dax: integrate *sync code & multi-order radix tree

Commit Message

Ross Zwisler April 14, 2016, 9:03 p.m. UTC

With the new addition of the multi-order radix tree support we can simplify
the DAX *sync PMD support a bit.  Instead of manually checking to see if
our index is covered by a PMD entry we can rely on the new radix tree to
return the PMD entry if present.

Signed-off-by: Ross Zwisler <[e-mail]>
---
 fs/dax.c | 31 +++++++++++--------------------
 1 file changed, 11 insertions(+), 20 deletions(-)

Comments

Ross Zwisler April 15, 2016, 8:10 p.m. UTC | #1

On Thu, Apr 14, 2016 at 03:03:24PM -0600, Ross Zwisler wrote:
> With the new addition of the multi-order radix tree support we can simplify
> the DAX *sync PMD support a bit.  Instead of manually checking to see if
> our index is covered by a PMD entry we can rely on the new radix tree to
> return the PMD entry if present.
> 
> Signed-off-by: Ross Zwisler <[e-mail]>

Sorry, I should have explicitly called out that this patch (indeed, all
multi-order uses of the radix tree) need to be built upon the recent
multi-order fixes from Matthew and me:

https://lkml.org/lkml/2016/4/14/481

I've posted a working tree here:

https://git.kernel.org/cgit/linux/kernel/git/zwisler/linux.git/log/?h=dax_mo_radix

Thanks,
- Ross
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch

diff --git a/fs/dax.c b/fs/dax.c
index 735a608..3f87fcc 100644
--- a/fs/dax.c
+++ b/fs/dax.c
@@ -40,6 +40,7 @@ 
 #define RADIX_DAX_SECTOR(entry) (((unsigned long)entry >> RADIX_DAX_SHIFT))
 #define RADIX_DAX_ENTRY(sector, pmd) ((void *)((unsigned long)sector << \
 		RADIX_DAX_SHIFT | (pmd ? RADIX_DAX_PMD : RADIX_DAX_PTE)))
+#define RADIX_DAX_ORDER(pmd) (pmd ? PMD_SHIFT - PAGE_SHIFT : 0)
 
 static long dax_map_atomic(struct block_device *bdev, struct blk_dax_ctl *dax)
 {
@@ -360,13 +361,11 @@  static int copy_user_bh(struct page *to, struct inode *inode,
 }
 
 #define NO_SECTOR -1
-#define DAX_PMD_INDEX(page_index) (page_index & (PMD_MASK >> PAGE_CACHE_SHIFT))
 
 static int dax_radix_entry(struct address_space *mapping, pgoff_t index,
 		sector_t sector, bool pmd_entry, bool dirty)
 {
 	struct radix_tree_root *page_tree = &mapping->page_tree;
-	pgoff_t pmd_index = DAX_PMD_INDEX(index);
 	int type, error = 0;
 	void *entry;
 
@@ -376,12 +375,6 @@  static int dax_radix_entry(struct address_space *mapping, pgoff_t index,
 
 	spin_lock_irq(&mapping->tree_lock);
 
-	entry = radix_tree_lookup(page_tree, pmd_index);
-	if (entry && RADIX_DAX_TYPE(entry) == RADIX_DAX_PMD) {
-		index = pmd_index;
-		goto dirty;
-	}
-
 	entry = radix_tree_lookup(page_tree, index);
 	if (entry) {
 		type = RADIX_DAX_TYPE(entry);
@@ -418,7 +411,8 @@  static int dax_radix_entry(struct address_space *mapping, pgoff_t index,
 		goto unlock;
 	}
 
-	error = radix_tree_insert(page_tree, index,
+	error = __radix_tree_insert(page_tree, index,
+			RADIX_DAX_ORDER(pmd_entry),
 			RADIX_DAX_ENTRY(sector, pmd_entry));
 	if (error)
 		goto unlock;
@@ -462,6 +456,13 @@  static int dax_writeback_one(struct block_device *bdev,
 		goto unlock;
 	}
 
+	/*
+	 * Even if dax_writeback_mapping_range() was given a wbc->range_start
+	 * in the middle of a PMD, the 'index' we are given will be aligned to
+	 * the start index of the PMD, as will the sector we pull from
+	 * 'entry'.  This allows us to flush for PMD_SIZE and not have to
+	 * worry about partial PMD writebacks.
+	 */
 	dax.sector = RADIX_DAX_SECTOR(entry);
 	dax.size = (type == RADIX_DAX_PMD ? PMD_SIZE : PAGE_SIZE);
 	spin_unlock_irq(&mapping->tree_lock);
@@ -502,12 +503,11 @@  int dax_writeback_mapping_range(struct address_space *mapping,
 		struct block_device *bdev, struct writeback_control *wbc)
 {
 	struct inode *inode = mapping->host;
-	pgoff_t start_index, end_index, pmd_index;
+	pgoff_t start_index, end_index;
 	pgoff_t indices[PAGEVEC_SIZE];
 	struct pagevec pvec;
 	bool done = false;
 	int i, ret = 0;
-	void *entry;
 
 	if (WARN_ON_ONCE(inode->i_blkbits != PAGE_SHIFT))
 		return -EIO;
@@ -517,15 +517,6 @@  int dax_writeback_mapping_range(struct address_space *mapping,
 
 	start_index = wbc->range_start >> PAGE_CACHE_SHIFT;
 	end_index = wbc->range_end >> PAGE_CACHE_SHIFT;
-	pmd_index = DAX_PMD_INDEX(start_index);
-
-	rcu_read_lock();
-	entry = radix_tree_lookup(&mapping->page_tree, pmd_index);
-	rcu_read_unlock();
-
-	/* see if the start of our range is covered by a PMD entry */
-	if (entry && RADIX_DAX_TYPE(entry) == RADIX_DAX_PMD)
-		start_index = pmd_index;
 
 	tag_pages_for_writeback(mapping, start_index, end_index);