* [dpdk-dev] [PATCH v2] net/ixgbe: fix l3fwd start failed on PF [not found] <20171102161031.167892-2-yanglong.wu@intel.com> @ 2017-11-20 1:53 ` Yanglong Wu 2017-11-20 2:29 ` [dpdk-dev] [PATCH v3] " Yanglong Wu 0 siblings, 1 reply; 13+ messages in thread From: Yanglong Wu @ 2017-11-20 1:53 UTC (permalink / raw) To: dev; +Cc: Yanglong Wu L3fwd start failed on PF, for tx_q check failed. That occured when the SRIOV is active and tx_q > rx_q. The tx_q is equal to nb_q_per_pool. The number of nb_q_per_pool should equeal to max number of queues supported by HW not nb_rx_q. Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to specific drivers) Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> --- drivers/net/ixgbe/ixgbe_ethdev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c index ae9c44421..0f0641da1 100644 --- a/drivers/net/ixgbe/ixgbe_ethdev.c +++ b/drivers/net/ixgbe/ixgbe_ethdev.c @@ -2180,7 +2180,7 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev *dev, uint16_t nb_rx_q) return -EINVAL; } - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 128 / RTE_ETH_DEV_SRIOV(dev).active; RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * nb_rx_q; return 0; -- 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
* [dpdk-dev] [PATCH v3] net/ixgbe: fix l3fwd start failed on PF 2017-11-20 1:53 ` [dpdk-dev] [PATCH v2] net/ixgbe: fix l3fwd start failed on PF Yanglong Wu @ 2017-11-20 2:29 ` Yanglong Wu 2017-11-20 2:40 ` [dpdk-dev] [PATCH v4] " Yanglong Wu 0 siblings, 1 reply; 13+ messages in thread From: Yanglong Wu @ 2017-11-20 2:29 UTC (permalink / raw) To: dev; +Cc: Yanglong Wu L3fwd start failed on PF, for tx_q check failed. That occurred when the SRIOV is active and tx_q > rx_q. The tx_q is equal to nb_q_per_pool. The number of nb_q_per_pool should equeal to max number of queues supported by HW not nb_rx_q. Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to specific drivers) Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> --- drivers/net/ixgbe/ixgbe_ethdev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c index ae9c44421..0f0641da1 100644 --- a/drivers/net/ixgbe/ixgbe_ethdev.c +++ b/drivers/net/ixgbe/ixgbe_ethdev.c @@ -2180,7 +2180,7 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev *dev, uint16_t nb_rx_q) return -EINVAL; } - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 128/RTE_ETH_DEV_SRIOV(dev).active; RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * nb_rx_q; return 0; -- 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
* [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on PF 2017-11-20 2:29 ` [dpdk-dev] [PATCH v3] " Yanglong Wu @ 2017-11-20 2:40 ` Yanglong Wu 2017-11-20 10:43 ` Ananyev, Konstantin 2018-01-08 3:06 ` [dpdk-dev] [PATCH v5] net/ixgbe: fix l3fwd start failed on Yanglong Wu 0 siblings, 2 replies; 13+ messages in thread From: Yanglong Wu @ 2017-11-20 2:40 UTC (permalink / raw) To: dev; +Cc: Yanglong Wu L3fwd start failed on PF, for tx_q check failed. That occurred when the SRIOV is active and tx_q > rx_q. The tx_q is equal to nb_q_per_pool. The number of nb_q_per_pool should equeal to max number of queues supported by HW not nb_rx_q. Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to specific drivers) Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> --- drivers/net/ixgbe/ixgbe_ethdev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c index ae9c44421..0f0641da1 100644 --- a/drivers/net/ixgbe/ixgbe_ethdev.c +++ b/drivers/net/ixgbe/ixgbe_ethdev.c @@ -2180,7 +2180,7 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev *dev, uint16_t nb_rx_q) return -EINVAL; } - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 128 / RTE_ETH_DEV_SRIOV(dev).active; RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * nb_rx_q; return 