Stefano Garzarella
2022-Aug-08 10:36 UTC
[RFC PATCH v3 4/9] vmci/vsock: use 'target' in notify_poll_in callback
On Wed, Aug 03, 2022 at 01:57:54PM +0000, Arseniy Krasnov wrote:>This callback controls setting of POLLIN,POLLRDNORM output bits of poll() >syscall,but in some cases,it is incorrectly to set it, when socket has >at least 1 bytes of available data. Use 'target' which is already exists >and equal to sk_rcvlowat in this case.Ditto as the previous patch. With that fixed: Reviewed-by: Stefano Garzarella <sgarzare at redhat.com> @Bryan, @Vishnu, if you're happy with this change, can you ack/review? Thanks, Stefano> >Signed-off-by: Arseniy Krasnov <AVKrasnov at sberdevices.ru> >--- > net/vmw_vsock/vmci_transport_notify.c | 8 ++++---- > net/vmw_vsock/vmci_transport_notify_qstate.c | 8 ++++---- > 2 files changed, 8 insertions(+), 8 deletions(-) > >diff --git a/net/vmw_vsock/vmci_transport_notify.c b/net/vmw_vsock/vmci_transport_notify.c >index d69fc4b595ad..852097e2b9e6 100644 >--- a/net/vmw_vsock/vmci_transport_notify.c >+++ b/net/vmw_vsock/vmci_transport_notify.c >@@ -340,12 +340,12 @@ vmci_transport_notify_pkt_poll_in(struct sock *sk, > { > struct vsock_sock *vsk = vsock_sk(sk); > >- if (vsock_stream_has_data(vsk)) { >+ if (vsock_stream_has_data(vsk) >= target) { > *data_ready_now = true; > } else { >- /* We can't read right now because there is nothing in the >- * queue. Ask for notifications when there is something to >- * read. >+ /* We can't read right now because there is not enough data >+ * in the queue. Ask for notifications when there is something >+ * to read. > */ > if (sk->sk_state == TCP_ESTABLISHED) { > if (!send_waiting_read(sk, 1)) >diff --git a/net/vmw_vsock/vmci_transport_notify_qstate.c b/net/vmw_vsock/vmci_transport_notify_qstate.c >index 0f36d7c45db3..12f0cb8fe998 100644 >--- a/net/vmw_vsock/vmci_transport_notify_qstate.c >+++ b/net/vmw_vsock/vmci_transport_notify_qstate.c >@@ -161,12 +161,12 @@ vmci_transport_notify_pkt_poll_in(struct sock *sk, > { > struct vsock_sock *vsk = vsock_sk(sk); > >- if (vsock_stream_has_data(vsk)) { >+ if (vsock_stream_has_data(vsk) >= target) { > *data_ready_now = true; > } else { >- /* We can't read right now because there is nothing in the >- * queue. Ask for notifications when there is something to >- * read. >+ /* We can't read right now because there is not enough data >+ * in the queue. Ask for notifications when there is something >+ * to read. > */ > if (sk->sk_state == TCP_ESTABLISHED) > vsock_block_update_write_window(sk); >-- >2.25.1
Vishnu Dasa
2022-Aug-19 04:49 UTC
[RFC PATCH v3 4/9] vmci/vsock: use 'target' in notify_poll_in callback
> On Aug 8, 2022, at 3:36 AM, Stefano Garzarella <sgarzare at redhat.com> wrote: > > On Wed, Aug 03, 2022 at 01:57:54PM +0000, Arseniy Krasnov wrote: >> This callback controls setting of POLLIN,POLLRDNORM output bits of poll() >> syscall,but in some cases,it is incorrectly to set it, when socket has >> at least 1 bytes of available data. Use 'target' which is already exists >> and equal to sk_rcvlowat in this case. > > Ditto as the previous patch. > With that fixed: > > Reviewed-by: Stefano Garzarella <sgarzare at redhat.com> > > @Bryan, @Vishnu, if you're happy with this change, can you ack/review?This patch looks good to me. Thank you, Arseniy for running the test with VMCI. I also ran some of our internal tests successfully with this patch series.> Thanks, > Stefano > >> >> Signed-off-by: Arseniy Krasnov <AVKrasnov at sberdevices.ru>Reviewed-by: Vishnu Dasa <vdasa at vmware.com>>> --- >> net/vmw_vsock/vmci_transport_notify.c | 8 ++++---- >> net/vmw_vsock/vmci_transport_notify_qstate.c | 8 ++++---- >> 2 files changed, 8 insertions(+), 8 deletions(-) >> >> diff --git a/net/vmw_vsock/vmci_transport_notify.c b/net/vmw_vsock/vmci_transport_notify.c >> index d69fc4b595ad..852097e2b9e6 100644 >> --- a/net/vmw_vsock/vmci_transport_notify.c >> +++ b/net/vmw_vsock/vmci_transport_notify.c >> @@ -340,12 +340,12 @@ vmci_transport_notify_pkt_poll_in(struct sock *sk, >> { >> struct vsock_sock *vsk = vsock_sk(sk); >> >> - if (vsock_stream_has_data(vsk)) { >> + if (vsock_stream_has_data(vsk) >= target) { >> *data_ready_now = true; >> } else { >> - /* We can't read right now because there is nothing in the >> - * queue. Ask for notifications when there is something to >> - * read. >> + /* We can't read right now because there is not enough data >> + * in the queue. Ask for notifications when there is something >> + * to read. >> */ >> if (sk->sk_state == TCP_ESTABLISHED) { >> if (!send_waiting_read(sk, 1)) >> diff --git a/net/vmw_vsock/vmci_transport_notify_qstate.c b/net/vmw_vsock/vmci_transport_notify_qstate.c >> index 0f36d7c45db3..12f0cb8fe998 100644 >> --- a/net/vmw_vsock/vmci_transport_notify_qstate.c >> +++ b/net/vmw_vsock/vmci_transport_notify_qstate.c >> @@ -161,12 +161,12 @@ vmci_transport_notify_pkt_poll_in(struct sock *sk, >> { >> struct vsock_sock *vsk = vsock_sk(sk); >> >> - if (vsock_stream_has_data(vsk)) { >> + if (vsock_stream_has_data(vsk) >= target) { >> *data_ready_now = true; >> } else { >> - /* We can't read right now because there is nothing in the >> - * queue. Ask for notifications when there is something to >> - * read. >> + /* We can't read right now because there is not enough data >> + * in the queue. Ask for notifications when there is something >> + * to read. >> */ >> if (sk->sk_state == TCP_ESTABLISHED) >> vsock_block_update_write_window(sk); >> -- >> 2.25.1