From: Xiubo Li <xiubli@xxxxxxxxxx>
When hit a file hole, will keep the header.assert_ver as 0, and
in MDS side it will check it to decide whether should it do a
RMW.
And always set the header.block_size to CEPH_FSCRYPT_BLOCK_SIZE,
because even in the hole case, the MDS will need to use this to
do the filer.truncate().
Signed-off-by: Xiubo Li <xiubli@xxxxxxxxxx>
---
Hi Jeff,
Please squash this patch to the previous "ceph: add truncate size handling support for fscrypt" commit in ceph-client/wip-fscrypt-size branch.
And also please sync the ceph PR, I have updated it too.
fs/ceph/inode.c | 14 ++++++++++++--
1 file changed, 12 insertions(+), 2 deletions(-)
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index 53b8e2ff3678..b4f7a4b4f15c 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -2312,6 +2312,12 @@ static int fill_fscrypt_truncate(struct inode *inode,
header.ver = 1;
header.compat = 1;
+ /*
+ * Always set the block_size to CEPH_FSCRYPT_BLOCK_SIZE,
+ * because in MDS it may need this to do the truncate.
+ */
+ header.block_size = cpu_to_le32(CEPH_FSCRYPT_BLOCK_SIZE);
+
/*
* If we hit a hole here, we should just skip filling
* the fscrypt for the request, because once the fscrypt
@@ -2327,15 +2333,19 @@ static int fill_fscrypt_truncate(struct inode *inode,
pos, i_size);
header.data_len = cpu_to_le32(8 + 8 + 4);
+
+ /*
+ * If the "assert_ver" is 0 means hitting a hole, and
+ * the MDS will use the it to check whether hitting a
+ * hole or not.
+ */
header.assert_ver = 0;
header.file_offset = 0;
- header.block_size = 0;
ret = 0;
} else {
header.data_len = cpu_to_le32(8 + 8 + 4 + CEPH_FSCRYPT_BLOCK_SIZE);
header.assert_ver = cpu_to_le64(objvers.objvers[0].objver);
header.file_offset = cpu_to_le64(orig_pos);
- header.block_size = cpu_to_le32(CEPH_FSCRYPT_BLOCK_SIZE);
/* truncate and zero out the extra contents for the last block */
memset(iov.iov_base + boff, 0, PAGE_SIZE - boff);
--
2.27.0
