Re: [dpdk-dev] [PATCH 1/2] net/ixgbe: fix mailbox interrupt handler

[Zhang, Qi Z]

> -Original Message-
> From: Dai, Wei
> Sent: Thursday, December 28, 2017 12:39 PM
> To: Zhang, Qi Z ; Lu, Wenzhuo
> 
> Cc: dev@dpdk.org; Wang, Liang-min ;
> sta...@dpdk.org
> Subject: RE: [PATCH 1/2] net/ixgbe: fix mailbox interrupt handler
> 
> According to step 7 & 8 in Table 7-74 PF-to-VF Messaging Flow in 82599
> datasheet, The calling of ixgbe_read_mbx( ) can be replaced by following lines
> to avoid touching ->mbx.v2p_mailbox, IXGBE_WRITE_REG(hw,
> IXGBE_VFMAILBOX, IXGBE_VFMAILBOX_ACK);

Thanks for the suggestion.

I think the patch try to keep things unchanged when a PF reset happen, but it 
is not going to cover the case that PF reset interrupt and a foreground VF to 
PF request happen simultaneously which is an knowing issue
Your suggestion may help on that, but it is not a complete fix. I think it will 
be better to have a separate patch that focus on it.

Regards
Qi
> 
> > -Original Message-
> > From: Zhang, Qi Z
> > Sent: Thursday, December 28, 2017 4:22 AM
> > To: Lu, Wenzhuo 
> > Cc: dev@dpdk.org; Dai, Wei ; Wang, Liang-min
> > ; Zhang, Qi Z ;
> > sta...@dpdk.org
> > Subject: [PATCH 1/2] net/ixgbe: fix mailbox interrupt handler
> >
> > Mailbox interrupt handler only take care of PF reset notification, for
> > other message ixgbe_read_mbx should not be called since it get chance
> > to break the foreground VF to PF communication.
> > This can be simply repeated by
> > testpmd>rx_vlan rm all 0
> >
> > Fixes: 77234603fba0 ("net/ixgbe: support VF mailbox interrupt for link
> > up/down")
> > Cc: sta...@dpdk.org
> >
> > Signed-off-by: Qi Zhang 
> > ---
> >  drivers/net/ixgbe/ixgbe_ethdev.c | 10 +++---
> >  1 file changed, 7 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c
> > b/drivers/net/ixgbe/ixgbe_ethdev.c
> > index ff19a56..02121f4 100644
> > --- a/drivers/net/ixgbe/ixgbe_ethdev.c
> > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c
> > @@ -8165,13 +8165,17 @@ static void ixgbevf_mbx_process(struct
> > rte_eth_dev *dev)
> > struct ixgbe_hw *hw =
> > IXGBE_DEV_PRIVATE_TO_HW(dev->data->dev_private);
> > u32 in_msg = 0;
> >
> > -   if (ixgbe_read_mbx(hw, &in_msg, 1, 0))
> > -   return;
> > +   /* peek the message first */
> > +   in_msg = IXGBE_READ_REG(hw, IXGBE_VFMBMEM);
> >
> > /* PF reset VF event */
> > -   if (in_msg == IXGBE_PF_CONTROL_MSG)
> > +   if (in_msg == IXGBE_PF_CONTROL_MSG) {
> > +   /* dummy mbx read to ack pf */
> > +   if (ixgbe_read_mbx(hw, &in_msg, 1, 0))
> > +   return;
> > _rte_eth_dev_callback_process(dev,
> > RTE_ETH_EVENT_INTR_RESET,
> >   NULL, NULL);
> > +   }
> >  }
> >
> >  static int
> > --
> > 2.7.4

Re: [dpdk-dev] [PATCH] i40e: fix flag sent to mac_address_write

[Xing, Beilei]

> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Igor Ryzhov
> Sent: Thursday, January 11, 2018 6:40 AM
> To: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH] i40e: fix flag sent to mac_address_write

Prefix with net/i40e:

> 
> Use the same value as in Linux driver.

Fixes line is needed.
Thanks for the patch.

> 
> Signed-off-by: Igor Ryzhov 
> ---
>  drivers/net/i40e/i40e_ethdev.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/drivers/net/i40e/i40e_ethdev.c b/drivers/net/i40e/i40e_ethdev.c
> index 285d92b..93b9dd0 100644
> --- a/drivers/net/i40e/i40e_ethdev.c
> +++ b/drivers/net/i40e/i40e_ethdev.c
> @@ -10795,8 +10795,7 @@ static void i40e_set_default_mac_addr(struct
> rte_eth_dev *dev,
>   return;
>   }
> 
> - /* Flags: 0x3 updates port address */
> - i40e_aq_mac_address_write(hw, 0x3, mac_addr->addr_bytes,
> NULL);
> + i40e_aq_mac_address_write(hw, I40E_AQC_WRITE_TYPE_LAA_WOL,
> +mac_addr->addr_bytes, NULL);
>  }
> 
>  static int
> --
> 2.6.4

[dpdk-dev] [PATCH 1/6] ethdev: add a function to look up Rx offload names

[Andrew Rybchenko]

From: Ivan Malov 

Commonly, drivers converted to the new offload API
may need to log unsupported offloads as a response
to wrong settings. From this perspective, it would
be convenient to have generic functions to look up
offload names. The patch adds such a helper for Rx.

Signed-off-by: Ivan Malov 
Signed-off-by: Andrew Rybchenko 

Cc: Thomas Monjalon 
Cc: Ferruh Yigit 
Cc: Shahaf Shuler 

---
 lib/librte_ether/rte_ethdev.c   | 42 +
 lib/librte_ether/rte_ethdev.h   | 15 
 lib/librte_ether/rte_ethdev_version.map |  6 +
 3 files changed, 63 insertions(+)

diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c
index b349599..3d09950 100644
--- a/lib/librte_ether/rte_ethdev.c
+++ b/lib/librte_ether/rte_ethdev.c
@@ -89,6 +89,32 @@ static const struct rte_eth_xstats_name_off 
rte_txq_stats_strings[] = {
 #define RTE_NB_TXQ_STATS (sizeof(rte_txq_stats_strings) /  \
sizeof(rte_txq_stats_strings[0]))
 
+#defineRTE_RX_OFFLOAD_BIT2STR(_name)   \
+   { DEV_RX_OFFLOAD_##_name, #_name }
+
+static const struct {
+   uint64_t offload;
+   const char *name;
+} rte_rx_offload_names[] = {
+   RTE_RX_OFFLOAD_BIT2STR(VLAN_STRIP),
+   RTE_RX_OFFLOAD_BIT2STR(IPV4_CKSUM),
+   RTE_RX_OFFLOAD_BIT2STR(UDP_CKSUM),
+   RTE_RX_OFFLOAD_BIT2STR(TCP_CKSUM),
+   RTE_RX_OFFLOAD_BIT2STR(TCP_LRO),
+   RTE_RX_OFFLOAD_BIT2STR(QINQ_STRIP),
+   RTE_RX_OFFLOAD_BIT2STR(OUTER_IPV4_CKSUM),
+   RTE_RX_OFFLOAD_BIT2STR(MACSEC_STRIP),
+   RTE_RX_OFFLOAD_BIT2STR(HEADER_SPLIT),
+   RTE_RX_OFFLOAD_BIT2STR(VLAN_FILTER),
+   RTE_RX_OFFLOAD_BIT2STR(VLAN_EXTEND),
+   RTE_RX_OFFLOAD_BIT2STR(JUMBO_FRAME),
+   RTE_RX_OFFLOAD_BIT2STR(CRC_STRIP),
+   RTE_RX_OFFLOAD_BIT2STR(SCATTER),
+   RTE_RX_OFFLOAD_BIT2STR(TIMESTAMP),
+   RTE_RX_OFFLOAD_BIT2STR(SECURITY),
+};
+
+#undef RTE_RX_OFFLOAD_BIT2STR
 
 /**
  * The user application callback description.
@@ -746,6 +772,22 @@ rte_eth_convert_rx_offloads(const uint64_t rx_offloads,
rxmode->security = 0;
 }
 
+const char *
+rte_eth_dev_rx_offload_name(uint64_t offload)
+{
+   const char *name = "UNKNOWN";
+   unsigned int i;
+
+   for (i = 0; i < RTE_DIM(rte_rx_offload_names); ++i) {
+   if (offload == rte_rx_offload_names[i].offload) {
+   name = rte_rx_offload_names[i].name;
+   break;
+   }
+   }
+
+   return name;
+}
+
 int
 rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_q, uint16_t nb_tx_q,
  const struct rte_eth_conf *dev_conf)
diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h
index f0eeefe..88ceffa 100644
--- a/lib/librte_ether/rte_ethdev.h
+++ b/lib/librte_ether/rte_ethdev.h
@@ -945,6 +945,11 @@ struct rte_eth_conf {
 DEV_RX_OFFLOAD_VLAN_FILTER | \
 DEV_RX_OFFLOAD_VLAN_EXTEND)
 
+/*
+ * If new Rx offload capabilities are defined, they also must be
+ * mentioned in rte_rx_offload_names in rte_ethdev.c file.
+ */
+
 /**
  * TX offload capabilities of a device.
  */
@@ -1922,6 +1927,16 @@ int rte_eth_dev_detach(uint16_t port_id, char *devname);
 uint32_t rte_eth_speed_bitflag(uint32_t speed, int duplex);
 
 /**
+ * Get DEV_RX_OFFLOAD_* flag name
+ *
+ * @param offload
+ *   Offload flag
+ * @return
+ *   Offload name or 'UNKNOWN' if the flag cannot be recognised
+ */
+const char *rte_eth_dev_rx_offload_name(uint64_t offload);
+
+/**
  * Configure an Ethernet device.
  * This function must be invoked first before any other function in the
  * Ethernet API. This function can also be re-invoked when a device is in the
diff --git a/lib/librte_ether/rte_ethdev_version.map 
b/lib/librte_ether/rte_ethdev_version.map
index e9681ac..03455cf 100644
--- a/lib/librte_ether/rte_ethdev_version.map
+++ b/lib/librte_ether/rte_ethdev_version.map
@@ -198,6 +198,12 @@ DPDK_17.11 {
 
 } DPDK_17.08;
 
+DPDK_18.02 {
+   global:
+
+   rte_eth_dev_rx_offload_name;
+} DPDK_17.11;
+
 EXPERIMENTAL {
global:
 
-- 
2.7.4

[dpdk-dev] [PATCH 0/6] net/sfc: convert to the new offload API

[Andrew Rybchenko]

May be it is too late to suggest a new API functions to ethdev,
but hopefully if the idea is accepted, it could be applied in the
current release cycle since these functions are trivial.

I'm not sure that rte_ethdev_version.map is updated correctly
since EXPERIMENTAL tag is present and I don't understand how it
should be handled.

In general for the transition period from old offload API to the
new one it would be useful to convert Tx offloads to/from txq_flags
in rte_eth_dev_info_get() for default_txconf and
rte_eth_tx_queue_info_get(). Unfortunately it was lost during
new offload API patches review. However, it would require testing
for all network PMDs and we decided to follow more conservative
approach and kept code to fill in txq_flags which should be simply
removed when txq_flags are removed.

Cc: Thomas Monjalon 
Cc: Ferruh Yigit 
Cc: Shahaf Shuler 

Ivan Malov (6):
  ethdev: add a function to look up Rx offload names
  ethdev: add a function to look up Tx offload names
  net/sfc: factor out function to report Rx capabilities
  net/sfc: convert to the new Rx offload API
  net/sfc: factor out function to report Tx capabilities
  net/sfc: convert to the new Tx offload API

 drivers/net/sfc/sfc_dp_tx.h |   2 +
 drivers/net/sfc/sfc_ethdev.c|  58 ++
 drivers/net/sfc/sfc_port.c  |   5 +-
 drivers/net/sfc/sfc_rx.c| 128 +++--
 drivers/net/sfc/sfc_rx.h|   3 +
 drivers/net/sfc/sfc_tx.c| 194 +---
 drivers/net/sfc/sfc_tx.h|   4 +
 lib/librte_ether/rte_ethdev.c   |  87 ++
 lib/librte_ether/rte_ethdev.h   |  30 +
 lib/librte_ether/rte_ethdev_version.map |   7 ++
 10 files changed, 396 insertions(+), 122 deletions(-)

-- 
2.7.4

[dpdk-dev] [PATCH 2/6] ethdev: add a function to look up Tx offload names

[Andrew Rybchenko]

From: Ivan Malov 

Commonly, drivers converted to the new offload API
may need to log unsupported offloads as a response
to wrong settings. From this perspective, it would
be convenient to have generic functions to look up
offload names. The patch adds such a helper for Tx.

Signed-off-by: Ivan Malov 
Signed-off-by: Andrew Rybchenko 

Cc: Thomas Monjalon 
Cc: Ferruh Yigit 
Cc: Shahaf Shuler 
---
 lib/librte_ether/rte_ethdev.c   | 45 +
 lib/librte_ether/rte_ethdev.h   | 15 +++
 lib/librte_ether/rte_ethdev_version.map |  1 +
 3 files changed, 61 insertions(+)

diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c
index 3d09950..071521e 100644
--- a/lib/librte_ether/rte_ethdev.c
+++ b/lib/librte_ether/rte_ethdev.c
@@ -116,6 +116,35 @@ static const struct {
 
 #undef RTE_RX_OFFLOAD_BIT2STR
 
+#defineRTE_TX_OFFLOAD_BIT2STR(_name)   \
+   { DEV_TX_OFFLOAD_##_name, #_name }
+
+static const struct {
+   uint64_t offload;
+   const char *name;
+} rte_tx_offload_names[] = {
+   RTE_TX_OFFLOAD_BIT2STR(VLAN_INSERT),
+   RTE_TX_OFFLOAD_BIT2STR(IPV4_CKSUM),
+   RTE_TX_OFFLOAD_BIT2STR(UDP_CKSUM),
+   RTE_TX_OFFLOAD_BIT2STR(TCP_CKSUM),
+   RTE_TX_OFFLOAD_BIT2STR(SCTP_CKSUM),
+   RTE_TX_OFFLOAD_BIT2STR(TCP_TSO),
+   RTE_TX_OFFLOAD_BIT2STR(UDP_TSO),
+   RTE_TX_OFFLOAD_BIT2STR(OUTER_IPV4_CKSUM),
+   RTE_TX_OFFLOAD_BIT2STR(QINQ_INSERT),
+   RTE_TX_OFFLOAD_BIT2STR(VXLAN_TNL_TSO),
+   RTE_TX_OFFLOAD_BIT2STR(GRE_TNL_TSO),
+   RTE_TX_OFFLOAD_BIT2STR(IPIP_TNL_TSO),
+   RTE_TX_OFFLOAD_BIT2STR(GENEVE_TNL_TSO),
+   RTE_TX_OFFLOAD_BIT2STR(MACSEC_INSERT),
+   RTE_TX_OFFLOAD_BIT2STR(MT_LOCKFREE),
+   RTE_TX_OFFLOAD_BIT2STR(MULTI_SEGS),
+   RTE_TX_OFFLOAD_BIT2STR(MBUF_FAST_FREE),
+   RTE_TX_OFFLOAD_BIT2STR(SECURITY),
+};
+
+#undef RTE_TX_OFFLOAD_BIT2STR
+
 /**
  * The user application callback description.
  *
@@ -788,6 +817,22 @@ rte_eth_dev_rx_offload_name(uint64_t offload)
return name;
 }
 
+const char *
+rte_eth_dev_tx_offload_name(uint64_t offload)
+{
+   const char *name = "UNKNOWN";
+   unsigned int i;
+
+   for (i = 0; i < RTE_DIM(rte_tx_offload_names); ++i) {
+   if (offload == rte_tx_offload_names[i].offload) {
+   name = rte_tx_offload_names[i].name;
+   break;
+   }
+   }
+
+   return name;
+}
+
 int
 rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_q, uint16_t nb_tx_q,
  const struct rte_eth_conf *dev_conf)
diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h
index 88ceffa..04a4b9c 100644
--- a/lib/librte_ether/rte_ethdev.h
+++ b/lib/librte_ether/rte_ethdev.h
@@ -980,6 +980,11 @@ struct rte_eth_conf {
  */
 #define DEV_TX_OFFLOAD_SECURITY 0x0002
 
+/*
+ * If new Tx offload capabilities are defined, they also must be
+ * mentioned in rte_tx_offload_names in rte_ethdev.c file.
+ */
+
 struct rte_pci_device;
 
 /**
@@ -1937,6 +1942,16 @@ uint32_t rte_eth_speed_bitflag(uint32_t speed, int 
duplex);
 const char *rte_eth_dev_rx_offload_name(uint64_t offload);
 
 /**
+ * Get DEV_TX_OFFLOAD_* flag name
+ *
+ * @param offload
+ *   Offload flag
+ * @return
+ *   Offload name or 'UNKNOWN' if the flag cannot be recognised
+ */
+const char *rte_eth_dev_tx_offload_name(uint64_t offload);
+
+/**
  * Configure an Ethernet device.
  * This function must be invoked first before any other function in the
  * Ethernet API. This function can also be re-invoked when a device is in the
diff --git a/lib/librte_ether/rte_ethdev_version.map 
b/lib/librte_ether/rte_ethdev_version.map
index 03455cf..45f393f 100644
--- a/lib/librte_ether/rte_ethdev_version.map
+++ b/lib/librte_ether/rte_ethdev_version.map
@@ -202,6 +202,7 @@ DPDK_18.02 {
global:
 
rte_eth_dev_rx_offload_name;
+   rte_eth_dev_tx_offload_name;
 } DPDK_17.11;
 
 EXPERIMENTAL {
-- 
2.7.4

[dpdk-dev] [PATCH 3/6] net/sfc: factor out function to report Rx capabilities

[Andrew Rybchenko]

From: Ivan Malov 

The patch adds a separate function to report supported
Rx capabilities because this function will be required
in more places across the code in the upcoming patches.

Signed-off-by: Ivan Malov 
Signed-off-by: Andrew Rybchenko 
---
 drivers/net/sfc/sfc_ethdev.c |  9 +
 drivers/net/sfc/sfc_rx.c | 17 +
 drivers/net/sfc/sfc_rx.h |  2 ++
 3 files changed, 20 insertions(+), 8 deletions(-)

diff --git a/drivers/net/sfc/sfc_ethdev.c b/drivers/net/sfc/sfc_ethdev.c
index af867a7..851b38b 100644
--- a/drivers/net/sfc/sfc_ethdev.c
+++ b/drivers/net/sfc/sfc_ethdev.c
@@ -104,14 +104,7 @@ sfc_dev_infos_get(struct rte_eth_dev *dev, struct 
rte_eth_dev_info *dev_info)
/* By default packets are dropped if no descriptors are available */
dev_info->default_rxconf.rx_drop_en = 1;
 
-   dev_info->rx_offload_capa =
-   DEV_RX_OFFLOAD_IPV4_CKSUM |
-   DEV_RX_OFFLOAD_UDP_CKSUM |
-   DEV_RX_OFFLOAD_TCP_CKSUM;
-
-   if ((encp->enc_tunnel_encapsulations_supported != 0) &&
-   (sa->dp_rx->features & SFC_DP_RX_FEAT_TUNNELS))
-   dev_info->rx_offload_capa |= DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM;
+   dev_info->rx_offload_capa = sfc_rx_get_dev_offload_caps(sa);
 
dev_info->tx_offload_capa =
DEV_TX_OFFLOAD_IPV4_CKSUM |
diff --git a/drivers/net/sfc/sfc_rx.c b/drivers/net/sfc/sfc_rx.c
index 6fa56b4..d35f4f7 100644
--- a/drivers/net/sfc/sfc_rx.c
+++ b/drivers/net/sfc/sfc_rx.c
@@ -762,6 +762,23 @@ sfc_rx_qstop(struct sfc_adapter *sa, unsigned int sw_index)
sfc_ev_qstop(rxq->evq);
 }
 
+uint64_t
+sfc_rx_get_dev_offload_caps(struct sfc_adapter *sa)
+{
+   const efx_nic_cfg_t *encp = efx_nic_cfg_get(sa->nic);
+   uint64_t caps = 0;
+
+   caps |= DEV_RX_OFFLOAD_IPV4_CKSUM;
+   caps |= DEV_RX_OFFLOAD_UDP_CKSUM;
+   caps |= DEV_RX_OFFLOAD_TCP_CKSUM;
+
+   if (encp->enc_tunnel_encapsulations_supported &&
+   (sa->dp_rx->features & SFC_DP_RX_FEAT_TUNNELS))
+   caps |= DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM;
+
+   return caps;
+}
+
 static int
 sfc_rx_qcheck_conf(struct sfc_adapter *sa, unsigned int rxq_max_fill_level,
   const struct rte_eth_rxconf *rx_conf)
diff --git a/drivers/net/sfc/sfc_rx.h b/drivers/net/sfc/sfc_rx.h
index 2cf75bc..cc9245f 100644
--- a/drivers/net/sfc/sfc_rx.h
+++ b/drivers/net/sfc/sfc_rx.h
@@ -142,6 +142,8 @@ void sfc_rx_qfini(struct sfc_adapter *sa, unsigned int 
sw_index);
 int sfc_rx_qstart(struct sfc_adapter *sa, unsigned int sw_index);
 void sfc_rx_qstop(struct sfc_adapter *sa, unsigned int sw_index);
 
+uint64_t sfc_rx_get_dev_offload_caps(struct sfc_adapter *sa);
+
 void sfc_rx_qflush_done(struct sfc_rxq *rxq);
 void sfc_rx_qflush_failed(struct sfc_rxq *rxq);
 
-- 
2.7.4

[dpdk-dev] [PATCH 5/6] net/sfc: factor out function to report Tx capabilities

[Andrew Rybchenko]

From: Ivan Malov 

The patch adds a separate function to report supported
Tx capabilities because this function will be required
in more places across the code in the upcoming patches.

Signed-off-by: Ivan Malov 
Signed-off-by: Andrew Rybchenko 
---
 drivers/net/sfc/sfc_ethdev.c | 13 +
 drivers/net/sfc/sfc_tx.c | 23 +++
 drivers/net/sfc/sfc_tx.h |  2 ++
 3 files changed, 26 insertions(+), 12 deletions(-)

diff --git a/drivers/net/sfc/sfc_ethdev.c b/drivers/net/sfc/sfc_ethdev.c
index 0244a0f..0fe9bf5 100644
--- a/drivers/net/sfc/sfc_ethdev.c
+++ b/drivers/net/sfc/sfc_ethdev.c
@@ -114,20 +114,12 @@ sfc_dev_infos_get(struct rte_eth_dev *dev, struct 
rte_eth_dev_info *dev_info)
dev_info->rx_offload_capa = sfc_rx_get_dev_offload_caps(sa) |
dev_info->rx_queue_offload_capa;
 
-   dev_info->tx_offload_capa =
-   DEV_TX_OFFLOAD_IPV4_CKSUM |
-   DEV_TX_OFFLOAD_UDP_CKSUM |
-   DEV_TX_OFFLOAD_TCP_CKSUM;
-
-   if (encp->enc_tunnel_encapsulations_supported != 0)
-   dev_info->tx_offload_capa |= DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM;
+   dev_info->tx_offload_capa = sfc_tx_get_dev_offload_caps(sa);
 
dev_info->default_txconf.txq_flags = ETH_TXQ_FLAGS_NOXSUMSCTP;
if ((~sa->dp_tx->features & SFC_DP_TX_FEAT_VLAN_INSERT) ||
!encp->enc_hw_tx_insert_vlan_enabled)
dev_info->default_txconf.txq_flags |= ETH_TXQ_FLAGS_NOVLANOFFL;
-   else
-   dev_info->tx_offload_capa |= DEV_TX_OFFLOAD_VLAN_INSERT;
 
if (~sa->dp_tx->features & SFC_DP_TX_FEAT_MULTI_SEG)
dev_info->default_txconf.txq_flags |= ETH_TXQ_FLAGS_NOMULTSEGS;
@@ -146,9 +138,6 @@ sfc_dev_infos_get(struct rte_eth_dev *dev, struct 
rte_eth_dev_info *dev_info)
}
 #endif
 
-   if (sa->tso)
-   dev_info->tx_offload_capa |= DEV_TX_OFFLOAD_TCP_TSO;
-
/* Initialize to hardware limits */
dev_info->rx_desc_lim.nb_max = EFX_RXQ_MAXNDESCS;
dev_info->rx_desc_lim.nb_min = EFX_RXQ_MINNDESCS;
diff --git a/drivers/net/sfc/sfc_tx.c b/drivers/net/sfc/sfc_tx.c
index ee22049..7504037 100644
--- a/drivers/net/sfc/sfc_tx.c
+++ b/drivers/net/sfc/sfc_tx.c
@@ -34,6 +34,29 @@
  */
 #define SFC_TX_QFLUSH_POLL_ATTEMPTS(2000)
 
+uint64_t
+sfc_tx_get_dev_offload_caps(struct sfc_adapter *sa)
+{
+   const efx_nic_cfg_t *encp = efx_nic_cfg_get(sa->nic);
+   uint64_t caps = 0;
+
+   caps |= DEV_TX_OFFLOAD_IPV4_CKSUM;
+   caps |= DEV_TX_OFFLOAD_UDP_CKSUM;
+   caps |= DEV_TX_OFFLOAD_TCP_CKSUM;
+
+   if (encp->enc_tunnel_encapsulations_supported)
+   caps |= DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM;
+
+   if ((sa->dp_tx->features & SFC_DP_TX_FEAT_VLAN_INSERT) &&
+   encp->enc_hw_tx_insert_vlan_enabled)
+   caps |= DEV_TX_OFFLOAD_VLAN_INSERT;
+
+   if (sa->tso)
+   caps |= DEV_TX_OFFLOAD_TCP_TSO;
+
+   return caps;
+}
+
 static int
 sfc_tx_qcheck_conf(struct sfc_adapter *sa, unsigned int txq_max_fill_level,
   const struct rte_eth_txconf *tx_conf)
diff --git a/drivers/net/sfc/sfc_tx.h b/drivers/net/sfc/sfc_tx.h
index c2b889f..fc9a9f7 100644
--- a/drivers/net/sfc/sfc_tx.h
+++ b/drivers/net/sfc/sfc_tx.h
@@ -129,6 +129,8 @@ void sfc_tx_qstop(struct sfc_adapter *sa, unsigned int 
sw_index);
 int sfc_tx_start(struct sfc_adapter *sa);
 void sfc_tx_stop(struct sfc_adapter *sa);
 
+uint64_t sfc_tx_get_dev_offload_caps(struct sfc_adapter *sa);
+
 /* From 'sfc_tso.c' */
 int sfc_efx_tso_alloc_tsoh_objs(struct sfc_efx_tx_sw_desc *sw_ring,
unsigned int txq_entries,
-- 
2.7.4

[dpdk-dev] [PATCH 4/6] net/sfc: convert to the new Rx offload API

[Andrew Rybchenko]

From: Ivan Malov 

Ethdev Rx offloads API has changed since:
commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API")
This commit support the new Rx offloads API.

Signed-off-by: Ivan Malov 
Signed-off-by: Andrew Rybchenko 
---
 drivers/net/sfc/sfc_ethdev.c |  27 +--
 drivers/net/sfc/sfc_port.c   |   5 +-
 drivers/net/sfc/sfc_rx.c | 111 +--
 drivers/net/sfc/sfc_rx.h |   1 +
 4 files changed, 103 insertions(+), 41 deletions(-)

diff --git a/drivers/net/sfc/sfc_ethdev.c b/drivers/net/sfc/sfc_ethdev.c
index 851b38b..0244a0f 100644
--- a/drivers/net/sfc/sfc_ethdev.c
+++ b/drivers/net/sfc/sfc_ethdev.c
@@ -104,7 +104,15 @@ sfc_dev_infos_get(struct rte_eth_dev *dev, struct 
rte_eth_dev_info *dev_info)
/* By default packets are dropped if no descriptors are available */
dev_info->default_rxconf.rx_drop_en = 1;
 
-   dev_info->rx_offload_capa = sfc_rx_get_dev_offload_caps(sa);
+   dev_info->rx_queue_offload_capa = sfc_rx_get_queue_offload_caps(sa);
+
+   /*
+* rx_offload_capa includes both device and queue offloads since
+* the latter may be requested on a per device basis which makes
+* sense when some offloads are needed to be set on all queues.
+*/
+   dev_info->rx_offload_capa = sfc_rx_get_dev_offload_caps(sa) |
+   dev_info->rx_queue_offload_capa;
 
dev_info->tx_offload_capa =
DEV_TX_OFFLOAD_IPV4_CKSUM |
@@ -882,7 +890,13 @@ sfc_dev_set_mtu(struct rte_eth_dev *dev, uint16_t mtu)
 * The driver does not use it, but other PMDs update jumbo_frame
 * flag and max_rx_pkt_len when MTU is set.
 */
-   dev->data->dev_conf.rxmode.jumbo_frame = (mtu > ETHER_MAX_LEN);
+   if (mtu > ETHER_MAX_LEN) {
+   struct rte_eth_rxmode *rxmode = &dev->data->dev_conf.rxmode;
+
+   rxmode->offloads |= DEV_RX_OFFLOAD_JUMBO_FRAME;
+   rxmode->jumbo_frame = 1;
+   }
+
dev->data->dev_conf.rxmode.max_rx_pkt_len = sa->port.pdu;
 
sfc_adapter_unlock(sa);
@@ -1045,8 +1059,13 @@ sfc_rx_queue_info_get(struct rte_eth_dev *dev, uint16_t 
rx_queue_id,
qinfo->conf.rx_free_thresh = rxq->refill_threshold;
qinfo->conf.rx_drop_en = 1;
qinfo->conf.rx_deferred_start = rxq_info->deferred_start;
-   qinfo->scattered_rx =
-   ((rxq_info->type_flags & EFX_RXQ_FLAG_SCATTER) != 0);
+   qinfo->conf.offloads = DEV_RX_OFFLOAD_IPV4_CKSUM |
+  DEV_RX_OFFLOAD_UDP_CKSUM |
+  DEV_RX_OFFLOAD_TCP_CKSUM;
+   if (rxq_info->type_flags & EFX_RXQ_FLAG_SCATTER) {
+   qinfo->conf.offloads |= DEV_RX_OFFLOAD_SCATTER;
+   qinfo->scattered_rx = 1;
+   }
qinfo->nb_desc = rxq_info->entries;
 
sfc_adapter_unlock(sa);
diff --git a/drivers/net/sfc/sfc_port.c b/drivers/net/sfc/sfc_port.c
index a48388d..c423f52 100644
--- a/drivers/net/sfc/sfc_port.c
+++ b/drivers/net/sfc/sfc_port.c
@@ -299,11 +299,12 @@ sfc_port_configure(struct sfc_adapter *sa)
 {
const struct rte_eth_dev_data *dev_data = sa->eth_dev->data;
struct sfc_port *port = &sa->port;
+   const struct rte_eth_rxmode *rxmode = &dev_data->dev_conf.rxmode;
 
sfc_log_init(sa, "entry");
 
-   if (dev_data->dev_conf.rxmode.jumbo_frame)
-   port->pdu = dev_data->dev_conf.rxmode.max_rx_pkt_len;
+   if (rxmode->offloads & DEV_RX_OFFLOAD_JUMBO_FRAME)
+   port->pdu = rxmode->max_rx_pkt_len;
else
port->pdu = EFX_MAC_PDU(dev_data->mtu);
 
diff --git a/drivers/net/sfc/sfc_rx.c b/drivers/net/sfc/sfc_rx.c
index d35f4f7..abc53fb 100644
--- a/drivers/net/sfc/sfc_rx.c
+++ b/drivers/net/sfc/sfc_rx.c
@@ -768,6 +768,8 @@ sfc_rx_get_dev_offload_caps(struct sfc_adapter *sa)
const efx_nic_cfg_t *encp = efx_nic_cfg_get(sa->nic);
uint64_t caps = 0;
 
+   caps |= DEV_RX_OFFLOAD_JUMBO_FRAME;
+   caps |= DEV_RX_OFFLOAD_CRC_STRIP;
caps |= DEV_RX_OFFLOAD_IPV4_CKSUM;
caps |= DEV_RX_OFFLOAD_UDP_CKSUM;
caps |= DEV_RX_OFFLOAD_TCP_CKSUM;
@@ -779,10 +781,62 @@ sfc_rx_get_dev_offload_caps(struct sfc_adapter *sa)
return caps;
 }
 
+uint64_t
+sfc_rx_get_queue_offload_caps(struct sfc_adapter *sa)
+{
+   uint64_t caps = 0;
+
+   if (sa->dp_rx->features & SFC_DP_RX_FEAT_SCATTER)
+   caps |= DEV_RX_OFFLOAD_SCATTER;
+
+   return caps;
+}
+
+static void
+sfc_rx_log_offloads(struct sfc_adapter *sa, const char *offload_group,
+   const char *verdict, uint64_t offloads)
+{
+   unsigned long long bit;
+
+   while ((bit = __builtin_ffsll(offloads)) != 0) {
+   uint64_t flag = (1ULL << --bit);
+
+   sfc_err(sa, "Rx %s offload %s %s", offload_group,
+   rte_eth_dev_rx_offload_name(flag), verdict);
+
+   offloads &= ~flag;
+   }

[dpdk-dev] [PATCH 6/6] net/sfc: convert to the new Tx offload API

[Andrew Rybchenko]

From: Ivan Malov 

Ethdev Tx offloads API has changed since:
commit cba7f53b717d ("ethdev: introduce Tx queue offloads API")
This commit support the new Tx offloads API.

The code which fills in txq_flags in default_txconf is preserved
because rte_eth_dev_info_get() lacks conversion between offloads
and txq_flags fields which means that a legacy application which
relies on default_txconf will fail to configure Tx queues in the
case when some bits in txq_flags are mandatory.

