public inbox for linux-s390@vger.kernel.org
 help / color / mirror / Atom feed
From: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
To: Jason Wang <jasowang@redhat.com>
Cc: virtualization@lists.linux.dev,
	"Richard Weinberger" <richard@nod.at>,
	"Anton Ivanov" <anton.ivanov@cambridgegreys.com>,
	"Johannes Berg" <johannes@sipsolutions.net>,
	"Michael S. Tsirkin" <mst@redhat.com>,
	"David S. Miller" <davem@davemloft.net>,
	"Eric Dumazet" <edumazet@google.com>,
	"Jakub Kicinski" <kuba@kernel.org>,
	"Paolo Abeni" <pabeni@redhat.com>,
	"Hans de Goede" <hdegoede@redhat.com>,
	"Ilpo Järvinen" <ilpo.jarvinen@linux.intel.com>,
	"Vadim Pasternak" <vadimp@nvidia.com>,
	"Bjorn Andersson" <andersson@kernel.org>,
	"Mathieu Poirier" <mathieu.poirier@linaro.org>,
	"Cornelia Huck" <cohuck@redhat.com>,
	"Halil Pasic" <pasic@linux.ibm.com>,
	"Eric Farman" <farman@linux.ibm.com>,
	"Heiko Carstens" <hca@linux.ibm.com>,
	"Vasily Gorbik" <gor@linux.ibm.com>,
	"Alexander Gordeev" <agordeev@linux.ibm.com>,
	"Christian Borntraeger" <borntraeger@linux.ibm.com>,
	"Sven Schnelle" <svens@linux.ibm.com>,
	"Alexei Starovoitov" <ast@kernel.org>,
	"Daniel Borkmann" <daniel@iogearbox.net>,
	"Jesper Dangaard Brouer" <hawk@kernel.org>,
	"John Fastabend" <john.fastabend@gmail.com>,
	"Benjamin Berg" <benjamin.berg@intel.com>,
	"Yang Li" <yang.lee@linux.alibaba.com>,
	linux-um@lists.infradead.org, netdev@vger.kernel.org,
	platform-driver-x86@vger.kernel.org,
	linux-remoteproc@vger.kernel.org, linux-s390@vger.kernel.org,
	kvm@vger.kernel.org, bpf@vger.kernel.org
Subject: Re: [PATCH vhost 04/17] virtio_ring: split: remove double check of the unmap ops
Date: Thu, 1 Feb 2024 14:43:46 +0800	[thread overview]
Message-ID: <1706769826.0586398-4-xuanzhuo@linux.alibaba.com> (raw)
In-Reply-To: <CACGkMEs-wUa_z_tGYEwBf7EVJAtuJdkX4HAdjqMXHEM1ys-gKQ@mail.gmail.com>

