Re: [PATCH RFC v7 02/14] fixup! vhost: option to fetch descriptors through an independent struct
From: Eugenio Pérez <eperezma@redhat.com>
Date: 2020-06-10 14:00:15
Also in:
kvm, lkml, netdev
Subsystem:
the rest, virtio host (vhost) · Maintainers:
Linus Torvalds, "Michael S. Tsirkin", Jason Wang
On Wed, 2020-06-10 at 07:35 -0400, Michael S. Tsirkin wrote:
quoted hunk ↗ jump to hunk
--- drivers/vhost/vhost.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c index 180b7b58c76b..11433d709651 100644 --- a/drivers/vhost/vhost.c +++ b/drivers/vhost/vhost.c@@ -2614,7 +2614,7 @@ int vhost_get_vq_desc_batch(struct vhost_virtqueue *vq, err_fetch: vq->ndescs = 0; - return ret; + return ret ? ret : vq->num; } EXPORT_SYMBOL_GPL(vhost_get_vq_desc_batch);
I'm able to lock the vhost_get_vq_desc running virtio_test with no arguments against this patch. It does not happen if it returns vq->num early, appended below. Let me know if you prefer your conditional at the end of the function and I will investigate the cause. Thanks!
diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
index 28f324fd77df..4d198994e7be 100644@@ -2350,7 +2345,9 @@ int vhost_get_vq_desc(struct vhost_virtqueue *vq, int ret = fetch_descs(vq); int i; - if (ret <= 0) + if (ret == 0) + return vq->num; + else if (ret < 0) goto err; /* Now convert to IOV */