Signed-off-by: Ivan Malov 
Signed-off-by: Andrew Rybchenko 
---
 drivers/net/sfc/sfc_dp_tx.h  |   2 +
 drivers/net/sfc/sfc_ethdev.c |  17 +++-
 drivers/net/sfc/sfc_tx.c | 179 +++
 drivers/net/sfc/sfc_tx.h |   2 +
 4 files changed, 131 insertions(+), 69 deletions(-)

diff --git a/drivers/net/sfc/sfc_dp_tx.h b/drivers/net/sfc/sfc_dp_tx.h
index f400755..75d72fe 100644
--- a/drivers/net/sfc/sfc_dp_tx.h
+++ b/drivers/net/sfc/sfc_dp_tx.h
@@ -41,6 +41,8 @@ struct sfc_dp_tx_qcreate_info {
unsigned intfree_thresh;
/** Transmit queue configuration flags */
unsigned intflags;
+   /** Offloads enabled on the transmit queue */
+   uint64_toffloads;
/** Tx queue size */
unsigned inttxq_entries;
/** Maximum size of data in the DMA descriptor */
diff --git a/drivers/net/sfc/sfc_ethdev.c b/drivers/net/sfc/sfc_ethdev.c
index 0fe9bf5..a86cff6 100644
--- a/drivers/net/sfc/sfc_ethdev.c
+++ b/drivers/net/sfc/sfc_ethdev.c
@@ -83,6 +83,7 @@ sfc_dev_infos_get(struct rte_eth_dev *dev, struct 
rte_eth_dev_info *dev_info)
 {
struct sfc_adapter *sa = dev->data->dev_private;
const efx_nic_cfg_t *encp = efx_nic_cfg_get(sa->nic);
+   uint64_t txq_offloads_def = 0;
 
sfc_log_init(sa, "entry");
 
@@ -114,7 +115,20 @@ sfc_dev_infos_get(struct rte_eth_dev *dev, struct 
rte_eth_dev_info *dev_info)
dev_info->rx_offload_capa = sfc_rx_get_dev_offload_caps(sa) |
dev_info->rx_queue_offload_capa;
 
-   dev_info->tx_offload_capa = sfc_tx_get_dev_offload_caps(sa);
+   dev_info->tx_queue_offload_capa = sfc_tx_get_queue_offload_caps(sa);
+
+   /*
+* tx_offload_capa includes both device and queue offloads since
+* the latter may be requested on a per device basis which makes
+* sense when some offloads are needed to be set on all queues.
+*/
+   dev_info->tx_offload_capa = sfc_tx_get_dev_offload_caps(sa) |
+   dev_info->tx_queue_offload_capa;
+
+   if (dev_info->tx_offload_capa & DEV_TX_OFFLOAD_MBUF_FAST_FREE)
+   txq_offloads_def |= DEV_TX_OFFLOAD_MBUF_FAST_FREE;
+
+   dev_info->default_txconf.offloads |= txq_offloads_def;
 
dev_info->default_txconf.txq_flags = ETH_TXQ_FLAGS_NOXSUMSCTP;
if ((~sa->dp_tx->features & SFC_DP_TX_FEAT_VLAN_INSERT) ||
@@ -1081,6 +1095,7 @@ sfc_tx_queue_info_get(struct rte_eth_dev *dev, uint16_t 
tx_queue_id,
memset(qinfo, 0, sizeof(*qinfo));
 
qinfo->conf.txq_flags = txq_info->txq->flags;
+   qinfo->conf.offloads = txq_info->txq->offloads;
qinfo->conf.tx_free_thresh = txq_info->txq->free_thresh;
qinfo->conf.tx_deferred_start = txq_info->deferred_start;
qinfo->nb_desc = txq_info->entries;
diff --git a/drivers/net/sfc/sfc_tx.c b/drivers/net/sfc/sfc_tx.c
index 7504037..757b03b 100644
--- a/drivers/net/sfc/sfc_tx.c
+++ b/drivers/net/sfc/sfc_tx.c
@@ -40,6 +40,26 @@ sfc_tx_get_dev_offload_caps(struct sfc_adapter *sa)
const efx_nic_cfg_t *encp = efx_nic_cfg_get(sa->nic);
uint64_t caps = 0;
 
+   if ((sa->dp_tx->features & SFC_DP_TX_FEAT_VLAN_INSERT) &&
+   encp->enc_hw_tx_insert_vlan_enabled)
+   caps |= DEV_TX_OFFLOAD_VLAN_INSERT;
+
+   if (sa->dp_tx->features & SFC_DP_TX_FEAT_MULTI_SEG)
+   caps |= DEV_TX_OFFLOAD_MULTI_SEGS;
+
+   if ((~sa->dp_tx->features & SFC_DP_TX_FEAT_MULTI_POOL) &&
+   (~sa->dp_tx->features & SFC_DP_TX_FEAT_REFCNT))
+   caps |= DEV_TX_OFFLOAD_MBUF_FAST_FREE;
+
+   return caps;
+}
+
+uint64_t
+sfc_tx_get_queue_offload_caps(struct sfc_adapter *sa)
+{
+   const efx_nic_cfg_t *encp = efx_nic_cfg_get(sa->nic);
+   uint64_t caps = 0;
+
caps |= DEV_TX_OFFLOAD_IPV4_CKSUM;
caps |= DEV_TX_OFFLOAD_UDP_CKSUM;
caps |= DEV_TX_OFFLOAD_TCP_CKSUM;
@@ -47,22 +67,55 @@ sfc_tx_get_dev_offload_caps(struct sfc_adapter *sa)
if (encp->enc_tunnel_encapsulations_supported)
caps |= DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM;
 
-   if ((sa->dp_tx->features & SFC_DP_TX_FEAT_VLAN_INSERT) &&
-   encp->enc_hw_tx_insert_vlan_enabled)
-   caps |= DEV_TX_OFFLOAD_VLAN_INSERT;
-
if (sa->tso)
caps |= DEV_TX_OFFLOAD_TCP_TSO;
 
return caps;
 }
 
+static void
+sfc_tx_log_offloads(struct sfc_adapter *sa, const char *

Re: [dpdk-dev] [dpdk-dev, v2] net/e1000: always enable receive and transmit

[Wenzhuo Lu]

Hi,

> Date: Thu, 19 Oct 2017 23:23:39 -0400
>
> From: Chas Williams 
>
> The transmit and receive controller state machines are only enabled after
> receiving an interrupt and the link status is now valid.  If an adapter
> is being used in conjunction with NC-SI, network controller sideband
> interface, the adapter may never get a link state change interrupt since
> the adapter's PHY is always link up and never changes state.
>
> To fix this, always enable and disable the transmit and receive with
> .dev_start and .dev_stop.  This is a better match for what is typically
> done with the other PMD's.  Since we may never get an interrupt to check
> the link state, we also poll once at the end of .dev_start to get the
> current link status.
>
> Signed-off-by: Chas Williams 
Acked-by: Wenzhuo Lu 

Re: [dpdk-dev] [PATCH] net/mlx5: fix parsing all-multicast from flow item

[Nélio Laranjeiro]

On Wed, Jan 10, 2018 at 11:51:53PM -0800, Yongseok Koh wrote:
> As the dst_mac of allmulti is already masked with the mask, it has 0x01 in
> the first octet. Checking the least significant bit only can't distinguish
> it from broadcast or IPv6 multicast.
>
> Fixes: bb47fb6e6067 ("net/mlx5: fix flow type for allmulti rules")
> Cc: sta...@dpdk.org
> 
> Signed-off-by: Yongseok Koh 
> ---
>  drivers/net/mlx5/mlx5_flow.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
> index 305b2ec01..d01c8069b 100644
> --- a/drivers/net/mlx5/mlx5_flow.c
> +++ b/drivers/net/mlx5/mlx5_flow.c
> @@ -1281,7 +1281,7 @@ mlx5_flow_create_eth(const struct rte_flow_item *item,
>   eth.val.ether_type &= eth.mask.ether_type;
>   }
>   mlx5_flow_create_copy(parser, ð, eth_size);
> - parser->allmulti = eth.val.dst_mac[0] & 1;
> + parser->allmulti = eth.val.dst_mac[0] == 0x01;
>   return 0;
>  }
>  
> -- 
> 2.11.0
> 

Seems you are introducing a bug, for broadcast Mac addresses, this will
not work i.e. 0xff != 0x01 but it as the multicast bit set.  From my
understanding, Verbs flow attribute must also be modified in such
situation.

Are you sure about this change?

Thanks,

-- 
Nélio Laranjeiro
6WIND

Re: [dpdk-dev] [PATCH] net/i40e: fix VSI MAC filter on primary address change

[Xing, Beilei]

Hi Igor,

Thanks for the catch, and glad to see your patch☺ It resolves a potential 
problem in PMD.
The patch looks OK for me except some minor comments (in another mail thread).

Best Regards,
Beilei

From: Igor Ryzhov [mailto:iryz...@nfware.com]
Sent: Thursday, January 11, 2018 6:47 AM
To: Zhang, Helin 
Cc: Xing, Beilei ; Olivier Matz 
; dev@dpdk.org; Wu, Jingjing ; 
sta...@dpdk.org; Laurent Hardy 
Subject: Re: [dpdk-dev] [PATCH] net/i40e: fix VSI MAC filter on primary address 
change

Hello everyone.

It's sad that my comments were unanswered.
I'm ok with the first two – they were mostly style-related.
But I made an investigation on the third one and it is a bug.

This is a description (from X710 datasheet) of flags sent to mac_address_write 
command:

By bits:
0-7 – Reserved
8 – MAC_MAG_EN
9 – LAA_WOL_PRESERVE
10-13 – Reserved
14-15 – Write type (00 – Update LAA only, 01 – Update LAA and WOL, 10 – Update 
port address, 11 – Reserved, but used in Linux to enable multicast magic packet 
wake up)

Current code uses 0x3 flag, apparently trying to update LAA, WOL and port 
address, but it sets first two bits instead of last two.
These bits are reserved, that's why it doesn't break anything.
At the same time, last two bits are set to zero, and the command changes LAA 
address only – it's enough to work in simple case.

The last question – which flag is correct to use – 01 (LAA + WOL) or 11 (LAA + 
WOL + port).
Linux driver uses the first one, and here is the patch to fix the issue:
https://dpdk.org/dev/patchwork/patch/33524/

Best regards,
Igor

On Wed, Jan 10, 2018 at 4:57 PM, Zhang, Helin 
mailto:helin.zh...@intel.com>> wrote:


> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On 
> Behalf Of Xing, Beilei
> Sent: Thursday, January 4, 2018 1:39 PM
> To: Olivier Matz; dev@dpdk.org; Wu, Jingjing
> Cc: sta...@dpdk.org; Laurent Hardy
> Subject: Re: [dpdk-dev] [PATCH] net/i40e: fix VSI MAC filter on primary 
> address
> change
>
>
>
> > -Original Message-
> > From: Olivier Matz 
> > [mailto:olivier.m...@6wind.com]
> > Sent: Wednesday, January 3, 2018 10:29 PM
> > To: dev@dpdk.org; Wu, Jingjing 
> > mailto:jingjing...@intel.com>>; Xing, Beilei
> > mailto:beilei.x...@intel.com>>
> > Cc: sta...@dpdk.org; Laurent Hardy 
> > mailto:laurent.ha...@6wind.com>>
> > Subject: [PATCH] net/i40e: fix VSI MAC filter on primary address
> > change
> >
> > When primary address mac is changed, the mac filters were not updated
> > in the VSI with the new mac addr and incoming packets with this
> > destination address are dropped by the hardware filters.
> >
> > This patch removes the VSI mac filter for the previous mac address and
> > adds a new one for new mac address.
> >
> > Fixes: e18e01e92c29 ("i40e: support default MAC address setting")
> > Cc: sta...@dpdk.org
> >
> > Signed-off-by: Laurent Hardy 
> > mailto:laurent.ha...@6wind.com>>
> > Signed-off-by: Olivier Matz 
> > mailto:olivier.m...@6wind.com>>
>
> Thanks for the fix.
> Acked-by: Beilei Xing mailto:beilei.x...@intel.com>>
Applied to dpdk-next-net-intel, thanks!

/Helin

[dpdk-dev] [RFC] New packet type query API

[Qiming Yang]

This RFC contains a proposal to add a new packet type query API to DPDK. 

Motivation
==
In current DPDK implementation, when received one packet, driver will lookup 
the packet type mapping table and transmit the result value to struct mbuf. 
Appilcation will analyze the mbuf->packet_type only when user want to show the 
packet type name. It will waste a lot of cycles, because sometimes user don't 
care what the packet type is, but the look-up will always exist. We think the 
packet type name is not needed all the time and the look-up is not needed 
neither. So we will add a flag, when enable the flag, the ptype value in 
descriptor will be passed through to mbuf->packet_type directly. So a new API 
to get packet name by values is added also, it will be called when user want to 
know what the packet type name is. So we can use the flag to control which 
value will transmit to mbuf(direct ptype value or ptype name value).

The current ptype expression depend on bit width(max 32 bit, used 28bit 
already), if we support more packet type in future, like next generation NIC 
support flexible pipeline, it will introduce some new packet types, the 
remaining bit may not enough. So I suppose to use the rte_flow items to present 
packet type.  The advantage of rte_flow item are: 
1. It's generic enough, every NIC can use these items to present their packet; 
2. it's flexible, can add new items unlimited to support new packet type.

Proposed solution

The new API used to query the exact packet type name with packet type value 
from descriptor.

/** Packet type name structure */
struct rte_eth_ptype_name  {
uint64_t ptype_value;
enum *name;
}

/**
*Query the exact packet type name with packet type values
*
* @param   port_id Port identifier of Ethernet device.
* @param   names  Pointer to packet type name buffer.
* @param   size   The size of ptype_values array (number of element).
* @param   ptype_valuesPointer to a table of ptype values.
* @return
*  - On success returns zero 
*  - On failure returns non-zero
*/
int rte_eth_ptype_name_get_by_value (uint8_t port_id, struct rte_eth_ptype_name 
*names, int size, uint64_t *ptype_values);

Add a new flag 'enable_ptype_direct' in structure rte_eth_rxmode to enable 
packet type get direct mode. The direct mode means the packet type value will 
pass through directly from HW to driver without lookup ptype mapping table. 
User can call function  rte_eth_ptype_name_get_by_value to query what the 
packet type value means when they need.

struct rte_eth_rxmode {
??? ???
uint16_t header_split : 1, /**< Header Split enable. */
hw_ip_checksum   : 1, /**< IP/UDP/TCP checksum offload enable. 
*/
hw_vlan_filter   : 1, /**< VLAN filter enable. */
hw_vlan_strip: 1, /**< VLAN strip enable. */
hw_vlan_extend   : 1, /**< Extended VLAN enable. */
jumbo_frame  : 1, /**< Jumbo Frame Receipt enable. */
hw_strip_crc : 1, /**< Enable CRC stripping by hardware. */
enable_scatter   : 1, /**< Enable scatter packets rx handler */
+   enable_ptype_direct  :1, /** < Enable packet type get direct 
mode. */
enable_lro   : 1; /**< Enable LRO */
};
 
This flag will be configured in dev_configure(), and can be queried by user 
through dev->data->dev_conf.rxmode.enable_ptype_direct. In receive function, 
driver will store HW's packet type value in mbuf->packet_type if direct mode is 
enabled, If not, maintain the existing code.

Driver maintains a new ptype_value and rte_flow item mapping table, when 
download a profile and support new packet type, the SW mapping table will be 
updated according to the ptype information analyzed from profile. 

Future work
===
Support to configure the packet type direct mode per queue.

Re: [dpdk-dev] [PATCH] net/e1000: add minimum support for Broadcom 54616 PHY

[Lu, Wenzhuo]

Hi Chas,

> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Chas Williams
> Sent: Thursday, December 7, 2017 7:55 AM
> To: dev@dpdk.org
> Cc: skh...@vmware.com; Charles (Chas) Williams ;
> Chas Williams 
> Subject: [dpdk-dev] [PATCH] net/e1000: add minimum support for Broadcom
> 54616 PHY
> 
> From: "Charles (Chas) Williams" 
> 
> If we find a Broadcom 54616, handle as a e1000_phy_none assuming that
> the NIC reset has initialized the PHY to a sane state.
> 
> Signed-off-by: Chas Williams 

> b/drivers/net/e1000/base/e1000_defines.h
> index dbc2bbb..e2101c1 100644
> --- a/drivers/net/e1000/base/e1000_defines.h
> +++ b/drivers/net/e1000/base/e1000_defines.h
> @@ -1274,6 +1274,7 @@ POSSIBILITY OF SUCH DAMAGE.
>  #define I350_I_PHY_ID0x015403B0
>  #define I210_I_PHY_ID0x01410C00
>  #define IGP04E1000_E_PHY_ID  0x02A80391
> +#define BCM54616_E_PHY_ID0x03625D10
TBH, normally we don't change the base code. I checked the kernel driver and 
don't find this PHY either. May I ask if it's an Intel NIC and what's the 
device ID? Thanks.

Re: [dpdk-dev] [PATCH v4 0/5] lib: add Port Representors

[Alex Rosenbaum]

Declan,

In all previous series I explained my objection for using the broker
framework for holding the VF rep port.
You guys just waved it away (and sent a new series) instead of really
giving a clear reason why they are required.

I think Yuanhan's suggestion gets ride of the broker model in a clean way.
It exposes the VF rep's and keeps much similar to port model to allow
both control and second phase data path. The flow steering actions
suggestion match much nicer with this as well.

PS: me and Yuanhan did exchange ideas regarding this previously, so
it's not surprising I agree with his suggestion.

Alex


On Thu, Jan 11, 2018 at 5:18 AM, Yuanhan Liu  wrote:
> On Mon, Jan 08, 2018 at 02:37:15PM +, Remy Horton wrote:
>> Port Representors provide a logical presentation in DPDK of VF (virtual
>> function) ports for the purposes of control and monitoring. Each port
>> representor device represents a single VF and is associated with it's
>> parent physical function (PF) PMD which provides the back-end hooks for
>> the representor device ops and defines the control domain to which that
>> port belongs. This allows to use existing DPDK APIs to monitor and control
>> the port without the need to create and maintain VF specific APIs.
>
> Firstly, I don't object this model. More precisely, I don't object for
> introducing a port to control the VFs. I even like this idea a bit, for
> at least it could get rid of some PMD specific APIs, say
> rte_pmd_i40e_set_vf_mac_addr, etc.
>
> However, I don't quite like this design, for a simple reason, it makes
> things way more complex:
>
> - new APIs are introduced.
> - new virtual PMD is required
> - broker concept
>
>
> I was thinking below should work:
>
> - Use the devargs for enabling the port (or VF) representor model.
>   For example: -w 04:00.0,vf_ports=0-3
>
>   The vf_ports is for telling the driver we are going to create VF
>   representors for VF 0 to VF 3. It could be a port mask like what
>   this patchset does.
>
> - Then inside the driver (say, i40e)
>
>   * allocate 1 DPDK ethdev port
>   * allocate 4 DPDK ethdev port, one for each VF specified by
> the vf_ports option. And these are VF representors.
>   * link the VF representors to the right VF port
> For example, for i40e, it should be (from patch 3):
> vf = &pf->vfs[representor->vport_id];
>   * set the proper ethdev callbacks for the VF representors.
>
> As you can see, none of above complexity is introduced.
>
> Probably you might find something are missing:
>
> - to identify the vf rep port for a specific VF.
>
>   Which could be addressed by the new devargs syntax we are proposing:
>   http://dpdk.org/ml/archives/dev/2017-December/084234.html
>
>   For example, the right VF rep port could be identified by below devarg:
>   bus=pci,id=04:00.0/class=eth,vf_id=0
>
> - the ability of hotplug
>
>   I think it could also be addressed by the new devargs syntax, also by
>   re-using the rte_eth_dev_attach() function:
>
>   rte_eth_dev_attach("bus=pci,id=04:00.0/class=eth,vf_id=4", &port_id);
>
>
> Thoughts?
>
> --yliu
>>
>> +-+   +---+  +---+
>> |Control Plane|   |   Data Plane  |  |   Data Plane  |
>> | Application |   |   Application |  |   Application |
>> +-+   +---+  +---+
>> | eth dev api |   |  eth dev api  |  |  eth dev api  |
>> +-+   +---+  +---+
>> +---+  +---+  +---+   +---+  +---+
>> |  PF0  |  | Port  |  | Port  |   |VF0 PMD|  |VF0 PMD|
>> |  PMD  <--+ Rep 0 |  | Rep 1 |   +---+  +--++
>> |   |  | PMD   |  | PMD   | |
>> +---+--^+  +---+  +-+-+ |
>> |  ||  ||
>> |  ++  ||
>> |  ||
>> |  ||
>> ++  |
>> |   |  HW (logical view)   | |  |
>> | --+--+ +---+ +---+---+ |  |
>> | |   PF   | |  VF0  | |  VF1  | |  |
>> | || |   | |   ++
>> | ++ +---+ +---+ |
>> | ++ |
>> | |VEB | |
>> | ++ |
>> | ++ |
>> | |  Port  | |
>> | |   0| |
>> | ++ |
>> ++
>>
>> The figure above shows a deployment where the PF is bound to a DPDK control
>> plane application which uses representor ports to manage the configuration 
>>

Re: [dpdk-dev] [PATCH v6 02/14] net/avf: initialization of avf PMD

[Ferruh Yigit]

On 1/11/2018 2:07 AM, Lu, Wenzhuo wrote:
> Hi Stephen,
> 
>> -Original Message-
>> From: Stephen Hemminger [mailto:step...@networkplumber.org]
>> Sent: Thursday, January 11, 2018 1:15 AM
>> To: Lu, Wenzhuo 
>> Cc: dev@dpdk.org; Wu, Jingjing 
>> Subject: Re: [dpdk-dev] [PATCH v6 02/14] net/avf: initialization of avf PMD
>>
>> On Wed, 10 Jan 2018 14:15:49 +0800
>> Wenzhuo Lu  wrote:
>>
>>> +
>>> +#define AVF_MAX_NUM_QUEUES   16
>>> +/* Vlan table size */
>>> +#define AVF_VLAN_TB_SIZE   (4096 / (CHAR_BIT * 
>>> sizeof(uint32_t)))
>>
>> You could use ETHER_MAX_VLAN_ID (which is 4095).
>> Also it is most efficient if bit tables use unsigned long to access.
> Thanks for the suggestion.
> I found this macro is useless. I'd like just removing it.
> 
> Hi Ferruh,
> As this patch set is accepted to next-net, I can sent a fixes patch for this 
> change. Is it OK? Would you like helping merge the fixes to the original 
> patch? Thanks.

Hi Wenzhuo,

That is OK, I can squash fixes on top original set in next-net.

Re: [dpdk-dev] [PATCH] bus/fslmc: add support for scanned device count

[Hemant Agrawal]

On 1/2/2018 6:38 PM, Shreyansh Jain wrote:

FSLMC bus detects a multiple type of logical objects representing
components of the datapath.

Using the type of device, a newly introduced API
rte_fslmc_get_device_count can return the count of devices
scanned of that device type.

Signed-off-by: Shreyansh Jain 
---
:: This patch is based on *net-next* tree.


..
Acked-by: Hemant Agrawal 

Re: [dpdk-dev] [PATCH v3 6/8] net/vdev_netvsc: skip routed netvsc probing

[Matan Azrad]

Hi Stephan

From: Stephen Hemminger, Wednesday, January 10, 2018 6:44 PM
> On Wed, 10 Jan 2018 15:07:14 +
> Matan Azrad  wrote:
> 
> > Hi Stephan
> >
> > From: Stephen Hemminger, Tuesday, January 9, 2018 8:51 PM
> > > To: Matan Azrad 
> > > Cc: Ferruh Yigit ; Thomas Monjalon
> > > ; dev@dpdk.org; Raslan Darawsheh
> > > 
> > > Subject: Re: [PATCH v3 6/8] net/vdev_netvsc: skip routed netvsc
> > > probing
> > >
> > > On Tue,  9 Jan 2018 14:47:31 +
> > > Matan Azrad  wrote:
> > >
> > > > NetVSC netdevices which are already routed should not be probed
> > > > because they are used for management purposes by the HyperV.
> > > >
> > > > prevent routed netvsc devices probing.
> > > >
> > > > Signed-off-by: Raslan Darawsheh 
> > > > Signed-off-by: Matan Azrad 
> > > > ---
> > > >  doc/guides/nics/vdev_netvsc.rst   |  2 +-
> > > >  drivers/net/vdev_netvsc/vdev_netvsc.c | 46
> > > > +++
> > > >  2 files changed, 47 insertions(+), 1 deletion(-)
> > > >
> > > > diff --git a/doc/guides/nics/vdev_netvsc.rst
> > > > b/doc/guides/nics/vdev_netvsc.rst index fde1fb8..f779862 100644
> > > > --- a/doc/guides/nics/vdev_netvsc.rst
> > > > +++ b/doc/guides/nics/vdev_netvsc.rst
> > > > @@ -87,4 +87,4 @@ The following device parameters are supported:
> > > >MAC address.
> > > >
> > > >  Not specifying either ``iface`` or ``mac`` makes this driver
> > > > attach itself to -all NetVSC interfaces found on the system.
> > > > +all unrouted NetVSC interfaces found on the system.
> > > > diff --git a/drivers/net/vdev_netvsc/vdev_netvsc.c
> > > > b/drivers/net/vdev_netvsc/vdev_netvsc.c
> > > > index 3d8895b..4295b92 100644
> > > > --- a/drivers/net/vdev_netvsc/vdev_netvsc.c
> > > > +++ b/drivers/net/vdev_netvsc/vdev_netvsc.c
> > > > @@ -38,6 +38,7 @@
> > > >  #define VDEV_NETVSC_PROBE_MS 1000
> > > >
> > > >  #define NETVSC_CLASS_ID "{f8615163-df3e-46c5-913f-f2d2f965ed0e}"
> > > > +#define NETVSC_MAX_ROUTE_LINE_SIZE 300
> > > >
> > > >  #define DRV_LOG(level, ...) \
> > > > rte_log(RTE_LOG_ ## level, \
> > > > @@ -192,6 +193,44 @@ static LIST_HEAD(, vdev_netvsc_ctx)
> > > > vdev_netvsc_ctx_list =  }
> > > >
> > > >  /**
> > > > + * Determine if a network interface has a route.
> > > > + *
> > > > + * @param[in] name
> > > > + *   Network device name.
> > > > + *
> > > > + * @return
> > > > + *   A nonzero value when interface has an route. In case of error,
> > > > + *   rte_errno is updated and 0 returned.
> > > > + */
> > > > +static int
> > > > +vdev_netvsc_has_route(const char *name) {
> > > > +   FILE *fp;
> > > > +   int ret = 0;
> > > > +   char route[NETVSC_MAX_ROUTE_LINE_SIZE];
> > > > +   char *netdev;
> > > > +
> > > > +   fp = fopen("/proc/net/route", "r");
> > > > +   if (!fp) {
> > > > +   rte_errno = errno;
> > > > +   return 0;
> > > > +   }
> > > > +   while (fgets(route, NETVSC_MAX_ROUTE_LINE_SIZE, fp) != NULL) {
> > > > +   netdev = strtok(route, "\t");
> > > > +   if (strcmp(netdev, name) == 0) {
> > > > +   ret = 1;
> > > > +   break;
> > > > +   }
> > > > +   /* Move file pointer to the next line. */
> > > > +   while (strchr(route, '\n') == NULL &&
> > > > +  fgets(route, NETVSC_MAX_ROUTE_LINE_SIZE, fp) !=
> > > NULL)
> > > > +   ;
> > > > +   }
> > > > +   fclose(fp);
> > > > +   return ret;
> > > > +}
> > >
> > > In many ways /proc/net/route is legacy intervace.
> > > And system may have 1 M routes.
> > >
> > > Maybe there is faster way to do this with netlink by looking to see
> > > if there is an address associated with the interface.
> >
> > Actually this is control path, we don't care about performance very much.
> > But I can get other idea here, Do you have suggestion?
> >
> > Thanks!
> >
> 
> Use netlink (or ioctl) to get interface address.
> If interface has an IPv4 or IPv6 (not link local), then skip it.

As I a little bit investigated I found that IPv6 getting is problematic by 
ioctl.
And using nelink for it, really doesn't  worth the effort.
So, I suggest to keep this code simple as is in spite of the optional high 
latency for this function, after all it is a control path.

[dpdk-dev] [PATCH 2/2] examples/ipsec-secgw: fix SPI byte order in flow item

[Nelio Laranjeiro]

SPI field is defined in the RFC2406 [1] as a big endian field it should be
provided in its final form to the drivers through RTE flow.

Fixes: ec17993a145a ("examples/ipsec-secgw: support security offload")
Cc: akhil.go...@nxp.com

Signed-off-by: Nelio Laranjeiro 

[1] https://tools.ietf.org/html/rfc2406
---
 examples/ipsec-secgw/ipsec.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/examples/ipsec-secgw/ipsec.c b/examples/ipsec-secgw/ipsec.c
index 580e09a3a..8df0f00ab 100644
--- a/examples/ipsec-secgw/ipsec.c
+++ b/examples/ipsec-secgw/ipsec.c
@@ -195,7 +195,7 @@ create_session(struct ipsec_ctx *ipsec_ctx, struct ipsec_sa 
*sa)
sa->pattern[2].type = RTE_FLOW_ITEM_TYPE_ESP;
sa->pattern[2].spec = &sa->esp_spec;
sa->pattern[2].mask = &rte_flow_item_esp_mask;
-   sa->esp_spec.hdr.spi = sa->spi;
+   sa->esp_spec.hdr.spi = rte_cpu_to_be_32(sa->spi);
 
sa->pattern[3].type = RTE_FLOW_ITEM_TYPE_END;
 
-- 
2.11.0

[dpdk-dev] [PATCH 1/2] net: fix ESP header byte ordering definition

[Nelio Laranjeiro]

ESP header is defined in the RFC2406 [1] as Big Endian fields it should use
the corresponding types in DPDK as well.

Fixes: d4b684f7197a ("net: add ESP header to generic flow steering")
Cc: bor...@mellanox.com

Signed-off-by: Nelio Laranjeiro 

[1] https://tools.ietf.org/html/rfc2406
---
 lib/librte_net/rte_esp.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/lib/librte_net/rte_esp.h b/lib/librte_net/rte_esp.h
index e228af092..148c06e09 100644
--- a/lib/librte_net/rte_esp.h
+++ b/lib/librte_net/rte_esp.h
@@ -49,8 +49,8 @@ extern "C" {
  * ESP Header
  */
 struct esp_hdr {
-   uint32_t spi;  /**< Security Parameters Index */
-   uint32_t seq;  /**< packet sequence number */
+   rte_be32_t spi;  /**< Security Parameters Index */
+   rte_be32_t seq;  /**< packet sequence number */
 } __attribute__((__packed__));
 
 #ifdef __cplusplus
-- 
2.11.0

[dpdk-dev] [PATCH] net/mlx5: fix all multi verification code position

[Nelio Laranjeiro]

All multi code should not be handled in exit part of the code but in the
mainline of the function.

Fixes: 4241d84c0a32 ("net/mlx5: fix flow type for allmulti rules")
Cc: rasl...@mellanox.com

Signed-off-by: Nelio Laranjeiro 
---
 drivers/net/mlx5/mlx5_flow.c | 20 ++--
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
index 334a4f4ba..f050ea713 100644
--- a/drivers/net/mlx5/mlx5_flow.c
+++ b/drivers/net/mlx5/mlx5_flow.c
@@ -1156,16 +1156,6 @@ priv_flow_convert(struct priv *priv,
attr->priority +
hash_rxq_init[parser->layer].flow_priority;
}
-exit_free:
-   /* Only verification is expected, all resources should be released. */
-   if (!parser->create) {
-   for (i = 0; i != hash_rxq_init_n; ++i) {
-   if (parser->queue[i].ibv_attr) {
-   rte_free(parser->queue[i].ibv_attr);
-   parser->queue[i].ibv_attr = NULL;
-   }
-   }
-   }
if (parser->allmulti &&
parser->layer == HASH_RXQ_ETH) {
for (i = 0; i != hash_rxq_init_n; ++i) {
@@ -1177,6 +1167,16 @@ priv_flow_convert(struct priv *priv,
IBV_FLOW_ATTR_MC_DEFAULT;
}
}
+exit_free:
+   /* Only verification is expected, all resources should be released. */
+   if (!parser->create) {
+   for (i = 0; i != hash_rxq_init_n; ++i) {
+   if (parser->queue[i].ibv_attr) {
+   rte_free(parser->queue[i].ibv_attr);
+   parser->queue[i].ibv_attr = NULL;
+   }
+   }
+   }
return ret;
 exit_enomem:
for (i = 0; i != hash_rxq_init_n; ++i) {
-- 
2.11.0

Re: [dpdk-dev] [PATCH v4 0/5] lib: add Port Representors

[Remy Horton]

On 11/01/2018 03:18, Yuanhan Liu wrote:
[..]

- Then inside the driver (say, i40e)

  * allocate 1 DPDK ethdev port
  * allocate 4 DPDK ethdev port, one for each VF specified by
the vf_ports option. And these are VF representors.
  * link the VF representors to the right VF port
For example, for i40e, it should be (from patch 3):
vf = &pf->vfs[representor->vport_id];
  * set the proper ethdev callbacks for the VF representors.

As you can see, none of above complexity is introduced.


What springs to mind is whether this would end up delegating to the 
individual PMDs tasks that in the patchset are handled centrally in the 
library. Question is where the balance should be, and as an aside what 
should be the base-line requirements for a PMD to be "port representable".


..Remy

[dpdk-dev] [PATCH] test/memzone: Fix test_memzone_basic & test_memzone_free issues.

[Phil Yang]

1. test_memzone_basic: No need to free a NULL memzone. It will cause test
termination.
2. test_memzone_free: Out of mz[RTE_MAX_MEMZONE] memory bound while
reserving memzone for mz. It will flush variable i. Fix by extend to
mz[RTE_MAX_MEMZONE + 1].

Signed-off-by: Phil Yang 
---
 test/test/test_memzone.c | 6 +-
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/test/test/test_memzone.c b/test/test/test_memzone.c
index 6e80977..24e29a7 100644
--- a/test/test/test_memzone.c
+++ b/test/test/test_memzone.c
@@ -251,10 +251,6 @@ test_memzone_reserve_flags(void)
printf("MEMZONE FLAG 2MB\n");
return -1;
}
-   if (rte_memzone_free(mz)) {
-   printf("Fail memzone free\n");
-   return -1;
-   }
}
 
if (hugepage_2MB_avail && hugepage_1GB_avail) {
@@ -746,7 +742,7 @@ test_memzone_bounded(void)
 static int
 test_memzone_free(void)
 {
-   const struct rte_memzone *mz[RTE_MAX_MEMZONE];
+   const struct rte_memzone *mz[RTE_MAX_MEMZONE + 1];
int i;
char name[20];
 
-- 
2.7.4

Re: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on VF

[Dai, Wei]

According to 82599 datasheet, VF working mode is controlled by its host PF.
If VF is working in RSS or DCB mode, it can use more than 1 queues.
That also means its host PF should enable RSS or DCB.
The PF Rx mode can be ETH_MQ_RX_VMDQ_RSS or ETH_MQ_RX_VMDQ_DCB.
So I think that DPDK app with PF cause this issue, not ixgbe PMD itself.

By the way, you can reply your patch mail to provide more info about how
You reproduce this issue to easy others' reviewing and testing your patch.

Thanks

> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Yanglong Wu
> Sent: Tuesday, January 9, 2018 2:32 PM
> To: dev@dpdk.org; Yang, Zhiyong 
> Cc: Wu, Yanglong 
> Subject: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on VF
> 
> VF can't run in multi queue mode, if nb_q_per_pool was set as 1.
> Nb_q_per_pool is passed through to max_rx_q and max_tx_q in VF.
> So if nb_q_per_pool is equal to 1, max_rx_q and max_tx_q can't be more
> than 1 and VF multi queue mode will fail.
> 
> Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check)
> 
> Signed-off-by: Yanglong Wu 
> ---
> v1~v3:
> coding style issue fix
> ---
> v4:
> Rework according to comments
> ---
>  drivers/net/ixgbe/ixgbe_ethdev.c | 2 --
>  1 file changed, 2 deletions(-)
> 
> diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c
> b/drivers/net/ixgbe/ixgbe_ethdev.c
> index baaeee5d9..6af4f8b2b 100644
> --- a/drivers/net/ixgbe/ixgbe_ethdev.c
> +++ b/drivers/net/ixgbe/ixgbe_ethdev.c
> @@ -2240,8 +2240,6 @@ ixgbe_check_mq_mode(struct rte_eth_dev *dev)
>   case ETH_MQ_RX_NONE:
>   /* if nothing mq mode configure, use default scheme */
>   dev->data->dev_conf.rxmode.mq_mode =
> ETH_MQ_RX_VMDQ_ONLY;
> - if (RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool > 1)
> - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 1;
>   break;
>   default: /* ETH_MQ_RX_DCB, ETH_MQ_RX_DCB_RSS or
> ETH_MQ_TX_DCB*/
>   /* SRIOV only works in VMDq enable mode */
> --
> 2.11.0

Re: [dpdk-dev] [PATCH 2/2] examples/ipsec-secgw: fix SPI byte order in flow item

[Akhil Goyal]

Acked-by: Akhil Goyal 

[dpdk-dev] [PATCH v1] net/i40e: fix packet type parser issue

[Beilei Xing]

Ptype mapping table will fail to update when loading
PPP profile, fix the issue via modifying metadata and
adding check.

Fixes: ab2e350c4f4b ("net/i40e: improve packet type parser")
Cc: sta...@dpdk.org

Signed-off-by: Beilei Xing 
---
 drivers/net/i40e/i40e_ethdev.c  | 2 +-
 drivers/net/i40e/rte_pmd_i40e.c | 6 --
 2 files changed, 5 insertions(+), 3 deletions(-)

diff --git a/drivers/net/i40e/i40e_ethdev.c b/drivers/net/i40e/i40e_ethdev.c
index a8e9ab2..b5ab036 100644
--- a/drivers/net/i40e/i40e_ethdev.c
+++ b/drivers/net/i40e/i40e_ethdev.c
@@ -11254,7 +11254,7 @@ i40e_update_customized_ptype(struct rte_eth_dev *dev, 
uint8_t *pkg,
continue;
memset(name, 0, sizeof(name));
strcpy(name, proto[n].name);
-   if (!strncmp(name, "PPPOE", 5))
+   if (!strncmp(name, "PPPoE", 5))
ptype_mapping[i].sw_ptype |=
RTE_PTYPE_L2_ETHER_PPPOE;
else if (!strncmp(name, "OIPV4", 5)) {
diff --git a/drivers/net/i40e/rte_pmd_i40e.c b/drivers/net/i40e/rte_pmd_i40e.c
index 55ae2fe..af33a7c 100644
--- a/drivers/net/i40e/rte_pmd_i40e.c
+++ b/drivers/net/i40e/rte_pmd_i40e.c
@@ -2053,7 +2053,8 @@ static int check_invalid_pkt_type(uint32_t pkt_type)
l2 != RTE_PTYPE_L2_ETHER_LLDP &&
l2 != RTE_PTYPE_L2_ETHER_NSH &&
l2 != RTE_PTYPE_L2_ETHER_VLAN &&
-   l2 != RTE_PTYPE_L2_ETHER_QINQ)
+   l2 != RTE_PTYPE_L2_ETHER_QINQ &&
+   l2 != RTE_PTYPE_L2_ETHER_PPPOE)
return -1;
 
if (l3 &&
@@ -2082,7 +2083,8 @@ static int check_invalid_pkt_type(uint32_t pkt_type)
tnl != RTE_PTYPE_TUNNEL_GENEVE &&
tnl != RTE_PTYPE_TUNNEL_GRENAT &&
tnl != RTE_PTYPE_TUNNEL_GTPC &&
-   tnl != RTE_PTYPE_TUNNEL_GTPU)
+   tnl != RTE_PTYPE_TUNNEL_GTPU &&
+   tnl != RTE_PTYPE_TUNNEL_L2TP)
return -1;
 
if (il2 &&
-- 
2.5.5

[dpdk-dev] [PATCH v2] net/i40e: fix flag sent to mac_address_write

[Igor Ryzhov]

Use the same value as in Linux driver.

