[dpdk-dev] [PATCH] net/enic: decrement Tx mbuf reference count before recycling

[dpdk-dev] [PATCH] net/enic: decrement Tx mbuf reference count before recycling

[John Daley]

In the Tx cleanup function, the reference count in mbufs to be
returned to the pool should to be decremented before they are
returned. Decrementing is not done by rte_mempool_put_bulk()
so it must be done separately using __rte_pktmbuf_prefree_seg().
If decrementing does not result in a 0 reference count the mbuf
is not returned to the pool and whatever has the last reference
is responsible for freeing.

Fixes: 36935afbc53c ("net/enic: refactor Tx mbuf recycling")
Reviewed-by: Nelson Escobar <neescoba@cisco.com>
Signed-off-by: John Daley <johndale@cisco.com>
---
Since reference counts are set to 0 when mbufs are reallocated from the
pool, and sending packets with reference count not equal to 1 is probably
an application error, this patch may not be critical. But a debug ASSERT
caught it and it would be nice to have it fixed in 16.07.

 drivers/net/enic/enic_rxtx.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/net/enic/enic_rxtx.c b/drivers/net/enic/enic_rxtx.c
index 5ac1d69..96be478 100644
--- a/drivers/net/enic/enic_rxtx.c
+++ b/drivers/net/enic/enic_rxtx.c
@@ -398,7 +398,14 @@ static inline void enic_free_wq_bufs(struct vnic_wq *wq, u16 completed_index)
 	pool = ((struct rte_mbuf *)buf->mb)->pool;
 	for (i = 0; i < nb_to_free; i++) {
 		buf = &wq->bufs[tail_idx];
-		m = (struct rte_mbuf *)(buf->mb);
+		m = __rte_pktmbuf_prefree_seg((struct rte_mbuf *)(buf->mb));
+		buf->mb = NULL;
+
+		if (unlikely(m == NULL)) {
+			tail_idx = enic_ring_incr(desc_count, tail_idx);
+			continue;
+		}
+
 		if (likely(m->pool == pool)) {
 			ENIC_ASSERT(nb_free < ENIC_MAX_WQ_DESCS);
 			free[nb_free++] = m;
@@ -409,7 +416,6 @@ static inline void enic_free_wq_bufs(struct vnic_wq *wq, u16 completed_index)
 			pool = m->pool;
 		}
 		tail_idx = enic_ring_incr(desc_count, tail_idx);
-		buf->mb = NULL;
 	}
 
 	rte_mempool_put_bulk(pool, (void **)free, nb_free);
-- 
2.7.0

[dpdk-dev] [PATCH] net/enic: increment filter failure counter

[John Daley]

One instance of a filter add failure was not incrementing the
the fail counter.

Fixes: 4c2c7bf41f5a ("net/enic: fix negative array index write")
Signed-off-by: John Daley <johndale@cisco.com>
---
 drivers/net/enic/enic_clsf.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/net/enic/enic_clsf.c b/drivers/net/enic/enic_clsf.c
index 3365176..e6f57be 100644
--- a/drivers/net/enic/enic_clsf.c
+++ b/drivers/net/enic/enic_clsf.c
@@ -218,6 +218,7 @@ int enic_fdir_add_fltr(struct enic *enic, struct rte_eth_fdir_filter *params)
 
 	pos = rte_hash_add_key(enic->fdir.hash, params);
 	if (pos < 0) {
+		enic->fdir.stats.f_add++;
 		dev_err(enic, "Add hash key failed\n");
 		return pos;
 	}
-- 
2.7.0

Re: [dpdk-dev] [PATCH] net/enic: increment filter failure counter

[Thomas Monjalon]

2016-07-08 15:22, John Daley:
> One instance of a filter add failure was not incrementing the
> the fail counter.
> 
> Fixes: 4c2c7bf41f5a ("net/enic: fix negative array index write")
> Signed-off-by: John Daley <johndale@cisco.com>

Applied, thanks

Re: [dpdk-dev] [PATCH] net/enic: decrement Tx mbuf reference count before recycling

[Olivier Matz]

Hi John,

On 07/09/2016 12:22 AM, John Daley wrote:
> In the Tx cleanup function, the reference count in mbufs to be
> returned to the pool should to be decremented before they are
> returned. Decrementing is not done by rte_mempool_put_bulk()
> so it must be done separately using __rte_pktmbuf_prefree_seg().
> If decrementing does not result in a 0 reference count the mbuf
> is not returned to the pool and whatever has the last reference
> is responsible for freeing.
> 
> Fixes: 36935afbc53c ("net/enic: refactor Tx mbuf recycling")
> Reviewed-by: Nelson Escobar <neescoba@cisco.com>
> Signed-off-by: John Daley <johndale@cisco.com>
> ---
> Since reference counts are set to 0 when mbufs are reallocated from the
> pool, and sending packets with reference count not equal to 1 is probably
> an application error, this patch may not be critical. But a debug ASSERT
> caught it and it would be nice to have it fixed in 16.07.

Sending a packet with refcnt != 1 is not an error. It can happen when
using mbuf clones. So indeed it would be better to have in 16.07.

