allocation,
and free mbuf when receive packets
and rx cached_cons must be rollbacked if fail to allocating mbuf,
found by Ciara Loftus
Signed-off-by: Li RongQing
Signed-off-by: Dongsheng Rong
---
V2: rollback rx cached_cons if mbuf failed to be allocated
V3: add comment when rollback rx
> > + if (rte_pktmbuf_alloc_bulk(umem->mb_pool, fq_bufs, nb_pkts)) {
> > + AF_XDP_LOG(DEBUG,
> > + "Failed to get enough buffers for fq.\n");
> > + /* rollback cached_cons which is added by
> > +* xsk_ring_prod__needs_wakeup
> > +*/
>
allocation,
and free mbuf when receive packets
and rx cached_cons must be rollbacked if fail to allocating mbuf,
found by Ciara Loftus
Signed-off-by: Li RongQing
Signed-off-by: Dongsheng Rong
---
V2: rollback rx cached_cons if mbuf failed to be allocated
V3: add comment when rollback rx
>
> Apologies for the delay, I missed your reply Li.
> With the data you've provided I think the patch is justified.
> I think the rollback requires some explanation in the code as it may not be
> immediately clear what is happening.
> I suggest a v3 with either a comment above the rollback, or a
> -Original Message-
> From: Loftus, Ciara [mailto:ciara.lof...@intel.com]
> Sent: Friday, October 02, 2020 12:24 AM
> To: Li,Rongqing
> Cc: dev@dpdk.org
> Subject: RE: [PATCH][v2] net/af_xdp: avoid to unnecessary allocation and free
> mbuf in rx path
>
>
Ping, should I resend this?
Thanks
-Li
Fixes: ae2a04864a9a "(net/bonding: reduce slave starvation on Rx poll)"
Thanks
-Li
From: Wei Hu (Xavier) [mailto:huwei...@chinasoftinc.com]
Sent: Wednesday, September 30, 2020 4:41 PM
To: Li,Rongqing
Cc: dev@dpdk.org; Wei Hu (Xavier)
Subject: Re: [dpdk-dev] [PATCH] net/bonding: fix
ping
thanks
-Li
allocation,
and free mbuf when receive packets
V2: rollback rx cached_cons if mbuf failed to be allocated
Signed-off-by: Li RongQing
Signed-off-by: Dongsheng Rong
---
drivers/net/af_xdp/rte_eth_af_xdp.c | 67 -
1 file changed, 29 insertions(+), 38 deletions
time RX, this is a kind of cache false share, low
performance.
so move active_slave from bond_dev_private to bond_rx_queue
make it as per queue variable
Signed-off-by: Li RongQing
Signed-off-by: Dongsheng Rong
---
drivers/net/bonding/eth_bond_private.h | 3 ++-
drivers/net/bonding
From: Dongsheng Rong
it should be bond_rx_queue in bond_ethdev_rx_burst_alb in
RX path, not bond_tx_queue
Fixes: 06fe78b98ccd ("bond: add mode 6")
Signed-off-by: Dongsheng Rong
Signed-off-by: Li RongQing
---
drivers/net/bonding/rte_eth_bond_pmd.c | 4 ++--
1 file changed, 2 insert
> -Original Message-
> From: Loftus, Ciara [mailto:ciara.lof...@intel.com]
> Sent: Friday, September 18, 2020 5:39 PM
> To: Li,Rongqing
> Cc: dev@dpdk.org
> Subject: RE: [dpdk-dev] [PATCH 2/2] af_xdp: avoid to unnecessary allocation
> and
> free mbuf
>
>
unable to reserve fill
queue because dpdk has not pakcets to receive, at last
deadlock will happen
so move reserve fill queue before xsk_ring_cons__peek
to fix it
Acked-by: Ciara Loftus
Signed-off-by: Li RongQing
Signed-off-by: Dongsheng Rong
Cc: sta...@dpdk.org
---
diff with v1:
cc sta
> -Original Message-
> From: Loftus, Ciara [mailto:ciara.lof...@intel.com]
> Sent: Friday, September 18, 2020 5:27 PM
> To: Li,Rongqing ; dev@dpdk.org
> Cc: sta...@dpdk.org
> Subject: RE: [dpdk-dev] [PATCH] af_xdp: avoid deadlock due to empty fill queue
>
> &g
> -Original Message-
> From: Loftus, Ciara [mailto:ciara.lof...@intel.com]
> Sent: Friday, September 18, 2020 5:39 PM
> To: Li,Rongqing
> Cc: dev@dpdk.org
> Subject: RE: [dpdk-dev] [PATCH 2/2] af_xdp: avoid to unnecessary allocation
> and
> free mbuf
>
>
> -Original Message-
> From: Loftus, Ciara [mailto:ciara.lof...@intel.com]
> Sent: Thursday, September 17, 2020 3:33 PM
> To: Li,Rongqing
> Cc: dev@dpdk.org
> Subject: RE: [dpdk-dev] [PATCH 1/2] af_xdp: not return a negative value in
> af_xdp_rx_zc
>
> >
unable to reserve fill
queue because dpdk has not pakcets to receive, at last
deadlock will happen
so move reserve fill queue before xsk_ring_cons__peek
to fix it
Signed-off-by: Li RongQing
---
drivers/net/af_xdp/rte_eth_af_xdp.c | 7 ---
1 file changed, 4 insertions(+), 3 deletions(-)
diff
Signed-off-by: Li RongQing
---
drivers/net/af_xdp/rte_eth_af_xdp.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/net/af_xdp/rte_eth_af_xdp.c
b/drivers/net/af_xdp/rte_eth_af_xdp.c
index 48824050e..de830825d 100644
--- a/drivers/net/af_xdp/rte_eth_af_xdp.c
af_xdp_rx_zc should always return the number of received packets,
and negative value will be as number of received packets, and
confuse the caller
Fixes: d8a210774e1d ("net/af_xdp: support unaligned umem chunks")
Cc: sta...@dpdk.org
Signed-off-by: Li RongQing
---
drivers/
optimize rx performance, by allocating mbuf based on result
of xsk_ring_cons__peek, to avoid to redundancy allocation,
and free mbuf when receive packets
Signed-off-by: Li RongQing
Signed-off-by: Dongsheng Rong
---
drivers/net/af_xdp/rte_eth_af_xdp.c | 64
af_xdp_rx_zc should always return the number of received packets,
and negative value will be as number of received packets, and
confuse the caller
Fixes: d8a210774e1d ("net/af_xdp: support unaligned umem chunks")
Cc: sta...@dpdk.org
Signed-off-by: Li RongQing
---
drivers/
21 matches
Mail list logo