[PATCH bpf V3 0/2] xsk: fix for xsk_poll writeable

[PATCH bpf V3 0/2] xsk: fix for xsk_poll writeable

[Xuan Zhuo]

V2:
   #2 patch made some changes following magnus' opinions.

V3:
   Regarding the function xskq_cons_present_entries, I think daniel are right,
   I have modified it.

Xuan Zhuo (2):
  xsk: replace datagram_poll by sock_poll_wait
  xsk: change the tx writeable condition

 net/xdp/xsk.c       | 20 ++++++++++++++++----
 net/xdp/xsk_queue.h |  6 ++++++
 2 files changed, 22 insertions(+), 4 deletions(-)

--
1.8.3.1

[PATCH bpf V3 1/2] xsk: replace datagram_poll by sock_poll_wait

[Xuan Zhuo]

datagram_poll will judge the current socket status (EPOLLIN, EPOLLOUT)
based on the traditional socket information (eg: sk_wmem_alloc), but
this does not apply to xsk. So this patch uses sock_poll_wait instead of
datagram_poll, and the mask is calculated by xsk_poll.

Signed-off-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Acked-by: Magnus Karlsson <magnus.karlsson@intel.com>
---
 net/xdp/xsk.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c
index b7b039b..9bbfd8a 100644
--- a/net/xdp/xsk.c
+++ b/net/xdp/xsk.c
@@ -471,11 +471,13 @@ static int xsk_sendmsg(struct socket *sock, struct msghdr *m, size_t total_len)
 static __poll_t xsk_poll(struct file *file, struct socket *sock,
 			     struct poll_table_struct *wait)
 {
-	__poll_t mask = datagram_poll(file, sock, wait);
+	__poll_t mask = 0;
 	struct sock *sk = sock->sk;
 	struct xdp_sock *xs = xdp_sk(sk);
 	struct xsk_buff_pool *pool;
 
+	sock_poll_wait(file, sock, wait);
+
 	if (unlikely(!xsk_is_bound(xs)))
 		return mask;
 
-- 
1.8.3.1

[PATCH bpf V3 2/2] xsk: change the tx writeable condition

[Xuan Zhuo]

Modify the tx writeable condition from the queue is not full to the
number of present tx queues is less than the half of the total number
of queues. Because the tx queue not full is a very short time, this will
cause a large number of EPOLLOUT events, and cause a large number of
process wake up.

Signed-off-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Acked-by: Magnus Karlsson <magnus.karlsson@intel.com>
---
 net/xdp/xsk.c       | 16 +++++++++++++---
 net/xdp/xsk_queue.h |  6 ++++++
 2 files changed, 19 insertions(+), 3 deletions(-)

diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c
index 9bbfd8a..6250447 100644
--- a/net/xdp/xsk.c
+++ b/net/xdp/xsk.c
@@ -211,6 +211,14 @@ static int __xsk_rcv(struct xdp_sock *xs, struct xdp_buff *xdp, u32 len,
 	return 0;
 }
 
+static bool xsk_tx_writeable(struct xdp_sock *xs)
+{
+	if (xskq_cons_present_entries(xs->tx) > xs->tx->nentries / 2)
+		return false;
+
+	return true;
+}
+
 static bool xsk_is_bound(struct xdp_sock *xs)
 {
 	if (READ_ONCE(xs->state) == XSK_BOUND) {
@@ -296,7 +304,8 @@ void xsk_tx_release(struct xsk_buff_pool *pool)
 	rcu_read_lock();
 	list_for_each_entry_rcu(xs, &pool->xsk_tx_list, tx_list) {
 		__xskq_cons_release(xs->tx);
-		xs->sk.sk_write_space(&xs->sk);
+		if (xsk_tx_writeable(xs))
+			xs->sk.sk_write_space(&xs->sk);
 	}
 	rcu_read_unlock();
 }
@@ -436,7 +445,8 @@ static int xsk_generic_xmit(struct sock *sk)
 
