2021-08-10 13:44:27

by Arseny Krasnov

Subject: [RFC PATCH v2 2/5] vhost/vsock: support MSG_EOR bit processing

It works in the same way as 'end-of-message' bit: if packet has
'EOM' bit, also check for 'EOR' bit.

Signed-off-by: Arseny Krasnov <[email protected]>
---
drivers/vhost/vsock.c | 12 ++++++++----
1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
index feaf650affbe..06fc132b13c8 100644
--- a/drivers/vhost/vsock.c
+++ b/drivers/vhost/vsock.c
@@ -114,7 +114,7 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
size_t nbytes;
size_t iov_len, payload_len;
int head;
- bool restore_flag = false;
+ uint32_t flags_to_restore = 0;

spin_lock_bh(&vsock->send_pkt_list_lock);
if (list_empty(&vsock->send_pkt_list)) {
@@ -187,7 +187,12 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
*/
if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOM) {
pkt->hdr.flags &= ~cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
- restore_flag = true;
+ flags_to_restore |= le32_to_cpu(VIRTIO_VSOCK_SEQ_EOM);
+
+ if (le32_to_cpu(pkt->hdr.flags & VIRTIO_VSOCK_SEQ_EOR)) {
+ pkt->hdr.flags &= ~cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
+ flags_to_restore |= le32_to_cpu(VIRTIO_VSOCK_SEQ_EOR);
+ }
}
}

@@ -224,8 +229,7 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
* to send it with the next available buffer.
*/
if (pkt->off < pkt->len) {
- if (restore_flag)
- pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
+ pkt->hdr.flags |= cpu_to_le32(flags_to_restore);

/* We are queueing the same virtio_vsock_pkt to handle
* the remaining bytes, and we want to deliver it
--
2.25.1

2021-08-11 09:08:19

by Stefano Garzarella

Subject: Re: [RFC PATCH v2 2/5] vhost/vsock: support MSG_EOR bit processing

On Tue, Aug 10, 2021 at 02:40:15PM +0300, Arseny Krasnov wrote:
>It works in the same way as 'end-of-message' bit: if packet has
>'EOM' bit, also check for 'EOR' bit.

Please describe all changes, e.g. the new variable to accumulate flags
to restore.

>
>Signed-off-by: Arseny Krasnov <[email protected]>
>---
> drivers/vhost/vsock.c | 12 ++++++++----
> 1 file changed, 8 insertions(+), 4 deletions(-)
>
>diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
>index feaf650affbe..06fc132b13c8 100644
>--- a/drivers/vhost/vsock.c
>+++ b/drivers/vhost/vsock.c
>@@ -114,7 +114,7 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
> size_t nbytes;
> size_t iov_len, payload_len;
> int head;
>- bool restore_flag = false;
>+ uint32_t flags_to_restore = 0;
>
> spin_lock_bh(&vsock->send_pkt_list_lock);
> if (list_empty(&vsock->send_pkt_list)) {
>@@ -187,7 +187,12 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
> */
> if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOM) {
> pkt->hdr.flags &= ~cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
>- restore_flag = true;
>+ flags_to_restore |= le32_to_cpu(VIRTIO_VSOCK_SEQ_EOM);
>+
>+ if (le32_to_cpu(pkt->hdr.flags & VIRTIO_VSOCK_SEQ_EOR)) {
>+ pkt->hdr.flags &= ~cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
>+ flags_to_restore |= le32_to_cpu(VIRTIO_VSOCK_SEQ_EOR);
^
I'm not sure this is needed, VIRTIO_VSOCK_SEQ_EOR is represented in the
cpu endianess.

I think here you can simpy do `flags_to_restore |= VIRTIO_VSOCK_SEQ_EOR`
then use `pkt->hdr.flags |= cpu_to_le32(flags_to_restore);` as you
already do.

>+ }
> }
> }
>
>@@ -224,8 +229,7 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
> * to send it with the next available buffer.
> */
> if (pkt->off < pkt->len) {
>- if (restore_flag)
>- pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
>+ pkt->hdr.flags |= cpu_to_le32(flags_to_restore);
>
> /* We are queueing the same virtio_vsock_pkt to handle
> * the remaining bytes, and we want to deliver it
>--
>2.25.1
>