Fixes: e18e01e92c29 ("i40e: support default MAC address setting")
Cc: sta...@dpdk.org

Signed-off-by: Igor Ryzhov 
---
v2:
* fix checkpatch warning (long line)
* fix commit subject
* add Fixes line
* CC to stable
---
 drivers/net/i40e/i40e_ethdev.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/i40e/i40e_ethdev.c b/drivers/net/i40e/i40e_ethdev.c
index 285d92b..055b9e8 100644
--- a/drivers/net/i40e/i40e_ethdev.c
+++ b/drivers/net/i40e/i40e_ethdev.c
@@ -10795,8 +10795,8 @@ static void i40e_set_default_mac_addr(struct 
rte_eth_dev *dev,
return;
}
 
-   /* Flags: 0x3 updates port address */
-   i40e_aq_mac_address_write(hw, 0x3, mac_addr->addr_bytes, NULL);
+   i40e_aq_mac_address_write(hw, I40E_AQC_WRITE_TYPE_LAA_WOL,
+ mac_addr->addr_bytes, NULL);
 }
 
 static int
-- 
2.6.4

Re: [dpdk-dev] [PATCH] net/i40e: fix VSI MAC filter on primary address change

[Igor Ryzhov]

Sent v2: https://dpdk.org/dev/patchwork/patch/33570/

On Thu, Jan 11, 2018 at 11:21 AM, Xing, Beilei 
wrote:

> Hi Igor,
>
>
>
> Thanks for the catch, and glad to see your patchJ It resolves a potential
> problem in PMD.
>
> The patch looks OK for me except some minor comments (in another mail
> thread).
>
>
>
> Best Regards,
>
> Beilei
>
>
>
> *From:* Igor Ryzhov [mailto:iryz...@nfware.com]
> *Sent:* Thursday, January 11, 2018 6:47 AM
> *To:* Zhang, Helin 
> *Cc:* Xing, Beilei ; Olivier Matz <
> olivier.m...@6wind.com>; dev@dpdk.org; Wu, Jingjing ;
> sta...@dpdk.org; Laurent Hardy 
>
> *Subject:* Re: [dpdk-dev] [PATCH] net/i40e: fix VSI MAC filter on primary
> address change
>
>
>
> Hello everyone.
>
>
>
> It's sad that my comments were unanswered.
>
> I'm ok with the first two – they were mostly style-related.
>
> But I made an investigation on the third one and it is a bug.
>
>
>
> This is a description (from X710 datasheet) of flags sent to
> mac_address_write command:
>
>
>
> By bits:
>
> 0-7 – Reserved
>
> 8 – MAC_MAG_EN
>
> 9 – LAA_WOL_PRESERVE
>
> 10-13 – Reserved
>
> 14-15 – Write type (00 – Update LAA only, 01 – Update LAA and WOL, 10 –
> Update port address, 11 – Reserved, but used in Linux to enable multicast
> magic packet wake up)
>
>
>
> Current code uses 0x3 flag, apparently trying to update LAA, WOL and port
> address, but it sets first two bits instead of last two.
>
> These bits are reserved, that's why it doesn't break anything.
>
> At the same time, last two bits are set to zero, and the command changes
> LAA address only – it's enough to work in simple case.
>
>
>
> The last question – which flag is correct to use – 01 (LAA + WOL) or 11
> (LAA + WOL + port).
>
> Linux driver uses the first one, and here is the patch to fix the issue:
>
> https://dpdk.org/dev/patchwork/patch/33524/
>
>
>
> Best regards,
>
> Igor
>
>
>
> On Wed, Jan 10, 2018 at 4:57 PM, Zhang, Helin 
> wrote:
>
>
>
> > -Original Message-
> > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Xing, Beilei
> > Sent: Thursday, January 4, 2018 1:39 PM
> > To: Olivier Matz; dev@dpdk.org; Wu, Jingjing
> > Cc: sta...@dpdk.org; Laurent Hardy
> > Subject: Re: [dpdk-dev] [PATCH] net/i40e: fix VSI MAC filter on primary
> address
> > change
> >
> >
> >
> > > -Original Message-
> > > From: Olivier Matz [mailto:olivier.m...@6wind.com]
> > > Sent: Wednesday, January 3, 2018 10:29 PM
> > > To: dev@dpdk.org; Wu, Jingjing ; Xing, Beilei
> > > 
> > > Cc: sta...@dpdk.org; Laurent Hardy 
> > > Subject: [PATCH] net/i40e: fix VSI MAC filter on primary address
> > > change
> > >
> > > When primary address mac is changed, the mac filters were not updated
> > > in the VSI with the new mac addr and incoming packets with this
> > > destination address are dropped by the hardware filters.
> > >
> > > This patch removes the VSI mac filter for the previous mac address and
> > > adds a new one for new mac address.
> > >
> > > Fixes: e18e01e92c29 ("i40e: support default MAC address setting")
> > > Cc: sta...@dpdk.org
> > >
> > > Signed-off-by: Laurent Hardy 
> > > Signed-off-by: Olivier Matz 
> >
> > Thanks for the fix.
> > Acked-by: Beilei Xing 
>
> Applied to dpdk-next-net-intel, thanks!
>
> /Helin
>
>
>

[dpdk-dev] [PATCH v5 02/11] eventdev: add API to perform self test

[Pavan Nikhilesh]

Add API to perform self test on the underlying event device driver.

Signed-off-by: Pavan Nikhilesh 
Acked-by: Harry van Haaren 
Acked-by: Jerin Jacob 
---
 lib/librte_eventdev/rte_eventdev.c   | 10 ++
 lib/librte_eventdev/rte_eventdev.h   | 12 
 lib/librte_eventdev/rte_eventdev_pmd.h   | 10 ++
 lib/librte_eventdev/rte_eventdev_version.map |  6 ++
 4 files changed, 38 insertions(+)

diff --git a/lib/librte_eventdev/rte_eventdev.c 
b/lib/librte_eventdev/rte_eventdev.c
index 035bd0017..85496e575 100644
--- a/lib/librte_eventdev/rte_eventdev.c
+++ b/lib/librte_eventdev/rte_eventdev.c
@@ -1114,6 +1114,16 @@ int rte_event_dev_xstats_reset(uint8_t dev_id,
return -ENOTSUP;
 }
 
+int rte_event_dev_selftest(uint8_t dev_id)
+{
+   RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL);
+   struct rte_eventdev *dev = &rte_eventdevs[dev_id];
+
+   if (dev->dev_ops->dev_selftest != NULL)
+   return (*dev->dev_ops->dev_selftest)();
+   return -ENOTSUP;
+}
+
 int
 rte_event_dev_start(uint8_t dev_id)
 {
diff --git a/lib/librte_eventdev/rte_eventdev.h 
b/lib/librte_eventdev/rte_eventdev.h
index 1bbea57db..9201a5af8 100644
--- a/lib/librte_eventdev/rte_eventdev.h
+++ b/lib/librte_eventdev/rte_eventdev.h
@@ -1787,6 +1787,18 @@ rte_event_dev_xstats_reset(uint8_t dev_id,
   const uint32_t ids[],
   uint32_t nb_ids);
 
+/**
+ * Trigger the eventdev self test.
+ *
+ * @param dev_id
+ *   The identifier of the device
+ * @return
+ *   - 0: Selftest successful
+ *   - -ENOTSUP if the device doesn't support selftest
+ *   - other values < 0 on failure.
+ */
+int rte_event_dev_selftest(uint8_t dev_id);
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/lib/librte_eventdev/rte_eventdev_pmd.h 
b/lib/librte_eventdev/rte_eventdev_pmd.h
index 6ed40cd38..7c39a040c 100644
--- a/lib/librte_eventdev/rte_eventdev_pmd.h
+++ b/lib/librte_eventdev/rte_eventdev_pmd.h
@@ -603,6 +603,13 @@ typedef int (*eventdev_eth_rx_adapter_stats_get)
 typedef int (*eventdev_eth_rx_adapter_stats_reset)
(const struct rte_eventdev *dev,
const struct rte_eth_dev *eth_dev);
+/**
+ * Start eventdev selftest.
+ *
+ * @return
+ *   Return 0 on success.
+ */
+typedef int (*eventdev_selftest)(void);
 
 /** Event device operations function pointer table */
 struct rte_eventdev_ops {
@@ -658,6 +665,9 @@ struct rte_eventdev_ops {
/**< Get ethernet Rx stats */
eventdev_eth_rx_adapter_stats_reset eth_rx_adapter_stats_reset;
/**< Reset ethernet Rx stats */
+
+   eventdev_selftest dev_selftest;
+   /**< Start eventdev Selftest */
 };
 
 /**
diff --git a/lib/librte_eventdev/rte_eventdev_version.map 
b/lib/librte_eventdev/rte_eventdev_version.map
index 108ae61fb..2aef470b5 100644
--- a/lib/librte_eventdev/rte_eventdev_version.map
+++ b/lib/librte_eventdev/rte_eventdev_version.map
@@ -68,3 +68,9 @@ DPDK_17.11 {
rte_event_eth_rx_adapter_stop;
 
 } DPDK_17.08;
+
+DPDK_18.02 {
+   global:
+
+   rte_event_dev_selftest;
+} DPDK_17.11;
-- 
2.15.1

[dpdk-dev] [PATCH v5 04/11] event/octeontx: modify octeontx eventdev test

[Pavan Nikhilesh]

Modify test_eventdev_octeontx to be standalone selftest independent of
test framework.

Signed-off-by: Pavan Nikhilesh 
---
 drivers/event/octeontx/ssovf_evdev_selftest.c | 425 ++
 1 file changed, 232 insertions(+), 193 deletions(-)

diff --git a/drivers/event/octeontx/ssovf_evdev_selftest.c 
b/drivers/event/octeontx/ssovf_evdev_selftest.c
index 8fddb4fd2..325c110c8 100644
--- a/drivers/event/octeontx/ssovf_evdev_selftest.c
+++ b/drivers/event/octeontx/ssovf_evdev_selftest.c
@@ -46,12 +46,21 @@
 #include 
 #include 
 #include 
+#include 
 
-#include "test.h"
+#include "ssovf_evdev.h"
 
 #define NUM_PACKETS (1 << 18)
 #define MAX_EVENTS  (16 * 1024)
 
+#define OCTEONTX_TEST_RUN(setup, teardown, test) \
+   octeontx_test_run(setup, teardown, test, #test)
+
+static int total;
+static int passed;
+static int failed;
+static int unsupported;
+
 static int evdev;
 static struct rte_mempool *eventdev_test_mempool;
 
@@ -79,11 +88,11 @@ static inline int
 seqn_list_update(int val)
 {
if (seqn_list_index >= NUM_PACKETS)
-   return TEST_FAILED;
+   return -1;
 
seqn_list[seqn_list_index++] = val;
rte_smp_wmb();
-   return TEST_SUCCESS;
+   return 0;
 }
 
 static inline int
@@ -93,11 +102,11 @@ seqn_list_check(int limit)
 
for (i = 0; i < limit; i++) {
if (seqn_list[i] != i) {
-   printf("Seqn mismatch %d %d\n", seqn_list[i], i);
-   return TEST_FAILED;
+   ssovf_log_dbg("Seqn mismatch %d %d", seqn_list[i], i);
+   return -1;
}
}
-   return TEST_SUCCESS;
+   return 0;
 }
 
 struct test_core_param {
@@ -114,20 +123,21 @@ testsuite_setup(void)
 
evdev = rte_event_dev_get_dev_id(eventdev_name);
if (evdev < 0) {
-   printf("%d: Eventdev %s not found - creating.\n",
+   ssovf_log_dbg("%d: Eventdev %s not found - creating.",
__LINE__, eventdev_name);
if (rte_vdev_init(eventdev_name, NULL) < 0) {
-   printf("Error creating eventdev %s\n", eventdev_name);
-   return TEST_FAILED;
+   ssovf_log_dbg("Error creating eventdev %s",
+   eventdev_name);
+   return -1;
}
evdev = rte_event_dev_get_dev_id(eventdev_name);
if (evdev < 0) {
-   printf("Error finding newly created eventdev\n");
-   return TEST_FAILED;
+   ssovf_log_dbg("Error finding newly created eventdev");
+   return -1;
}
}
 
-   return TEST_SUCCESS;
+   return 0;
 }
 
 static void
@@ -177,31 +187,32 @@ _eventdev_setup(int mode)
512, /* Use very small mbufs */
rte_socket_id());
if (!eventdev_test_mempool) {
-   printf("ERROR creating mempool\n");
-   return TEST_FAILED;
+   ssovf_log_dbg("ERROR creating mempool");
+   return -1;
}
 
ret = rte_event_dev_info_get(evdev, &info);
-   TEST_ASSERT_SUCCESS(ret, "Failed to get event dev info");
-   TEST_ASSERT(info.max_num_events >= (int32_t)MAX_EVENTS,
-   "max_num_events=%d < max_events=%d",
-   info.max_num_events, MAX_EVENTS);
+   RTE_TEST_ASSERT_SUCCESS(ret, "Failed to get event dev info");
+   RTE_TEST_ASSERT(info.max_num_events >= (int32_t)MAX_EVENTS,
+   "ERROR max_num_events=%d < max_events=%d",
+   info.max_num_events, MAX_EVENTS);
 
devconf_set_default_sane_values(&dev_conf, &info);
if (mode == TEST_EVENTDEV_SETUP_DEQUEUE_TIMEOUT)
dev_conf.event_dev_cfg |= RTE_EVENT_DEV_CFG_PER_DEQUEUE_TIMEOUT;
 
ret = rte_event_dev_configure(evdev, &dev_conf);
-   TEST_ASSERT_SUCCESS(ret, "Failed to configure eventdev");
+   RTE_TEST_ASSERT_SUCCESS(ret, "Failed to configure eventdev");
 
uint32_t queue_count;
-   TEST_ASSERT_SUCCESS(rte_event_dev_attr_get(evdev,
+   RTE_TEST_ASSERT_SUCCESS(rte_event_dev_attr_get(evdev,
RTE_EVENT_DEV_ATTR_QUEUE_COUNT,
&queue_count), "Queue count get failed");
 
if (mode == TEST_EVENTDEV_SETUP_PRIORITY) {
if (queue_count > 8) {
-   printf("test expects the unique priority per queue\n");
+   ssovf_log_dbg(
+   "test expects the unique priority per queue");
return -ENOTSUP;
}
 
@@ -216,35 +227,39 @@ _eventdev_setup(int mode)
 
ret = rte_event_queue_default_conf_get(evdev, i,
 

[dpdk-dev] [PATCH v5 03/11] event/octeontx: move eventdev octeontx test to driver

[Pavan Nikhilesh]

Move octeontx eventdev specific test (test_eventdev_octeontx.c) to
driver/event/octeontx.

Signed-off-by: Pavan Nikhilesh 
Acked-by: Jerin Jacob 
---
 .../event/octeontx/ssovf_evdev_selftest.c| 0
 test/test/Makefile   | 1 -
 2 files changed, 1 deletion(-)
 rename test/test/test_eventdev_octeontx.c => 
drivers/event/octeontx/ssovf_evdev_selftest.c (100%)

diff --git a/test/test/test_eventdev_octeontx.c 
b/drivers/event/octeontx/ssovf_evdev_selftest.c
similarity index 100%
rename from test/test/test_eventdev_octeontx.c
rename to drivers/event/octeontx/ssovf_evdev_selftest.c
diff --git a/test/test/Makefile b/test/test/Makefile
index bb54c9808..87e3169d2 100644
--- a/test/test/Makefile
+++ b/test/test/Makefile
@@ -211,7 +211,6 @@ SRCS-y += test_eventdev.c
 SRCS-y += test_event_ring.c
 SRCS-y += test_event_eth_rx_adapter.c
 SRCS-$(CONFIG_RTE_LIBRTE_PMD_SW_EVENTDEV) += test_eventdev_sw.c
-SRCS-$(CONFIG_RTE_LIBRTE_PMD_OCTEONTX_SSOVF) += test_eventdev_octeontx.c
 endif
 
 SRCS-$(CONFIG_RTE_LIBRTE_KVARGS) += test_kvargs.c
-- 
2.15.1

[dpdk-dev] [PATCH v5 01/11] eal: add common test assert macros

[Pavan Nikhilesh]

Adding common test assertion macros for unit testing.
Replaced common macros in test/test.h with new RTE_TEST_ASSERT_* macros.

Signed-off-by: Pavan Nikhilesh 
Acked-by: Jerin Jacob 
---

 v5 Changes:
  - rebase patchset
  - remove duplication between test/test/test.h and rte_test.h by mapping
  RTE_TEST_ASSERT* macros to test.h

 v4 Changes:
  - add SPDX licence tags.
  - change octeontx selftest name to 'ssovf_evdev_selftest'

 v3 Changes:
  - add eventdev driver specific selftest to test/test

 v2 Changes:
  - remove duplications of test macros.
  - add selftest to test/test.
  - remove selftest devargs from sw eventdev.

 lib/librte_eal/common/Makefile   |  2 +-
 lib/librte_eal/common/include/rte_test.h | 46 +
 test/test/test.h | 70 +---
 3 files changed, 58 insertions(+), 60 deletions(-)
 create mode 100644 lib/librte_eal/common/include/rte_test.h

diff --git a/lib/librte_eal/common/Makefile b/lib/librte_eal/common/Makefile
index 9effd0d45..eba1059f2 100644
--- a/lib/librte_eal/common/Makefile
+++ b/lib/librte_eal/common/Makefile
@@ -43,7 +43,7 @@ INC += rte_hexdump.h rte_devargs.h rte_bus.h rte_dev.h
 INC += rte_pci_dev_feature_defs.h rte_pci_dev_features.h
 INC += rte_malloc.h rte_keepalive.h rte_time.h
 INC += rte_service.h rte_service_component.h
-INC += rte_bitmap.h rte_vfio.h
+INC += rte_bitmap.h rte_vfio.h rte_test.h

 GENERIC_INC := rte_atomic.h rte_byteorder.h rte_cycles.h rte_prefetch.h
 GENERIC_INC += rte_spinlock.h rte_memcpy.h rte_cpuflags.h rte_rwlock.h
diff --git a/lib/librte_eal/common/include/rte_test.h 
b/lib/librte_eal/common/include/rte_test.h
new file mode 100644
index 0..89e47f47a
--- /dev/null
+++ b/lib/librte_eal/common/include/rte_test.h
@@ -0,0 +1,46 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2015 Cavium, Inc
+ */
+
+#ifndef _RTE_TEST_H_
+#define _RTE_TEST_H_
+
+#include 
+
+/* Before including rte_test.h file you can define
+ * RTE_TEST_TRACE_FAILURE(_file, _line, _func) macro to better trace/debug test
+ * failures. Mostly useful in development phase.
+ */
+#ifndef RTE_TEST_TRACE_FAILURE
+#define RTE_TEST_TRACE_FAILURE(_file, _line, _func)
+#endif
+
+
+#define RTE_TEST_ASSERT(cond, msg, ...) do {  \
+   if (!(cond)) {\
+   RTE_LOG(DEBUG, EAL, "Test assert %s line %d failed: " \
+   msg "\n", __func__, __LINE__, ##__VA_ARGS__); \
+   RTE_TEST_TRACE_FAILURE(__FILE__, __LINE__, __func__); \
+   return -1;\
+   } \
+} while (0)
+
+#define RTE_TEST_ASSERT_EQUAL(a, b, msg, ...) \
+   RTE_TEST_ASSERT(a == b, msg, ##__VA_ARGS__)
+
+#define RTE_TEST_ASSERT_NOT_EQUAL(a, b, msg, ...) \
+   RTE_TEST_ASSERT(a != b, msg, ##__VA_ARGS__)
+
+#define RTE_TEST_ASSERT_SUCCESS(val, msg, ...) \
+   RTE_TEST_ASSERT(val == 0, msg, ##__VA_ARGS__)
+
+#define RTE_TEST_ASSERT_FAIL(val, msg, ...) \
+   RTE_TEST_ASSERT(val != 0, msg, ##__VA_ARGS__)
+
+#define RTE_TEST_ASSERT_NULL(val, msg, ...) \
+   RTE_TEST_ASSERT(val == NULL, msg, ##__VA_ARGS__)
+
+#define RTE_TEST_ASSERT_NOT_NULL(val, msg, ...) \
+   RTE_TEST_ASSERT(val != NULL, msg, ##__VA_ARGS__)
+
+#endif /* _RTE_TEST_H_ */
diff --git a/test/test/test.h b/test/test/test.h
index 08ffe949c..0ea18586c 100644
--- a/test/test/test.h
+++ b/test/test/test.h
@@ -38,7 +38,6 @@
 #include 

 #include 
-#include 

 #define TEST_SUCCESS  (0)
 #define TEST_FAILED  (-1)
@@ -50,23 +49,13 @@
 # define TEST_TRACE_FAILURE(_file, _line, _func)
 #endif

-#define TEST_ASSERT(cond, msg, ...) do { \
-   if (!(cond)) {   \
-   printf("TestCase %s() line %d failed: "  \
-   msg "\n", __func__, __LINE__, ##__VA_ARGS__);   
 \
-   TEST_TRACE_FAILURE(__FILE__, __LINE__, __func__);\
-   return TEST_FAILED;  \
-   }\
-} while (0)
+#define RTE_TEST_TRACE_FAILURE TEST_TRACE_FAILURE

-#define TEST_ASSERT_EQUAL(a, b, msg, ...) do {   \
-   if (!(a == b)) { \
-   printf("TestCase %s() line %d failed: "  \
-   msg "\n", __func__, __LINE__, ##__VA_ARGS__);   
 \
-   TEST_TRACE_FAILURE(__FILE__, __LINE__, __func__);\
-   return TEST_FAILED;  \
-   }\
-} while (0)
+#include 
+
+#define TEST_ASSERT RTE_TEST_ASSERT
+
+#define T

[dpdk-dev] [PATCH v5 05/11] event/octeontx: update octeontx eventdev selftest ops

[Pavan Nikhilesh]

Update octeontx eventdev ops to invoke selftest when application
invokes `rte_event_dev_selftest`.

Signed-off-by: Pavan Nikhilesh 
Acked-by: Jerin Jacob 
---
 drivers/event/octeontx/Makefile  | 3 ++-
 drivers/event/octeontx/ssovf_evdev.c | 2 ++
 drivers/event/octeontx/ssovf_evdev.h | 4 
 3 files changed, 8 insertions(+), 1 deletion(-)

diff --git a/drivers/event/octeontx/Makefile b/drivers/event/octeontx/Makefile
index fdf1b7385..9a6ffdfb1 100644
--- a/drivers/event/octeontx/Makefile
+++ b/drivers/event/octeontx/Makefile
@@ -42,7 +42,7 @@ CFLAGS += -I$(RTE_SDK)/drivers/mempool/octeontx/
 CFLAGS += -I$(RTE_SDK)/drivers/net/octeontx/
 
 LDLIBS += -lrte_eal -lrte_eventdev -lrte_mempool_octeontx
-LDLIBS += -lrte_bus_pci
+LDLIBS += -lrte_bus_pci -lrte_mempool -lrte_mbuf
 LDLIBS += -lrte_bus_vdev
 
 EXPORT_MAP := rte_pmd_octeontx_ssovf_version.map
@@ -54,6 +54,7 @@ LIBABIVER := 1
 #
 SRCS-$(CONFIG_RTE_LIBRTE_PMD_OCTEONTX_SSOVF) += ssovf_worker.c
 SRCS-$(CONFIG_RTE_LIBRTE_PMD_OCTEONTX_SSOVF) += ssovf_evdev.c
+SRCS-$(CONFIG_RTE_LIBRTE_PMD_OCTEONTX_SSOVF) += ssovf_evdev_selftest.c
 
 ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y)
 CFLAGS_ssovf_worker.o += -fno-prefetch-loop-arrays
diff --git a/drivers/event/octeontx/ssovf_evdev.c 
b/drivers/event/octeontx/ssovf_evdev.c
index b80a6c0ae..0a96fbd7f 100644
--- a/drivers/event/octeontx/ssovf_evdev.c
+++ b/drivers/event/octeontx/ssovf_evdev.c
@@ -613,6 +613,8 @@ static const struct rte_eventdev_ops ssovf_ops = {
.eth_rx_adapter_start = ssovf_eth_rx_adapter_start,
.eth_rx_adapter_stop = ssovf_eth_rx_adapter_stop,
 
+   .dev_selftest = test_eventdev_octeontx,
+
.dump = ssovf_dump,
.dev_start= ssovf_start,
.dev_stop = ssovf_stop,
diff --git a/drivers/event/octeontx/ssovf_evdev.h 
b/drivers/event/octeontx/ssovf_evdev.h
index b093a3e73..72a980447 100644
--- a/drivers/event/octeontx/ssovf_evdev.h
+++ b/drivers/event/octeontx/ssovf_evdev.h
@@ -57,6 +57,9 @@
 #define ssovf_log_err(fmt, args...) \
RTE_LOG(ERR, EVENTDEV, "[%s] %s() " fmt "\n", \
RTE_STR(EVENTDEV_NAME_OCTEONTX_PMD), __func__, ## args)
+#define ssovf_log_selftest(fmt, args...) \
+   RTE_LOG(INFO, EVENTDEV, "[%s] %s() " fmt "\n", \
+   RTE_STR(EVENTDEV_NAME_OCTEONTX_PMD), __func__, ## args)
 
 #define SSO_MAX_VHGRP (64)
 #define SSO_MAX_VHWS  (32)
@@ -196,5 +199,6 @@ uint16_t ssows_deq_timeout_burst(void *port, struct 
rte_event ev[],
uint16_t nb_events, uint64_t timeout_ticks);
 void ssows_flush_events(struct ssows *ws, uint8_t queue_id);
 void ssows_reset(struct ssows *ws);
+int test_eventdev_octeontx(void);
 
 #endif /* __SSOVF_EVDEV_H__ */
-- 
2.15.1

[dpdk-dev] [PATCH v5 06/11] event/octeontx: add selftest to device arguments

[Pavan Nikhilesh]

Add selftest as a device argument that can be enabled by suppling
'self_test=1' as a vdev parameter

--vdev="event_octeontx,self_test=1"

The selftest is run after vdev creation is successfully
complete.

Signed-off-by: Pavan Nikhilesh 
Acked-by: Jerin Jacob 
---
 drivers/event/octeontx/Makefile  |  2 +-
 drivers/event/octeontx/ssovf_evdev.c | 41 
 drivers/event/octeontx/ssovf_evdev.h |  2 ++
 3 files changed, 44 insertions(+), 1 deletion(-)

diff --git a/drivers/event/octeontx/Makefile b/drivers/event/octeontx/Makefile
index 9a6ffdfb1..ff2a11793 100644
--- a/drivers/event/octeontx/Makefile
+++ b/drivers/event/octeontx/Makefile
@@ -42,7 +42,7 @@ CFLAGS += -I$(RTE_SDK)/drivers/mempool/octeontx/
 CFLAGS += -I$(RTE_SDK)/drivers/net/octeontx/
 
 LDLIBS += -lrte_eal -lrte_eventdev -lrte_mempool_octeontx
-LDLIBS += -lrte_bus_pci -lrte_mempool -lrte_mbuf
+LDLIBS += -lrte_bus_pci -lrte_mempool -lrte_mbuf -lrte_kvargs
 LDLIBS += -lrte_bus_vdev
 
 EXPORT_MAP := rte_pmd_octeontx_ssovf_version.map
diff --git a/drivers/event/octeontx/ssovf_evdev.c 
b/drivers/event/octeontx/ssovf_evdev.c
index 0a96fbd7f..f092d352b 100644
--- a/drivers/event/octeontx/ssovf_evdev.c
+++ b/drivers/event/octeontx/ssovf_evdev.c
@@ -38,6 +38,7 @@
 #include 
 #include 
 #include 
+#include 
 #include 
 #include 
 #include 
@@ -593,6 +594,15 @@ ssovf_close(struct rte_eventdev *dev)
return 0;
 }
 
+static int
+ssovf_selftest(const char *key __rte_unused, const char *value,
+   void *opaque)
+{
+   int *flag = opaque;
+   *flag = !!atoi(value);
+   return 0;
+}
+
 /* Initialize and register event driver with DPDK Application */
 static const struct rte_eventdev_ops ssovf_ops = {
.dev_infos_get= ssovf_info_get,
@@ -630,7 +640,14 @@ ssovf_vdev_probe(struct rte_vdev_device *vdev)
struct rte_eventdev *eventdev;
static int ssovf_init_once;
const char *name;
+   const char *params;
int ret;
+   int selftest = 0;
+
+   static const char *const args[] = {
+   SSOVF_SELFTEST_ARG,
+   NULL
+   };
 
name = rte_vdev_device_name(vdev);
/* More than one instance is not supported */
@@ -639,6 +656,28 @@ ssovf_vdev_probe(struct rte_vdev_device *vdev)
return -EINVAL;
}
 
+   params = rte_vdev_device_args(vdev);
+   if (params != NULL && params[0] != '\0') {
+   struct rte_kvargs *kvlist = rte_kvargs_parse(params, args);
+
+   if (!kvlist) {
+   ssovf_log_info(
+   "Ignoring unsupported params supplied '%s'",
+   name);
+   } else {
+   int ret = rte_kvargs_process(kvlist,
+   SSOVF_SELFTEST_ARG,
+   ssovf_selftest, &selftest);
+   if (ret != 0) {
+   ssovf_log_err("%s: Error in selftest", name);
+   rte_kvargs_free(kvlist);
+   return ret;
+   }
+   }
+
+   rte_kvargs_free(kvlist);
+   }
+
eventdev = rte_event_pmd_vdev_init(name, sizeof(struct ssovf_evdev),
rte_socket_id());
if (eventdev == NULL) {
@@ -689,6 +728,8 @@ ssovf_vdev_probe(struct rte_vdev_device *vdev)
edev->max_event_ports);
 
ssovf_init_once = 1;
+   if (selftest)
+   test_eventdev_octeontx();
return 0;
 
 error:
diff --git a/drivers/event/octeontx/ssovf_evdev.h 
b/drivers/event/octeontx/ssovf_evdev.h
index 72a980447..018fbdfa5 100644
--- a/drivers/event/octeontx/ssovf_evdev.h
+++ b/drivers/event/octeontx/ssovf_evdev.h
@@ -117,6 +117,8 @@
 #define SSO_GRP_GET_PRIORITY  0x7
 #define SSO_GRP_SET_PRIORITY  0x8
 
+#define SSOVF_SELFTEST_ARG   ("selftest")
+
 /*
  * In Cavium OcteonTX SoC, all accesses to the device registers are
  * implictly strongly ordered. So, The relaxed version of IO operation is
-- 
2.15.1

[dpdk-dev] [PATCH v5 08/11] event/sw: modify eventdev software test

[Pavan Nikhilesh]

Modify test_eventdev_sw to be standalone selftest independent of test
framework.

Signed-off-by: Pavan Nikhilesh 
Acked-by: Harry van Haaren 
---
 drivers/event/sw/sw_evdev_selftest.c | 70 ++--
 1 file changed, 36 insertions(+), 34 deletions(-)

diff --git a/drivers/event/sw/sw_evdev_selftest.c 
b/drivers/event/sw/sw_evdev_selftest.c
index 4108b00bf..0e84b2003 100644
--- a/drivers/event/sw/sw_evdev_selftest.c
+++ b/drivers/event/sw/sw_evdev_selftest.c
@@ -52,7 +52,7 @@
 #include 
 #include 
 
-#include "test.h"
+#include "sw_evdev.h"
 
 #define MAX_PORTS 16
 #define MAX_QIDS 16
@@ -3042,7 +3042,7 @@ worker_loopback(struct test *t, uint8_t 
disable_implicit_release)
 
 static struct rte_mempool *eventdev_func_mempool;
 
-static int
+int
 test_sw_eventdev(void)
 {
struct test *t = malloc(sizeof(struct test));
@@ -3053,25 +3053,25 @@ test_sw_eventdev(void)
 */
release_ev.op = RTE_EVENT_OP_RELEASE;
 
-   const char *eventdev_name = "event_sw0";
+   const char *eventdev_name = "event_sw";
evdev = rte_event_dev_get_dev_id(eventdev_name);
if (evdev < 0) {
printf("%d: Eventdev %s not found - creating.\n",
__LINE__, eventdev_name);
if (rte_vdev_init(eventdev_name, NULL) < 0) {
printf("Error creating eventdev\n");
-   return -1;
+   goto test_fail;
}
evdev = rte_event_dev_get_dev_id(eventdev_name);
if (evdev < 0) {
printf("Error finding newly created eventdev\n");
-   return -1;
+   goto test_fail;
}
}
 
if (rte_event_dev_service_id_get(evdev, &t->service_id) < 0) {
printf("Failed to get service ID for software event dev\n");
-   return -1;
+   goto test_fail;
}
 
rte_service_runstate_set(t->service_id, 1);
@@ -3088,7 +3088,7 @@ test_sw_eventdev(void)
rte_socket_id());
if (!eventdev_func_mempool) {
printf("ERROR creating mempool\n");
-   return -1;
+   goto test_fail;
}
}
t->mbuf_pool = eventdev_func_mempool;
@@ -3096,146 +3096,146 @@ test_sw_eventdev(void)
ret = test_single_directed_packet(t);
if (ret != 0) {
printf("ERROR - Single Directed Packet test FAILED.\n");
-   return ret;
+   goto test_fail;
}
printf("*** Running Directed Forward Credit test...\n");
ret = test_directed_forward_credits(t);
if (ret != 0) {
printf("ERROR - Directed Forward Credit test FAILED.\n");
-   return ret;
+   goto test_fail;
}
printf("*** Running Single Load Balanced Packet test...\n");
ret = single_packet(t);
if (ret != 0) {
printf("ERROR - Single Packet test FAILED.\n");
-   return ret;
+   goto test_fail;
}
printf("*** Running Unordered Basic test...\n");
ret = unordered_basic(t);
if (ret != 0) {
printf("ERROR -  Unordered Basic test FAILED.\n");
-   return ret;
+   goto test_fail;
}
printf("*** Running Ordered Basic test...\n");
ret = ordered_basic(t);
if (ret != 0) {
printf("ERROR -  Ordered Basic test FAILED.\n");
-   return ret;
+   goto test_fail;
}
printf("*** Running Burst Packets test...\n");
ret = burst_packets(t);
if (ret != 0) {
printf("ERROR - Burst Packets test FAILED.\n");
-   return ret;
+   goto test_fail;
}
printf("*** Running Load Balancing test...\n");
ret = load_balancing(t);
if (ret != 0) {
printf("ERROR - Load Balancing test FAILED.\n");
-   return ret;
+   goto test_fail;
}
printf("*** Running Prioritized Directed test...\n");
ret = test_priority_directed(t);
if (ret != 0) {
printf("ERROR - Prioritized Directed test FAILED.\n");
-   return ret;
+   goto test_fail;
}
printf("*** Running Prioritized Atomic test...\n");
ret = test_priority_atomic(t);
if (ret != 0) {
printf("ERROR - Prioritized Atomic test FAILED.\n");
-   return ret;
+   goto test_fail;
}
 
printf("*** Running Prioritized Ordered test...\n");
ret = test_priority_ordered(t);
if (ret != 0) {
printf("ERROR - Prioritized Ordered test FAILED.\n");
-   return ret;
+   goto test_fail;
}
printf("*** Runnin

[dpdk-dev] [PATCH v5 10/11] test: register eventdev selftest

[Pavan Nikhilesh]

Signed-off-by: Pavan Nikhilesh 
Acked-by: Jerin Jacob 
---
 test/test/test_eventdev.c | 22 ++
 1 file changed, 22 insertions(+)

diff --git a/test/test/test_eventdev.c b/test/test/test_eventdev.c
index f8ee1bef6..7465251dc 100644
--- a/test/test/test_eventdev.c
+++ b/test/test/test_eventdev.c
@@ -1018,4 +1018,26 @@ test_eventdev_common(void)
return unit_test_suite_runner(&eventdev_common_testsuite);
 }
 
+static int
+test_eventdev_selftest_impl(const char *pmd, const char *opts)
+{
+   rte_vdev_init(pmd, opts);
+   return rte_event_dev_selftest(rte_event_dev_get_dev_id(pmd));
+}
+
+static int
+test_eventdev_selftest_sw(void)
+{
+   return test_eventdev_selftest_impl("event_sw", "");
+}
+
+static int
+test_eventdev_selftest_octeontx(void)
+{
+   return test_eventdev_selftest_impl("event_octeontx", "");
+}
+
 REGISTER_TEST_COMMAND(eventdev_common_autotest, test_eventdev_common);
+REGISTER_TEST_COMMAND(eventdev_selftest_sw, test_eventdev_selftest_sw);
+REGISTER_TEST_COMMAND(eventdev_selftest_octeontx,
+   test_eventdev_selftest_octeontx);
-- 
2.15.1

[dpdk-dev] [PATCH v5 09/11] event/sw: update software eventdev selftest ops

[Pavan Nikhilesh]

Update software eventdev ops to invoke selftest when application invokes
`rte_event_dev_selftest`.