 out:
 	if (sent_frame)
-		sk->sk_write_space(sk);
+		if (xsk_tx_writeable(xs))
+			sk->sk_write_space(sk);
 
 	mutex_unlock(&xs->mutex);
 	return err;
@@ -493,7 +503,7 @@ static __poll_t xsk_poll(struct file *file, struct socket *sock,
 
 	if (xs->rx && !xskq_prod_is_empty(xs->rx))
 		mask |= EPOLLIN | EPOLLRDNORM;
-	if (xs->tx && !xskq_cons_is_full(xs->tx))
+	if (xs->tx && xsk_tx_writeable(xs))
 		mask |= EPOLLOUT | EPOLLWRNORM;
 
 	return mask;
diff --git a/net/xdp/xsk_queue.h b/net/xdp/xsk_queue.h
index cdb9cf3..9e71b9f 100644
--- a/net/xdp/xsk_queue.h
+++ b/net/xdp/xsk_queue.h
@@ -264,6 +264,12 @@ static inline bool xskq_cons_is_full(struct xsk_queue *q)
 		q->nentries;
 }
 
+static inline u32 xskq_cons_present_entries(struct xsk_queue *q)
+{
+	/* No barriers needed since data is not accessed */
+	return READ_ONCE(q->ring->producer) - READ_ONCE(q->ring->consumer);
+}
+
 /* Functions for producers */
 
 static inline bool xskq_prod_is_full(struct xsk_queue *q)
-- 
1.8.3.1

Re: [PATCH bpf V3 1/2] xsk: replace datagram_poll by sock_poll_wait

[Magnus Karlsson]

On Tue, Dec 1, 2020 at 3:00 PM Xuan Zhuo <xuanzhuo@linux.alibaba.com> wrote:
>
> datagram_poll will judge the current socket status (EPOLLIN, EPOLLOUT)
> based on the traditional socket information (eg: sk_wmem_alloc), but
> this does not apply to xsk. So this patch uses sock_poll_wait instead of
> datagram_poll, and the mask is calculated by xsk_poll.
>

Could you please add a Fixes label here as this is a bug fix? It will
help the persons backporting this. Here is the line that goes just
above your Sign-off-by.

Fixes: c497176cb2e4 ("xsk: add Rx receive functions and poll support")

Thanks: Magnus

> Signed-off-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
> Acked-by: Magnus Karlsson <magnus.karlsson@intel.com>
> ---
>  net/xdp/xsk.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c
> index b7b039b..9bbfd8a 100644
> --- a/net/xdp/xsk.c
> +++ b/net/xdp/xsk.c
> @@ -471,11 +471,13 @@ static int xsk_sendmsg(struct socket *sock, struct msghdr *m, size_t total_len)
>  static __poll_t xsk_poll(struct file *file, struct socket *sock,
>                              struct poll_table_struct *wait)
>  {
> -       __poll_t mask = datagram_poll(file, sock, wait);
> +       __poll_t mask = 0;
>         struct sock *sk = sock->sk;
>         struct xdp_sock *xs = xdp_sk(sk);
>         struct xsk_buff_pool *pool;
>
> +       sock_poll_wait(file, sock, wait);
> +
>         if (unlikely(!xsk_is_bound(xs)))
>                 return mask;
>
> --
> 1.8.3.1
>

Re: [PATCH bpf V3 2/2] xsk: change the tx writeable condition

[Magnus Karlsson]

On Tue, Dec 1, 2020 at 2:59 PM Xuan Zhuo <xuanzhuo@linux.alibaba.com> wrote:
>
> Modify the tx writeable condition from the queue is not full to the
> number of present tx queues is less than the half of the total number
> of queues. Because the tx queue not full is a very short time, this will
> cause a large number of EPOLLOUT events, and cause a large number of
> process wake up.