On Wed, 31 Jan 2024 17:12:22 +0800, Jason Wang <jasowang@redhat.com> wrote:
> On Tue, Jan 30, 2024 at 7:42 PM Xuan Zhuo <xuanzhuo@linux.alibaba.com> wrote:
> >
> > In the functions vring_unmap_one_split and
> > vring_unmap_one_split_indirect,
> > multiple checks are made whether unmap is performed and whether it is
> > INDIRECT.
> >
> > These two functions are usually called in a loop, and we should put the
> > check outside the loop.
> >
> > And we unmap the descs with VRING_DESC_F_INDIRECT on the same path with
> > other descs, that make the thing more complex. If we distinguish the
> > descs with VRING_DESC_F_INDIRECT before unmap, thing will be clearer.
> >
> > 1. only one desc of the desc table is used, we do not need the loop
> > 2. the called unmap api is difference from the other desc
> > 3. the vq->premapped is not needed to check
> > 4. the vq->indirect is not needed to check
> > 5. the state->indir_desc must not be null
> >
> > Signed-off-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
> > ---
> >  drivers/virtio/virtio_ring.c | 80 ++++++++++++++++++------------------
> >  1 file changed, 39 insertions(+), 41 deletions(-)
> >
> > diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c
> > index dd03bc5a81fe..2b41fdbce975 100644
> > --- a/drivers/virtio/virtio_ring.c
> > +++ b/drivers/virtio/virtio_ring.c
> > @@ -452,9 +452,6 @@ static void vring_unmap_one_split_indirect(const struct vring_virtqueue *vq,
> >  {
> >         u16 flags;
> >
> > -       if (!vring_need_unmap_buffer(vq))
> > -               return;
> > -
> >         flags = virtio16_to_cpu(vq->vq.vdev, desc->flags);
> >
> >         dma_unmap_page(vring_dma_dev(vq),
> > @@ -472,27 +469,12 @@ static unsigned int vring_unmap_one_split(const struct vring_virtqueue *vq,
> >
> >         flags = extra[i].flags;
> >
> > -       if (flags & VRING_DESC_F_INDIRECT) {
> > -               if (!vq->use_dma_api)
> > -                       goto out;
> > -
> > -               dma_unmap_single(vring_dma_dev(vq),
> > -                                extra[i].addr,
> > -                                extra[i].len,
> > -                                (flags & VRING_DESC_F_WRITE) ?
> > -                                DMA_FROM_DEVICE : DMA_TO_DEVICE);
> > -       } else {
> > -               if (!vring_need_unmap_buffer(vq))
> > -                       goto out;
> > -
> > -               dma_unmap_page(vring_dma_dev(vq),
> > -                              extra[i].addr,
> > -                              extra[i].len,
> > -                              (flags & VRING_DESC_F_WRITE) ?
> > -                              DMA_FROM_DEVICE : DMA_TO_DEVICE);
> > -       }
> > +       dma_unmap_page(vring_dma_dev(vq),
> > +                      extra[i].addr,
> > +                      extra[i].len,
> > +                      (flags & VRING_DESC_F_WRITE) ?
> > +                      DMA_FROM_DEVICE : DMA_TO_DEVICE);
> >
> > -out:
> >         return extra[i].next;
> >  }
> >
> > @@ -660,7 +642,7 @@ static inline int virtqueue_add_split(struct virtqueue *_vq,
> >                         vq, desc, total_sg * sizeof(struct vring_desc),
> >                         DMA_TO_DEVICE);
> >                 if (vring_mapping_error(vq, addr)) {
> > -                       if (vq->premapped)
> > +                       if (!vring_need_unmap_buffer(vq))
> >                                 goto free_indirect;
> >
> >                         goto unmap_release;
> > @@ -713,6 +695,9 @@ static inline int virtqueue_add_split(struct virtqueue *_vq,
> >         return 0;
> >
> >  unmap_release:
> > +
> > +       WARN_ON(!vring_need_unmap_buffer(vq));
> > +
> >         err_idx = i;
> >
> >         if (indirect)
> > @@ -774,34 +759,42 @@ static void detach_buf_split(struct vring_virtqueue *vq, unsigned int head,
> >  {
> >         unsigned int i, j;
> >         __virtio16 nextflag = cpu_to_virtio16(vq->vq.vdev, VRING_DESC_F_NEXT);
> > +       u16 flags;
> >
> >         /* Clear data ptr. */
> >         vq->split.desc_state[head].data = NULL;
> > +       flags = vq->split.desc_extra[head].flags;
> >
> >         /* Put back on free list: unmap first-level descriptors and find end */
> >         i = head;
> >
> > -       while (vq->split.vring.desc[i].flags & nextflag) {
> > -               vring_unmap_one_split(vq, i);
> > -               i = vq->split.desc_extra[i].next;
> > -               vq->vq.num_free++;
> > -       }
> > -
> > -       vring_unmap_one_split(vq, i);
> > -       vq->split.desc_extra[i].next = vq->free_head;
> > -       vq->free_head = head;
> > +       if (!(flags & VRING_DESC_F_INDIRECT)) {
>
> So during add we do:
>
>         if (!indirect && vring_need_unmap_buffer(vq))
>                 vq->split.desc_extra[prev & (vq->split.vring.num - 1)].flags &=
>                         ~VRING_DESC_F_NEXT;


This does not affect this patch.

