Re: [dpdk-dev] [PATCH v2] net/ixgbe: check jumbo frame enable parameter

["Zhang, Qi Z" <qi.z.zhang@intel.com>]

Hi Wei:

> -----Original Message-----
> From: Zhao1, Wei
> Sent: Thursday, November 1, 2018 1:40 AM
> To: dev@dpdk.org
> Cc: stable@dpdk.org; Zhang, Qi Z <qi.z.zhang@intel.com>;
> yamashita.hideyuki@po.ntt-tx.co.jp; stephen@networkplumber.org
> Subject: RE: [PATCH v2] net/ixgbe: check jumbo frame enable parameter
> 
> Add the background for this patch.
> https://mails.dpdk.org/archives/dev/2018-November/117771.html
> https://mails.dpdk.org/archives/dev/2018-November/117772.html
> 
> 
> 
> > -----Original Message-----
> > From: Zhao1, Wei
> > Sent: Thursday, November 1, 2018 4:29 PM
> > To: Zhao1, Wei <wei.zhao1@intel.com>; dev@dpdk.org
> > Cc: stable@dpdk.org; Zhang, Qi Z <qi.z.zhang@intel.com>;
> > yamashita.hideyuki@po.ntt-tx.co.jp; stephen@networkplumber.org
> > Subject: [PATCH v2] net/ixgbe: check jumbo frame enable parameter
> >
> > There is necessary to do some check of max packet size boundary for
> > code safe in order to avoid error in register setting when enable jumbo.
> >
> > Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
> >
> > ---
> >
> > v2:
> > -add more log infomation and fix build issue
> > ---
> >  drivers/net/ixgbe/ixgbe_ethdev.c |  4 +---
> > drivers/net/ixgbe/ixgbe_ethdev.h |  2 ++
> >  drivers/net/ixgbe/ixgbe_rxtx.c   | 31
> > +++++++++++++++++++++++++++++++
> >  3 files changed, 34 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c
> > b/drivers/net/ixgbe/ixgbe_ethdev.c
> > index 5a2c351..d65a911 100644
> > --- a/drivers/net/ixgbe/ixgbe_ethdev.c
> > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c
> > @@ -173,8 +173,6 @@ static int
> > ixgbe_dev_queue_stats_mapping_set(struct rte_eth_dev *eth_dev,
> >  					     uint8_t is_rx);
> >  static int ixgbe_fw_version_get(struct rte_eth_dev *dev, char
> *fw_version,
> >  				 size_t fw_size);
> > -static void ixgbe_dev_info_get(struct rte_eth_dev *dev,
> > -			       struct rte_eth_dev_info *dev_info);
> >  static const uint32_t *ixgbe_dev_supported_ptypes_get(struct
> > rte_eth_dev *dev);  static void ixgbevf_dev_info_get(struct
> > rte_eth_dev *dev,
> >  				 struct rte_eth_dev_info *dev_info); @@ -
> > 3692,7 +3690,7 @@ static int ixgbevf_dev_xstats_get_names(__rte_unused
> > struct rte_eth_dev *dev,
> >  		return 0;
> >  }
> >
> > -static void
> > +void
> >  ixgbe_dev_info_get(struct rte_eth_dev *dev, struct rte_eth_dev_info
> > *dev_info)  {
> >  	struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev); diff --
> > git a/drivers/net/ixgbe/ixgbe_ethdev.h
> > b/drivers/net/ixgbe/ixgbe_ethdev.h
> > index d0b9396..a474be4 100644
> > --- a/drivers/net/ixgbe/ixgbe_ethdev.h
> > +++ b/drivers/net/ixgbe/ixgbe_ethdev.h
> > @@ -727,6 +727,8 @@ int ixgbe_action_rss_same(const struct
> > rte_flow_action_rss *comp,
> >  			  const struct rte_flow_action_rss *with);  int
> > ixgbe_config_rss_filter(struct rte_eth_dev *dev,
> >  		struct ixgbe_rte_flow_rss_conf *conf, bool add);
> > +void ixgbe_dev_info_get(struct rte_eth_dev *dev,
> > +			       struct rte_eth_dev_info *dev_info);
> >
> >  static inline int
> >  ixgbe_ethertype_filter_lookup(struct ixgbe_filter_info *filter_info,
> > diff --git a/drivers/net/ixgbe/ixgbe_rxtx.c
> > b/drivers/net/ixgbe/ixgbe_rxtx.c index 2f0262a..94051ca 100644
> > --- a/drivers/net/ixgbe/ixgbe_rxtx.c
> > +++ b/drivers/net/ixgbe/ixgbe_rxtx.c
> > @@ -4814,6 +4814,34 @@ void __attribute__((cold))
> >  	return 0;
> >  }
> >
> > +static int __attribute__((cold))
> > +ixgbe_dev_jumboenable_check(struct rte_eth_dev *dev,
> > +				uint16_t max_rx_pkt_len)

it's better to return Boolean for this function

> > +{
> > +	struct rte_eth_dev_info dev_info;
> > +
> > +	ixgbe_dev_info_get(dev, &dev_info);
> > +
> > +	/* check that max packet size is within the allowed range */
> > +	if (max_rx_pkt_len < ETHER_MIN_MTU) {
> > +		PMD_INIT_LOG(ERR, "max packet size is too small.");
> > +		return -EINVAL;
> > +	}

Why this is only needed for DEV_RX_OFFLOAD_JUMBO_FRAME? 
And we will compare to ETHER_MAX_LEN which is a bigger value, this looks redundant. 

> > +
> > +	if (max_rx_pkt_len > dev_info.max_rx_pktlen) {
> > +		PMD_INIT_LOG(ERR, "max packet size is too big.");
> > +		return -EINVAL;
> > +	}

dev_info.max_rx_pktlen = 15872, it should be replaced by a macro, then we don't need to call ixgbe_dev_info_get here.
and why this is only needed for DEV_RX_OFFLOAD_JUMBO_FRAME?
> > +
> > +	/* check jumbo mode if needed */
> > +	if (max_rx_pkt_len < ETHER_MAX_LEN) {
> > +		PMD_INIT_LOG(ERR, "No need to enable jumbo.");
> > +		return -EINVAL;

And I agree with Stephen's comment,
Looks all of these is common for all drivers, it should be moved to ether layer.

> > +	}
> > +
> > +	return 0;
> > +}
> > +
> >  /*
> >   * Initializes Receive Unit.
> >   */
> > @@ -4865,6 +4893,9 @@ int __attribute__((cold))
> >  	 * Configure jumbo frame support, if any.
> >  	 */
> >  	if (rx_conf->offloads & DEV_RX_OFFLOAD_JUMBO_FRAME) {
> > +		if (ixgbe_dev_jumboenable_check(dev, rx_conf-
> > >max_rx_pkt_len))
> > +			return -EINVAL;
> > +
> >  		hlreg0 |= IXGBE_HLREG0_JUMBOEN;
> >  		maxfrs = IXGBE_READ_REG(hw, IXGBE_MAXFRS);
> >  		maxfrs &= 0x0000FFFF;
> > --
> > 1.8.3.1