0; -- 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on PF 2017-11-20 2:40 ` [dpdk-dev] [PATCH v4] " Yanglong Wu @ 2017-11-20 10:43 ` Ananyev, Konstantin 2017-11-21 2:17 ` Wu, Yanglong 2018-01-07 15:34 ` Zhang, Helin 2018-01-08 3:06 ` [dpdk-dev] [PATCH v5] net/ixgbe: fix l3fwd start failed on Yanglong Wu 1 sibling, 2 replies; 13+ messages in thread From: Ananyev, Konstantin @ 2017-11-20 10:43 UTC (permalink / raw) To: Wu, Yanglong, dev; +Cc: Wu, Yanglong > -----Original Message----- > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Yanglong Wu > Sent: Monday, November 20, 2017 2:40 AM > To: dev@dpdk.org > Cc: Wu, Yanglong <yanglong.wu@intel.com> > Subject: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on PF > > L3fwd start failed on PF, for tx_q check failed. > That occurred when the SRIOV is active and tx_q > rx_q. > The tx_q is equal to nb_q_per_pool. The number of nb_q_per_pool > should equeal to max number of queues supported by HW not nb_rx_q. > > Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to specific drivers) > > Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> > --- > drivers/net/ixgbe/ixgbe_ethdev.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c > index ae9c44421..0f0641da1 100644 > --- a/drivers/net/ixgbe/ixgbe_ethdev.c > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c > @@ -2180,7 +2180,7 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev *dev, uint16_t nb_rx_q) > return -EINVAL; > } > > - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; > + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 128 / RTE_ETH_DEV_SRIOV(dev).active; Please use appropriate macros instead of constants (s/128/IXGBE_MAX_RX_QUEUE_NUM). > RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * nb_rx_q; Shouldn't we also change that line to: RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool; ? Konstantin > > return 0; > -- > 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on PF 2017-11-20 10:43 ` Ananyev, Konstantin @ 2017-11-21 2:17 ` Wu, Yanglong 2018-01-07 15:34 ` Zhang, Helin 1 sibling, 0 replies; 13+ messages in thread From: Wu, Yanglong @ 2017-11-21 2:17 UTC (permalink / raw) To: Ananyev, Konstantin, dev Hi, Yes, I think so, we can change it. nb_rx_q just means the number of queue used for a port and RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool means max queue can be used. -----Original Message----- From: Ananyev, Konstantin Sent: Monday, November 20, 2017 6:44 PM To: Wu, Yanglong <yanglong.wu@intel.com>; dev@dpdk.org Cc: Wu, Yanglong <yanglong.wu@intel.com> Subject: RE: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on PF > -----Original Message----- > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Yanglong Wu > Sent: Monday, November 20, 2017 2:40 AM > To: dev@dpdk.org > Cc: Wu, Yanglong <yanglong.wu@intel.com> > Subject: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on PF > > L3fwd start failed on PF, for tx_q check failed. > That occurred when the SRIOV is active and tx_q > rx_q. > The tx_q is equal to nb_q_per_pool. The number of nb_q_per_pool should > equeal to max number of queues supported by HW not nb_rx_q. > > Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to > specific drivers) > > Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> > --- > drivers/net/ixgbe/ixgbe_ethdev.