Signed-off-by: Pavan Nikhilesh 
Acked-by: Harry van Haaren 
---
 drivers/event/sw/Makefile   | 2 ++
 drivers/event/sw/sw_evdev.c | 2 ++
 drivers/event/sw/sw_evdev.h | 1 +
 3 files changed, 5 insertions(+)

diff --git a/drivers/event/sw/Makefile b/drivers/event/sw/Makefile
index 2f2b67bac..149258748 100644
--- a/drivers/event/sw/Makefile
+++ b/drivers/event/sw/Makefile
@@ -44,6 +44,7 @@ CFLAGS += -Wno-missing-field-initializers
 endif
 endif
 LDLIBS += -lrte_eal -lrte_eventdev -lrte_kvargs -lrte_ring
+LDLIBS += -lrte_mempool -lrte_mbuf
 LDLIBS += -lrte_bus_vdev
 
 # library version
@@ -57,6 +58,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_PMD_SW_EVENTDEV) += sw_evdev.c
 SRCS-$(CONFIG_RTE_LIBRTE_PMD_SW_EVENTDEV) += sw_evdev_worker.c
 SRCS-$(CONFIG_RTE_LIBRTE_PMD_SW_EVENTDEV) += sw_evdev_scheduler.c
 SRCS-$(CONFIG_RTE_LIBRTE_PMD_SW_EVENTDEV) += sw_evdev_xstats.c
+SRCS-$(CONFIG_RTE_LIBRTE_PMD_SW_EVENTDEV) += sw_evdev_selftest.c
 
 # export include files
 SYMLINK-y-include +=
diff --git a/drivers/event/sw/sw_evdev.c b/drivers/event/sw/sw_evdev.c
index 1ef634093..f40597aae 100644
--- a/drivers/event/sw/sw_evdev.c
+++ b/drivers/event/sw/sw_evdev.c
@@ -794,6 +794,8 @@ sw_probe(struct rte_vdev_device *vdev)
.xstats_get_names = sw_xstats_get_names,
.xstats_get_by_name = sw_xstats_get_by_name,
.xstats_reset = sw_xstats_reset,
+
+   .dev_selftest = test_sw_eventdev,
};
 
static const char *const args[] = {
diff --git a/drivers/event/sw/sw_evdev.h b/drivers/event/sw/sw_evdev.h
index d08f7d042..b1d2c371c 100644
--- a/drivers/event/sw/sw_evdev.h
+++ b/drivers/event/sw/sw_evdev.h
@@ -334,5 +334,6 @@ int sw_xstats_reset(struct rte_eventdev *dev,
const uint32_t ids[],
uint32_t nb_ids);
 
+int test_sw_eventdev(void);
 
 #endif /* _SW_EVDEV_H_ */
-- 
2.15.1

[dpdk-dev] [PATCH v5 07/11] event/sw: move eventdev software test to driver

[Pavan Nikhilesh]

Move software eventdev specific test (test_eventdev_sw) to
driver/event/sw/.

Signed-off-by: Pavan Nikhilesh 
Acked-by: Harry van Haaren 
---
 test/test/test_eventdev_sw.c => drivers/event/sw/sw_evdev_selftest.c | 0
 test/test/Makefile   | 1 -
 2 files changed, 1 deletion(-)
 rename test/test/test_eventdev_sw.c => drivers/event/sw/sw_evdev_selftest.c 
(100%)

diff --git a/test/test/test_eventdev_sw.c b/drivers/event/sw/sw_evdev_selftest.c
similarity index 100%
rename from test/test/test_eventdev_sw.c
rename to drivers/event/sw/sw_evdev_selftest.c
diff --git a/test/test/Makefile b/test/test/Makefile
index 87e3169d2..c551ad964 100644
--- a/test/test/Makefile
+++ b/test/test/Makefile
@@ -210,7 +210,6 @@ ifeq ($(CONFIG_RTE_LIBRTE_EVENTDEV),y)
 SRCS-y += test_eventdev.c
 SRCS-y += test_event_ring.c
 SRCS-y += test_event_eth_rx_adapter.c
-SRCS-$(CONFIG_RTE_LIBRTE_PMD_SW_EVENTDEV) += test_eventdev_sw.c
 endif
 
 SRCS-$(CONFIG_RTE_LIBRTE_KVARGS) += test_kvargs.c
-- 
2.15.1

[dpdk-dev] [PATCH v5 11/11] doc: update eventdev documentation

[Pavan Nikhilesh]

Update octeontx eventdev documentation to include new selftest feature.

Signed-off-by: Pavan Nikhilesh 
Acked-by: Marko Kovacevic 
---
 doc/guides/eventdevs/octeontx.rst | 13 +
 1 file changed, 13 insertions(+)

diff --git a/doc/guides/eventdevs/octeontx.rst 
b/doc/guides/eventdevs/octeontx.rst
index cef004a20..bf2985be1 100644
--- a/doc/guides/eventdevs/octeontx.rst
+++ b/doc/guides/eventdevs/octeontx.rst
@@ -114,6 +114,19 @@ Example:
 
 ./your_eventdev_application --vdev="event_octeontx"
 
+
+Selftest
+
+
+The functionality of octeontx eventdev can be verified using this option,
+various unit and functional tests are run to verify the sanity.
+The tests are run once the vdev creation is successfully complete.
+
+.. code-block:: console
+
+--vdev="event_octeontx,self_test=1"
+
+
 Limitations
 ---
 
-- 
2.15.1

Re: [dpdk-dev] [PATCH v5 01/11] eal: add common test assert macros

[Pavan Nikhilesh]

+Cc: tho...@monjalon.net
On Thu, Jan 11, 2018 at 03:51:46PM +0530, Pavan Nikhilesh wrote:
> Adding common test assertion macros for unit testing.
> Replaced common macros in test/test.h with new RTE_TEST_ASSERT_* macros.
>
> Signed-off-by: Pavan Nikhilesh 
> Acked-by: Jerin Jacob 
> ---
>
>  v5 Changes:
>   - rebase patchset
>   - remove duplication between test/test/test.h and rte_test.h by mapping
>   RTE_TEST_ASSERT* macros to test.h
>
>  v4 Changes:
>   - add SPDX licence tags.
>   - change octeontx selftest name to 'ssovf_evdev_selftest'
>
>  v3 Changes:
>   - add eventdev driver specific selftest to test/test
>
>  v2 Changes:
>   - remove duplications of test macros.
>   - add selftest to test/test.
>   - remove selftest devargs from sw eventdev.
>
>  lib/librte_eal/common/Makefile   |  2 +-
>  lib/librte_eal/common/include/rte_test.h | 46 +
>  test/test/test.h | 70 
> +---
>  3 files changed, 58 insertions(+), 60 deletions(-)
>  create mode 100644 lib/librte_eal/common/include/rte_test.h
>
> diff --git a/lib/librte_eal/common/Makefile b/lib/librte_eal/common/Makefile
> index 9effd0d45..eba1059f2 100644
> --- a/lib/librte_eal/common/Makefile
> +++ b/lib/librte_eal/common/Makefile
> @@ -43,7 +43,7 @@ INC += rte_hexdump.h rte_devargs.h rte_bus.h rte_dev.h
>  INC += rte_pci_dev_feature_defs.h rte_pci_dev_features.h
>  INC += rte_malloc.h rte_keepalive.h rte_time.h
>  INC += rte_service.h rte_service_component.h
> -INC += rte_bitmap.h rte_vfio.h
> +INC += rte_bitmap.h rte_vfio.h rte_test.h
>
>  GENERIC_INC := rte_atomic.h rte_byteorder.h rte_cycles.h rte_prefetch.h
>  GENERIC_INC += rte_spinlock.h rte_memcpy.h rte_cpuflags.h rte_rwlock.h
> diff --git a/lib/librte_eal/common/include/rte_test.h 
> b/lib/librte_eal/common/include/rte_test.h
> new file mode 100644
> index 0..89e47f47a
> --- /dev/null
> +++ b/lib/librte_eal/common/include/rte_test.h
> @@ -0,0 +1,46 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(c) 2015 Cavium, Inc
> + */
> +
> +#ifndef _RTE_TEST_H_
> +#define _RTE_TEST_H_
> +
> +#include 
> +
> +/* Before including rte_test.h file you can define
> + * RTE_TEST_TRACE_FAILURE(_file, _line, _func) macro to better trace/debug 
> test
> + * failures. Mostly useful in development phase.
> + */
> +#ifndef RTE_TEST_TRACE_FAILURE
> +#define RTE_TEST_TRACE_FAILURE(_file, _line, _func)
> +#endif
> +
> +
> +#define RTE_TEST_ASSERT(cond, msg, ...) do { 
>  \
> + if (!(cond)) {\
> + RTE_LOG(DEBUG, EAL, "Test assert %s line %d failed: " \
> + msg "\n", __func__, __LINE__, ##__VA_ARGS__); \
> + RTE_TEST_TRACE_FAILURE(__FILE__, __LINE__, __func__); \
> + return -1;\
> + } \
> +} while (0)
> +
> +#define RTE_TEST_ASSERT_EQUAL(a, b, msg, ...) \
> + RTE_TEST_ASSERT(a == b, msg, ##__VA_ARGS__)
> +
> +#define RTE_TEST_ASSERT_NOT_EQUAL(a, b, msg, ...) \
> + RTE_TEST_ASSERT(a != b, msg, ##__VA_ARGS__)
> +
> +#define RTE_TEST_ASSERT_SUCCESS(val, msg, ...) \
> + RTE_TEST_ASSERT(val == 0, msg, ##__VA_ARGS__)
> +
> +#define RTE_TEST_ASSERT_FAIL(val, msg, ...) \
> + RTE_TEST_ASSERT(val != 0, msg, ##__VA_ARGS__)
> +
> +#define RTE_TEST_ASSERT_NULL(val, msg, ...) \
> + RTE_TEST_ASSERT(val == NULL, msg, ##__VA_ARGS__)
> +
> +#define RTE_TEST_ASSERT_NOT_NULL(val, msg, ...) \
> + RTE_TEST_ASSERT(val != NULL, msg, ##__VA_ARGS__)
> +
> +#endif /* _RTE_TEST_H_ */
> diff --git a/test/test/test.h b/test/test/test.h
> index 08ffe949c..0ea18586c 100644
> --- a/test/test/test.h
> +++ b/test/test/test.h
> @@ -38,7 +38,6 @@
>  #include 
>
>  #include 
> -#include 
>
>  #define TEST_SUCCESS  (0)
>  #define TEST_FAILED  (-1)
> @@ -50,23 +49,13 @@
>  # define TEST_TRACE_FAILURE(_file, _line, _func)
>  #endif
>
> -#define TEST_ASSERT(cond, msg, ...) do { \
> - if (!(cond)) {   \
> - printf("TestCase %s() line %d failed: "  \
> - msg "\n", __func__, __LINE__, ##__VA_ARGS__);   
>  \
> - TEST_TRACE_FAILURE(__FILE__, __LINE__, __func__);\
> - return TEST_FAILED;  \
> - }\
> -} while (0)
> +#define RTE_TEST_TRACE_FAILURE TEST_TRACE_FAILURE
>
> -#define TEST_ASSERT_EQUAL(a, b, msg, ...) do {   \
> - if (!(a == b)) { \
> - printf("TestCase %s() line %d failed: "  \
> - msg "\n", __func__, __LINE__, ##__VA_ARGS__);   
>  \
> -

Re: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on VF

[Wu, Yanglong]

Yes, this problem is caused by PF passing a wrong number of queue to VF. So 
basically, I removed  model check of "case ETH_MQ_RX_NONE". It seem right after 
I do that but I'am also not very sure the changing is really right. So thank to 
your comments, I will redebug this problem.

-Original Message-
From: Dai, Wei 
Sent: Thursday, January 11, 2018 5:48 PM
To: Wu, Yanglong ; dev@dpdk.org; Yang, Zhiyong 

Cc: Wu, Yanglong 
Subject: RE: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on VF

According to 82599 datasheet, VF working mode is controlled by its host PF.
If VF is working in RSS or DCB mode, it can use more than 1 queues.
That also means its host PF should enable RSS or DCB.
The PF Rx mode can be ETH_MQ_RX_VMDQ_RSS or ETH_MQ_RX_VMDQ_DCB.
So I think that DPDK app with PF cause this issue, not ixgbe PMD itself.

By the way, you can reply your patch mail to provide more info about how You 
reproduce this issue to easy others' reviewing and testing your patch.

Thanks

> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Yanglong Wu
> Sent: Tuesday, January 9, 2018 2:32 PM
> To: dev@dpdk.org; Yang, Zhiyong 
> Cc: Wu, Yanglong 
> Subject: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on VF
> 
> VF can't run in multi queue mode, if nb_q_per_pool was set as 1.
> Nb_q_per_pool is passed through to max_rx_q and max_tx_q in VF.
> So if nb_q_per_pool is equal to 1, max_rx_q and max_tx_q can't be more 
> than 1 and VF multi queue mode will fail.
> 
> Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check)
> 
> Signed-off-by: Yanglong Wu 
> ---
> v1~v3:
> coding style issue fix
> ---
> v4:
> Rework according to comments
> ---
>  drivers/net/ixgbe/ixgbe_ethdev.c | 2 --
>  1 file changed, 2 deletions(-)
> 
> diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c
> b/drivers/net/ixgbe/ixgbe_ethdev.c
> index baaeee5d9..6af4f8b2b 100644
> --- a/drivers/net/ixgbe/ixgbe_ethdev.c
> +++ b/drivers/net/ixgbe/ixgbe_ethdev.c
> @@ -2240,8 +2240,6 @@ ixgbe_check_mq_mode(struct rte_eth_dev *dev)
>   case ETH_MQ_RX_NONE:
>   /* if nothing mq mode configure, use default scheme */
>   dev->data->dev_conf.rxmode.mq_mode = 
> ETH_MQ_RX_VMDQ_ONLY;
> - if (RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool > 1)
> - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 1;
>   break;
>   default: /* ETH_MQ_RX_DCB, ETH_MQ_RX_DCB_RSS or ETH_MQ_TX_DCB*/
>   /* SRIOV only works in VMDq enable mode */
> --
> 2.11.0

Re: [dpdk-dev] [PATCH v5 01/11] eal: add common test assert macros

[Thomas Monjalon]

11/01/2018 11:29, Pavan Nikhilesh:
> +Cc: tho...@monjalon.net
> On Thu, Jan 11, 2018 at 03:51:46PM +0530, Pavan Nikhilesh wrote:
> > Adding common test assertion macros for unit testing.
> > Replaced common macros in test/test.h with new RTE_TEST_ASSERT_* macros.
> >
> > Signed-off-by: Pavan Nikhilesh 
> > Acked-by: Jerin Jacob 
> > ---
> >
> >  v5 Changes:
> >   - rebase patchset
> >   - remove duplication between test/test/test.h and rte_test.h by mapping
> >   RTE_TEST_ASSERT* macros to test.h

Very good, thanks.

Acked-by: Thomas Monjalon 

Re: [dpdk-dev] [PATCH v6 01/23] eventtimer: add event timer adapter API

[Pavan Nikhilesh]

On Wed, Jan 10, 2018 at 06:20:52PM -0600, Erik Gabriel Carrillo wrote:
> Signed-off-by: Erik Gabriel Carrillo 

Consider giving credit to the original authors.

> ---
>  lib/librte_eventdev/Makefile  |   1 +
>  lib/librte_eventdev/rte_event_timer_adapter.h | 566 
> ++
>  lib/librte_eventdev/rte_eventdev.h|   4 +-
>  3 files changed, 569 insertions(+), 2 deletions(-)
>  create mode 100644 lib/librte_eventdev/rte_event_timer_adapter.h
>
> diff --git a/lib/librte_eventdev/Makefile b/lib/librte_eventdev/Makefile
> index 7fd78c7..685b474 100644
> --- a/lib/librte_eventdev/Makefile
> +++ b/lib/librte_eventdev/Makefile
> @@ -27,6 +27,7 @@ SYMLINK-y-include += rte_eventdev_pmd_pci.h
>  SYMLINK-y-include += rte_eventdev_pmd_vdev.h
>  SYMLINK-y-include += rte_event_ring.h
>  SYMLINK-y-include += rte_event_eth_rx_adapter.h
> +SYMLINK-y-include += rte_event_timer_adapter.h
>
>  # versioning export map
>  EXPORT_MAP := rte_eventdev_version.map
> diff --git a/lib/librte_eventdev/rte_event_timer_adapter.h 
> b/lib/librte_eventdev/rte_event_timer_adapter.h
> new file mode 100644
> index 000..7d967e6
> --- /dev/null
> +++ b/lib/librte_eventdev/rte_event_timer_adapter.h
> @@ -0,0 +1,566 @@
> +/*-
> + *   BSD LICENSE
> + *
> + *   Copyright 2017 Cavium, Inc.
> + *   Copyright(c) 2017 Intel Corporation. All rights reserved.
> + *
> + *   Redistribution and use in source and binary forms, with or without
> + *   modification, are permitted provided that the following conditions
> + *   are met:
> + *
> + * * Redistributions of source code must retain the above copyright
> + *   notice, this list of conditions and the following disclaimer.
> + * * Redistributions in binary form must reproduce the above copyright
> + *   notice, this list of conditions and the following disclaimer in
> + *   the documentation and/or other materials provided with the
> + *   distribution.
> + * * Neither the name of Intel Corporation nor the names of its
> + *   contributors may be used to endorse or promote products derived
> + *   from this software without specific prior written permission.
> + *
> + *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
> + *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
> + *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
> + *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
> + *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
> + *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
> + *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
> + *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
> + *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
> + *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
> + *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
> + */

Use SPDX licence tags

> +#ifndef __RTE_EVENT_TIMER_ADAPTER_H__
> +#define __RTE_EVENT_TIMER_ADAPTER_H__
> +
> +/**
> + * @file
> + *
> + * RTE Event Timer Adapter

> + */
> +
> +#ifdef __cplusplus
> +extern "C" {
> +#endif
> +
> +#include 
> +#include 
> +
> +#include "rte_eventdev.h"
> +
> +#define RTE_EVENT_TIMER_ADAPTER_NUM_MAX 64
> +
> +/**
> + * @warning
> + * @b EXPERIMENTAL: this enum may change without prior notice

Please make sure that Experimental tags usage is inline with the new tool
http://dpdk.org/dev/patchwork/patch/32234.

Cheers,
Pavan

> + *
> + * Timer adapter clock source
> + */
> +enum rte_event_timer_adapter_clk_src {
> + RTE_EVENT_TIMER_ADAPTER_CPU_CLK,
> + /**< Use CPU clock as the clock source. */
> + RTE_EVENT_TIMER_ADAPTER_EXT_CLK0,
> + /**< Platform dependent external clock source 0. */
> + RTE_EVENT_TIMER_ADAPTER_EXT_CLK1,
> + /**< Platform dependent external clock source 1. */
> + RTE_EVENT_TIMER_ADAPTER_EXT_CLK2,
> + /**< Platform dependent external clock source 2. */
> + RTE_EVENT_TIMER_ADAPTER_EXT_CLK3,
> + /**< Platform dependent external clock source 3. */
> +};
> +

Re: [dpdk-dev] [PATCH 00/11] net/vhostpci: A new vhostpci PMD supporting VM2VM scenario

[Yang, Zhiyong]

Hi Maxime, all, 

> -Original Message-
> From: Maxime Coquelin [mailto:maxime.coque...@redhat.com]
> Sent: Tuesday, December 19, 2017 7:15 PM
> To: Yang, Zhiyong ; dev@dpdk.org;
> y...@fridaylinux.org
> Cc: Wang, Wei W ; Tan, Jianfeng
> 
> Subject: Re: [PATCH 00/11] net/vhostpci: A new vhostpci PMD supporting
> VM2VM scenario
> 
> Hi Zhiyong,
> 
> On 11/30/2017 10:46 AM, Zhiyong Yang wrote:
> > Vhostpci PMD is a new type driver working in guest OS which has
> > ability to drive the vhostpci modern pci device, which is a new virtio 
> > device.
> >
> > The following linking is about vhostpci design:
> >
> > An initial device design is presented at KVM Forum'16:
> > http://www.linux-kvm.org/images/5/55/02x07A-Wei_Wang-Design_of-
> Vhost-p
> > ci.pdf The latest device design and implementation will be posted to
> > the QEMU community soon.
> >
> > Vhostpci PMD works in pair with virtio-net PMD to achieve
> > point-to-point communication between VMs. DPDK already has
> > virtio/vhost user PMD pair to implement RX/TX packets between
> > guest/host scenario. However, for VM2VM use cases, Virtio PMD needs to
> > transmit pkts from VM1 to host OS firstly by vhost user port, then
> > transmit pkts to the 2nd VM by virtio PMD port again. Virtio/Vhostpci
> > PMD pair can implement shared memory to receive/trasmit packets
> > directly between two VMs. Currently, the entire memory of the virtio-net
> side VM is shared to the vhost-pci side VM, and mapped via device BAR2,
> and the first 4KB area of BAR2 is reserved to store the metadata.
> >
> > The vhostpci/virtio PMD working processing is the following:
> >
> > 1.VM1 startup with vhostpci device, bind the device to DPDK in the
> > guest1, launch the DPDK testpmd, then waiting for the remote memory
> > info (the VM2 shares memory, memory regions and vring info).
> >
> > 2.VM2 startup with virtio-net device, bind the virito-net to DPDK in
> > the VM2, run testpmd using virtio PMD.
> >
> > 3.vhostpci device negotiate virtio message with virtio-net device via
> > socket as vhost user/virtio-net do that.
> >
> > 4.Vhostpci device gets VM2's memory region and vring info and write
> > the metadata to VM2's shared memory.
> >
> > 5.When the metadata is ready to be read by the Vhostpci PMD, the PMD
> > will receive a config interrupt with LINK_UP set in the status config.
> >
> > 6.Vhostpci PMD and Virtio PMD can transmit/receive the packets.
> >
> > How to test?
> >
> > 1. launch VM1 with vhostpci device.
> > qemu/x86_64-softmmu/qemu-system-x86_64 -cpu host -M pc -enable-
> kvm \
> > -smp 16,threads=1,sockets=1 -m 8G -mem-prealloc -realtime mlock=on \
> > -object memory-backend-file,id=mem,size=8G,mem-
> path=/dev/hugepages, \
> > share=on -numa node,memdev=mem -drive
> > if=virtio,file=/root/vhost-pci/guest1.img,format=raw \ -kernel
> > /opt/guest_kernel -append 'root=/dev/vda1 ro default_hugepagesz=1G
> > hugepagesz=1G \
> > hugepages=2 console=ttyS0,115200,8n1 3' -netdev
> > tap,id=net1,br=br0,script=/etc/qemu-ifup \ -chardev
> > socket,id=slave1,server,wait=off, path=/opt/vhost-pci-slave1 -device
> > vhost-pci-net-pci, \
> > chardev=slave1 \
> > -nographic
> >
> > 2. bind vhostpci device to dpdk using igb_uio.
> > startup dpdk
> > ./x86_64-native-linuxapp-gcc/app/testpmd -c 0x3 -n 4 -- -i
> >
> > 3. launch VM2 with virtio-net device.
> >
> > qemu/x86_64-softmmu/qemu-system-x86_64 -cpu host -M pc -enable-
> kvm \
> > -smp 4,threads=1,sockets=1 -m 8G -mem-prealloc -realtime mlock=on \
> > -object
> > memory-backend-file,id=mem,size=8G,mem-
> path=/dev/hugepages,share=on \
> > -numa node,memdev=mem -drive
> > if=virtio,file=/root/vhost-pci/guest2.img,format=raw \ -net none
> > -no-hpet -kernel /opt/guest_kernel \ -append 'root=/dev/vda1 ro
> > default_hugepagesz=1G hugepagesz=1G hugepages=2
> > console=ttyS0,115200,8n1 3' \ -chardev
> > socket,id=sock2,path=/opt/vhost-pci-slave1 \ -netdev
> > type=vhost-user,id=net2,chardev=sock2,vhostforce \ -device
> > virtio-net-pci,mac=52:54:00:00:00:02,netdev=net2 \ -nographic
> >
> > 4.bind virtio-net to dpdk using igb_uio run dpdk
> >
> > ./x86_64-native-linuxapp-gcc/app/testpmd -c 0x3 -n 4 --socket-mem
> > 512,0 \
> > -- -i --rxq=1 --txq=1 --nb-cores=1
> >
> > 5. vhostpci PMD run "start"
> >
> > 6. virtio PMD side run "start tx_first"
> >
> > loopback testing can work.
> >
> > note:
> > 1. only support igb_uio for now.
> > 2. vhostpci device is a modern pci device. vhostpci PMD only supports
> > mergable mode. Virtio device side must be mergable mode.
> > 3. vhostpci PMD supports one queue pair for now.
> >
> > Zhiyong Yang (11):
> >drivers/net: add vhostpci PMD base files
> >net/vhostpci: public header files
> >net/vhostpci: add debugging log macros
> >net/vhostpci: add basic framework
> >net/vhostpci: add queue setup
> >net/vhostpci: add support for link status change
> >net/vhostpci: get remote memory region and vring info
> >net/vhostpci: add RX function
> >net/vhostpci: add TX functio

Re: [dpdk-dev] [PATCH v6 05/23] eventtimer: add adapter allocation definitions

[Pavan Nikhilesh]

On Wed, Jan 10, 2018 at 06:20:56PM -0600, Erik Gabriel Carrillo wrote:
> Add definitions for the functions that allocate and deallocate adapter
> instances in the default software implementation.
>
> Signed-off-by: Erik Gabriel Carrillo 
> ---
>  lib/Makefile  |   2 +-
>  lib/librte_eventdev/Makefile  |   2 +-
>  lib/librte_eventdev/rte_event_timer_adapter.c | 138 
> +-
>  lib/librte_eventdev/rte_event_timer_adapter.h |   2 +-
>  4 files changed, 139 insertions(+), 5 deletions(-)
>
> diff --git a/lib/Makefile b/lib/Makefile
> index 4202702..4c53f8c 100644
> --- a/lib/Makefile
> +++ b/lib/Makefile
> @@ -29,7 +29,7 @@ DEPDIRS-librte_security := librte_eal librte_mempool 
> librte_ring librte_mbuf
>  DEPDIRS-librte_security += librte_ether
>  DEPDIRS-librte_security += librte_cryptodev
>  DIRS-$(CONFIG_RTE_LIBRTE_EVENTDEV) += librte_eventdev
> -DEPDIRS-librte_eventdev := librte_eal librte_ring librte_ether librte_hash
> +DEPDIRS-librte_eventdev := librte_eal librte_ring librte_ether librte_hash 
> librte_mempool
>  DIRS-$(CONFIG_RTE_LIBRTE_VHOST) += librte_vhost
>  DEPDIRS-librte_vhost := librte_eal librte_mempool librte_mbuf librte_ether
>  DIRS-$(CONFIG_RTE_LIBRTE_HASH) += librte_hash
> diff --git a/lib/librte_eventdev/Makefile b/lib/librte_eventdev/Makefile
> index 8f11a79..e68f888 100644
> --- a/lib/librte_eventdev/Makefile
> +++ b/lib/librte_eventdev/Makefile
> @@ -13,7 +13,7 @@ LIBABIVER := 3
>  # build flags
>  CFLAGS += -O3
>  CFLAGS += $(WERROR_FLAGS)
> -LDLIBS += -lrte_eal -lrte_ring -lrte_ethdev -lrte_hash
> +LDLIBS += -lrte_eal -lrte_ring -lrte_ethdev -lrte_hash -lrte_mempool
>
>  # library source files
>  SRCS-y += rte_eventdev.c
> diff --git a/lib/librte_eventdev/rte_event_timer_adapter.c 
> b/lib/librte_eventdev/rte_event_timer_adapter.c
> index 540be95..9c4ba1c 100644
> --- a/lib/librte_eventdev/rte_event_timer_adapter.c
> +++ b/lib/librte_eventdev/rte_event_timer_adapter.c
> @@ -31,11 +31,17 @@
>   */
>
>  #include 
> +#include 
>
>  #include 
>  #include 
>  #include 
>  #include 
> +#include 
> +#include 
> +#include 
> +#include 
> +#include 
>
>  #include "rte_eventdev.h"
>  #include "rte_eventdev_pmd.h"
> @@ -163,6 +169,11 @@ rte_event_timer_adapter_create_ext(
>
>   adapter_id = conf->timer_adapter_id;
>
> + if (adapter_id >= RTE_EVENT_TIMER_ADAPTER_NUM_MAX) {
> + rte_errno = -EINVAL;
> + return NULL;
> + }
> +
>   /* Check adapter ID not already allocated */
>   adapter = &adapters[adapter_id];
>   if (adapter->allocated) {
> @@ -412,18 +423,141 @@ rte_event_timer_cancel_burst(const struct 
> rte_event_timer_adapter *adapter,
>   * Software event timer adapter ops definitions
>   */
>
> +struct rte_event_timer_adapter_sw_data {
> + /* Number of outstanding timers managed by event adapter. */
> + int nb_armed_evtims;
> + /* Identifier of service executing timer management logic. */
> + uint32_t service_id;
> + /* Ring containing messages to arm or cancel event timers */
> + struct rte_ring *msg_ring;
> + /* Mempool containing msg objects */
> + struct rte_mempool *msg_pool;
> + /* Mempool containing timer objects */
> + struct rte_mempool *tim_pool;
> +};

Don't use rte_ prefix for internal data.

> +
> +enum msg_type {MSG_TYPE_ARM, MSG_TYPE_CANCEL};
> +
> +struct msg {
> + enum msg_type type;
> + struct rte_event_timer *evtim;
> +};
> +
> +static int
> +sw_event_timer_adapter_service_func(void *arg)
> +{
> + RTE_SET_USED(arg);
> + return 0;
> +}
> +
>  static int
>  sw_event_timer_adapter_init(struct rte_event_timer_adapter *adapter)
>  {
> - RTE_SET_USED(adapter);
> + int ret;
> + struct rte_event_timer_adapter_sw_data *sw_data;
> + uint64_t nb_timers;

> +
> + char pool_name[RTE_RING_NAMESIZE];
> + snprintf(pool_name, RTE_RING_NAMESIZE, "sw_evtim_adap_msg_pool_%"PRIu8,
> +  adapter->data->id);
> + sw_data->msg_pool = rte_mempool_create(pool_name, nb_timers,
> +sizeof(struct msg), 32, 0, NULL,
> +NULL, NULL, NULL,
> +adapter->data->socket_id, 0);
> + if (sw_data->msg_pool == NULL) {
> + rte_errno = ENOMEM;
> + return -1;
> + }
> +
> + snprintf(pool_name, RTE_RING_NAMESIZE, "sw_evtim_adap_tim_pool_%"PRIu8,
> +  adapter->data->id);
> + sw_data->tim_pool = rte_mempool_create(pool_name, nb_timers,
> +sizeof(struct rte_timer), 32, 0,
> +NULL, NULL, NULL, NULL,
> +adapter->data->socket_id, 0);
> + if (sw_data->tim_pool == NULL) {
> + printf("Could not allocate tim mempool\n");
> + return -1;
> + }

Any specific reason for having seperate

Re: [dpdk-dev] [PATCH v5 11/11] doc: update eventdev documentation

[Jerin Jacob]

-Original Message-
> Date: Thu, 11 Jan 2018 15:51:56 +0530
> From: Pavan Nikhilesh 
> To: jerin.ja...@caviumnetworks.com, harry.van.haa...@intel.com,
>  gage.e...@intel.com, liang.j...@intel.com
> Cc: dev@dpdk.org, Pavan Nikhilesh 
> Subject: [dpdk-dev] [PATCH v5 11/11] doc: update eventdev documentation
> X-Mailer: git-send-email 2.14.1
> 
> Update octeontx eventdev documentation to include new selftest feature.
> 
> Signed-off-by: Pavan Nikhilesh 
> Acked-by: Marko Kovacevic 

Applied this series to dpdk-next-eventdev/master. Thanks.

Re: [dpdk-dev] [PATCH v6 10/23] eventtimer: remove service id entry from info structure

[Pavan Nikhilesh]

On Wed, Jan 10, 2018 at 06:21:01PM -0600, Erik Gabriel Carrillo wrote:
> Signed-off-by: Erik Gabriel Carrillo 
> ---
>  lib/librte_eventdev/rte_event_timer_adapter.c | 11 +--
>  lib/librte_eventdev/rte_event_timer_adapter.h |  2 --
>  2 files changed, 1 insertion(+), 12 deletions(-)
>
> diff --git a/lib/librte_eventdev/rte_event_timer_adapter.c 
> b/lib/librte_eventdev/rte_event_timer_adapter.c
> index 68748be..a4c8012 100644
> --- a/lib/librte_eventdev/rte_event_timer_adapter.c
> +++ b/lib/librte_eventdev/rte_event_timer_adapter.c
> @@ -609,15 +609,6 @@ sw_event_timer_adapter_stop(const struct 
> rte_event_timer_adapter *adapter)
>   return rte_service_component_runstate_set(sw_data->service_id, 0);
>  }
>
> -static void
> -sw_event_timer_adapter_get_info(const struct rte_event_timer_adapter 
> *adapter,
> - struct rte_event_timer_adapter_info *adapter_info)
> -{
> - struct rte_event_timer_adapter_sw_data *sw_data;
> - sw_data = adapter->data->adapter_priv;
> - adapter_info->service_id = sw_data->service_id;
> -}
> -

Removing get_info is not a good idea, the application using event timer might
need some info about capabilities.
Thinking form sw_event_timer perspective I think we need to have a caps to check
if it supports DISTRIBUTED scheduling similar to sw event dev (or) expose
RTE_EVENT_TIMER_ADAPTER_CAP_INTERNAL_PORT.

>  static int
>  sw_event_timer_arm_burst(const struct rte_event_timer_adapter *adapter,
>struct rte_event_timer **evtims,
> @@ -661,7 +652,7 @@ const struct rte_event_timer_adapter_ops 
> sw_event_adapter_timer_ops = {
>   .uninit = sw_event_timer_adapter_uninit,
>   .start = sw_event_timer_adapter_start,
>   .stop = sw_event_timer_adapter_stop,
> - .get_info = sw_event_timer_adapter_get_info,
> + .get_info = NULL,
>   .arm_burst = sw_event_timer_arm_burst,
>   .arm_tmo_tick_burst = sw_event_timer_arm_tmo_tick_burst,
>   .cancel_burst = sw_event_timer_cancel_burst,
> diff --git a/lib/librte_eventdev/rte_event_timer_adapter.h 
> b/lib/librte_eventdev/rte_event_timer_adapter.h
> index 84d3c39..8d29cfc 100644
> --- a/lib/librte_eventdev/rte_event_timer_adapter.h
> +++ b/lib/librte_eventdev/rte_event_timer_adapter.h
> @@ -283,8 +283,6 @@ struct rte_event_timer_adapter_info {
>   /**< Event timer adapter capabilities */
>   int16_t event_dev_port_id;
>   /**< Event device port ID, if applicable */
> - int32_t service_id;
> - /**< Service ID, if applicable */
>  };
>
>  /**
> --
> 2.6.4
>

[dpdk-dev] [PATCH v2 0/3] crypto/dpaa_sec: performance optimizations

[Akhil Goyal]

Following changes are added to improve performance.
1. optimize virtual to physical address conversion
2. support for multiple sessions in a single queue pair
3. support for ipsec protocol offload

changes in v2:
 - incorporated comments from Hemant
 - split the patchset to remove dependency on bus/dpaa patch
Note:
1. This patchset is now independent of the patches on net subtree.
2. bus/dpaa patch is already applied to net subtree.
3. The last patch in v1 of this series will be sent separately as it will
   be dependent on the net subtree.

Akhil Goyal (1):
  crypto/dpaa_sec: support ipsec protocol offload

Hemant Agrawal (2):
  crypto/dpaa_sec: optimize virt to phy conversion
  crypto/dpaa_sec: support multiple sessions per qp

 doc/guides/cryptodevs/features/dpaa_sec.ini |   1 +
 drivers/crypto/dpaa_sec/dpaa_sec.c  | 591 
 drivers/crypto/dpaa_sec/dpaa_sec.h  | 135 +--
 3 files changed, 613 insertions(+), 114 deletions(-)

-- 
2.9.3

[dpdk-dev] [PATCH v2 1/3] crypto/dpaa_sec: optimize virt to phy conversion

[Akhil Goyal]

From: Hemant Agrawal 

Context memory is allocated from mempool. Ideally
it will get all memory from single segment, so simple offset
calculation is used for address conversion for such addresses
from context memory.

