Hi,
> -----Original Message-----
> From: Yuanhan Liu [mailto:yuanhan....@linux.intel.com]
> Sent: Friday, December 30, 2016 2:40 PM
> To: Tan, Jianfeng
> Cc: dev@dpdk.org; step...@networkplumber.org
> Subject: Re: [PATCH v2 8/9] examples/l3fwd: add parse-ptype option
>
> On Thu, Dec 29, 2016 at 07:30:42AM +0000, Jianfeng Tan wrote:
> > To support those devices that do not provide packet type info when
> > receiving packets, add a new option, --parse-ptype, to analyze
> > packet type in the Rx callback.
>
> I think this would be needed for all PMD drivers don't have the PTYPE
> support. For these, --parse-ptype looks like a mandatory option in
> the l3fwd example. I didn't find such option given in your test guide
> though, which is weird. Mistake?
Oops, my fault, it should be used with this option. As I just cared about if
packets are received, instead of what types of packets are received, so I
missed that. I'll fix it.
>
> Besides that, is there a way to query whether a PMD supports PTYPE
> or not?
Yes, we have API rte_eth_dev_get_supported_ptypes() to do that. This patch is
to emulate the commit 71a7e2424e07 ("examples/l3fwd: fix using packet type
blindly").
As we talked offline, I'll leverage this API to query if device support needed
ptypes, if not, register callback to analysis ptypes. This avoids to use
another option. But for record, this also leads to un-consistent behavior with
l3fwd.
Thanks,
Jianfeng
> If not, we should add a software one by our own. This could
> be done without introducing yet another option.
>
> >
> > Signed-off-by: Jianfeng Tan <jianfeng....@intel.com>
> > ---
> > examples/l3fwd-power/main.c | 60
> ++++++++++++++++++++++++++++++++++++++++++++-
> > 1 file changed, 59 insertions(+), 1 deletion(-)
> >
> > diff --git a/examples/l3fwd-power/main.c b/examples/l3fwd-
> power/main.c
> > index b65d683..44843ec 100644
> > --- a/examples/l3fwd-power/main.c
> > +++ b/examples/l3fwd-power/main.c
> > @@ -164,6 +164,8 @@ static uint32_t enabled_port_mask = 0;
> > static int promiscuous_on = 0;
> > /* NUMA is enabled by default. */
> > static int numa_on = 1;
> > +static int parse_ptype; /**< Parse packet type using rx callback, and */
> > + /**< disabled by default */
> >
> > enum freq_scale_hint_t
> > {
> > @@ -607,6 +609,48 @@ get_ipv4_dst_port(struct ipv4_hdr *ipv4_hdr,
> uint8_t portid,
> > #endif
> >
> > static inline void
> > +parse_ptype_one(struct rte_mbuf *m)
> > +{
> > + struct ether_hdr *eth_hdr;
> > + uint32_t packet_type = RTE_PTYPE_UNKNOWN;
> > + uint16_t ether_type;
> > +
> > + eth_hdr = rte_pktmbuf_mtod(m, struct ether_hdr *);
> > + ether_type = eth_hdr->ether_type;
> > + if (ether_type == rte_cpu_to_be_16(ETHER_TYPE_IPv4))
> > + packet_type |= RTE_PTYPE_L3_IPV4_EXT_UNKNOWN;
> > + else if (ether_type == rte_cpu_to_be_16(ETHER_TYPE_IPv6))
> > + packet_type |= RTE_PTYPE_L3_IPV6_EXT_UNKNOWN;
>
> BTW, we have rte_net_get_ptype(). Looks like a better option?
>
> --yliu