For the same reason, I also wonder if enic_free_wq_buf() should also be
updated with:

-       rte_mempool_put(mbuf->pool, mbuf);
+       rte_pktmbuf_free(mbuf);


Regards,
Olivier

Re: [dpdk-dev] [PATCH] net/enic: decrement Tx mbuf reference count before recycling

[John Daley (johndale)]

> -----Original Message-----
> From: Olivier Matz [mailto:olivier.matz@6wind.com]
> Sent: Monday, July 11, 2016 3:04 AM
> To: John Daley (johndale) <johndale@cisco.com>; dev@dpdk.org
> Cc: bruce.richardson@intel.com
> Subject: Re: [dpdk-dev] [PATCH] net/enic: decrement Tx mbuf reference
> count before recycling
> 
> Hi John,
> 
> On 07/09/2016 12:22 AM, John Daley wrote:
> > In the Tx cleanup function, the reference count in mbufs to be
> > returned to the pool should to be decremented before they are
> > returned. Decrementing is not done by rte_mempool_put_bulk() so it
> > must be done separately using __rte_pktmbuf_prefree_seg().
> > If decrementing does not result in a 0 reference count the mbuf is not
> > returned to the pool and whatever has the last reference is
> > responsible for freeing.
> >
> > Fixes: 36935afbc53c ("net/enic: refactor Tx mbuf recycling")
> > Reviewed-by: Nelson Escobar <neescoba@cisco.com>
> > Signed-off-by: John Daley <johndale@cisco.com>
> > ---
> > Since reference counts are set to 0 when mbufs are reallocated from
> > the pool, and sending packets with reference count not equal to 1 is
> > probably an application error, this patch may not be critical. But a
> > debug ASSERT caught it and it would be nice to have it fixed in 16.07.
> 
> Sending a packet with refcnt != 1 is not an error. It can happen when using
> mbuf clones. So indeed it would be better to have in 16.07.
> 
> For the same reason, I also wonder if enic_free_wq_buf() should also be
> updated with:
> 
> -       rte_mempool_put(mbuf->pool, mbuf);
> +       rte_pktmbuf_free(mbuf);
That is a very good point, thank you. I'll use rte_pktmubf_free_seg(mbuf) though, since we are walking an array of all mbuf segments. V2 coming momentarily.
-john
> 
> 
> Regards,
> Olivier

[dpdk-dev] [PATCH v2] net/enic: decrement Tx mbuf reference count before recycling

[John Daley]

In the burst Tx cleanup function, the reference count in mbufs
returned to the pool should to be decremented before they are
returned. Decrementing is not done by rte_mempool_put_bulk()
so it must be done separately using __rte_pktmbuf_prefree_seg().

Also when returning unsent buffers when the device is stopped
use rte_mbuf_free_seg() instead of rte_mempool_put() so that
reference counts are properly decremented.

Fixes: 36935afbc53c ("net/enic: refactor Tx mbuf recycling")

Reviewed-by: Nelson Escobar <neescoba@cisco.com>
Signed-off-by: John Daley <johndale@cisco.com>
---
v2: Use rte_pktmbuf_free_seg when returning mbufs when the device is
stoped. Suggested by Olivier Matz <olivier.matz@6wind.com>.

 drivers/net/enic/enic_main.c |  2 +-
 drivers/net/enic/enic_rxtx.c | 10 ++++++++--
 2 files changed, 9 insertions(+), 3 deletions(-)

diff --git a/drivers/net/enic/enic_main.c b/drivers/net/enic/enic_main.c
index d8669cc..54aaa25 100644
--- a/drivers/net/enic/enic_main.c
+++ b/drivers/net/enic/enic_main.c
@@ -106,7 +106,7 @@ static void enic_free_wq_buf(struct vnic_wq_buf *buf)
 {
 	struct rte_mbuf *mbuf = (struct rte_mbuf *)buf->mb;
 
-	rte_mempool_put(mbuf->pool, mbuf);
+	rte_pktmbuf_free_seg(mbuf);
 	buf->mb = NULL;
 }
 
diff --git a/drivers/net/enic/enic_rxtx.c b/drivers/net/enic/enic_rxtx.c
index 2f4a08c..845a8e6 100644
--- a/drivers/net/enic/enic_rxtx.c
+++ b/drivers/net/enic/enic_rxtx.c
@@ -398,7 +398,14 @@ static inline void enic_free_wq_bufs(struct vnic_wq *wq, u16 completed_index)
 	pool = ((struct rte_mbuf *)buf->mb)->pool;
 	for (i = 0; i < nb_to_free; i++) {
 		buf = &wq->bufs[tail_idx];
-		m = (struct rte_mbuf *)(buf->mb);
+		m = __rte_pktmbuf_prefree_seg((struct rte_mbuf *)(buf->mb));
+		buf->mb = NULL;
+
+		if (unlikely(m == NULL)) {
+			tail_idx = enic_ring_incr(desc_count, tail_idx);
+			continue;
+		}
+
 		if (likely(m->pool == pool)) {
 			RTE_ASSERT(nb_free < ENIC_MAX_WQ_DESCS);
 			free[nb_free++] = m;
@@ -409,7 +416,6 @@ static inline void enic_free_wq_bufs(struct vnic_wq *wq, u16 completed_index)
 			pool = m->pool;
 		}
 		tail_idx = enic_ring_incr(desc_count, tail_idx);
-		buf->mb = NULL;
 	}
 
 	rte_mempool_put_bulk(pool, (void **)free, nb_free);
-- 
2.7.0

Re: [dpdk-dev] [PATCH v2] net/enic: decrement Tx mbuf reference count before recycling

[Thomas Monjalon]

2016-07-11 12:45, John Daley:
> In the burst Tx cleanup function, the reference count in mbufs
> returned to the pool should to be decremented before they are
> returned. Decrementing is not done by rte_mempool_put_bulk()
> so it must be done separately using __rte_pktmbuf_prefree_seg().
> 
> Also when returning unsent buffers when the device is stopped
> use rte_mbuf_free_seg() instead of rte_mempool_put() so that
> reference counts are properly decremented.
> 
> Fixes: 36935afbc53c ("net/enic: refactor Tx mbuf recycling")
> 
> Reviewed-by: Nelson Escobar <neescoba@cisco.com>
> Signed-off-by: John Daley <johndale@cisco.com>

Applied, thanks