Signed-off-by: Hemant Agrawal 
Acked-by: Akhil Goyal 
---
 drivers/crypto/dpaa_sec/dpaa_sec.c | 27 ++-
 drivers/crypto/dpaa_sec/dpaa_sec.h |  1 +
 2 files changed, 19 insertions(+), 9 deletions(-)

diff --git a/drivers/crypto/dpaa_sec/dpaa_sec.c 
b/drivers/crypto/dpaa_sec/dpaa_sec.c
index ad1b309..157eace 100644
--- a/drivers/crypto/dpaa_sec/dpaa_sec.c
+++ b/drivers/crypto/dpaa_sec/dpaa_sec.c
@@ -80,6 +80,8 @@ dpaa_sec_alloc_ctx(dpaa_sec_session *ses)
dcbz_64(&ctx->job.sg[SG_CACHELINE_3]);
 
ctx->ctx_pool = ses->ctx_pool;
+   ctx->vtop_offset = (uint64_t) ctx
+   - rte_mempool_virt2iova(ctx);
 
return ctx;
 }
@@ -104,6 +106,13 @@ dpaa_mem_vtop(void *vaddr)
return (rte_iova_t)(NULL);
 }
 
+/* virtual address conversin when mempool support is available for ctx */
+static inline phys_addr_t
+dpaa_mem_vtop_ctx(struct dpaa_sec_op_ctx *ctx, void *vaddr)
+{
+   return (uint64_t)vaddr - ctx->vtop_offset;
+}
+
 static inline void *
 dpaa_mem_ptov(rte_iova_t paddr)
 {
@@ -563,7 +572,7 @@ build_auth_only(struct rte_crypto_op *op, dpaa_sec_session 
*ses)
if (is_decode(ses)) {
/* need to extend the input to a compound frame */
sg->extension = 1;
-   qm_sg_entry_set64(sg, dpaa_mem_vtop(&cf->sg[2]));
+   qm_sg_entry_set64(sg, dpaa_mem_vtop_ctx(ctx, &cf->sg[2]));
sg->length = sym->auth.data.length + ses->digest_length;
sg->final = 1;
cpu_to_hw_sg(sg);
@@ -577,7 +586,7 @@ build_auth_only(struct rte_crypto_op *op, dpaa_sec_session 
*ses)
cpu_to_hw_sg(sg);
 
/* let's check digest by hw */
-   start_addr = dpaa_mem_vtop(old_digest);
+   start_addr = dpaa_mem_vtop_ctx(ctx, old_digest);
sg++;
qm_sg_entry_set64(sg, start_addr);
sg->length = ses->digest_length;
@@ -631,7 +640,7 @@ build_cipher_only(struct rte_crypto_op *op, 
dpaa_sec_session *ses)
sg->extension = 1;
sg->final = 1;
sg->length = sym->cipher.data.length + ses->iv.length;
-   qm_sg_entry_set64(sg, dpaa_mem_vtop(&cf->sg[2]));
+   qm_sg_entry_set64(sg, dpaa_mem_vtop_ctx(ctx, &cf->sg[2]));
cpu_to_hw_sg(sg);
 
sg = &cf->sg[2];
@@ -677,7 +686,7 @@ build_cipher_auth_gcm(struct rte_crypto_op *op, 
dpaa_sec_session *ses)
/* input */
rte_prefetch0(cf->sg);
sg = &cf->sg[2];
-   qm_sg_entry_set64(&cf->sg[1], dpaa_mem_vtop(sg));
+   qm_sg_entry_set64(&cf->sg[1], dpaa_mem_vtop_ctx(ctx, sg));
if (is_encode(ses)) {
qm_sg_entry_set64(sg, dpaa_mem_vtop(IV_ptr));
sg->length = ses->iv.length;
@@ -722,7 +731,7 @@ build_cipher_auth_gcm(struct rte_crypto_op *op, 
dpaa_sec_session *ses)
   ses->digest_length);
sg++;
 
-   qm_sg_entry_set64(sg, dpaa_mem_vtop(ctx->digest));
+   qm_sg_entry_set64(sg, dpaa_mem_vtop_ctx(ctx, ctx->digest));
sg->length = ses->digest_length;
length += sg->length;
sg->final = 1;
@@ -736,7 +745,7 @@ build_cipher_auth_gcm(struct rte_crypto_op *op, 
dpaa_sec_session *ses)
 
/* output */
sg++;
-   qm_sg_entry_set64(&cf->sg[0], dpaa_mem_vtop(sg));
+   qm_sg_entry_set64(&cf->sg[0], dpaa_mem_vtop_ctx(ctx, sg));
qm_sg_entry_set64(sg,
dst_start_addr + sym->aead.data.offset - ses->auth_only_len);
sg->length = sym->aead.data.length + ses->auth_only_len;
@@ -788,7 +797,7 @@ build_cipher_auth(struct rte_crypto_op *op, 
dpaa_sec_session *ses)
/* input */
rte_prefetch0(cf->sg);
sg = &cf->sg[2];
-   qm_sg_entry_set64(&cf->sg[1], dpaa_mem_vtop(sg));
+   qm_sg_entry_set64(&cf->sg[1], dpaa_mem_vtop_ctx(ctx, sg));
if (is_encode(ses)) {
qm_sg_entry_set64(sg, dpaa_mem_vtop(IV_ptr));
sg->length = ses->iv.length;
@@ -818,7 +827,7 @@ build_cipher_auth(struct rte_crypto_op *op, 
dpaa_sec_session *ses)
   ses->digest_length);
sg++;
 
-   qm_sg_entry_set64(sg, dpaa_mem_vtop(ctx->digest));
+   qm_sg_entry_set64(sg, dpaa_mem_vtop_ctx(ctx, ctx->digest));
sg->length = ses->digest_length;
length += sg->length;
sg->final = 1;
@@ -832,7 +841,7 @@ build_cipher_auth(struct rte_crypto_op *op, 
dpaa_sec_session *ses)
 
/* output */
sg++;
-   qm_sg_entry_set64(&cf->sg[0], dpaa_mem_vtop(sg));
+   qm_sg_entry_set64(&cf->sg[0], dpaa_mem_vtop_ctx(ctx, sg));
qm_sg_entry_set64(sg, dst_start_addr + sym->ciphe

[dpdk-dev] [PATCH v2 3/3] crypto/dpaa_sec: support ipsec protocol offload

[Akhil Goyal]

Signed-off-by: Akhil Goyal 
Acked-by: Hemant Agrawal 
---
 doc/guides/cryptodevs/features/dpaa_sec.ini |   1 +
 drivers/crypto/dpaa_sec/dpaa_sec.c  | 410 ++--
 drivers/crypto/dpaa_sec/dpaa_sec.h  |  62 -
 3 files changed, 450 insertions(+), 23 deletions(-)

diff --git a/doc/guides/cryptodevs/features/dpaa_sec.ini 
b/doc/guides/cryptodevs/features/dpaa_sec.ini
index 0e8f5b2..deab53a 100644
--- a/doc/guides/cryptodevs/features/dpaa_sec.ini
+++ b/doc/guides/cryptodevs/features/dpaa_sec.ini
@@ -7,6 +7,7 @@
 Symmetric crypto   = Y
 Sym operation chaining = Y
 HW Accelerated = Y
+Protocol offload   = Y
 
 ;
 ; Supported crypto algorithms of the 'dpaa_sec' crypto driver.
diff --git a/drivers/crypto/dpaa_sec/dpaa_sec.c 
b/drivers/crypto/dpaa_sec/dpaa_sec.c
index f433f0a..d7b6f39 100644
--- a/drivers/crypto/dpaa_sec/dpaa_sec.c
+++ b/drivers/crypto/dpaa_sec/dpaa_sec.c
@@ -15,6 +15,7 @@
 #include 
 #include 
 #include 
+#include 
 #include 
 #include 
 #include 
@@ -196,8 +197,19 @@ dqrr_out_fq_cb_rx(struct qman_portal *qm __always_unused,
 * sg[1] for input
 */
job = dpaa_mem_ptov(qm_fd_addr_get64(fd));
+
ctx = container_of(job, struct dpaa_sec_op_ctx, job);
ctx->fd_status = fd->status;
+   if (ctx->op->sess_type == RTE_CRYPTO_OP_SECURITY_SESSION) {
+   struct qm_sg_entry *sg_out;
+   uint32_t len;
+
+   sg_out = &job->sg[0];
+   hw_sg_to_cpu(sg_out);
+   len = sg_out->length;
+   ctx->op->sym->m_src->pkt_len = len;
+   ctx->op->sym->m_src->data_len = len;
+   }
dpaa_sec_ops[dpaa_sec_op_nb++] = ctx->op;
dpaa_sec_op_ending(ctx);
 
@@ -261,7 +273,13 @@ static inline int is_aead(dpaa_sec_session *ses)
 static inline int is_auth_cipher(dpaa_sec_session *ses)
 {
return ((ses->cipher_alg != RTE_CRYPTO_CIPHER_NULL) &&
-   (ses->auth_alg != RTE_CRYPTO_AUTH_NULL));
+   (ses->auth_alg != RTE_CRYPTO_AUTH_NULL) &&
+   (ses->proto_alg != RTE_SECURITY_PROTOCOL_IPSEC));
+}
+
+static inline int is_proto_ipsec(dpaa_sec_session *ses)
+{
+   return (ses->proto_alg == RTE_SECURITY_PROTOCOL_IPSEC);
 }
 
 static inline int is_encode(dpaa_sec_session *ses)
@@ -282,27 +300,39 @@ caam_auth_alg(dpaa_sec_session *ses, struct alginfo 
*alginfo_a)
ses->digest_length = 0;
break;
case RTE_CRYPTO_AUTH_MD5_HMAC:
-   alginfo_a->algtype = OP_ALG_ALGSEL_MD5;
+   alginfo_a->algtype =
+   (ses->proto_alg == RTE_SECURITY_PROTOCOL_IPSEC) ?
+   OP_PCL_IPSEC_HMAC_MD5_96 : OP_ALG_ALGSEL_MD5;
alginfo_a->algmode = OP_ALG_AAI_HMAC;
break;
case RTE_CRYPTO_AUTH_SHA1_HMAC:
-   alginfo_a->algtype = OP_ALG_ALGSEL_SHA1;
+   alginfo_a->algtype =
+   (ses->proto_alg == RTE_SECURITY_PROTOCOL_IPSEC) ?
+   OP_PCL_IPSEC_HMAC_SHA1_96 : OP_ALG_ALGSEL_SHA1;
alginfo_a->algmode = OP_ALG_AAI_HMAC;
break;
case RTE_CRYPTO_AUTH_SHA224_HMAC:
-   alginfo_a->algtype = OP_ALG_ALGSEL_SHA224;
+   alginfo_a->algtype =
+   (ses->proto_alg == RTE_SECURITY_PROTOCOL_IPSEC) ?
+   OP_PCL_IPSEC_HMAC_SHA1_160 : OP_ALG_ALGSEL_SHA224;
alginfo_a->algmode = OP_ALG_AAI_HMAC;
break;
case RTE_CRYPTO_AUTH_SHA256_HMAC:
-   alginfo_a->algtype = OP_ALG_ALGSEL_SHA256;
+   alginfo_a->algtype =
+   (ses->proto_alg == RTE_SECURITY_PROTOCOL_IPSEC) ?
+   OP_PCL_IPSEC_HMAC_SHA2_256_128 : OP_ALG_ALGSEL_SHA256;
alginfo_a->algmode = OP_ALG_AAI_HMAC;
break;
case RTE_CRYPTO_AUTH_SHA384_HMAC:
-   alginfo_a->algtype = OP_ALG_ALGSEL_SHA384;
+   alginfo_a->algtype =
+   (ses->proto_alg == RTE_SECURITY_PROTOCOL_IPSEC) ?
+   OP_PCL_IPSEC_HMAC_SHA2_384_192 : OP_ALG_ALGSEL_SHA384;
alginfo_a->algmode = OP_ALG_AAI_HMAC;
break;
case RTE_CRYPTO_AUTH_SHA512_HMAC:
-   alginfo_a->algtype = OP_ALG_ALGSEL_SHA512;
+   alginfo_a->algtype =
+   (ses->proto_alg == RTE_SECURITY_PROTOCOL_IPSEC) ?
+   OP_PCL_IPSEC_HMAC_SHA2_512_256 : OP_ALG_ALGSEL_SHA512;
alginfo_a->algmode = OP_ALG_AAI_HMAC;
break;
default:
@@ -317,15 +347,21 @@ caam_cipher_alg(dpaa_sec_session *ses, struct alginfo 
*alginfo_c)
case RTE_CRYPTO_CIPHER_NULL:
break;
case RTE_CRYPTO_CIPHER_AES_CBC:
-   alginfo_c->algtype = OP_ALG_ALGSEL_AES;
+   alginfo_c->algtype =
+   (ses->proto_alg == RTE_SECURITY_

[dpdk-dev] [PATCH v2 2/3] crypto/dpaa_sec: support multiple sessions per qp

[Akhil Goyal]

From: Hemant Agrawal 

Signed-off-by: Hemant Agrawal 
Acked-by: Akhil Goyal 
---
 drivers/crypto/dpaa_sec/dpaa_sec.c | 154 +
 drivers/crypto/dpaa_sec/dpaa_sec.h |  74 +-
 2 files changed, 145 insertions(+), 83 deletions(-)

diff --git a/drivers/crypto/dpaa_sec/dpaa_sec.c 
b/drivers/crypto/dpaa_sec/dpaa_sec.c
index 157eace..f433f0a 100644
--- a/drivers/crypto/dpaa_sec/dpaa_sec.c
+++ b/drivers/crypto/dpaa_sec/dpaa_sec.c
@@ -43,6 +43,9 @@ static uint8_t cryptodev_driver_id;
 static __thread struct rte_crypto_op **dpaa_sec_ops;
 static __thread int dpaa_sec_op_nb;
 
+static int
+dpaa_sec_attach_sess_q(struct dpaa_sec_qp *qp, dpaa_sec_session *sess);
+
 static inline void
 dpaa_sec_op_ending(struct dpaa_sec_op_ctx *ctx)
 {
@@ -151,15 +154,6 @@ dpaa_sec_init_rx(struct qman_fq *fq_in, rte_iova_t hwdesc,
/* Clear FQ options */
memset(&fq_opts, 0x00, sizeof(struct qm_mcc_initfq));
 
-   flags = QMAN_FQ_FLAG_LOCKED | QMAN_FQ_FLAG_DYNAMIC_FQID |
-   QMAN_FQ_FLAG_TO_DCPORTAL;
-
-   ret = qman_create_fq(0, flags, fq_in);
-   if (unlikely(ret != 0)) {
-   PMD_INIT_LOG(ERR, "qman_create_fq failed");
-   return ret;
-   }
-
flags = QMAN_INITFQ_FLAG_SCHED;
fq_opts.we_mask = QM_INITFQ_WE_DESTWQ | QM_INITFQ_WE_CONTEXTA |
  QM_INITFQ_WE_CONTEXTB;
@@ -171,9 +165,11 @@ dpaa_sec_init_rx(struct qman_fq *fq_in, rte_iova_t hwdesc,
 
fq_in->cb.ern  = ern_sec_fq_handler;
 
+   PMD_INIT_LOG(DEBUG, "in-%x out-%x", fq_in->fqid, fqid_out);
+
ret = qman_init_fq(fq_in, flags, &fq_opts);
if (unlikely(ret != 0))
-   PMD_INIT_LOG(ERR, "qman_init_fq failed");
+   PMD_INIT_LOG(ERR, "qman_init_fq failed %d", ret);
 
return ret;
 }
@@ -357,7 +353,7 @@ dpaa_sec_prep_cdb(dpaa_sec_session *ses)
 {
struct alginfo alginfo_c = {0}, alginfo_a = {0}, alginfo = {0};
uint32_t shared_desc_len = 0;
-   struct sec_cdb *cdb = &ses->qp->cdb;
+   struct sec_cdb *cdb = &ses->cdb;
int err;
 #if RTE_BYTE_ORDER == RTE_BIG_ENDIAN
int swap = false;
@@ -877,12 +873,10 @@ dpaa_sec_enqueue_op(struct rte_crypto_op *op,  struct 
dpaa_sec_qp *qp)
ses = (dpaa_sec_session *)get_session_private_data(op->sym->session,
cryptodev_driver_id);
 
-   if (unlikely(!qp->ses || qp->ses != ses)) {
-   qp->ses = ses;
-   ses->qp = qp;
-   ret = dpaa_sec_prep_cdb(ses);
-   if (ret)
-   return ret;
+   if (unlikely(!ses->qp || ses->qp != qp)) {
+   PMD_INIT_LOG(DEBUG, "sess->qp - %p qp %p", ses->qp, qp);
+   if (dpaa_sec_attach_sess_q(qp, ses))
+   return -1;
}
 
/*
@@ -918,7 +912,7 @@ dpaa_sec_enqueue_op(struct rte_crypto_op *op,  struct 
dpaa_sec_qp *qp)
if (auth_only_len)
fd.cmd = 0x8000 | auth_only_len;
do {
-   ret = qman_enqueue(&qp->inq, &fd, 0);
+   ret = qman_enqueue(ses->inq, &fd, 0);
} while (ret != 0);
 
return 0;
@@ -1134,43 +1128,82 @@ dpaa_sec_aead_init(struct rte_cryptodev *dev 
__rte_unused,
return 0;
 }
 
-static int
-dpaa_sec_qp_attach_sess(struct rte_cryptodev *dev, uint16_t qp_id, void *ses)
+static struct qman_fq *
+dpaa_sec_attach_rxq(struct dpaa_sec_dev_private *qi)
 {
-   dpaa_sec_session *sess = ses;
-   struct dpaa_sec_qp *qp;
+   unsigned int i;
 
-   PMD_INIT_FUNC_TRACE();
+   for (i = 0; i < qi->max_nb_sessions; i++) {
+   if (qi->inq_attach[i] == 0) {
+   qi->inq_attach[i] = 1;
+   return &qi->inq[i];
+   }
+   }
+   PMD_DRV_LOG(ERR, "All ses session in use %x", qi->max_nb_sessions);
+
+   return NULL;
+}
 
-   qp = dev->data->queue_pairs[qp_id];
-   if (qp->ses != NULL) {
-   PMD_INIT_LOG(ERR, "qp in-use by another session\n");
-   return -EBUSY;
+static int
+dpaa_sec_detach_rxq(struct dpaa_sec_dev_private *qi, struct qman_fq *fq)
+{
+   unsigned int i;
+
+   for (i = 0; i < qi->max_nb_sessions; i++) {
+   if (&qi->inq[i] == fq) {
+   qi->inq_attach[i] = 0;
+   return 0;
+   }
}
+   return -1;
+}
+
+static int
+dpaa_sec_attach_sess_q(struct dpaa_sec_qp *qp, dpaa_sec_session *sess)
+{
+   int ret;
 
-   qp->ses = sess;
sess->qp = qp;
+   ret = dpaa_sec_prep_cdb(sess);
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Unable to prepare sec cdb");
+   return -1;
+   }
 
-   return dpaa_sec_prep_cdb(sess);
+   ret = dpaa_sec_init_rx(sess->inq, dpaa_mem_vtop(&sess->cdb),
+  qman_fq_fqid(&qp->outq));
+   if (ret)
+   PMD_DRV_LOG(ERR, "Unable to init

Re: [dpdk-dev] [PATCH v6 12/23] eventtimer: add event timer arm/cancel function definitions

[Pavan Nikhilesh]

On Wed, Jan 10, 2018 at 06:21:03PM -0600, Erik Gabriel Carrillo wrote:
> Signed-off-by: Erik Gabriel Carrillo 
> ---
>  lib/Makefile  |   2 +-
>  lib/librte_eventdev/Makefile  |   2 +-
>  lib/librte_eventdev/rte_event_timer_adapter.c | 155 
> +++---
>  mk/rte.app.mk |   2 +-
>  4 files changed, 141 insertions(+), 20 deletions(-)
>
> diff --git a/lib/Makefile b/lib/Makefile
> index 4c53f8c..c2bee80 100644
> --- a/lib/Makefile
> +++ b/lib/Makefile
> @@ -29,7 +29,7 @@ DEPDIRS-librte_security := librte_eal librte_mempool 
> librte_ring librte_mbuf
>  DEPDIRS-librte_security += librte_ether
>  DEPDIRS-librte_security += librte_cryptodev
>  DIRS-$(CONFIG_RTE_LIBRTE_EVENTDEV) += librte_eventdev
> -DEPDIRS-librte_eventdev := librte_eal librte_ring librte_ether librte_hash 
> librte_mempool
> +DEPDIRS-librte_eventdev := librte_eal librte_ring librte_ether librte_hash 
> librte_mempool librte_timer
>  DIRS-$(CONFIG_RTE_LIBRTE_VHOST) += librte_vhost
>  DEPDIRS-librte_vhost := librte_eal librte_mempool librte_mbuf librte_ether
>  DIRS-$(CONFIG_RTE_LIBRTE_HASH) += librte_hash
> diff --git a/lib/librte_eventdev/Makefile b/lib/librte_eventdev/Makefile
> index e68f888..6e95528 100644
> --- a/lib/librte_eventdev/Makefile
> +++ b/lib/librte_eventdev/Makefile
> @@ -13,7 +13,7 @@ LIBABIVER := 3
>  # build flags
>  CFLAGS += -O3
>  CFLAGS += $(WERROR_FLAGS)
> -LDLIBS += -lrte_eal -lrte_ring -lrte_ethdev -lrte_hash -lrte_mempool
> +LDLIBS += -lrte_eal -lrte_ring -lrte_ethdev -lrte_hash -lrte_mempool 
> -lrte_timer
>
>  # library source files
>  SRCS-y += rte_eventdev.c
> diff --git a/lib/librte_eventdev/rte_event_timer_adapter.c 
> b/lib/librte_eventdev/rte_event_timer_adapter.c
> index a4c8012..38e52cb 100644
> --- a/lib/librte_eventdev/rte_event_timer_adapter.c
> +++ b/lib/librte_eventdev/rte_event_timer_adapter.c
> @@ -609,16 +609,91 @@ sw_event_timer_adapter_stop(const struct 
> rte_event_timer_adapter *adapter)
>   return rte_service_component_runstate_set(sw_data->service_id, 0);
>  }
>
> +static __rte_always_inline void
> +swap(struct rte_event_timer **evtims, int i, int j)
> +{
> + struct rte_event_timer *tmp;
> +
> + tmp = evtims[i];
> + evtims[i] = evtims[j];
> + evtims[j] = tmp;
> +}
> +
> +static __rte_always_inline int
> +__sw_event_timer_arm_burst(const struct rte_event_timer_adapter *adapter,
> +struct rte_event_timer **evtims,
> +uint16_t nb_evtims)
> +{
> + int i, n, mark, nb_fails = 0;
> + int fails[nb_evtims];
> + struct rte_event_timer_adapter_sw_data *sw_data;
> + struct msg *msgs[nb_evtims];
> + struct rte_timer *tims[nb_evtims];
> +
> + sw_data = adapter->data->adapter_priv;
> +
> + n = rte_mempool_get_bulk(sw_data->msg_pool, (void **)msgs, nb_evtims);
> + if (n < 0) {
> + rte_errno = ENOMEM;
> + return 0;
> + }
> +
> + n = rte_mempool_get_bulk(sw_data->tim_pool, (void **)tims, nb_evtims);
> + if (n < 0) {
> + rte_errno = ENOMEM;

Need to free msg objs that were dequeued prior (Although this has a very low
chance of happening.)
As mentioned before consider having a single pool.

> + return 0;
> + }
> +
> + for (i = 0; i < nb_evtims; i++) {
> + rte_timer_init(tims[i]);
> + evtims[i]->impl_opaque[0] = (uintptr_t)tims[i];
> + evtims[i]->impl_opaque[1] = (uintptr_t)adapter;
> +
> + msgs[i]->evtim = evtims[i];
> + msgs[i]->type = MSG_TYPE_ARM;
> + }
> +
> + n = rte_ring_enqueue_burst(sw_data->msg_ring, (void **)msgs, nb_evtims,
> +NULL);
> + if (n < nb_evtims) {
> + rte_mempool_put_bulk(sw_data->msg_pool, (void **)&msgs[n],
> +  nb_evtims - n);
> + rte_mempool_put_bulk(sw_data->tim_pool, (void **)&tims[n],
> +  nb_evtims - n);
> + }
> +
> + for (i = 0; i < n; i++) {
> + /* Wait until state is updated */
> + while (evtims[i]->state == RTE_EVENT_TIMER_NOT_ARMED ||
> +evtims[i]->state == RTE_EVENT_TIMER_CANCELED)
> + ;
> +
> + /* Note any failures */
> + if (evtims[i]->state != RTE_EVENT_TIMER_ARMED) {
> + fails[nb_fails++] = i;
> + rte_errno = EINVAL;
> + }
> +
> + /* TODO: handle the case of consecutive arm requests;  the
> +  * second request can erroneously see success from the first
> +  */
> + }
> +
> + /* Move the failures to the end of the array */
> + for (i = 0, mark = n - 1; i < nb_fails; i++, mark--)
> + swap(evtims, fails[i], mark);
> +
> + n = mark + 1;
> +
> + return n;
> +}
> +
>  static int
>  sw_event_timer_arm_burst(const struct rt

[dpdk-dev] [PATCH v2] crypto/dpaa_sec: rewrite Rx/Tx path

[Akhil Goyal]

Rx and Tx patch are rewritten with improved internal APIs
to improve performance.

Signed-off-by: Akhil Goyal 
Signed-off-by: Nipun Gupta 
---
changes in v2:
Incorporated comments from Hemant.
split patch from the patch series to remove dependancy of other patches on 
bus/dpaa patch
Note:
This patch is dependent on [1](already merged in net subtree) and [2](under 
review)

[1] c7b3891464c9 ("bus/dpaa: support for enqueue frames of multiple queues")
[2]  http://dpdk.org/ml/archives/dev/2018-January/086674.html

 drivers/crypto/dpaa_sec/dpaa_sec.c | 259 +
 drivers/crypto/dpaa_sec/dpaa_sec.h |   3 +-
 2 files changed, 153 insertions(+), 109 deletions(-)

diff --git a/drivers/crypto/dpaa_sec/dpaa_sec.c 
b/drivers/crypto/dpaa_sec/dpaa_sec.c
index 66828b5..586ebb6 100644
--- a/drivers/crypto/dpaa_sec/dpaa_sec.c
+++ b/drivers/crypto/dpaa_sec/dpaa_sec.c
@@ -537,46 +537,66 @@ dpaa_sec_prep_cdb(dpaa_sec_session *ses)
return 0;
 }
 
-static inline unsigned int
-dpaa_volatile_deq(struct qman_fq *fq, unsigned int len, bool exact)
-{
-   unsigned int pkts = 0;
-   int ret;
-   struct qm_mcr_queryfq_np np;
-   enum qman_fq_state state;
-   uint32_t flags;
-   uint32_t vdqcr;
-
-   qman_query_fq_np(fq, &np);
-   if (np.frm_cnt) {
-   vdqcr = QM_VDQCR_NUMFRAMES_SET(len);
-   if (exact)
-   vdqcr |= QM_VDQCR_EXACT;
-   ret = qman_volatile_dequeue(fq, 0, vdqcr);
-   if (ret)
-   return 0;
-   do {
-   pkts += qman_poll_dqrr(len);
-   qman_fq_state(fq, &state, &flags);
-   } while (flags & QMAN_FQ_STATE_VDQCR);
-   }
-   return pkts;
-}
-
 /* qp is lockless, should be accessed by only one thread */
 static int
 dpaa_sec_deq(struct dpaa_sec_qp *qp, struct rte_crypto_op **ops, int nb_ops)
 {
struct qman_fq *fq;
+   unsigned int pkts = 0;
+   int ret;
+   struct qm_dqrr_entry *dq;
 
fq = &qp->outq;
-   dpaa_sec_op_nb = 0;
-   dpaa_sec_ops = ops;
+   ret = qman_set_vdq(fq, (nb_ops > DPAA_MAX_DEQUEUE_NUM_FRAMES) ?
+   DPAA_MAX_DEQUEUE_NUM_FRAMES : nb_ops);
+   if (ret)
+   return 0;
+
+   do {
+   const struct qm_fd *fd;
+   struct dpaa_sec_job *job;
+   struct dpaa_sec_op_ctx *ctx;
+   struct rte_crypto_op *op;
+
+   dq = qman_dequeue(fq);
+   if (!dq)
+   continue;
+
+   fd = &dq->fd;
+   /* sg is embedded in an op ctx,
+* sg[0] is for output
+* sg[1] for input
+*/
+   job = dpaa_mem_ptov(qm_fd_addr_get64(fd));
+
+   ctx = container_of(job, struct dpaa_sec_op_ctx, job);
+   ctx->fd_status = fd->status;
+   op = ctx->op;
+   if (op->sess_type == RTE_CRYPTO_OP_SECURITY_SESSION) {
+   struct qm_sg_entry *sg_out;
+   uint32_t len;
+
+   sg_out = &job->sg[0];
+   hw_sg_to_cpu(sg_out);
+   len = sg_out->length;
+   op->sym->m_src->pkt_len = len;
+   op->sym->m_src->data_len = len;
+   }
+   if (!ctx->fd_status) {
+   op->status = RTE_CRYPTO_OP_STATUS_SUCCESS;
+   } else {
+   printf("\nSEC return err: 0x%x", ctx->fd_status);
+   op->status = RTE_CRYPTO_OP_STATUS_ERROR;
+   }
+   ops[pkts++] = op;
 
-   if (unlikely(nb_ops > DPAA_SEC_BURST))
-   nb_ops = DPAA_SEC_BURST;
+   /* report op status to sym->op and then free the ctx memeory */
+   rte_mempool_put(ctx->ctx_pool, (void *)ctx);
 
-   return dpaa_volatile_deq(fq, nb_ops, 1);
+   qman_dqrr_consume(fq, dq);
+   } while (fq->flags & QMAN_FQ_STATE_VDQCR);
+
+   return pkts;
 }
 
 /**
@@ -949,95 +969,118 @@ build_proto(struct rte_crypto_op *op, dpaa_sec_session 
*ses)
return cf;
 }
 
-static int
-dpaa_sec_enqueue_op(struct rte_crypto_op *op,  struct dpaa_sec_qp *qp)
-{
-   struct dpaa_sec_job *cf;
-   dpaa_sec_session *ses;
-   struct qm_fd fd;
-   int ret;
-   uint32_t auth_only_len = op->sym->auth.data.length -
-   op->sym->cipher.data.length;
-
-   if (op->sess_type == RTE_CRYPTO_OP_WITH_SESSION)
-   ses = (dpaa_sec_session *)get_session_private_data(
-   op->sym->session, cryptodev_driver_id);
-   else if (op->sess_type == RTE_CRYPTO_OP_SECURITY_SESSION)
-   ses = (dpaa_sec_session *)get_sec_session_private_data(
-   op->sym->sec_session);
-   else
-   return -ENOTSUP;

[dpdk-dev] [PATCH] examples/ipsec-secgw: fix corner case for spi value

[Akhil Goyal]

application is using index 0 of SA table as error,
with current value of IPSEC_SA_MAX_ENTRIES(128) it can
not support SA with spi = 128, as it uses sa_idx = 0
in the SA table.

With this patch, sa_idx = 0 can also be used.

PS: spi = 0 is an invalid SPI and application throws error
for it.

Fixes: d299106e8e31 ("examples/ipsec-secgw: add IPsec sample application")

Signed-off-by: Akhil Goyal 
---
 examples/ipsec-secgw/ipsec-secgw.c | 7 ---
 examples/ipsec-secgw/sa.c  | 2 ++
 2 files changed, 6 insertions(+), 3 deletions(-)

diff --git a/examples/ipsec-secgw/ipsec-secgw.c 
b/examples/ipsec-secgw/ipsec-secgw.c
index d6f7b36..4d50e72 100644
--- a/examples/ipsec-secgw/ipsec-secgw.c
+++ b/examples/ipsec-secgw/ipsec-secgw.c
@@ -420,7 +420,8 @@ inbound_sp_sa(struct sp_ctx *sp, struct sa_ctx *sa, struct 
traffic_type *ip,
}
 
sa_idx = ip->res[i] & PROTECT_MASK;
-   if (sa_idx == 0 || !inbound_sa_check(sa, m, sa_idx)) {
+   if (sa_idx >= IPSEC_SA_MAX_ENTRIES ||
+   !inbound_sa_check(sa, m, sa_idx)) {
rte_pktmbuf_free(m);
continue;
}
@@ -485,9 +486,9 @@ outbound_sp(struct sp_ctx *sp, struct traffic_type *ip,
for (i = 0; i < ip->num; i++) {
m = ip->pkts[i];
sa_idx = ip->res[i] & PROTECT_MASK;
-   if ((ip->res[i] == 0) || (ip->res[i] & DISCARD))
+   if (ip->res[i] & DISCARD)
rte_pktmbuf_free(m);
-   else if (sa_idx != 0) {
+   else if (sa_idx < IPSEC_SA_MAX_ENTRIES) {
ipsec->res[ipsec->num] = sa_idx;
ipsec->pkts[ipsec->num++] = m;
} else /* BYPASS */
diff --git a/examples/ipsec-secgw/sa.c b/examples/ipsec-secgw/sa.c
index 1ae4e2a..25ceea1 100644
--- a/examples/ipsec-secgw/sa.c
+++ b/examples/ipsec-secgw/sa.c
@@ -240,6 +240,8 @@ parse_sa_tokens(char **tokens, uint32_t n_tokens,
APP_CHECK_TOKEN_IS_NUM(tokens, 1, status);
if (status->status < 0)
return;
+   if (atoi(tokens[1]) == INVALID_SPI)
+   return;
rule->spi = atoi(tokens[1]);
 
for (ti = 2; ti < n_tokens; ti++) {
-- 
2.9.3

Re: [dpdk-dev] [PATCH v6 13/23] eventtimer: add adapter service definition

[Pavan Nikhilesh]

On Wed, Jan 10, 2018 at 06:21:04PM -0600, Erik Gabriel Carrillo wrote:
> Define the callback function for the service that corresponds to an
> adapter instance, as well as the callback for expired timers that the
> service manages.
>
> Signed-off-by: Erik Gabriel Carrillo 
> ---
>  lib/librte_eventdev/rte_event_timer_adapter.c | 198 
> +-
>  lib/librte_eventdev/rte_event_timer_adapter.h |   2 +-
>  2 files changed, 198 insertions(+), 2 deletions(-)
>
> diff --git a/lib/librte_eventdev/rte_event_timer_adapter.c 
> b/lib/librte_eventdev/rte_event_timer_adapter.c
> index 38e52cb..0266ad5 100644
> --- a/lib/librte_eventdev/rte_event_timer_adapter.c
> +++ b/lib/librte_eventdev/rte_event_timer_adapter.c
> @@ -40,8 +40,10 @@
>  #include 
>  #include 
>  #include 
> +#include 
>  #include 
>  #include 
> +#include 
>
>  #include "rte_eventdev.h"
>  #include "rte_eventdev_pmd.h"
> @@ -460,10 +462,198 @@ struct msg {
>   struct rte_event_timer *evtim;
>  };

> + if (n != 1 && rte_errno == -ENOSPC) {
> + /* If we couldn't enqueue because the event port was
> +  * backpressured, put the timer back in the skiplist with an
> +  * immediate expiry value so we can process it again on the
> +  * next iteration.
> +  */
> + rte_timer_reset_sync(tim, SINGLE, 0, rte_lcore_id(),
> +  sw_event_timer_cb, evtim);
> + } else {
> + sw_data->nb_armed_evtims--;
> + rte_wmb();

Any reason for using barrier here?. IMO smp_wmb() would be more than sufficient
or use atomics.

> + evtim->state = RTE_EVENT_TIMER_NOT_ARMED;
> + rte_mempool_put(sw_data->tim_pool, (void **)&tim);
> + }
> +}
> +
> +static __rte_always_inline uint64_t
> +get_timeout_cycles(struct rte_event_timer *evtim,
> +struct rte_event_timer_adapter *adapter)
> +{
> + uint64_t timeout_ns;
> +
> + timeout_ns = evtim->timeout_ticks * adapter->data->conf.timer_tick_ns;
> +#define NSECPERSEC 1E9
> + return timeout_ns * rte_get_timer_hz() / NSECPERSEC;
> +
> +}
> +
> +/* Check that event timer timeout value is in range */
> +static __rte_always_inline int
> +check_timeout(struct rte_event_timer *evtim,
> +   const struct rte_event_timer_adapter *adapter)
> +{
> + uint64_t tmo_nsec = evtim->timeout_ticks *
> + adapter->data->conf.timer_tick_ns;
> +
> + return  (tmo_nsec > adapter->data->conf.max_tmo_ns) ? -1
> + : (tmo_nsec < adapter->data->conf.timer_tick_ns) ? -2
> + : 0;

Consider simplifying this for readability.

> +}
> +
> +/* Check that event timer event queue sched type matches destination event 
> queue
> + * sched type
> + */
> +static __rte_always_inline int
> +check_destination_event_queue(struct rte_event_timer *evtim,
> +   const struct rte_event_timer_adapter *adapter)


> +
> +#define NB_OBJS 32
>  static int
>  sw_event_timer_adapter_service_func(void *arg)
>  {
> - RTE_SET_USED(arg);
> + int i, num_msgs, ret;
> + uint64_t cycles;
> + uint16_t nb_events;
> + struct rte_event_timer_adapter *adapter;
> + struct rte_event_timer_adapter_sw_data *sw_data;
> + struct rte_event_timer *evtim = NULL;
> + struct rte_timer *tim = NULL;
> + struct msg *msg, *msgs[NB_OBJS];
> +
> + adapter = arg;
> + sw_data = adapter->data->adapter_priv;
> +
> + while (!rte_ring_empty(sw_data->msg_ring)) {
> + num_msgs = rte_ring_dequeue_burst(sw_data->msg_ring,
> +   (void **)msgs, NB_OBJS, NULL);
> +
> + for (i = 0; i < num_msgs; i++) {
> + msg = msgs[i];
> + evtim = msg->evtim;
> +
> + tim = (struct rte_timer *)evtim->impl_opaque[0];
> + RTE_ASSERT(tim != NULL);
> +
> + switch (msg->type) {
> + case MSG_TYPE_ARM:
> + if (validate_event_timer(evtim, adapter) < 0) {
> + rte_mempool_put(sw_data->tim_pool,
> + (void **)&tim);
> + continue;
> + }
> +
> + /* Checks passed; set an rte_timer */
> + cycles = get_timeout_cycles(msg->evtim,
> + adapter);
> + rte_timer_reset_sync(tim, cycles, SINGLE,
> +  rte_lcore_id(),
> +  sw_event_timer_cb,
> +  msg->evtim);
> +
> + sw_data->nb_armed_evtims++;
> + rte_wmb();

Same as above comment.

