[dpdk-dev] [dpdk-stable] [PATCH v2] net/i40e: fix compile error

[Zhang, Qi Z]

Hi Ferruh:

> -----Original Message-----
> From: Yigit, Ferruh
> Sent: Monday, February 20, 2017 7:44 PM
> To: Zhang, Qi Z <qi.z.zhang at intel.com>; Wu, Jingjing <jingjing.wu at intel.com>;
> Zhang, Helin <helin.zhang at intel.com>
> Cc: dev at dpdk.org; stable at dpdk.org
> Subject: Re: [dpdk-stable] [PATCH v2] net/i40e: fix compile error
> 
> On 2/20/2017 12:08 AM, Qi Zhang wrote:
> > This patch fix the compile error when
> > RTE_LIBRTE_I40E_RX_ALLOW_BULK_ALLOC
> > is disabled.
> 
> Compile error:
>   CC i40e_rxtx_vec_sse.o
> .../drivers/net/i40e/i40e_rxtx_vec_sse.c:72:26:
> error: no member named 'fake_mbuf' in 'struct i40e_rx_queue'
>             rxep[i].mbuf = &rxq->fake_mbuf;
>                             ~~~  ^
> 
> >
> > Fixes: 9ed94e5bb04e ("i40e: add vector Rx")
> > Cc: stable at dpdk.org
> >
> > Signed-off-by: Qi Zhang <qi.z.zhang at intel.com>
> > ---
> >  drivers/net/i40e/i40e_rxtx.c | 2 +-
> >  drivers/net/i40e/i40e_rxtx.h | 2 +-
> >  2 files changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/net/i40e/i40e_rxtx.c
> > b/drivers/net/i40e/i40e_rxtx.c index 48429cc..40f4a45 100644
> > --- a/drivers/net/i40e/i40e_rxtx.c
> > +++ b/drivers/net/i40e/i40e_rxtx.c
> > @@ -2209,11 +2209,11 @@ i40e_reset_rx_queue(struct i40e_rx_queue
> *rxq)
> >  	for (i = 0; i < len * sizeof(union i40e_rx_desc); i++)
> >  		((volatile char *)rxq->rx_ring)[i] = 0;
> >
> > -#ifdef RTE_LIBRTE_I40E_RX_ALLOW_BULK_ALLOC
> >  	memset(&rxq->fake_mbuf, 0x0, sizeof(rxq->fake_mbuf));
> >  	for (i = 0; i < RTE_PMD_I40E_RX_MAX_BURST; ++i)
> >  		rxq->sw_ring[rxq->nb_rx_desc + i].mbuf = &rxq->fake_mbuf;
> >
> > +#ifdef RTE_LIBRTE_I40E_RX_ALLOW_BULK_ALLOC
> 
> Hi Qi,
> 
> Just to double check, below update in i40e_rxtx.h is enough to fix the build
> error.
> 
> Can you please confirm assigning fake_mbuf to sw_ring[] is required step,
> independent from RTE_LIBRTE_I40E_RX_ALLOW_BULK_ALLOC enabled or not?

I think it's necessary to assign fake_mbuf for couple additional entries in sw_ring 
Because there is no place to initialize these entries (i40e_rxq_rearm only cover entries round in nb_rx_desc) while _recv_raw_pkts_vec require these entries in some situation.(though only RTE_I40E_DESCS_PER_LOOP-1 entries are required, but we can reuse the code)

> And it can be good to document this in commit log.
OK, I will add it in v3
> 
> Thanks,
> ferruh
> 
> 
> >  	rxq->rx_nb_avail = 0;
> >  	rxq->rx_next_avail = 0;
> >  	rxq->rx_free_trigger = (uint16_t)(rxq->rx_free_thresh - 1); diff
> > --git a/drivers/net/i40e/i40e_rxtx.h b/drivers/net/i40e/i40e_rxtx.h
> > index 9df8a56..4f7f319 100644
> > --- a/drivers/net/i40e/i40e_rxtx.h
> > +++ b/drivers/net/i40e/i40e_rxtx.h
> > @@ -119,11 +119,11 @@ struct i40e_rx_queue {
> >  	uint16_t nb_rx_hold; /**< number of held free RX desc */
> >  	struct rte_mbuf *pkt_first_seg; /**< first segment of current packet */
> >  	struct rte_mbuf *pkt_last_seg; /**< last segment of current packet
> > */
> > +	struct rte_mbuf fake_mbuf; /**< dummy mbuf */
> >  #ifdef RTE_LIBRTE_I40E_RX_ALLOW_BULK_ALLOC
> >  	uint16_t rx_nb_avail; /**< number of staged packets ready */
> >  	uint16_t rx_next_avail; /**< index of next staged packets */
> >  	uint16_t rx_free_trigger; /**< triggers rx buffer allocation */
> > -	struct rte_mbuf fake_mbuf; /**< dummy mbuf */
> >  	struct rte_mbuf *rx_stage[RTE_PMD_I40E_RX_MAX_BURST * 2];  #endif
> >
> >