* [dpdk-dev] [PATCH] net/ixgbe: check jumbo frame enable parameter @ 2018-11-01 7:55 wei zhao 2018-11-01 8:28 ` [dpdk-dev] [PATCH v2] " wei zhao 2018-11-01 17:46 ` [dpdk-dev] [PATCH] " Stephen Hemminger 0 siblings, 2 replies; 7+ messages in thread From: wei zhao @ 2018-11-01 7:55 UTC (permalink / raw) To: wei.zhao1, dev; +Cc: stable, qi.z.zhang There is necessary to do some check of max packet size boundary for code safe when enable jumbo frame. Signed-off-by: Wei Zhao <wei.zhao1@intel.com> --- drivers/net/ixgbe/ixgbe_rxtx.c | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+) diff --git a/drivers/net/ixgbe/ixgbe_rxtx.c b/drivers/net/ixgbe/ixgbe_rxtx.c index 2f0262a..d304dee 100644 --- a/drivers/net/ixgbe/ixgbe_rxtx.c +++ b/drivers/net/ixgbe/ixgbe_rxtx.c @@ -4814,6 +4814,34 @@ void __attribute__((cold)) return 0; } +int __attribute__((cold)) +ixgbe_dev_jumboenable_check(struct rte_eth_dev *dev, + uint16_t max_rx_pkt_len) +{ + struct rte_eth_dev_info dev_info; + + ixgbe_dev_info_get(dev, &dev_info); + + /* check that max packet size is within the allowed range */ + if (max_rx_pkt_len < ETHER_MIN_MTU) { + PMD_INIT_LOG(ERR, "max packet size is too small."); + return -EINVAL; + } + + if (max_rx_pkt_len > dev_info.max_rx_pktlen) { + PMD_INIT_LOG(ERR, "max packet size is too big."); + return -EINVAL; + } + + /* check jumbo mode if needed */ + if (max_rx_pkt_len < ETHER_MAX_LEN) { + PMD_INIT_LOG(ERR, "No need to enable jumbo."); + return -EINVAL; + } + + return 0; +} + /* * Initializes Receive Unit. */ @@ -4865,6 +4893,9 @@ int __attribute__((cold)) * Configure jumbo frame support, if any. */ if (rx_conf->offloads & DEV_RX_OFFLOAD_JUMBO_FRAME) { + if (ixgbe_dev_jumboenable_check(dev, rx_conf->max_rx_pkt_len)) + return -EINVAL; + hlreg0 |= IXGBE_HLREG0_JUMBOEN; maxfrs = IXGBE_READ_REG(hw, IXGBE_MAXFRS); maxfrs &= 0x0000FFFF; -- 1.8.3.1 ^ permalink raw reply [flat|nested] 7+ messages in thread
* [dpdk-dev] [PATCH v2] net/ixgbe: check jumbo frame enable parameter 2018-11-01 7:55 [dpdk-dev] [PATCH] net/ixgbe: check jumbo frame enable parameter wei zhao @ 2018-11-01 8:28 ` wei zhao 2018-11-01 8:40 ` Zhao1, Wei 2018-11-01 17:46 ` [dpdk-dev] [PATCH] " Stephen Hemminger 1 sibling, 1 reply; 7+ messages in thread From: wei zhao @ 2018-11-01 8:28 UTC (permalink / raw) To: wei.zhao1, dev; +Cc: stable, qi.z.zhang, yamashita.hideyuki, stephen There is necessary to do some check of max packet size boundary for code safe in order to avoid error in register setting when enable jumbo. Signed-off-by: Wei Zhao <wei.zhao1@intel.com> --- v2: -add more log infomation and fix build issue --- drivers/net/ixgbe/ixgbe_ethdev.c | 4 +--- drivers/net/ixgbe/ixgbe_ethdev.h | 2 ++ drivers/net/ixgbe/ixgbe_rxtx.c | 31 +++++++++++++++++++++++++++++++ 3 files changed, 34 insertions(+), 3 deletions(-) diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c index 5a2c351..d65a911 100644 --- a/drivers/net/ixgbe/ixgbe_ethdev.c +++ b/drivers/net/ixgbe/ixgbe_ethdev.c @@ -173,8 +173,6 @@ static int ixgbe_dev_queue_stats_mapping_set(struct rte_eth_dev *eth_dev, uint8_t