[Virtio-fs] [PATCH 1/5] virtiofsd: skip unnecessary vu_queue_get_avail_bytes()
Liu Bo
bo.liu at linux.alibaba.com
Fri Jul 26 21:35:40 UTC 2019
On Fri, Jul 26, 2019 at 10:10:59AM +0100, Stefan Hajnoczi wrote:
> When debug output is disabled there is no need to calculate the number
> of in/out bytes available.
>
> There is also no need to skip a request if there are 0 out bytes. The
> request parsing code already handles invalid requests.
>
> Signed-off-by: Stefan Hajnoczi <stefanha at redhat.com>
> ---
> contrib/virtiofsd/fuse_virtio.c | 13 +++++--------
> 1 file changed, 5 insertions(+), 8 deletions(-)
>
> diff --git a/contrib/virtiofsd/fuse_virtio.c b/contrib/virtiofsd/fuse_virtio.c
> index 083e4fc317..d543c6d30f 100644
> --- a/contrib/virtiofsd/fuse_virtio.c
> +++ b/contrib/virtiofsd/fuse_virtio.c
> @@ -507,18 +507,16 @@ static void *fv_queue_thread(void *opaque)
> ret = pthread_rwlock_rdlock(&qi->virtio_dev->vu_dispatch_rwlock);
> assert(ret == 0); /* there is no possible error case */
>
> - /* out is from guest, in is too guest */
> - unsigned int in_bytes, out_bytes;
> - vu_queue_get_avail_bytes(dev, q, &in_bytes, &out_bytes, ~0, ~0);
> + if (se->debug) {
> + /* out is from guest, in is too guest */
> + unsigned int in_bytes, out_bytes;
> + vu_queue_get_avail_bytes(dev, q, &in_bytes, &out_bytes, ~0, ~0);
>
> - if (se->debug)
> fuse_debug("%s: Queue %d gave evalue: %zx available: in: %u out: %u\n",
> __func__, qi->qidx, (size_t)evalue, in_bytes,
> out_bytes);
> -
> - if (!out_bytes) {
> - goto next;
> }
> +
> while (1) {
> bool allocated_bufv = false;
> struct fuse_bufvec bufv;
> @@ -708,7 +706,6 @@ static void *fv_queue_thread(void *opaque)
> elem = NULL;
> }
>
> -next:
> pthread_rwlock_unlock(&qi->virtio_dev->vu_dispatch_rwlock);
> }
> pthread_mutex_destroy(&ch.lock);
> --
> 2.21.0
>
> _______________________________________________
> Virtio-fs mailing list
> Virtio-fs at redhat.com
> https://www.redhat.com/mailman/listinfo/virtio-fs
Reviewed-by: Liu Bo <bo.liu at linux.alibaba.com>
thanks,
-liubo
More information about the Virtio-fs
mailing list