On Mon, 2022-04-25 at 15:54 -0400, Jeff Layton wrote:
> Currently when we create a file, we spin up an xattr buffer to send
> along with the create request. If we end up doing an async create
> however, then we currently pass down a zero-length xattr buffer.
>
> Fix the code to send down the xattr buffer in req->r_pagelist. If the
> xattrs span more than a page, however give up and don't try to do an
> async create.
>
> Fixes: 9a8d03ca2e2c ("ceph: attempt to do async create when possible")
> URL: https://bugzilla.redhat.com/show_bug.cgi?id=2063929
> Reported-by: John Fortin <fortinj66@xxxxxxxxx>
> Reported-by: Sri Ramanujam <sri@xxxxxxxxxxxx>
> Signed-off-by: Jeff Layton <jlayton@xxxxxxxxxx>
> ---
> fs/ceph/file.c | 16 +++++++++++++---
> 1 file changed, 13 insertions(+), 3 deletions(-)
>
> diff --git a/fs/ceph/file.c b/fs/ceph/file.c
> index 6c9e837aa1d3..8c8226c0feac 100644
> --- a/fs/ceph/file.c
> +++ b/fs/ceph/file.c
> @@ -629,9 +629,15 @@ static int ceph_finish_async_create(struct inode *dir, struct dentry *dentry,
> iinfo.change_attr = 1;
> ceph_encode_timespec64(&iinfo.btime, &now);
>
> - iinfo.xattr_len = ARRAY_SIZE(xattr_buf);
> - iinfo.xattr_data = xattr_buf;
> - memset(iinfo.xattr_data, 0, iinfo.xattr_len);
> + if (req->r_pagelist) {
> + iinfo.xattr_len = req->r_pagelist->length;
> + iinfo.xattr_data = req->r_pagelist->mapped_tail;
> + } else {
> + /* fake it */
> + iinfo.xattr_len = ARRAY_SIZE(xattr_buf);
> + iinfo.xattr_data = xattr_buf;
> + memset(iinfo.xattr_data, 0, iinfo.xattr_len);
> + }
>
> in.ino = cpu_to_le64(vino.ino);
> in.snapid = cpu_to_le64(CEPH_NOSNAP);
> @@ -743,6 +749,10 @@ int ceph_atomic_open(struct inode *dir, struct dentry *dentry,
> err = ceph_security_init_secctx(dentry, mode, &as_ctx);
> if (err < 0)
> goto out_ctx;
> + /* Async create can't handle more than a page of xattrs */
> + if (as_ctx.pagelist &&
> + !list_is_singular(&as_ctx.pagelist->head))
> + try_async = false;
> } else if (!d_in_lookup(dentry)) {
> /* If it's not being looked up, it's negative */
> return -ENOENT;
Oh, I meant to mark this for stable as well. Xiubo, can you do that when
you merge it?
Thanks,
--
Jeff Layton <jlayton@xxxxxxxxxx>