is_rx); static int ixgbe_fw_version_get(struct rte_eth_dev *dev, char *fw_version, size_t fw_size); -static void ixgbe_dev_info_get(struct rte_eth_dev *dev, - struct rte_eth_dev_info *dev_info); static const uint32_t *ixgbe_dev_supported_ptypes_get(struct rte_eth_dev *dev); static void ixgbevf_dev_info_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info); @@ -3692,7 +3690,7 @@ static int ixgbevf_dev_xstats_get_names(__rte_unused struct rte_eth_dev *dev, return 0; } -static void +void ixgbe_dev_info_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info) { struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev); diff --git a/drivers/net/ixgbe/ixgbe_ethdev.h b/drivers/net/ixgbe/ixgbe_ethdev.h index d0b9396..a474be4 100644 --- a/drivers/net/ixgbe/ixgbe_ethdev.h +++ b/drivers/net/ixgbe/ixgbe_ethdev.h @@ -727,6 +727,8 @@ int ixgbe_action_rss_same(const struct rte_flow_action_rss *comp, const struct rte_flow_action_rss *with); int ixgbe_config_rss_filter(struct rte_eth_dev *dev, struct ixgbe_rte_flow_rss_conf *conf, bool add); +void ixgbe_dev_info_get(struct rte_eth_dev *dev, + struct rte_eth_dev_info *dev_info); static inline int ixgbe_ethertype_filter_lookup(struct ixgbe_filter_info *filter_info, diff --git a/drivers/net/ixgbe/ixgbe_rxtx.c b/drivers/net/ixgbe/ixgbe_rxtx.c index 2f0262a..94051ca 100644 --- a/drivers/net/ixgbe/ixgbe_rxtx.c +++ b/drivers/net/ixgbe/ixgbe_rxtx.c @@ -4814,6 +4814,34 @@ void __attribute__((cold)) return 0; } +static int __attribute__((cold)) +ixgbe_dev_jumboenable_check(struct rte_eth_dev *dev, + uint16_t max_rx_pkt_len) +{ + struct rte_eth_dev_info dev_info; + + ixgbe_dev_info_get(dev, &dev_info); + + /* check that max packet size is within the allowed range */ + if (max_rx_pkt_len < ETHER_MIN_MTU) { + PMD_INIT_LOG(ERR, "max packet size is too small."); + return -EINVAL; + } + + if (max_rx_pkt_len > dev_info.max_rx_pktlen) { + PMD_INIT_LOG(ERR, "max packet size is too big."); + return -EINVAL; + } + + /* check jumbo mode if needed */ + if (max_rx_pkt_len < ETHER_MAX_LEN) { + PMD_INIT_LOG(ERR, "No need to enable jumbo."); + return -EINVAL; + } + + return 0; +} + /* * Initializes Receive Unit. */ @@ -4865,6 +4893,9 @@ int __attribute__((cold)) * Configure jumbo frame support, if any. */ if (rx_conf->offloads & DEV_RX_OFFLOAD_JUMBO_FRAME) { + if (ixgbe_dev_jumboenable_check(dev, rx_conf->max_rx_pkt_len)) + return -EINVAL; + hlreg0 |= IXGBE_HLREG0_JUMBOEN; maxfrs = IXGBE_READ_REG(hw, IXGBE_MAXFRS); maxfrs &= 0x0000FFFF; -- 1.8.3.1 ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [dpdk-dev] [PATCH v2] net/ixgbe: check jumbo frame enable parameter 2018-11-01 8:28 ` [dpdk-dev] [PATCH v2] " wei zhao @ 2018-11-01 8:40 ` Zhao1, Wei 2018-11-05 16:34 ` Zhang, Qi Z 0 siblings, 1 reply; 7+ messages in thread From: Zhao1, Wei @ 2018-11-01 8:40 UTC (permalink / raw) To: dev; +Cc: stable, Zhang, Qi Z, yamashita.hideyuki, stephen Add the background for this patch. https://mails.dpdk.org/archives/dev/2018-November/117771.html https://mails.dpdk.org/archives/dev/2018-November/117772.html > -----Original Message----- > From: Zhao1, Wei > Sent: Thursday, November 1, 2018 4:29 PM > To: Zhao1, Wei <wei.zhao1@intel.com>; dev@dpdk.org > Cc: stable@dpdk.org; Zhang, Qi Z <qi.z.zhang@intel.com>; > yamashita.hideyuki@po.ntt-tx.co.jp; stephen@networkplumber.org > Subject: [PATCH v2] net/ixgbe: check jumbo frame enable parameter > > There is necessary to do some check of max packet size boundary for code > safe in order to avoid error in register setting when enable jumbo. > > Signed-off-by: Wei Zhao <wei.zhao1@intel.com> > > --- > > v2: > -add more log infomation and fix build issue > --- > drivers/net/ixgbe/ixgbe_ethdev.c | 4 +--- > drivers/net/ixgbe/ixgbe_ethdev.h | 2 ++ > drivers/net/ixgbe/ixgbe_rxtx.c | 31 > +++++++++++++++++++++++++++++++ > 3 files changed, 34 insertions(+), 3 deletions(-) > > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c > b/drivers/net/ixgbe/ixgbe_ethdev.c > index 5a2c351..d65a911 100644 > --- a/drivers/net/ixgbe/ixgbe_ethdev.c > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c > @@ -173,8 +173,6 @@ static int > ixgbe_dev_queue_stats_mapping_set(struct rte_eth_dev *eth_dev, > uint8_t is_rx); > static int ixgbe_fw_version_get(struct rte_eth_dev *dev, char *fw_version, > size_t fw_size); > -static void ixgbe_dev_info_get(struct rte_eth_dev *dev, > - struct rte_eth_dev_info *dev_info); > static const uint32_t *ixgbe_dev_supported_ptypes_get(struct > rte_eth_dev *dev); static void ixgbevf_dev_info_get(struct rte_eth_dev > *dev, > struct rte_eth_dev_info *dev_info); @@ - > 3692,7 +3690,7 @@ static int ixgbevf_dev_xstats_get_names(__rte_unused > struct rte_eth_dev *dev, > return 0; > } > > -static void > +void > ixgbe_dev_info_get(struct rte_eth_dev *dev, struct rte_eth_dev_info > *dev_info) { > struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev); diff -- > git a/drivers/net/ixgbe/ixgbe_ethdev.h b/drivers/net/ixgbe/ixgbe_ethdev.h > index d0b9396..a474be4 100644 > --- a/drivers/net/ixgbe/ixgbe_ethdev.h > +++ b/drivers/net/ixgbe/ixgbe_ethdev.h > @@ -727,6 +727,8 @@ int ixgbe_action_rss_same(const struct > rte_flow_action_rss *comp, > const struct rte_flow_action_rss *with); int > ixgbe_config_rss_filter(struct rte_eth_dev *dev, > struct ixgbe_rte_flow_rss_conf *conf, bool add); > +void ixgbe_dev_info_get(struct rte_eth_dev *dev, > + struct rte_eth_dev_info *dev_info); > > static inline int > ixgbe_ethertype_filter_lookup(struct ixgbe_filter_info *filter_info, diff --git > a/drivers/net/ixgbe/ixgbe_rxtx.c b/drivers/net/ixgbe/ixgbe_rxtx.c index > 2f0262a..94051ca 100644 > --- a/drivers/net/ixgbe/ixgbe_rxtx.c > +++ b/drivers/net/ixgbe/ixgbe_rxtx.c > @@ -4814,6 +4814,34 @@ void __attribute__((cold)) > return 0; > } > > +static int __attribute__((cold)) > +ixgbe_dev_jumboenable_check(struct rte_eth_dev *dev, > + uint16_t max_rx_pkt_len) > +{ > + struct rte_eth_dev_info dev_info; > + > + ixgbe_dev_info_get(dev, &dev_info); > + > + /* check that max packet size is within the allowed range */ > + if (max_rx_pkt_len < ETHER_MIN_MTU) { > + PMD_INIT_LOG(ERR, "max packet size is too small."); > + return -EINVAL; > + } > + > + if (max_rx_pkt_len > dev_info.max_rx_pktlen) { > + PMD_INIT_LOG(ERR, "max packet size is too big."); > + return -EINVAL; > + } > + > + /* check jumbo mode if needed */ > + if (max_rx_pkt_len < ETHER_MAX_LEN) { > + PMD_INIT_LOG(ERR, "No need to enable jumbo."); > + return -EINVAL; > + } > + > + return 0; > +} > + > /* > * Initializes Receive Unit. > */ > @@ -4865,6 +4893,9 @@ int __attribute__((cold)) > * Configure jumbo frame support, if any. > */ > if (rx_conf->offloads & DEV_RX_OFFLOAD_JUMBO_FRAME) { > + if (ixgbe_dev_jumboenable_check(dev, rx_conf- > >max_rx_pkt_len)) > + return -EINVAL; > + > hlreg0 |= IXGBE_HLREG0_JUMBOEN; > maxfrs = IXGBE_READ_REG(hw, IXGBE_MAXFRS); > maxfrs &= 0x0000FFFF; > -- > 1.8.3.1 ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [dpdk-dev] [PATCH v2] net/ixgbe: check jumbo frame enable parameter 2018-11-01 8:40 ` Zhao1, Wei @ 2018-11-05 16:34 ` Zhang, Qi Z 2018-11-06 2:49 ` Zhao1, Wei 0 siblings, 1 reply; 7+ messages in thread From: Zhang, Qi Z @ 2018-11-05 16:34 UTC (permalink / raw) To: Zhao1, Wei, dev; +Cc: stable, yamashita.hideyuki, stephen Hi Wei: > -----Original Message----- > From: Zhao1, Wei > Sent: Thursday, November 1, 2018 1:40 AM > To: dev@dpdk.org > Cc: stable@dpdk.org; Zhang, Qi Z <qi.z.zhang@intel.com>; > yamashita.hideyuki@po.ntt-tx.co.jp; stephen@networkplumber.org > Subject: RE: [PATCH v2] net/ixgbe: check jumbo frame enable parameter > > Add the background for this patch. > https://mails.dpdk.org/archives/dev/2018-November/117771.html > https://mails.dpdk.org/archives/dev/2018-November/117772.html > > > > > -----Original Message----- > > From: Zhao1, Wei > > Sent: Thursday, November 1, 2018 4:29 PM > > To: Zhao1, Wei <wei.zhao1@intel.com>; dev@dpdk.org > > Cc: stable@dpdk.org; Zhang, Qi Z <qi.z.zhang@intel.com>; > > yamashita.hideyuki@po.ntt-tx.co.jp; stephen@networkplumber.org > > Subject: [PATCH v2] net/ixgbe: check jumbo frame enable parameter > > > > There is necessary to do some check of max packet size boundary for > > code safe in order to avoid error in register setting when enable jumbo. > > > > Signed-off-by: Wei Zhao <wei.zhao1@intel.com> > > > > --- > > > > v2: > > -add more log infomation and fix build issue > > --- > > drivers/net/ixgbe/ixgbe_ethdev.c | 4 +--- > > drivers/net/ixgbe/ixgbe_ethdev.h | 2 ++ > > drivers/net/ixgbe/ixgbe_rxtx.c | 31 > > +++++++++++++++++++++++++++++++ > > 3 files changed, 34 insertions(+), 3 deletions(-) > > > > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c > > b/drivers/net/ixgbe/ixgbe_ethdev.c > > index 5a2c351..d65a911 100644 > > --- a/drivers/net/ixgbe/ixgbe_ethdev.c > > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c > > @@ -173,8 +173,6 @@ static int > > ixgbe_dev_queue_stats_mapping_set(struct rte_eth_dev *eth_dev, > > uint8_t is_rx); > > static int ixgbe_fw_version_get(struct rte_eth_dev *dev, char > *fw_version, > > size_t fw_size); > > -static void ixgbe_dev_info_get(struct rte_eth_dev *dev, > > - struct rte_eth_dev_info *dev_info); > > static const uint32_t *ixgbe_dev_supported_ptypes_get(struct > > rte_eth_dev *dev); static void ixgbevf_dev_info_get(struct > > rte_eth_dev *dev, > > struct rte_eth_dev_info *dev_info); @@ - > > 3692,7 +3690,7 @@ static int ixgbevf_dev_xstats_get_names(__rte_unused > > struct rte_eth_dev *dev, > > return 0; > > } > > > > -static void > > +void > > ixgbe_dev_info_get(struct rte_eth_dev *dev, struct rte_eth_dev_info > > *dev_info) { > > struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev); diff -- > > git a/drivers/net/ixgbe/ixgbe_ethdev.h > > b/drivers/net/ixgbe/ixgbe_ethdev.h > > index d0b9396..a474be4 100644 > > --- a/drivers/net/ixgbe/ixgbe_ethdev.h > > +++ b/drivers/net/ixgbe/ixgbe_ethdev.h > > @@ -727,6 +727,8 @@ int ixgbe_action_rss_same(const struct > > rte_flow_action_rss *comp, > > const struct rte_flow_action_rss *with); int > > ixgbe_config_rss_filter(struct rte_eth_dev *dev, > > struct ixgbe_rte_flow_rss_conf *conf, bool add); > > +void ixgbe_dev_info_get(struct rte_eth_dev *dev, > > + struct rte_eth_dev_info *dev_info); > > > > static inline int > > ixgbe_ethertype_filter_lookup(struct ixgbe_filter_info *filter_info, > > diff --git a/drivers/net/ixgbe/ixgbe_rxtx.c > > b/drivers/net/ixgbe/ixgbe_rxtx.c index 2f0262a..94051ca 100644 > > --- a/drivers/net/ixgbe/ixgbe_rxtx.c > > +++ b/drivers/net/ixgbe/ixgbe_rxtx.c > > @@ -4814,6 +4814,34 @@ void __attribute__((cold)) > > return 0; > > } > > > > +static int __attribute__((cold)) > > +ixgbe_dev_jumboenable_check(struct rte_eth_dev *dev, > > + uint16_t max_rx_pkt_len) it's better to return Boolean for this function > > +{ > > + struct rte_eth_dev_info dev_info; > > + > > + ixgbe_dev_info_get(dev, &dev_info); > > + > > + /* check that max packet size is within the allowed range */ > > + if (max_rx_pkt_len < ETHER_MIN_MTU) { > > + PMD_INIT_LOG(ERR, "max packet size is too small."); > > + return -EINVAL; > > + } Why this is only needed for DEV_RX_OFFLOAD_JUMBO_FRAME? And we will compare to ETHER_MAX_LEN which is a bigger value, this looks redundant. > > + > > + if (max_rx_pkt_len > dev_info.max_rx_pktlen) { > > + PMD_INIT_LOG(ERR, "max packet size is too big."); > > + return -EINVAL; > > + } dev_info.max_rx_pktlen = 15872, it should be replaced by a macro, then we don't need to call ixgbe_dev_info_get here. and why this is only needed for DEV_RX_OFFLOAD_JUMBO_FRAME? > > + > > + /* check jumbo mode if needed */ > > + if (max_rx_pkt_len < ETHER_MAX_LEN) { > > + PMD_INIT_LOG(ERR, "No need to enable jumbo."); > > + return -EINVAL; And I agree with Stephen's comment, Looks all of these is common for all drivers, it should be moved to ether layer. > > + } > > + > > + return 0; > > +} > > + > > /* > > * Initializes Receive Unit. > > */ > > @@ -4865,6 +4893,9 @@ int __attribute__((cold)) > > * Configure jumbo frame support, if any. > > */ > > if (rx_conf->offloads & DEV_RX_OFFLOAD_JUMBO_FRAME) { > > + if (ixgbe_dev_jumboenable_check(dev, rx_conf- > > >max_rx_pkt_len)) > > + return -EINVAL; > > + > > hlreg0 |= IXGBE_HLREG0_JUMBOEN; > > maxfrs = IXGBE_READ_REG(hw, IXGBE_MAXFRS); > > maxfrs &= 0x0000FFFF; > > -- > > 