aboutsummaryrefslogtreecommitdiff
path: root/drivers/net/ethernet/intel
AgeCommit message (Collapse)AuthorFilesLines
2021-03-22iavf: Support Ethernet Type Flow Director filtersHaiyue Wang3-1/+42
Support the addition and deletion of Ethernet filters. Supported fields are: proto Supported flow-types are: ether Example usage: ethtool -N ens787f0v0 flow-type ether proto 0x8863 action 6 ethtool -N ens787f0v0 flow-type ether proto 0x8864 action 7 Signed-off-by: Haiyue Wang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22iavf: Support IPv6 Flow Director filtersHaiyue Wang3-0/+246
Support the addition and deletion of IPv6 filters. Supported fields are: src-ip, dst-ip, src-port, dst-port and l4proto Supported flow-types are: tcp6, udp6, sctp6, ip6, ah6, esp6 Example usage: ethtool -N ens787f0v0 flow-type tcp6 src-ip 2001::2 \ dst-ip CDCD:910A:2222:5498:8475:1111:3900:2020 \ tclass 1 src-port 22 dst-port 23 action 7 L2TPv3 over IP with 'Session ID' 17: ethtool -N ens787f0v0 flow-type ip6 l4proto 115 l4data 17 action 7 Signed-off-by: Haiyue Wang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22iavf: Support IPv4 Flow Director filtersHaiyue Wang5-1/+881
Support the addition and deletion of IPv4 filters. Supported fields are: src-ip, dst-ip, src-port, dst-port and l4proto Supported flow-types are: tcp4, udp4, sctp4, ip4, ah4, esp4 Example usage: ethtool -N ens787f0v0 flow-type tcp4 src-ip 192.168.0.20 \ dst-ip 192.168.0.21 tos 4 src-port 22 dst-port 23 action 8 L2TPv3 over IP with 'Session ID' 17: ethtool -N ens787f0v0 flow-type ip4 l4proto 115 l4data 17 action 3 Signed-off-by: Haiyue Wang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22iavf: Add framework to enable ethtool ntuple filtersHaiyue Wang4-1/+257
Enable ethtool ntuple filter support on the VF driver using the virtchnl interface to the PF driver and the Flow director functionality in the hardware. Signed-off-by: Haiyue Wang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Check FDIR program status for AVFQi Zhang7-16/+541
Enable returning FDIR completion status by checking the ctrl_vsi Rx queue descriptor value. To enable returning FDIR completion status from ctrl_vsi Rx queue, COMP_Queue and COMP_Report of FDIR filter programming descriptor needs to be properly configured. After program request sent to ctrl_vsi Tx queue, ctrl_vsi Rx queue interrupt will be triggered and completion status will be returned. Driver will first issue request in ice_vc_fdir_add_fltr(), then pass FDIR context to the background task in interrupt service routine ice_vc_fdir_irq_handler() and finally deal with them in ice_flush_fdir_ctx(). ice_flush_fdir_ctx() will check the descriptor's value, fdir context, and then send back virtual channel message to VF by calling ice_vc_add_fdir_fltr_post(). An additional timer will be setup in case of hardware interrupt timeout. Signed-off-by: Yahui Cao <[email protected]> Signed-off-by: Brett Creeley <[email protected]> Signed-off-by: Jesse Brandeburg <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Add more FDIR filter type for AVFQi Zhang4-18/+545
FDIR for AVF can forward - L2TPV3 packets by matching session id. - IPSEC ESP packets by matching security parameter index. - IPSEC AH packets by matching security parameter index. - NAT_T ESP packets by matching security parameter index. - Any PFCP session packets(s field is 1). Signed-off-by: Yahui Cao <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Add GTPU FDIR filter for AVFQi Zhang4-0/+218
Add new FDIR filter type to forward GTPU packets by matching TEID or QFI. The filter is only enabled when COMMS DDP package is downloaded. Signed-off-by: Yahui Cao <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Add non-IP Layer2 protocol FDIR filter for AVFQi Zhang6-4/+73
Add new filter type that allow forward non-IP Ethernet packets base on its ethertype. The filter is only enabled when COMMS DDP package is loaded. Signed-off-by: Yahui Cao <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Add new actions support for VF FDIRQi Zhang3-1/+25
Add two new actions support for VF FDIR: A passthrough action does not specify the destination queue, but just allow the packet go to next pipeline stage, a typical use cases is combined with a software mark (FDID) action. Allow specify a 2^n continuous queues as the destination of a FDIR rule. Packet distribution is based on current RSS configure. Signed-off-by: Yahui Cao <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Add FDIR pattern action parser for VFQi Zhang1-2/+345