1. this just considers the VRING_DESC_F_NEXT of desc_extra.flags
2. the desc_extra.flags is updated by virtqueue_add_desc_split()

So for desc_extra.flags & VRING_DESC_F_INDIRECT, that is right.

Thanks.


>
> Then using flags here unconditionally seems not reliable.
>
> I post a patch to store flags unconditionally at:
>
> https://lore.kernel.org/all/20220224122655-mutt-send-email-mst@kernel.org/
>
> > +               while (vq->split.vring.desc[i].flags & nextflag) {
> > +                       if (vring_need_unmap_buffer(vq))
> > +                               vring_unmap_one_split(vq, i);
> > +                       i = vq->split.desc_extra[i].next;
> > +                       vq->vq.num_free++;
> > +               }
> >
> > -       /* Plus final descriptor */
> > -       vq->vq.num_free++;
> > +               if (vring_need_unmap_buffer(vq))
> > +                       vring_unmap_one_split(vq, i);
> >
> > -       if (vq->indirect) {
> > +               if (ctx)
> > +                       *ctx = vq->split.desc_state[head].indir_desc;
> > +       } else {
> >                 struct vring_desc *indir_desc =
> >                                 vq->split.desc_state[head].indir_desc;
> >                 u32 len;
> >
> > -               /* Free the indirect table, if any, now that it's unmapped. */
> > -               if (!indir_desc)
> > -                       return;
> > +               if (vq->use_dma_api) {
> > +                       struct vring_desc_extra *extra = vq->split.desc_extra;
> > +
> > +                       dma_unmap_single(vring_dma_dev(vq),
> > +                                        extra[i].addr,
> > +                                        extra[i].len,
> > +                                        (flags & VRING_DESC_F_WRITE) ?
> > +                                        DMA_FROM_DEVICE : DMA_TO_DEVICE);
> > +               }
>
> Note that there's a following
>
> BUG_ON(!(vq->split.desc_extra[head].flags &
>                                 VRING_DESC_F_INDIRECT));
>
> Which I think we can remove.
>
> Thanks
>

  parent reply	other threads:[~2024-02-01  6:48 UTC|newest]