1.8.3.1 ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [dpdk-dev] [PATCH v2] net/ixgbe: check jumbo frame enable parameter 2018-11-05 16:34 ` Zhang, Qi Z @ 2018-11-06 2:49 ` Zhao1, Wei 0 siblings, 0 replies; 7+ messages in thread From: Zhao1, Wei @ 2018-11-06 2:49 UTC (permalink / raw) To: Zhang, Qi Z, dev; +Cc: stable, yamashita.hideyuki, stephen > -----Original Message----- > From: Zhang, Qi Z > Sent: Tuesday, November 6, 2018 12:34 AM > To: Zhao1, Wei <wei.zhao1@intel.com>; dev@dpdk.org > Cc: stable@dpdk.org; yamashita.hideyuki@po.ntt-tx.co.jp; > stephen@networkplumber.org > Subject: RE: [PATCH v2] net/ixgbe: check jumbo frame enable parameter > > Hi Wei: > > > -----Original Message----- > > From: Zhao1, Wei > > Sent: Thursday, November 1, 2018 1:40 AM > > To: dev@dpdk.org > > Cc: stable@dpdk.org; Zhang, Qi Z <qi.z.zhang@intel.com>; > > yamashita.hideyuki@po.ntt-tx.co.jp; stephen@networkplumber.org > > Subject: RE: [PATCH v2] net/ixgbe: check jumbo frame enable parameter > > > > Add the background for this patch. > > https://mails.dpdk.org/archives/dev/2018-November/117771.html > > https://mails.dpdk.org/archives/dev/2018-November/117772.html > > > > > > > > > -----Original Message----- > > > From: Zhao1, Wei > > > Sent: Thursday, November 1, 2018 4:29 PM > > > To: Zhao1, Wei <wei.zhao1@intel.com>; dev@dpdk.org > > > Cc: stable@dpdk.org; Zhang, Qi Z <qi.z.zhang@intel.com>; > > > yamashita.hideyuki@po.ntt-tx.co.jp; stephen@networkplumber.org > > > Subject: [PATCH v2] net/ixgbe: check jumbo frame enable parameter > > > > > > There is necessary to do some check of max packet size boundary for > > > code safe in order to avoid error in register setting when enable jumbo. > > > > > > Signed-off-by: Wei Zhao <wei.zhao1@intel.com> > > > > > > --- > > > > > > v2: > > > -add more log infomation and fix build issue > > > --- > > > drivers/net/ixgbe/ixgbe_ethdev.c | 4 +--- > > > drivers/net/ixgbe/ixgbe_ethdev.h | 2 ++ > > > drivers/net/ixgbe/ixgbe_rxtx.c | 31 > > > +++++++++++++++++++++++++++++++ > > > 3 files changed, 34 insertions(+), 3 deletions(-) > > > > > > diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c > > > b/drivers/net/ixgbe/ixgbe_ethdev.c > > > index 5a2c351..d65a911 100644 > > > --- a/drivers/net/ixgbe/ixgbe_ethdev.c > > > +++ b/drivers/net/ixgbe/ixgbe_ethdev.c > > > @@ -173,8 +173,6 @@ static int > > > ixgbe_dev_queue_stats_mapping_set(struct rte_eth_dev *eth_dev, > > > uint8_t is_rx); > > > static int ixgbe_fw_version_get(struct rte_eth_dev *dev, char > > *fw_version, > > > size_t fw_size); > > > -static void ixgbe_dev_info_get(struct rte_eth_dev *dev, > > > - struct rte_eth_dev_info *dev_info); > > > static const uint32_t *ixgbe_dev_supported_ptypes_get(struct > > > rte_eth_dev *dev); static void ixgbevf_dev_info_get(struct > > > rte_eth_dev *dev, > > > struct rte_eth_dev_info *dev_info); @@ - > > > 3692,7 +3690,7 @@ static int > > > ixgbevf_dev_xstats_get_names(__rte_unused > > > struct rte_eth_dev *dev, > > > return 0; > > > } > > > > > > -static void > > > +void > > > ixgbe_dev_info_get(struct