[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[UNIKRAFT PATCH RFCv4 30/35] plat/virtio_9p: Remove excessive check for max virtqueue size



For qemu virtio mmio read with VIRTIO_MMIO_QUEUE_NUM_MAX, it will
unconditionally return VIRTQUEUE_MAX_SIZE(1024) unlike virtio pci.
But it can set the number of queue size when seting up the virtqueue.
See vm_setup_vq which will be fixed in coming patch.

So the excessive check in virtio 9p is not necessary.

Signed-off-by: Jia He <justin.he@xxxxxxx>
---
 plat/drivers/virtio/virtio_9p.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/plat/drivers/virtio/virtio_9p.c b/plat/drivers/virtio/virtio_9p.c
index 045ce42..aead7a2 100644
--- a/plat/drivers/virtio/virtio_9p.c
+++ b/plat/drivers/virtio/virtio_9p.c
@@ -307,17 +307,15 @@ static int virtio_9p_vq_alloc(struct virtio_9p_device *d)
 
        d->hwvq_id = 0;
        if (unlikely(qdesc_size != NUM_SEGMENTS)) {
-               uk_pr_err(DRIVER_NAME": Expected %d descriptors, found %d 
(virtqueue %"
+               uk_pr_info(DRIVER_NAME": Expected %d descriptors, found %d 
(virtqueue %"
                          PRIu16")\n", NUM_SEGMENTS, qdesc_size, d->hwvq_id);
-               rc = -EINVAL;
-               goto exit;
        }
 
        uk_sglist_init(&d->sg, ARRAY_SIZE(d->sgsegs), &d->sgsegs[0]);
 
        d->vq = virtio_vqueue_setup(d->vdev,
                                    d->hwvq_id,
-                                   qdesc_size,
+                                   NUM_SEGMENTS,
                                    virtio_9p_recv,
                                    a);
        if (unlikely(PTRISERR(d->vq))) {
-- 
2.17.1




 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.