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c > b/drivers/net/ixgbe/ixgbe_ethdev.c > index ae9c44421..0f0641da1 100644 > --- a/drivers/net/ixgbe/ixgbe_ethdev.c > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c > @@ -2180,7 +2180,7 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev *dev, uint16_t nb_rx_q) > return -EINVAL; > } > > - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; > + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 128 / > +RTE_ETH_DEV_SRIOV(dev).active; Please use appropriate macros instead of constants (s/128/IXGBE_MAX_RX_QUEUE_NUM). > RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * nb_rx_q; Shouldn't we also change that line to: RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool; ? Konstantin > > return 0; > -- > 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on PF 2017-11-20 10:43 ` Ananyev, Konstantin 2017-11-21 2:17 ` Wu, Yanglong @ 2018-01-07 15:34 ` Zhang, Helin 1 sibling, 0 replies; 13+ messages in thread From: Zhang, Helin @ 2018-01-07 15:34 UTC (permalink / raw) To: Wu, Yanglong; +Cc: Wu, Yanglong, Ananyev, Konstantin, dev Hi Yanglong Please follow the comments, and get all of them well addressed. My comments is you need to use in-reply-to when you send any updated versions of a patch. Regards, Helin > -----Original Message----- > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Ananyev, > Konstantin > Sent: Monday, November 20, 2017 6:44 PM > To: Wu, Yanglong; dev@dpdk.org > Cc: Wu, Yanglong > Subject: Re: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on PF > > > > > -----Original Message----- > > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Yanglong Wu > > Sent: Monday, November 20, 2017 2:40 AM > > To: dev@dpdk.org > > Cc: Wu, Yanglong <yanglong.wu@intel.com> > > Subject: [dpdk-dev] [PATCH v4] net/ixgbe: fix l3fwd start failed on PF > > > > L3fwd start failed on PF, for tx_q check failed. > > That occurred when the SRIOV is active and tx_q > rx_q. > > The tx_q is equal to nb_q_per_pool. The number of nb_q_per_pool should > > equeal to max number of queues supported by HW not nb_rx_q. > > > > Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to > > specific drivers) > > > > Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> > > --- > > drivers/net/ixgbe/ixgbe_ethdev.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c > > b/drivers/net/ixgbe/ixgbe_ethdev.c > > index ae9c44421..0f0641da1 100644 > > --- a/drivers/net/ixgbe/ixgbe_ethdev.c > > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c > > @@ -2180,7 +2180,7 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev > *dev, uint16_t nb_rx_q) > > return -EINVAL; > > } > > > > - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; > > + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 128 / > > +RTE_ETH_DEV_SRIOV(dev).active; > > Please use appropriate macros instead of constants > (s/128/IXGBE_MAX_RX_QUEUE_NUM). > > > > RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * > nb_rx_q; > > > Shouldn't we also change that line to: > RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * > RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool; > ? > > Konstantin > > > > > return 0; > > -- > > 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
* [dpdk-dev] [PATCH v5] net/ixgbe: fix l3fwd start failed on 2017-11-20 2:40 ` [dpdk-dev] [PATCH v4] " Yanglong Wu 2017-11-20 10:43 ` Ananyev, Konstantin @ 2018-01-08 3:06 ` Yanglong Wu 2018-01-08 11:54 ` Ananyev, Konstantin 2018-01-10 1:51 ` [dpdk-dev] [PATCH v6] net/ixgbe: fix tx_q check failed on PF Yanglong Wu 1 sibling, 2 replies; 13+ messages in thread From: Yanglong Wu @ 2018-01-08 3:06 UTC (permalink / raw) To: dev; +Cc: konstantin.ananyev, Yanglong Wu L3fwd start failed on PF, for tx_q check failed. That occurred when the SRIOV is active and tx_q > rx_q. The tx_q is equal to nb_q_per_pool. The number of nb_q_per_pool should equeal to max number of queues supported by HW not nb_rx_q. Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to specific drivers) Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> --- v5: Rework according to comments --- drivers/net/ixgbe/ixgbe_ethdev.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c index ff19a564a..baaeee5d9 100644 --- a/drivers/net/ixgbe/ixgbe_ethdev.c +++ b/drivers/net/ixgbe/ixgbe_ethdev.c @@ -95,6 +95,9 @@ /* Timer value included in XOFF frames. */ #define IXGBE_FC_PAUSE 0x680 +/*Default value of Max Rx Queue*/ +#define IXGBE_MAX_RX_QUEUE_NUM 128 + #define IXGBE_LINK_DOWN_CHECK_TIMEOUT 4000 /* ms */ #define IXGBE_LINK_UP_CHECK_TIMEOUT 1000 /* ms */ #define IXGBE_VMDQ_NUM_UC_MAC 4096 /* Maximum nb. of UC MAC addr. */ @@ -2194,9 +2197,10 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev *dev, uint16_t nb_rx_q) return -EINVAL; } - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; - RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * nb_rx_q; - + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = + IXGBE_MAX_RX_QUEUE_NUM / RTE_ETH_DEV_SRIOV(dev).active; + RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = + pci_dev->max_vfs * RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool; return 0; } -- 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [dpdk-dev] [PATCH v5] net/ixgbe: fix l3fwd start failed on 2018-01-08 3:06 ` [dpdk-dev] [PATCH v5] net/ixgbe: fix l3fwd start failed on Yanglong Wu @ 2018-01-08 11:54 ` Ananyev, Konstantin 2018-01-09 2:29 ` Wu, Yanglong 2018-01-10 1:51 ` [dpdk-dev] [PATCH v6] net/ixgbe: fix tx_q check failed on PF Yanglong Wu 1 sibling, 1 reply; 13+ messages in thread From: Ananyev, Konstantin @ 2018-01-08 11:54 UTC (permalink / raw) To: Wu, Yanglong, dev > -----Original Message----- > From: Wu, Yanglong > Sent: Monday, January 8, 2018 3:06 AM > To: dev@dpdk.org > Cc: Ananyev, Konstantin <konstantin.ananyev@intel.com>; Wu, Yanglong <yanglong.wu@intel.com> > Subject: [PATCH v5] net/ixgbe: fix l3fwd start failed on > > L3fwd start failed on PF, for tx_q check failed. > That occurred when the SRIOV is active and tx_q > rx_q. > The tx_q is equal to nb_q_per_pool. The number of nb_q_per_pool > should equeal to max number of queues supported by HW not nb_rx_q. But then 2 (or more cores) could try to TX packets through the same TX queue? Why not just fil to start gracefully (call rte_exit() or so) if such situation occurred? Konstantin > > Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to > specific drivers) > > Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> > --- > v5: > Rework according to comments > --- > drivers/net/ixgbe/ixgbe_ethdev.c | 10 +++++++--- > 1 file changed, 7 insertions(+), 3 deletions(-) > > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c > index ff19a564a..baaeee5d9 100644 > --- a/drivers/net/ixgbe/ixgbe_ethdev.c > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c > @@ -95,6 +95,9 @@ > /* Timer value included in XOFF frames. */ > #define IXGBE_FC_PAUSE 0x680 > > +/*Default value of Max Rx Queue*/ > +#define IXGBE_MAX_RX_QUEUE_NUM 128 > + > #define IXGBE_LINK_DOWN_CHECK_TIMEOUT 4000 /* ms */ > #define IXGBE_LINK_UP_CHECK_TIMEOUT 1000 /* ms */ > #define IXGBE_VMDQ_NUM_UC_MAC 4096 /* Maximum nb. of UC MAC addr. */ > @@ -2194,9 +2197,10 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev *dev, uint16_t nb_rx_q) > return -EINVAL; > } > > - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; > - RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * nb_rx_q; > - > + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = > + IXGBE_MAX_RX_QUEUE_NUM / RTE_ETH_DEV_SRIOV(dev).active; > + RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = > + pci_dev->max_vfs * RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool; > return 0; > } > > -- > 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [dpdk-dev] [PATCH