rte_eth_dev *dev, struct rte_eth_dev_info > > > *dev_info) { > > > struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev); diff -- > > > git a/drivers/net/ixgbe/ixgbe_ethdev.h > > > b/drivers/net/ixgbe/ixgbe_ethdev.h > > > index d0b9396..a474be4 100644 > > > --- a/drivers/net/ixgbe/ixgbe_ethdev.h > > > +++ b/drivers/net/ixgbe/ixgbe_ethdev.h > > > @@ -727,6 +727,8 @@ int ixgbe_action_rss_same(const struct > > > rte_flow_action_rss *comp, > > > const struct rte_flow_action_rss *with); int > > > ixgbe_config_rss_filter(struct rte_eth_dev *dev, > > > struct ixgbe_rte_flow_rss_conf *conf, bool add); > > > +void ixgbe_dev_info_get(struct rte_eth_dev *dev, > > > + struct rte_eth_dev_info *dev_info); > > > > > > static inline int > > > ixgbe_ethertype_filter_lookup(struct ixgbe_filter_info > > > *filter_info, diff --git a/drivers/net/ixgbe/ixgbe_rxtx.c > > > b/drivers/net/ixgbe/ixgbe_rxtx.c index 2f0262a..94051ca 100644 > > > --- a/drivers/net/ixgbe/ixgbe_rxtx.c > > > +++ b/drivers/net/ixgbe/ixgbe_rxtx.c > > > @@ -4814,6 +4814,34 @@ void __attribute__((cold)) > > > return 0; > > > } > > > > > > +static int __attribute__((cold)) > > > +ixgbe_dev_jumboenable_check(struct rte_eth_dev *dev, > > > + uint16_t max_rx_pkt_len) > > it's better to return Boolean for this function Ok > > > > +{ > > > + struct rte_eth_dev_info dev_info; > > > + > > > + ixgbe_dev_info_get(dev, &dev_info); > > > + > > > + /* check that max packet size is within the allowed range */ > > > + if (max_rx_pkt_len < ETHER_MIN_MTU) { > > > + PMD_INIT_LOG(ERR, "max packet size is too small."); > > > + return -EINVAL; > > > + } > > Why this is only needed for DEV_RX_OFFLOAD_JUMBO_FRAME? > And we will compare to ETHER_MAX_LEN which is a bigger value, this looks > redundant. Ok, I will check that > > > > + > > > + if (max_rx_pkt_len > dev_info.max_rx_pktlen) { > > > + PMD_INIT_LOG(ERR, "max packet size is too big."); > > > + return -EINVAL; > > > + } > > dev_info.max_rx_pktlen = 15872, it should be replaced by a macro, then we > don't need to call ixgbe_dev_info_get here. > and why this is only needed for DEV_RX_OFFLOAD_JUMBO_FRAME? > > > + > > > + /* check jumbo mode if needed */ > > > + if (max_rx_pkt_len < ETHER_MAX_LEN) { > > > + PMD_INIT_LOG(ERR, "No need to enable jumbo."); > > > + return -EINVAL; > > And I agree with Stephen's comment, > Looks all of these is common for all drivers, it should be moved to ether layer. Ok， I will move to rte layer > > > > + } > > > + > > > + return 0; > > > +} > > > + > > > /* > > > * Initializes Receive Unit. > > > */ > > > @@ -4865,6 +4893,9 @@ int __attribute__((cold)) > > > * Configure jumbo frame support, if any. > > > */ > > > if (rx_conf->offloads & DEV_RX_OFFLOAD_JUMBO_FRAME) { > > > + if (ixgbe_dev_jumboenable_check(dev, rx_conf- > > > >max_rx_pkt_len)) > > > + return -EINVAL; > > > + > > > hlreg0 |= IXGBE_HLREG0_JUMBOEN; > > > maxfrs = IXGBE_READ_REG(hw, IXGBE_MAXFRS); > > > maxfrs &= 0x0000FFFF; > > > -- > > > 1.8.3.1 ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [dpdk-dev] [PATCH] net/ixgbe: check