> + 

Re: [dpdk-dev] [PATCH v3 09/12] app/eventdev: add pipeline queue worker functions

[Van Haaren, Harry]

> From: Pavan Nikhilesh [mailto:pbhagavat...@caviumnetworks.com]
> Sent: Wednesday, January 10, 2018 8:17 PM
> To: Van Haaren, Harry ;
> jerin.ja...@caviumnetworks.com; santosh.shu...@caviumnetworks.com; Eads,
> Gage ; hemant.agra...@nxp.com; nipun.gu...@nxp.com; Ma,
> Liang J 
> Cc: dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH v3 09/12] app/eventdev: add pipeline queue
> worker functions
> 
> On Wed, Jan 10, 2018 at 04:53:53PM +, Van Haaren, Harry wrote:
> > Replying to self...
> >
> > > -Original Message-
> > > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Van Haaren, Harry
> > > Sent: Wednesday, January 10, 2018 4:45 PM
> > > To: Pavan Nikhilesh ;
> > > jerin.ja...@caviumnetworks.com; santosh.shu...@caviumnetworks.com; Eads,
> > > Gage ; hemant.agra...@nxp.com; nipun.gu...@nxp.com;
> Ma,
> > > Liang J 
> > > Cc: dev@dpdk.org
> > > Subject: Re: [dpdk-dev] [PATCH v3 09/12] app/eventdev: add pipeline
> queue
> > > worker functions
> > >
> > > > From: Pavan Nikhilesh [mailto:pbhagavat...@caviumnetworks.com]
> > > > Sent: Wednesday, January 10, 2018 2:52 PM
> > > > To: jerin.ja...@caviumnetworks.com; santosh.shu...@caviumnetworks.com;
> Van
> > > > Haaren, Harry ; Eads, Gage
> > > > ; hemant.agra...@nxp.com; nipun.gu...@nxp.com;
> Ma,
> > > > Liang J 
> > > > Cc: dev@dpdk.org; Pavan Nikhilesh 
> > > > Subject: [dpdk-dev] [PATCH v3 09/12] app/eventdev: add pipeline queue
> > > worker
> > > > functions
> > > >
> > >
> > > 
> > >
> > >
> > > > +static __rte_always_inline void
> > > > +pipeline_tx_pkt_safe(struct rte_mbuf *mbuf)
> > > > +{
> > > > +   while (rte_eth_tx_burst(mbuf->port, 0, &mbuf, 1) != 1)
> > > > +   rte_pause();
> > > > +}
> > >
> > > re safe, see comment below
> > >
> > > > +
> > > > +static __rte_always_inline void
> > > > +pipeline_tx_pkt_unsafe(struct rte_mbuf *mbuf, struct test_pipeline
> *t)
> > > > +{
> > > > +   rte_spinlock_t *lk = &t->tx_lk[mbuf->port];
> > > > +
> > > > +   rte_spinlock_lock(lk);
> > > > +   pipeline_tx_pkt_safe(mbuf);
> > > > +   rte_spinlock_unlock(lk);
> > > > +}
> > >
> > > IIRC usually the "Safe" version of a function has extra
> locks/protection,
> > > while the "normal" version has better performance, but less-error-
> checking.
> > >
> > > Here, the "unsafe" function does the extra locking. If looking from the
> HW
> > > POV, that makes sense, but I think its inverted from most existing
> code...
> > >
> > > Happy to be proved wrong here .. ?
> > >
> > > 
> >
> >
> > Thinking a little more about this, also in light of patch 11/12 of this
> series.
> >
> > The code here has a "safe" and "unsafe" version of TX. This involves
> adding a spinlock inside the code, which is being locked/unlocked before
> doing the actual TX action.
> >
> > I don't understand why this is necessary? DPDK's general stance on locking
> for data-path is DPDK functions do not provide locks, and that application
> level must implement thread-synchronization if it is required.
> >
> > In this case, the app/eventdev can be considered an App, but I don't like
> the idea of providing a sample application and code that duplicates core
> functionality with safe/unsafe versions..
> >
> 
> Some PMD's (net/octeontx) have capability to do multi-thread safe Tx where
> no
> thread-synchronization is required. This is exposed via the offload flag
> 'DEV_TX_OFFLOAD_MT_LOCKFREE'.

Yes understood.


> So, the _safe Tx functions are selected based on the above offload
> capability
> and when the capability is absent _unsafe Tx functions are selected i.e.
> synchronized Tx via spin locks based on the Egress port id.


This part changes the current behavior of the sample app.

Currently there is a (SINGLE_LINK | ATOMIC) stage at the end of the pipeline, 
which performs this "many-to-one" action, allowing a single core to dequeue all 
TX traffic, and perform the TX operation in a lock-free manner.

Changing this to a locking mechanism is going to hurt performance on platforms 
that do not support TX_OFFLOAD_MT_LOCKFREE.

In my opinion, the correct fix is to alter the overall pipeline, and always use 
lockless TX. Examples below;

NO TX_OFFLOAD_MT_LOCKFREE:

   Eth RX adapter -> stage 1 -> stage 2...(N-1) -> stage N -> stage TX (Atomic 
| SINGLE_LINK) -> eth TX


WITH TX_OFFLOAD_MT_LOCKFREE:

   Eth RX adapter -> stage 1 -> stage 2...(N-1) -> stage N -> eth TX MT Capable


By configuring the pipeline based on MT_OFFLOAD_LOCKFREE capability flag, and 
adding the SINGLE_LINK at the end if required, we can support both models 
without resorting to locked TX functions.

I think this will lead to a cleaner and more performant solution.

Re: [dpdk-dev] [PATCH v6 15/23] eventtimer: add buffering of timer expiry events

[Pavan Nikhilesh]

On Wed, Jan 10, 2018 at 06:21:06PM -0600, Erik Gabriel Carrillo wrote:
> Buffer timer expiry events generated while walking a "run list"
> in rte_timer_manage, and burst enqueue them to an event device
> to the extent possible.
>

IMO in some cases this adds a lot of delay between expiries and events being
published to event dev. For example, having long expiry interval (default 300
seconds for mac expiry) the expired entries would remain in the buffer till 32
other entries expire.


> Signed-off-by: Erik Gabriel Carrillo 
> ---
>  lib/librte_eventdev/rte_event_timer_adapter.c | 118 
> +++---
>  1 file changed, 108 insertions(+), 10 deletions(-)
>

Re: [dpdk-dev] [PATCH v6 19/23] test: exercise event timer arm and expiry

[Pavan Nikhilesh]

On Wed, Jan 10, 2018 at 06:21:10PM -0600, Erik Gabriel Carrillo wrote:
> Add a test that creates an event timer and detects the generation of a
> timer expiry event being scheduled through the software event device.
>
> Signed-off-by: Erik Gabriel Carrillo 
> ---
>  test/test/test_event_timer_adapter.c | 106 
> +++
>  1 file changed, 106 insertions(+)
>

Please consider following common_code -> sw_dev -> test -> doc so that it would
be easy to review. Some patches could be squashed into one.

Re: [dpdk-dev] [PATCH v2 2/6] ethdev: add port ownership

[Ananyev, Konstantin]

Hi Matan,

> 
> Hi Konstantin
> 
> From: Ananyev, Konstantin, Wednesday, January 10, 2018 3:36 PM
> > Hi Matan,
> >
> > Few comments from me below.
> > BTW, do you plan to add ownership mandatory check in control path
> > functions that change port configuration?
> 
> No.

So it still totally voluntary usage and application nneds to be changed
to exploit it?
Apart from RTE_FOR_EACH_DEV() change proposed by Gaetan?

> 
> 
> > Konstantin
> >
> > > -Original Message-
> > > From: Matan Azrad [mailto:ma...@mellanox.com]
> > > Sent: Sunday, January 7, 2018 9:46 AM
> > > To: Thomas Monjalon ; Gaetan Rivet
> > > ; Wu, Jingjing 
> > > Cc: dev@dpdk.org; Neil Horman ; Richardson,
> > > Bruce ; Ananyev, Konstantin
> > > 
> > > Subject: [PATCH v2 2/6] ethdev: add port ownership
> > >
> > > The ownership of a port is implicit in DPDK.
> > > Making it explicit is better from the next reasons:
> > > 1. It will define well who is in charge of the port usage synchronization.
> > > 2. A library could work on top of a port.
> > > 3. A port can work on top of another port.
> > >
> > > Also in the fail-safe case, an issue has been met in testpmd.
> > > We need to check that the application is not trying to use a port
> > > which is already managed by fail-safe.
> > >
> > > A port owner is built from owner id(number) and owner name(string)
> > > while the owner id must be unique to distinguish between two identical
> > > entity instances and the owner name can be any name.
> > > The name helps to logically recognize the owner by different DPDK
> > > entities and allows easy debug.
> > > Each DPDK entity can allocate an owner unique identifier and can use
> > > it and its preferred name to owns valid ethdev ports.
> > > Each DPDK entity can get any port owner status to decide if it can
> > > manage the port or not.
> > >
> > > The mechanism is synchronized for both the primary process threads and
> > > the secondary processes threads to allow secondary process entity to
> > > be a port owner.
> > >
> > > Add a sinchronized ownership mechanism to DPDK Ethernet devices to
> > > avoid multiple management of a device by different DPDK entities.
> > >
> > > The current ethdev internal port management is not affected by this
> > > feature.
> > >
> > > Signed-off-by: Matan Azrad 
> > > ---
> > >  doc/guides/prog_guide/poll_mode_drv.rst |  14 ++-
> > >  lib/librte_ether/rte_ethdev.c   | 206
> > ++--
> > >  lib/librte_ether/rte_ethdev.h   |  89 ++
> > >  lib/librte_ether/rte_ethdev_version.map |  12 ++
> > >  4 files changed, 311 insertions(+), 10 deletions(-)
> >
> >
> > >
> > >
> > > diff --git a/lib/librte_ether/rte_ethdev.c
> > > b/lib/librte_ether/rte_ethdev.c index 684e3e8..0e12452 100644
> > > --- a/lib/librte_ether/rte_ethdev.c
> > > +++ b/lib/librte_ether/rte_ethdev.c
> > > @@ -70,7 +70,10 @@
> > >
> > >  static const char *MZ_RTE_ETH_DEV_DATA = "rte_eth_dev_data";  struct
> > > rte_eth_dev rte_eth_devices[RTE_MAX_ETHPORTS];
> > > +/* ports data array stored in shared memory */
> > >  static struct rte_eth_dev_data *rte_eth_dev_data;
> > > +/* next owner identifier stored in shared memory */ static uint16_t
> > > +*rte_eth_next_owner_id;
> > >  static uint8_t eth_dev_last_created_port;
> > >
> > >  /* spinlock for eth device callbacks */ @@ -82,6 +85,9 @@
> > >  /* spinlock for add/remove tx callbacks */  static rte_spinlock_t
> > > rte_eth_tx_cb_lock = RTE_SPINLOCK_INITIALIZER;
> > >
> > > +/* spinlock for eth device ownership management stored in shared
> > > +memory */ static rte_spinlock_t *rte_eth_dev_ownership_lock;
> > > +
> > >  /* store statistics names and its offset in stats structure  */
> > > struct rte_eth_xstats_name_off {
> > >   char name[RTE_ETH_XSTATS_NAME_SIZE]; @@ -153,14 +159,18 @@
> > enum {  }
> > >
> > >  static void
> > > -rte_eth_dev_data_alloc(void)
> > > +rte_eth_dev_share_data_alloc(void)
> > >  {
> > >   const unsigned flags = 0;
> > >   const struct rte_memzone *mz;
> > > + const unsigned int data_size = RTE_MAX_ETHPORTS *
> > > + sizeof(*rte_eth_dev_data);
> > >
> > >   if (rte_eal_process_type() == RTE_PROC_PRIMARY) {
> > > + /* Allocate shared memory for port data and ownership */
> > >   mz = rte_memzone_reserve(MZ_RTE_ETH_DEV_DATA,
> > > - RTE_MAX_ETHPORTS *
> > sizeof(*rte_eth_dev_data),
> > > + data_size + sizeof(*rte_eth_next_owner_id)
> > +
> > > + sizeof(*rte_eth_dev_ownership_lock),
> > >   rte_socket_id(), flags);
> > >   } else
> > >   mz = rte_memzone_lookup(MZ_RTE_ETH_DEV_DATA);
> > > @@ -168,9 +178,17 @@ enum {
> > >   rte_panic("Cannot allocate memzone for ethernet port
> > data\n");
> > >
> > >   rte_eth_dev_data = mz->addr;
> > > - if (rte_eal_process_type() == RTE_PROC_PRIMARY)
> > > - memset(rte_eth_dev_data, 0,
> > > -  

Re: [dpdk-dev] [PATCH 2/4] drivers/net: add drivers for Cavium NICs to meson build

[Pavan Nikhilesh]

On Tue, Jan 09, 2018 at 05:29:36PM +, Bruce Richardson wrote:
> On Sun, Dec 31, 2017 at 05:32:01AM +0530, Pavan Nikhilesh wrote:
> > Add Cavium octeontx and thunder nicvf to meson build infrastructure.
> >
> > Signed-off-by: Pavan Nikhilesh 
> > ---
> >  drivers/net/meson.build |  3 ++-
> >  drivers/net/octeontx/base/meson.build   | 12 
> >  drivers/net/octeontx/meson.build| 10 ++
> >  drivers/net/thunderx/Makefile   |  2 +-
> >  drivers/net/thunderx/base/meson.build   | 17 
> > +
> >  drivers/net/thunderx/meson.build|  9 +
> >  ...x_nicvf_version.map => rte_pmd_thunderx_version.map} |  0
> >  7 files changed, 51 insertions(+), 2 deletions(-)
> >  create mode 100644 drivers/net/octeontx/base/meson.build
> >  create mode 100644 drivers/net/octeontx/meson.build
> >  create mode 100644 drivers/net/thunderx/base/meson.build
> >  create mode 100644 drivers/net/thunderx/meson.build
> >  rename drivers/net/thunderx/{rte_pmd_thunderx_nicvf_version.map => 
> > rte_pmd_thunderx_version.map} (100%)
> >
>
> Some minor comments below.
>
> > diff --git a/drivers/net/meson.build b/drivers/net/meson.build
> > index 72e1d3578..b132d3558 100644
> > --- a/drivers/net/meson.build
> > +++ b/drivers/net/meson.build
> > @@ -3,7 +3,8 @@
> >
> >  drivers = ['af_packet', 'bonding',
> > 'e1000', 'fm10k', 'i40e', 'ixgbe',
> > -   'null', 'pcap', 'ring']
> > +   'null', 'pcap', 'ring',
> > +   'octeontx', 'thunderx']
> >  std_deps = ['ethdev', 'kvargs'] # 'ethdev' also pulls in mbuf, net, eal etc
> >  std_deps += ['bus_pci'] # very many PMDs depend on PCI, so make std
> >  std_deps += ['bus_vdev']# same with vdev bus
> > diff --git a/drivers/net/octeontx/base/meson.build 
> > b/drivers/net/octeontx/base/meson.build
> > new file mode 100644
> > index 0..637483d68
> > --- /dev/null
> > +++ b/drivers/net/octeontx/base/meson.build
> > @@ -0,0 +1,12 @@
> > +sources = [
> > +   'octeontx_bgx.c',
> > +   'octeontx_pkivf.c',
> > +   'octeontx_pkovf.c'
> > +]
> > +
> > +base_lib = static_library('octeontx_base', sources,
> > +   c_args: cflags,
> > +   dependencies: static_rte_bus_pci,
> > +   include_directories: mempool_octeontx_dir)
> > +
> > +base_objs = base_lib.extract_all_objects()
> > diff --git a/drivers/net/octeontx/meson.build 
> > b/drivers/net/octeontx/meson.build
> > new file mode 100644
> > index 0..f263a4b51
> > --- /dev/null
> > +++ b/drivers/net/octeontx/meson.build
> > @@ -0,0 +1,10 @@
> > +subdir('base')
> > +objs = [base_objs]
> > +
> > +sources = files('octeontx_rxtx.c',
> > +   'octeontx_ethdev.c'
> > +   )
> > +
> > +deps += ['mempool_octeontx', 'eventdev']
> > +
> > +includes += include_directories('base')
> > diff --git a/drivers/net/thunderx/Makefile b/drivers/net/thunderx/Makefile
> > index e50e1ad8c..3e4501d34 100644
> > --- a/drivers/net/thunderx/Makefile
> > +++ b/drivers/net/thunderx/Makefile
> > @@ -44,7 +44,7 @@ LDLIBS += -lrte_eal -lrte_mbuf -lrte_mempool -lrte_ring
> >  LDLIBS += -lrte_ethdev -lrte_net -lrte_kvargs
> >  LDLIBS += -lrte_bus_pci
> >
> > -EXPORT_MAP := rte_pmd_thunderx_nicvf_version.map
> > +EXPORT_MAP := rte_pmd_thunderx_version.map
> >
> >  LIBABIVER := 1
> >
> > diff --git a/drivers/net/thunderx/base/meson.build 
> > b/drivers/net/thunderx/base/meson.build
> > new file mode 100644
> > index 0..867a6d661
> > --- /dev/null
> > +++ b/drivers/net/thunderx/base/meson.build
> > @@ -0,0 +1,17 @@
> > +sources = [
> > +   'nicvf_bsvf.c',
> > +   'nicvf_hw.c',
> > +   'nicvf_mbox.c'
> > +]
> > +
> > +depends = ['ethdev']
> > +static_obj = []
> > +foreach d: depends
> > +   static_obj += [get_variable('static_rte_' + d)]
> > +endforeach
> > +
> > +base_lib = static_library('nicvf_base', sources,
> > +   c_args: cflags,
> > +   dependencies: static_obj)
> > +
>
> Given there is only one dependency here, I think it's probably best just
> to do like you did with the octeontx base code, and put it explicitly in
> the dependencies line, rather than bothering with a loop.

Agreed, will remove the loop and include directly.

>
> > +base_objs = base_lib.extract_all_objects()
> > diff --git a/drivers/net/thunderx/meson.build 
> > b/drivers/net/thunderx/meson.build
> > new file mode 100644
> > index 0..17852f732
> > --- /dev/null
> > +++ b/drivers/net/thunderx/meson.build
> > @@ -0,0 +1,9 @@
> > +subdir('base')
> > +objs = [base_objs]
> > +
> > +sources = files('nicvf_ethdev.c',
> > +   'nicvf_svf.c',
> > +   'nicvf_rxtx.c')
> > +cflags += '-Ofast'
>
> Is there a reason this is explicitly needed? The optimization flags for
> the DPDK build as a whole should be covered by the meson top-level
> flags, based on the build type. With this given here, changing the build
> type to "debug" or "debugoptimized" may not work as expected for the
> driver.

I do

Re: [dpdk-dev] [PATCH v5] sched: make RED scaling configurable

[Dumitrescu, Cristian]

Hi Alan,

More issues and questions below:

...

> diff --git a/lib/librte_sched/rte_red.c b/lib/librte_sched/rte_red.h
I suggest we use uint32_t for rte_red_scaling and rte_red_max_threshold, even 
though their values can fit into uint8_t and uint16_t respectively.


> diff --git a/lib/librte_sched/rte_red.c b/lib/librte_sched/rte_red.c
> +int
> +rte_red_set_scaling(uint16_t max_red_queue_length)
> +{
> + int8_t count;
> +
> + if (rte_red_init_done)
> + /**
> +  * Can't change the scaling once the red table has been
> +  * computed.
> +  */
> + return -1;

Is there a reason why we cannot simply reset the scaling here?

> +
> + if (max_red_queue_length < RTE_RED_MIN_QUEUE_LENGTH)
> + return -2;
> +
> + if (max_red_queue_length > RTE_RED_MAX_QUEUE_LENGTH)
> + return -3;
> +
> + if (!rte_is_power_of_2(max_red_queue_length))
> + return -4;
> +
> + count = 0;
> + while (max_red_queue_length != 0) {
> + max_red_queue_length >>= 1;
> + count++;
> + }

This does not look right to me. I think you want to compute the log2 of 
max_red_queue_length here, but your result (count) is bigger by 1 than it 
should be, right?
When max_red_queue_length = RTE_RED_DEFAULT_QUEUE_LENGTH = 1024, the result 
should be: count = RTE_RED_SCALING = 10, not 11.

I suggest you use rte_bsf32() function for this purpose.

> +
> + rte_red_scaling -= count - RTE_RED_SCALING;

Why not simply: rte_red_scaling = count?

> + rte_red_max_threshold = max_red_queue_length - 1;
> + return 0;
> +}
> +

> diff --git a/lib/librte_sched/rte_sched_version.map
> b/lib/librte_sched/rte_sched_version.map
> index 3aa159a..262bece 100644
> --- a/lib/librte_sched/rte_sched_version.map
> +++ b/lib/librte_sched/rte_sched_version.map
> @@ -29,3 +29,10 @@ DPDK_2.1 {
>   rte_sched_port_pkt_read_color;
> 
>  } DPDK_2.0;
> +
> +DPDK_17.08 {
> + global;
> +
> + __rte_red_reset;
> + rte_red_set_scaling;
> +} DPDK_2.1;

You need to put the correct DPDK release number here.

You also need to do more work to make sure the share library support  is not 
broken:
-need to increase LIBABIVER in Makefile
-need to update the library .so number in release notes
-need to check that build does not fail for shared libraries

(You can use this patch as example: https://dpdk.org/dev/patchwork/patch/33109/ 
)

> diff --git a/test/test/test_red.c b/test/test/test_red.c
The RED  test code fails to link, please fix:

export RTE_TARGET=x86_64-native-linuxapp-gcc

[ ]$ ~/git_dpdk_red/dpdk$ make -C test/test
make: Entering directory 'dpdk/test/test'
...
   LD test
test_red.o: In function `test_rte_red_init':
test_red.c:(.text+0x2cd): undefined reference to `__rte_red_reset'
test_red.c:(.text+0x2da): undefined reference to `rte_red_set_scaling'
test_red.c:(.text+0x2ed): undefined reference to `rte_red_max_threshold'
test_red.o: In function `increase_actual_qsize':
test_red.c:(.text+0x48c): undefined reference to `rte_red_scaling'
test_red.c:(.text+0x533): undefined reference to `rte_red_scaling'
test_red.o: In function `enqueue_dequeue_func':
test_red.c:(.text+0x70c): undefined reference to `rte_red_scaling'
test_red.c:(.text+0x7a5): undefined reference to `rte_red_scaling'
test_red.o: In function `increase_average_qsize':
test_red.c:(.text+0x95c): undefined reference to `rte_red_scaling'
test_red.o:test_red.c:(.text+0x9f1): more undefined references to 
`rte_red_scaling' follow
collect2: error: ld returned 1 exit status
dpdk/mk/rte.app.mk:306: recipe for target 'test' failed
make: *** [test] Error 1
make: Leaving directory 'dpdk/test/test'

Regards,
Cristian

Re: [dpdk-dev] [PATCH 3/4] event/octeontx: add octeontx event device to meson build

[Pavan Nikhilesh]

On Tue, Jan 09, 2018 at 05:34:00PM +, Bruce Richardson wrote:
> On Sun, Dec 31, 2017 at 05:32:02AM +0530, Pavan Nikhilesh wrote:
> > Add Cavium octeontx to meson build and change order of drivers built
> > from event->mempool->net to mempool->net->event to resolve dependency.
> >
> > Signed-off-by: Pavan Nikhilesh 
> > ---
> >  drivers/event/meson.build| 2 +-
> >  drivers/event/octeontx/meson.build   | 5 
> > +
> >  ...octeontx_ssovf_version.map => rte_pmd_octeontx_event_version.map} | 0
> >  drivers/meson.build  | 2 +-
> >  4 files changed, 7 insertions(+), 2 deletions(-)
> >  create mode 100644 drivers/event/octeontx/meson.build
> >  rename drivers/event/octeontx/{rte_pmd_octeontx_ssovf_version.map => 
> > rte_pmd_octeontx_event_version.map} (100%)
> >
> > diff --git a/drivers/event/meson.build b/drivers/event/meson.build
> > index 437d80bfd..d7bc48545 100644
> > --- a/drivers/event/meson.build
> > +++ b/drivers/event/meson.build
> > @@ -1,7 +1,7 @@
> >  # SPDX-License-Identifier: BSD-3-Clause
> >  # Copyright(c) 2017 Intel Corporation
> >
> > -drivers = ['skeleton', 'sw']
> > +drivers = ['skeleton', 'sw', 'octeontx']
> >  std_deps = ['eventdev', 'kvargs']
> >  config_flag_fmt = 'RTE_LIBRTE_@0@_EVENTDEV_PMD'
> >  driver_name_fmt = 'rte_pmd_@0@_event'
> > diff --git a/drivers/event/octeontx/meson.build 
> > b/drivers/event/octeontx/meson.build
> > new file mode 100644
> > index 0..fa46a67b5
> > --- /dev/null
> > +++ b/drivers/event/octeontx/meson.build
> > @@ -0,0 +1,5 @@
> > +sources = files('ssovf_evdev.c',
> > +   'ssovf_worker.c'
> > +)
> > +
> > +deps += ['mempool_octeontx', 'bus_vdev', 'pmd_octeontx']
> > diff --git a/drivers/event/octeontx/rte_pmd_octeontx_ssovf_version.map 
> > b/drivers/event/octeontx/rte_pmd_octeontx_event_version.map
> > similarity index 100%
> > rename from drivers/event/octeontx/rte_pmd_octeontx_ssovf_version.map
> > rename to drivers/event/octeontx/rte_pmd_octeontx_event_version.map
> > diff --git a/drivers/meson.build b/drivers/meson.build
> > index f5009aa2e..52d7176a3 100644
> > --- a/drivers/meson.build
> > +++ b/drivers/meson.build
> > @@ -1,7 +1,7 @@
> >  # SPDX-License-Identifier: BSD-3-Clause
> >  # Copyright(c) 2017 Intel Corporation
> >
> > -driver_classes = ['bus', 'crypto', 'event', 'mempool', 'net']
> > +driver_classes = ['bus', 'crypto', 'mempool', 'net', 'event']
>
> This change is ok with me, but I think you need to add an explicit
> comment in the file - not just in the comment message - as to why the
> order is not alphabetical.
>
> In fact, it might be good to split out each type on it's own line with a
> comment, since bus should almost certainly go first, then probably
> mempool - since we might have crypto drivers in future that depend on
> it, and then finally the other types.

Agreed, will follow the order dictated in make filesystem.
>
> >
> >  foreach class:driver_classes
> > drivers = []
> > --
> > 2.15.1
> >

[dpdk-dev] [PATCH] net/i40e: fix VF reset stats crash

[David C Harton]

Calling i40evf_dev_xstats_reset can sometimes crash.

Fixed issue by adding a pstats NULL check.

Fixes: da61cd0849766 ("i40evf: add extended stats")
   8210e9e0d805e ("net/i40e: fix clear xstats bug in VF")
Cc: wei.zh...@intel.com

Signed-off-by: David C Harton 
---
 drivers/net/i40e/i40e_ethdev_vf.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/i40e/i40e_ethdev_vf.c 
b/drivers/net/i40e/i40e_ethdev_vf.c
index b96d77a..9c773fe 100644
--- a/drivers/net/i40e/i40e_ethdev_vf.c
+++ b/drivers/net/i40e/i40e_ethdev_vf.c
@@ -923,7 +923,8 @@ i40evf_dev_xstats_reset(struct rte_eth_dev *dev)
i40evf_query_stats(dev, &pstats);
 
/* set stats offset base on current values */
-   vf->vsi.eth_stats_offset = *pstats;
+   if (pstats != NULL)
+   vf->vsi.eth_stats_offset = *pstats;
 }
 
 static int i40evf_dev_xstats_get_names(__rte_unused struct rte_eth_dev *dev,
-- 
2.10.3.dirty

Re: [dpdk-dev] [PATCH] maintainers: update for enic

[Ferruh Yigit]

On 1/10/2018 9:17 AM, John Daley wrote:
> Welcome Hyong Youb Kim.
> 
> Signed-off-by: John Daley 

Series applied to dpdk-next-net/master, thanks.

Welcome Hyong Youb Kim.

[dpdk-dev] [PATCH] net/sfc/base: change license to BSD-3-Clause

[Andrew Rybchenko]

Change license from BSD-2-Clause-FreeBSD to BSD-3-Clause.
Bump copyright year.

Fixes: 0b6f6260ac6c ("net/sfc/base: change license to BSD-3-Clause")
Cc: Ferruh Yigit 

Signed-off-by: Andrew Rybchenko 
---
Could you squash the patch into changeset specified in Fixes tag.

 drivers/net/sfc/base/efx_tunnel.c | 30 +++---
 1 file changed, 3 insertions(+), 27 deletions(-)

diff --git a/drivers/net/sfc/base/efx_tunnel.c 
b/drivers/net/sfc/base/efx_tunnel.c
index 8a8ac32..25fa976 100644
--- a/drivers/net/sfc/base/efx_tunnel.c
+++ b/drivers/net/sfc/base/efx_tunnel.c
@@ -1,31 +1,7 @@
-/*
- * Copyright (c) 2017 Solarflare Communications Inc.
- * All rights reserved.
- *
- * Redistribution and use in source and binary forms, with or without
- * modification, are permitted provided that the following conditions are met:
- *
- * 1. Redistributions of source code must retain the above copyright notice,
- *this list of conditions and the following disclaimer.
- * 2. Redistributions in binary form must reproduce the above copyright notice,
- *this list of conditions and the following disclaimer in the documentation
- *and/or other materials provided with the distribution.
+/* SPDX-License-Identifier: BSD-3-Clause
  *
- * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
- * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
- * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
- * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
- * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
- * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
- * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
- * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
- * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
- * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
- * EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
- *
- * The views and conclusions contained in the software and documentation are
- * those of the authors and should not be interpreted as representing official
- * policies, either expressed or implied, of the FreeBSD Project.
+ * Copyright (c) 2007-2018 Solarflare Communications Inc.
+ * All rights reserved.
  */
 
 #include "efx.h"
-- 
2.7.4

Re: [dpdk-dev] [PATCH 1/4] drivers/mempool: add octeontx mempool driver to meson build

[Pavan Nikhilesh]

On Tue, Jan 09, 2018 at 05:35:28PM +, Bruce Richardson wrote:
> On Sun, Dec 31, 2017 at 05:32:00AM +0530, Pavan Nikhilesh wrote:
> > Add octeontx hardware mempool driver to meson build.
> >
> > Signed-off-by: Pavan Nikhilesh 
> > ---
> >  drivers/mempool/meson.build  | 2 +-
> >  drivers/mempool/octeontx/meson.build | 8 
> >  2 files changed, 9 insertions(+), 1 deletion(-)
> >  create mode 100644 drivers/mempool/octeontx/meson.build
> >
> > diff --git a/drivers/mempool/meson.build b/drivers/mempool/meson.build
> > index a47b1d9b1..599185609 100644
> > --- a/drivers/mempool/meson.build
> > +++ b/drivers/mempool/meson.build
> > @@ -1,7 +1,7 @@
> >  # SPDX-License-Identifier: BSD-3-Clause
> >  # Copyright(c) 2017 Intel Corporation
> >
> > -drivers = ['ring', 'stack']
> > +drivers = ['ring', 'stack', 'octeontx']
> >  std_deps = ['mempool']
> >  config_flag_fmt = 'RTE_LIBRTE_@0@_MEMPOOL'
> >  driver_name_fmt = 'rte_mempool_@0@'
> > diff --git a/drivers/mempool/octeontx/meson.build 
> > b/drivers/mempool/octeontx/meson.build
> > new file mode 100644
> > index 0..57ddf58b7
> > --- /dev/null
> > +++ b/drivers/mempool/octeontx/meson.build
> > @@ -0,0 +1,8 @@
> > +sources = files('octeontx_ssovf.c',
> > +   'octeontx_mbox.c',
> > +   'octeontx_fpavf.c',
> > +   'rte_mempool_octeontx.c'
> > +)
> > +deps += ['mbuf', 'bus_pci']
> > +
> > +mempool_octeontx_dir = include_directories('.')
>
> Is this variable used in later patches, I don't see it on a quick scan?
> Normally we just use the dependency objects to manage paths to include
> directories, and we don't need to explicitly have include variables.

I was using this in 'drivers/net/octeontx/base' while experimenting will remove
in v2.

>
> /Bruce

Re: [dpdk-dev] [PATCH 1/4] drivers/mempool: add octeontx mempool driver to meson build

[Pavan Nikhilesh]

On Tue, Jan 09, 2018 at 05:35:28PM +, Bruce Richardson wrote:
> On Sun, Dec 31, 2017 at 05:32:00AM +0530, Pavan Nikhilesh wrote:
> > Add octeontx hardware mempool driver to meson build.
> >
> > Signed-off-by: Pavan Nikhilesh 
> > ---
> >  drivers/mempool/meson.build  | 2 +-
> >  drivers/mempool/octeontx/meson.build | 8 
> >  2 files changed, 9 insertions(+), 1 deletion(-)
> >  create mode 100644 drivers/mempool/octeontx/meson.build
> >
> > diff --git a/drivers/mempool/meson.build b/drivers/mempool/meson.build
> > index a47b1d9b1..599185609 100644
> > --- a/drivers/mempool/meson.build
> > +++ b/drivers/mempool/meson.build
> > @@ -1,7 +1,7 @@
> >  # SPDX-License-Identifier: BSD-3-Clause
> >  # Copyright(c) 2017 Intel Corporation
> >
> > -drivers = ['ring', 'stack']
> > +drivers = ['ring', 'stack', 'octeontx']
> >  std_deps = ['mempool']
> >  config_flag_fmt = 'RTE_LIBRTE_@0@_MEMPOOL'
> >  driver_name_fmt = 'rte_mempool_@0@'
> > diff --git a/drivers/mempool/octeontx/meson.build 
> > b/drivers/mempool/octeontx/meson.build
> > new file mode 100644
> > index 0..57ddf58b7
> > --- /dev/null
> > +++ b/drivers/mempool/octeontx/meson.build
> > @@ -0,0 +1,8 @@
> > +sources = files('octeontx_ssovf.c',
> > +   'octeontx_mbox.c',
> > +   'octeontx_fpavf.c',
> > +   'rte_mempool_octeontx.c'
> > +)
> > +deps += ['mbuf', 'bus_pci']
> > +
> > +mempool_octeontx_dir = include_directories('.')
>
> Is this variable used in later patches, I don't see it on a quick scan?
> Normally we just use the dependency objects to manage paths to include
> directories, and we don't need to explicitly have include variables.
>

I was using it in 'drivers/net/octeontx/base' while experimenting with meson, I
will remove it in v2.

> /Bruce

[dpdk-dev] [PATCH v2] net/i40e: fix VF reset stats crash

[David C Harton]

Calling i40evf_dev_xstats_reset can sometimes crash.

Fixed issue by adding a pstats NULL check.

Fixes: 8210e9e0d805e ("net/i40e: fix clear xstats bug in VF")
Cc: wei.zh...@intel.com

Signed-off-by: David C Harton 
---

v2:
* Fixed fixes line in commit message.


 drivers/net/i40e/i40e_ethdev_vf.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/i40e/i40e_ethdev_vf.c 
b/drivers/net/i40e/i40e_ethdev_vf.c
index b96d77a..9c773fe 100644
--- a/drivers/net/i40e/i40e_ethdev_vf.c
+++ b/drivers/net/i40e/i40e_ethdev_vf.c
@@ -923,7 +923,8 @@ i40evf_dev_xstats_reset(struct rte_eth_dev *dev)
i40evf_query_stats(dev, &pstats);
 
/* set stats offset base on current values */
-   vf->vsi.eth_stats_offset = *pstats;
+   if (pstats != NULL)
+   vf->vsi.eth_stats_offset = *pstats;
 }
 
 static int i40evf_dev_xstats_get_names(__rte_unused struct rte_eth_dev *dev,
-- 
2.10.3.dirty

[dpdk-dev] [PATCH] app/testpmd: add meter to the actions table

[Tomasz Duszynski]

Since METER action is supported by the testpmd application
suitable entry should exist in flow actions information table.

Without that testpmd will return error on adding a new flow to
the list of flows attached to a given port.