Thread overview: 48+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-01-30 11:42 [PATCH vhost 00/17] virtio: drivers maintain dma info for premapped vq Xuan Zhuo
2024-01-30 11:42 ` [PATCH vhost 01/17] virtio_ring: introduce vring_need_unmap_buffer Xuan Zhuo
2024-01-31  9:12   ` Jason Wang
2024-01-30 11:42 ` [PATCH vhost 02/17] virtio_ring: packed: remove double check of the unmap ops Xuan Zhuo
2024-01-31  9:12   ` Jason Wang
2024-01-31  9:58     ` Xuan Zhuo
2024-01-30 11:42 ` [PATCH vhost 03/17] virtio_ring: packed: structure the indirect desc table Xuan Zhuo
2024-01-31  9:12   ` Jason Wang
2024-01-31 10:00     ` Xuan Zhuo
2024-02-01  0:41       ` Jason Wang
2024-01-30 11:42 ` [PATCH vhost 04/17] virtio_ring: split: remove double check of the unmap ops Xuan Zhuo
2024-01-31  9:12   ` Jason Wang
2024-01-31  9:43     ` Michael S. Tsirkin
2024-02-01  5:47       ` Jason Wang
2024-02-01  6:43     ` Xuan Zhuo [this message]
2024-01-30 11:42 ` [PATCH vhost 05/17] virtio_ring: split: structure the indirect desc table Xuan Zhuo
2024-01-31  9:12   ` Jason Wang
2024-01-30 11:42 ` [PATCH vhost 06/17] virtio_ring: no store dma info when unmap is not needed Xuan Zhuo
2024-01-31  9:12   ` Jason Wang
2024-02-01  6:04     ` Xuan Zhuo
2024-02-02  3:04       ` Jason Wang
2024-01-30 11:42 ` [PATCH vhost 07/17] virtio: find_vqs: pass struct instead of multi parameters Xuan Zhuo
2024-01-31  9:12   ` Jason Wang
2024-02-01  3:00     ` Xuan Zhuo
2024-02-20 14:18       ` Halil Pasic
2024-02-21  2:03         ` Xuan Zhuo
2024-01-30 11:42 ` [PATCH vhost 08/17] virtio: vring_new_virtqueue(): " Xuan Zhuo
2024-01-31  9:12   ` Jason Wang
2024-01-31 11:03   ` Ilpo Järvinen
2024-02-01  2:24     ` Xuan Zhuo
2024-01-30 11:42 ` [PATCH vhost 09/17] virtio_ring: reuse the parameter struct of find_vqs() Xuan Zhuo
2024-01-30 11:42 ` [PATCH vhost 10/17] virtio: find_vqs: add new parameter premapped Xuan Zhuo
2024-01-30 11:42 ` [PATCH vhost 11/17] virtio_ring: export premapped to driver by struct virtqueue Xuan Zhuo
2024-01-31  9:12   ` Jason Wang
2024-02-01  3:17     ` Xuan Zhuo
2024-01-30 11:42 ` [PATCH vhost 12/17] virtio_net: set premapped mode by find_vqs() Xuan Zhuo
2024-01-30 11:42 ` [PATCH vhost 13/17] virtio_ring: remove api of setting vq premapped Xuan Zhuo
2024-01-30 11:42 ` [PATCH vhost 14/17] virtio_ring: introduce dma map api for page Xuan Zhuo
2024-01-30 11:42 ` [PATCH vhost 15/17] virtio_net: unify the code for recycling the xmit ptr Xuan Zhuo
2024-01-30 11:42 ` [PATCH vhost 16/17] virtio_net: rename free_old_xmit_skbs to free_old_xmit Xuan Zhuo
2024-01-30 11:42 ` [PATCH vhost 17/17] virtio_net: sq support premapped mode Xuan Zhuo
2024-01-31  9:12   ` Jason Wang
2024-02-01  3:21     ` Xuan Zhuo
2024-02-01  5:36       ` Jason Wang
2024-02-01  5:56         ` Xuan Zhuo
2024-02-02  3:06           ` Jason Wang
2024-02-01  7:43 ` [PATCH vhost 00/17] virtio: drivers maintain dma info for premapped vq Zhu Yanjun
2024-02-01  9:57   ` Xuan Zhuo

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1706769826.0586398-4-xuanzhuo@linux.alibaba.com \
    --to=xuanzhuo@linux.alibaba.com \
    --cc=agordeev@linux.ibm.com \
    --cc=andersson@kernel.org \
    --cc=anton.ivanov@cambridgegreys.com \
    --cc=ast@kernel.org \
    --cc=benjamin.berg@intel.com \
    --cc=borntraeger@linux.ibm.com \
    --cc=bpf@vger.kernel.org \
    --cc=cohuck@redhat.com \
    --cc=daniel@iogearbox.net \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=farman@linux.ibm.com \
    --cc=gor@linux.ibm.com \
    --cc=hawk@kernel.org \
    --cc=hca@linux.ibm.com \
    --cc=hdegoede@redhat.com \
    --cc=ilpo.jarvinen@linux.intel.com \
    --cc=jasowang@redhat.com \
    --cc=johannes@sipsolutions.net \
    --cc=john.fastabend@gmail.com \
    --cc=kuba@kernel.org \
    --cc=kvm@vger.kernel.org \
    --cc=linux-remoteproc@vger.kernel.org \
    --cc=linux-s390@vger.kernel.org \
    --cc=linux-um@lists.infradead.org \
    --cc=mathieu.poirier@linaro.org \
    --cc=mst@redhat.com \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=pasic@linux.ibm.com \
    --cc=platform-driver-x86@vger.kernel.org \
    --cc=richard@nod.at \
    --cc=svens@linux.ibm.com \
    --cc=vadimp@nvidia.com \
    --cc=virtualization@lists.linux.dev \
    --cc=yang.lee@linux.alibaba.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox