| Message ID | 1394603618-1044-6-git-send-email-jeffrey.t.kirsher@intel.com |
|---|---|
| State | Changes Requested, archived |
| Delegated to: | David Miller |
| Headers | show |
On Wed, Mar 12, 2014 at 7:53 AM, Jeff Kirsher <jeffrey.t.kirsher@intel.com> wrote: > This patch adds the necessary defines, variables, and stats updating > procedures to use the commands implemented in debugfs. Guys, Do we have a clear criteria when it makes sense to have debugfs for network driver and when it doesn't, for example we got this cold shower http://marc.info/?l=linux-netdev&m=135959539719935&w=2 when attempting to do so and I am not fully clear (or want to sharpen the sow) how to review a concrete design/need for debugs to fall into yes, it makes sense, or no, it doesn't b/c there are existing mechanisms that can be used or extended for that matter. Or. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
> -----Original Message----- > From: netdev-owner@vger.kernel.org [mailto:netdev-owner@vger.kernel.org] > On Behalf Of Or Gerlitz > Sent: Wednesday, March 12, 2014 8:58 AM > To: Kirsher, Jeffrey T; David Miller; Andy Gospodarek > Cc: Wyborny, Carolyn; netdev@vger.kernel.org; sassmann@redhat.com; Josh > Hay; Amir Vadai; Yevgeny Petrilin > Subject: Re: [net-next 05/13] igb: Add support for debugfs > > On Wed, Mar 12, 2014 at 7:53 AM, Jeff Kirsher <jeffrey.t.kirsher@intel.com> > wrote: > > This patch adds the necessary defines, variables, and stats updating > > procedures to use the commands implemented in debugfs. > > Guys, > > Do we have a clear criteria when it makes sense to have debugfs for network > driver and when it doesn't, for example we got this cold shower > http://marc.info/?l=linux-netdev&m=135959539719935&w=2 when attempting > to do so and I am not fully clear (or want to sharpen the sow) how to review a > concrete design/need for debugs to fall into yes, it makes sense, or no, it > doesn't b/c there are existing mechanisms that can be used or extended for that > matter. > > Or. > -- > To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a > message to majordomo@vger.kernel.org More majordomo info at > http://vger.kernel.org/majordomo-info.html Josh and I tried very hard to not reimplement anything already available in some other tool or mechanism and used the examples we already have for this in our other drivers. That being said, we'll abide and follow any clear criteria provided. Thanks, Carolyn Carolyn Wyborny Linux Development Networking Division Intel Corporation -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
From: "Wyborny, Carolyn" <carolyn.wyborny@intel.com> Date: Wed, 12 Mar 2014 16:35:49 +0000 > Josh and I tried very hard to not reimplement anything already > available in some other tool or mechanism and used the examples we > already have for this in our other drivers. That being said, we'll > abide and follow any clear criteria provided. I don't think what's being done here is appropriate. ethtool can provide the extended statistics. We already have a way to dump registers in ethtool as well. If you want to add fine-grained register read support and write support, that would be a nice addition to ethtool. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
> -----Original Message----- > From: David Miller [mailto:davem@davemloft.net] > Sent: Wednesday, March 12, 2014 12:47 PM > To: Wyborny, Carolyn > Cc: or.gerlitz@gmail.com; Kirsher, Jeffrey T; gospo@redhat.com; > netdev@vger.kernel.org; sassmann@redhat.com; hayja@mail.uc.edu; > amirv@mellanox.com; yevgenyp@mellanox.com > Subject: Re: [net-next 05/13] igb: Add support for debugfs > > From: "Wyborny, Carolyn" <carolyn.wyborny@intel.com> > Date: Wed, 12 Mar 2014 16:35:49 +0000 > > > Josh and I tried very hard to not reimplement anything already > > available in some other tool or mechanism and used the examples we > > already have for this in our other drivers. That being said, we'll > > abide and follow any clear criteria provided. > > I don't think what's being done here is appropriate. > > ethtool can provide the extended statistics. > > We already have a way to dump registers in ethtool as well. If you want to add > fine-grained register read support and write support, that would be a nice > addition to ethtool. Ok, I'll remove the register outputs. Is there any objection to the other functionality provided? Thanks, Carolyn -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
From: "Wyborny, Carolyn" <carolyn.wyborny@intel.com> Date: Wed, 12 Mar 2014 20:13:53 +0000 >> -----Original Message----- >> From: David Miller [mailto:davem@davemloft.net] >> Sent: Wednesday, March 12, 2014 12:47 PM >> To: Wyborny, Carolyn >> Cc: or.gerlitz@gmail.com; Kirsher, Jeffrey T; gospo@redhat.com; >> netdev@vger.kernel.org; sassmann@redhat.com; hayja@mail.uc.edu; >> amirv@mellanox.com; yevgenyp@mellanox.com >> Subject: Re: [net-next 05/13] igb: Add support for debugfs >> >> From: "Wyborny, Carolyn" <carolyn.wyborny@intel.com> >> Date: Wed, 12 Mar 2014 16:35:49 +0000 >> >> > Josh and I tried very hard to not reimplement anything already >> > available in some other tool or mechanism and used the examples we >> > already have for this in our other drivers. That being said, we'll >> > abide and follow any clear criteria provided. >> >> I don't think what's being done here is appropriate. >> >> ethtool can provide the extended statistics. >> >> We already have a way to dump registers in ethtool as well. If you want to add >> fine-grained register read support and write support, that would be a nice >> addition to ethtool. > > Ok, I'll remove the register outputs. Is there any objection to the other functionality provided? I said the statistics can be done with ethtool, so I object to everything. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
> -----Original Message----- > From: David Miller [mailto:davem@davemloft.net] > Sent: Wednesday, March 12, 2014 1:24 PM > To: Wyborny, Carolyn > Cc: or.gerlitz@gmail.com; Kirsher, Jeffrey T; gospo@redhat.com; > netdev@vger.kernel.org; sassmann@redhat.com; hayja@mail.uc.edu; > amirv@mellanox.com; yevgenyp@mellanox.com > Subject: Re: [net-next 05/13] igb: Add support for debugfs > > From: "Wyborny, Carolyn" <carolyn.wyborny@intel.com> > Date: Wed, 12 Mar 2014 20:13:53 +0000 > > >> -----Original Message----- > >> From: David Miller [mailto:davem@davemloft.net] > >> Sent: Wednesday, March 12, 2014 12:47 PM > >> To: Wyborny, Carolyn > >> Cc: or.gerlitz@gmail.com; Kirsher, Jeffrey T; gospo@redhat.com; > >> netdev@vger.kernel.org; sassmann@redhat.com; hayja@mail.uc.edu; > >> amirv@mellanox.com; yevgenyp@mellanox.com > >> Subject: Re: [net-next 05/13] igb: Add support for debugfs > >> > >> From: "Wyborny, Carolyn" <carolyn.wyborny@intel.com> > >> Date: Wed, 12 Mar 2014 16:35:49 +0000 > >> > >> > Josh and I tried very hard to not reimplement anything already > >> > available in some other tool or mechanism and used the examples we > >> > already have for this in our other drivers. That being said, we'll > >> > abide and follow any clear criteria provided. > >> > >> I don't think what's being done here is appropriate. > >> > >> ethtool can provide the extended statistics. > >> > >> We already have a way to dump registers in ethtool as well. If you > >> want to add fine-grained register read support and write support, > >> that would be a nice addition to ethtool. > > > > Ok, I'll remove the register outputs. Is there any objection to the other > functionality provided? > > I said the statistics can be done with ethtool, so I object to everything. Understood. Withdrawing patchset. Will take suggestion for ethtool work. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/net/ethernet/intel/igb/e1000_82575.h b/drivers/net/ethernet/intel/igb/e1000_82575.h index 2a721a1..806e730 100644 --- a/drivers/net/ethernet/intel/igb/e1000_82575.h +++ b/drivers/net/ethernet/intel/igb/e1000_82575.h @@ -116,6 +116,18 @@ union e1000_adv_rx_desc { #define E1000_RXDADV_STAT_TS 0x10000 /* Pkt was time stamped */ #define E1000_RXDADV_STAT_TSIP 0x08000 /* timestamp in packet */ +#define E1000_RXDADV_PKTTYPE_ILMASK 0x00F0 +#define E1000_RXDADV_PKTTYPE_TLMASK 0x0F00 +#define E1000_RXDADV_PKTTYPE_NONE 0x0000 +#define E1000_RXDADV_PKTTYPE_IPV4 0x0010 /* IPV4 hdr present */ +#define E1000_RXDADV_PKTTYPE_IPV4_EX 0x0020 /* IPV4 hdr + extensions */ +#define E1000_RXDADV_PKTTYPE_IPV6 0x0040 /* IPV6 hdr present */ +#define E1000_RXDADV_PKTTYPE_IPV6_EX 0x0080 /* IPV6 + extensions */ +#define E1000_RXDADV_PKTTYPE_TCP 0x0100 /* TCP hdr present */ +#define E1000_RXDADV_PKTTYPE_UDP 0x0200 /* UDP hdr present */ +#define E1000_RXDADV_PKTTYPE_SCTP 0x0400 /* SCTP hdr present */ +#define E1000_RXDADV_PKTTYPE_NFS 0x0800 /* NDS hdr present */ + /* Transmit Descriptor - Advanced */ union e1000_adv_tx_desc { struct { diff --git a/drivers/net/ethernet/intel/igb/e1000_defines.h b/drivers/net/ethernet/intel/igb/e1000_defines.h index 393c896..59303ef 100644 --- a/drivers/net/ethernet/intel/igb/e1000_defines.h +++ b/drivers/net/ethernet/intel/igb/e1000_defines.h @@ -430,7 +430,7 @@ /* Extended Interrupt Cause Set */ /* E1000_EITR_CNT_IGNR is only for 82576 and newer */ #define E1000_EITR_CNT_IGNR 0x80000000 /* Don't reset counters on write */ - +#define E1000_EITR_INTERVAL 0x00007FFC /* Transmit Descriptor Control */ /* Enable the counting of descriptors still to be processed. */ @@ -920,6 +920,8 @@ #define E1000_EEER_LPI_FC 0x00040000 /* EEE Enable on FC */ #define E1000_EEE_SU_LPI_CLK_STP 0X00800000 /* EEE LPI Clock Stop */ #define E1000_EEER_EEE_NEG 0x20000000 /* EEE capability nego */ +#define E1000_EEER_RX_LPI_STATUS 0x40000000 /* RX in LPI state */ +#define E1000_EEER_TX_LPI_STATUS 0x80000000 /* TX in LPI state */ #define E1000_EEE_LP_ADV_ADDR_I350 0x040F /* EEE LP Advertisement */ #define E1000_EEE_LP_ADV_DEV_I210 7 /* EEE LP Adv Device */ #define E1000_EEE_LP_ADV_ADDR_I210 61 /* EEE LP Adv Register */ diff --git a/drivers/net/ethernet/intel/igb/e1000_regs.h b/drivers/net/ethernet/intel/igb/e1000_regs.h index abdd935..7e17181 100644 --- a/drivers/net/ethernet/intel/igb/e1000_regs.h +++ b/drivers/net/ethernet/intel/igb/e1000_regs.h @@ -368,6 +368,8 @@ #define E1000_IPCNFG 0x0E38 /* Internal PHY Configuration */ #define E1000_EEER 0x0E30 /* Energy Efficient Ethernet */ #define E1000_EEE_SU 0X0E34 /* EEE Setup */ +#define E1000_TLPIC 0x4148 /* EEE TX LPI count */ +#define E1000_RLPIC 0x414C /* EEE RX LPI count */ #define E1000_EMIADD 0x10 /* Extended Memory Indirect Address */ #define E1000_EMIDATA 0x11 /* Extended Memory Indirect Data */ #define E1000_MMDAC 13 /* MMD Access Control */ diff --git a/drivers/net/ethernet/intel/igb/igb.h b/drivers/net/ethernet/intel/igb/igb.h index 68b32d4a..a8c4fca 100644 --- a/drivers/net/ethernet/intel/igb/igb.h +++ b/drivers/net/ethernet/intel/igb/igb.h @@ -224,6 +224,18 @@ struct igb_rx_queue_stats { u64 alloc_failed; }; +struct igb_rx_pkt_stats { + u64 ipv4_packets; /* IPv4 headers processed */ + u64 ipv4e_packets; /* IPv4E headers processed */ + u64 ipv6_packets; /* IPv6 headers processed */ + u64 ipv6e_packets; /* IPv6e headers processed */ + u64 tcp_packets; /* TCP headers processed */ + u64 udp_packets; /* UDP headers processed */ + u64 sctp_packets; /* SCTP headers processed */ + u64 nfs_packets; /* NFS headers processed */ + u64 other_packets; +}; + struct igb_ring_container { struct igb_ring *ring; /* pointer to linked list of rings */ unsigned int total_bytes; /* total bytes processed this int */ @@ -268,6 +280,7 @@ struct igb_ring { struct { struct sk_buff *skb; struct igb_rx_queue_stats rx_stats; + struct igb_rx_pkt_stats pkt_stats; struct u64_stats_sync rx_syncp; }; }; @@ -460,6 +473,8 @@ struct igb_adapter { #ifdef CONFIG_DEBUG_FS struct dentry *igb_dbg_adapter; #endif /* CONFIG_DEBUG_FS */ + + int devrc; }; #define IGB_FLAG_HAS_MSI (1 << 0) diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c index a709caf..6cc9487 100644 --- a/drivers/net/ethernet/intel/igb/igb_main.c +++ b/drivers/net/ethernet/intel/igb/igb_main.c @@ -2012,6 +2012,8 @@ void igb_reset(struct igb_adapter *adapter) igb_ptp_reset(adapter); igb_get_phy_info(hw); + + adapter->devrc++; } static netdev_features_t igb_fix_features(struct net_device *netdev, @@ -2498,6 +2500,7 @@ static int igb_probe(struct pci_dev *pdev, const struct pci_device_id *ent) /* reset the hardware with the new settings */ igb_reset(adapter); + adapter->devrc = 0; /* Init the I2C interface */ err = igb_init_i2c(adapter); @@ -6923,11 +6926,53 @@ static void igb_process_skb_fields(struct igb_ring *rx_ring, struct sk_buff *skb) { struct net_device *dev = rx_ring->netdev; + __le16 pkt_info = rx_desc->wb.lower.lo_dword.pkt_info; + bool other = false; igb_rx_hash(rx_ring, rx_desc, skb); igb_rx_checksum(rx_ring, rx_desc, skb); + switch (pkt_info & E1000_RXDADV_PKTTYPE_ILMASK) { + case E1000_RXDADV_PKTTYPE_IPV4: + rx_ring->pkt_stats.ipv4_packets++; + break; + case E1000_RXDADV_PKTTYPE_IPV4_EX: + rx_ring->pkt_stats.ipv4e_packets++; + break; + case E1000_RXDADV_PKTTYPE_IPV6: + rx_ring->pkt_stats.ipv6_packets++; + break; + case E1000_RXDADV_PKTTYPE_IPV6_EX: + rx_ring->pkt_stats.ipv6e_packets++; + break; + case E1000_RXDADV_PKTTYPE_NONE: + other = true; + default: + break; + } + + switch (pkt_info & E1000_RXDADV_PKTTYPE_TLMASK) { + case E1000_RXDADV_PKTTYPE_TCP: + rx_ring->pkt_stats.tcp_packets++; + break; + case E1000_RXDADV_PKTTYPE_UDP: + rx_ring->pkt_stats.udp_packets++; + break; + case E1000_RXDADV_PKTTYPE_SCTP: + rx_ring->pkt_stats.sctp_packets++; + break; + case E1000_RXDADV_PKTTYPE_NFS: + rx_ring->pkt_stats.nfs_packets++; + break; + case E1000_RXDADV_PKTTYPE_NONE: + if (other) + rx_ring->pkt_stats.other_packets++; + break; + default: + break; + } + igb_ptp_rx_hwtstamp(rx_ring, rx_desc, skb); if ((dev->features & NETIF_F_HW_VLAN_CTAG_RX) &&