[dpdk-dev] [PATCH 0/3] Support for flow director behavior "other" on Intel FVL NIC

[dpdk-dev] [PATCH 0/3] Support for flow director behavior "other" on Intel FVL NIC

[Andrey Chilikin]

This patch set adds 1 new flow director behavior "other" on Intel X(L)710 NIC.
When this mode is selected flow director will direct packet to LAN while the queue is defined by other filters. This can be used to extract flexible payload to RX desriptor with the flow director filter while targeted queue will be defined by other filters, for example, by hash filter (RSS).

Andrey Chilikin (3):
  librte_ether: add RTE_ETH_FDIR_OTHER for flow director behavior
  i40e: add RTE_ETH_FDIR_OTHER processing for flow director behavior
  ixgbe: add check for supported flow director behavior

 drivers/net/i40e/i40e_fdir.c    |   12 ++++++++++--
 drivers/net/ixgbe/ixgbe_fdir.c  |    3 ++-
 lib/librte_ether/rte_eth_ctrl.h |    1 +
 3 files changed, 13 insertions(+), 3 deletions(-)

-- 
1.7.4.1

[dpdk-dev] [PATCH 1/3] librte_ether: add RTE_ETH_FDIR_OTHER for flow director behavior

[Andrey Chilikin]

Add new flow director behavior RTE_ETH_FDIR_OTHER to assign a queue by other filters

Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com>
---
 lib/librte_ether/rte_eth_ctrl.h |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/lib/librte_ether/rte_eth_ctrl.h b/lib/librte_ether/rte_eth_ctrl.h
index 26b7b33..3acf501 100644
--- a/lib/librte_ether/rte_eth_ctrl.h
+++ b/lib/librte_ether/rte_eth_ctrl.h
@@ -417,6 +417,7 @@ struct rte_eth_fdir_input {
 enum rte_eth_fdir_behavior {
 	RTE_ETH_FDIR_ACCEPT = 0,
 	RTE_ETH_FDIR_REJECT,
+	RTE_ETH_FDIR_OTHER,
 };
 
 /**
-- 
1.7.4.1

[dpdk-dev] [PATCH 2/3] i40e: add RTE_ETH_FDIR_OTHER processing for flow director behavior

[Andrey Chilikin]

Add support for RTE_ETH_FDIR_OTHER flow director behavior so output queue is assigned by other filters

Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com>
---
 drivers/net/i40e/i40e_fdir.c |   12 ++++++++++--
 1 files changed, 10 insertions(+), 2 deletions(-)

diff --git a/drivers/net/i40e/i40e_fdir.c b/drivers/net/i40e/i40e_fdir.c
index c9ce98f..e2de4c5 100644
--- a/drivers/net/i40e/i40e_fdir.c
+++ b/drivers/net/i40e/i40e_fdir.c
@@ -1103,8 +1103,16 @@ i40e_fdir_filter_programming(struct i40e_pf *pf,
 
 	if (fdir_action->behavior == RTE_ETH_FDIR_REJECT)
 		dest = I40E_FILTER_PROGRAM_DESC_DEST_DROP_PACKET;
-	else
-		dest = I40E_FILTER_PROGRAM_DESC_DEST_DIRECT_PACKET_QINDEX;
+	else if (fdir_action->behavior == RTE_ETH_FDIR_ACCEPT)
+ 		dest = I40E_FILTER_PROGRAM_DESC_DEST_DIRECT_PACKET_QINDEX;
+	else if (fdir_action->behavior == RTE_ETH_FDIR_OTHER)
+		dest = I40E_FILTER_PROGRAM_DESC_DEST_DIRECT_PACKET_OTHER;
+	else {
+		PMD_DRV_LOG(ERR, "Failed to program FDIR filter:"
+			    " unsupported fdir behavior.");
+		return -EINVAL;
+	}
+
 	fdirdp->dtype_cmd_cntindex |= rte_cpu_to_le_32((dest <<
 				I40E_TXD_FLTR_QW1_DEST_SHIFT) &
 				I40E_TXD_FLTR_QW1_DEST_MASK);
-- 
1.7.4.1

[dpdk-dev] [PATCH 3/3] ixgbe: add check for supported flow director behaviors

[Andrey Chilikin]

Handle only supported flow director behaviors

Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com>
---
 drivers/net/ixgbe/ixgbe_fdir.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/drivers/net/ixgbe/ixgbe_fdir.c b/drivers/net/ixgbe/ixgbe_fdir.c
index 5c8b833..cf0e8be 100644
--- a/drivers/net/ixgbe/ixgbe_fdir.c
+++ b/drivers/net/ixgbe/ixgbe_fdir.c
@@ -959,7 +959,8 @@ ixgbe_add_del_fdir_filter(struct rte_eth_dev *dev,
 				" signature mode.");
 			return -EINVAL;
 		}
-	} else if (fdir_filter->action.rx_queue < IXGBE_MAX_RX_QUEUE_NUM)
+	} else if (fdir_filter->action.behavior == RTE_ETH_FDIR_ACCEPT &&
+			fdir_filter->action.rx_queue < IXGBE_MAX_RX_QUEUE_NUM)
 		queue = (uint8_t)fdir_filter->action.rx_queue;
 	else
 		return -EINVAL;
-- 
1.7.4.1

Re: [dpdk-dev] [PATCH 1/3] librte_ether: add RTE_ETH_FDIR_OTHER for flow director behavior

[Bruce Richardson]

On Tue, Sep 29, 2015 at 03:26:38PM +0100, Andrey Chilikin wrote:
> Add new flow director behavior RTE_ETH_FDIR_OTHER to assign a queue by other filters
> 
> Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com>
> ---
>  lib/librte_ether/rte_eth_ctrl.h |    1 +
>  1 files changed, 1 insertions(+), 0 deletions(-)
> 
> diff --git a/lib/librte_ether/rte_eth_ctrl.h b/lib/librte_ether/rte_eth_ctrl.h
> index 26b7b33..3acf501 100644
> --- a/lib/librte_ether/rte_eth_ctrl.h
> +++ b/lib/librte_ether/rte_eth_ctrl.h
> @@ -417,6 +417,7 @@ struct rte_eth_fdir_input {
>  enum rte_eth_fdir_behavior {
>  	RTE_ETH_FDIR_ACCEPT = 0,
>  	RTE_ETH_FDIR_REJECT,
> +	RTE_ETH_FDIR_OTHER,
>  };
>  
>  /**
> -- 
> 1.7.4.1
> 
Is "OTHER" meant to be a "NOOP" or a generic hold-all for a set of other possible
behaviours. From the description, I would assume NOOP - in which case would
RTE_ETH_FDIR_NOOP not be a better name?

/Bruce

Re: [dpdk-dev] [PATCH 1/3] librte_ether: add RTE_ETH_FDIR_OTHER for flow director behavior

[Chilikin, Andrey]

> -----Original Message-----
> From: Richardson, Bruce
> Sent: Tuesday, September 29, 2015 3:38 PM
> To: Chilikin, Andrey
> Cc: dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH 1/3] librte_ether: add RTE_ETH_FDIR_OTHER
> for flow director behavior
> 
> On Tue, Sep 29, 2015 at 03:26:38PM +0100, Andrey Chilikin wrote:
> > Add new flow director behavior RTE_ETH_FDIR_OTHER to assign a queue by
> > other filters
> >
> > Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com>
> > ---
> >  lib/librte_ether/rte_eth_ctrl.h |    1 +
> >  1 files changed, 1 insertions(+), 0 deletions(-)
> >
> > diff --git a/lib/librte_ether/rte_eth_ctrl.h
> > b/lib/librte_ether/rte_eth_ctrl.h index 26b7b33..3acf501 100644
> > --- a/lib/librte_ether/rte_eth_ctrl.h
> > +++ b/lib/librte_ether/rte_eth_ctrl.h
> > @@ -417,6 +417,7 @@ struct rte_eth_fdir_input {  enum
> > rte_eth_fdir_behavior {
> >  	RTE_ETH_FDIR_ACCEPT = 0,
> >  	RTE_ETH_FDIR_REJECT,
> > +	RTE_ETH_FDIR_OTHER,
> >  };
> >
> >  /**
> > --
> > 1.7.4.1
> >
> Is "OTHER" meant to be a "NOOP" or a generic hold-all for a set of other
> possible behaviours. From the description, I would assume NOOP - in which
> case would RTE_ETH_FDIR_NOOP not be a better name?
> 
> /Bruce
"NOOP" sounds like no operation at all, but FD still performs matching and extracts flexible payload to the RX descriptor, it only skips queue assignment which is done by other filters with lower priority. Maybe RTE_ETH_FDIR_PASSTHRU?

Regards,
Andrey

Re: [dpdk-dev] [PATCH 1/3] librte_ether: add RTE_ETH_FDIR_OTHER for flow director behavior

[Bruce Richardson]

On Tue, Sep 29, 2015 at 03:48:32PM +0100, Chilikin, Andrey wrote:
> > -----Original Message-----
> > From: Richardson, Bruce
> > Sent: Tuesday, September 29, 2015 3:38 PM
> > To: Chilikin, Andrey
> > Cc: dev@dpdk.org
> > Subject: Re: [dpdk-dev] [PATCH 1/3] librte_ether: add RTE_ETH_FDIR_OTHER
> > for flow director behavior
> > 
> > On Tue, Sep 29, 2015 at 03:26:38PM +0100, Andrey Chilikin wrote:
> > > Add new flow director behavior RTE_ETH_FDIR_OTHER to assign a queue by
> > > other filters
> > >
> > > Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com>
> > > ---
> > >  lib/librte_ether/rte_eth_ctrl.h |    1 +
> > >  1 files changed, 1 insertions(+), 0 deletions(-)
> > >
> > > diff --git a/lib/librte_ether/rte_eth_ctrl.h
> > > b/lib/librte_ether/rte_eth_ctrl.h index 26b7b33..3acf501 100644
> > > --- a/lib/librte_ether/rte_eth_ctrl.h
> > > +++ b/lib/librte_ether/rte_eth_ctrl.h
> > > @@ -417,6 +417,7 @@ struct rte_eth_fdir_input {  enum
> > > rte_eth_fdir_behavior {
> > >  	RTE_ETH_FDIR_ACCEPT = 0,
> > >  	RTE_ETH_FDIR_REJECT,
> > > +	RTE_ETH_FDIR_OTHER,
> > >  };
> > >
> > >  /**
> > > --
> > > 1.7.4.1
> > >
> > Is "OTHER" meant to be a "NOOP" or a generic hold-all for a set of other
> > possible behaviours. From the description, I would assume NOOP - in which
> > case would RTE_ETH_FDIR_NOOP not be a better name?
> > 
> > /Bruce
> "NOOP" sounds like no operation at all, but FD still performs matching and 
> extracts flexible payload to the RX descriptor, it only skips queue assignment
> which is done by other filters with lower priority. Maybe RTE_ETH_FDIR_PASSTHRU?
> 
> Regards,
> Andrey

PASSTHRU seems good to me. It's clearer than "OTHER" anyway :-)

/Bruce

[dpdk-dev] [PATCH v2 0/3] Support for flow director behavior "passthru" on Intel FVL NIC

[Andrey Chilikin]

This patch set adds new flow director behavior "passthru" on Intel X(L)710 NIC.
When this mode is selected flow director will direct packet to LAN while the queue is defined by other filters. This can be used to extract flexible payload to RX desriptor with the flow director filter while targeted queue will be defined by other filters, for example, by hash filter (RSS).

v2: rename RTE_ETH_FDIR_OTHER to RTE_ETH_FDIR_PASSTHRU

Andrey Chilikin (3):
  librte_ether: add RTE_ETH_FDIR_PASSTHRU for flow director behavior
  i40e: add RTE_ETH_FDIR_PASSTHRU processing for flow director behavior
  ixgbe: add check for supported flow director behaviors

 drivers/net/i40e/i40e_fdir.c    |   12 ++++++++++--
 drivers/net/ixgbe/ixgbe_fdir.c  |    3 ++-
 lib/librte_ether/rte_eth_ctrl.h |    1 +
 3 files changed, 13 insertions(+), 3 deletions(-)

-- 
1.7.4.1

[dpdk-dev] [PATCH v2 1/3] librte_ether: add RTE_ETH_FDIR_PASSTHRU for flow director behavior

[Andrey Chilikin]

Add new flow director behavior RTE_ETH_FDIR_PASSTHRU to assign a queue by other filters

Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com>
---
 lib/librte_ether/rte_eth_ctrl.h |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/lib/librte_ether/rte_eth_ctrl.h b/lib/librte_ether/rte_eth_ctrl.h
index 26b7b33..a0a6aab 100644
--- a/lib/librte_ether/rte_eth_ctrl.h
+++ b/lib/librte_ether/rte_eth_ctrl.h
@@ -417,6 +417,7 @@ struct rte_eth_fdir_input {
 enum rte_eth_fdir_behavior {
 	RTE_ETH_FDIR_ACCEPT = 0,
 	RTE_ETH_FDIR_REJECT,
+	RTE_ETH_FDIR_PASSTHRU,
 };
 
 /**
-- 
1.7.4.1

[dpdk-dev] [PATCH v2 2/3] i40e: add RTE_ETH_FDIR_PASSTHRU processing for flow director behavior

[Andrey Chilikin]

Add support for RTE_ETH_FDIR_PASSTHRU flow director behavior so output queue is assigned by other filters

Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com>
---
 drivers/net/i40e/i40e_fdir.c |   12 ++++++++++--
 1 files changed, 10 insertions(+), 2 deletions(-)

diff --git a/drivers/net/i40e/i40e_fdir.c b/drivers/net/i40e/i40e_fdir.c
index c9ce98f..45a372c 100644
--- a/drivers/net/i40e/i40e_fdir.c
+++ b/drivers/net/i40e/i40e_fdir.c
@@ -1103,8 +1103,16 @@ i40e_fdir_filter_programming(struct i40e_pf *pf,
 
 	if (fdir_action->behavior == RTE_ETH_FDIR_REJECT)
 		dest = I40E_FILTER_PROGRAM_DESC_DEST_DROP_PACKET;
-	else
-		dest = I40E_FILTER_PROGRAM_DESC_DEST_DIRECT_PACKET_QINDEX;
+	else if (fdir_action->behavior == RTE_ETH_FDIR_ACCEPT)
+ 		dest = I40E_FILTER_PROGRAM_DESC_DEST_DIRECT_PACKET_QINDEX;
+	else if (fdir_action->behavior == RTE_ETH_FDIR_PASSTHRU)
+		dest = I40E_FILTER_PROGRAM_DESC_DEST_DIRECT_PACKET_OTHER;
+	else {
+		PMD_DRV_LOG(ERR, "Failed to program FDIR filter:"
+			    " unsupported fdir behavior.");
+		return -EINVAL;
+	}
+
 	fdirdp->dtype_cmd_cntindex |= rte_cpu_to_le_32((dest <<
 				I40E_TXD_FLTR_QW1_DEST_SHIFT) &
 				I40E_TXD_FLTR_QW1_DEST_MASK);
-- 
1.7.4.1

[dpdk-dev] [PATCH v2 3/3] ixgbe: add check for supported flow director behaviors

[Andrey Chilikin]

Handle only supported flow director behaviors

Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com>
---
 drivers/net/ixgbe/ixgbe_fdir.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/drivers/net/ixgbe/ixgbe_fdir.c b/drivers/net/ixgbe/ixgbe_fdir.c
index 5c8b833..cf0e8be 100644
--- a/drivers/net/ixgbe/ixgbe_fdir.c
+++ b/drivers/net/ixgbe/ixgbe_fdir.c
@@ -959,7 +959,8 @@ ixgbe_add_del_fdir_filter(struct rte_eth_dev *dev,
 				" signature mode.");
 			return -EINVAL;
 		}
-	} else if (fdir_filter->action.rx_queue < IXGBE_MAX_RX_QUEUE_NUM)
+	} else if (fdir_filter->action.behavior == RTE_ETH_FDIR_ACCEPT &&
+			fdir_filter->action.rx_queue < IXGBE_MAX_RX_QUEUE_NUM)
 		queue = (uint8_t)fdir_filter->action.rx_queue;
 	else
 		return -EINVAL;
-- 
1.7.4.1

Re: [dpdk-dev] [PATCH v2 0/3] Support for flow director behavior "passthru" on Intel FVL NIC

[Wu, Jingjing]

> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Andrey Chilikin
> Sent: Wednesday, September 30, 2015 9:43 PM
> To: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH v2 0/3] Support for flow director behavior
> "passthru" on Intel FVL NIC
> 
> This patch set adds new flow director behavior "passthru" on Intel X(L)710
> NIC.
> When this mode is selected flow director will direct packet to LAN while the
> queue is defined by other filters. This can be used to extract flexible payload
> to RX desriptor with the flow director filter while targeted queue will be
> defined by other filters, for example, by hash filter (RSS).
> 
> v2: rename RTE_ETH_FDIR_OTHER to RTE_ETH_FDIR_PASSTHRU
> 
> Andrey Chilikin (3):
>   librte_ether: add RTE_ETH_FDIR_PASSTHRU for flow director behavior
>   i40e: add RTE_ETH_FDIR_PASSTHRU processing for flow director behavior
>   ixgbe: add check for supported flow director behaviors
> 
>  drivers/net/i40e/i40e_fdir.c    |   12 ++++++++++--
>  drivers/net/ixgbe/ixgbe_fdir.c  |    3 ++-
>  lib/librte_ether/rte_eth_ctrl.h |    1 +
>  3 files changed, 13 insertions(+), 3 deletions(-)
> 
> --
Acked-by: Jingjing Wu <jingjing.wu@intel.com>

> 1.7.4.1

Re: [dpdk-dev] [PATCH v2 0/3] Support for flow director behavior "passthru" on Intel FVL NIC

[Thomas Monjalon]

> > This patch set adds new flow director behavior "passthru" on Intel X(L)710
> > NIC.
> > When this mode is selected flow director will direct packet to LAN while the
> > queue is defined by other filters. This can be used to extract flexible payload
> > to RX desriptor with the flow director filter while targeted queue will be
> > defined by other filters, for example, by hash filter (RSS).
> > 
> > v2: rename RTE_ETH_FDIR_OTHER to RTE_ETH_FDIR_PASSTHRU
> 
> Acked-by: Jingjing Wu <jingjing.wu@intel.com>

Applied, thanks