On Tue, Mar 1, 2011 at 6:38 AM, Aneesh Kumar K.V
<aneesh.ku...@linux.vnet.ibm.com> wrote:
> Signed-off-by: Aneesh Kumar K.V <aneesh.ku...@linux.vnet.ibm.com>
> ---
> hw/9pfs/virtio-9p.c | 31 +++++++++++++++++++++++++++++++
> hw/9pfs/virtio-9p.h | 2 ++
> 2 files changed, 33 insertions(+), 0 deletions(-)
>
> diff --git a/hw/9pfs/virtio-9p.c b/hw/9pfs/virtio-9p.c
> index c4b0198..882f4f3 100644
> --- a/hw/9pfs/virtio-9p.c
> +++ b/hw/9pfs/virtio-9p.c
> @@ -1978,6 +1978,36 @@ static void v9fs_fsync(V9fsState *s, V9fsPDU *pdu)
> v9fs_post_do_fsync(s, pdu, err);
> }
>
> +static void v9fs_post_do_syncfs(V9fsState *s, V9fsPDU *pdu, int err)
> +{
> + if (err == -1) {
> + err = -errno;
> + }
> + complete_pdu(s, pdu, err);
> +}
> +
> +static void v9fs_syncfs(V9fsState *s, V9fsPDU *pdu)
> +{
> + int err;
> + int32_t fid;
> + size_t offset = 7;
> + V9fsFidState *fidp;
> +
> + pdu_unmarshal(pdu, offset, "d", &fid);
> + fidp = lookup_fid(s, fid);
> + if (fidp == NULL) {
> + err = -ENOENT;
> + v9fs_post_do_syncfs(s, pdu, err);
> + return;
> + }
> + /*
> + * We don't have per file system syncfs
> + * So just return success
> + */
> + err = 0;
> + v9fs_post_do_syncfs(s, pdu, err);
> +}
Please explain the semantics of P9_TSYNCFS. Won't returning success
without doing anything lead to data integrity issues?
It seems unnecessary to split v9fs_post_do_syncfs() into its own
function since there is no blocking point here and a callback will not
be needed.
Stefan
