Re: [PATCH 1/2] vsock/virtio: Move rx_buf_nr and rx_buf_max_nr initialization position
From: Stefano Garzarella <sgarzare@redhat.com>
Date: 2025-02-10 16:16:13
Also in:
kvm, virtualization
On Fri, Feb 07, 2025 at 01:20:32PM +0800, Junnan Wu wrote:
From: Ying Gao <redacted> In function virtio_vsock_probe, it initializes the variables "rx_buf_nr" and "rx_buf_max_nr", but in function virtio_vsock_restore it doesn't. Move the initizalition position into function virtio_vsock_vqs_start. Once executing s2r twice in a row without
s2r ? suspend 2 ram? Please define the acronym, it was hard for me to understand (the code helped me).
initializing rx_buf_nr and rx_buf_max_nr, the rx_buf_max_nr increased to three times vq->num_free, at this time, in function virtio_transport_rx_work, the conditions to fill rx buffer (rx_buf_nr < rx_buf_max_nr / 2) can't be met.
Please add a Fixes tag, in this case I think it should be:
Fixes: bd50c5dc182b ("vsock/virtio: add support for device suspend/resume")
but please, double check.
Signed-off-by: Ying Gao <redacted> Signed-off-by: Junnan Wu <redacted> --- net/vmw_vsock/virtio_transport.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
I find the commit title/description a bit hard to understand, please take a look at: https://www.kernel.org/doc/html/latest/process/submitting-patches.html#describe-your-changes In this case I'd write something like this: vsock/virtio: initialize rx_buf_nr and rx_buf_max_nr when resuming [Describe the symptom] When executing suspend/resume twice in a row, ... [Describe the issue] `rx_buf_nr` and `rx_buf_max_nr` are initialized only in virtio_vsock_probe(), but they should be reset whenever virtqueues are recreated, like after a suspend/resume. ... [Desribe the fix, what this patch does] Move the `rx_buf_nr` and `rx_buf_max_nr` initialization in virtio_vsock_vqs_init(), so we are sure that they are properly initialized, every time we initialize the virtqueues, either when we load the driver or after a suspend/resume. ...
quoted hunk ↗ jump to hunk
diff --git a/net/vmw_vsock/virtio_transport.c b/net/vmw_vsock/virtio_transport.c index b58c3818f284..9eefd0fba92b 100644 --- a/net/vmw_vsock/virtio_transport.c +++ b/net/vmw_vsock/virtio_transport.c@@ -688,6 +688,8 @@ static void virtio_vsock_vqs_start(struct virtio_vsock *vsock)
I think it is better to move the initialization of those fields in virtio_vsock_vqs_init().
quoted hunk ↗ jump to hunk
mutex_unlock(&vsock->tx_lock); mutex_lock(&vsock->rx_lock); + vsock->rx_buf_nr = 0; + vsock->rx_buf_max_nr = 0; virtio_vsock_rx_fill(vsock); vsock->rx_run = true; mutex_unlock(&vsock->rx_lock);@@ -779,8 +781,6 @@ static int virtio_vsock_probe(struct virtio_device *vdev)vsock->vdev = vdev; - vsock->rx_buf_nr = 0; - vsock->rx_buf_max_nr = 0; atomic_set(&vsock->queued_replies, 0);
Should we also move `queued_replies` ? Thanks, Stefano
mutex_init(&vsock->tx_lock); -- 2.34.1