From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx1.redhat.com (mx1.redhat.com [209.132.183.28]) by dpdk.org (Postfix) with ESMTP id 08C6C1B81C for ; Mon, 17 Dec 2018 17:10:04 +0100 (CET) Received: from smtp.corp.redhat.com (int-mx08.intmail.prod.int.phx2.redhat.com [10.5.11.23]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 1A47280502; Mon, 17 Dec 2018 16:10:03 +0000 (UTC) Received: from [10.36.112.10] (unknown [10.36.112.10]) by smtp.corp.redhat.com (Postfix) with ESMTPS id 4988B3DA7; Mon, 17 Dec 2018 16:10:01 +0000 (UTC) To: Jens Freimann , dev@dpdk.org Cc: tiwei.bie@intel.com, Gavin.Hu@arm.com References: <20181214155916.1142-1-jfreimann@redhat.com> <20181214155916.1142-3-jfreimann@redhat.com> From: Maxime Coquelin Message-ID: Date: Mon, 17 Dec 2018 17:09:59 +0100 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.3.1 MIME-Version: 1.0 In-Reply-To: <20181214155916.1142-3-jfreimann@redhat.com> Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.84 on 10.5.11.23 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.27]); Mon, 17 Dec 2018 16:10:03 +0000 (UTC) Subject: Re: [dpdk-dev] [PATCH v13 02/10] net/virtio: add packed virtqueue helpers X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 17 Dec 2018 16:10:04 -0000 On 12/14/18 4:59 PM, Jens Freimann wrote: > Add helper functions to set/clear and check descriptor flags. > > Signed-off-by: Jens Freimann > --- > drivers/net/virtio/virtio_pci.h | 6 +++ > drivers/net/virtio/virtqueue.h | 91 ++++++++++++++++++++++++++++++++- > 2 files changed, 95 insertions(+), 2 deletions(-) > > diff --git a/drivers/net/virtio/virtio_pci.h b/drivers/net/virtio/virtio_pci.h > index 4c975a531..b22b62dad 100644 > --- a/drivers/net/virtio/virtio_pci.h > +++ b/drivers/net/virtio/virtio_pci.h > @@ -315,6 +315,12 @@ vtpci_with_feature(struct virtio_hw *hw, uint64_t bit) > return (hw->guest_features & (1ULL << bit)) != 0; > } > > +static inline int > +vtpci_packed_queue(struct virtio_hw *hw) > +{ > + return vtpci_with_feature(hw, VIRTIO_F_RING_PACKED); > +} > + > /* > * Function declaration from virtio_pci.c > */ > diff --git a/drivers/net/virtio/virtqueue.h b/drivers/net/virtio/virtqueue.h > index d4e0858e4..19fabae0b 100644 > --- a/drivers/net/virtio/virtqueue.h > +++ b/drivers/net/virtio/virtqueue.h > @@ -251,6 +251,44 @@ struct virtio_tx_region { > __attribute__((__aligned__(16))); > }; > > +static inline void > +_set_desc_avail(struct vring_packed_desc *desc, int wrap_counter) > +{ > + desc->flags |= VRING_DESC_F_AVAIL(wrap_counter) | > + VRING_DESC_F_USED(!wrap_counter); > +} > + > +static inline void > +set_desc_avail(struct virtqueue *vq, struct vring_packed_desc *desc) > +{ > + _set_desc_avail(desc, vq->avail_wrap_counter); > +} > + > +static inline int > +desc_is_used(struct vring_packed_desc *desc, struct virtqueue *vq) > +{ > + uint16_t used, avail, flags; > + > + flags = desc->flags; > + used = !!(flags & VRING_DESC_F_USED(1)); > + avail = !!(flags & VRING_DESC_F_AVAIL(1)); > + > + return avail == used && used == vq->used_wrap_counter; > +} > + > + > +static inline void > +vring_desc_init_packed(struct virtqueue *vq, int n) > +{ > + int i; > + for (i = 0; i < n - 1; i++) { > + vq->ring_packed.desc_packed[i].id = i; > + vq->vq_descx[i].next = i + 1; > + } > + vq->ring_packed.desc_packed[i].id = i; > + vq->vq_descx[i].next = VQ_RING_DESC_CHAIN_END; > +} > + > /* Chain all the descriptors in the ring with an END */ > static inline void > vring_desc_init(struct vring_desc *dp, uint16_t n) > @@ -262,13 +300,59 @@ vring_desc_init(struct vring_desc *dp, uint16_t n) > dp[i].next = VQ_RING_DESC_CHAIN_END; > } > > +/** > + * Tell the backend not to interrupt us. > + */ > +static inline void > +virtqueue_disable_intr_packed(struct virtqueue *vq) > +{ > + uint16_t *event_flags = &vq->ring_packed.driver_event->desc_event_flags; > + > + if (*event_flags != RING_EVENT_FLAGS_DISABLE) > + *event_flags = RING_EVENT_FLAGS_DISABLE; Why not just setting the value unconditionally? IIUC, it is not called in the hot path? > +} > + > + > /** > * Tell the backend not to interrupt us. > */ > static inline void > virtqueue_disable_intr(struct virtqueue *vq) > { > - vq->vq_ring.avail->flags |= VRING_AVAIL_F_NO_INTERRUPT; > + if (vtpci_packed_queue(vq->hw)) > + virtqueue_disable_intr_packed(vq); > + else > + vq->vq_ring.avail->flags |= VRING_AVAIL_F_NO_INTERRUPT; > +} > + > +/** > + * Tell the backend to interrupt. Implementation for packed virtqueues. > + */ > +static inline void > +virtqueue_enable_intr_packed(struct virtqueue *vq) > +{ > + uint16_t *off_wrap = &vq->ring_packed.driver_event->desc_event_off_wrap; > + uint16_t *event_flags = &vq->ring_packed.driver_event->desc_event_flags; > + > + *off_wrap = vq->vq_used_cons_idx | > + ((uint16_t)(vq->used_wrap_counter << 15)); Don't we want to set off_wrap only if VIRTIO_RING_F_EVENT_IDX negotiated? > + > + if (vq->event_flags_shadow == RING_EVENT_FLAGS_DISABLE) { > + virtio_wmb(); > + vq->event_flags_shadow = > + vtpci_with_feature(vq->hw, VIRTIO_RING_F_EVENT_IDX) ? > + RING_EVENT_FLAGS_DESC : RING_EVENT_FLAGS_ENABLE; > + *event_flags = vq->event_flags_shadow; > + } > +} > + > +/** > + * Tell the backend to interrupt. Implementation for split virtqueues. > + */ > +static inline void > +virtqueue_enable_intr_split(struct virtqueue *vq) > +{ > + vq->vq_ring.avail->flags &= (~VRING_AVAIL_F_NO_INTERRUPT); > } > > /** > @@ -277,7 +361,10 @@ virtqueue_disable_intr(struct virtqueue *vq) > static inline void > virtqueue_enable_intr(struct virtqueue *vq) > { > - vq->vq_ring.avail->flags &= (~VRING_AVAIL_F_NO_INTERRUPT); > + if (vtpci_packed_queue(vq->hw)) > + virtqueue_enable_intr_packed(vq); > + else > + virtqueue_enable_intr_split(vq); > } > > /** >