Re: [dpdk-dev] [PATCH v2] net/tap: allow user MAC to be passed as args

["Varghese, Vipin" <vipin.varghese@intel.com>]

Hi Pascal,

Please find my observation inline

> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Varghese, Vipin
> Sent: Monday, February 5, 2018 3:59 PM
> To: Pascal Mazon <pascal.mazon@6wind.com>; dev@dpdk.org
> Cc: Yigit, Ferruh <ferruh.yigit@intel.com>; Jain, Deepak K
> <deepak.k.jain@intel.com>
> Subject: Re: [dpdk-dev] [PATCH v2] net/tap: allow user MAC to be passed as args
> 
> Hi Pascal,
> 
> Thanks for the update, I see your point instead of mixing up with int32_t and
> uint64_t; converting to ' struct ether_addr' makes sense.
> 
> I will check internally and thanks once again for the suggestion.
> 
> Thanks
> Vipin Varghese
> 
> > -----Original Message-----
> > From: Pascal Mazon [mailto:pascal.mazon@6wind.com]
> > Sent: Monday, February 5, 2018 3:55 PM
> > To: dev@dpdk.org; Varghese, Vipin <vipin.varghese@intel.com>
> > Cc: Yigit, Ferruh <ferruh.yigit@intel.com>; Jain, Deepak K
> > <deepak.k.jain@intel.com>; Pascal Mazon <pascal.mazon@6wind.com>
> > Subject: [PATCH v2] net/tap: allow user MAC to be passed as args
> >
> > From: Vipin Varghese <vipin.varghese@intel.com>
> >
> > Allow TAP PMD to pass user desired MAC address as argument.
> > The argument value is processed as string, where each 2 bytes are
> > converted to HEX MAC address after validation.
> >
> > Signed-off-by: Vipin Varghese <vipin.varghese@intel.com>
> > Signed-off-by: Pascal Mazon <pascal.mazon@6wind.com>
> > ---
> >
> > Hi Vipin,
> >
> > I suggest the following patch for your MAC address argument, if that suits you.
> > I removed ETH_TAP_MAC_STR_FXD and ETH_TAP_MAC_STR_USR that were
> not
> > really useful. I find it nicer to deal with MAC mostly inside set_mac_type().
> > What do you think? Can you test that it fits your needs?
> >
> > Best regards,
> > Pascal
> >
> >  doc/guides/nics/tap.rst       |  6 ++++
> >  drivers/net/tap/rte_eth_tap.c | 80
> > +++++++++++++++++++++++++++++-----------
> > ---
> >  2 files changed, 60 insertions(+), 26 deletions(-)
> >
> > diff --git a/doc/guides/nics/tap.rst b/doc/guides/nics/tap.rst index
> > dc6f834ca0a1..6b083c846b21 100644
> > --- a/doc/guides/nics/tap.rst
> > +++ b/doc/guides/nics/tap.rst

<Snipped>

> > +
> > +	while ((mac_byte != NULL) &&
> > +			strspn(mac_byte, ETH_TAP_CMP_MAC_FMT) &&

Added ' strspn(mac_byte + 1, ETH_TAP_CMP_MAC_FMT) &&'; because 2 character validation was not taking place

<Snipped>

> >
> >  /* Open a TAP interface device.
> > @@ -1594,7 +1622,7 @@ rte_pmd_tap_probe(struct rte_vdev_device *dev)
> >  	int speed;
> >  	char tap_name[RTE_ETH_NAME_MAX_LEN];
> >  	char remote_iface[RTE_ETH_NAME_MAX_LEN];
> > -	int fixed_mac_type = 0;
> > +	struct ether_addr user_mac = {0};

Changed to ' struct ether_addr user_mac = { .addr_bytes={0}};'; to prevent compilation failure.

> >
> >  	name = rte_vdev_device_name(dev);
> >  	params = rte_vdev_device_args(dev);
> > @@ -1640,7 +1668,7 @@ rte_pmd_tap_probe(struct rte_vdev_device *dev)
> >  				ret = rte_kvargs_process(kvlist,
> >  							 ETH_TAP_MAC_ARG,
> >  							 &set_mac_type,
> > -							 &fixed_mac_type);
> > +							 &user_mac);
> >  				if (ret == -1)
> >  					goto leave;
> >  			}
> > @@ -1651,7 +1679,7 @@ rte_pmd_tap_probe(struct rte_vdev_device *dev)
> >  	RTE_LOG(NOTICE, PMD, "Initializing pmd_tap for %s as %s\n",
> >  		name, tap_name);
> >
> > -	ret = eth_dev_tap_create(dev, tap_name, remote_iface,
> > fixed_mac_type);
> > +	ret = eth_dev_tap_create(dev, tap_name, remote_iface, &user_mac);
> >
> >  leave:
> >  	if (ret == -1) {
> > @@ -1716,5 +1744,5 @@ RTE_PMD_REGISTER_ALIAS(net_tap, eth_tap);
> > RTE_PMD_REGISTER_PARAM_STRING(net_tap,
> >  			      ETH_TAP_IFACE_ARG "=<string> "
> >  			      ETH_TAP_SPEED_ARG "=<int> "
> > -			      ETH_TAP_MAC_ARG "=" ETH_TAP_MAC_FIXED " "
> > +			      ETH_TAP_MAC_ARG "=" ETH_TAP_MAC_ARG_FMT
> >  			      ETH_TAP_REMOTE_ARG "=<string>");
> > --
> > 2.11.0

With these changes it is reviewed from my end.