jumbo frame enable parameter 2018-11-01 7:55 [dpdk-dev] [PATCH] net/ixgbe: check jumbo frame enable parameter wei zhao 2018-11-01 8:28 ` [dpdk-dev] [PATCH v2] " wei zhao @ 2018-11-01 17:46 ` Stephen Hemminger 2018-11-05 9:13 ` Zhao1, Wei 1 sibling, 1 reply; 7+ messages in thread From: Stephen Hemminger @ 2018-11-01 17:46 UTC (permalink / raw) To: wei zhao; +Cc: dev, stable, qi.z.zhang On Thu, 1 Nov 2018 15:55:26 +0800 wei zhao <wei.zhao1@intel.com> wrote: > + /* check that max packet size is within the allowed range */ > + if (max_rx_pkt_len < ETHER_MIN_MTU) { > + PMD_INIT_LOG(ERR, "max packet size is too small."); > + return -EINVAL; > + } > + > + if (max_rx_pkt_len > dev_info.max_rx_pktlen) { > + PMD_INIT_LOG(ERR, "max packet size is too big."); > + return -EINVAL; > + } > + > + /* check jumbo mode if needed */ > + if (max_rx_pkt_len < ETHER_MAX_LEN) { > + PMD_INIT_LOG(ERR, "No need to enable jumbo."); > + return -EINVAL; > + } > + Yes, these checks are needed and for other devices as well. Why not add them into rte_ethdev instead? ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [dpdk-dev] [PATCH] net/ixgbe: check jumbo frame enable parameter 2018-11-01 17:46 ` [dpdk-dev] [PATCH] " Stephen Hemminger @ 2018-11-05 9:13 ` Zhao1, Wei 0 siblings, 0 replies; 7+ messages in thread From: Zhao1, Wei @ 2018-11-05 9:13 UTC (permalink / raw) To: Stephen Hemminger; +Cc: dev, stable, Zhang, Qi Z > -----Original Message----- > From: Stephen Hemminger [mailto:stephen@networkplumber.org] > Sent: Friday, November 2, 2018 1:46 AM > To: Zhao1, Wei <wei.zhao1@intel.com> > Cc: dev@dpdk.org; stable@dpdk.org; Zhang, Qi Z <qi.z.zhang@intel.com> > Subject: Re: [dpdk-dev] [PATCH] net/ixgbe: check jumbo frame enable > parameter > > On Thu, 1 Nov 2018 15:55:26 +0800 > wei zhao <wei.zhao1@intel.com> wrote: > > > + /* check that max packet size is within the allowed range */ > > + if (max_rx_pkt_len < ETHER_MIN_MTU) { > > + PMD_INIT_LOG(ERR, "max packet size is too small."); > > + return -EINVAL; > > + } > > + > > + if (max_rx_pkt_len > dev_info.max_rx_pktlen) { > > + PMD_INIT_LOG(ERR, "max packet size is too big."); > > + return -EINVAL; > > + } > > + > > + /* check jumbo mode if needed */ > > + if (max_rx_pkt_len < ETHER_MAX_LEN) { > > + PMD_INIT_LOG(ERR, "No need to enable jumbo."); > > + return -EINVAL; > > + } > > + > > Yes, these checks are needed and for other devices as well. > Why not add them into rte_ethdev instead? I am not familiar with NIC except ixgbe and i40e, I am not sure property of other NIC and rationality of these check.... ^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2018-11-06 2:49 UTC | newest] Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2018-11-01 7:55 [dpdk-dev] [PATCH] net/ixgbe: check jumbo frame enable parameter wei zhao 2018-11-01 8:28 ` [dpdk-dev] [PATCH v2] " wei zhao 2018-11-01 8:40 ` Zhao1, Wei 2018-11-05 16:34 ` Zhang, Qi Z 2018-11-06 2:49 ` Zhao1, Wei 2018-11-01 17:46 ` [dpdk-dev] [PATCH] " Stephen Hemminger 2018-11-05 9:13 ` Zhao1, Wei
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).