Thread (28 messages) 28 messages, 7 authors, 2021-02-25

[PATCH 2/7] fsdax: Introduce dax_copy_edges() for CoW

flat view
STALE2057d

From: Shiyang Ruan <hidden>
Date: 2021-02-07 17:11:01
Also in: linux-btrfs, linux-fsdevel, lkml, nvdimm, ocfs2-devel
Subsystem: filesystem direct access (dax), filesystems (vfs and infrastructure), the rest · Maintainers: Dan Williams, Alexander Viro, Christian Brauner, Linus Torvalds

dax_copy_edges() is a helper functions performs a copy from one part of
the device to another for data not page aligned.

Signed-off-by: Goldwyn Rodrigues <redacted>
Signed-off-by: Shiyang Ruan <redacted>
---
 fs/dax.c | 41 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 41 insertions(+)
diff --git a/fs/dax.c b/fs/dax.c
index b012b2db7ba2..ea4e8a434900 100644
--- a/fs/dax.c
+++ b/fs/dax.c
@@ -1038,6 +1038,47 @@ static int dax_iomap_direct_access(struct iomap *iomap, loff_t pos, size_t size,
 	return rc;
 }
 
+/*
+ * Copies the part of the pages not included in the write, but required for CoW
+ * because offset/offset+length are not page aligned.
+ */
+static int dax_copy_edges(loff_t pos, loff_t length, struct iomap *srcmap,
+			  void *daddr, bool pmd)
+{
+	size_t page_size = pmd ? PMD_SIZE : PAGE_SIZE;
+	loff_t offset = pos & (page_size - 1);
+	size_t size = ALIGN(offset + length, page_size);
+	loff_t end = pos + length;
+	loff_t pg_end = round_up(end, page_size);
+	void *saddr = 0;
+	int ret = 0;
+
+	ret = dax_iomap_direct_access(srcmap, pos, size, &saddr, NULL);
+	if (ret)
+		return ret;
+	/*
+	 * Copy the first part of the page
+	 * Note: we pass offset as length
+	 */
+	if (offset) {
+		if (saddr)
+			ret = copy_mc_to_kernel(daddr, saddr, offset);
+		else
+			memset(daddr, 0, offset);
+	}
+
+	/* Copy the last part of the range */
+	if (end < pg_end) {
+		if (saddr)
+			ret = copy_mc_to_kernel(daddr + offset + length,
+			       saddr + offset + length,	pg_end - end);
+		else
+			memset(daddr + offset + length, 0, pg_end - end);
+	}
+
+	return ret;
+}
+
 /*
  * The user has performed a load from a hole in the file.  Allocating a new
  * page in the file would cause excessive storage usage for workloads with
-- 
2.30.0


Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help