Signed-off-by: Tomasz Duszynski 
---
 app/test-pmd/config.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/app/test-pmd/config.c b/app/test-pmd/config.c
index 0a84481..4ad19fb 100644
--- a/app/test-pmd/config.c
+++ b/app/test-pmd/config.c
@@ -1038,6 +1038,7 @@ static const struct {
MK_FLOW_ACTION(RSS, sizeof(struct rte_flow_action_rss)), /* +queue[] */
MK_FLOW_ACTION(PF, 0),
MK_FLOW_ACTION(VF, sizeof(struct rte_flow_action_vf)),
+   MK_FLOW_ACTION(METER, sizeof(struct rte_flow_action_meter)),
 };
 
 /** Compute storage space needed by action configuration. */
-- 
2.7.4

[dpdk-dev] [pull-request] next-qos 18.02 pre-rc1

[Cristian Dumitrescu]

The following changes since commit f06125c07d6203a84e9b242c62d6a8e532a5c51d:

  doc: add metering and policing in testpmd guide (2018-01-10 00:54:52 +0100)

are available in the git repository at:

  http://dpdk.org/git/next/dpdk-next-qos 

for you to fetch changes up to 9550cc5fe3408b33d53673c723a5d08259b35a28:

  lib/librte_meter: add meter configuration profile (2018-01-11 11:17:18 +)


Cristian Dumitrescu (1):
  lib/librte_meter: add meter configuration profile

 doc/guides/rel_notes/deprecation.rst   |   3 -
 doc/guides/rel_notes/release_18_02.rst |   2 +-
 .../pipeline/pipeline_flow_actions_be.c|  25 ++-
 examples/qos_meter/main.c  |  39 -
 examples/qos_meter/main.h  |  32 ++--
 lib/librte_meter/Makefile  |   2 +-
 lib/librte_meter/rte_meter.c   |  93 ++
 lib/librte_meter/rte_meter.h   | 195 ++---
 lib/librte_meter/rte_meter_version.map |   8 +
 test/test/test_meter.c | 194 +++-
 10 files changed, 392 insertions(+), 201 deletions(-)

Re: [dpdk-dev] DPDK-pktgen not giving expected throughput

[Wiles, Keith]

> On Jan 11, 2018, at 12:45 AM, Shailja Pandey  wrote:
> 
> I have also tried with multiple cpus using the command line-
> 
> ./app/app/x86_64-native-linuxapp-gcc/pktgen -c 0x -n 4 -- -T -P -m 
> "[0:4].0,[5:8].1,[9-12].2,[13-16].3” 

What is port 0 and 1, I guess that port 2 and 3 are the 2x40 card. I guess I 
need the ‘lspci | grep Ether’ output too.

Lets assume port 2 and 3 are the 2 ports of 40G and the NIC is on PCI bus0 
attached to NUMA node 0.

May need to blacklist the first two port 0 and 1 to remove them from being used.

./app/app/x86_64-native-linuxapp-gcc/pktgen -c 0x -n 4 -- -T -P -m 
“[2,4,6:8,10,12].0, [14,16,18:20,22,24].1”

Now you have 6 cores per port 3 on TX and 3 on RX on each port. You maybe be 
able to get away with 4 per port with the speed of the machine you have.

Next we have to use the range command to make sure the packets are transmitted 
with a varying set of 5 tuples.

Do a help command an look at the range commands, there is an example in the 
test directory.

To see the range information use ‘page range’ and then you can modify each port.

After you are happy with the configuration use the ‘save ’ command to 
save the configuration. Then you can use ‘load ’ or add it to the 
command line with the -f  option.




> 
> On Thursday 11 January 2018 10:20 AM, Shailja Pandey wrote:
>> The command line for Pktgen is-
>> 
>> ./app/app/x86_64-native-linuxapp-gcc/pktgen -c 0xfff -n 4 -- -T -P -m 
>> "[2].0,[4].1,[8].2,[10].3"
>> 
>> Configuration of the machine is-
>> 
>> Architecture:  x86_64
>> CPU op-mode(s):32-bit, 64-bit
>> Byte Order:Little Endian
>> CPU(s):32
>> On-line CPU(s) list:   0-31
>> Thread(s) per core:2
>> Core(s) per socket:8
>> Socket(s): 2
>> NUMA node(s):  2
>> Vendor ID: GenuineIntel
>> CPU family:6
>> Model: 63
>> Model name:Intel(R) Xeon(R) CPU E5-2640 v3 @ 2.60GHz
>> Stepping:  2
>> CPU MHz:   1204.226
>> CPU max MHz:   3400.
>> CPU min MHz:   1200.
>> BogoMIPS:  5195.08
>> Virtualization:VT-x
>> L1d cache: 32K
>> L1i cache: 32K
>> L2 cache:  256K
>> L3 cache:  20480K
>> NUMA node0 CPU(s): 0,2,4,6,8,10,12,14,16,18,20,22,24,26,28,30
>> NUMA node1 CPU(s): 1,3,5,7,9,11,13,15,17,19,21,23,25,27,29,31
>> 
>> Linux XeonE5 4.4.0-93-generic
>> 
>> Thanks!
>> 
>> 
>> 
>> 
>> 
>> >
>>  On Jan 10, 2018, at 7:36 AM, Shailja Pandey > > wrote:
>> 
>> >
>>  
>> 
>> >
>>  Hi,
>> 
>> >
>>  
>> 
>> >
>>  We are performing experiments on Dell Poweredge R430 server, which is based 
>> on Haswell architecture based xeon-2640 v3 processor. We have attached XL 
>> 710 NIC(2x40 GbE) to the machine and expect 59 Mpps packet generation per 
>> port. But pktgen is able to generate only 21 Mpps per port and we are not 
>> sure about the problem.
>> 
>> >
>>  
>> 
>> >
>>  We are using DPDK version 16.07 and DPDK-pktgen version 3.1.0. Is there any 
>> way to generate packets from pktgen at 118 Mpps ?
>> 
>> 
>> What is the command line for Pktgen?
>> 
>> Using a single CPU you should be able to generate about 21Mpps, which tells 
>> me the configuration is not correct.
>> 
>> Pktgen needs N number of TX and RX cores to generate more traffic, plus the 
>> packets need to be spread across the cores using RSS. This means pktgen 
>> needs to be setup to generate packets with the 5 tuple difference to get all 
>> of the cores to receive the packets.
>> 
>> I need to know the command line and the configuration of the machine to help.
>> >
>>  
>> 
>> >
>>  -- 
>> 
>> >
>>  
>> 
>> >
>>  Thanks,
>> 
>> >
>>  Shailja
>> 
>> >
>>  
>> 
>> 
>> Regards,
>> Keith
>> 
>> -- 
>> 
>> Thanks,
>> Shailja
>> 
> 
> -- 
> 
> Thanks,
> Shailja
> 

Regards,
Keith

Re: [dpdk-dev] [PATCH v3 09/12] app/eventdev: add pipeline queue worker functions

[Pavan Nikhilesh]

On Thu, Jan 11, 2018 at 12:17:38PM +, Van Haaren, Harry wrote:
 > >
 
> > > Thinking a little more about this, also in light of patch 11/12 of this
> > series.
> > >
> > > The code here has a "safe" and "unsafe" version of TX. This involves
> > adding a spinlock inside the code, which is being locked/unlocked before
> > doing the actual TX action.
> > >
> > > I don't understand why this is necessary? DPDK's general stance on locking
> > for data-path is DPDK functions do not provide locks, and that application
> > level must implement thread-synchronization if it is required.
> > >
> > > In this case, the app/eventdev can be considered an App, but I don't like
> > the idea of providing a sample application and code that duplicates core
> > functionality with safe/unsafe versions..
> > >
> >
> > Some PMD's (net/octeontx) have capability to do multi-thread safe Tx where
> > no
> > thread-synchronization is required. This is exposed via the offload flag
> > 'DEV_TX_OFFLOAD_MT_LOCKFREE'.
>
> Yes understood.
>
>
> > So, the _safe Tx functions are selected based on the above offload
> > capability
> > and when the capability is absent _unsafe Tx functions are selected i.e.
> > synchronized Tx via spin locks based on the Egress port id.
>
>
> This part changes the current behavior of the sample app.
>
> Currently there is a (SINGLE_LINK | ATOMIC) stage at the end of the pipeline, 
> which performs this "many-to-one" action, allowing a single core to dequeue 
> all TX traffic, and perform the TX operation in a lock-free manner.
>
> Changing this to a locking mechanism is going to hurt performance on 
> platforms that do not support TX_OFFLOAD_MT_LOCKFREE.
>
> In my opinion, the correct fix is to alter the overall pipeline, and always 
> use lockless TX. Examples below;
>
> NO TX_OFFLOAD_MT_LOCKFREE:
>
>Eth RX adapter -> stage 1 -> stage 2...(N-1) -> stage N -> stage TX 
> (Atomic | SINGLE_LINK) -> eth TX

Agreed, when we detect that tx is not lockfree the workers would just forward
the events to  (Atomic | SINGLE_LINK) event queue which would be dequeued by a
service(mt_unsafe) and Tx them lockfree.

>
>
> WITH TX_OFFLOAD_MT_LOCKFREE:
>
>Eth RX adapter -> stage 1 -> stage 2...(N-1) -> stage N -> eth TX MT 
> Capable

The current lockfree pipeline would remain the same.
>
>
> By configuring the pipeline based on MT_OFFLOAD_LOCKFREE capability flag, and 
> adding the SINGLE_LINK at the end if required, we can support both models 
> without resorting to locked TX functions.
>
> I think this will lead to a cleaner and more performant solution.
>

Thoughts?

Pavan.

> 

Re: [dpdk-dev] [PATCH v4] lib/librte_meter: add meter configuration profile

[Dumitrescu, Cristian]

> -Original Message-
> From: Singh, Jasvinder
> Sent: Monday, January 8, 2018 3:44 PM
> To: dev@dpdk.org
> Cc: Dumitrescu, Cristian 
> Subject: [PATCH v4] lib/librte_meter: add meter configuration profile
> 
> From: Cristian Dumitrescu 
> 
> This patch adds support for meter configuration profiles.
> Benefits: simplified configuration procedure, improved performance.
> 

...

> 
> Signed-off-by: Cristian Dumitrescu 
> Signed-off-by: Jasvinder Singh 
> ---
> V4:
> - change commit message
> 

Applied to next-qos tree.

[dpdk-dev] [PATCH V10 1/2] eal: add uevent monitor api and callback func

[Jeff Guo]

This patch aim to add a general uevent mechanism in eal device layer,
to enable all linux kernel object uevent monitoring, user could use these
APIs to monitor and read out the device status info that sent from the
kernel side, then corresponding to handle it, such as when detect hotplug
uevent type, user could detach or attach the device, and more it benefit
to use to do smoothly fail safe work.

About uevent monitoring:
a: add one epolling to poll the netlink socket, to monitor the uevent of
   the device.
b: add enum of rte_eal_dev_event_type and struct of rte_eal_uevent.
c: add below APIs in rte eal device layer.
   rte_dev_callback_register
   rte_dev_callback_unregister
   _rte_dev_callback_process
   rte_dev_evt_mntr_start
   rte_dev_evt_mntr_stop

Signed-off-by: Jeff Guo 
---
V10->V9:
a.fix prefix issue.
b.use a common callback lists for all device and all type to replace
add callback parameter into device struct.
c.delete some unuse part.
---
 lib/librte_eal/bsdapp/eal/eal_dev.c |  38 ++
 lib/librte_eal/common/eal_common_dev.c  | 124 ++
 lib/librte_eal/common/include/rte_dev.h | 101 +++
 lib/librte_eal/linuxapp/eal/Makefile|   1 +
 lib/librte_eal/linuxapp/eal/eal_dev.c   | 223 
 5 files changed, 487 insertions(+)
 create mode 100644 lib/librte_eal/bsdapp/eal/eal_dev.c
 create mode 100644 lib/librte_eal/linuxapp/eal/eal_dev.c

diff --git a/lib/librte_eal/bsdapp/eal/eal_dev.c 
b/lib/librte_eal/bsdapp/eal/eal_dev.c
new file mode 100644
index 000..32c17e8
--- /dev/null
+++ b/lib/librte_eal/bsdapp/eal/eal_dev.c
@@ -0,0 +1,38 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 Intel Corporation
+ */
+
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+
+#include "eal_thread.h"
+
+int
+rte_dev_evt_mntr_start(void)
+{
+   RTE_LOG(ERR, EAL, "Not support event monitor for FreeBSD\n");
+   return -1;
+}
+
+int
+rte_dev_evt_mntr_stop(void)
+{
+   RTE_LOG(ERR, EAL, "Not support event monitor for FreeBSD\n");
+   return -1;
+}
diff --git a/lib/librte_eal/common/eal_common_dev.c 
b/lib/librte_eal/common/eal_common_dev.c
index dda8f58..b602535 100644
--- a/lib/librte_eal/common/eal_common_dev.c
+++ b/lib/librte_eal/common/eal_common_dev.c
@@ -42,9 +42,31 @@
 #include 
 #include 
 #include 
+#include 
+#include 
 
 #include "eal_private.h"
 
+/* spinlock for device callbacks */
+static rte_spinlock_t rte_dev_event_lock = RTE_SPINLOCK_INITIALIZER;
+
+/**
+ * The user application callback description.
+ *
+ * It contains callback address to be registered by user application,
+ * the pointer to the parameters for callback, and the event type.
+ */
+struct rte_dev_event_callback {
+   TAILQ_ENTRY(rte_dev_event_callback) next; /**< Callbacks list */
+   rte_dev_event_cb_fn cb_fn;/**< Callback address */
+   void *cb_arg;   /**< Callback parameter */
+   char *dev_name; /**< device name */
+   uint32_t active;/**< Callback is executing */
+};
+
+/* A genaral callback for all registerd devices */
+static struct rte_dev_event_cb_list dev_event_cbs;
+
 static int cmp_detached_dev_name(const struct rte_device *dev,
const void *_name)
 {
@@ -234,3 +256,105 @@ int rte_eal_hotplug_remove(const char *busname, const 
char *devname)
rte_eal_devargs_remove(busname, devname);
return ret;
 }
+
+int
+rte_dev_callback_register(char *dev_name,
+   rte_dev_event_cb_fn cb_fn, void *cb_arg)
+{
+   struct rte_dev_event_callback *event_cb = NULL;
+
+   rte_spinlock_lock(&rte_dev_event_lock);
+
+   if (TAILQ_EMPTY(&(dev_event_cbs)))
+   TAILQ_INIT(&(dev_event_cbs));
+
+   TAILQ_FOREACH(event_cb, &(dev_event_cbs), next) {
+   if (event_cb->cb_fn == cb_fn &&
+   event_cb->cb_arg == cb_arg &&
+   event_cb->dev_name == dev_name)
+   break;
+   }
+
+   /* create a new callback. */
+   if (event_cb == NULL) {
+   /* allocate a new user callback entity */
+   event_cb = malloc(sizeof(struct rte_dev_event_callback));
+   if (event_cb != NULL) {
+   event_cb->cb_fn = cb_fn;
+   event_cb->cb_arg = cb_arg;
+   event_cb->dev_name = dev_name;
+   }
+   TAILQ_INSERT_TAIL(&(dev_event_cbs), event_cb, next);
+   }
+
+   rte_spinlock_unlock(&rte_dev_event_lock);
+   return (event_cb == NULL) ? -1 : 0;
+}
+
+int
+rte_dev_callback_unregister(char *dev_name,
+   rte_dev_event_cb_fn cb_fn, void *cb_arg)
+{
+   int ret;
+   struct rte_dev_event_callback *event_cb, *next;
+
+   if (!cb_fn || dev_name == NULL)
+   re

[dpdk-dev] [PATCH V10 2/2] eal: add uevent pass and process function

[Jeff Guo]

In order to handle the uevent which have been detected from the kernel
side, add uevent process function, let hot plug event to be example to
show uevent mechanism how to pass the uevent and process the uevent.

About uevent passing and processing, add below functions in linux eal
dev layer. FreeBSD not support uevent ,so let it to be void and do not
implement in function.
a.dev_uev_parse
b.dev_uev_receive
c.dev_uev_process

Signed-off-by: Jeff Guo 
---
V10->V9:
delete some unuse part
---
 lib/librte_eal/common/include/rte_dev.h |  23 +++
 lib/librte_eal/linuxapp/eal/eal_dev.c   | 109 +++-
 2 files changed, 130 insertions(+), 2 deletions(-)

diff --git a/lib/librte_eal/common/include/rte_dev.h 
b/lib/librte_eal/common/include/rte_dev.h
index fea037a..a3166f7 100644
--- a/lib/librte_eal/common/include/rte_dev.h
+++ b/lib/librte_eal/common/include/rte_dev.h
@@ -51,6 +51,23 @@ extern "C" {
 
 #include 
 
+#define RTE_EAL_UEV_MSG_LEN 4096
+#define RTE_EAL_UEV_MSG_ELEM_LEN 128
+
+enum rte_dev_state {
+   RTE_DEV_UNDEFINED,  /**< unknown device state */
+   RTE_DEV_FAULT,  /**< device fault or error */
+   RTE_DEV_PARSED, /**< device have been parsed on bus*/
+   RTE_DEV_PROBED, /**< devcie have been probed driver  */
+};
+
+enum rte_dev_subsystem {
+   RTE_DEV_SUBSYSTEM_UIO,
+   RTE_DEV_SUBSYSTEM_VFIO,
+   RTE_DEV_SUBSYSTEM_PCI,
+   RTE_DEV_SUBSYSTEM_MAX
+};
+
 /**
  * The device event type.
  */
@@ -61,10 +78,16 @@ enum rte_dev_event_type {
RTE_DEV_EVENT_MAX   /**< max value of this enum */
 };
 
+enum event_monitor_netlink_group {
+   RTE_DEV_EVENT_MONITOR_KERNEL,
+   RTE_DEV_EVENT_MONITOR_UDEV,
+};
+
 struct rte_dev_event {
enum rte_dev_event_type type;   /**< device event type */
int subsystem;  /**< subsystem id */
char *devname;  /**< device name */
+   enum event_monitor_netlink_group group; /**< device netlink group */
 };
 
 typedef int (*rte_dev_event_cb_fn)(enum rte_dev_event_type event,
diff --git a/lib/librte_eal/linuxapp/eal/eal_dev.c 
b/lib/librte_eal/linuxapp/eal/eal_dev.c
index bc32aab..1d0ec33 100644
--- a/lib/librte_eal/linuxapp/eal/eal_dev.c
+++ b/lib/librte_eal/linuxapp/eal/eal_dev.c
@@ -79,10 +79,115 @@ dev_monitor_enable(int netlink_fd)
return -1;
 }
 
+static void
+dev_uev_parse(const char *buf, struct rte_dev_event *event)
+{
+   char action[RTE_EAL_UEV_MSG_ELEM_LEN];
+   char subsystem[RTE_EAL_UEV_MSG_ELEM_LEN];
+   char dev_path[RTE_EAL_UEV_MSG_ELEM_LEN];
+   char pci_slot_name[RTE_EAL_UEV_MSG_ELEM_LEN];
+   int i = 0;
+
+   memset(action, 0, RTE_EAL_UEV_MSG_ELEM_LEN);
+   memset(subsystem, 0, RTE_EAL_UEV_MSG_ELEM_LEN);
+   memset(dev_path, 0, RTE_EAL_UEV_MSG_ELEM_LEN);
+   memset(pci_slot_name, 0, RTE_EAL_UEV_MSG_ELEM_LEN);
+
+   while (i < RTE_EAL_UEV_MSG_LEN) {
+   for (; i < RTE_EAL_UEV_MSG_LEN; i++) {
+   if (*buf)
+   break;
+   buf++;
+   }
+   if (!strncmp(buf, "libudev", 7)) {
+   buf += 7;
+   i += 7;
+   event->group = RTE_DEV_EVENT_MONITOR_UDEV;
+   }
+   if (!strncmp(buf, "ACTION=", 7)) {
+   buf += 7;
+   i += 7;
+   snprintf(action, sizeof(action), "%s", buf);
+   } else if (!strncmp(buf, "DEVPATH=", 8)) {
+   buf += 8;
+   i += 8;
+   snprintf(dev_path, sizeof(dev_path), "%s", buf);
+   } else if (!strncmp(buf, "SUBSYSTEM=", 10)) {
+   buf += 10;
+   i += 10;
+   snprintf(subsystem, sizeof(subsystem), "%s", buf);
+   } else if (!strncmp(buf, "PCI_SLOT_NAME=", 14)) {
+   buf += 14;
+   i += 14;
+   snprintf(pci_slot_name, sizeof(subsystem), "%s", buf);
+   event->devname = pci_slot_name;
+   }
+   for (; i < RTE_EAL_UEV_MSG_LEN; i++) {
+   if (*buf == '\0')
+   break;
+   buf++;
+   }
+   }
+
+   if (!strncmp(subsystem, "pci", 3))
+   event->subsystem = RTE_DEV_SUBSYSTEM_UIO;
+   if (!strncmp(action, "add", 3))
+   event->type = RTE_DEV_EVENT_ADD;
+   if (!strncmp(action, "remove", 6))
+   event->type = RTE_DEV_EVENT_REMOVE;
+   event->devname = pci_slot_name;
+}
+
+static int
+dev_uev_receive(int fd, struct rte_dev_event *uevent)
+{
+   int ret;
+   char buf[RTE_EAL_UEV_MSG_LEN];
+
+   memset(uevent, 0, sizeof(struct rte_dev_event));
+   memset(buf, 0, RTE_EAL_UEV_MSG_LEN);
+
+   ret = recv(fd, buf, RTE_EAL_UE

Re: [dpdk-dev] [pull-request] next-qos 18.02 pre-rc1

[Thomas Monjalon]

11/01/2018 14:51, Cristian Dumitrescu:
> Cristian Dumitrescu (1):
>   lib/librte_meter: add meter configuration profile

Only one commit?

I am sure there is more pending in patchwork.

Re: [dpdk-dev] [PATCH v2 3/3] crypto/dpaa_sec: support ipsec protocol offload

[De Lara Guarch, Pablo]

Hi Akhil,

> -Original Message-
> From: Akhil Goyal [mailto:akhil.go...@nxp.com]
> Sent: Thursday, January 11, 2018 11:33 AM
> To: dev@dpdk.org
> Cc: De Lara Guarch, Pablo ;
> hemant.agra...@nxp.com; Akhil Goyal 
> Subject: [PATCH v2 3/3] crypto/dpaa_sec: support ipsec protocol offload
> 
> Signed-off-by: Akhil Goyal 
> Acked-by: Hemant Agrawal 

I think it is worth adding this change in release notes.

Thanks,
Pablo

Re: [dpdk-dev] [PATCH 1/4] drivers/mempool: add octeontx mempool driver to meson build

[Pavan Nikhilesh]

On Tue, Jan 09, 2018 at 05:41:23PM +, Bruce Richardson wrote:
> On Sun, Dec 31, 2017 at 05:32:00AM +0530, Pavan Nikhilesh wrote:
> > Add octeontx hardware mempool driver to meson build.
> >
> > Signed-off-by: Pavan Nikhilesh 
> > ---
>
> On the set as a whole, it looks good to me, just the few minor nits pointed
> out in other emails. Can you do a V2 with those fixed? Also, I notice
> that you are missing License tags on the new files created, those should
> probably be added too in V2.

At the time of sending the patchset we didnt have approval for SPDX licence
tags so, left them bare will update the tags in v2.

>
> Thanks,
> /Bruce

Thanks,
Pavan.

Re: [dpdk-dev] [PATCH 1/6] test: fix memory leak in bitmap test

[Dumitrescu, Cristian]

> -Original Message-
> From: Burakov, Anatoly
> Sent: Friday, December 22, 2017 10:12 AM
> To: dev@dpdk.org
> Cc: Dumitrescu, Cristian ;
> pbhagavat...@caviumnetworks.com
> Subject: [PATCH 1/6] test: fix memory leak in bitmap test
> 
> Fixes: c7e4a134e769 ("test: verify bitmap operations")
> Cc: pbhagavat...@caviumnetworks.com
> Signed-off-by: Anatoly Burakov 
> ---
>  test/test/test_bitmap.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/test/test/test_bitmap.c b/test/test/test_bitmap.c
> index 5c9eee9..7045d33 100644
> --- a/test/test/test_bitmap.c
> +++ b/test/test/test_bitmap.c
> @@ -186,6 +186,9 @@ test_bitmap(void)
>   if (test_bitmap_scan_operations(bmp) < 0)
>   return TEST_FAILED;
> 
> + rte_bitmap_free(bmp);
> + rte_free(mem);
> +
>   return TEST_SUCCESS;
>  }
> 
> --
> 2.7.4

Acked-by: Cristian Dumitrescu 

Re: [dpdk-dev] [PATCH 5/6] test: fix memory leak in table autotest

[Dumitrescu, Cristian]

> -Original Message-
> From: Burakov, Anatoly
> Sent: Friday, December 22, 2017 10:12 AM
> To: dev@dpdk.org
> Cc: Dumitrescu, Cristian ; sta...@dpdk.org
> Subject: [PATCH 5/6] test: fix memory leak in table autotest
> 
> Always deallocate allocated resources after the test is done.
> 
> Fixes: 5205954791cb ("app/test: packet framework unit tests")
> Cc: cristian.dumitre...@intel.com
> Cc: sta...@dpdk.org
> Signed-off-by: Anatoly Burakov 
> ---
>  test/test/test_table.c | 44 
>  1 file changed, 28 insertions(+), 16 deletions(-)
> 
> diff --git a/test/test/test_table.c b/test/test/test_table.c


Acked-by: Cristian Dumitrescu 

Re: [dpdk-dev] [PATCH] test: fix uninitialized parameter in table autotest

[Dumitrescu, Cristian]

> -Original Message-
> From: Burakov, Anatoly
> Sent: Thursday, December 21, 2017 3:53 PM
> To: dev@dpdk.org
> Cc: Dumitrescu, Cristian ; Kerlin, Marcin
> ; sta...@dpdk.org
> Subject: [PATCH] test: fix uninitialized parameter in table autotest
> 
> delete_bulk() copies metadata to pointers provided by the entries
> parameter, but in the unit test, they are uninitialized, leading
> to rte_table attempting to memcpy into random garbage pointers.
> 
> Memsetting pointer table to zero will prevent that from happening.
> 
> Fixes: 48f2543cf0a8 ("app/test: add bulk adding and deleting")
> Cc: marcinx.ker...@intel.com
> Cc: sta...@dpdk.org
> Signed-off-by: Anatoly Burakov 
> ---
>  test/test/test_table_acl.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/test/test/test_table_acl.c b/test/test/test_table_acl.c
> index 08c100f..6fcf4cc 100644
> --- a/test/test/test_table_acl.c
> +++ b/test/test/test_table_acl.c
> @@ -532,6 +532,8 @@ setup_acl_pipeline(void)
>   struct rte_pipeline_table_entry *table_entries[5];
>   int key_found[5];
> 
> + memset(table_entries, 0, sizeof(table_entries));
> +
>   for (n = 0; n < 5; n++) {
>   memset(&keys[n], 0, sizeof(struct
> rte_table_acl_rule_delete_params));
>   key_array[n] = &keys[n];
> --
> 2.7.4

Acked-by: Cristian Dumitrescu 

Re: [dpdk-dev] [PATCH V9 1/5] eal: add uevent monitor api and callback func

[Guo, Jia]

On 1/11/2018 9:43 AM, Thomas Monjalon wrote:

Hi,

Thanks for splitting the patches.
I will review the first one today. Please see below.

10/01/2018 10:12, Jeff Guo:

--- /dev/null
+++ b/lib/librte_eal/bsdapp/eal/eal_dev.c
+int
+rte_dev_monitor_start(void)
+{
+   return -1;
+}
+
+int
+rte_dev_monitor_stop(void)
+{
+   return -1;
+}

You should add a log to show it is not supported.

ok.

--- /dev/null
+++ b/lib/librte_eal/bsdapp/eal/include/exec-env/rte_dev.h
+#ifndef _RTE_DEV_H_
+#error "don't include this file directly, please include generic "
+#endif

Why creating different rte_dev.h for BSD and Linux?
This is an API, it should be the same.

if no need at this time, combine it to a file.

+/**
+ * Start the device uevent monitoring.
+ *
+ * @param none
+ * @return
+ *   - On success, zero.
+ *   - On failure, a negative value.
+ */
+int
+rte_dev_monitor_start(void);
+
+/**
+ * Stop the device uevent monitoring .
+ *
+ * @param none
+ * @return
+ *   - On success, zero.
+ *   - On failure, a negative value.
+ */
+
+int
+rte_dev_monitor_stop(void);
--- a/lib/librte_eal/common/eal_common_dev.c
+++ b/lib/librte_eal/common/eal_common_dev.c
@@ -42,9 +42,32 @@
  #include 
  #include 
  #include 
+#include 
+#include 
  
  #include "eal_private.h"
  
+/* spinlock for device callbacks */

+static rte_spinlock_t rte_dev_cb_lock = RTE_SPINLOCK_INITIALIZER;

Please rename to rte_dev_event_lock.
Let's use rte_dev_event_ prefix consistently.

make consistently, agree.

+ * The user application callback description.
+ *
+ * It contains callback address to be registered by user application,
+ * the pointer to the parameters for callback, and the event type.
+ */
+struct rte_eal_dev_callback {

Rename to rte_dev_event?

+   TAILQ_ENTRY(rte_eal_dev_callback) next; /**< Callbacks list */
+   rte_eal_dev_cb_fn cb_fn;/**< Callback address */

Rename to rte_dev_event_callback?

+   void *cb_arg;   /**< Parameter for callback */

Comment should be about opaque context.

+   void *ret_param;/**< Return parameter */
+   enum rte_dev_event_type event;  /**< device event type */
+   uint32_t active;/**< Callback is executing */

Why active is needed?

avoid the lock when unregistered  callback.

+};
+
+/* A genaral callback for all new devices be added onto the bus */
+static struct rte_eal_dev_callback *dev_add_cb;

It should not be a different callback for new devices.
You must allow registering the callback for all and new devices.
Please look how it's done for ethdev:
https://dpdk.org/patch/32900/
the aim to use this special callback is because when new device add onto 
the bus, no device instance to store the callback. i saw ethdev 
solution, that is base on port but that would not make sense in rte 
device layer. so
i try to abandon add callback in rte device, replace of add device name 
into callback , please see my v10 patch.

+int
+rte_dev_callback_register(struct rte_device *device,
+   enum rte_dev_event_type event,
+   rte_eal_dev_cb_fn cb_fn, void *cb_arg)
+{

Why passing an event type at registration?
I think the event processing dispatch must be done in the callback,
not at registration.
make sense, just register all type for device ,and let eal to pass the 
event.

+   /* allocate a new interrupt callback entity */
+   user_cb = rte_zmalloc("eal device event",
+   sizeof(*user_cb), 0);

No need to use rte_malloc here.
Please check this callback API patch:
https://dpdk.org/patch/33144/
could be better to concentration the code. but if you could tell me why 
not use rte_zmalloc.

--- a/lib/librte_eal/common/include/rte_dev.h
+++ b/lib/librte_eal/common/include/rte_dev.h
+enum uev_monitor_netlink_group {
+   UEV_MONITOR_KERNEL,
+   UEV_MONITOR_UDEV,
+};

Please keep a namespace prefix like RTE_DEV_EVENT_ (same for enum name).
Some comments are missing for these constants.

+/**
+ * The device event type.
+ */
+enum rte_dev_event_type {
+   RTE_DEV_EVENT_UNKNOWN,  /**< unknown event type */
+   RTE_DEV_EVENT_ADD,  /**< device being added */
+   RTE_DEV_EVENT_REMOVE,
+   /**< device being removed */
+   RTE_DEV_EVENT_CHANGE,
+   /**< device status being changed,
+* etc charger percent
+*/

What means status changed?
What means charger percent?
status changed means that object path change or other more,  charger 
percent just a example for some kobject status.  so i don't think we 
should explicit identify all , i will  delete it until we want to use it.

+   RTE_DEV_EVENT_MOVE, /**< device sysfs path being moved */

sysfs is Linux specific


+   RTE_DEV_EVENT_ONLINE,   /**< device being enable */

You mean a device can be

Re: [dpdk-dev] [PATCH 0/3] AVF optimization

[Ferruh Yigit]

On 1/11/2018 6:52 AM, Wenzhuo Lu wrote:
> The optimization for AVF code.
> As the AVF patches is in the next-net repo but not merged
> to the master branch, proposed these patches that can be
> merged to the previouse ones.
> Sorry for the inconvenience.
> 
> Wenzhuo Lu (3):
>   net/avf/base: add spinlock functions
>   net/avf: remove spinlock functions
>   net/avf: remove useless code

Series squashed into relevant commit in next-net, thanks.

Please double check latest patches.

Re: [dpdk-dev] [PATCH 1/5] net/ixgbevf: unregister irq handler when other interrupts not allowed.

[Tonghao Zhang]

On Thu, Jan 11, 2018 at 2:55 PM, Xing, Beilei  wrote:
>
>
>> -Original Message-
>> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Tonghao Zhang
>> Sent: Friday, January 5, 2018 10:11 PM
>> To: dev@dpdk.org
>> Cc: Tonghao Zhang 
>> Subject: [dpdk-dev] [PATCH 1/5] net/ixgbevf: unregister irq handler when
>> other interrupts not allowed.
>>
>> When we bind the ixgbe VF (e.g 82599 card) to igb_uio and enable the rx-
>> interrupt, there will be more than one epoll_wait on intr_handle.fd.
>> One is in "eal-intr-thread" thread, and the others are in the thread which 
>> call
>> the "rte_epoll_wait". The problem is that sometiems "eal-intr-thread"
>> thread will process the rx interrupt, and then rte_epoll_wait can't get the
>> event any more, and the packets may be lost.
>>
>> We should unregister the status interrupt handler in "eal-intr-thread"
>> thread and the ixgbe pf is in the same case.
>>
>> Signed-off-by: Tonghao Zhang 
>
> The patch looks OK for me, the commit log can be reworded, for example, "we" 
> can be removed.
Yes

> Apart from that:
> Acked-by: Beilei Xing 
Thanks for your reviews. I will repost v2
>
>> ---
>>  drivers/net/ixgbe/ixgbe_ethdev.c | 15 +++
>>  1 file changed, 15 insertions(+)
>>
>> diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c
>> b/drivers/net/ixgbe/ixgbe_ethdev.c
>> index 43e0132..e67389f 100644
>> --- a/drivers/net/ixgbe/ixgbe_ethdev.c
>> +++ b/drivers/net/ixgbe/ixgbe_ethdev.c
>> @@ -5049,6 +5049,15 @@ static int
>> ixgbevf_dev_xstats_get_names(__rte_unused struct rte_eth_dev *dev,
>>   }
>>   ixgbevf_configure_msix(dev);
>>
>> + if (!rte_intr_allow_others(intr_handle)) {
>> + rte_intr_callback_unregister(intr_handle,
>> +  ixgbevf_dev_interrupt_handler,
>> +  dev);
>> + if (dev->data->dev_conf.intr_conf.lsc != 0)
>> + PMD_INIT_LOG(INFO, "lsc won't enable because of"
>> +  " no intr multiplex");
>> + }
>> +
>>   /* When a VF port is bound to VFIO-PCI, only miscellaneous interrupt
>>* is mapped to VFIO vector 0 in eth_ixgbevf_dev_init( ).
>>* If previous VFIO interrupt mapping setting in
>> eth_ixgbevf_dev_init( ) @@ -5091,6 +5100,12 @@ static int
>> ixgbevf_dev_xstats_get_names(__rte_unused struct rte_eth_dev *dev,
>>
>>   ixgbe_dev_clear_queues(dev);
>>
>> + if (!rte_intr_allow_others(intr_handle))
>> + /* resume to the default handler */
>> + rte_intr_callback_register(intr_handle,
>> +ixgbevf_dev_interrupt_handler,
>> +(void *)dev);
>> +
>>   /* Clean datapath event and queue/vec mapping */
>>   rte_intr_efd_disable(intr_handle);
>>   if (intr_handle->intr_vec != NULL) {
>> --
>> 1.8.3.1
>

Re: [dpdk-dev] [PATCH] examples/ipsec-secgw: add cryptodev mask option

[De Lara Guarch, Pablo]

> -Original Message-
> From: Akhil Goyal [mailto:akhil.go...@nxp.com]
> Sent: Thursday, December 14, 2017 6:52 AM
> To: dev@dpdk.org
> Cc: De Lara Guarch, Pablo ;
> hemant.agra...@nxp.com; Gonzalez Monroy, Sergio
> ; Nicolau, Radu
> ; Akhil Goyal 
> Subject: [PATCH] examples/ipsec-secgw: add cryptodev mask option
> 
> Previously, ipsec-secgw application did not give user the flexibility to 
> decide
> which crypto device(s) will be used.
> 
> In this patch, a new cryptodev_mask option is added to the application.
> Same as portmask, the cryptodev_mask avails the user to mask out the
> unwanted crypto devices in the system.
> 
> This patch is similar to the support added in l2fwd-crypto
> (d2797f51cc63: examples/l2fwd-crypto: add cryptodev mask option)
> 
> Signed-off-by: Akhil Goyal 

Applied to dpdk-next-crypto.
Thanks,

Pablo

Re: [dpdk-dev] [PATCH] vhost: support Explicit Congestion Notification

[Yuanhan Liu]

On Wed, Nov 22, 2017 at 11:19:42AM +0800, Jiayu Hu wrote:
> In virtio, Explicit Congestion Notification (ECN) includes two parts:
> guest ECN and host ECN. Guest ECN means the frontend can handle TSO
> packets which have ECN set, and host ECN means the backend can handle
> TSO packets which have ECN set.
> 
> The ECN features are rarely used. However, virtio-net enables them by
> default, and vhost-net support both. To make live migration from
> vhost-net to vhost-user possible, this patch announces to support
> guest and host ECN in vhost-user.
> 
> Signed-off-by: Jiayu Hu 

Applied to dpdk-next-virtio.

Thanks.

--yliu

Re: [dpdk-dev] [PATCH 2/5] net/ixgbevf: set the inter-interrupt interval for EITR.