And the Fixes label here should be:

Fixes: 35fcde7f8deb ("xsk: support for Tx")

> Signed-off-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
> Acked-by: Magnus Karlsson <magnus.karlsson@intel.com>
> ---
>  net/xdp/xsk.c       | 16 +++++++++++++---
>  net/xdp/xsk_queue.h |  6 ++++++
>  2 files changed, 19 insertions(+), 3 deletions(-)
>
> diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c
> index 9bbfd8a..6250447 100644
> --- a/net/xdp/xsk.c
> +++ b/net/xdp/xsk.c
> @@ -211,6 +211,14 @@ static int __xsk_rcv(struct xdp_sock *xs, struct xdp_buff *xdp, u32 len,
>         return 0;
>  }
>
> +static bool xsk_tx_writeable(struct xdp_sock *xs)
> +{
> +       if (xskq_cons_present_entries(xs->tx) > xs->tx->nentries / 2)
> +               return false;
> +
> +       return true;
> +}
> +
>  static bool xsk_is_bound(struct xdp_sock *xs)
>  {
>         if (READ_ONCE(xs->state) == XSK_BOUND) {
> @@ -296,7 +304,8 @@ void xsk_tx_release(struct xsk_buff_pool *pool)
>         rcu_read_lock();
>         list_for_each_entry_rcu(xs, &pool->xsk_tx_list, tx_list) {
>                 __xskq_cons_release(xs->tx);
> -               xs->sk.sk_write_space(&xs->sk);
> +               if (xsk_tx_writeable(xs))
> +                       xs->sk.sk_write_space(&xs->sk);
>         }
>         rcu_read_unlock();
>  }
> @@ -436,7 +445,8 @@ static int xsk_generic_xmit(struct sock *sk)
>
>  out:
>         if (sent_frame)
> -               sk->sk_write_space(sk);
> +               if (xsk_tx_writeable(xs))
> +                       sk->sk_write_space(sk);
>
>         mutex_unlock(&xs->mutex);
>         return err;
> @@ -493,7 +503,7 @@ static __poll_t xsk_poll(struct file *file, struct socket *sock,
>
>         if (xs->rx && !xskq_prod_is_empty(xs->rx))
>                 mask |= EPOLLIN | EPOLLRDNORM;
> -       if (xs->tx && !xskq_cons_is_full(xs->tx))
> +       if (xs->tx && xsk_tx_writeable(xs))
>                 mask |= EPOLLOUT | EPOLLWRNORM;
>
>         return mask;
> diff --git a/net/xdp/xsk_queue.h b/net/xdp/xsk_queue.h
> index cdb9cf3..9e71b9f 100644
> --- a/net/xdp/xsk_queue.h
> +++ b/net/xdp/xsk_queue.h
> @@ -264,6 +264,12 @@ static inline bool xskq_cons_is_full(struct xsk_queue *q)
>                 q->nentries;
>  }
>
> +static inline u32 xskq_cons_present_entries(struct xsk_queue *q)
> +{
> +       /* No barriers needed since data is not accessed */
> +       return READ_ONCE(q->ring->producer) - READ_ONCE(q->ring->consumer);
> +}
> +
>  /* Functions for producers */
>
>  static inline bool xskq_prod_is_full(struct xsk_queue *q)
> --
> 1.8.3.1
>

Re: [PATCH bpf V3 0/2] xsk: fix for xsk_poll writeable

[patchwork-bot+netdevbpf]

Hello:

This series was applied to bpf/bpf.git (refs/heads/master):

On Tue,  1 Dec 2020 21:56:56 +0800 you wrote:
> V2:
>    #2 patch made some changes following magnus' opinions.
> 
> V3:
>    Regarding the function xskq_cons_present_entries, I think daniel are right,
>    I have modified it.
> 
> [...]

Here is the summary with links:
  - [bpf,V3,1/2] xsk: replace datagram_poll by sock_poll_wait
    https://git.kernel.org/bpf/bpf/c/f5da54187e33
  - [bpf,V3,2/2] xsk: change the tx writeable condition
    https://git.kernel.org/bpf/bpf/c/3413f04141aa

You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html