_t end = iocb->ki_pos + iov_iter_count(from) - 1;
+	pos = iocb->ki_pos;
+	count = iov_iter_count(from);
 
-		ret = invalidate_inode_pages2_range(mapping,
-						    iocb->ki_pos >> PAGE_SHIFT,
-						    end >> PAGE_SHIFT);
-		WARN_ON_ONCE(ret);
-	}
+	trace_xfs_file_dax_write(ip, count, pos);
 
-	if (iolock == XFS_IOLOCK_EXCL && !unaligned_io) {
-		xfs_rw_ilock_demote(ip, XFS_IOLOCK_EXCL);
-		iolock = XFS_IOLOCK_SHARED;
+	ret = iomap_dax_rw(iocb, from, &xfs_iomap_ops);
+	if (ret > 0 && iocb->ki_pos > i_size_read(inode)) {
+		i_size_write(inode, iocb->ki_pos);
+		error = xfs_setfilesize(ip, pos, ret);
 	}
 
-	trace_xfs_file_dax_write(ip, iov_iter_count(from), iocb->ki_pos);
-
-	data = *from;
-	ret = dax_do_io(iocb, inode, &data, xfs_get_blocks_direct,
-			xfs_end_io_direct_write, 0);
-	if (ret > 0) {
-		iocb->ki_pos += ret;
-		iov_iter_advance(from, ret);
-	}
 out:
 	xfs_rw_iunlock(ip, iolock);
-	return ret;
+	return error ? error : ret;
 }
 
 STATIC ssize_t
@@ -1514,7 +1469,7 @@ xfs_filemap_page_mkwrite(
 	xfs_ilock(XFS_I(inode), XFS_MMAPLOCK_SHARED);
 
 	if (IS_DAX(inode)) {
-		ret = dax_mkwrite(vma, vmf, xfs_get_blocks_dax_fault);
+		ret = iomap_dax_fault(vma, vmf, &xfs_iomap_ops);
 	} else {
 		ret = iomap_page_mkwrite(vma, vmf, &xfs_iomap_ops);
 		ret = block_page_mkwrite_return(ret);
@@ -1548,7 +1503,7 @@ xfs_filemap_fault(
 		 * changes to xfs_get_blocks_direct() to map unwritten extent
 		 * ioend for conversion on read-only mappings.
 		 */
-		ret = dax_fault(vma, vmf, xfs_get_blocks_dax_fault);
+		ret = iomap_dax_fault(vma, vmf, &xfs_iomap_ops);
 	} else
 		ret = filemap_fault(vma, vmf);
 	xfs_iunlock(XFS_I(inode), XFS_MMAPLOCK_SHARED);

fs/xfs/xfs_iomap.c

@@ -934,11 +934,13 @@ xfs_iomap_write_unwritten(
 	return error;
 }
 
-static inline bool imap_needs_alloc(struct xfs_bmbt_irec *imap, int nimaps)
+static inline bool imap_needs_alloc(struct inode *inode,
+		struct xfs_bmbt_irec *imap, int nimaps)
 {
 	return !nimaps ||
 		imap->br_startblock == HOLESTARTBLOCK ||
-		imap->br_startblock == DELAYSTARTBLOCK;
+		imap->br_startblock == DELAYSTARTBLOCK ||
+		(IS_DAX(inode) && ISUNWRITTEN(imap));
 }
 
 static int
@@ -954,16 +956,18 @@ xfs_file_iomap_begin(
 	struct xfs_bmbt_irec	imap;
 	xfs_fileoff_t		offset_fsb, end_fsb;
 	int			nimaps = 1, error = 0;
+	unsigned		lockmode;
 
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return -EIO;
 
-	if ((flags & IOMAP_WRITE) && !xfs_get_extsz_hint(ip)) {
+	if ((flags & IOMAP_WRITE) &&
+	    !IS_DAX(inode) && !xfs_get_extsz_hint(ip)) {
 		return xfs_file_iomap_begin_delay(inode, offset, length, flags,
 				iomap);
 	}
 
-	xfs_ilock(ip, XFS_ILOCK_EXCL);
+	lockmode = xfs_ilock_data_map_shared(ip);
 
 	ASSERT(offset <= mp->m_super->s_maxbytes);
 	if ((xfs_fsize_t)offset + length > mp->m_super->s_maxbytes)
@@ -974,11 +978,11 @@ xfs_file_iomap_begin(
 	error = xfs_bmapi_read(ip, offset_fsb, end_fsb - offset_fsb, &imap,
 			       &nimaps, XFS_BMAPI_ENTIRE);
 	if (error) {
-		xfs_iunlock(ip, XFS_ILOCK_EXCL);
+		xfs_iunlock(ip, lockmode);
 		return error;
 	}
 
-	if ((flags & IOMAP_WRITE) && imap_needs_alloc(&imap, nimaps)) {
+	if ((flags & IOMAP_WRITE) && imap_needs_alloc(inode, &imap, nimaps)) {
 		/*
 		 * We cap the maximum length we map here to MAX_WRITEBACK_PAGES
 		 * pages to keep the chunks of work done where somewhat symmetric
@@ -994,17 +998,19 @@ xfs_file_iomap_begin(
 		 * xfs_iomap_write_direct() expects the shared lock. It
 		 * is unlocked on return.
 		 */
-		xfs_ilock_demote(ip, XFS_ILOCK_EXCL);
+		if (lockmode == XFS_ILOCK_EXCL)
+			xfs_ilock_demote(ip, lockmode);
 		error = xfs_iomap_write_direct(ip, offset, length, &imap,
 				nimaps);
 		if (error)
 			return error;
 
+		iomap->flags = IOMAP_F_NEW;
 		trace_xfs_iomap_alloc(ip, offset, length, 0, &imap);
 	} else {
 		ASSERT(nimaps);
 
-		xfs_iunlock(ip, XFS_ILOCK_EXCL);
+		xfs_iunlock(ip, lockmode);
 		trace_xfs_iomap_found(ip, offset, length, 0, &imap);
 	}
 

include/linux/dax.h

@@ -6,13 +6,19 @@
 #include <linux/radix-tree.h>
 #include <asm/pgtable.h>
 
+struct iomap_ops;
+
 /* We use lowest available exceptional entry bit for locking */
 #define RADIX_DAX_ENTRY_LOCK (1 << RADIX_TREE_EXCEPTIONAL_SHIFT)
 
+ssize_t iomap_dax_rw(struct kiocb *iocb, struct iov_iter *iter,
+		struct iomap_ops *ops);
 ssize_t dax_do_io(struct kiocb *, struct inode *, struct iov_iter *,
 		  get_block_t, dio_iodone_t, int flags);
 int dax_zero_page_range(struct inode *, loff_t from, unsigned len, get_block_t);
 int dax_truncate_page(struct inode *, loff_t from, get_block_t);
+int iomap_dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf,
+			struct iomap_ops *ops);
 int dax_fault(struct vm_area_struct *, struct vm_fault *, get_block_t);
 int dax_delete_mapping_entry(struct address_space *mapping, pgoff_t index);
 void dax_wake_mapping_entry_waiter(struct address_space *mapping,

include/linux/iomap.h

@@ -23,6 +23,7 @@ struct vm_fault;
  */
 #define IOMAP_F_MERGED	0x01	/* contains multiple blocks/extents */
 #define IOMAP_F_SHARED	0x02	/* block shared with another file */
+#define IOMAP_F_NEW	0x04	/* blocks have been newly allocated */
 
 /*
  * Magic value for blkno: