[dpdk-dev] [PATCH v5] ethdev: add return code to rte_eth_stats_reset()

David Harton (dharton) dharton at cisco.com
Wed Sep 20 16:01:23 CEST 2017



> -----Original Message-----
> From: Ferruh Yigit [mailto:ferruh.yigit at intel.com]
> 
> On 9/1/2017 3:26 AM, David Harton wrote:
> > Some devices do not support reset of eth stats.  An application may
> > need to know not to clear shadow stats if the device cannot.
> >
> > rte_eth_stats_reset is updated to provide a return code to share
> > whether the device supports reset or not.
> >
> > Signed-off-by: David Harton <dharton at cisco.com>
> > ---
> >
> > v5:
> > * squashed doc patch
> > * moved rel_note change from ABI to API section
> >
> > v4:
> > * commented return values
> >
> > v3:
> > * overcame noob errors and figured out patch challenged
> >
> > v2:
> > * fixed soft tab issue inserted while moving changes
> >
> >  doc/guides/rel_notes/release_17_11.rst | 13 +++++++++++++
> >  lib/librte_ether/rte_ethdev.c          |  8 +++++---
> >  lib/librte_ether/rte_ethdev.h          |  6 +++++-
> >  3 files changed, 23 insertions(+), 4 deletions(-)
> >
> > diff --git a/doc/guides/rel_notes/release_17_11.rst
> > b/doc/guides/rel_notes/release_17_11.rst
> > index 22df4fd..6282667 100644
> > --- a/doc/guides/rel_notes/release_17_11.rst
> > +++ b/doc/guides/rel_notes/release_17_11.rst
> > @@ -110,6 +110,19 @@ API Changes
> >     Also, make sure to start the actual text at the margin.
> >     =========================================================
> >
> > +* **Modified the return type of rte_eth_stats_reset.**
> > +
> > +  Changed return type of ``rte_eth_stats_reset`` from ``void`` to
> > + ``int``  so the caller may know whether a device supports the
> > + operation or not  and if the operation was carried out.
> > +
> 
> > +* **Modified the vlan_offload_set_t function prototype in the ethdev
> > +library.**
> > +
> > +  Changed the function prototype of ``vlan_offload_set_t``.  The
> > + return value  has been changed from ``void`` to ``int`` so the
> > + caller to knows whether  the backing device supports the operation
> > + or if the operation was  successfully performed.
> > +
> 
> Is this addition to the document related to this patch?

Good catch.  No. :(

I must have mishandled the rebase I did to update this patch.  V6 coming.  
Would be great if you could re-ACK afterwards so this one can move.

Thanks,
Dave

> 
> >  * **Modified the vlan_offload_set_t function prototype in the ethdev
> > library.**
> >
> >    Changed the function prototype of ``vlan_offload_set_t``.  The
> > return value diff --git a/lib/librte_ether/rte_ethdev.c
> > b/lib/librte_ether/rte_ethdev.c index 05e52b8..f0f1775 100644
> > --- a/lib/librte_ether/rte_ethdev.c
> > +++ b/lib/librte_ether/rte_ethdev.c
> > @@ -1341,17 +1341,19 @@ struct rte_eth_dev *
> >  	return 0;
> >  }
> >
> > -void
> > +int
> >  rte_eth_stats_reset(uint8_t port_id)
> >  {
> >  	struct rte_eth_dev *dev;
> >
> > -	RTE_ETH_VALID_PORTID_OR_RET(port_id);
> > +	RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
> >  	dev = &rte_eth_devices[port_id];
> >
> > -	RTE_FUNC_PTR_OR_RET(*dev->dev_ops->stats_reset);
> > +	RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->stats_reset, -ENOTSUP);
> >  	(*dev->dev_ops->stats_reset)(dev);
> >  	dev->data->rx_mbuf_alloc_failed = 0;
> > +
> > +	return 0;
> >  }
> >
> >  static int
> > diff --git a/lib/librte_ether/rte_ethdev.h
> > b/lib/librte_ether/rte_ethdev.h index 7254fd0..9110725 100644
> > --- a/lib/librte_ether/rte_ethdev.h
> > +++ b/lib/librte_ether/rte_ethdev.h
> > @@ -2246,8 +2246,12 @@ int rte_eth_tx_queue_setup(uint8_t port_id,
> uint16_t tx_queue_id,
> >   *
> >   * @param port_id
> >   *   The port identifier of the Ethernet device.
> > + * @return
> > + *   - (0) if device notified to reset stats.
> > + *   - (-ENOTSUP) if hardware doesn't support.
> > + *   - (-ENODEV) if *port_id* invalid.
> >   */
> > -void rte_eth_stats_reset(uint8_t port_id);
> > +int rte_eth_stats_reset(uint8_t port_id);
> >
> >  /**
> >   * Retrieve names of extended statistics of an Ethernet device.
> >



More information about the dev mailing list