v5] net/ixgbe: fix l3fwd start failed on 2018-01-08 11:54 ` Ananyev, Konstantin @ 2018-01-09 2:29 ` Wu, Yanglong 2018-01-09 12:39 ` Ananyev, Konstantin 0 siblings, 1 reply; 13+ messages in thread From: Wu, Yanglong @ 2018-01-09 2:29 UTC (permalink / raw) To: Ananyev, Konstantin, Lu, Wenzhuo; +Cc: dev Hi, Konstantin Thanks for your comments! Do you means that tx_q is must less than rx_q when the SRIOV is active and if not, the application case will not be supported? Do you think my patch will cause to 2 (or more cores) could try to TX packets through the same TX queue? And as far as I know, the way of core using tx_q queue is depend on the application (e.g. in l3fwd tx_q equal to number of core) and multi core use same tx_q is not suggested for locker is needed in this situation. So why do you think my patch will lead to multi core using same queue? Yanglong Wu -----Original Message----- From: Ananyev, Konstantin Sent: Monday, January 8, 2018 7:55 PM To: Wu, Yanglong <yanglong.wu@intel.com>; dev@dpdk.org Subject: RE: [PATCH v5] net/ixgbe: fix l3fwd start failed on > -----Original Message----- > From: Wu, Yanglong > Sent: Monday, January 8, 2018 3:06 AM > To: dev@dpdk.org > Cc: Ananyev, Konstantin <konstantin.ananyev@intel.com>; Wu, Yanglong > <yanglong.wu@intel.com> > Subject: [PATCH v5] net/ixgbe: fix l3fwd start failed on > > L3fwd start failed on PF, for tx_q check failed. > That occurred when the SRIOV is active and tx_q > rx_q. > The tx_q is equal to nb_q_per_pool. The number of nb_q_per_pool should > equeal to max number of queues supported by HW not nb_rx_q. But then 2 (or more cores) could try to TX packets through the same TX queue? Why not just fil to start gracefully (call rte_exit() or so) if such situation occurred? Konstantin > > Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to > specific drivers) > > Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> > --- > v5: > Rework according to comments > --- > drivers/net/ixgbe/ixgbe_ethdev.c | 10 +++++++--- > 1 file changed, 7 insertions(+), 3 deletions(-) > > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c > b/drivers/net/ixgbe/ixgbe_ethdev.c > index ff19a564a..baaeee5d9 100644 > --- a/drivers/net/ixgbe/ixgbe_ethdev.c > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c > @@ -95,6 +95,9 @@ > /* Timer value included in XOFF frames. */ #define IXGBE_FC_PAUSE > 0x680 > > +/*Default value of Max Rx Queue*/ > +#define IXGBE_MAX_RX_QUEUE_NUM 128 > + > #define IXGBE_LINK_DOWN_CHECK_TIMEOUT 4000 /* ms */ > #define IXGBE_LINK_UP_CHECK_TIMEOUT 1000 /* ms */ > #define IXGBE_VMDQ_NUM_UC_MAC 4096 /* Maximum nb. of UC MAC addr. */ > @@ -2194,9 +2197,10 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev *dev, uint16_t nb_rx_q) > return -EINVAL; > } > > - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; > - RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * nb_rx_q; > - > + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = > + IXGBE_MAX_RX_QUEUE_NUM / RTE_ETH_DEV_SRIOV(dev).active; > + RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = > + pci_dev->max_vfs * RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool; > return 0; > } > > -- > 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [dpdk-dev] [PATCH v5] net/ixgbe: fix l3fwd start failed on 2018-01-09 2:29 ` Wu, Yanglong @ 2018-01-09 12:39 ` Ananyev, Konstantin 0 siblings, 0 replies; 13+ messages in thread From: Ananyev, Konstantin @ 2018-01-09 12:39 UTC (permalink / raw) To: Wu, Yanglong, Lu, Wenzhuo; +Cc: dev Hi Yanglong, Sorry, I was just confused by the description of the patch. The code itself seems ok... probably need to rephrase the description (remove mentioning l3fwd?). Konstantin > Hi, Konstantin > > Thanks for your comments! > Do you means that tx_q is must less than rx_q when the SRIOV is active and if not, the application case will not be supported? > Do you think my patch will cause to 2 (or more cores) could try to TX packets through the same TX queue? And as far as I know, the way of > core using tx_q queue is depend on the application (e.g. in l3fwd tx_q equal to number of core) and multi core use same tx_q is not > suggested for locker is needed in this situation. So why do you think my patch will lead to multi core using same queue? > > Yanglong Wu > > > -----Original Message----- > From: Ananyev, Konstantin > Sent: Monday, January 8, 2018 7:55 PM > To: Wu, Yanglong <yanglong.wu@intel.com>; dev@dpdk.org > Subject: RE: [PATCH v5] net/ixgbe: fix l3fwd start failed on > > > > > -----Original Message----- > > From: Wu, Yanglong > > Sent: Monday, January 8, 2018 3:06 AM > > To: dev@dpdk.org > > Cc: Ananyev, Konstantin <konstantin.ananyev@intel.com>; Wu, Yanglong > > <yanglong.wu@intel.com> > > Subject: [PATCH v5] net/ixgbe: fix l3fwd start failed on > > > > L3fwd start failed on PF, for tx_q check failed. > > That occurred when the SRIOV is active and tx_q > rx_q. > > The tx_q is equal to nb_q_per_pool. The number of nb_q_per_pool should > > equeal to max number of queues supported by HW not nb_rx_q. > > But then 2 (or more cores) could try to TX packets through the same TX queue? > Why not just fil to start gracefully (call rte_exit() or so) if such situation occurred? > Konstantin > > > > > Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to > > specific drivers) > > > > Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> > > --- > > v5: > > Rework according to comments > > --- > > drivers/net/ixgbe/ixgbe_ethdev.c | 10 +++++++--- > > 1 file changed, 7 insertions(+), 3 deletions(-) > > > > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c > > b/drivers/net/ixgbe/ixgbe_ethdev.c > > index ff19a564a..baaeee5d9 100644 > > --- a/drivers/net/ixgbe/ixgbe_ethdev.c > > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c > > @@ -95,6 +95,9 @@ > > /* Timer value included in XOFF frames. */ #define IXGBE_FC_PAUSE > > 0x680 > > > > +/*Default value of Max Rx Queue*/ > > +#define IXGBE_MAX_RX_QUEUE_NUM 128 > > + > > #define IXGBE_LINK_DOWN_CHECK_TIMEOUT 4000 /* ms */ > > #define IXGBE_LINK_UP_CHECK_TIMEOUT 1000 /* ms */ > > #define IXGBE_VMDQ_NUM_UC_MAC 4096 /* Maximum nb. of UC MAC addr. */ > > @@ -2194,9 +2197,10 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev *dev, uint16_t nb_rx_q) > > return -EINVAL; > > } > > > > - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; > > - RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * nb_rx_q; > > - > > + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = > > + IXGBE_MAX_RX_QUEUE_NUM / RTE_ETH_DEV_SRIOV(dev).active; > > + RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = > > + pci_dev->max_vfs * RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool; > > return 0; > > } > > > > -- > > 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
* [dpdk-dev] [PATCH v6] net/ixgbe: fix tx_q check failed on PF 2018-01-08 3:06 ` [dpdk-dev] [PATCH v5] net/ixgbe: fix l3fwd start failed on Yanglong Wu 2018-01-08 11:54 ` Ananyev, Konstantin @ 2018-01-10 1:51 ` Yanglong Wu 2018-01-10 1:57 ` Ananyev, Konstantin 1 sibling, 1 reply; 13+ messages in thread From: Yanglong Wu @ 2018-01-10 1:51 UTC (permalink / raw) To: dev, konstantin.ananyev, helin.zhang; +Cc: Yanglong Wu Tx_q check failed on PF that occurred when the SRIOV is active and tx_q > rx_q. The tx_q is equal to nb_q_per_pool. The number of nb_q_per_pool should equeal to max number of queues supported by HW not nb_rx_q. Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to specific drivers) Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> --- v5: Rework according to comments --- v6: rework commit log --- drivers/net/ixgbe/ixgbe_ethdev.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c index ff19a564a..baaeee5d9 100644 --- a/drivers/net/ixgbe/ixgbe_ethdev.c +++ b/drivers/net/ixgbe/ixgbe_ethdev.c @@ -95,6 +95,9 @@ /* Timer value included in XOFF frames. */ #define IXGBE_FC_PAUSE 0x680 +/*Default value of Max Rx Queue*/ +#define IXGBE_MAX_RX_QUEUE_NUM 128 + #define IXGBE_LINK_DOWN_CHECK_TIMEOUT 4000 /* ms */ #define IXGBE_LINK_UP_CHECK_TIMEOUT 1000 /* ms */ #define IXGBE_VMDQ_NUM_UC_MAC 4096 /* Maximum nb. of UC MAC addr. */ @@ -2194,9 +2197,10 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev *dev, uint16_t nb_rx_q) return -EINVAL; } - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; - RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * nb_rx_q; - + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = + IXGBE_MAX_RX_QUEUE_NUM / RTE_ETH_DEV_SRIOV(dev).active; + RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = + pci_dev->max_vfs * RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool; return 0; } -- 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [dpdk-dev] [PATCH v6] net/ixgbe: fix tx_q check failed on PF 2018-01-10 1:51 ` [dpdk-dev] [PATCH v6] net/ixgbe: fix tx_q check failed on PF Yanglong Wu @ 2018-01-10 1:57 ` Ananyev, Konstantin 2018-01-10 2:23 ` Zhang, Helin 0 siblings, 1 reply; 13+ messages in thread From: Ananyev, Konstantin @ 2018-01-10 1:57 UTC (permalink / raw) To: Wu, Yanglong, dev, Zhang, Helin > -----Original Message----- > From: Wu, Yanglong > Sent: Wednesday, January 10, 2018 1:52 AM > To: dev@dpdk.org; Ananyev, Konstantin <konstantin.ananyev@intel.com>; Zhang, Helin <helin.zhang@intel.com> > Cc: Wu, Yanglong <yanglong.wu@intel.com> > Subject: [PATCH v6] net/ixgbe: fix tx_q check failed on PF > > Tx_q check failed on PF that occurred when the SRIOV is active > and tx_q > rx_q. The tx_q is equal to nb_q_per_pool. The number > of nb_q_per_pool should equeal to max number of queues supported > by HW not nb_rx_q. > > Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to > specific drivers) > > Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> > --- > v5: > Rework according to comments > --- > v6: > rework commit log > --- > drivers/net/ixgbe/ixgbe_ethdev.c | 10 +++++++--- > 1 file changed, 7 insertions(+), 3 deletions(-) > > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c > index ff19a564a..baaeee5d9 100644 > --- a/drivers/net/ixgbe/ixgbe_ethdev.c > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c > @@ -95,6 +95,9 @@ > /* Timer value included in XOFF frames. */ > #define IXGBE_FC_PAUSE 0x680 > > +/*Default value of Max Rx Queue*/ > +#define IXGBE_MAX_RX_QUEUE_NUM 128 > + > #define IXGBE_LINK_DOWN_CHECK_TIMEOUT 4000 /* ms */ > #define IXGBE_LINK_UP_CHECK_TIMEOUT 1000 /* ms */ > #define IXGBE_VMDQ_NUM_UC_MAC 4096 /* Maximum nb. of UC MAC addr. */ > @@ -2194,9 +2197,10 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev *dev, uint16_t nb_rx_q) > return -EINVAL; > } > > - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; > - RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * nb_rx_q; > - > + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = > + IXGBE_MAX_RX_QUEUE_NUM / RTE_ETH_DEV_SRIOV(dev).active; > + RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = > + pci_dev->max_vfs * RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool; > return 0; > } > > -- Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com> > 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [dpdk-dev] [PATCH