Thread (28 messages) 28 messages, 5 authors, 2021-06-10

Re: [PATCH 04/14] mm: Add functions to lock invalidate_lock for two mappings

From: "Darrick J. Wong" <djwong@kernel.org>
Date: 2021-06-07 15:46:25
Also in: ceph-devel, linux-cifs, linux-f2fs-devel, linux-fsdevel, linux-mm, linux-xfs

On Mon, Jun 07, 2021 at 04:52:14PM +0200, Jan Kara wrote:
Some operations such as reflinking blocks among files will need to lock
invalidate_lock for two mappings. Add helper functions to do that.

Signed-off-by: Jan Kara <jack@suse.cz>
Straightforward lift from xfs, though now with vfs lock ordering
rules...

Reviewed-by: Darrick J. Wong <djwong@kernel.org>

--D
quoted hunk ↗ jump to hunk
---
 include/linux/fs.h |  6 ++++++
 mm/filemap.c       | 38 ++++++++++++++++++++++++++++++++++++++
 2 files changed, 44 insertions(+)
diff --git a/include/linux/fs.h b/include/linux/fs.h
index d8afbc9661d7..ddc11bafc183 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -849,6 +849,12 @@ static inline void filemap_invalidate_unlock_shared(
 void lock_two_nondirectories(struct inode *, struct inode*);
 void unlock_two_nondirectories(struct inode *, struct inode*);
 
+void filemap_invalidate_lock_two(struct address_space *mapping1,
+				 struct address_space *mapping2);
+void filemap_invalidate_unlock_two(struct address_space *mapping1,
+				   struct address_space *mapping2);
+
+
 /*
  * NOTE: in a 32bit arch with a preemptable kernel and
  * an UP compile the i_size_read/write must be atomic
diff --git a/mm/filemap.c b/mm/filemap.c
index c8e7e451d81e..b8e9bccecd9f 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -1009,6 +1009,44 @@ struct page *__page_cache_alloc(gfp_t gfp)
 EXPORT_SYMBOL(__page_cache_alloc);
 #endif
 
+/*
+ * filemap_invalidate_lock_two - lock invalidate_lock for two mappings
+ *
+ * Lock exclusively invalidate_lock of any passed mapping that is not NULL.
+ *
+ * @mapping1: the first mapping to lock
+ * @mapping2: the second mapping to lock
+ */
+void filemap_invalidate_lock_two(struct address_space *mapping1,
+				 struct address_space *mapping2)
+{
+	if (mapping1 > mapping2)
+		swap(mapping1, mapping2);
+	if (mapping1)
+		down_write(&mapping1->invalidate_lock);
+	if (mapping2 && mapping1 != mapping2)
+		down_write_nested(&mapping2->invalidate_lock, 1);
+}
+EXPORT_SYMBOL(filemap_invalidate_lock_two);
+
+/*
+ * filemap_invalidate_unlock_two - unlock invalidate_lock for two mappings
+ *
+ * Unlock exclusive invalidate_lock of any passed mapping that is not NULL.
+ *
+ * @mapping1: the first mapping to unlock
+ * @mapping2: the second mapping to unlock
+ */
+void filemap_invalidate_unlock_two(struct address_space *mapping1,
+				   struct address_space *mapping2)
+{
+	if (mapping1)
+		up_write(&mapping1->invalidate_lock);
+	if (mapping2 && mapping1 != mapping2)
+		up_write(&mapping2->invalidate_lock);
+}
+EXPORT_SYMBOL(filemap_invalidate_unlock_two);
+
 /*
  * In order to wait for pages to become available there must be
  * waitqueues associated with pages. By using a hash table of
-- 
2.26.2
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help