Add basic FDIR flow list and pattern / action parse functions for VF. Signed-off-by: Yahui Cao <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Enable FDIR Configure for AVFQi Zhang9-6/+1094
The virtual channel is going to be extended to support FDIR and RSS configure from AVF. New data structures and OP codes will be added, the patch enable the FDIR part. To support above advanced AVF feature, we need to figure out what kind of data structure should be passed from VF to PF to describe an FDIR rule or RSS config rule. The common part of the requirement is we need a data structure to represent the input set selection of a rule's hash key. An input set selection is a group of fields be selected from one or more network protocol layers that could be identified as a specific flow. For example, select dst IP address from an IPv4 header combined with dst port from the TCP header as the input set for an IPv4/TCP flow. The patch adds a new data structure virtchnl_proto_hdrs to abstract a network protocol headers group which is composed of layers of network protocol header(virtchnl_proto_hdr). A protocol header contains a 32 bits mask (field_selector) to describe which fields are selected as input sets, as well as a header type (enum virtchnl_proto_hdr_type). Each bit is mapped to a field in enum virtchnl_proto_hdr_field guided by its header type. +------------+-----------+------------------------------+ | | Proto Hdr | Header Type A | | | +------------------------------+ | | | BIT 31 | ... | BIT 1 | BIT 0 | | |-----------+------------------------------+ |Proto Hdrs | Proto Hdr | Header Type B | | | +------------------------------+ | | | BIT 31 | ... | BIT 1 | BIT 0 | | |-----------+------------------------------+ | | Proto Hdr | Header Type C | | | +------------------------------+ | | | BIT 31 | ... | BIT 1 | BIT 0 | | |-----------+------------------------------+ | | .... | +-------------------------------------------------------+ All fields in enum virtchnl_proto_hdr_fields are grouped with header type and the value of the first field of a header type is always 32 aligned. enum proto_hdr_type { header_type_A = 0; header_type_B = 1; .... } enum proto_hdr_field { /* header type A */ header_A_field_0 = 0, header_A_field_1 = 1, header_A_field_2 = 2, header_A_field_3 = 3, /* header type B */ header_B_field_0 = 32, // = header_type_B << 5 header_B_field_0 = 33, header_B_field_0 = 34 header_B_field_0 = 35, .... }; So we have: proto_hdr_type = proto_hdr_field / 32 bit offset = proto_hdr_field % 32 To simply the protocol header's operations, couple help macros are added. For example, to select src IP and dst port as input set for an IPv4/UDP flow. we have: struct virtchnl_proto_hdr hdr[2]; VIRTCHNL_SET_PROTO_HDR_TYPE(&hdr[0], IPV4) VIRTCHNL_ADD_PROTO_HDR_FIELD(&hdr[0], IPV4, SRC) VIRTCHNL_SET_PROTO_HDR_TYPE(&hdr[1], UDP) VIRTCHNL_ADD_PROTO_HDR_FIELD(&hdr[1], UDP, DST) The byte array is used to store the protocol header of a training package. The byte array must be network order. The patch added virtual channel support for iAVF FDIR add/validate/delete filter. iAVF FDIR is Flow Director for Intel Adaptive Virtual Function which can direct Ethernet packets to the queues of the Network Interface Card. Add/delete command is adding or deleting one rule for each virtual channel message, while validate command is just verifying if this rule is valid without any other operations. To add or delete one rule, driver needs to config TCAM and Profile, build training packets which contains the input set value, and send the training packets through FDIR Tx queue. In addition, driver needs to manage the software context to avoid adding duplicated rules, deleting non-existent rule, input set conflicts and other invalid cases. NOTE: Supported pattern/actions and their parse functions are not be included in this patch, they will be added in a separate one. Signed-off-by: Jeff Guo <[email protected]> Signed-off-by: Yahui Cao <[email protected]> Signed-off-by: Simei Su <[email protected]> Signed-off-by: Beilei Xing <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Add support for per VF ctrl VSI enablingQi Zhang5-11/+129
We are going to enable FDIR configure for AVF through virtual channel. The first step is to add helper functions to support control VSI setup. A control VSI will be allocated for a VF when AVF creates its first FDIR rule through ice_vf_ctrl_vsi_setup(). The patch will also allocate FDIR rule space for VF's control VSI. If a VF asks for flow director rules, then those should come entirely from the best effort pool and not from the guaranteed pool. The patch allow a VF VSI to have only space in the best effort rules. Signed-off-by: Xiaoyun Li <[email protected]> Signed-off-by: Yahui Cao <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Enhanced IPv4 and IPv6 flow filterQi Zhang2-5/+53
Separate IPv4 and IPv6 ptype bit mask table into 2 tables: with or without L4 protocols. When a flow filter without any l4 type is specified, the ICE_FLOW_SEG_HDR_IPV_OTHER flag can be used to describe if user want to create a IP rule target for all IP packet or just IP packet without l4 header. Signed-off-by: Dan Nowlin <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Support to separate GTP-U uplink and downlinkQi Zhang4-15/+241
To apply different input set for GTP-U packet with or without extend header as well as GTP-U uplink and downlink, we need to add TCAM mask matching capability. This allows comprehending different PTYPE attributes by examining flags from the parser. Using this method, different profiles can be used by examining flag values from the parser. Signed-off-by: Dan Nowlin <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Add more advanced protocol support in flow filterQi Zhang3-6/+397
Add more protocol support in flow filter, these include PPPoE, L2TPv3, GTP, PFCP, ESP and AH. Signed-off-by: Ting Xu <[email protected]> Signed-off-by: Yahui Cao <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Support non word aligned input set fieldQi Zhang8-36/+623
To support FDIR input set with protocol field like DSCP, TTL, PROT, etc. which is not word aligned, we need to enable field vector masking. Signed-off-by: Dan Nowlin <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-22ice: Add more basic protocol support for flow filterQi Zhang5-6/+220
Add more protocol and field support for flow filter include: ETH, VLAN, ICMP, ARP and TCP flag. Signed-off-by: Kevin Scott <[email protected]> Signed-off-by: Qi Zhang <[email protected]> Tested-by: Chen Bo <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-19e1000e: Mark e1000e_pm_prepare() as __maybe_unusedWei Yongjun1-1/+1
The function e1000e_pm_prepare() may have no callers depending on configuration, so it must be marked __maybe_unused to avoid harmless warning: drivers/net/ethernet/intel/e1000e/netdev.c:6926:12: warning: 'e1000e_pm_prepare' defined but not used [-Wunused-function] 6926 | static int e1000e_pm_prepare(struct device *dev) | ^~~~~~~~~~~~~~~~~ Fixes: ccf8b940e5fd ("e1000e: Leverage direct_complete to speed up s2ram") Reported-by: Hulk Robot <[email protected]> Signed-off-by: Wei Yongjun <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-19igc: Assign boolean values to a bool variableJiapeng Zhong1-8/+8
Fix the following coccicheck warnings: ./drivers/net/ethernet/intel/igc/igc_main.c:4961:2-14: WARNING: Assignment of 0/1 to bool variable. ./drivers/net/ethernet/intel/igc/igc_main.c:4955:2-14: WARNING: Assignment of 0/1 to bool variable. ./drivers/net/ethernet/intel/igc/igc_main.c:4933:1-13: WARNING: Assignment of 0/1 to bool variable. ./drivers/net/ethernet/intel/igc/igc_main.c:4592:1-24: WARNING: Assignment of 0/1 to bool variable. ./drivers/net/ethernet/intel/igc/igc_main.c:4438:2-25: WARNING: Assignment of 0/1 to bool variable. ./drivers/net/ethernet/intel/igc/igc_main.c:4396:2-25: WARNING: Assignment of 0/1 to bool variable. ./drivers/net/ethernet/intel/igc/igc_main.c:4018:2-25: WARNING: Assignment of 0/1 to bool variable. Reported-by: Abaci Robot <[email protected]> Signed-off-by: Jiapeng Zhong <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-19igc: Remove unused MII_CR_LOOPBACKSasha Neftin1-1/+0
MII_CR_LOOPBACK masks not in use in i225 device and can be removed. Signed-off-by: Sasha Neftin <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-19igc: Remove unused MII_CR_SPEEDSasha Neftin1-3/+0
Force PHY speed not supported for i225 devices. MII_CR_SPEED masks not in use in i225 device and can be removed. Signed-off-by: Sasha Neftin <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-19igc: Remove unused MII_CR_RESETSasha Neftin1-1/+0
MII_CR_RESET mask not in use in i225 device and can be removed Signed-off-by: Sasha Neftin <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-19igb: check timestamp validityJesse Brandeburg3-14/+32
Add a couple of checks to make sure timestamping is on and that the timestamp value from DMA is valid. This avoids any functional issues that could come from a misinterpreted time stamp. One of the functions changed doesn't need a return value added because there was no value in checking from the calling locations. While here, fix a couple of reverse christmas tree issues next to the code being changed. Fixes: f56e7bba22fa ("igb: Pull timestamp from fragment before adding it to skb") Fixes: 9cbc948b5a20 ("igb: add XDP support") Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Dave Switzer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-19igb: Fix duplicate include guardTom Seewald1-3/+3
The include guard "_E1000_HW_H_" is used by two separate header files in two different drivers (e1000/e1000_hw.h and igb/e1000_hw.h). Using the same include guard macro in more than one header file may cause unexpected behavior from the compiler. Fix this by renaming the duplicate guard in the igb driver. Fixes: 9d5c824399de ("igb: PCI-Express 82575 Gigabit Ethernet driver") Signed-off-by: Tom Seewald <[email protected]> Reviewed-by: Jesse Brandeburg <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-19e1000e: Fix duplicate include guardTom Seewald1-3/+3
The include guard "_E1000_HW_H_" is used by header files in three different drivers (e1000/e1000_hw.h, e1000e/hw.h, and igb/e1000_hw.h). Using the same include guard macro in more than one header file may cause unexpected behavior from the compiler. Fix the duplicate include guard in the e1000e driver by renaming it. Fixes: bc7f75fa9788 ("[E1000E]: New pci-express e1000 driver (currently for ICH9 devices only)") Signed-off-by: Tom Seewald <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Acked-by: Sasha Neftin <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-18bpf, devmap: Move drop error path to devmap for XDP_REDIRECTLorenzo Bianconi4-28/+24
We want to change the current ndo_xdp_xmit drop semantics because it will allow us to implement better queue overflow handling. This is working towards the larger goal of a XDP TX queue-hook. Move XDP_REDIRECT error path handling from each XDP ethernet driver to devmap code. According to the new APIs, the driver running the ndo_xdp_xmit pointer, will break tx loop whenever the hw reports a tx error and it will just return to devmap caller the number of successfully transmitted frames. It will be devmap responsibility to free dropped frames. Move each XDP ndo_xdp_xmit capable driver to the new APIs: - veth - virtio-net - mvneta - mvpp2 - socionext - amazon ena - bnxt - freescale (dpaa2, dpaa) - xen-frontend - qede - ice - igb - ixgbe - i40e - mlx5 - ti (cpsw, cpsw-new) - tun - sfc Signed-off-by: Lorenzo Bianconi <[email protected]> Signed-off-by: Daniel Borkmann <[email protected]> Reviewed-by: Ioana Ciornei <[email protected]> Reviewed-by: Ilias Apalodimas <[email protected]> Reviewed-by: Camelia Groza <[email protected]> Acked-by: Edward Cree <[email protected]> Acked-by: Jesper Dangaard Brouer <[email protected]> Acked-by: Shay Agroskin <[email protected]> Link: https://lore.kernel.org/bpf/ed670de24f951cfd77590decf0229a0ad7fd12f6.1615201152.git.lorenzo@kernel.org
2021-03-17intel: Update drivers to use ethtool_sprintfAlexander Duyck4-101/+50
Update the Intel drivers to make use of ethtool_sprintf. The general idea is to reduce code size and overhead by replacing the repeated pattern of string printf statements and ETH_STRING_LEN counter increments. Signed-off-by: Alexander Duyck <[email protected]> Acked-by: Jesse Brandeburg <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-03-16Merge branch '40GbE' of ↵David S. Miller3-12/+22
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue Tony Nguyen says: ==================== 40GbE Intel Wired LAN Driver Updates 2021-03-16 This series contains updates to i40e, ixgbe, and ice drivers. Magnus Karlsson says: Optimize run_xdp_zc() for the XDP program verdict being XDP_REDIRECT in the xsk zero-copy path. This path is only used when having AF_XDP zero-copy on and in that case most packets will be directed to user space. This provides around 100k extra packets in throughput on my server when running l2fwd in xdpsock. ==================== Signed-off-by: David S. Miller <[email protected]>
2021-03-16Merge branch '1GbE' of ↵David S. Miller1-2/+10
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue Tony Nguyen says: ==================== 1GbE Intel Wired LAN Driver Updates 2021-03-15 This series contains updates to e1000e only. Chen Yu says: The NIC is put in runtime suspend status when there is no cable connected. As a result, it is safe to keep non-wakeup NIC in runtime suspended during s2ram because the system does not rely on the NIC plug event nor WoL to wake up the system. Besides that, unlike the s2idle, s2ram does not need to manipulate S0ix settings during suspend. ==================== Acked-by: Jakub Kicinski <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-03-15e1000e: Remove the runtime suspend restriction on CNP+Chen Yu1-1/+1
Although there is platform issue of runtime suspend support on CNP, it would be more flexible to let the user decide whether to disable runtime or not because: 1. This can be done in userspace via echo on > /sys/devices/pci0000\:00/0000\:00\:1f.d/power/control 2. More and more NICs would support runtime suspend, disabling the runtime suspend on them by default would impact the validation. Only disable runtime suspend on CNP in case of any user space regression. Signed-off-by: Chen Yu <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-15e1000e: Leverage direct_complete to speed up s2ramChen Yu1-1/+9
The NIC is put in runtime suspend status when there is no cable connected. As a result, it is safe to keep non-wakeup NIC in runtime suspended during s2ram because the system does not rely on the NIC plug event nor WoL to wake up the system. Besides that, unlike the s2idle, s2ram does not need to manipulate S0ix settings during suspend. This patch introduces the .prepare() for e1000e so that if the NIC is runtime suspended the subsequent suspend/resume hooks will be skipped so as to speed up the s2ram. The pm core will check whether the NIC is a wake up device so there's no need to check it again in .prepare(). DPM_FLAG_SMART_PREPARE flag should be set during probe to ask the pci subsystem to honor the driver's prepare() result. Besides, the NIC remains runtime suspended after resumed from s2ram as there is no need to resume it. Tested on i7-2600K with 82579V NIC Before the patch: e1000e 0000:00:19.0: pci_pm_suspend+0x0/0x160 returned 0 after 225146 usecs e1000e 0000:00:19.0: pci_pm_resume+0x0/0x90 returned 0 after 140588 usecs After the patch: echo disabled > //sys/devices/pci0000\:00/0000\:00\:19.0/power/wakeup becomes 0 usecs because the hooks will be skipped. Suggested-by: Kai-Heng Feng <[email protected]> Signed-off-by: Chen Yu <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-15net: ethernet: intel: igb: Typo fix in the file igb_main.cBhaskar Chowdhury1-1/+1
s/structue/structure/ Signed-off-by: Bhaskar Chowdhury <[email protected]> Acked-by: Randy Dunlap <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-03-15ice: optimize for XDP_REDIRECT in xsk pathMagnus Karlsson1-4/+8
Optimize ice_run_xdp_zc() for the XDP program verdict being XDP_REDIRECT in the xsk zero-copy path. This path is only used when having AF_XDP zero-copy on and in that case most packets will be directed to user space. This provides a little over 100k extra packets in throughput on my server when running l2fwd in xdpsock. Signed-off-by: Magnus Karlsson <[email protected]> Tested-by: George Kuruvinakunnel <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-15ixgbe: optimize for XDP_REDIRECT in xsk pathMagnus Karlsson1-4/+7
Optimize ixgbe_run_xdp_zc() for the XDP program verdict being XDP_REDIRECT in the xsk zero-copy path. This path is only used when having AF_XDP zero-copy on and in that case most packets will be directed to user space. This provides a little under 100k extra packets in throughput on my server when running l2fwd in xdpsock. Signed-off-by: Magnus Karlsson <[email protected]> Tested-by: Vishakha Jambekar <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-15i40e: optimize for XDP_REDIRECT in xsk pathMagnus Karlsson1-4/+7
Optimize i40e_run_xdp_zc() for the XDP program verdict being XDP_REDIRECT in the xsk zero-copy path. This path is only used when having AF_XDP zero-copy on and in that case most packets will be directed to user space. This provides a little over 100k extra packets in throughput on my server when running l2fwd in xdpsock. Signed-off-by: Magnus Karlsson <[email protected]> Tested-by: George Kuruvinakunnel <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-12igb: avoid premature Rx buffer reuseLi RongQing1-7/+15
Igb needs a similar fix as commit 75aab4e10ae6a ("i40e: avoid premature Rx buffer reuse") The page recycle code, incorrectly, relied on that a page fragment could not be freed inside xdp_do_redirect(). This assumption leads to that page fragments that are used by the stack/XDP redirect can be reused and overwritten. To avoid this, store the page count prior invoking xdp_do_redirect(). Longer explanation: Intel NICs have a recycle mechanism. The main idea is that a page is split into two parts. One part is owned by the driver, one part might be owned by someone else, such as the stack. t0: Page is allocated, and put on the Rx ring +--------------- used by NIC ->| upper buffer (rx_buffer) +--------------- | lower buffer +--------------- page count == USHRT_MAX rx_buffer->pagecnt_bias == USHRT_MAX t1: Buffer is received, and passed to the stack (e.g.) +--------------- | upper buff (skb) +--------------- used by NIC ->| lower buffer (rx_buffer) +--------------- page count == USHRT_MAX rx_buffer->pagecnt_bias == USHRT_MAX - 1 t2: Buffer is received, and redirected +--------------- | upper buff (skb) +--------------- used by NIC ->| lower buffer (rx_buffer) +--------------- Now, prior calling xdp_do_redirect(): page count == USHRT_MAX rx_buffer->pagecnt_bias == USHRT_MAX - 2 This means that buffer *cannot* be flipped/reused, because the skb is still using it. The problem arises when xdp_do_redirect() actually frees the segment. Then we get: page count == USHRT_MAX - 1 rx_buffer->pagecnt_bias == USHRT_MAX - 2 From a recycle perspective, the buffer can be flipped and reused, which means that the skb data area is passed to the Rx HW ring! To work around this, the page count is stored prior calling xdp_do_redirect(). Fixes: 9cbc948b5a20 ("igb: add XDP support") Signed-off-by: Li RongQing <[email protected]> Reviewed-by: Alexander Duyck <[email protected]> Tested-by: Vishakha Jambekar <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-12ixgbe: move headroom initialization to ixgbe_configure_rx_ringMaciej Fijalkowski1-1/+2
ixgbe_rx_offset(), that is supposed to initialize the Rx buffer headroom, relies on __IXGBE_RX_BUILD_SKB_ENABLED flag. Currently, the callsite of mentioned function is placed incorrectly within ixgbe_setup_rx_resources() where Rx ring's build skb flag is not set yet. This causes the XDP_REDIRECT to be partially broken due to inability to create xdp_frame in the headroom space, as the headroom is 0. Fix this by moving ixgbe_rx_offset() to ixgbe_configure_rx_ring() after the flag setting, which happens to be set in ixgbe_set_rx_buffer_len. Fixes: c0d4e9d223c5 ("ixgbe: store the result of ixgbe_rx_offset() onto ixgbe_ring") Signed-off-by: Maciej Fijalkowski <[email protected]> Tested-by: Vishakha Jambekar <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-12ice: move headroom initialization to ice_setup_rx_ctxMaciej Fijalkowski2-17/+18
ice_rx_offset(), that is supposed to initialize the Rx buffer headroom, relies on ICE_RX_FLAGS_RING_BUILD_SKB flag as well as XDP prog presence. Currently, the callsite of mentioned function is placed incorrectly within ice_setup_rx_ring() where Rx ring's build skb flag is not set yet. This causes the XDP_REDIRECT to be partially broken due to inability to create xdp_frame in the headroom space, as the headroom is 0. Fix this by moving ice_rx_offset() to ice_setup_rx_ctx() after the flag setting. Fixes: f1b1f409bf79 ("ice: store the result of ice_rx_offset() onto ice_ring") Signed-off-by: Maciej Fijalkowski <[email protected]> Tested-by: Kiran Bhandare <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-12i40e: move headroom initialization to i40e_configure_rx_ringMaciej Fijalkowski2-12/+13
i40e_rx_offset(), that is supposed to initialize the Rx buffer headroom, relies on I40E_RXR_FLAGS_BUILD_SKB_ENABLED flag. Currently, the callsite of mentioned function is placed incorrectly within i40e_setup_rx_descriptors() where Rx ring's build skb flag is not set yet. This causes the XDP_REDIRECT to be partially broken due to inability to create xdp_frame in the headroom space, as the headroom is 0. For the record, below is the call graph: i40e_vsi_open i40e_vsi_setup_rx_resources i40e_setup_rx_descriptors i40e_rx_offset() <-- sets offset to 0 as build_skb flag is set below i40e_vsi_configure_rx i40e_configure_rx_ring set_ring_build_skb_enabled(ring) <-- set build_skb flag Fix this by moving i40e_rx_offset() to i40e_configure_rx_ring() after the flag setting. Fixes: f7bb0d71d658 ("i40e: store the result of i40e_rx_offset() onto i40e_ring") Reported-by: Jesper Dangaard Brouer <[email protected]> Co-developed-by: Jesper Dangaard Brouer <[email protected]> Signed-off-by: Jesper Dangaard Brouer <[email protected]> Signed-off-by: Maciej Fijalkowski <[email protected]> Acked-by: Jesper Dangaard Brouer <[email protected]> Tested-by: Jesper Dangaard Brouer <[email protected]> Tested-by: Kiran Bhandare <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-12ice: fix napi work done reporting in xsk pathMagnus Karlsson2-7/+9
Fix the wrong napi work done reporting in the xsk path of the ice driver. The code in the main Rx processing loop was written to assume that the buffer allocation code returns true if all allocations where successful and false if not. In contrast with all other Intel NIC xsk drivers, the ice_alloc_rx_bufs_zc() has the inverted logic messing up the work done reporting in the napi loop. This can be fixed either by inverting the return value from ice_alloc_rx_bufs_zc() in the function that uses this in an incorrect way, or by changing the return value of ice_alloc_rx_bufs_zc(). We chose the latter as it makes all the xsk allocation functions for Intel NICs behave in the same way. My guess is that it was this unexpected discrepancy that gave rise to this bug in the first place. Fixes: 5bb0c4b5eb61 ("ice, xsk: Move Rx allocation out of while-loop") Reported-by: Maciej Fijalkowski <[email protected]> Signed-off-by: Magnus Karlsson <[email protected]> Tested-by: Kiran Bhandare <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-11e1000e: Fix error handling in e1000_set_d0_lplu_state_82571Dinghao Liu1-0/+2
There is one e1e_wphy() call in e1000_set_d0_lplu_state_82571 that we have caught its return value but lack further handling. Check and terminate the execution flow just like other e1e_wphy() in this function. Fixes: bc7f75fa9788 ("[E1000E]: New pci-express e1000 driver (currently for ICH9 devices only)") Signed-off-by: Dinghao Liu <[email protected]> Acked-by: Sasha Neftin <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-11e1000e: add rtnl_lock() to e1000_reset_taskVitaly Lifshits1-1/+5
A possible race condition was found in e1000_reset_task, after discovering a similar issue in igb driver via commit 024a8168b749 ("igb: reinit_locked() should be called with rtnl_lock"). Added rtnl_lock() and rtnl_unlock() to avoid this. Fixes: bc7f75fa9788 ("[E1000E]: New pci-express e1000 driver (currently for ICH9 devices only)") Suggested-by: Jakub Kicinski <[email protected]> Signed-off-by: Vitaly Lifshits <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-11igc: Fix igc_ptp_rx_pktstamp()Andre Guedes2-33/+41
The comment describing the timestamps layout in the packet buffer is wrong and the code is actually retrieving the timestamp in Timer 1 reference instead of Timer 0. This hasn't been a big issue so far because hardware is configured to report both timestamps using Timer 0 (see IGC_SRRCTL register configuration in igc_ptp_enable_rx_timestamp() helper). This patch fixes the comment and the code so we retrieve the timestamp in Timer 0 reference as expected. This patch also takes the opportunity to get rid of the hw.mac.type check since it is not required. Fixes: 81b055205e8ba ("igc: Add support for RX timestamping") Signed-off-by: Andre Guedes <[email protected]> Signed-off-by: Vedang Patel <[email protected]> Signed-off-by: Jithu Joseph <[email protected]> Reviewed-by: Maciej Fijalkowski <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-11igc: Fix Supported Pause Frame Link SettingMuhammad Husaini Zulkifli1-0/+3
The Supported Pause Frame always display "No" even though the Advertised pause frame showing the correct setting based on the pause parameters via ethtool. Set bit in link_ksettings to "Supported" for Pause Frame. Before output: Supported pause frame use: No Expected output: Supported pause frame use: Symmetric Fixes: 8c5ad0dae93c ("igc: Add ethtool support") Signed-off-by: Muhammad Husaini Zulkifli <[email protected]> Reviewed-by: Malli C <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Acked-by: Sasha Neftin <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-11igc: Fix Pause Frame AdvertisingMuhammad Husaini Zulkifli1-3/+1
Fix Pause Frame Advertising when getting the advertisement via ethtool. Remove setting the "advertising" bit in link_ksettings during default case when Tx and Rx are in off state with Auto Negotiate off. Below is the original output of advertisement link during Tx and Rx off: Advertised pause frame use: Symmetric Receive-only Expected output: Advertised pause frame use: No Fixes: 8c5ad0dae93c ("igc: Add ethtool support") Signed-off-by: Muhammad Husaini Zulkifli <[email protected]> Reviewed-by: Malli C <[email protected]> Acked-by: Sasha Neftin <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-11igc: reinit_locked() should be called with rtnl_lockSasha Neftin1-0/+9
This commit applies to the igc_reset_task the same changes that were applied to the igb driver in commit 024a8168b749 ("igb: reinit_locked() should be called with rtnl_lock") and fix possible race in reset subtask. Fixes: 0507ef8a0372 ("igc: Add transmit and receive fastpath and interrupt handlers") Suggested-by: Jakub Kicinski <[email protected]> Signed-off-by: Sasha Neftin <[email protected]> Tested-by: Dvora Fuxbrumer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-10net: add a helper to avoid issues with HW TX timestamping and SO_TXTIMEVladimir Oltean2-2/+2
As explained in commit 29d98f54a4fe ("net: enetc: allow hardware timestamping on TX queues with tc-etf enabled"), hardware TX timestamping requires an skb with skb->tstamp = 0. When a packet is sent with SO_TXTIME, the skb->skb_mstamp_ns corrupts the value of skb->tstamp, so the drivers need to explicitly reset skb->tstamp to zero after consuming the TX time. Create a helper named skb_txtime_consumed() which does just that. All drivers which offload TC_SETUP_QDISC_ETF should implement it, and it would make it easier to assess during review whether they do the right thing in order to be compatible with hardware timestamping or not. Suggested-by: Vinicius Costa Gomes <[email protected]> Signed-off-by: Vladimir Oltean <[email protected]> Acked-by: Vinicius Costa Gomes <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-03-05net: intel: iavf: fix error return code of iavf_init_get_resources()Jia-Ju Bai1-1/+2
When iavf_process_config() fails, no error return code of iavf_init_get_resources() is assigned. To fix this bug, err is assigned with the return value of iavf_process_config(), and then err is checked. Reported-by: TOTE Robot <[email protected]> Signed-off-by: Jia-Ju Bai <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-03-04ixgbe: Fix memleak in ixgbe_configure_clsu32Dinghao Liu1-2/+4
When ixgbe_fdir_write_perfect_filter_82599() fails, input allocated by kzalloc() has not been freed, which leads to memleak. Signed-off-by: Dinghao Liu <[email protected]> Reviewed-by: Paul Menzel <[email protected]> Tested-by: Tony Brelinski <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-03-04ixgbe: fail to create xfrm offload of IPsec tunnel mode SAAntony Antony2-0/+10
Based on talks and indirect references ixgbe IPsec offlod do not support IPsec tunnel mode offload. It can only support IPsec transport mode offload. Now explicitly fail when creating non transport mode SA with offload to avoid false performance expectations. Fixes: 63a67fe229ea ("ixgbe: add ipsec offload add and remove SA") Signed-off-by: Antony Antony <[email protected]> Acked-by: Shannon Nelson <[email protected]> Tested-by: Tony Brelinski <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.