On Tue, Jan 12, 2016 at 10:44:51AM +0100, Jan Kara wrote:

On Thu 07-01-16 22:27:52, Ross Zwisler wrote:

quoted

When we get a DAX PMD fault for a write it is possible that there could be
some number of 4k zero pages already present for the same range that were
inserted to service reads from a hole.  These 4k zero pages need to be
unmapped from the VMAs and removed from the struct address_space radix tree
before the real DAX PMD entry can be inserted.

For PTE faults this same use case also exists and is handled by a
combination of unmap_mapping_range() to unmap the VMAs and
delete_from_page_cache() to remove the page from the address_space radix
tree.

For PMD faults we do have a call to unmap_mapping_range() (protected by a
buffer_new() check), but nothing clears out the radix tree entry.  The
buffer_new() check is also incorrect as the current ext4 and XFS filesystem
code will never return a buffer_head with BH_New set, even when allocating
new blocks over a hole.  Instead the filesystem will zero the blocks
manually and return a buffer_head with only BH_Mapped set.

Fix this situation by removing the buffer_new() check and adding a call to
truncate_inode_pages_range() to clear out the radix tree entries before we
insert the DAX PMD.

Signed-off-by: Ross Zwisler <redacted>
Reported-by: Dan Williams <redacted>
Tested-by: Dan Williams <redacted>

Just two nits below. Nothing serious so you can add:

Reviewed-by: Jan Kara <jack@suse.cz>

Cool, thank you for the review!

quoted

---
 fs/dax.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/fs/dax.c b/fs/dax.c
index 513bba5..5b84a46 100644
--- a/fs/dax.c
+++ b/fs/dax.c

@@ -589,6 +589,7 @@ int __dax_pmd_fault(struct vm_area_struct *vma, unsigned long address,
 	bool write = flags & FAULT_FLAG_WRITE;
 	struct block_device *bdev;
 	pgoff_t size, pgoff;
+	loff_t lstart, lend;
 	sector_t block;
 	int result = 0;
 

@@ -643,15 +644,13 @@ int __dax_pmd_fault(struct vm_area_struct *vma, unsigned long address,
 		goto fallback;
 	}
 
-	/*
-	 * If we allocated new storage, make sure no process has any
-	 * zero pages covering this hole
-	 */
-	if (buffer_new(&bh)) {
-		i_mmap_unlock_read(mapping);
-		unmap_mapping_range(mapping, pgoff << PAGE_SHIFT, PMD_SIZE, 0);
-		i_mmap_lock_read(mapping);
-	}
+	/* make sure no process has any zero pages covering this hole */
+	lstart = pgoff << PAGE_SHIFT;
+	lend = lstart + PMD_SIZE - 1; /* inclusive */
+	i_mmap_unlock_read(mapping);

Just a nit but is there reason why we grab i_mmap_lock_read(mapping) only
to release it a few lines below? The bh checks inside the locked region
don't seem to rely on i_mmap_lock...

I think we can probably just take it when we're done with the truncate() -
I'll fix for v9.

quoted

+	unmap_mapping_range(mapping, lstart, PMD_SIZE, 0);
+	truncate_inode_pages_range(mapping, lstart, lend);

These two calls can be shortened as:

truncate_pagecache_range(inode, lstart, lend);

Nice.  I'll change it for v9.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>