[Tonghao Zhang]

On Thu, Jan 11, 2018 at 3:26 PM, Xing, Beilei  wrote:
>
>
>> -Original Message-
>> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Tonghao Zhang
>> Sent: Friday, January 5, 2018 10:11 PM
>> To: dev@dpdk.org
>> Cc: Tonghao Zhang 
>> Subject: [dpdk-dev] [PATCH 2/5] net/ixgbevf: set the inter-interrupt interval
>> for EITR.
>>
>> Set EITR interval as default. This patch can improve the performance when
>> we enable the rx-intrrupt to process the packets because we hope rx-
>> intrrupt reduce CPU.
>>
>> The 200us value of EITR makes the performance better with the low CPU.
>>
>> Signed-off-by: Tonghao Zhang 
>> ---
>>  drivers/net/ixgbe/ixgbe_ethdev.c | 3 +++
>>  1 file changed, 3 insertions(+)
>>
>> diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c
>> b/drivers/net/ixgbe/ixgbe_ethdev.c
>> index e67389f..b20cab9 100644
>> --- a/drivers/net/ixgbe/ixgbe_ethdev.c
>> +++ b/drivers/net/ixgbe/ixgbe_ethdev.c
>> @@ -5780,6 +5780,9 @@ static void ixgbevf_set_vfta_all(struct rte_eth_dev
>> *dev, bool on)
>>   if (vector_idx < base + intr_handle->nb_efd - 1)
>>   vector_idx++;
>>   }
>> +
>> + IXGBE_WRITE_REG(hw, IXGBE_VTEITR(IXGBE_MISC_VEC_ID),
>> + IXGBE_EITR_INTERVAL_US(200) |
>
> Is it possible to configure interval via DPDK configuration? Just like 
> CONFIG_RTE_LIBRTE_I40E_ITR_INTERVAL for i40e.
> Maybe the value is good for you but not suitable for other users.

Thanks  for you tips.
>
> Besides, seems the patches in the patchset have no dependency, needn't put 
> them into one patchset, and then the Acked patch can be applied quickly.
>
>> IXGBE_EITR_CNT_WDIS);
>>  }
>>
>>  /**
>> --
>> 1.8.3.1
>

[dpdk-dev] [PATCH] doc: ethdev ABI change deprecation notice

[Kirill Rybalchenko]

Signed-off-by: Kirill Rybalchenko 
---
 doc/guides/rel_notes/deprecation.rst | 6 ++
 1 file changed, 6 insertions(+)

diff --git a/doc/guides/rel_notes/deprecation.rst 
b/doc/guides/rel_notes/deprecation.rst
index 13e8543..f0589ea 100644
--- a/doc/guides/rel_notes/deprecation.rst
+++ b/doc/guides/rel_notes/deprecation.rst
@@ -45,6 +45,12 @@ Deprecation Notices
   Target release for removal of the legacy API will be defined once most
   PMDs have switched to rte_flow.
 
+* ethdev: anounce ABI change
+  The size of variables flow_types_mask in rte_eth_fdir_info structure,
+  sym_hash_enable_mask and valid_bit_mask in rte_eth_hash_global_conf structure
+  will be increased from 32 to 64 bits to fulfill hardware requirements.
+  This change will break existing ABI as size of the structures will increase.
+
 * i40e: The default flexible payload configuration which extracts the first 16
   bytes of the payload for RSS will be deprecated starting from 18.02. If
   required the previous behavior can be configured using existing flow
-- 
2.5.5

Re: [dpdk-dev] [PATCH v2 2/6] ethdev: add port ownership

[Matan Azrad]

Hi Konstantin

From: Ananyev, Konstantin, Thursday, January 11, 2018 2:40 PM
> Hi Matan,
> 
> >
> > Hi Konstantin
> >
> > From: Ananyev, Konstantin, Wednesday, January 10, 2018 3:36 PM
> > > Hi Matan,
> > >
> > > Few comments from me below.
> > > BTW, do you plan to add ownership mandatory check in control path
> > > functions that change port configuration?
> >
> > No.
> 
> So it still totally voluntary usage and application nneds to be changed to
> exploit it?
> Apart from RTE_FOR_EACH_DEV() change proposed by Gaetan?
> 

Also RTE_FOR_EACH_DEV() change proposed by Gaetan is not protected because 2 
DPDK entities can get the same port while using it.
As I wrote in the log\docs and as discussed a lot in the first version:
The new synchronization rules are:
1. The port allocation and port release synchronization will be
   managed by ethdev.
2. The port usage synchronization will be managed by the port owner.
3. The port ownership API synchronization(also with port creation) will be 
managed by ethdev.
5. DPDK entity which want to use a port must take ownership before.

Ethdev should not protect 2 and 4 according these rules.

> > > Konstantin
> > >
> > > > -Original Message-
> > > > From: Matan Azrad [mailto:ma...@mellanox.com]
> > > > Sent: Sunday, January 7, 2018 9:46 AM
> > > > To: Thomas Monjalon ; Gaetan Rivet
> > > > ; Wu, Jingjing 
> > > > Cc: dev@dpdk.org; Neil Horman ;
> Richardson,
> > > > Bruce ; Ananyev, Konstantin
> > > > 
> > > > Subject: [PATCH v2 2/6] ethdev: add port ownership
> > > >
> > > > The ownership of a port is implicit in DPDK.
> > > > Making it explicit is better from the next reasons:
> > > > 1. It will define well who is in charge of the port usage 
> > > > synchronization.
> > > > 2. A library could work on top of a port.
> > > > 3. A port can work on top of another port.
> > > >
> > > > Also in the fail-safe case, an issue has been met in testpmd.
> > > > We need to check that the application is not trying to use a port
> > > > which is already managed by fail-safe.
> > > >
> > > > A port owner is built from owner id(number) and owner name(string)
> > > > while the owner id must be unique to distinguish between two
> > > > identical entity instances and the owner name can be any name.
> > > > The name helps to logically recognize the owner by different DPDK
> > > > entities and allows easy debug.
> > > > Each DPDK entity can allocate an owner unique identifier and can
> > > > use it and its preferred name to owns valid ethdev ports.
> > > > Each DPDK entity can get any port owner status to decide if it can
> > > > manage the port or not.
> > > >
> > > > The mechanism is synchronized for both the primary process threads
> > > > and the secondary processes threads to allow secondary process
> > > > entity to be a port owner.
> > > >
> > > > Add a sinchronized ownership mechanism to DPDK Ethernet devices to
> > > > avoid multiple management of a device by different DPDK entities.
> > > >
> > > > The current ethdev internal port management is not affected by
> > > > this feature.
> > > >
> > > > Signed-off-by: Matan Azrad 
> > > > ---
> > > >  doc/guides/prog_guide/poll_mode_drv.rst |  14 ++-
> > > >  lib/librte_ether/rte_ethdev.c   | 206
> > > ++--
> > > >  lib/librte_ether/rte_ethdev.h   |  89 ++
> > > >  lib/librte_ether/rte_ethdev_version.map |  12 ++
> > > >  4 files changed, 311 insertions(+), 10 deletions(-)
> > >
> > >
> > > >
> > > >
> > > > diff --git a/lib/librte_ether/rte_ethdev.c
> > > > b/lib/librte_ether/rte_ethdev.c index 684e3e8..0e12452 100644
> > > > --- a/lib/librte_ether/rte_ethdev.c
> > > > +++ b/lib/librte_ether/rte_ethdev.c
> > > > @@ -70,7 +70,10 @@
> > > >
> > > >  static const char *MZ_RTE_ETH_DEV_DATA = "rte_eth_dev_data";
> > > > struct rte_eth_dev rte_eth_devices[RTE_MAX_ETHPORTS];
> > > > +/* ports data array stored in shared memory */
> > > >  static struct rte_eth_dev_data *rte_eth_dev_data;
> > > > +/* next owner identifier stored in shared memory */ static
> > > > +uint16_t *rte_eth_next_owner_id;
> > > >  static uint8_t eth_dev_last_created_port;
> > > >
> > > >  /* spinlock for eth device callbacks */ @@ -82,6 +85,9 @@
> > > >  /* spinlock for add/remove tx callbacks */  static rte_spinlock_t
> > > > rte_eth_tx_cb_lock = RTE_SPINLOCK_INITIALIZER;
> > > >
> > > > +/* spinlock for eth device ownership management stored in shared
> > > > +memory */ static rte_spinlock_t *rte_eth_dev_ownership_lock;
> > > > +
> > > >  /* store statistics names and its offset in stats structure  */
> > > > struct rte_eth_xstats_name_off {
> > > > char name[RTE_ETH_XSTATS_NAME_SIZE]; @@ -153,14 +159,18 @@
> > > enum {  }
> > > >
> > > >  static void
> > > > -rte_eth_dev_data_alloc(void)
> > > > +rte_eth_dev_share_data_alloc(void)
> > > >  {
> > > > const unsigned flags = 0;
> > > > const struct rte_memzone *mz;
> > > > +   const unsigned int data_size = RTE_MAX_ETHPORTS *
> >

Re: [dpdk-dev] [PATCH] net/i40e: fix VF reset stats crash

[Van Haaren, Harry]

> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of David C Harton
> Sent: Thursday, January 11, 2018 1:24 PM
> To: Wu, Jingjing ; Xing, Beilei
> 
> Cc: dev@dpdk.org; David C Harton ; Zhao1, Wei
> 
> Subject: [dpdk-dev] [PATCH] net/i40e: fix VF reset stats crash
> 
> Calling i40evf_dev_xstats_reset can sometimes crash.
> 
> Fixed issue by adding a pstats NULL check.

Perhaps add a note here, that the root cause of (pstats == NULL)
is a valid error return from i40evf_query_stats(), if the vf
command fails to execute.

Hence, we are not hiding symptoms of a different issue here,
this NULL check is required to validate the returned values.


> Fixes: da61cd0849766 ("i40evf: add extended stats")
>8210e9e0d805e ("net/i40e: fix clear xstats bug in VF")
> Cc: wei.zh...@intel.com
> 
> Signed-off-by: David C Harton 


Acked-by: Harry van Haaren 


> ---
>  drivers/net/i40e/i40e_ethdev_vf.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/i40e/i40e_ethdev_vf.c
> b/drivers/net/i40e/i40e_ethdev_vf.c
> index b96d77a..9c773fe 100644
> --- a/drivers/net/i40e/i40e_ethdev_vf.c
> +++ b/drivers/net/i40e/i40e_ethdev_vf.c
> @@ -923,7 +923,8 @@ i40evf_dev_xstats_reset(struct rte_eth_dev *dev)
>   i40evf_query_stats(dev, &pstats);
> 
>   /* set stats offset base on current values */
> - vf->vsi.eth_stats_offset = *pstats;
> + if (pstats != NULL)
> + vf->vsi.eth_stats_offset = *pstats;
>  }
> 
>  static int i40evf_dev_xstats_get_names(__rte_unused struct rte_eth_dev
> *dev,
> --
> 2.10.3.dirty

Re: [dpdk-dev] [PATCH] net/i40e: fix VF reset stats crash

[David Harton (dharton)]

> -Original Message-
> From: Van Haaren, Harry [mailto:harry.van.haa...@intel.com]
> 
> > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of David C Harton
> >
> > Calling i40evf_dev_xstats_reset can sometimes crash.
> >
> > Fixed issue by adding a pstats NULL check.
> 
> Perhaps add a note here, that the root cause of (pstats == NULL) is a
> valid error return from i40evf_query_stats(), if the vf command fails to
> execute.
> 
> Hence, we are not hiding symptoms of a different issue here, this NULL
> check is required to validate the returned values.
>
 
NULL is returned if the call fails.  
If desired I can capture the return code and base the check off the rc instead.
That's the pattern used in a couple other places.

Thanks,
Dave

Re: [dpdk-dev] [PATCH] net/i40e: fix VF reset stats crash

[Van Haaren, Harry]

> From: David Harton (dharton) [mailto:dhar...@cisco.com]
> Sent: Thursday, January 11, 2018 2:59 PM
> To: Van Haaren, Harry ; Wu, Jingjing
> ; Xing, Beilei 
> Cc: dev@dpdk.org; Zhao1, Wei 
> Subject: RE: [dpdk-dev] [PATCH] net/i40e: fix VF reset stats crash
> 
> 
> > -Original Message-
> > From: Van Haaren, Harry [mailto:harry.van.haa...@intel.com]
> >
> > > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of David C Harton
> > >
> > > Calling i40evf_dev_xstats_reset can sometimes crash.
> > >
> > > Fixed issue by adding a pstats NULL check.
> >
> > Perhaps add a note here, that the root cause of (pstats == NULL) is a
> > valid error return from i40evf_query_stats(), if the vf command fails to
> > execute.
> >
> > Hence, we are not hiding symptoms of a different issue here, this NULL
> > check is required to validate the returned values.
> >
> 
> NULL is returned if the call fails.
> If desired I can capture the return code and base the check off the rc
> instead.
> That's the pattern used in a couple other places.

If you prefer that implementation of the check, you can keep my Ack, I don't 
mind much.

I reviewed as in my experience a NULL pointer can often be detected earlier, 
but that wasn't the case here.

PS: Apologies I didn't reply to v2, was reading "up" and didn't check for 
newer..

Re: [dpdk-dev] [pull-request] next-qos 18.02 pre-rc1

[Dumitrescu, Cristian]

> -Original Message-
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Thursday, January 11, 2018 2:13 PM
> To: Dumitrescu, Cristian 
> Cc: dev@dpdk.org; Singh, Jasvinder 
> Subject: Re: [pull-request] next-qos 18.02 pre-rc1
> 
> 11/01/2018 14:51, Cristian Dumitrescu:
> > Cristian Dumitrescu (1):
> >   lib/librte_meter: add meter configuration profile
> 
> Only one commit?
> 
> I am sure there is more pending in patchwork.

Hi Thomas,

Nothing else ready to go in next-qos at this point.

I am working with Alan Dewar on his 3 patch sets, currently awaiting updated 
version. Likely  2 of them to be ready soon, so we can target inclusion into 
RC2.

Regards,
Cristian

[dpdk-dev] [PATCH v3] net/i40e: fix VF reset stats crash

[David C Harton]

Calling i40evf_dev_xstats_reset can sometimes crash.

Fixed issue by checking return code before using pstats.

Fixes: 8210e9e0d805e ("net/i40e: fix clear xstats bug in VF")
Cc: wei.zh...@intel.com

Signed-off-by: David C Harton 
---

v2:
* Fixed fixes line in commit message.

v3:
* Modified conditional to check ret code instead of pointer.


 drivers/net/i40e/i40e_ethdev_vf.c | 6 --
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/net/i40e/i40e_ethdev_vf.c 
b/drivers/net/i40e/i40e_ethdev_vf.c
index b96d77a..d5a9edb 100644
--- a/drivers/net/i40e/i40e_ethdev_vf.c
+++ b/drivers/net/i40e/i40e_ethdev_vf.c
@@ -916,14 +916,16 @@ i40evf_update_stats(struct i40e_vsi *vsi,
 static void
 i40evf_dev_xstats_reset(struct rte_eth_dev *dev)
 {
+   int ret;
struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private);
struct i40e_eth_stats *pstats = NULL;
 
/* read stat values to clear hardware registers */
-   i40evf_query_stats(dev, &pstats);
+   ret = i40evf_query_stats(dev, &pstats);
 
/* set stats offset base on current values */
-   vf->vsi.eth_stats_offset = *pstats;
+   if (ret == 0)
+   vf->vsi.eth_stats_offset = *pstats;
 }
 
 static int i40evf_dev_xstats_get_names(__rte_unused struct rte_eth_dev *dev,
-- 
2.10.3.dirty

Re: [dpdk-dev] [PATCH v6 22/23] doc: add event timer adapter section to programmer's guide

[Kovacevic, Marko]

> Signed-off-by: Erik Gabriel Carrillo 
> ---
>  doc/guides/prog_guide/event_timer_adapter.rst | 301
> ++
>  doc/guides/prog_guide/index.rst   |   1 +
>  2 files changed, 302 insertions(+)
>  create mode 100644 doc/guides/prog_guide/event_timer_adapter.rst

<...> 

Acked-by: Marko Kovacevic 

Re: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on VF

[Wu, Yanglong]

Hi daiwei:

After I double checked with this bug, the problem can be described as follow:
When we start PF using testpmd like this: testpmd -c 0x3 -n 4 
--socket-mem=1024,1024 --file-prefix=pf -w :05:00.0 -w :05:00.1 -- -i. 
The number of queue signed to each port would be 1 and the  rxmode.mq_mode of 
ixgbe eth_dev  would be set as  ETH_MQ_RX_NONE. And this config info is make 
sense, so until now there is no problem, right?
And according to the code:
case ETH_MQ_RX_NONE:
>   /* if nothing mq mode configure, use default scheme */
>   dev->data->dev_conf.rxmode.mq_mode = 
> ETH_MQ_RX_VMDQ_ONLY;
> - if (RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool > 1)
> - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 1;
>   break;
The nb_q_per_pool of (RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool  would be set as 1, 
that means for VF the number of max queue can be used would be only 1. 

And next we start VF using l3fwd:
l3fwd/build/l3fwd -c 0xc -n 4 -w :05:10.0 -w :05:10.1 -- -p 0x3 
--config '(0,0,18),(1,0,19)' --parse-ptype
The number of rx_q equal to 1 and there is no problem. But, in l3fwd, the 
number of tx_q would equal to the number of core used in l3fwd, so here number 
of tx_q equal to 2 but as described above, the max number of queue can be used 
in VF is just only 1. So there is problem and l3fwd will check failed.  


So there is two ways to fix this bug, the first is using this patch, don't 
assign RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool as 1 when mq_mode = 
ETH_MQ_RX_VMDQ_ONLY.
And the second is that the number of tx_q would not assigned as the number of 
cores used in l3fwd.

Which one do you think is make sense? 




Yanglong


-Original Message-
From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Wu, Yanglong
Sent: Thursday, January 11, 2018 6:30 PM
To: Dai, Wei ; dev@dpdk.org
Subject: Re: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on VF

Yes, this problem is caused by PF passing a wrong number of queue to VF. So 
basically, I removed  model check of "case ETH_MQ_RX_NONE". It seem right after 
I do that but I'am also not very sure the changing is really right. So thank to 
your comments, I will redebug this problem.

-Original Message-
From: Dai, Wei
Sent: Thursday, January 11, 2018 5:48 PM
To: Wu, Yanglong ; dev@dpdk.org; Yang, Zhiyong 

Cc: Wu, Yanglong 
Subject: RE: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on VF

According to 82599 datasheet, VF working mode is controlled by its host PF.
If VF is working in RSS or DCB mode, it can use more than 1 queues.
That also means its host PF should enable RSS or DCB.
The PF Rx mode can be ETH_MQ_RX_VMDQ_RSS or ETH_MQ_RX_VMDQ_DCB.
So I think that DPDK app with PF cause this issue, not ixgbe PMD itself.

By the way, you can reply your patch mail to provide more info about how You 
reproduce this issue to easy others' reviewing and testing your patch.

Thanks

> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Yanglong Wu
> Sent: Tuesday, January 9, 2018 2:32 PM
> To: dev@dpdk.org; Yang, Zhiyong 
> Cc: Wu, Yanglong 
> Subject: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on VF
> 
> VF can't run in multi queue mode, if nb_q_per_pool was set as 1.
> Nb_q_per_pool is passed through to max_rx_q and max_tx_q in VF.
> So if nb_q_per_pool is equal to 1, max_rx_q and max_tx_q can't be more 
> than 1 and VF multi queue mode will fail.
> 
> Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check)
> 
> Signed-off-by: Yanglong Wu 
> ---
> v1~v3:
> coding style issue fix
> ---
> v4:
> Rework according to comments
> ---
>  drivers/net/ixgbe/ixgbe_ethdev.c | 2 --
>  1 file changed, 2 deletions(-)
> 
> diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c
> b/drivers/net/ixgbe/ixgbe_ethdev.c
> index baaeee5d9..6af4f8b2b 100644
> --- a/drivers/net/ixgbe/ixgbe_ethdev.c
> +++ b/drivers/net/ixgbe/ixgbe_ethdev.c
> @@ -2240,8 +2240,6 @@ ixgbe_check_mq_mode(struct rte_eth_dev *dev)
>   case ETH_MQ_RX_NONE:
>   /* if nothing mq mode configure, use default scheme */
>   dev->data->dev_conf.rxmode.mq_mode = 
> ETH_MQ_RX_VMDQ_ONLY;
> - if (RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool > 1)
> - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 1;
>   break;
>   default: /* ETH_MQ_RX_DCB, ETH_MQ_RX_DCB_RSS or ETH_MQ_TX_DCB*/
>   /* SRIOV only works in VMDq enable mode */
> --
> 2.11.0

Re: [dpdk-dev] [PATCH 4/5] net/ixgbevf: add check for rte_intr_enable.

[Tonghao Zhang]

On Thu, Jan 11, 2018 at 3:06 PM, Xing, Beilei  wrote:
>
>
>> -Original Message-
>> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Tonghao Zhang
>> Sent: Friday, January 5, 2018 10:11 PM
>> To: dev@dpdk.org
>> Cc: Tonghao Zhang 
>> Subject: [dpdk-dev] [PATCH 4/5] net/ixgbevf: add check for rte_intr_enable.
>
> The patch is not only for ixgbevf, but also for ixgbe, right?
> so how about changing the title with net/ixgbe started?
>
OK

>> When we bind the ixgbevf to vfio and call the rte_eth_dev_rx_intr_enable
>> and rte_eth_dev_rx_intr_disable frequently, the interrupt setting
>> (msi_set_mask_bit) will take more CPU as show below. rte_intr_enable call
>> the ioctl to map the fd to interrupts frequently.
>>
>> perf top:
>> 5.45%  [kernel]   [k] msi_set_mask_bit
>>
>> It is unnecessary to call the rte_intr_enable in
>> ixgbe_dev_rx_queue_intr_enable. because the fds has been mapped to
>> interrupt and not unmapped in ixgbe_dev_rx_queue_intr_disable.
>>
>> This patch add checks for using VFIO.  With the patch, msi_set_mask_bit is
>> not listed in perl any more. Any suggestion will be welcome.
>>
>> Signed-off-by: Tonghao Zhang 
>> ---
>>  drivers/net/ixgbe/ixgbe_ethdev.c | 9 +++--
>>  1 file changed, 7 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c
>> b/drivers/net/ixgbe/ixgbe_ethdev.c
>> index e929235..79e4097 100644
>> --- a/drivers/net/ixgbe/ixgbe_ethdev.c
>> +++ b/drivers/net/ixgbe/ixgbe_ethdev.c
>> @@ -5610,7 +5610,9 @@ static void ixgbevf_set_vfta_all(struct rte_eth_dev
>> *dev, bool on)
>>   RTE_SET_USED(queue_id);
>>   IXGBE_WRITE_REG(hw, IXGBE_VTEIMS, intr->mask);
>>
>> - rte_intr_enable(intr_handle);
>> + if (intr_handle->type == RTE_INTR_HANDLE_UIO ||
>> + intr_handle->type == RTE_INTR_HANDLE_UIO_INTX)
>> + rte_intr_enable(intr_handle);
>
> For igb_uio, did you check if it's necessary to call rte_intr_enable every 
> time? Since rte interrupt is not disabled during 
> ixgbevf_dev_rx_queue_intr_disable.

good, remove the rte_intr_enable directly, because intr has been
disable/enable via writing the register in
ixgbevf_dev_rx_queue_intr_disable/enable.
>>
>>   return 0;
>>  }
>> @@ -5659,7 +5661,10 @@ static void ixgbevf_set_vfta_all(struct
>> rte_eth_dev *dev, bool on)
>>   mask &= (1 << (queue_id - 32));
>>   IXGBE_WRITE_REG(hw, IXGBE_EIMS_EX(1), mask);
>>   }
>> - rte_intr_enable(intr_handle);
>> +
>> + if (intr_handle->type == RTE_INTR_HANDLE_UIO ||
>> + intr_handle->type == RTE_INTR_HANDLE_UIO_INTX)
>> + rte_intr_enable(intr_handle);
>
> The same comment as above.
>
>>
>>   return 0;
>>  }
>> --
>> 1.8.3.1
>

[dpdk-dev] [PATCH 0/5] net/mrvl: fix recently found pmd issues

[Tomasz Duszynski]

Following commits provide fixes for a few recently
found NET MRVL PMD issues.

Natalie Samsonov (5):
  net/mrvl: fix multiple probe issue
  net/mrvl: fix hif objects allocation
  net/mrvl: fix oversize bpool handling
  net/mrvl: fix shadow queue tail and size calculations
  net/mrvl: keep shadow txqs inside pmd txq

 drivers/net/mrvl/mrvl_ethdev.c | 340 ++---
 drivers/net/mrvl/mrvl_ethdev.h |   3 -
 drivers/net/mrvl/mrvl_qos.c|   2 +-
 3 files changed, 217 insertions(+), 128 deletions(-)

--
2.7.4

[dpdk-dev] [PATCH 2/5] net/mrvl: fix hif objects allocation

[Tomasz Duszynski]

From: Natalie Samsonov 

1. Add checking for non-EAL threads.

2. Create hif objects on first use since sometimes on probe not all
   lcores are initialized and can be added later.
   In this case the hif objects for later cores were not created and
   this caused system crash.

Fixes: 0ddc9b8 ("net/mrvl: add net PMD skeleton")

Signed-off-by: Natalie Samsonov 
---
 drivers/net/mrvl/mrvl_ethdev.c | 181 -
 1 file changed, 107 insertions(+), 74 deletions(-)

diff --git a/drivers/net/mrvl/mrvl_ethdev.c b/drivers/net/mrvl/mrvl_ethdev.c
index 8b3865b..5c3b700 100644
--- a/drivers/net/mrvl/mrvl_ethdev.c
+++ b/drivers/net/mrvl/mrvl_ethdev.c
@@ -190,6 +190,59 @@ mrvl_reserve_bit(int *bitmap, int max)
return n;
 }

+static int
+mrvl_init_hif(int core_id)
+{
+   struct pp2_hif_params params;
+   char match[MRVL_MATCH_LEN];
+   int ret;
+
+   ret = mrvl_reserve_bit(&used_hifs, MRVL_MUSDK_HIFS_MAX);
+   if (ret < 0) {
+   RTE_LOG(ERR, PMD, "Failed to allocate hif %d\n", core_id);
+   return ret;
+   }
+
+   snprintf(match, sizeof(match), "hif-%d", ret);
+   memset(¶ms, 0, sizeof(params));
+   params.match = match;
+   params.out_size = MRVL_PP2_AGGR_TXQD_MAX;
+   ret = pp2_hif_init(¶ms, &hifs[core_id]);
+   if (ret) {
+   RTE_LOG(ERR, PMD, "Failed to initialize hif %d\n", core_id);
+   return ret;
+   }
+
+   return 0;
+}
+
+static inline struct pp2_hif*
+mrvl_get_hif(struct mrvl_priv *priv, int core_id)
+{
+   int ret;
+
+   if (likely(hifs[core_id] != NULL))
+   return hifs[core_id];
+
+   rte_spinlock_lock(&priv->lock);
+
+   ret = mrvl_init_hif(core_id);
+   if (ret < 0) {
+   RTE_LOG(ERR, PMD, "Failed to allocate hif %d\n", core_id);
+   goto out;
+   }
+
+   if (core_id < mrvl_lcore_first)
+   mrvl_lcore_first = core_id;
+
+   if (core_id > mrvl_lcore_last)
+   mrvl_lcore_last = core_id;
+out:
+   rte_spinlock_unlock(&priv->lock);
+
+   return hifs[core_id];
+}
+
 /**
  * Configure rss based on dpdk rss configuration.
  *
@@ -551,8 +604,15 @@ static void
 mrvl_flush_bpool(struct rte_eth_dev *dev)
 {
struct mrvl_priv *priv = dev->data->dev_private;
+   struct pp2_hif *hif;
uint32_t num;
int ret;
+   unsigned int core_id = rte_lcore_id();
+
+   if (core_id == LCORE_ID_ANY)
+   core_id = 0;
+
+   hif = mrvl_get_hif(priv, core_id);

ret = pp2_bpool_get_num_buffs(priv->bpool, &num);
if (ret) {
@@ -564,8 +624,7 @@ mrvl_flush_bpool(struct rte_eth_dev *dev)
struct pp2_buff_inf inf;
uint64_t addr;

-   ret = pp2_bpool_get_buff(hifs[rte_lcore_id()], priv->bpool,
-&inf);
+   ret = pp2_bpool_get_buff(hif, priv->bpool, &inf);
if (ret)
break;

@@ -1173,9 +1232,19 @@ mrvl_fill_bpool(struct mrvl_rxq *rxq, int num)
struct buff_release_entry entries[MRVL_PP2_TXD_MAX];
struct rte_mbuf *mbufs[MRVL_PP2_TXD_MAX];
int i, ret;
-   unsigned int core_id = rte_lcore_id();
-   struct pp2_hif *hif = hifs[core_id];
-   struct pp2_bpool *bpool = rxq->priv->bpool;
+   unsigned int core_id;
+   struct pp2_hif *hif;
+   struct pp2_bpool *bpool;
+
+   core_id = rte_lcore_id();
+   if (core_id == LCORE_ID_ANY)
+   core_id = 0;
+
+   hif = mrvl_get_hif(rxq->priv, core_id);
+   if (!hif)
+   return -1;
+
+   bpool = rxq->priv->bpool;

ret = rte_pktmbuf_alloc_bulk(rxq->mp, mbufs, num);
if (ret)
@@ -1311,8 +1380,15 @@ mrvl_rx_queue_release(void *rxq)
struct mrvl_rxq *q = rxq;
struct pp2_ppio_tc_params *tc_params;
int i, num, tc, inq;
+   struct pp2_hif *hif;
+   unsigned int core_id = rte_lcore_id();

-   if (!q)
+   if (core_id == LCORE_ID_ANY)
+   core_id = 0;
+
+   hif = mrvl_get_hif(q->priv, core_id);
+
+   if (!q || !hif)
return;

tc = q->priv->rxq_map[q->queue_id].tc;
@@ -1323,7 +1399,7 @@ mrvl_rx_queue_release(void *rxq)
struct pp2_buff_inf inf;
uint64_t addr;

-   pp2_bpool_get_buff(hifs[rte_lcore_id()], q->priv->bpool, &inf);
+   pp2_bpool_get_buff(hif, q->priv->bpool, &inf);
addr = cookie_addr_high | inf.cookie;
rte_pktmbuf_free((struct rte_mbuf *)addr);
}
@@ -1599,9 +1675,12 @@ mrvl_rx_pkt_burst(void *rxq, struct rte_mbuf **rx_pkts, 
uint16_t nb_pkts)
struct pp2_bpool *bpool;
int i, ret, rx_done = 0;
int num;
+   struct pp2_hif *hif;
unsigned int core_id = rte_lcore_id();

-   if (unlikely(!q->priv->ppio))
+   hif = mrvl_get_hif(q->priv, core_id);
+
+   if (unlike

[dpdk-dev] [PATCH 1/5] net/mrvl: fix multiple probe issue

[Tomasz Duszynski]

From: Natalie Samsonov 

MUSDK library initialization and cleanup should be done once.
This commit fixes that by doing necessary initialization once the first
port is probed and cleanup once the last port is removed.

Fixes: 0ddc9b8 ("net/mrvl: add net PMD skeleton")

Signed-off-by: Natalie Samsonov 
---
 drivers/net/mrvl/mrvl_ethdev.c | 78 +-
 drivers/net/mrvl/mrvl_ethdev.h |  3 --
 drivers/net/mrvl/mrvl_qos.c|  2 +-
 3 files changed, 56 insertions(+), 27 deletions(-)

diff --git a/drivers/net/mrvl/mrvl_ethdev.c b/drivers/net/mrvl/mrvl_ethdev.c
index 127ce44..8b3865b 100644
--- a/drivers/net/mrvl/mrvl_ethdev.c
+++ b/drivers/net/mrvl/mrvl_ethdev.c
@@ -111,6 +111,11 @@ struct pp2_bpool 
*mrvl_port_to_bpool_lookup[RTE_MAX_ETHPORTS];
 int mrvl_port_bpool_size[PP2_NUM_PKT_PROC][PP2_BPOOL_NUM_POOLS][RTE_MAX_LCORE];
 uint64_t cookie_addr_high = MRVL_COOKIE_ADDR_INVALID;

+struct mrvl_ifnames {
+   const char *names[PP2_NUM_ETH_PPIO * PP2_NUM_PKT_PROC];
+   int idx;
+};
+
 /*
  * To use buffer harvesting based on loopback port shadow queue structure
  * was introduced for buffers information bookkeeping.
@@ -156,10 +161,9 @@ struct mrvl_txq {
  */
 struct mrvl_shadow_txq shadow_txqs[RTE_MAX_ETHPORTS][RTE_MAX_LCORE];

-/** Number of ports configured. */
-int mrvl_ports_nb;
 static int mrvl_lcore_first;
 static int mrvl_lcore_last;
+static int mrvl_dev_num;

 static inline int
 mrvl_get_bpool_size(int pp2_id, int pool_id)
@@ -584,8 +588,10 @@ mrvl_dev_stop(struct rte_eth_dev *dev)
mrvl_dev_set_link_down(dev);
mrvl_flush_rx_queues(dev);
mrvl_flush_tx_shadow_queues(dev);
-   if (priv->qos_tbl)
+   if (priv->qos_tbl) {
pp2_cls_qos_tbl_deinit(priv->qos_tbl);
+   priv->qos_tbl = NULL;
+   }
pp2_ppio_deinit(priv->ppio);
priv->ppio = NULL;
 }
@@ -2070,6 +2076,7 @@ mrvl_eth_dev_create(struct rte_vdev_device *vdev, const 
char *name)

eth_dev->rx_pkt_burst = mrvl_rx_pkt_burst;
eth_dev->tx_pkt_burst = mrvl_tx_pkt_burst;
+   eth_dev->data->kdrv = RTE_KDRV_NONE;
eth_dev->data->dev_private = priv;
eth_dev->device = &vdev->device;
eth_dev->dev_ops = &mrvl_ops;
@@ -2103,6 +2110,7 @@ mrvl_eth_dev_destroy(const char *name)

priv = eth_dev->data->dev_private;
pp2_bpool_deinit(priv->bpool);
+   used_bpools[priv->pp_id] &= ~(1 << priv->bpool_bit);
rte_free(priv);
rte_free(eth_dev->data->mac_addrs);
rte_eth_dev_release_port(eth_dev);
@@ -2126,9 +2134,9 @@ static int
 mrvl_get_ifnames(const char *key __rte_unused, const char *value,
 void *extra_args)
 {
-   const char **ifnames = extra_args;
+   struct mrvl_ifnames *ifnames = extra_args;

-   ifnames[mrvl_ports_nb++] = value;
+   ifnames->names[ifnames->idx++] = value;

return 0;
 }
@@ -2177,6 +2185,8 @@ mrvl_deinit_hifs(void)
if (hifs[i])
pp2_hif_deinit(hifs[i]);
}
+   used_hifs = MRVL_MUSDK_HIFS_RESERVED;
+   memset(hifs, 0, sizeof(hifs));
 }

 static void mrvl_set_first_last_cores(int core_id)
@@ -2201,7 +2211,7 @@ static int
 rte_pmd_mrvl_probe(struct rte_vdev_device *vdev)
 {
struct rte_kvargs *kvlist;
-   const char *ifnames[PP2_NUM_ETH_PPIO * PP2_NUM_PKT_PROC];
+   struct mrvl_ifnames ifnames;
int ret = -EINVAL;
uint32_t i, ifnum, cfgnum, core_id;
const char *params;
@@ -2215,21 +2225,34 @@ rte_pmd_mrvl_probe(struct rte_vdev_device *vdev)
return -EINVAL;

ifnum = rte_kvargs_count(kvlist, MRVL_IFACE_NAME_ARG);
-   if (ifnum > RTE_DIM(ifnames))
+   if (ifnum > RTE_DIM(ifnames.names))
goto out_free_kvlist;

+   ifnames.idx = 0;
rte_kvargs_process(kvlist, MRVL_IFACE_NAME_ARG,
   mrvl_get_ifnames, &ifnames);

-   cfgnum = rte_kvargs_count(kvlist, MRVL_CFG_ARG);
-   if (cfgnum > 1) {
-   RTE_LOG(ERR, PMD, "Cannot handle more than one config file!\n");
-   goto out_free_kvlist;
-   } else if (cfgnum == 1) {
-   rte_kvargs_process(kvlist, MRVL_CFG_ARG,
-  mrvl_get_qoscfg, &mrvl_qos_cfg);
+
+   /*
+* The below system initialization should be done only once,
+* on the first provided configuration file
+*/
+   if (!mrvl_qos_cfg) {
+   cfgnum = rte_kvargs_count(kvlist, MRVL_CFG_ARG);
+   RTE_LOG(INFO, PMD, "Parsing config file!\n");
+   if (cfgnum > 1) {
+   RTE_LOG(ERR, PMD, "Cannot handle more than one config 
file!\n");
+   goto out_free_kvlist;
+   } else if (cfgnum == 1) {
+   rte_kvargs_process(kvlist, MRVL_CFG_ARG,
+  mrvl_get_qoscfg, &mrvl_qos_cfg);
+   }
}

+   if (mrvl_d