Re: [PATCH 2/2] vhost-net: revert vhost_exceeds_maxpend logic to its original

[Koichiro Den]

On Mon, 2017-08-21 at 20:40 +0800, Jason Wang wrote:
> 
> On 2017年08月21日 11:06, Jason Wang wrote:
> > 
> > 
> > On 2017年08月19日 14:41, Koichiro Den wrote:
> > > To depend on vq.num and the usage of VHOST_MAX_PEND is not succinct
> > > and in some case unexpected, so revert its logic part only.
> > 
> > Hi:
> > 
> > Could you explain a little bit more on the case that is was not 
> > sufficent?
> > 
> > Thanks
> 
> Just have another thought.
> 
> I wonder whether or not just use ulimit(memlock) is better here. It 
> looks more flexible.
> 
> Thanks
It sounds a better approach, though at present I have got no idea how much
portion it should be.

Re: [PATCH 2/2] vhost-net: revert vhost_exceeds_maxpend logic to its original

[Koichiro Den]

On Mon, 2017-08-21 at 11:06 +0800, Jason Wang wrote:
> 
> On 2017年08月19日 14:41, Koichiro Den wrote:
> > To depend on vq.num and the usage of VHOST_MAX_PEND is not succinct
> > and in some case unexpected, so revert its logic part only.
> 
> Hi:
> 
> Could you explain a little bit more on the case that is was not sufficent?
It's just because vq.num could theoretically be around VHOST_MAX_PEND, that
could lead to an unexpected situation. Plus, I thought its name and usage is not
intrinsic. Its actual functioning in normal vq.num == 256 case is I think good
enough. Thanks.
> 
> Thanks
> 
> > 
> > Signed-off-by: Koichiro Den 
> > ---
> >   drivers/vhost/net.c | 8 ++--
> >   1 file changed, 6 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
> > index 06d044862e58..99cf99b308a7 100644
> > --- a/drivers/vhost/net.c
> > +++ b/drivers/vhost/net.c
> > @@ -433,11 +433,15 @@ static int vhost_net_tx_get_vq_desc(struct vhost_net
> > *net,
> >   
> >   static bool vhost_exceeds_maxpend(struct vhost_net *net)
> >   {
> > +   int num_pends;
> >     struct vhost_net_virtqueue *nvq = >vqs[VHOST_NET_VQ_TX];
> >     struct vhost_virtqueue *vq = >vq;
> >   
> > -   return (nvq->upend_idx + vq->num - VHOST_MAX_PEND) % UIO_MAXIOV
> > -   == nvq->done_idx;
> > +   num_pends = likely(nvq->upend_idx >= nvq->done_idx) ?
> > +   (nvq->upend_idx - nvq->done_idx) :
> > +   (nvq->upend_idx + UIO_MAXIOV - nvq->done_idx);
> > +
> > +   return num_pends > VHOST_MAX_PEND;
> >   }
> >   
> >   /* Expects to be always run from workqueue - which acts as
> 
> 

Re: [PATCH 2/2] vhost-net: revert vhost_exceeds_maxpend logic to its original

[Jason Wang]

On 2017年08月21日 11:06, Jason Wang wrote:



On 2017年08月19日 14:41, Koichiro Den wrote:

To depend on vq.num and the usage of VHOST_MAX_PEND is not succinct
and in some case unexpected, so revert its logic part only.


Hi:

Could you explain a little bit more on the case that is was not 
sufficent?


Thanks


Just have another thought.

I wonder whether or not just use ulimit(memlock) is better here. It 
looks more flexible.


Thanks

Re: [PATCH 2/2] vhost-net: revert vhost_exceeds_maxpend logic to its original

[Jason Wang]

On 2017年08月19日 14:41, Koichiro Den wrote:

To depend on vq.num and the usage of VHOST_MAX_PEND is not succinct
and in some case unexpected, so revert its logic part only.


Hi:

Could you explain a little bit more on the case that is was not sufficent?

Thanks



Signed-off-by: Koichiro Den 
---
  drivers/vhost/net.c | 8 ++--
  1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 06d044862e58..99cf99b308a7 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -433,11 +433,15 @@ static int vhost_net_tx_get_vq_desc(struct vhost_net *net,
  
  static bool vhost_exceeds_maxpend(struct vhost_net *net)

  {
+   int num_pends;
struct vhost_net_virtqueue *nvq = >vqs[VHOST_NET_VQ_TX];
struct vhost_virtqueue *vq = >vq;
  
-	return (nvq->upend_idx + vq->num - VHOST_MAX_PEND) % UIO_MAXIOV

-   == nvq->done_idx;
+   num_pends = likely(nvq->upend_idx >= nvq->done_idx) ?
+   (nvq->upend_idx - nvq->done_idx) :
+   (nvq->upend_idx + UIO_MAXIOV - nvq->done_idx);
+
+   return num_pends > VHOST_MAX_PEND;
  }
  
  /* Expects to be always run from workqueue - which acts as

[PATCH 2/2] vhost-net: revert vhost_exceeds_maxpend logic to its original

[Koichiro Den]

To depend on vq.num and the usage of VHOST_MAX_PEND is not succinct
and in some case unexpected, so revert its logic part only.

Signed-off-by: Koichiro Den 
---
 drivers/vhost/net.c | 8 ++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 06d044862e58..99cf99b308a7 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -433,11 +433,15 @@ static int vhost_net_tx_get_vq_desc(struct vhost_net *net,
 
 static bool vhost_exceeds_maxpend(struct vhost_net *net)
 {
+   int num_pends;
struct vhost_net_virtqueue *nvq = >vqs[VHOST_NET_VQ_TX];
struct vhost_virtqueue *vq = >vq;
 
-   return (nvq->upend_idx + vq->num - VHOST_MAX_PEND) % UIO_MAXIOV
-   == nvq->done_idx;
+   num_pends = likely(nvq->upend_idx >= nvq->done_idx) ?
+   (nvq->upend_idx - nvq->done_idx) :
+   (nvq->upend_idx + UIO_MAXIOV - nvq->done_idx);
+
+   return num_pends > VHOST_MAX_PEND;
 }
 
 /* Expects to be always run from workqueue - which acts as
-- 
2.9.4