| Message ID | 20220817135718.2553-4-qtxuning1999@sjtu.edu.cn (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | In order support for virtio_ring, vhost and vsock. | expand |
On Wed, Aug 17, 2022 at 3:58 PM Guo Zhi <qtxuning1999@sjtu.edu.cn> wrote: > > Vsock uses buffers in order, and for tx driver doesn't have to > know the length of the buffer. So we can do a batch for vsock if > in order negotiated, only write one used ring for a batch of buffers > > Signed-off-by: Guo Zhi <qtxuning1999@sjtu.edu.cn> > --- > drivers/vhost/vsock.c | 9 ++++++++- > 1 file changed, 8 insertions(+), 1 deletion(-) > > diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c > index 368330417bde..b0108009c39a 100644 > --- a/drivers/vhost/vsock.c > +++ b/drivers/vhost/vsock.c > @@ -500,6 +500,7 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work *work) > int head, pkts = 0, total_len = 0; > unsigned int out, in; > bool added = false; > + int last_head = -1; > > mutex_lock(&vq->mutex); > > @@ -551,10 +552,16 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work *work) > else > virtio_transport_free_pkt(pkt); > > - vhost_add_used(vq, head, 0); > + if (!vhost_has_feature(vq, VIRTIO_F_IN_ORDER)) > + vhost_add_used(vq, head, 0); > + else > + last_head = head; > added = true; > } while(likely(!vhost_exceeds_weight(vq, ++pkts, total_len))); > > + /* If in order feature negotiaged, we can do a batch to increase performance */ > + if (vhost_has_feature(vq, VIRTIO_F_IN_ORDER) && last_head != -1) > + vhost_add_used(vq, last_head, 0); Expanding my previous mail on patch 1, you can also use this in vsock tx queue code. This way, no modifications to vhost.c functions are needed. Thanks! > no_more_replies: > if (added) > vhost_signal(&vsock->dev, vq); > -- > 2.17.1 >
----- Original Message ----- > From: "eperezma" <eperezma@redhat.com> > To: "Guo Zhi" <qtxuning1999@sjtu.edu.cn> > Cc: "jasowang" <jasowang@redhat.com>, "sgarzare" <sgarzare@redhat.com>, "Michael Tsirkin" <mst@redhat.com>, "netdev" > <netdev@vger.kernel.org>, "linux-kernel" <linux-kernel@vger.kernel.org>, "kvm list" <kvm@vger.kernel.org>, > "virtualization" <virtualization@lists.linux-foundation.org> > Sent: Thursday, August 18, 2022 2:19:29 PM > Subject: Re: [RFC v2 3/7] vsock: batch buffers in tx > On Wed, Aug 17, 2022 at 3:58 PM Guo Zhi <qtxuning1999@sjtu.edu.cn> wrote: >> >> Vsock uses buffers in order, and for tx driver doesn't have to >> know the length of the buffer. So we can do a batch for vsock if >> in order negotiated, only write one used ring for a batch of buffers >> >> Signed-off-by: Guo Zhi <qtxuning1999@sjtu.edu.cn> >> --- >> drivers/vhost/vsock.c | 9 ++++++++- >> 1 file changed, 8 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c >> index 368330417bde..b0108009c39a 100644 >> --- a/drivers/vhost/vsock.c >> +++ b/drivers/vhost/vsock.c >> @@ -500,6 +500,7 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work >> *work) >> int head, pkts = 0, total_len = 0; >> unsigned int out, in; >> bool added = false; >> + int last_head = -1; >> >> mutex_lock(&vq->mutex); >> >> @@ -551,10 +552,16 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work >> *work) >> else >> virtio_transport_free_pkt(pkt); >> >> - vhost_add_used(vq, head, 0); >> + if (!vhost_has_feature(vq, VIRTIO_F_IN_ORDER)) >> + vhost_add_used(vq, head, 0); >> + else >> + last_head = head; >> added = true; >> } while(likely(!vhost_exceeds_weight(vq, ++pkts, total_len))); >> >> + /* If in order feature negotiaged, we can do a batch to increase >> performance */ >> + if (vhost_has_feature(vq, VIRTIO_F_IN_ORDER) && last_head != -1) >> + vhost_add_used(vq, last_head, 0); > > Expanding my previous mail on patch 1, you can also use this in vsock > tx queue code. This way, no modifications to vhost.c functions are > needed. > > Thanks! As replied in patch 1, no modification to vhost is not feasible. Thanks! > >> no_more_replies: >> if (added) >> vhost_signal(&vsock->dev, vq); >> -- >> 2.17.1 >>
在 2022/8/17 21:57, Guo Zhi 写道: > Vsock uses buffers in order, and for tx driver doesn't have to > know the length of the buffer. So we can do a batch for vsock if > in order negotiated, only write one used ring for a batch of buffers > > Signed-off-by: Guo Zhi <qtxuning1999@sjtu.edu.cn> > --- > drivers/vhost/vsock.c | 9 ++++++++- > 1 file changed, 8 insertions(+), 1 deletion(-) > > diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c > index 368330417bde..b0108009c39a 100644 > --- a/drivers/vhost/vsock.c > +++ b/drivers/vhost/vsock.c > @@ -500,6 +500,7 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work *work) > int head, pkts = 0, total_len = 0; > unsigned int out, in; > bool added = false; > + int last_head = -1; > > mutex_lock(&vq->mutex); > > @@ -551,10 +552,16 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work *work) > else > virtio_transport_free_pkt(pkt); > > - vhost_add_used(vq, head, 0); > + if (!vhost_has_feature(vq, VIRTIO_F_IN_ORDER)) > + vhost_add_used(vq, head, 0); > + else > + last_head = head; > added = true; > } while(likely(!vhost_exceeds_weight(vq, ++pkts, total_len))); > > + /* If in order feature negotiaged, we can do a batch to increase performance */ > + if (vhost_has_feature(vq, VIRTIO_F_IN_ORDER) && last_head != -1) > + vhost_add_used(vq, last_head, 0); I may miss something but spec said "The device then skips forward in the ring according to the size of the batch. ". I don't see how it is done here. Thanks > no_more_replies: > if (added) > vhost_signal(&vsock->dev, vq);
----- Original Message ----- > From: "jasowang" <jasowang@redhat.com> > To: "Guo Zhi" <qtxuning1999@sjtu.edu.cn>, "eperezma" <eperezma@redhat.com>, "sgarzare" <sgarzare@redhat.com>, "Michael > Tsirkin" <mst@redhat.com> > Cc: "netdev" <netdev@vger.kernel.org>, "linux-kernel" <linux-kernel@vger.kernel.org>, "kvm list" <kvm@vger.kernel.org>, > "virtualization" <virtualization@lists.linux-foundation.org> > Sent: Thursday, August 25, 2022 3:08:58 PM > Subject: Re: [RFC v2 3/7] vsock: batch buffers in tx > 在 2022/8/17 21:57, Guo Zhi 写道: >> Vsock uses buffers in order, and for tx driver doesn't have to >> know the length of the buffer. So we can do a batch for vsock if >> in order negotiated, only write one used ring for a batch of buffers >> >> Signed-off-by: Guo Zhi <qtxuning1999@sjtu.edu.cn> >> --- >> drivers/vhost/vsock.c | 9 ++++++++- >> 1 file changed, 8 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c >> index 368330417bde..b0108009c39a 100644 >> --- a/drivers/vhost/vsock.c >> +++ b/drivers/vhost/vsock.c >> @@ -500,6 +500,7 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work >> *work) >> int head, pkts = 0, total_len = 0; >> unsigned int out, in; >> bool added = false; >> + int last_head = -1; >> >> mutex_lock(&vq->mutex); >> >> @@ -551,10 +552,16 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work >> *work) >> else >> virtio_transport_free_pkt(pkt); >> >> - vhost_add_used(vq, head, 0); >> + if (!vhost_has_feature(vq, VIRTIO_F_IN_ORDER)) >> + vhost_add_used(vq, head, 0); >> + else >> + last_head = head; >> added = true; >> } while(likely(!vhost_exceeds_weight(vq, ++pkts, total_len))); >> >> + /* If in order feature negotiaged, we can do a batch to increase performance >> */ >> + if (vhost_has_feature(vq, VIRTIO_F_IN_ORDER) && last_head != -1) >> + vhost_add_used(vq, last_head, 0); > > > I may miss something but spec said "The device then skips forward in the > ring according to the size of the batch. ". > > I don't see how it is done here. > > Thanks > It can skip them in __vhost_add_used_n if _F_IN_ORDER is negotiated. last_used_idx will be added by size of the batch. > >> no_more_replies: >> if (added) >> vhost_signal(&vsock->dev, vq);
diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c index 368330417bde..b0108009c39a 100644 --- a/drivers/vhost/vsock.c +++ b/drivers/vhost/vsock.c @@ -500,6 +500,7 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work *work) int head, pkts = 0, total_len = 0; unsigned int out, in; bool added = false; + int last_head = -1; mutex_lock(&vq->mutex); @@ -551,10 +552,16 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work *work) else virtio_transport_free_pkt(pkt); - vhost_add_used(vq, head, 0); + if (!vhost_has_feature(vq, VIRTIO_F_IN_ORDER)) + vhost_add_used(vq, head, 0); + else + last_head = head; added = true; } while(likely(!vhost_exceeds_weight(vq, ++pkts, total_len))); + /* If in order feature negotiaged, we can do a batch to increase performance */ + if (vhost_has_feature(vq, VIRTIO_F_IN_ORDER) && last_head != -1) + vhost_add_used(vq, last_head, 0); no_more_replies: if (added) vhost_signal(&vsock->dev, vq);
Vsock uses buffers in order, and for tx driver doesn't have to know the length of the buffer. So we can do a batch for vsock if in order negotiated, only write one used ring for a batch of buffers Signed-off-by: Guo Zhi <qtxuning1999@sjtu.edu.cn> --- drivers/vhost/vsock.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)