v6] net/ixgbe: fix tx_q check failed on PF 2018-01-10 1:57 ` Ananyev, Konstantin @ 2018-01-10 2:23 ` Zhang, Helin 0 siblings, 0 replies; 13+ messages in thread From: Zhang, Helin @ 2018-01-10 2:23 UTC (permalink / raw) To: Ananyev, Konstantin, Wu, Yanglong, dev > -----Original Message----- > From: Ananyev, Konstantin > Sent: Wednesday, January 10, 2018 9:57 AM > To: Wu, Yanglong; dev@dpdk.org; Zhang, Helin > Subject: RE: [PATCH v6] net/ixgbe: fix tx_q check failed on PF > > > > > -----Original Message----- > > From: Wu, Yanglong > > Sent: Wednesday, January 10, 2018 1:52 AM > > To: dev@dpdk.org; Ananyev, Konstantin <konstantin.ananyev@intel.com>; > > Zhang, Helin <helin.zhang@intel.com> > > Cc: Wu, Yanglong <yanglong.wu@intel.com> > > Subject: [PATCH v6] net/ixgbe: fix tx_q check failed on PF > > > > Tx_q check failed on PF that occurred when the SRIOV is active and > > tx_q > rx_q. The tx_q is equal to nb_q_per_pool. The number of > > nb_q_per_pool should equeal to max number of queues supported by HW > > not nb_rx_q. > > > > Fixes: 27b609cbd1c6 (ethdev: move the multi-queue mode check to > > specific drivers) > > > > Signed-off-by: Yanglong Wu <yanglong.wu@intel.com> > > --- > > v5: > > Rework according to comments > > --- > > v6: > > rework commit log > > --- > > drivers/net/ixgbe/ixgbe_ethdev.c | 10 +++++++--- > > 1 file changed, 7 insertions(+), 3 deletions(-) > > > > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c > > b/drivers/net/ixgbe/ixgbe_ethdev.c > > index ff19a564a..baaeee5d9 100644 > > --- a/drivers/net/ixgbe/ixgbe_ethdev.c > > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c > > @@ -95,6 +95,9 @@ > > /* Timer value included in XOFF frames. */ #define IXGBE_FC_PAUSE > > 0x680 > > > > +/*Default value of Max Rx Queue*/ > > +#define IXGBE_MAX_RX_QUEUE_NUM 128 > > + > > #define IXGBE_LINK_DOWN_CHECK_TIMEOUT 4000 /* ms */ > > #define IXGBE_LINK_UP_CHECK_TIMEOUT 1000 /* ms */ > > #define IXGBE_VMDQ_NUM_UC_MAC 4096 /* Maximum nb. of UC > MAC addr. */ > > @@ -2194,9 +2197,10 @@ ixgbe_check_vf_rss_rxq_num(struct rte_eth_dev > *dev, uint16_t nb_rx_q) > > return -EINVAL; > > } > > > > - RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q; > > - RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = pci_dev->max_vfs * > nb_rx_q; > > - > > + RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = > > + IXGBE_MAX_RX_QUEUE_NUM / > RTE_ETH_DEV_SRIOV(dev).active; > > + RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx = > > + pci_dev->max_vfs * RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool; > > return 0; > > } > > > > -- > > Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com> Applied to dpdk-next-net-intel, with commit log changes. Thanks! /Helin > > > 2.11.0 ^ permalink raw reply [flat|nested] 13+ messages in thread
end of thread, other threads:[~2018-01-10 2:23 UTC | newest]
Thread overview: 13+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
[not found] <20171102161031.167892-2-yanglong.wu@intel.com>
2017-11-20 1:53 ` [dpdk-dev] [PATCH v2] net/ixgbe: fix l3fwd start failed on PF Yanglong Wu
2017-11-20 2:29 ` [dpdk-dev] [PATCH v3] " Yanglong Wu
2017-11-20 2:40 ` [dpdk-dev] [PATCH v4] " Yanglong Wu
2017-11-20 10:43 ` Ananyev, Konstantin
2017-11-21 2:17 ` Wu, Yanglong
2018-01-07 15:34 ` Zhang, Helin
2018-01-08 3:06 ` [dpdk-dev] [PATCH v5] net/ixgbe: fix l3fwd start failed on Yanglong Wu
2018-01-08 11:54 ` Ananyev, Konstantin
2018-01-09 2:29 ` Wu, Yanglong
2018-01-09 12:39 ` Ananyev, Konstantin
2018-01-10 1:51 ` [dpdk-dev] [PATCH v6] net/ixgbe: fix tx_q check failed on PF Yanglong Wu
2018-01-10 1:57 ` Ananyev, Konstantin
2018-01-10 2:23 ` Zhang, Helin
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).